National Academies Press: OpenBook

Assessing Changing Food Consumption Patterns (1981)

Chapter: Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data

« Previous: Appendix A: Background Papers for Workshop on Methods for the Collection of Aggregate Data on Food Consumption
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 87
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 88
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 89
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 90
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 91
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 92
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 93
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 94
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 95
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 96
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 97
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 98
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 99
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 100
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 101
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 102
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 103
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 104
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 105
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 106
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 107
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 108
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 109
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 110
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 111
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 112
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 113
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 114
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 115
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 116
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 117
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 118
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 119
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 120
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 121
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 122
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 123
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 124
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 125
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 126
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 127
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 128
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 129
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 130
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 131
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 132
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 133
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 134
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 135
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 136
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 137
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 138
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 139
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 140
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 141
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 142
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 143
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 144
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 145
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 146
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 147
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 148
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 149
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 150
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 151
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 152
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 153
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 154
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 155
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 156
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 157
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 158
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 159
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 160
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 161
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 162
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 163
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 164
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 165
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 166
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 167
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 168
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 169
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 170
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 171
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 172
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 173
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 174
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 175
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 176
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 177
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 178
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 179
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 180
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 181
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 182
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 183
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 184
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 185
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 186
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 187
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 188
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 189
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 190
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 191
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 192
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 193
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 194
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 195
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 196
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 197
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 198
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 199
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 200
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 201
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 202
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 203
Suggested Citation:"Appendix B: Background Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data." National Research Council. 1981. Assessing Changing Food Consumption Patterns. Washington, DC: The National Academies Press. doi: 10.17226/380.
×
Page 204

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

APPENDIX BBackground Papers for Workshop on Evaluation of Methods for Obtaining Food Consumption Data 1

Dietary Methodology CHARLOTTE M. YOUNG It is of interest to find two branches of government suddenly showing inter- est in dietary methodology and willing to invest some money in solution to its problems. Literally millions of dollars have been spent on research in which dietary data have been related to various measures of nutritional status. Often the significance of the dietary methods used has been relatively unknown, and it has been difficult to figure out why the method used was chosen. Thirty some years ago I remember questioning the basis for choos- ing dietary methods used in certain studies and finally concluded it usually seemed to be expediency. These observations led me into some feeble ef- forts at looking at dietary methodology, which we did, with no special funds, as a side issue to another study. At that point I left an active interest in the field in pursuit of seeking answers to other practical questions in human nutrition. STATE OF THE ART For years I have tried to encourage nutritionists to make contributions to this field. The "take" has been very limited. Much of the basic work was done 20 to 40 years ago. Though for 20 years there has been a great need for dietary methodology for epidemiological studies dealing with the relation- ship between dietary intake of an individual and his biochemical or health status, we have not come very far. Many reviews of various depth and value concerning dietary survey methods have been written. I shall not attempt to compete with these; instead, I have prepared a list of many of the reviews 89

JO CHARLOTTE M. YOUNG including symposia on the subject which occurred in the 1950's and 1960's. The reader may refer to these for details. Recently I have been impressed that the government has become very interested in dietary methodology. You will find the extensive review of Burk and Pao (1976) as they contemplated the 1975 Household Food Con- sumption Survey, the most recent in a series made about every 10 years by the U.S. Department of Agriculture. It is my understanding that several other sections of the government wished through this study to be able to obtain dietary information on both households and individuals. Hence in about 1975-76 the Consumer and Food Economics Institute of the USDA contracted with Response Analysis of Princeton, N.J., to do an extensive field trial after suggesting a series of methodologies to be tested. As I recall, the task was twofold: ( 1) to recommend procedures for a nationwide study of household food use and individual food intake with justification for such recommendations and (2) to recommend whether a pilot study was needed before proceeding with the nationwide data collection, together with a rationale for the conclusion reached. In this study some nine different methods received field trials. Eight of the methods included an effort to get information on household food usage by either some form of 7-day recall, diary, or interviews requiring different types of preparation or training. Four of the methods included data on individual family members using diary or recall for 1 to 3 days from the homemaker for all individuals in the household or from each individual subject as respondent. If the reports can be made available to you, the Committee and its consultants might well wish to examine them. None of the methods proposed incorporated a direct way of measuring its validity or reliability, since there was no observation of all food used or repeat observations in the same households. The Response Analysis survey is probably the most comprehensive study on dietary methodology of which I am aware. It is interesting that so quickly after these studies the Food and Drug Administration has asked the Food and Nutrition Board to conduct an in-depth study with some of the same directions. Such opportunities were what some of us dreamed of years ago. Today it is fortunate that diverse professions have developed interest in dietary methodology so that your Committee will have access not only to the talents of nutritionists of various backgrounds, but also epidemiologists, an- thropologists, economists, sociologists, physicians, statisticians and com- puter scientists. DIETARY STATUS, NOT NUTRITIONAL STATUS It is imperative to remember that in measuring dietary status one is not measuring nutritional status. In using dietary assessments one should be

Dietary Methodology 91 quite clear as to their limitations, realizing that dietary studies give no direct measurement of nutritional status. They give only presumptive evidence. If food intake is quite poor, nutritional status may be below desirable limits. If nutrient intake appears to meet certain dietary standards and if the individual has no factors that may adversely condition his nutritional needs, he proba- bly will be well nourished. Dietary studies should be referred to as diet studies, not nutritional studies. In interpreting dietary information, one must bear in mind that there are pronounced individual differences in nutritional requirements; that our knowledge of absolute nutritional requirements is relatively meager; that a whole series of factors may condition an individual's nutrient needs by either interfering with the ingestion, absorption, or utilization of a nutrient or by increasing his nutritional needs through increased requirements, excretion, or destruction of the nutrient. These so-called "conditioning fac- tors" may be discovered by medical histories and examination as part of a nutritional assessment but may be completely unknown in a dietary study. Furthermore, we have little information on how slight differences in nutrient intake affect the health of individuals, as well as little knowledge as to what level of failure to meet dietary allowances food intake affects the health of the individual. Additional laboratory and clinical tests are necessary before nutri- tional status is established. Given these limitations, dietary studies do have many uses. In survey work alone, a description of food intake may be valuable in interpreting nutritional findings. The studies help to identify apparent dietary deficien- cies or excesses. They also form a concrete basis for action programs, for in the long run, therapy will need to be interpreted into feeding programs. Also, valid and reliable information on dietary intake of an individual over a sufficient period of time may be important in relation to various biochemical and clinical measurements of his nutritional status. AREAS IN WHICH WORK IS STILL NEEDED Though there is need for work in the entire field of dietary methodology, certain areas stand out as especially critical: 1. Validity, which to my understanding does not include the refinements or subdivisions of Burk and Pao. To me, validity means the degree to which the method is a true measure of what the investigator wishes to describe. What one wishes to measure of course depends upon one's objectives. Burk and Pao describe five aspects of validity: accuracy, concurrent, construct, content, and predictive validities. The aspect most usually considered I sus- pect would come under content validity, i.e., does the method give the

92 CHARLOTTE M. YOUNG "usual" picture of the subject's eating at the time in which the investigator is interested. Too little information is available on validity, and probably "true" valid- ity cannot be measured. In the early studies efforts were made to determine if two or more different methods gave comparable results or whether similar population samples in the same frame of time gave similar results. In more recent and sophisticated times, there have been an extremely limited number of studies in which actual intake (observed without the subject's knowledge) has been compared to the reported intake whether by some form of record by the subject or by recall interview. Some have only been of a single meal. The ethical question always arises as to whether the subject should be aware that he is being observed. Factors that interfere with getting the desired "usual" picture include: (a) the tendency to eat differently during a record period either deliberately or unconsciously as a matter of convenience; (b) the desire to please or outwit the surveyor, which can lead either to omit- ting certain items or reporting certain foods the subject believes he should eat; (c) the general attitude to the questioner or to surveys in general; (d) ability or willingness to keep accurate records; (e) the condition of the subject's mem- ory or willingness to use it; (f) food choice if the subject knows he is being observed; and (g) the mere act of record-keeping. Some individuals, such as homemakers, certain kinds of students, or certain food faddists, may be much more aware of what they eat than others. 2. Reliability, which to me means repeatability or true reproducibility or the error variation in collecting and processing dietary data. Some people confuse reliability as I am using it with validity. Most reliability studies have related to such matters as how accurately actual intake has been re- ported, errors due to the use of food tables, the differences in nutritional value between calculated and chemically analyzed diets, whether reported studies on the same population in a limited period of time gave similar results, and errors in sampling as induced by failure to get random samples when the objective would indicate its desirability but practical cir- cumstances preclude random sampling. 3. Intraindividual variability in food or nutrient intake, which is espe- cially important in determining the length of time that an individual needs to be studied for a true picture of intake be it current or past. There are more of these studies than those related to validity, but they still are quite limited. 4. Interindividual variability, which is of particular interest in population studies or group averages to determine the size of sample needed for the purpose. More has been done here than in the case of intraindividual varia- bility. In recent years the role of a statistician in the planning stages of a study as well as the later analytical stages has been appreciated by nutritionists.

Dietary Methodology 93 Statisticians are extremely valuable in making the investigator clearly define his objectives and then in assisting in the many decisions that must be made such as sampling methods, sample size, appropriate quality control, setting up appropriate analyses, and interpretations of the results once available. 5. Changes in inter- and intraindividual variability over time and par- ticularly in relation to seasonal variation. 6. Epidemiological methods suitable for large-scale surveys of the re- lationship of certain dietary or nutrient intakes or habits to disease condi- tions, two of which are of great current interest: coronary heart disease and cancer. Since such studies are often based on relatively large numbers of subjects, special techniques need to be developed or others adapted to the purpose. Five or six groups have attempted to develop various forms of short questionnaires and to check their validity and reliability. Hankin and her colleagues in a period of over 10 years (first in California and later in Hawaii) have developed and tested such a questionnaire in relation to car- diovascular disease, and more recently, to cancer. Previous techniques have been expensive and tedious and often could not be applied to all kinds of population groups. Also, often they have~concentrated on only one facet of dietary pattern, i.e., the level of nutrient intake, with little information on other parameters of food intake such as specific food usage, changes in food patterns, spacing or apportionment of food over the day, environment, speed of eating, etc. 7. Need for more studies of response rates and factors that affect them. 8. Need for studies of actual time and costs involved in the use of differ- ent methods. 9. Development of methods for use by relatively untrained, inexperi- enced interviewers, or nonnutritionists. 10. The role of different disciplines in the study of dietary habits. The role of the statistician was mentioned earlier, but what about the others? We must recognize our need for them at appropriate times as well as their need in certain studies for the advice of an expert nutritionist. It is again part of the increasing recognition of the interdisciplinary nature of much research concerned with very personal aspects of human beings. WHAT WE HAVE LEARNED; DECISIONS TO BE MADE First and foremost, it is clear that the objectives of the dietary study must be clearly defined, for the objective determines the appropriate methods to be used in collecting, processing and interpreting the dietary data. Past studies would indicate that this point has not always been appreciated. Secondly, it is clear on the basis of past dietary studies that there is a need

94 CHARLOTTE M. YOUNG for a clear-cut distinction between methods to be used for the average intake of a group for group comparisons and those suitable for defining the dietary intake of an individual. Dietary methodology studies have demonstrated the much smaller cost in time, money, and subject cooperation to obtain the mean intake of a group. There has been increasing realization of the time, cost, subject cooperation, and degree of precision necessary to obtain a valid and reliable measure of the intake of an individual. Even after the objective is clearly defined, many decisions must be made with regard to methodology before a dietary study is undertaken, particu- larly studies concerned with the intake of individuals. Decisions with regard to the collection of data center on sampling, the schedule or form to be used and/or the interviewer, the instructions or supervision to be given, the time period to be covered, the timing of the recording process or recall period, the methods to be used in determining the amounts of food used, and whether food intake only or food habits as well are of interest. Questions of sample size and type are fundamental. Does the sample need to be random so that one can describe a larger universe from which the sample was drawn? What limitations will be imposed by variation in the willingness of selected subjects to cooperate? Or can the sample be nonran- dom if one is concerned with data on only the sample individuals and the characteristics of the larger universe bearing upon food consumption? The size of the sample varies with the objective and the method chosen but also on the classifications to be made in analyzing the data. For group averages a larger sample may be needed to include interindividual variability; the lower the variability in food consumption, the smaller the number needed for stable averages. The method of sampling should be clearly stated as well as information relative to those unwilling to or unable to cooperate. Too often this point has been neglected in reports. It is well to remember that sampling is a field of statistics in itself. What time period should be covered? Are we interested in current intake, immediately past intake for a fixed period, or the usual or characteristic intake over an extended period in the present or past? In what kind of dietary information are we interested? In nutrients, which ones? Foods? If so, which ones? Are we concerned with quantitative aspects or only the presence or absence of certain foods in the diet? Is it a food pattern in which we are interested? Are we concerned with food habits such as time, place, and circumstances of eating; with whom, regularity, food patterns, seasonings, waste, peculiarities of preparation, etc.? How will amounts be reported? By weight, by household measures, by estimations of size of portions, or merely by the frequency with which some item is eaten? From whom shall such information be obtained, i.e., who will be the respondent? The individual subject, the person responsible for the

Dietary Methodology 95 subject's food, or both, or an observer? If the latter, what kind of training is needed? The major determinant of these decisions is the objective of the study, but also to be considered are the resources available in terms of time, money, trained personnel, and the availability of a sufficient supply of suitable subjects (in Burk and Pao's terms, the respondent burden, field survey costs, and data-processing costs). NECESSITY FOR PRETESTING METHODOLOGY All evidence points to the necessity of pretesting the method on the particu- lar population before undertaking a large study. As indicated earlier, possi- ble methods vary with age, sex, culture, and education of the subjects as well as the complexity of the food resources available to the participant. Age is a factor not only for the very young, who cannot write or report, but also for the elderly, where memory may be affected. Education and culture may affect the ability to write and how much interviewers must be involved. Inter- and intraindividual variability should be determined as an indica- tion of the number of subjects needed as well as the time period that must be covered. It is also an opportunity to determine if there is a seasonal variation in intake. All too often the results are disastrous when pretesting has not been done. Quality control for data collection with training of interviewers and careful standardization must be planned. In advance of the study, decisions also must be made with regard to methods of analysis, such as whether one is interested in diet patterns, frequency of intake of particular foods, or specific nutrients. If individual nutrients are to be calculated, what methods and what tables are suitable? If chemically analyzed, what methods? Interpretation or evaluation of ade- quacy then follows. For precision, even when food tables are used, there is increasing emphasis on chemical analyses of certain local foods to check values included in nationally used tables. Some investigators believe that before a major study is begun, the form of anticipated tables of results should be set up to be sure proper data and needed numbers of subjects are being used to meet the objective. TYPES OF METHODS AVAILABLE In general there are two types of dietary studies and within each type a variety of methods may be used: those concerned with food usage of families or institutions sharing common food supplies and those concerned with food usage of individuals. Information is obtained either by food rec- ords or by recall of what has been eaten.

96 Household or Institution Methods Three methods are available: CHARLOTTE M. YOUNG 1. Food accounts- a simple running description of food purchased, re- ceived as gifts, or produced for household use over a given period of time. These are not precise and rarely used today. 2. Food records a weighted inventory of foods on hand at the begin- ning and close of the study together with a day-to-day record of food brought into the home or institution over the period of the study with or without an accounting of kitchen and plate waste or food fed to pets. This is the method the USDA used in its early household consumption studies. 3. Food lists a method by which an interviewer obtains from the person responsible for the food an estimate of the quantities of food used over a given period of time. This is the method used by the USDA in the more recent household food consumption studies. The methods vary widely in cost of collecting and processing the data and to some extent the uses to which the data may be put. It is well to remember that there is no merit in using a more elaborate or expensive method than is necessary to obtain the data needed to meet the defined objective of the study. Individual Methods Dietary data on individuals are collected either (1) to obtain average nutrient intake, food intake, or food habits of groups for comparison with other groups or (2) to obtain nutrient intake of a given individual for correlation with clinical or biochemical measurements obtained on that individual. Studies vary from a qualitative type of food habit inquiry to those of a very much more precise quantitative nature. Each type of study has its use; the important thing is to consider carefully what kind of facts are needed for a given purpose and which method provides these facts at the lowest cost. Methods used with individuals include the following: 1. An estimation by recall, in which the subject, or in the case of young children perhaps the parent, recalls the food intake over the previous 24 hours or longer, with dependence on memory. 2. Records of food eaten by an individual kept by weights, household measurements, or by estimated quantities over a stated period of time. 3. Dietary history, in which by recall or repeated food records or both the interviewer aims to discover the usual eating patterns over a relatively long

Dietary Methodology 97 period of time. It is a time-consuming process requiring professional per- sonnel. 4. More recently foodfrequency questionnaires, either self-administered or interviewer-administered, have been studied as a means of acquiring information on general dietary intake or specific foods or nutrients over a longer period of time at less expense and with less personnel resources for epidemiological studies in which people may be grouped in extremes of intake. Unfortunately, a number of these initial efforts have not been done with typical American cultures. 5. Weighted intake for precise measurements during which all food eaten is carefully weighted and nutritive value either calculated from food tables or determined by laboratory analysis of duplicate samples; usually this method is used only in research groups with special facilities for collection and analysis. FACTORS INVOLVED IN CHOICE OF METHODS OR STATE OF THE ART WITH REGARD TO EACH METHOD Three statements of Keys (1968, 1979) may well be borne in mind as we consider the factors involved in choice of method. The first is that good surveys are difficult, costly, and require professional expertise in planning, operation, and interpretation. The second is that little is known of the long- time intraindividual variability of diet or that diet at any one time is rep- resentative over longer periods of time; and third, that even the best dietary surveys cannot provide answers about biochemical relationships. Many comments may be made about the various dietary methods based on heterogenous dietary methodology studies that have been made over the last 40 years. More than most reviewers, Burk and Pao (1976) have brought them together in one place in relation to each type of method and is the latest review now in print. Hence the following discussion draws heavily on their review. Greater detail may be found in the review itself. Food Accounts This method is not widely used in this country. Great Britain and Israel are two countries with considerable experience with the method. In general, the food account gives a record of food entering the house or institution in a given period but not whether it is actually consumed in that period. It is less accurate than other household or institution methods but is felt by reviewers to give a good picture of the general diet. There is usually no indication of food discarded for spoilage, plate waste, other waste or fed to animals. Sampling can be large since less is demanded of respondents. However,

98 CHARLOTTE M. YOUNG sampling is dependent on the willingness and the ability of the household to keep a written record (Sudman and Ferber, 1971~. It is reported in one study that the lowest cooperation was found in one- or two-member households or where the head of the household was 55 years or older, or where there was less than an eighth grade education or income less than $2,000. Such house- holds might be suspicious of surveys or find record-keeping difficult or too time-consuming. Though recording is not as detailed and time-consuming as other record methods, not all families will participate or are able to do so. It is possible that the diet is less likely to be altered by the food account method than by other more intrusive methods. Little information is available on the respondent burden or field survey costs except that two or three visits per week are made to homes by field personnel. No data are available on processing costs. Food Records Information with regard to food record and list recall methods comes largely from studies done by the USDA, which has used the methods widely and has changed for its 10-year food consumption studies from the use of the food record (referred to by Burk and Pao as inventory record method) to the list recall method. Inventory record and list recall procedures have been compared in un- published studies by Murray (1970) and Grossman and Popka (1976~. The latter provides the best basis for appraisal of the two methods. The reader is referred to the Burk and Pao review for details of the studies too extensive to be given here. The food record method requires both considerable respondent coopera- tion and considerable interviewer time since food inventories must be taken previous to the onset of the record-keeping period and at the end of it. Records must be kept of all food entering the household during the specified period as well as quantities in weight and price expenditures. Random sampling of households using this method is difficult or almost impossible, and therefore results cannot usually be generalized. The nonre- sponse rate is much higher than the list recall method, and fewer small and elderly households are likely to participate. Sampling is disturbed not only by the initial refusal to participate, but also by failure to keep the records once started and by incompleteness of the reporting of key items. In one study it is reported that the quantity of food, money value, and nutritive values obtained with the food record method averaged 20 percent less than those derived from the list recall procedures. For these methods there is no direct comparison to actual values; comparison between methods is the best information available.

Dietary Methodology 99 As to reliability, it is reported that based on quantities of major food groups food records and list recall give similar values. However, the list recall method gives slightly higher percentage standard errors. As to its repeatability it has been reported that similar mean values for food con- sumption have been obtained by replication of surveys. As to validity, there is the problem of possible changes in food patterns because of the burden of weighing and measuring, or conforming to per- ceived culturally approved patterns. Considerably higher field survey costs are involved with the food record due to the use of more interviewer time for the beginning and ending inven- tories and the visits in between, the total average time given as 53/4 hours. There are also greater costs of processing data since some 12 hours were taken to prepare records for computer versus only about 5~/~ hours for the recall schedules. Also it was reported that there were 3.7 times as many editing and processing errors per set of food records as per list recall schedules. Some reviewers make the point that in Europe weighing of ingredients rather than use of household measurements is fairly common, and con- sequently there may be fewer problems with the food record method. In any case, the lower respondent rate in sampling is a major handicap in the use of the family or institution food record method. Food List or List Recall The food list or list recall method usually involves only one visit by the interviewer and is reported to take an average of 2 hours if no memory aids are involved or 2~/2 hours if memory aids are used. A major advantage is a higher response rate so that a more adequate and representative sample may be attained than in the case of the food record method. Small households and older households are reported as more likely to respond to the list recall method. Burk and Pao indicate that researchers experienced in surveying homemakers for food data insist most homemakers are able to recall house- hold use between two time boundaries if they are related to the cycle of food activities. Furthermore, major food shopping has been shown to be done by the week, or less often, and usually by the homemaker. The method also has the advantage that the interviewer can immediately evaluate responses and probe for additional information when needed. Quality and completeness of information such as sizes of containers, brands, etc., can be improved by the interchange between interviewer and respondent and observation at the home site. It is interesting that the use of memory aids seemed to annoy many

100 CHARLOTTE M. YOUNG respondents, and, in a comparison of list recall interviews where memory aids were used in contrast to where they were not, little effect was shown on the amounts recalled. As indicated earlier, the field survey costs were lower for the list recall method since the average interviewer time was less than half that for the food record method. One would wonder, too, if costs would not be less in achieving the desired respondent rate because of the lower response rate for the food record method. The computer-related processing costs were less than half for the list recall method as for the food record method; in addition, much more time was spent to rectify the greater number of editing and processing errors per set of food record schedules as contrasted to the list recall schedule. Thus, in this country for household or institutional studies the list recall method in recent years has taken precedence over the food record method, particularly in government circles (USDA, 1972b). INDIVIDUAL METHODS Burk and Pao conclude that no one method is consistently advantageous over all others and that researchers have to decide what trade-offs are most relevant to their objectives. Recall The 24-hour recall is often used to get a general diet pattern on a group basis; it has the disadvantage of relying on memory. However, memory for recalls limited to 24 hours appears quite good except perhaps in older people. Campbell and Dodds (1967) found there was a significant difference between older and younger subjects in calories forgotten and that the factor of age was more important than education. They concluded that before the shorter, more economical 24-hour recall is used with older people where the menu is not known, modifications will be needed if reliable data are to be obtained. They suggest that a checklist or the assistance of a spouse or younger family member may be required. The 24-hour recall is particularly good for group studies because it saves the risk of individuals changing their customary food patterns and, in most cases, because of the higher response rate due to the light respondent bur- den. There is the problem of the subject's inability to estimate certain sizes accurately, which may be helped to some extent by the interviewer's use of models, glasses, bowls, and spoons of various sizes to aid in the estimation of portion sizes. There are problems of both under- and overestimation and some report a greater tendency toward underestimation of intakes.

Dietary Methodology 101 A great advantage is the greater response rate, which, on a group basis, gives greater representative and predictive validity. Also costs are lower because of less personnel time, though processing costs will be higher to convert estimates to weights. On an individual basis, the place of the 24-hour recall still needs to be established. Obviously, the 24-hour recall of intake cannot be correlated with clinical or biochemical measurements on the individual that result from intake over a much longer period of time. It is conceivable that repeated 24-hour recalls over a substantial period of time might be useful for this purpose. Reliability as related to sampling appears to be high in terms of satisfac- tory samples obtained (Abraham et al., 1974; Swanson et al., 1959; USDA, 1972a) though less satisfactory samples have been reported by others (Mor- gan, 1959; USDHEW, 19721. Reliability (repeatability) has been good in terms of actual consumption for groups of children (Emmons and Hayes, 1973; Samuelson, 1970) with advance notice helping recall (Meredith et al., 1951~. The same study indi- cated more underreporting with increased number of items with 86 hos- pitalized lactating women. Linusson et al. (1974), comparing recall infor- mation with quantities eaten at all three meals as determined by weighing by the interviewer, found the recall method appeared fairly accurate for qual- itative estimation of averages for population groups but not highly valid for ascertaining quantity of food consumed. Regression analyses between recall and actual intake showed for all food groups a tendency to overestimate actual intake when consumption was low and to underestimate it when it was high. In a study by Gersovitz et al. (1978), the internal validity of a 24-hour recall and 7-day dietary record was investigated among a group of noninstitutionalized elderly subjects who were participating in a congregate meals plan. Internal validity was assessed by comparing reported intake with unobtrusively obtained data on actual intake for the single meal per day at the congregate meal site. Paired test results for both the 24-hour recall and the 7-day record suggest that both methods provide equally accurate esti- mates of the mean intake. Results suggested that the recall is prone to overreporting low intakes and underreporting high intakes, a pattern refer- red to as the "flat slope syndrome," the main implications of which are a downward bias in the number of subjects with extremely high or extremely low intakes. Balogh et al. (1971) in Israel with volunteer male clerical and adminis- trative workers found that repeated 24-hour recalls were valuable aids in the difficult and complex area of classifying individual dietary intakes and should be used more frequently, particularly for those populations and for

102 CHARLOTTE M. YOUNG those specific nutrients for which the range of variability is not extreme. They state that when variability is extreme, the method of repeated 24-hour recall interview will reveal the situation and point toward the hard fact that no method short of extensive daily sampling is likely to be successful in accurately categorizing large numbers of individuals with respect to dietary intake. Frank et al. (1977) have proposed an improved 24-hour recall technique to be used with school children that requires well-trained nutritionists, graduated food models, and standardized probing. Two different nutrition- ists as interviewers for the same children produce results that are not sig- nificantly different. The interview, however, took 30 to 40 minutes per child. Madden et al. (1976), using a sample of subjects 60 years or older, compared actual intake with 24-hour recall intakes. By paired test no sig- nificant difference was found between mean recall and mean actual intake for any of eight nutrients except calories. Using regression analysis, results indicated that for calories, protein, and vitamin A small intakes tended to be overreported and large intakes underreported. According to Burk and Pao's classification, the accuracy aspect of valid- ity, i.e., the absence of systemic error, found that for group averages the 24-hour recall could substitute for an estimated 7-day record (Young et al., 1952b). Adelson (1960) found the 7-day record of intelligent adult men gave similar results to the 7-day weighed records. However, Trulson (1954) found recalls by children gave lower averages than 7-day estimated records. Campbell and Dodds (1967) found evidence of underestimation but the ability to remember in their older subjects increased by probing. Burk and Pao's concurrent validity, i.e., two measures of the same con- cept, examines studies of weighed records in relationship to recalls. Bransby et al. (1948) with school boys in Britain, and Thomson (1958) with pregnant women in Britain, found recalls gave lower estimates and, for Bransby et al., in some cases higher. Adelson (1960) for adults, and Samuelson (1970) for children in Sweden, found acceptable comparisons between the two. In construct validity, i.e., the degree to which variability in concept is measured, for various groups of children (Eppright et al., 19521; Iowa, South Dakota, and North Central women (Swanson et al., 1959~; USDA unpublished data ~ 1965~; and HANES data on women age 60 and over (Abra- ham et al., 1974), relative standard errors depended somewhat on group size, but 24-hour recall compared favorably with estimated record. For representative validity, i.e., the measure or level taken as an indicator of a more general level, the 24-hour recall appeared to yield better rep- resentation than other methods, because response rates were higher (Mongeau, 1974~.

Dietary Methodology For predictive validity, the good response rate for 24-hour recall contrib- utes to its usefulness (USDA, 1972a). Response burden is lightened by the use of the 24-hour recall (Young et al., 1952b) and field survey costs are less if only one visit is necessary (Adelson, 1960~; data-processing costs are increased when household mea- sures must be converted to weights (Adelson, 1960~. 103 Records (EstimatedJ Response rate for the dietary record method is apt to be less than the 24-hour recall, because some people will not take on the burden of record-keeping and some may be unable to keep records. Food survey costs will vary with the number of interviewer visits re- quired. Data-processing costs include the necessity of converting estimated amounts to weights or to common household measures. Some investigators believe that a 1-day record in other cases, a 3-day record will give a reliable average dietary intake of a group if enough individuals are included. However, they are not adequate for measurements for individuals. There is no common agreement on the period required in this case, and it does vary with which nutrients one wishes to measure. A 7-day record has been most commonly used, but there is evidence because of intraindividual variability and seasonal variability that it probably is not adequate. Longer periods are difficult to obtain and further reduce response rates. Very few long-term studies have been done (Chappell, 1955; Young et al ., 1953b; Yudkin, 1951~. For some population groups, the day of record-keeping as well as the day of the week makes a substantial difference. More research is needed on the number, spacing, and section of days of record-keeping. When greater variety of foodstuffs and smaller samples are used, longer periods of time are necessary. With limited variety of foods, shorter periods may be used. The accuracy of estimated records may be questioned. One of the big handicaps in many estimated records is the inability of subjects to estimate portion sizes of food accurately particularly for certain types of food (Young et al., 1953a). Reliability as related to sampling is more difficult than for the 24-hour recall. A random sample is difficult to obtain because of the relatively low response rate though it is considered less difficult than the weighed record (Marr,19711. On a group basis, 1 day was considered satsifactory by Chalmers et al. 1952) with a variety of population groups, but others considered consecutive days to be necessary (Eppright et al., 1955, 19724. Estimated records gave higher nutritive values than weighed records (Eppright et al.,

104 CHARLOTTE M. YOUNG 1952) but results by other investigators were inconsistent (Bransby et al., 1948; Young et al., 1953a). Still other investigators found correlation be- tween consecutive days (Trulson, 1951), but Chalmers et al. (1952) found no correlation. Evidence of actual variability was found by Celender (1963) and Young et al. (1953b). As far as accuracy is concerned there was an upward bias at the start as reported by Celender (1963) and Young et al. (1953a). Several studies report bias because of inability to estimate portion sizes accurately (Young et al., 1953a); others indicate that some respondents are unwilling to report accurately (Paul et al., 19631. Still others found accuracy was influenced by interviewers (Church et al., 1954; Steele et al., 1951~. An effect of day of record-keeping was reported by Cellier and Hankin (1963), Eppright et al. (1952), and Leverton and Marsh (19391. Eppright et al. (1952) and Owen et al. (1974) reported concurrent valid- ity with positive correlation of dietary findings with biochemical measures. In terms of construct validity Young et al. (1952b) found that for group averages a 1-day record was sufficient, but for the estimate of an indi- vidual's intake, a longer period is necessary. The length of time required varied with the particular nutrient. In terms of representative validity, several investigators reported that con- secutive 3-day records did not give the same results as 1 day, 7 days, or combinations of nonconsecutive days (Eppright et al., 1952; Hankin et al., 1967; Trulson, 1951~. Others have shown that weekend days give different results than week days (Cellier and Hankin, 1963; Leverton and Marsh, 19391. Representativeness of data is also affected by the nonresponse rate of the sample as shown by Owen et al. (1974) in the national preschool nutrition survey. Eppright et al. (1972), Owen et al. (1974), and other surveys have demonstrated good predictive validity of food and nutrition intake measured with estimated records. Response rate is reduced when the estimated food record method is used because of the burden to the subject as demonstrated by both Mongeau (1974) and Young et al. (1952b). Field survey costs are increased if more than one interviewer visit is required (Eppright et al., 1972; Owen et al., 1974), and data-processing costs are increased when household measures must be converted to weights or common measures. Dietary History The dietary history method was developed by Burke (1947) as a research tool to estimate usual or average food and nutrient intake over a period of time so it might be related to growth or other clinical measurements. The

Dietary Methodology rationale was that such measurements are the result of earlier, long-time food habits, not current habits alone. As used by Burke and later her stu- dent, Beat (1967), in growth and development studies of children, repeated dietary studies were done on the same children and were useful in classify- ing them into groups. Both 3-day records and recall of usual intake starting with the past 24 hours were used together with subsequent cross-check with a food list partially as a means of probing. Their subjects usually knew in advance histories were to be taken and therefore were alerted to being more aware of their food intake than has been the case of some subsequent uses of the method. Having observed Burke taking dietary histories with her quiet, nonsuggestive method, which put the subject at ease, I am convinced that in her hands it was quite different than has been practiced by others. The method requires a highly skilled and trained interviewer and usually takes an hour or longer to obtain. Obviously it cannot be done by just anyone with just any subject. In some cases of people with fairly unstructured eating patterns, it may be impossible. However, for long-term studies, or one concerned with immediate past intake, it is one of the few possibilities if skilled personnel and sufficient time are available. Several investigators have been able to get satisfactory samples with the use of the dietary history method (Burke, 1947; Beat, 1967; Christakis et al., 1968; Dawber et al., 1962; Mann earl., 1962; Young et al., 1952a). Results relative to repeatability have varied with different investigators and different types of subjects. Reshef and Epstein (1972) found good cor- relations for repeat interviews using the same nutritionist; Huenemann and Turner (1942) found that subjects really did not know what they ate. Good repeatability for group averages did not necessarily imply good estimates for individuals (Calender, 1963; Reed and Burke, 1954; Trulson and McCann, 1959~. Several studies have shown that the diet history is likely to give higher estimates of intake than other methods (Young et al., 1952b; Trul- son, 195 1; Stevens et al., 1963; Celender, 19631. In some cases the subjects had not been alerted to being aware of their dietary intake. Since in most cases the dietary history method has not been used with probability samples, it would rate poorly in terms of representativeness. The respondent burden is little for the dietary history method but the effort is great for the interviewer (den Hartog et al., 1965~. Field survey costs are high in terms of needing highly skilled and specially trained nu- tritionists with approximately an hour or longer for each interview, and considerably longer to edit interviews and prepare them for data processing. 105 FOOD FREQUENCY METHOD FOR RECALL OR PAST FOOD INTAKE Increasingly various forms of dietary questionnaires have been developed for use in epidemiologic studies to give an indication how often certain

106 CHARLOTTE M. YOUNG foods are eaten in a given period of time. These questionnaires may be given by interview or they may be self-administered. One of the earliest was developed by Stefanik and Trulson (1962), which took 20-25 minutes by interview. Results were compared with either recent 7-day records or previ- ous dietary histories. The investigators felt at the group level and at the individual level of analysis the short questionnaire gave generally equivalent estimates of the qualitative consumption of foods. Very shortly afterwards, Abramson et al. (1963) reported a study de- signed to determine if the frequency with which specific foods were taken might be a reasonable index of the usual quantity of these foods eaten per week; both sets of data were obtained by interview. They then tried to test the value of the method empirically by seeking associations between the frequency data and the subject's hemoglobin level. It was not possible to obtain quantitative data by actual measurement for comparison with the frequency data, and the reliability of the method, that is, the extent to which a repeated interview promoted similar information to the first, was not tested. Data were obtained from 30-minute interviews with a heterogeneous sample of 60 Jewish women aged 17 to 39 in the fifth and eighth months of preg- nancy. The findings suggest that variations in size of average servings do not outweigh effects of differences in frequency. The authors felt that with the heterogeneous populations the food frequency interview did not give results close enough to quantitative intake to justify the choice of the method in studies of the diet of individuals, but it was close enough to warrant its consideration of moderately sizable groups in an epidemiologic study. A main limitation of the method they felt to be its inability to produce data on individual nutrients. Balogh et al. ( 1968), in a long-term prospective investigation of ischemic heart disease of adult Israeli males 40 years and older, developed a short dietary questionnaire of 15 minutes duration. It was based on a scoring system of frequency x nutrients x amount and a week period. Estimation of portion size was aided by food models and a food list was used instead of meal patterns. The validity was tested with 49 persons against a modified Burke dietary history and a combined week's record with weighing of rep- resentative food items. The authors felt the questionnaire to be a practical tool for the assessment of dietary intake in the prospective study. Epstein et al. (1970), in another Israeli study using groups of men and women, some born in North Africa and others in Europe, compared a short dietary questionnaire with a research dietary history of the Burke type. There was a moderate or low correlation between the data obtained by the same individuals by the two methods. The finding suggested that the valid- ity of the short questionnaire method varied with sex, age, ethnic group, and

Dietary Methodology education and was lowest among persons with a varied diet. The mean daily intake of nutrients was lower by the short method. The lack of high correla- tion between data obtained by the two methods was taken as evidence that the short questionnaire is not a very valid indicator of the diet of an indi- vidual. In 1972, Reshef and Epstein reported that the reliability of the questionnaire did not vary in the different sex and country of birth groups. None of the mean values for the food items was significantly different in two interviews. In addition, variability of the diet did not affect the reliability. Browe et al. (1966a) reported the use of a questionnaire to estimate the dietary intake of 1,514 male civil service employees being followed in an epidemiologic study of cardiovascular disease in New York State. The sur- vey was done to classify the men's diets in relation to serum lipid levels, other clinical data being gathered and subsequent occurrence and clinical course of heart disease. The advantages of the method include ease of administration, relatively low cost, minimum need for specially trained personnel, and ease of application to large-scale surveys. The disadvantages are mainly those typified by all nonobjective methods for assessing dietary intake. Yudkin (1966) used a 15-20-minute clinic interview questionnaire with 23 heart patients to assess sugar intake. Comparison was made to 7-day food records in household measures made by the subjects a few days later. Yud- kin felt the short questionnaire gave average results that agreed well with the 7-day records. Stead (1968) compared two different modifications of the Burke dietary history method for estimate of average nutrient intake of a group of 30 white 12-year-old school girls in Pretoria. The results obtained by the two methods agreed fairly well in the case of almost all nutrients studied. Hankin and others have worked with developing a self-administered short questionnaire to estimate accurately the dietary intakes of participants in cardiovascular epidemiologic studies in San Francisco with Japanese- American men and later in Honolulu since 1967 (Hankin and Huenemann, 1967; Hankin et al., 1967; Hankin et al., 1970) and later for studies of gastrointestinal cancer (Hankin et al., 1975; Hankin et al., 19781. The ques- tionnaires were based on 7-day measured food records. In formulating the questionnaires, food groups were used. For group purposes a limited number of food items that took 30 minutes per questionnaire predicted fairly well the average intake of the men for the nutrients considered. To use the method, a preliminary investigation with a sample of the population to be studied is necessary to establish standards, to identify particular items eaten in definable serving units, and to provide data for comparable regression equations ap- propriate to the population. 107

108 WEIGHED RECORD METHOD CHARLOTTE M. YOUNG A weighed record is a history of all foods eaten by an individual in a specified period with the amounts of each determined by weighing. It is a most reliable and most expensive method and the only method for absolute accuracy of quantity of food eaten. But the weighing procedure limits the size of sample and tends to disrupt customary routines. It is most ideal for clinical studies; yet it can be used only for current intake and does not give information on long-term intakes. Since foods are in weighed portions, pro- cessing costs are less, but more personnel time may be involved in field survey costs. No satisfactory evaluations of this method for large surveys were found in the literature using weighed and analyzed records, probably because of the great cost in time and money, need for literate and highly cooperative subjects and the small number of persons who can be handled (Marr, 1971; Mongeau, 19741. Random samples are very difficult to obtain because of low response rate (Burrill et al., 1959; den Hartog et al., 1965; Leverton and Pazur, 1957~. Greatest variation was found in day to day records, but 7 days was consid- ered sufficiently reliable for group averages (Adelson, 1960; Widdowson, 1947~. However, 1 week was not considered a reliable estimate of an indi- vidual's intake (Chappell, 1955; Fry et al., 1963; Huenemann and Turner, 1942; Young etal., 1952a, 1953; Yudkin, 1951~. From an accuracy aspect of validity, Marr (1971) considers the weighed record the most accurate measure. Yudkin (1951) felt there was an upward bias at the start, whereas den Hartog et al. (1965) and Ohlson et al. (1950) felt there was a downward bias if it was too much trouble for the respondent. If a portion size is missing an average size is inserted. Dieckmann et al. (1951), in a study of pregnant women, found weighed portion sizes did not conform to those postulated in printed tables, whereas Beegle et al. (1954) with a group of mature women found mean and median amounts eaten of any given food were relatively close. Use of food composition tables for nutrient content was found to give acceptable accuracy (Widdowson and McCance, 1943~; others report less than acceptable accuracy (Leverton, 1937; Manalo and Jones, 1966; Walberg and Adams, 1965) . As far as concurrent validity is concerned, Eppright et al. (1952) found that averages for estimated records exceeded means from weighed records. However, Dieckmann et al. (1952) and Huenemann and Turner (1942) found positive correlations with biochemical measures. As far as concurrent validity (i.e., the degree to which variability in concept is measured) is concerned, Marr (1971) stated that the weighed record gives a precise measure of intake, but not necessarily of customary

Dietary Methodology 109 consumption. In Italy (Fidanza and Fidanza-Alberti, 1967) and in England (Heady, 1961), researchers felt that records for 3-5 days capture most of the information. Keys et al . (1966' and Yudkin ~ 1951) felt that less than 1 week was too short a period. Representative validity because of the low response rates adversely af- fects generalization of results when the weighed record is used (Adelson, 1961; Thomson, 1958), and predictability is low for large-scale surveys. The respondent burden for the weighed method is heavy (Eppnght et al., 1952; den Hartog et al., 19651. Field survey costs are higher if more than one interviewer visit is required (Adelson, 1960; den Hartog et al., 1965; Mart, 1971~. However, data-processing costs are decreased because foods do not require conversion to common weights. S UMMARY There is obviously a need for more methodology studies, since there are still many unanswered questions. It is apparent from the diversity of results that pretesting is basic with the population group to be studied before a major survey is undertaken. It would be helpful if the results of such pretesting were made available in the literature. The greatest need is a means of finding out past individual intake for correlation with physical and biochemical measurements on the individual. REFERENCES AND BIBLIOGRAPHY Abraham, S., F. Lowenstein, and C. Johnson. 1974. Preliminary findings of the first health and nutrition examination survey, United States, 1971-72: Dietary intake and biochemical findings. U.S. Department of Health, Education, and Welfare, National Center for Health Statistics, Public Health Service. DHEW Publ. No. (HRA) 74-1219-1. 183 pp. Abramson, J. H., C. Slome, and C. Kosovsky. 1963. Food frequency interview as an epidemiological tool. Am. J. Public Health 53:1093. Adelson, S. F. 1960. Some problems in collecting dietary data from individuals. J. Am. Diet. Assoc. 36:453. Adelson, S. F. 1961. Practical procedures for dietary surveys. Pages 85-96 in Proc. 3rd Intl. Congr. Dietet. Newman Books, London. Adelson, S. F., E. Asp, and I. Noble. 1961. Household records of food used and discarded. A pilot study in St. Paul. J. Am. Diet. Assoc. 39:578. Anderson, R. K., and H. R. Sandstead. 1947. Nutritional appraisal and demonstration program of the U.S . Public Health Service. J. Am. Diet. Assoc. 23: 1201. Assessment of nutritional status and food consumption surveys. 1974. In J. E. Somogyi, Ruschhkon-Zurich, and A. Szczygiel, eds. Proc. Eleventh Symp. Group Europ. Nutr. Warsaw, Apr. 9-13, 1972. Bibl. Nutr. Diet. 20. Baker, H., O. Frank, S. Feingold, et al. 1967. Vitamins, total cholesterol and triglycerides in 642 New York City school children. Am. J. Clin. Nutr. 20:850.

110 CHARLOTTE M. YOUNG Balogh, M., J. H. Medalie, H. Smith, and J. J. Groen. 1968. The development of a dietary questionnaire for ischemic heart disease survey. Isr. J. Med. Sci . 4: 195. Balogh, M., H. A. Kahn, and J. H. Medalie. 1971. Random repeat 24-hour dietary recalls. Am. J. Clin. Nutr. 24:304. Bavly, S. 1976. Levels of nutrition in Israel, 1968-69. Family Expenditure Survey Part II. Ministry Educ. Culture, Coll. Nutr. Home Econ., Cent. Burl Stat., Jerusalem. 102 pp. Beal, V. A. 1967. The nutritional history in longitudinal research. J. Am. Diet. Assoc. 51:426. Becker, B. G., B. P. Indik, and A. M. Beeualkes. 1960. Dietary intake methodologies A review. Univ. Mich. Res. Inst. Tech. Rep. Proj. 03188, Ann Arbor, Mich. Bebb, H. T., T. B. Houser, J. C. Witchi, and A. S. Littell. 1972. Nutritive content of the usual diets of 82 men. J. Am. Diet. Assoc. 61:507. Beegle, R. M., P. H. Roberts, N. Howard, et al. 1954. Weights of foods eaten per meal by 242 women 30-92 years of age. Mich. Agric. Exp. Stn. Tech. Bull. 244, No.-Cent. Region. Publ. 37. 32 pp. Bigwood, E. J. 1939. Guiding principles for studies on the nutrition of populations. League of Nations Health Organization, Geneva. Bransby, E. R., C. G. Daubney, and J. King. 1948. Comparison of results obtained by different methods of individual dietary survey. Brit. J. Nutr. 2:89-110. Browe, J. H., R. M. Gofstein, D. M. Morlley, and M. C. McCarthy. 1966a. Diet and heart disease study in the cardiovascular health center. 1. A questionnaire and its application in assessing dietary intake. J. Am. Diet. Assoc. 48:95. Browe, J. H., D. M. Morlley, et al. 1966b. Diet and heart disease study in the cardiovascular health center, 2. Construction of a food table and its use. J. Am. Diet. Assoc. 48:101. Browe, J. H., D. M. Morlley, V. M. Logrillo, and J. T. Doyle. 1967. Diet and heart disease study in the cardiovascular health center. 3. Dietary intake and physical activity of male civil service employees. J. Am. Diet. Assoc. 50:376. Bryan, A. H., and E. L. Anderson. 1960. Retrospective dietary interviewing. A technique. J. Am. Diet. Assoc. 37:558. Bureau of Human Nutrition and Home Economics. 1948. Summary of Proceedings of a Conference on the Study of Food Intake. U.S. Department of Agriculture, Washington, D. C., October 25 -28, 1948. Mimeo. Burk, M. C., and E. M. Pao.1976. Methods for large scale surveys of household and individual diets. Home Econ. Res. Rep. No. 40. U.S. Government Printing Office, Washington, D.C., November 1976. Burke, B. S. 1947. Thedietary history as a research tool. Am. Diet. Assoc. J. 23:1041-1046. Burrill, L. M., B. Alsup, C. Schuck, etal. 1959. Evaluation of the self-chosen weighted diets of 402 women 30-97 years of age in the seven North Central states. S. Dak. State Coll., Home Econ. Dept., Agric. Exp. Stn. Bull. 478, No.-Cent. Region. Publ. 105. 19 pp. Campbell, V. A., and M. L. Dodds. 1967. Collecting dietary information from groups of older people. Am. Diet. Assoc. J. 51(1):29-33. Celendar, I. H. 1963. Reliability of methods of recording dietary intake in women at constant weight. University Microfilms, Ann Arbor, Mich. 161 pp. Cellier, K. M., and M. E. Hankin. 1963. Studies of nutrition in pregnancy. I. Some consider- ations in collecting dietary information. Am. J. Clin. Nutr. 13:55-62. Chalmers, F. W., M. M. Clayton, L. O. Gates, et al. 1952. The dietary record How many and which days? Chamberlain, K., and M. Pike. 1948. An experimental study of the accuracy of a method of survey of individual diets not directly based on weighing. Scientific Advisors Division, Ministry of Food, London. Mimeo to Author.

Dietary Methodology 111 Chappell, G. M. 1955. Long-term individual dietary surveys. Br. J. Nutr. 9:323-339. Christakis, G., A. Miridjanian, L. Nath, et al. 1968. A nutritional epidemiologic investigation of 642 New York City children. Am. J. Clin. Nutr. 21:107. Church, H. N., M. M. Clayton, C. M. Young, and W. D. Foster. 1954. Can different inter- viewers obtain comparable dietary survey data? J. Am. Diet. Assoc. 30:777. Clark, F. 1969. The 1965-66 food consumption survey: Scope, methodology, and highlights. Pages 43-56 in Using food surveys in consumer education. N.C. State Univ. Agric. Policy Inst., Raleigh. Clark, F. 1974. Recent food consumption surveys and their uses. Fed. Proc. 33:2270. Committee on Nutrition Surveys. 1949. Nutrition surveys: Their techniques and value. Na- tional Research Council, Bull. 117, National Academy of Sciences, Washington, D.C., May 1949. (See Section II, Dietary surveys and their interpretations, pp. 11-47.) Darby, W. J. 1947. The influence of some recent studies in the interpretation of the findings of dietary surveys. J. Am. Diet. Assoc. 23:204. Dawber, T., G. Pearson, P. Anderson, et al. 1962. Dietary assessment in the epidemiologic study of coronary heart disease: The Framingham study. II. Reliability of measurement. Am. J. Clin. Nutr. 11:226. den Hartog, C., Th. Van Schaik, L. Dalderup, et al. 1965. The diet of volunteers participating in a long term epidemiological field survey on coronary heart disease at Zutphen, The Nether- lands. Voeding 26:184. Dieckmann,W. J.,D. F.Turner,E. J.Meiller,etal. l951.Observationsonproteinintakeand health of mother and baby. II. Food intake. J. Am. Diet. Assoc. 27:1053. Dierks, E. C., and L. M. Morse. 1965. Food habits and food intakes of preschool children. J. Am. Diet. Assoc. 47:292. Dietary methodologies. 1973. In Nutritional assessment in health programs. Am. J. Public Health 63(Suppl.):11-18. Durnin, J. V. G. A., and E. C. Blake. 1962. A comparison of the food intake of elderly women estimated by individual inventory and the National Food Survey methods. Br. J. Nutr. 16:261. Eads, M. G., and A. P. Meredith. 1948. Nutrition studies. II. Methods of collecting dietary data. Public Health Rep. 63:777. Eagles, J. A., M. G. Whiting, and R. E. Olson. 1966. Dietary appraisal. Problems in proces- sing dietary data. Am. J. Clin. Nutr. 19:1. Emmons, L., and M. Hayes. 1973. Accuracy of 24-hour recalls of young children. J. Am. Diet. Assoc. 62:409. Eppright, E. S., M. B. Patton, A. Marlatt, and M. Hathaway. 1952. Dietary study methods. V. Some problems in collecting dietary information about groups of children. J. Am. Diet. Assoc. 28:43. Eppright, E. S., E. Jebe, V. Sidwell, et al. 1955. Nutrition of 9-, 10- and 11-year-old public school children in Iowa, Kansas and Ohio. Iowa State Coll. Iowa Agric. Exp. Stn. Bull. 434:616. Eppright, E. S., H. M. Fox, B. A. Fryer, et al. 1972. Nutrition of infants and preschool children in the north central region of the United States of America. World Rev. Nutr. Diet. 14:269. Epstein, L. M., A. Reshef, B. Abramson, and O. Bialik. 1970. Validity of a short dietary questionnaire. Isr. J. Med. Sci. 6:589. Exton-Smith, A. N., and R. R. Stanton. 1965. Report of an investigation into the diet of elderly women living alone. Edwards Hospital Fund for London. Fidanza, F. 1971. Sources of error in dietary surveys. Bibl. Nutr. Dieta 1972. 20:105. Fidanza, F., and A. Fidanza-Alberti. 1967. Dietary surveys in connection with the epidemiol-

112 CHARLOTTE M. YOUNG ogy of heart disease. Reliability, sources of variation, and other data from nine surveys in Italy. Voeding 28:244. Flores, M. 1972. Methodology in nutrition surveys among preschool children. Arch. Latinoam. Nutr. 22:359. Frank, G. C., G. S. Berenson, P. E. Schilling, and M. C. Moore. 1977. Adapting the 24-hr. recall for epidemiologic studies of school children. J. Am. Diet. Assoc. 71:26. Fry, P. C., H. M. Fox, and H. Linkswiler. 1963. Nutrient intakes of healthy older women. J. Am. Diet. Assoc. 42:218. Futrell, M. F., L. Kilgore, and F. Windham. 1971. Nutritional status of negro preschool children in Mississippi. Evaluation of HOP index. J. Am. Diet. Assoc. 59:218. George, P. S., and G. A. King. 1961. Consumer demand for food commodities in the United States with projections for 1980. Calif. Univ. Agric. Sci., Giannini Found. Monogr. 26. 161 pp. Gersovitz, M., J. P. Madden, and H. S. Wright. 1978. Validity of the 24-hr. dietary recall and seven-day record for group comparisons. J. Am. Diet. Assoc. 73:48. Grant-Whiting, M., and R. M. Leverton. 1960. Reliability of dietary appraisal. Comparison between laboratory analysis and calculation from tables of food values. Am. J. Public Health 50:815. Gray, C. E., and N. R. Blackman. 1947. More high school student diets evaluated. J. Home Econ. 39:505. Groover, M. E., Jr., L. Boone, P. C. Houk, and S. Wolf. 1967. Problems in quantitation of dietary survey. J. Am. Med. Assoc. 201:86. Grossman, E., and D. Popka. 1976. Methodology of a household food consumption survey in Cincinnati, Ohio, Nov. 1969-Jan. 1970. Unpublished. Guggenham, K., J. Ilan, T. Vogo, and G. Mindel. 1960. Nutritional status and food consump- tion of pregnant women in a development area of Israel. Br. J. Nutr. 14:347. Guthrie, H. A. 1963. Nutritional intake of infants. J. Am. Diet. Assoc. 43: 120. Hankin, J. H., and R. Huenemann. 1967. A short dietary method for epidemiologic studies. I. Developing standard methods for interpreting seven-day measured food records. J. Am. Diet. Assoc. 50:487. Hankin, J. H., W. E. Reynolds, and S. Margen. 1967. A short dietary method for epidemiologic studies. II. Variability of measured nutrient intakes. Am. J. Clin. Nutr. 20:935. Hankin, J. H., R. A. Stallones, and H. B. Messenger. 1968. A short dietary method for epidemiologic studies. III. Development of questionnaire. Am. J. Epidemiol. 87:285. Hankin, J. H., H. B. Messenger, and R. A. Stallones. 1970. A short dietary method for epidemiologic studies. IV. Evaluation of questionnaire. Am. J. Epidemiol. 91:562. Hankin, J. H., G. G. Rhoads, and G. A. Glober. 1975. A dietary method for epidemiologic study of gastrointestinal cancer. Am. J. Clin. Nutr. 28:1055-1061. Hankin, J. H ., M. S . Rawlings, and A. Nomura.1978. Assessment of a short dietary method for a prospective study on cancer. Am. J. Clin. Nutr. 31:355. Heady, J. A. 1961. Diets of bank clerks. Development of a method of classifying the diets of individuals for use in epidemiological studies. J. R. Stat. Soc. Ser. A (Gen.) 125:336. Hiemstra, S. 1968. Food consumption, prices, expenditures. U.S. Dep. Agric., Econ. Res. Serv. Agric. Econ. Dep. Rep. 138. 193 pp. Huenemann, R., and D. Turner. 1942. Methods of dietary investigation. J. Am. Diet. Assoc. 18:562. Huenemann, R. L., F. E. French, and J. M. Bierman. 1961. Diets of pregnant women in Kauai, Hawaii. J. Am. Diet. Assoc. 39:567. Hummel, F. C., e~ al. 1942. Chemical composition of twenty-two common foods and compari-

Dietary Methodology son of analytical and calculated values of diets. J. Nutr. 24:41. 113 Hunter, G., J. Kastetic, and M. Ball. 1948. Assessment of diets: Analysis versus computation from food tables. Can. J. Res. Sec. E 26:367. ICNND. Manual for nutrition surveys, 2nd ed. 1963. U.S. Government Printing Office, Washington, D.C. Chap. VII, Dietary studies, pp. 159-230 and 249-260. Kaser, M., et al. 1947. A comparison of the calculated and determined caloric and vitamin content of mixed diets. Am. J. Hyg. 46:297. Keys, A. 1968. Dietary survey methods. In den Hartog, C., K. Buzina, F. Fidanza, A. Keys, and P. Roine, eds. Studies of cardiovascular epidemiology of heart disease. M. Wyt & Sons, Rotterdam. Pp. 9-29. Keys, A. 1979. Dietary survey methods. In R. Levy, B. Rifkind, B. Dennis, and N. Ernst, eds. Nutrition, lipids, and coronary heart disease. Raven Press, New York, Pp. 1-23. Koehne, M. 1935. Probable accuracy of dietary studies. J. Am. Diet. Assoc. 11: 105. LeBovit, C., and D. Baker. 1965. Food consumption and dietary levels of older households in Rochester, New York. U.S. Dep. Agric. Home Econ. Res. Rep. 25. 91 pp. Lechtig, A., C. Yarbrough, R. Martorell, H. Delgado, and R. D. Klein. 1976. The one day recall dietary survey: A review of its usefulness to estimate protein and calorie intake. Arch. Latinoam. Nutr. 26:1243. Leitch, I., and F. C. Aitken. 1949. Technique and interpretations of dietary surveys. Nutr. Abst. Rev . 19:507. Leverton, R. 1937. A comparison of values obtained by calculation and by analysis for iron content of 85 mixed diets. J. Am. Diet. Assoc. 13:139. Leverton, R. M., and A. G. Marsh. 1939. Comparison of food intakes for weekdays and for Saturday and Sunday. J. Home Econ. 31:111. Leverton, R. M., and J. Pazur. 1957. Food practices and nutritional status of typical Nebraska families. Nebr. Univ. Agric. Exp. Stn. Misc. Publ. 5. 52 pp. Linusson, E. E., D. Sanjur, and E. C. Erickson. 1974. Validating the 24-hour recall method as a dietary survey tool. Arch. Latinoam. Nutr. 24:277. Lockhart. E. E., et al. 1944. Study of nutritional quality of dietaries by chemical analysis. J. Am. Diet. Assoc. 20:742. Madden, J. P., S. J. Goodman, and H. A. Guthrie. 1976. Validity of the 24-hr. recall. Analysis of data obtained from elderly subjects. J. Am. Diet. Assoc. 68:143. Majonner, L., and Y. Hall. 1968. The national diet heart study assessment of dietary adher- ence. J. Am. Diet. Assoc. 52:288. Manalo, R., and J. Jones. 1966. The content of constant diets: A comparison between analyzed and calculated values. Am. J. Clin. Nutr. 18~39. Mann, G., G. Pearson, T. Gordon, and T. Dawber. 1962. Diet and cardiovascular disease in the Framingham study. I. Measurement of dietary intake. Am. J. Clin. Nutr. 11:200. Marr, J. W. 1971. Individual dietary surveys: Purposes and methods. World Rev. Nutr. Diet. 13:105. Marr, J. W. 1973. Dietary survey methods. In Individual and group aspects. Proc. R. Soc. Med. 66:639. Marr,J. W.,J. A.Heady,andJ. N.Morris. 1961.Towardsamethodforlarge-scaleindividual diet surveys. Pages 85-91 in Proc. 3rd Int. Congr. Dietet. Newman Books, London. McCann, M. B., M. F. Trulson, and F. J. Stare. 1961. Follow-up study of serum cholesterol, diet and physical findings of Italian-American factory workers. Am. J. Clin. Nutr. 9:351. McCay, C. M., et al. 1945. Nutritive value of food served in some large Navy messes. J. Am. Diet. Assoc. 21:88. McHenry, E. W., H. P. Ferguson, and J. Gurland. 1945. Sources of errors Can. J. Public Health 36:355. in dietary surveys.

14 CHARLOTTE M. YOUNG Meredith, A., A. Matthews, M. Zickefoose, et al. 1951. How well do school children recall what they have eaten? J. Am. Diet. Assoc. 27:749. Mickelsen, O., et al. 1947. The determination of fat and its relation to the calculated caloric value of diets. J. Am. Diet. Assoc. 23:952. Ministry of Agriculture, Fisheries and Food, National Food Survey Committee. 1972. House- hold food consumptions and expenditures. Her Majesty's Stationery Office, London. 218 pp. Mongeau, E. 1974. Dietary assessment of nutritional status. The assessment of nutritional status, Proc. Miles Symp., June 25, 1973. Saskatchewan University, Canada. Moore, M. C., B. C. Judlin, and P. McA. Kennemur. 1967. Using graduated food models in taking dietary histories. J. Am. Diet. Assoc. 51:447. Morgan, A. F., ed. 1959. Nutritional status. Calif. Agric. Exp. Stn. Bull. 769. 131 pp. Morris, J. N. 1970. Uses of epidemiology. Livingstone, Edinburgh. Morrison, S. D., C. F. Russell, and J. Stevenson. 1949. Estimating food intake by question- ing and weighing: A one day survey of eight subjects. Br. J. Nutr. 3:V. Murray, J. 1970. USDA food consumption surveys A historical review with special reference to the development and testing of collection methods. Unpublished. Norris, T. 1949. Dietary surveys: Their technique and interpretation. FAG Nutritional Studies No. 4, Washington, D.C., December 1949. 108 pp. Ohlson, M. A., L. Jackson, J. Bock, et al. 1950. Nutrition and dietary habits of aging women. Am. J. Public Health 40:1101. Owen, G., P. Garry, K. Kram, et al. 1969. Nutritional status of preschool children. A pilot study. Am. J. Clin. Nuts 22:1444. Owen, G., K. M. Kram, P. J. Garry, et al. 1974. A study of nutritional status of preschool children in the United States, 1968-70. Pediatrics 53(Pt. II, Suppl.):597. Pao, E. M., and M. C. Burk. 1975.` Portion sizes and day's intakes of selected foods. U.S. Dep. Agric. ARS-NE-67. 70 pp. Patterson, L. 1971. Dietary intake and physical development of Phoenix area children. J. Am. Diet. Assoc. 59:106. Patterson, J. M., and E. W. McHenry. 1941. Errors with calculation of the nutritive value of food intake. I. Comparison of calculated and determined amounts of calories, protein and fat. Can. J. Public Health 32:362. Paul, O., M. H. Lepper, W. H. Phelan, et al. 1963. A longitudinal study of coronary heart disease. Circulation 28:20. Pekkarinen, M. 1970. Methodology in the collection of food consumption data. World Rev. Nutr. Diet. 12:145. Pekkarinen, M., S. Kivioja, and L. Jortikka. 1967. A comparison of the food intake of rural families estimated by a one-day recall and precise weighing method. Voeding 28:470. Pekkarinen, M., S. Kivioja, and L. Jortikka. 1968. A comparison of the food intake of rural families estimated by one-day recall and precise weighing methods. In C. den Hartog, K. Buzina, F. Fidanza, A. Keys, and P. Roine, eds. Dietary studies and epidemiology of heart diseases. M. Wyt & Sons, Rotterdam, The Hague. Pitcher, H. L., J. M. Leichseming, S. P. Frommes, and L. M. Norris. 1961. Comparison of analyzed and calculated energy values of food intakes of adult women. Metabolism 10:475. Proceedings of Conference on Methods for Evaluating Nutritional Status of Mothers, Infants and Children. 1947. Res. Lab. Children's Fund of Michigan, Detroit. 93 pp. Mimeo. Quackenbush, G. G., and J. D. Shaffer. 1960. Collecting food purchase data by Consumer Panel A methodological report on MSU Consumer Panel, 1951-58. Mich. State Univ. Agric. Exp. Stn. Tech. Bull. 279. 73 pp. Reagan, B., and E. Grossman. 1951. Rural levels of living in Lee and Jones Counties, Mississippi, 1945 and a comparison of two methods of data collection. U. S. Dep. Agric. Inform. Bull. 41 . 164 pp .

Dietary Methodology 115 Reddy, S. 1968. Practical problems in obtaining valid and reliable information on household food utilization. J. Trop. Pediatr. 14:66. Reed, R. B., and B. S. Burke. 1954. Collection and analysis of dietary intake data. Am. J. Public Health 44: 1015. Reshef, A., and L. Epstein. 1972. Reliability of a dietary questionnaire. Am. J. Clin. Nutr. 25:91. Response Analysis Corporation. 1976. Report of findings food methodology study. Prepared for Consumer and Food Economics Institute, U. S. Department of Agriculture, Hyattsville, Md. Roberts, L. J., and M. Waite. 1925. A dietary study made in a day nursery by the individual method. J. Home Econ. 17:80; 142. Roine, P., and M. Pekkarinen. 1968. Methodology of dietary surveys. Int. Z. Vitaminforch. 11:31. Samuelson, G. 1970. An epidemiological study of child health and nutrition in a northern Swedish county. II. Methodological study of the recall technique. Nutr. Metab. 12:321. Samuelson, G. 1971. An epidemiological study of child health and nutrition in a northern Swedish county. I. Food consumption survey. Acta Paediatr. Scand. Suppl. 214. 44 pp. Schaefer, A. E. 1966. Assessment of nutritional status: Food intake studies. Page 217 in G. H. Beaton and E. W. McHenry, eds. Nutrition: A comprehensive treatise, Vol. III., Chap. 6, Academic Press, New York. Semmons, E. M., and E. W. McHenry. 1944. Errors in the calculation of the nutritive value of food intake. IV. Comparison of calculated and determined amounts of calcium. Can. J. Public Health 34:286. Sprauve, M. E., and M. L. Dodds. 1965. Dietary survey of adolescents in the Virgin Islands. J. Am. Diet. Assoc. 47:287. Stead, R. M. 1968. Assessment of methods of dietary appraisal. A comparison of two differ- ent modifications of Burke's dietary history method. S. Afr. Med. J. 42:961. Steele, B. F., and R. D. Tucker. 1952. Influence of dietary interpretations as calculated nutritive value of the diet. Page 78 in C. M. Young, et al. Cooperative nutritional status studies in the Northeast Region. III. Dietary methodology studies. Univ. Mass. Agric. Exp. Stn. Bull. No. 469. Steele, B. F., R. E. Franklin, V. L. Smudski, and C. M. Young. 1951. Use of checked seven-day records in a dietary survey. J. Am. Diet. Assoc. 27:957. Steele, B. F., V. L. Smudski, R. E. Franklin, and C. M. Young. 1952. Comparison between three methods of assessing nutrient intake of children. Page 59 in C. M. Young, et al. Cooperative nutritional status studies in the Northeast Region. III. Dietary methodology studies. Univ. Mass. Agric. Exp. Stn. Bull. No. 469. Stefanik, P. A., and M. F. Trulson. 1962. Determining the frequency intakes of foods in large group studies. Am. J. Clin. Nutr. 11:335. Stefferud, A., ed. 1959. Food: The yearbook of agriculture, 1959. U.S. Government Printing Office, Washington, D.C. Stevens, H., R. Bleiler, and M. Ohlson. 1963. Dietary intake of five groups of subjects. J. Am. Diet. Assoc. 42:387. Stock, A. L., and E. F. Wheeler. 1972. Evaluation of meals cooked by large-scale methods: A comparison of chemical analysis and calculation from food tables. Br. J. Nutr. 27:439. Strong, J. A., and R. Passmore. 1967. Adaption to overfeeding in man. Proc. Nutr. Soc. 26:163. Sudman, S., and R. Ferber. 1971. Experiments in obtaining consumer expenditures by diary methods. Am. Stat. Assoc. J. 66(336):725-735.

116 CHARLOTTE M. YOUNG Swanson, P., E. W. Willis, E. Jebe, et al. 1959. Food intakes of 2,189 women in five north central states. Iowa State Coll., Iowa Agric. Home Econ. Exp. Stn. Res. Bull. 468. Pp. 476-499. Symposium American Dietetic Association. 1951-1952. Dietary study methods, October 1950. J. Am. Diet. Assoc. 27:558, 745, 838, 968; 28:43. Symposium- American Institute of Nutrition. 1950. Symposium on Nutritional Appraisal of Human Subjects, Atlantic City, April 20, 1950. Maynard, L. A. Evaluation of dietary survey methods. Fed. Proc. 9:598. Symposium American Public Health Association. 1960-1961. Methodology for dietary studies in epidemiological surveys, 1959. Am. J. Public Health 50:803, 815, 824; 51:70. Symposium on Recent Advances in the Appraisal of the Nutrient Intake and Nutritional Status of Man. 1962. MIT, Boston, March 1962. Am. J. Clin. Nutr. 11:335 -404. Tasker, J., et al. 1967. Dietary survey by weighment method A comparison of random day, 3-day and 7-day period. Indian J. Med. Res. 55:90. Taylor, E.C. 1970. The impact of computer analysis on the design of dietary surveys. Proc. Nutr. Soc. 29:135. Third International Congress of Dietetics Proceedings. 1961. Wm. Bytes & Sons, Ltd., Brad- ford, Yorkshire, England. Thomas, R. V., et al. 1950a. Nutritional status of children. XII. Evaluation by computing the food intake of a group and by weighing and analyzing food eaten by representative subjects. J. Am. Diet. Assoc. 26:788. Thomas, R. V., M. Rutledge, E. Beach, et al. 1950b. Nutritional status of children. XIII. Accuracy of calculated intakes of food components with respect to analytical values. J. Am. Diet. Assoc. 26:889. Thomson, A. M. 1958. Diet in pregnancy. 1. Dietary survey technique and the nutritive value of diets taken by primigravidae. Br. J. Nutr. 12:446. Tinsley, W. V. 1947. Development of instruments for evaluating food practices, nutrition information, and school lunch programs and their use in nutrition education at the elemen- tary level. Doctor's thesis, Univ. of Minn., St. Paul. 333 pp. (Abstr. in J. Home Econ. 41:514, 1949.) Toscani, V. 1948. Comparison of analyzed with calculated diets. Food Res. 13:187. Trulson, M. F. 1951. Assessment of methods of dietary intake. Synopsis from Ph.D. thesis. Mimeo to author. Trulson, M. F. 1954. Assessment of dietary study methods. I. Comparison of methods for obtaining data for clinical work. J. Am. Diet. Assoc. 30:991. Trulson, M.F. 1955. Assessment of dietary study methods. II. Variability of eating practices and determination of sample size and duration of dietary surveys. J. Am. Diet. Assoc. 31:797. Trulson, M. F. 1960a. Report of Diet Methodology Committee. In Conference on Methodol- ogy in Epidemiological Studies in Cardiovascular Disease, Princeton, N.J., April 24-26, 1959. Am. J. Public Health Suppl. 50. 124 pp. Trulson, M. F. 1960b. Subcommittee on Methodology for Diet Appraisal. Pages 39-52 in Conference on Methodology in Epidemiological Studies in Cardiovascular Disease, Prince- ton, N.J., April 24-26, 1959. Am. J. Public Health Suppl. 50: October 1960. Trulson, M. F., and M. B. McCann. 1959. Comparison of dietary survey methods. J. Am. Diet. Assoc. 35:672. U.S. Department of Agriculture, Agricultural Research Service. 1972a. Food and nutrient intake of individuals in the United States, spring 1965. U.S. Dep. Agric. Household Food Consumption Survey, 1965-66, Rep. No. 11. 289 pp. U.S. Department of Agriculture, Agricultural Research Service. 1972b. Food consumption of households in the United States, seasons and year, 1965-66. U.S. Dep. Agric. Household Food Consumption Survey, 1965 -66. Rep. No. 12. 217 pp.

Dietary Methodology 117 U.S. Department of Health, Education, and Welfare, Health Services and Mental Health Administration, Center for Disease Control. 1972. Ten-state survey, 1968-70. DHEW Publ. No. (HSM) 72-8133. 340 pp. Van Schaik, Th. F. S. M. 1968. Practical problems in estimating dietary intake at home. Int. Z. Vitaminforsch. 11:43. Viteri, F., and B. Torim. 1975. Ingestion calorica y trahajo fisico de obreros agricos en Guatemala. Effect de la suplementacion alimentaria y su lager en los programos de salud. Boll Of. San. Pan. 78:58. Wait, B., and L. J. Roberts. 1932. Studies in the food requirement of adolescent girls. II. Daily food requirements in the energy intake of the individual. J. Am. Diet. Assoc. 8:323. Wakefield, L. N. 1966. The interview technique in research Sources of bias. J. Home Econ. 58:640. Walberg, S., and W. Adams. 1965. Calculated and determined values in constant diets. J. Am. Diet. Assoc. 47:37. Wertz, A. W., M. E. Lojkin, E. H. Morse, et al. 1952. Comparison of determined and calculated amounts of eight nutrients in one day's food intake of twenty-one subjects. In C. M. Young, et al. Cooperative nutritional status studies in the Northeast Region. III. Contributions to dietary methodology studies. Univ. Mass. Agric. Exp. Stn. Bull. No. 469. White, H. S. 1969. Inorganic elements in weighed diets of girls and young women. J. Am. Diet. Assoc. 55:38. Widdowson, E. M. 1947. A study of individual children's diets. Med. Res. Counc. Spec. Rep. Ser. No. 257. His Majesty's Stationery Office, London. 196 pp. Widdowson, E. M., and R. A. McCance. 1943. Food tables: Their scope and limitations. Lancet 1:230. Wiehl, D. G., and R. Reed. 1960. Development of new or improved dietary methods for epidemiological investigations. Am. J. Public Health 50:824. Youland, D. M., et al. 1976. Practices and problems in HANES. Dietary data methodology. J. Am. Diet. Assoc. 68:22. Young, C. M. 1965. Comparison of results of dietary surveys made by different methods. Pages 119-126 in New horizons in dietetics. Proc. 4th Int. Congr. Dietet. Ivar Hoeggstrom Tryckeri AB, Stockholm. Young, C. M., and E. W. McHenry. 1942-1943. Errors in the calculation of the nutritive value of food intake. II. Comparison of calculated and determined amounts of ascorbic acid. III. Comparison of calculated and determined amounts of iron. Can. J. Public Health 33:224, 34:367. Young, C. M., and K. Musgrave. 1951. The use of the dietary scorecard. J. Am. Diet. Assoc. 27:745. Young, C. M., et al. 1951a. Nutritional status survey, Groton Township, New York. II. Nutrient usage of families and individuals. J. Am. Diet. Assoc. 26:776. Young, C. M., V. Smudski, and B. Steele. l951b. Fall and spring diets in school children in New York State. J. Am. Diet. Assoc. 27:289. Young, C. M., F. W. Chalmers, H. N. Church, et al. 1952a. A comparison of dietary study methods. I. Dietary history vs. seven-day record. J. Am. Diet. Assoc. 28:124. Young, C. M., G. C. Hagan, R. E. Tucker, and W. D. Foster. 1952b. A comparison of dietary study methods. II. Dietary history vs. seven-day record vs. 24-hour recall. J. Am. Diet. Assoc. 28:218. Young, C. M., et al. 1952c. Cooperative nutritional status studies in the Northeast Region. III. Contributions to dietary methodology studies. Univ. Mass. Agric. Exp. Stn. Bull. No. 469, Amherst. 95 pp. Young, C. M., F. Chalmers, H. Church, et al. 1953a. Subject's estimation of food intake and calculated nutritive value of the diet. J. Am. Diet. Assoc. 29:1216.

8 CHARLOTTE M. YOUNG Young, C. M., R. Franklin, W. Foster, and B. Steele. 1953b. Weekly variation in nutrient intake of young adults. J. Am. Diet. Assoc. 29:459. Yudkin, J. 1951. Dietary surveys: Variation in weekly intake of nutrients. Br. J. Nutr. 5:177. Yudkin, J., and J. Roddy. 1966. Assessment of sugar intake: Validity of the questionnaire method. Br. J. Nutr. 20:807. Ziffer, H., O. Frank, G. Christakis, et al. 1967. Data analysis strategy for nutritional surveys of 642 New York City school children. Am. J. Clin. Nutr. 20:858.

Mail Diary Method for Collecting Food Purchasing and Food Usage Information from Consumer Panels I. J. ABRAMS INTRODUCTION The Market Research Corporation of American (MRCA) has specialized in collecting information from large samples on household purchases of gro- cery products, textile products, and other frequently bought consumer prod- ucts using the mail diary techniques. Similarly, it has collected detail infor- mation on household food preparation and on individual food intake. The National Consumer Panel (NCP) consists of a nationally distributed sample of 7,500 households, reporting every week in a printed diary their purchases of selected foods and other grocery products, on a continuing basis for extended periods of time. The National Household Menu Census (MC) Study, usually conducted every 5 years, is based on a subsample of 4,000 households selected from the NCP. Each study spans a period of 12 months, in which each household reports for 14 consecutive days all food preparation and consumption at home and away from home, using separate daily diaries. The NCP has been operating commerically since 1942, and the Menu Census Service has been repeated each 5 years since 1957, including an interim one-half sample size study conducted in 1975. These two services are designed to satisfy the information requirements of food-processing and marketing companies and of their advertising agencies. Reports are sold on a syndicated basis, whereby the costs are shared by the multiple users of the same information. 119

20 THE NATIONAL CONSUMER PANEL Objectives I. I. ABRAMS NCP measures the end of the food distribution channel, where the product enters the home of the consumer. It thus identifies who buys the product, how often, at what price, and from which store. Since these measurements are made continuously over time, NCP iS able to identify all the competing brands, forms, and other product variations that are available to the con- sumer in a given food category and the switching patterns that consum- ers make in purchase choices among the different brands over time. The data from NCP are usually summarized on a monthly and quarterly cycle, with special analyses performed on a semiannual or an annual basis. The data are used to track consumer purchases in a product category over time; to evaluate the effectiveness of various marketing, promotion, advertising, and pricing strategies employed by the client or by its competitors; to monitor the "life cycle" of various brands or products; to evaluate new entries into the market; to spot target segments exhibiting special consumer preferences; to identify opportunities for new products; to evaluate potential acquisitions of individual brands or of food manfacturing companies; and in general to provide a continuously updated data base suitable for diagnosing problems encountered in the market place, devising solutions to these problems, and methods of evaluating the success of such solutions. The essential characteristic of these data is that they are collected lon- gitudinally from a large sample of individual households, on a frequent basis, for predefined produce categories, on a sufficient detail basis to identify brands, types, subtypes, forms, packages, package sizes, individual prices paid by the consumers, the use of coupons or other purchase incen- tives, and the identity of the stores from which the purchases were made. Of course, this information represents consumption by the entire household and is not usually broken down by each individual member of that household, except in special cases. In this respect, the information is still at an aggre- gate level, and not at the individual eater level. Sample Design and Maintenance The basic design of NCP iS a stratified probability sample, incorporating 440 primary sample units. These consist of about 220 different locations repre- senting all large cities, most of the intermediate-size cities, and samples from the smaller cities, rural communities, and farm areas. Most of these geographical areas are then divided into two cells, representing small versus large households. Since the sample is operated continuously over time, the

Mail Diary Method 121 main function of this cell structure is to provide a framework for maintaining the balance of the sample to the population and to control all replacement recruiting efforts. This maintenance or replacement recruiting is purposive in nature and usually consists of replacing a lost household with another one living in the same area and having similar demographic characteristics. Households that move from their original area are maintained in the sample and when necessary are later replaced by new households that live in the area to which they moved. Thus, NCP continues to diffuse across the entire United States from its primary sampling unit. In fact, its distribution antici- pates the results that are later obtained through updates in the U.S. Census. Recruiting of replacement households is done in several ways. The more frequent method is by mail, using telephone lists. Personal interviewing may be used when the need arises to replace special kinds of households. At times, telephone recruiting may be used. Since it is usually difficult to recruit newly formed young households, NCP resorts to recruiting children of existing panel members when they leave home and set up their own house- holds. NCP does not accept volunteers or references provided by current panel members. Recruited households are trained for a period of from 1 to 3 months prior to their inclusion in the production sample. During that period, their weekly diaries are examined carefully for accuracy and completeness of recording, and write-backs to the households are used to inform them of any errors or misinterpretation of instructions. Compensation Reporting households are compensated with a variable schedule of "points," which are redeemable for gifts through a premium catalog, simi- lar to green stamps or other such saving stamps. A nominal number of points is given for returning a diary, whether on time or late, regardless of whether any purchases are reported in it. Bonuses are given for prompt reporting and for returning all diaries in the month. Periodically, NCP households receive additional questionnaires, and once a year they receive the basic Household Classification Questionnaire. Additional points are always given for such supplementary questionnaires. A special effort is made to encourage the continuity of panel membership of hard-to-keep young households by pro- viding gifts at the birthdates of small children. Weekly Diaries and Other Questionnaires The weekly diary has been recently redesigned to permit tearing out and returning only pages in which the household reported purchases of specified

122 I. J. ABRAMS products. This is an effort to control the mail costs, which constitutes a significant portion of the operations cost of a weekly diary. The front cover of the diary captures from the household information on every trip to any store that sells products contained in the diary. Households are encouraged to record each trip as soon as it is completed, in order to remind the house- hold to report purchases of products contained in the diary. The date and identity of the stores is reported only on the face of the diary, and the sequential number of the store trip is then reported on the first column of each entry. A standard entry inside the diary provides space for reporting the brand of the specific product; some product details unique to that category; the number of packages purchased; the size of each package; the unit or total price paid; and any special coupons, discounts, or store deals. Great care is exercised in the description of the products to be reported, through examples provided to further describe that product within various check boxes for that entry, through the description of other characteristics of the purchase, and in the positioning of each diary section in relation to other sections on that page and on other pages of the diary. When new sections are added to the diary, they are usually printed on the bottom half of the front cover, in order to bring their attention to the panel members, and later the section is inserted into an appropriate location within the diary. Changes in the positions of various items in the diary are made very infrequently, so that households are able to remember the locations of most products that must be entered, and thus reduce the total effort required in maintaining such diaries. A detailed Index of Products to be entered is printed on the inside front cover of the diary, to help the reporter find the location where to enter a specific item, as well as to inform the reporter whether or not a specific product should be reported in the diary. Frozen and refrigerated products are positioned in the early pages of the diary, since they are usually unpacked and stored first when the household returns from a shopping trip, followed by shelf stable canned and packaged products, and concluded with household and personal care products. Space is provided for entering purchases that cannot fit in the space allocated in the body of the diary and also occasionally for reporting products that could be entered in one of several places, but which the household is not sure where to report. Space is also provided at the end for reporting changes in the composition of the household that week. The Household Classification Questionnaire is administered each April to all NCP households. A similar form is also completed by each household when it is recruited. Detail information is reported on the income, educa- tion, and other demographic characteristics of each household member; and information on ownership of several appliances, on the type of residence, and on other characteristics that are required for classifying the household

Mail Diary Method 123 into the many categories that are used by manufacturers and by advertising agencies for analyzing patterns of household food purchasing. It is, of course, also used in balancing the sample to the population. Even though the questionnaire contains several sensitive personal items such as income, age, and the like, almost all households report this information on a regular basis. Occasionally a custom or temporary test market panel is recruited in locations specified by the client, where a new product is being tested by the manufacturer or by its competitors. Households for that market are recruited independently of NCP and are maintained for a limited period of time, usu- ally from 6 to 18 months. They report in a very small diary, which contains, in addition to the product category of interest, several other categories, whose functions are to mask the desired category, and thereby prevent overstimulating the panel members to purchase or explore items under test. Households that complete their test panel assignments may later be recruited into the National Consumer Panel if it is deficient in these kinds of house- holds in the particular market. Data Transcription and Editing All entries in those diary sections that are currently sold are transcribed by coders for machine processing. The diary sections are formatted to collect specific product characteristics that are needed by the client, and code structures are devised to capture this information in a unique way. Since all purchases of the specified category are reported by panel members, regard- less of what region of the country they live in, and from which store or outlet they purchased the item, including home deliveries, door-to-door salesman, mail order, home grown or canned, etc., the code structures are frequently extensive and detailed. They are continuously updated as new products or variations of existing products are reported by household members. These code structures also include information on known package sizes by brands, price ranges, outlets in which specific brands may be obtained, such as private label, and the like, and this information is then used to machine edit the transcribed information. All errors are returned to the coding department for verification or correction, and occasionally require follow-up with the reporting households. The diaries are usually mailed back on the Sunday or Monday following the reporting week, and most diaries are received by us by the following Friday. The entries are transcribed and checked by the end of the second week following the reporting week, at which time "flash" reports can be produced if needed. Reports are produced for two 4-week months, followed by a 5-week month each quarter. The quarters consist of 13 weeks each, adding up to a 52-week year. For some clients, the reporting is made each

124 I. J. ABRAMS 4-week month, or each 12-week quarter, resulting in 13 monthly reports per year, synchronized with the reporting cycle of SAMI. Reports Monthly and quarterly reports are usually delivered to clients 2 to 3 weeks following the close of the data period covered in these reports. Monthly and quarterly reports contain brand shares based on pounds or packages bought by households during the reporting period, projected to regions and U.S. totals; the number and percent of households who bought the product at least once during the time periods; the average frequency of purchasing occasions of each item; the average price paid; the average price paid when the pur- chase was on a special deal or store sale; the average quantity per buying household; the average size of a purchase; and other similar "buyer" statis- tics. Special reports frequently include statistics on share of total household requirements that were satisfied by each key product type or brand; percent of households who repeated purchasing the brand from the previous quarter or the previous year; and the amount of purchases of a particular category that were switched from one brand to another over two time periods. A substantial repertoire of special studies has been developed for proces- sing longitudinal consumer purchase information, in order to address unique problems that manufacturers encounter in the market place. Some of these analytical techniques are designed to identify target populations of "heavy users, " or "loyal buyers of a brand," or "promising potential buyers" of a particular brand or type. Special attention has been devoted to predicting the rate of adoption of a new product in the market, and ultimate success or failure of a new product, and the level that it may attain in the second or third year following its introduction, based on as short an observation period as possible. Validity and A ccuracy of the Data NCP and test panels are based on samples of ''Households'' following U.S. Census definitions. They usually exclude persons living in institutions, such as schools, dormitories, hospitals, residence hotels, and the armed forces. They do include households composed of single individuals, or of unrelated individuals sharing a common housing unit. Therefore, information reported by these panels is representative only of food purchases made by the corres- ponding household segment of the population. This excludes about 5 percent of the population. Since all reporting is done by mail and in writing, the samples exclude households that are illiterate, or those containing no potential reporter who

Mail Diary Method 125 writes and reads English. Also, since recruiting is frequently done from mailing lists extracted from telephone directories, the sample tends to be deficient in households that do not own telephones. In general, the projected sample is reasonably well balanced to U.S. Census on most key demographic characteristics. Since panel members have to be willing to report their purchases of a large number of product categories on a weekly basis, over extended periods of time, the sample will tend to have very few members in which the homemaker, or other potential reporter, is unwilling to undertake such a task. The willingness to do this work is, in a broad sense, correlated with the interest of the homemaker in the task of purchasing foods, or in other aspects related to maintaining a home, to shopping, and to reporting on such activities. Although the sample has representative members from every strata of the population, cooperation is lowest among single young persons, mothers with small children, persons with very low income, and especially those living in central cities. By contrast, households in middle-income-class suburbia, or with a full-time homemaker, tend to cooperate more readily when recruited, and then tend to stay in the panel for much longer periods of time. Incentives for cooperation vary widely and are not necessarily related to the level of compensation for this work. Of course, many homemakers consider the gifts that they get through cooperation to be "independently earned income." Some homemakers view cooperation as a method of influencing manufacturers to produce the kind of products they prefer. On the whole, data projected from NCP match reasonably well levels and trends of purchases known from independent sources. However, some pro- ducts suffer from some degree of undercoverage, which may be caused in part by a forgetfulness of the panel members to report all purchases of the product and in part from underpurchasing by panel members of a specific product because it is, in the eyes of a more skilled or experienced or know- ledgeable consumer, more expensive or less economical. Underreporting can occasionally result from imprecise or ambiguous format and description of a product in the diary. This may occur especially for products whose food category is ambiguous and that actually do not belong clearly to any specific category. In fact, manufacturers try to identify new product categories and market new products, which do not have to compete with well-established brands. It is therefore necessary at times to experiment in the way the item is described in the diary before panel households learn to report the item properly. Similarly, overcoverage may occur because panel cooperators tend to buy the more economical sizes, or the brands that offer more value for the money, or to redeem coupons and buy items on sale more often, and the like. However, because these tendencies are relatively constant over time, their main effect may be on the level of the projected data, but not as

126 I. J. ABRAMS much on trends and other aspects of the information being collected. Be- cause a great deal is known about these biases in the data, it is possible to anticipate them and to account for them in producing continuing or special studies. The accuracy of any item of information depends to a large part on the degree of variability in purchase patterns from one household to the next, and therefore on the size of the sample and on the percent of households who regularly buy that item. The smaller the number of households buying the item, the higher the variability in the computed average amount per buying household. Since the number of households buying any particular item in- creases for longer time periods, it is therefore occasionally possible to obtain the desired level of accuracy in an estimated purchase incidence of fre- quency for a brand at a semiannual or an annual level, when it is too small on a monthly or quarterly basis. When additional accuracy is required for a monthly or a quarterly time period in a given area, it is occasionally possible to augment the NCP in a particular region by adding "test panel" households to the sample. In general, the size of test market panels is usually deter- mined by the level of accuracy that is needed in the reported information. THE NATIONAL HOUSEHOLD MENU CENSUS SERVICE Objectives The Menu Census Service is designed to collect detail information on the ultimate disposition of all food products in terms of their consumption by individuals as obtained, or their use in preparation of other dishes to be eaten by individuals. In this respect, the Menu Census observes the true end of the food distribution system from the farm to its final use. From the point of view of food intake by individuals, information collected by the National Consumer Panel still represents an aggregate measurement at that point in the distribution system where the in-home inventory is replenished. This in-home inventory can be substantial, and foods may remain there for ex- tended periods of time. Although liquid milk, fresh meat, bread, and similar perishable products are generally consumed within a few days from their entry into the home, others, such as powdered milk, canned soup, spices and seasonings, dehydrated vegetables, instant mixes, and the like, are fre- quently stocked by the household for weeks, months, or even years after purchase, and are thus further removed from individual eating. Some of the primary uses of data collected in the Menu Census Studies are: (a) Early identification of trends in food usage by type, form, packag- ing, etc., which may lead to new product innovation, product modifications, new marketing strategies applied to existing products, new advertising ap-

Mail Diary Method 127 preaches, and the like. (b) A better understanding of the way specific pro- ducts or forms of a product are used or served by household or by individu- als, leading to the identification of population segments with special needs, and thus suggesting more effective marketing or advertising techniques, or to modifications of the form or the packaging of the product. (c) A determi- nation of the customer base of a given food category, leading to a decision to enter that market, or to acquire an existing manufacturer or brand in the market, or to discontinue participation in that category. (d) More recently, the Menu Census has been used to evaluate the exposure of individuals, by various age-groups, or by other characteristics, to specific food additives from the diet, in connection with regulatory activities of various gov- ernmental agencies, such as FDA, FTC, USDA, and others; the data have been used by FDA and the National Academy of Sciences to compute individual intake of food additives in the GRAS survey, PHASE Il. and are currently being used in the PHASE Ill survey. (e) The data have been used by FDA for ad hoc studies of the exposure of individuals to various direct or indirect additives to foods; they have similarly been used by commercial clients to evaluate the levels of exposure of individuals to alternate food additives, such as sweeteners, colorings, flavorings, and the like, in the search for a replace- ment to a currently used substance whose continued use may later be re- stricted. (f) They have been used to define foods that belong to selected categories or subcategories, in terms of their substitutability, in judicial hearings involving two competing manufacturers, or involving the govern- ment and a manufacturer. (g) They have been used to evaluate the nutritional contributions of groups of specific foods, or of meals containing specific foods, or of the dietary status of individuals with differing food consumption patterns. Sample Design, Recruiting, Training, Daily Diary and Other Forms, Compensation The Menu Census is usually conducted once in 5 years, from July through the following June. It was begun in July 1957, and the last study was completed in June 1978. An interim study was also conducted during the calendar year of 1975. Each study consists of 4,000 households, selected from the NCP, including whenever possible only households that reported in NCP for at least 6 months. The 1975 interim study contained only 2,000 households, most of whom were drawn from the previous Fourth Menu Census Study, which was conducted during July 1972 through June 1973. Table 1 shows a comparison between the demographics of the Sixth Menu Census sample and the U.S. Census. Each household reports all food preparation and consumption at home and

128 I. J. ABRAMS TABLE 1 Sixth National Household Menu Census, Composition of the Sample, Annual Summary, July 1 977-June 1978- Households and Household Members Distributed by Demographic Characteristics Households Persons U.S. Census in Sample in Sample Households Household Characteristics No. % No. % % U.S. TOTAL 4,000 100.0 11,150 100.0 100.0 Census Areas Northeast 916 22.9 2,618 23.5 23.0 North Central 1,159 29.0 3,317 29.7 26.6 South 1,195 29.9 3,305 29.6 32.0 West 730 18.3 1,910 17.1 18.4 Census Regions-9 Way New England (NE) 205 5.1 587 5.3 5.6 Mid-Atlantic (NE) 711 17.8 2,031 18.2 17.4 East North Central (NC) 798 20.0 2,308 20.7 18.7 West North Central (NC) 361 9.0 1,009 9.0 7.9 South Atlantic (S) 609 15.2 1,679 15.1 15.9 East South Atlantic (S) 236 5.9 642 5.8 6.2 West South Atlantic (S) 350 8.8 984 8.8 9.9 Mountain (W) 188 4.7 495 4.4 4.6 Pacific (W) 542 13.6 1,415 12.7 13.8 Metro Area Size-8 Way Farm 165 4.1 466 4.2 3.9 Under 2,500 487 12.2 1,391 12.5 11.4 2,500-49,999 460 11.5 1,270 11.4 10.0 50,000-249,999 323 8.1 897 8.0 8.1 2SO,000-499,999 428 10.7 1,231 11.0 11.0 500,000-999,999 466 11.7 1,268 11.4 12.7 1-2Mill. 561 14.0 1,508 13.5 13.7 2Mill. &over 1,110 27.8 3,119 28.0 28.3 Household Income 6 Way Under $5,000 610 15.3 1,026 9.2 17.8 $5,000-$8,999 729 18.2 1,644 14.7 17.4 $9,000-$12,999 750 18.8 2,120 19.0 17.1 $13,000-$16,999 661 16.5 2,196 19.7 15.1 $17,000-$22,999 652 16.3 2,153 19.3 16.5 $23,000 & over 598 15.0 2,011 18.0 16.1 Education Hh. Head 3 Way Grammar (under 9 yr) 609 15.2 1,409 12.6 20.6 High (9-12 yr) 1,861 46.5 5,299 47.5 48.3 College (over 12 yr) 1,530 38.3 4,442 39.8 31.1

Mail Diary Method TABLE 1 Continued 129 Households and Household Members Distributed by Demographic Characteristics Households Persons in Sample in Sample Household Characteristics No. To No. % % Occupation Hh. Head 4 Way White collar 1,455 36.4 4,505 40.4 32.5 Blue collar 1,219 30.5 4,098 36.8 33.1 Farmer 79 2.0 254 2.3 2.2 Not a worker 1,247 31.2 2,293 20.6 32.2 Homemaker Employment 2 Way Employed (6 mo or more) 1,445 36.1 3,928 35.2 NA Not employed (less than 6 mo) 2,555 63.9 7,222 64.8 NA Household Size 4 Way 1 Person 791 19.8 791 7.1 20.6 2 Persons 1,317 32.9 2,634 23.6 30.6 3-4 Persons 1,359 34.0 4,771 42.8 32.9 5 or more 533 13.3 2,954 26.5 15.9 Presence of Children 6 Way None 2,345 58.6 4,263 38.2 57.8 Under 6 yr only 505 12.6 1,819 16.3 12.3 6-12 yr only 277 6.9 1,047 9.4 6.5 13-17 yr only 317 7.9 1,215 10.9 8.7 6-12 & 13-17 yr 232 5.8 1,196 10.7 6.7 Under 6 & other age-groups 324 8.1 1,610 14.4 8.0 Age of Homemaker 6 Way Under 25 yr 223 5.6 655 5.9 11.3 25-34 yr 886 22.2 3,167 28.4 22.4 35-44 yr 569 14.2 2,373 21.3 16.7 45-54yr 635 15.9 1,970 17.7 17.3 55-64 yr 757 18.9 1,534 13.8 14.9 65 yr & over 930 23.3 1,451 13.0 17.4 Race 2 Way White 3,653 91.3 10,091 90.5 88.4 Nonwhite 347 8.7 1,059 9.5 11.6 Seasonal Totals Jul.-Sept. 1977 1,000 25.0 2,776 24.9 Oct.-Dec. 1977 1,000 25.0 2,718 24.4 Jan.-Mar. 1978 1,000 25.0 2,921 26.2 Apr.-Jun. 1978 1,000 25.0 2,735 24.5 U.S. Census Households

130 TABLE 1 Continued I. J. ABRAMS Households and Household Members Distributed by Demographic Characteristics Households Persons in Sample in Sample Household Characteristics No. % No. So So U.S. Census Households Individuals by Age and Sex: Male Total 3,142 78.6 5,127 46.0 48.6 Under 2 yr 169 4.2 171 1.5 1.5 2-5 yr 408 10.2 447 4.0 2.9 6-12yr 501 12.5 603 5.4 5.8 13-17 yr 382 9.6 471 4.2 4.9 18-24 yr 380 9.5 426 3.8 6.6a 25-34 yr 791 19.8 797 7.1 7.5 35 44 yr 561 14.0 561 5.0 5.3 45-54 yr 528 13.2 529 4.7 5.2 55-64 yr 523 13.1 523 4.7 4.5 65 yr & over 595 14.9 599 5.4 4.4 Female Total 3,950 98.8 6,023 54.0 51.4 Under 2 yr 159 4.0 160 1.4 1.4 2-Syr 375 9.4 426 3.8 2.8 6-12yr 493 12.3 606 5.4 5.6 13-17 yr 350 8.8 428 3.8 4.7 18-24 yr 436 10.9 466 4.2 6.6a 25-34yr 917 22.9 919 8.2 7.7 35 44 yr 571 14.3 572 5.1 5.5 45-54 yr 645 16.1 648 5.8 5.6 55-64 yr 758 19.0 763 6.8 5.1 65 yr & over 1,002 25.1 1,035 9.3 6.4 NOTE: The U.S. Census Population Estimates are based on residents of the United States (includes institu- tional population). a Includes all residents, whereas the Menu Census is based on household members only (excludes institu- tional population). away from home for 14 consecutive days in daily diaries. Households are distributed uniformly throughout the year, with about 11 households starting on each day of the year. Thus, there are about 153 households reporting on each day of the year. Each quarter contains exactly 1,000 households. The sample is balanced within each quarter, as closely as possible, to the U.S. Census by various demographic characteristics, including region, household size, household income, and the like. Households are invited to cooperate about 1 month before they will be needed, at which time they receive a yellow completed sample diary and a

Mail Diary Method 131 trial blank diary. When cooperators complete a trial diary, they are informed of any errors they may have made in reporting. When started, they are sent all 14 daily diaries and are instructed to begin reporting on a specified date. Incoming diaries are checked as received for accuracy and completeness, and write-backs are sent to the households to elicit additional details about the food used, or the nature of such use, as needed. Following the fourteenth day, the household receives a diet questionnaire, in which it reports information about each individual member of the house- hold, concerning weight, specific diets, foods encouraged to eat and encour- aged to avoid, and the consumption of vitamin supplements. In addition, the homemaker receives a questionnaire dealing with attitude, awareness, and interest in various subjects related to food preparation and usage and to life-style. The households are compensated with points for returning each diary or other form and for completing the entire study. These points tie in to their continuing compensation as members of NCP and therefore leverages on this other compensation that the household receives. Transcription, Editing, and Processing All information reported by the households in the daily diary and in the other questionnaires is coded, edited, and reported on a quarterly and on an annual basis. All foods are classified into about 65 major food categories and into as many as 99 different specific food items within each major category. Each item is further classified by three additional ways, depending on the specific major category. These additional classifications may refer to the presence of a particular additive, coating, flavoring, etc., or to the particular type of dish or way of serving, or to the specific form or source of the food as received by the household. In addition, the package type in which the food came is classified, as well as the method of cooking or preparation, and the appliance used in prepara- tion. Not only do we code every end dish as eaten, but also every ingredient used in its preparation. All transcribed information is machine-edited against the code structure, to ensure the validity of codes assigned to particular products and to ensure that all ingredients of a particular dish are properly connected to that dish. In addition, checks are made to ensure that every household member is ac- counted for in each main meal of every day. For example, if a person is recorded as having eaten a meal away from home, a check is made to verify that some food is reported as eaten by that person away from home. The resulting transcribed records on tape form a complete image of the informa- tion reported by the household for each meal for every day. Thus, any

132 I. J. ABRAMS analysis that could be specified on the basis of information written by the household in the diary could be carried out from the tape file. Reports Reports are produced at the end of each quarter and are delivered to sub- scribers about 4 months following the report period. Annual reports are processed and delivered immediately following the fourth quarter reports. Most subscribers follow up receipt of quarterly and annual reports with custom studies performed on various parts of the data base. These custom studies are usually executed with 2 to 6 weeks of authorization. They fre- quently involve the production of reports similar to the seven base reports, but executed on smaller or more finely defined subgroups of foods, or on special target households. These target households may be defined by the incidence of their usage of some foods in a particular way. For example, a study may be performed on the demographic classifications of households using both bottled soft drinks as well as powdered mixes in a 2-week period. Alternately, a study may be carried out, say, on the kind of starches served along with roast beef during the evening meal. As described above, special studies are also carried out on the cumulative exposure of individuals to specific additives or to particular nutrients from all foods eaten. Special analyses are carried out for many years following the execution of any one study and frequently involve developing trends in the pattern of use of a particular food product over 10 or 15 years. Validity, Completeness, Accuracy, and Biases in the Data It is relatively difficult to check the validity of Menu Census data for any particular product, because of the lack of easily obtainable independent reference information of similar detail. On the whole, based on 20 years of use by a large number of food processors and marketers, the Menu Census has proved to be a reliable reflection of patterns and trends in food prepara- tion and usage. As a rough check, some clients have estimated the coverage of specific product categories in the Menu Census as follows. An estimated average amount per eating occasion was multiplied by the frequency of eating of a specific product as reported in the Menu Census, and the resulting projected pounds of consumption in a 2-week period was compared to independently obtained average amount sold in an average 2-week period in that year. Such comparisons usually indicated a reasonable coverage of sales by the pro- jected Menu Census data. However, it is known that certain product categories are not well reported in the Menu Census. These include products

Mail Diary Method 133 with a high frequency of consumption outside the direct knowledge of the homemaker or the reporter, such as snack foods, candy, alcoholic bever- ages, and the like. Similarly, consumption of candy and other snacks instead of regular school lunches by children away from home may not be properly reported to the homemaker. Forgetfulness may also affect reporting, espe- cially of foods eaten by other household members away from home. The quality of reporting in the Menu Census is significantly enhanced by the fact that all members are drawn from the National Consumer Panel and are required to have participated in that panel for over 6 months whenever possible. Thus, reporting of food preparation and consumption is done by "experienced reporters," who have learned how to enter information in diaries in specified columns and sections, and to identify foods, packages, brands, etc., using "standard" terminology rapidly and accurately. Of course, as noted above, NCP households tend to be more interested in food purchasing, in economy, in "housekeeping," etc. Nevertheless, it seems unlikely that such characteristics would carry through in any appreci- able degree to affecting adversely patterns of food preparation and eating. It seems likely that these biases may contribute at most to second- or third- order effects in information that is of significant use in the Menu Census. Correlation Between Food Purchase and Usage by Menu Census Households The Menu Census diary does not request the reporting of quantities used on any serving or eating occasion. Therefore, average amount per eating occa- sion must be obtained from other sources when Menu Census data is used to quantify amounts of food additives or other substances consumed through the food by individuals of various age-groups. The current method uses averages computed from the USDA 1965 Survey of Consumption by Indi- viduals. We will shortly be able to use the data that has been collected by USDA on food intake by individuals in 1977-78. Another method for quantifying the average amount per eating occasion may be available by correlating the average quantities of specific foods purchased by Menu Census households, as reported by them in the weekly NCP diaries in the same time periods during which they also reported food usage in the Menu Census daily diaries. For example, during the Fourth Menu Census, households reported their purchases of cottage cheese every week and also reported all usage occasions of cottage cheese during a 2- week period in the daily diaries. These two separate sources of information could be correlated, using a standard regression analysis model, for all 4,000 households, and thereby derive average estimated amounts per eating occasion for persons classified into several different age-groups. Since cot-

34 I. J. ABRAMS tage cheese is a frequently purchased product, it appears that such estimates would be reasonably stable. A project is currently under way to develop such estimates for a large number of food categories whose purchases were reported in NCP during the period from July 1972 through June 1973, overlapping the Fourth Menu Census period. No results are as yet available from this study. If this study results in reliable estimates of consumption, then a good link would have been established between purchase information and consumption informa- tion for the same household. Therefore, one could track food consumption information indirectly by estimating it from purchase data. Since purchase data is usually less expensive to collect, one could collect Menu Census information only periodically, and in between one could follow up with observations of purchases of selected key products, from which estimates would be derived of the consumption of these same products by individuals . . in venous age-groups. In addition, since the relation between food purchases by households and movements of foods earlier in the pipeline is already well understood, one could use higher levels of aggregation of food movement at these earlier points in the distribution system to monitor emerging changes in food con- sumption. This possible relationship between aggregate and detail informa- tion has previously been discussed in the first workshop of this committee.

Collecting Data on American Food Consumption Patterns: An Anthropological Perspective CLAIRE MONOD CASSIDY According to Samuel Beck, English propaganda that the French lived on frogs and salad were implicity believed. Nathaniel Tracy, to entertain properly the admiral and officers of the French fleet, had~the swamps of Cambridge searched for green frogs, which were served whole in the soup at a formal dinner. The first officer who struck one with his spoon fished it out, held it up, and exclaimed: ''Moe Dieu! une grenouille!" The fascinated company embarrassed Tracy by roaring with laughter. R. O. Cummings, The American and His Food, 1940:30. INTRODUCTION The purpose of this paper is to consider how the anthropological perspective can contribute to the development of superior methods for collecting data on U.S. ("American") food consumption patterns. I will review selected liter- ature related to (1) anthropological studies of food habits among U.S. Americans and (2) theoretical orientations and techniques employed by an- thropologists that are of utility in the assessment of consumption patterns in urbanized society. ANTHROPOLOGY IN NUTRITION The anthropologist is fundamentally interested in variability (its sources in the social, cultural, historical, and biological spheres) and change (its patterning, constancy, rate alterations). These subjects are approached clas- sically using variants of the cross-cultural comparative method. Though 135

136 CLAIRE MONOD CASSIDY anthropologists formerly concentrated their research in prehistoric and "exotic" or non-Western societies, contemporary interest increasingly focuses on urban and urbanizing peoples worldwide. Sociocultural an- thropology developed and refined the now widely used nonquantitative technique of participant observation for eliciting information on behavior, beliefs and expectations from respondents. Montgomery ~ 1978:43) summarizes orientations to research he finds shared by anthropologists. His five points are worth quoting at length: 1) The recognition of the necessity to understand people in their own terms; 2) The maintenance of a healthy suspicion that what people say or believe bears no neces- sary relationship to what they do; 3) The recognition that much of ''culture" or "cultural behavior" is expressed in non-verbal ways; 4) The recognition that some of "culture" or "cultural behavior" is "unconscious" in that its existence, perfor- mance, or patterning is manifested without the actor's conscious or express aware- ness; 5) The recognition that, in any group of people, significant variations among and within individuals with respect to both beliefs and actions are to be expected. Cancian (1977) lists characteristics of anthropological science that make it especially effective in gleaning or disseminating information to/from non-Western peoples. (Here I shall use "non-Western" not as a geographic term but as one distinguishing philosophical positions not characterized by faith in and dependence upon "science" and "rationalism," hence poten- tially present among Americans.) Anthropologists are, then, scientists who are interested in discovering exceptions, in gaining width of perspective (holism) and in practicing patiently paced research strategies. They are often antiestablishment (or, at least, skeptical of the reality of "universal" truths). Anthropologists have long been interested in foods, ways of using foods ("foodways, " "food habits"), and nutrition. Earlier studies (e.g., Richards 1932, 1939; DuBois 1944; Honigmann 1961; Firth 1966; Rappaport 1968) considered foods as part of a larger ethnological picture and provided good data on attitudes and dietary content but typically little quantitative data adequate for assessing nutritional status. The latter subject has become of greater concern recently, especially with the emergence of the subdiscipline of nutritional anthropology. * Since numerous able reviews of this subject and *Nutritional anthropology focusses primarily on questions of nutrition and emphasizes the biocultural ecology approach. Students of the anthropology of food use the semiotic and folkloric approaches and are interested primarily in habits and beliefs associated with food use. The Committee on Nutritional Anthropology was founded in 1974. It is a branch of the Society for Medical Anthropology, recognized by the American Anthropological Association. Its newsletter is the CommuNicAtor. Within the International Union of Nutritional Scientists (NUNS) is the International Committee on Nutritional Anthropology (ICNA), while within the

Collecting Data on A merican Food Consumption Patterns 137 its history have recently appeared, I will not review it further here (Wilson, 1973; Netting, 1974; Arnott, 1976; Fitzgerald, 1976; Freedman, 1977; Johnston, 1977; Haas and Harrison, 1977; Montgomery, 1978; Montgomery and Bennett, 1978; Wilson, 19781. Anthropologists interested in foods and nutrition have used several theoretical onentations, including the historical (e.g., Cummings, 1940; Simoons, 1967; Chang, 1977~; folkloric (analysis of material culture as- sociated with food preparation, use, and storage); humanitarian (expressed as a concern with world hunger, which leads to consideration of food habits primarily as they affect the probability of malnutrition); biocultural; ecol- ogy; and semiotics. The biocultural ecology approach requires that foods and food habits be analyzed with respect to their capacity to promote adap- tation to the natural environment. The functional interpretation argues that under normal circumstances human groups will tend to find and practice food habits that optimize food distribution and maximize food value. An important implication of this approach is that the apparently "irrational" food habit may, when interpreted with greater care, prove rational. Take, for example, the refusal of milk ("nature's most nearly perfect food"!) by many malnourished peoples. This act appeared irrational until the genetic distribution of adult lactase insufficiency was discovered (McCraken, 19711. The hot-cold dichotomy, a health and nutrition patterning system probably descendant from Greek ideas of humoral pathology and still im- portant in many areas of the world, is currently believed to have no basis in physiological fact. But Lindenbaum (1977) mentions in passing emerging data that contradicts this interpretation. In the semiotic approach, "Food has been studied not for itself, but as a medium for social and for cognitive expression . . . " (Singer 1978:31. Singer (1978) defines three research goals for nutrition in anthropology, including in ascending inclusive order: (1) to use semiotic insights "the various treatments of food as symbol" to understand "eating . . . as an expressive act which . . . is trying to say something about something"; (2) to study "how the cultural system fits with the food production and nutri- tional systems"; and (3) to analyze the historical transformations of food systems. ''Internal dialectic, external pressure, and the necessities of survi- val may combine to lead to a changing fit between cultural, productive, and International Union of Anthropological and Ethnological Scientists (ICAES) iS the Commission on the Anthropology of Foods (ICAF) and the Committee on the Anthropology of Foods. Montgomery and Bennett (1978) list two journals and four newsletters devoted to the subjects of food, nutrition, and anthropology; more recent appearances include Appetite, The Journal for Research on Intake, Its Control and Its Consequences (international); The Digest, a News- letter for the Interdisciplinary Study of Food (Graduate Group in Folklore and Folklife, Uni- versity of Pennsylvania); and Foodtalk (San Francisco).

138 CLAIRE MONOD CASSIDY biological spheres thus creating new ways of eating and of feeding" (1978:3-4~. These ideas are significant to the analysis of U.S. food con- sumption patterns because they suggest both how broad must be our approach to collection and how potentially useful the final product may be if we can isolate and identify the more significant sources of pressure for change (or stability) and use these for prediction or intervention. FACTORS ASSOCIATED WITH FOOD CONSUMPTION What Is A Food? A food is at once a substance that provides energy or nutrients to maintain body function, a substance capable of assuaging hunger pangs, and a sub- stance defined as "edible." These meanings are not necessarily equally recognized by users, and in practice the third meaning is usually the most significant to the largest number of people. Edibility itself is not a simple concept, a fact recognized by the Commit- tee on Food Habits of the National Research Council (1945), which said: . . . [from the class of all] edible materials [some] are classified as inedible, edible by animals, edible by human beings, but not my kind of human being, edible by human beings such as self, and finally edible by self. These classifications are further reinforced by various sorts of attitudes that materials which are not eaten are defiling, wicked to eat, coarsening, would alter one's status, etc. The list in the second part of this section suggests the large number of factors that (potentially) affect the actual consumption of food in the family and by individuals. The terms "food habits" and "floodways" are sometimes distinguished (Cussler and DeGive, 1952), but recently have been used interchangeably as "habits of a group that reflect the way a culture standardizes behavior of the individual in the group in relation to food, so that the group comes to have a common pattern of eating" (Lowenberg et al., 1974: 1 17~. Leininger (1969) attempts to distinguish food universals, valid cross- culturally from nonuniversals. Leininger's nine food universals are charac- teristics of food use presumed to be present and functioning in all human groups. Food is: ( 1) "used to provide body energy and statisfy biophysiological hunger"; (2) used to initiate and maintain interpersonal relationships with friends, kinsmen, and strangers; (3) a determinant of the "nature and extent of interpersonal distance between people;" (4) used to express social and religious ideas; (S) used to "illustrate social status, pres- tige, group achievements"; (6) used to help individuals cope with psychological stresses and needs; (7) used to reward, punish, or otherwise

Collecting Data on American Food Consumption Patterns 139 influence the behavior of others; (8) used to "influence the political and economic status of a group"; and (9) used to "detect, treat and prevent social, physical and cultural behavior deviations and illness manifestations" (Leininger, 1969:passim). The nonuniversals, in contrast, assort food use by (1) culture-associated preferences (affected by various taboos, symbols, social structure, environ- mental availability, etc.~; (2) relationship to the supernatural, magic, sor- cery, and curing (or health maintenance); (3) social stratification; (4) ap- pearance and manner of serving; (5) the "internal metabolic environment and its interaction with the external environment" (Leininger, 1969:passim). When we ask what is used as food by lay people, we find the list includes some items formally considered nonnutritive, such as edible earths, laundry starch, ice cubes, caffeinated beverages, hard liquor, coca leaf or betel nut, tobacco, and so forth (cf., Wilson, 1978~. Wison considers these worth including in any survey of food habits because some may add important amount of nutrients (e.g., calcium from the typical betel nut preparation), while others may interfere with good nutritional status (e.g., cigarettes, vocal. Choice and Availability of Food in the Family The actual choice of food made by families and individuals is governed by many factors, some of which are listed in Table 1. The list is garnered, in part, from earlier attempts to assess what information must necessarily be collected to wholly understand food consumption patterns, American or other (cf., Eggan and Pijcan, 1943; Manual for the Study of Food Habits, 1945; Mead, 1964; Leininger, 1969; Christakis, 1963; Jerome, 1969a; Pangborn, 1975; Haas and Harrison, 1977; Wilson, 1978~. The fact that these factors appear es a list does not mean they do not interact with each other; in fact, they overlap and share feedback relationships, the details of which remain to be fully worked out. METH ODS A number of attempts have been made to detail methodologies appropri- ate to the collection of food consumption data (The Problem of Changing Food Habits,` 1943; Manual for the Study of Food Habits, 1945; Mead, 1955; Christakis, 1973), while several authors have made suggestions of information that might be included in such methodologies (e.g., Eggan and Pijcan, 1943; Mead, 1949, 1964; Lee, 1957; Pangborn, 1975; Jerome, 1969a,b; Haas and Harrison, 1977; Montgomery, 1978; Wilson, 1978; cf., Table 1~.

140 TABLE 1 Factors Affecting Food Choice and Availability CLAIRE MONOD CASSIDY Category Item Comment Biologic Genetics of food choice Limits of human adaptability Ecologic Soil fertility, water micronutrient content, ecozone variety, climate Populations differ genetically in ways significant to food choice, e.g., lactase deficiency and adult milk use (Bunce, 1969; Davis, 1969; McCracken, 1971); PTC tasting and cabbage family attractiveness (Fischer, 1967) Age, sex, altitude, temperature, body size and composition; significance of agriculture as a recent alteration in food source and type (cf., Stini, 1971; Haas and Harrison, 1977) Seasonal variation in the food supply Economic Production, Politics of food supply; societal stability; fossil transportation, & fuel availability distribution Family income Storage facilities Sociologic Classifications Rate of change in society Household Acquisition familial decisions Ethnicity; religion; class; caste; sex; age; occupation; education; rural-urban Local homogeneity of culture; isolation; access to advertising and nutrition information; access to alternate food sources Purchasing power; technical ability to acquire (shopping habits, gardening, hunting, gathering knowledge); technical ability to prepare (status kitchen, food preparer education); family structure Distribution Daily work schedule; sex, age, and health of . . declslons members; food-sharing behavior; idiocultures Cognitive Conceptual- Sacred, ordinary, and "fun" foods; sexuality izations of enhancing foods; health-promoting foods; edibility and health-damaging foods; allowable combinations food of foods; folk classifications of foods; categorizations color and shape associations. a Body image Preference for "fat" or "lean" shapes by age, sex; demonstrations of social status by body habitue (Cassidy, in press) a In the United States (and elsewhere), "white" chicken meat has higher status value than "dark,'' white rice than brown, white bread than dark (though we are possibly witnessing a change in status for some of these). Mead ( 1949) tells how shape of the serving dish may take precedence over its contents, hence nonmeat combinations will be accepted as meal "main dishes" if they are served in the appropriate container.

Collecting Data on American Food Consumption Patterns 141 American society is vast, complex (i.e., multiple interacting networks of production, distribution, education, control, and accession), heterogeneous, mobile, individualized. We face a large problem when we decide to develop a methodology that will distill from this mass a utilitarian (i.e., usable for purposes of prediction) description of American food consumption patterns. Anthropologists have classicaly concentrated their research in small and non-Western societies. There were two good reasons for this. A small soci- ety is "graspable" researchers can feel some confieance that the whole (at least most) of its character Anthropologists have classically concentrated their research in small and non-Western societies. There were two good reasons for this. A small soci- ety is "graspable"- researchers can feel some confidence that the whole (at least most) of its character can be discovered and perhaps described and understood; its boundaries are defined, hence so is the necessary extent of the anthropological inquiry. Second, the non-Western society is (usually) different enough from their own society to force the researchers to take little for granted, to rediscover, and to question their own categories of the "nor- mal," "ordinary," and "natural." The two advantages of working in the small and the distant are missing when we choose to analyze~our own society. Therefore, in developing a methodology, we must pay particular attention to (1) identifying biases (cultural, educational, paradigmatic) that affect our ability to recognize significant data and to design effective research strategies and (2) deciding how much data we want to collect. Identifying Biases The failure to recognize biases can cause incomplete data collection or the development of hypotheses based on invalid assumptions. In either case, the predictive value of findings is significantly reduced. For example, in Belize, Lowland Maya feast periodically on a pig's head. Men, the butchers and for these occasions the cooks, serve themselves the muscle portions, while women receive the fat, eyes, and brain. During my fieldwork there, I was told by another anthropologist that this distribution pattern was "another example of males overpowering females" here, specifically of taking the nutritionally better portions for themselves. This interpretation reflects bias and is probably proportionately invalid. The biases include the American one against organ foods and fat but towards protein (especially muscle) foods, and an ethnological one, which, expressed as an assumption, reads "men hold most of the power worldwide." This latter has the evidence of many earlier ethnographic works behind it and appears to have been "prov- en" by recent research that shows women of child-bearing age frequently

142 CLAIRE MONOD CASSIDY receive disproportionately unbalanced diets (cf., Rosenberg, 19731. Yet it has not, in my opinion, received adequate scrutiny. In any case, because of the unrecognized presence of these biases, this researcher reached premature closure on an ethnological problem. Alternate interpretations include phrasing the male-female power relationship in a form accessible to re- search, such as, "Where is the locus of women's power?"; and recognizing the important nutritional components of fat, eyes, and brain (fatty acids, fat-soluble vitamins), which are, in fact, relatively scarce in the typical daily Lowland Mayan diet. In this way we are led to see that the food distribution favors women nutritionally, and at the same time symbolically reifies an important masculine-feminine distinction, that of strength (meat is a "strong" food in Belize). Finally, because fat is highly desirable in most nonindustrialized societies (and among subgroups in our own society), the provision of fat to women cannot imply any male lack of regard for women. Other assumptions or tendencies of mind that are common and neces- sary to avoid in developing methodology are the very subtle ones of: · considering the norms of one's own group as normal (i.e., confusing norm and normal), which easily leads to viewing variability as abnormality or, in nutritional terms, as inadequacy; · considering the group with which we identify (e.g., white middle class, nutrition professionals) as homogeneous with regard to food con- sumption patterns; or · accepting as real the existence of reference persons, modal per- sonalities, culture types, and so forth. For both of the latter it is important to recognize that the amount of in- tracultural diversity is frequently greater than the amount of intercultural diversity (Pelto and Pelto, 1975y, especially as the detailed examination of groups inevitably leads to heightened sensitivity to subtleties of preference and use. Arensberg and Niehoff (1975) distinguish characteristics of American culture that are particularly important as they affect food use or our ability to design research. Though racially, ethnically, religiously, and occupation- ally heterogeneous, Americans express a rather narrow range of opinion on moral, political, economic, and social subjects. They are quite conforming with regard to language, hygiene, dress, diet, basic skills, land use, com- munity settlement, and recreation, facts Arensberg and Niehoff associate with our efficient mass education system. Status differences are present, but not very marked, and fall mainly along lines of occupation, education, and financial worth. There is considerable geographic and occupational mobil- ity. The family is commonly nuclear, with few children. In spirit the country is strongly secular and rationalistic, and displays strong interest in comfort

Collecting Data on American Food Consumption Patterns 143 (often expressed in terms of cleanliness), achievement, and success. The latter may be had by effort and efficient use of time (i.e., speed, precision). Anyone can succeed (egalitarianism); personal liberties exist and should not be infringed upon (individualism). Humans are almost infinitely perfectible; the person should attempt to improve himself. Progress not only may be had, but is almost a constant; the future orientation is also expressed in the accent on youthfulness and in an overwhelming concern with providing for children. Humanitarianism or concern for the welfare of others is highly valued, but is usually expressed impersonally. Twofold judgements (work: play, morel: immoral, civilized:primitive, practical:impractical, adequate: inadequate) are characteristic this puts the world of values into absolutes and tends to force users into positions of exclusiveness. "Science" is a belief system and method that has been, in the last 100 years, so acceptable to Americans as to have become almost a faith. It is based in rationalism, and profits from American emphases on speed, accu- racy, and the future orientation. What is labelled "scientific" is also almost inevitably labelled "progressive." However, the twofold judgement is re- placed in science by the concept of continua, an emphasis on gradual varia- tion, and often by a desire to avoid judgement except in flexible (rejectable) terms. Contrasts to modal American and nonmodal American and non-Western systems may be found in Arensberg and Niehoff (1975), Jelliffe and Jelliffe (1976), and Cancian (1977~. Jerome (1969a) distinguishes seven "themes" in American culture (indi- vidualism, democracy, capitalism, industrialism, pluralism, youthfulness, leisure) and shows how these affect food habits. For example, individualism and the emphasis on self-expression are shown through autonomy in food selection and consumption and the use of "image" foods that symbolize reward, reassurance, nostalgia, power, strength, etc. To ask Americans to give up certain foods for therapeutic or preventive purposes is to ask them to forsake independence and autonomy, even the democratic process; depen- dency is much feared by Americans. Capitalism stimulates demand through advertising; competition builds available variety of foodstuffs. Pluralism means that certain foods and food habits are at the core of the diet and are recognized as "American" by outsiders (e.g., hamburgers, sliced white bread, fruit drinks, carbonated beverages, apples, milk), while others (sec- ondary food habits) reflect the basic eating patterns of subcultures (regional, ethnic, religious) or the incorporation of nontraditional food items or whole cuisines into total foodways (accessory eating patterns). The interest in leisure is played out through the increasing use of restaurants and snack foods. Youthfulness is emphasized by the avoidance of fat and "fattening foods" and the use of low-calorie or "health" foods. Much work on American food habits, though able to provide guidance, is

144 CLAIRE MONOD CASSIDY nevertheless rather impressionistic (e.g., Mead, 1943; Joffe, 1943; Lee, 1957; Mitchell et al., 1968; Lowenberg, 1974~. However, some details on local groups has been provided through the important and interesting work of Lewin, 1943; Bennett et al., 1942; Jerome, 1969b, 1975. Defining Research Boundaries Food consumption analysis can be carried out at several levels of increasing precision, including the national, regional, community, neighborhood, household, and individual (the "idioculture" of food habits). The job of this workshop is to detail methods to collect information on the household and the individual, hence our research boundaries are already partially specified. Nevertheless, we must decide (1) how to conceptualize and categorize the enormous amount of information available on households and individuals and (2) how to discover what potential information we will consider irrele- vant to our purpose, for research on cultural subjects is almost infinitely extendable. There are at least four reasons for collecting data on American food consumption habits: · to learn more about American culture; · to assess the nutritional adequacy of diets and status of the population; · to identify the malnourished (presumably to intervene); and · to introduce "new " foods or recipes or alter proportions to dietary items for economic or public health reasons. A weighting of the importance of these four (which I do not wish to imply by this list) would provide a first step toward identifying research boundaries. Included as problems here are the needs for: · a measure of homogeneity in food use; this could be approached through the use of the core, secondary core, etc., categorizations as quantified by Jerome ~ 1975~; · some built-in characteristics to ensure flexibility so that change may be identified and described. Some Minor Points Included here are some ideas that simply do not fit handily elsewhere. Sexist or racist terminology must be avoided in designing research in- struments. For example, Christakis' (1963) dietary methodology section distinguishes between "homemakers" who are female and "heads of

Collecting Data on American Food Consumption Patterns 145 household" who are male and on question forms does not provide for working women (a majority of American women) or one-parent families. The decision to classify families on the basis of male occupation alone can be severely biasing. In the same source, all infants and children are referred to as "he" while their caretakers are called "she." New categories to include "health," "natural," and "ethnic" foods and use of "food supplements" (e.g., yeast, "Tiger's milk," vitamins, etc.) need to be developed. Instruments such as "Basic 7" or "Basic 4" charts should contain in each category-a broad enough range of foods to include those familiar to people of all ethnicities. The categories should be defined in nutritional terms, not in terms of cultural ideals. Research instruments, where possible, should be preceded for the com- puter. Computer programs can be developed early to handle the data; some sophisticated ones that permit mapping (e.g., of subtleties of food use countrywide) are now fairly readily available. It is important to reach pre- decisions on the handling of subjective data before mass surveys begin. Techniques Used by Anthropologists Included in this section are a small number of papers by anthropologists or nutritionists with anthropological training that describe novel, or employ adaptations of classical, techniques and that provide insights of utility to our present purpose. Wilson (1978) points out that the approaches used by anthropologists are considerably individualized and suggests that there has been little attempt at standardization because so few such studies have been done thus far. The classical anthropological technique is participant observation. This technique 'requires the researcher to live with and participate in the life- styles of study populations. The potentials for discovering subtleties and for detailing the systematics of cultural behavior are excellent, particularly in small and well-bounded populations such as those having band or tribe organization, or among people identified by community or caste member- ship. The technique requires highly trained personnel and the devotion of large blocks of time (months, years) to research. It becomes progressively less manageable as population size or sociopolitical complexity rise. It does not typically provide quantified data. These characteristics are technical liabilities in urbanized societies where information on large groups is wanted quickly and cheaply. However, a number of adaptations of partici- pant observation have been developed over the years that overcome many of these problems. These adaptations involve, first, selecting a specific problem-topic on

146 CLAIRE MONOD CASSIDY which to focus research. Techniques from other disciplines are sometimes added to the research plan to guide interviews or to provide quantified data. Thus: (1) Respondents may be followed and observed with regard to the problem-topic intensively for brief periods without the use of interviews. Wilson (1974) recommends this approach as superior for collecting dietary data on toddlers who cannot respond well to a recall technique. A possible drawback, especially in a mobile society, is that if the child is taken elsewhere the researcher may not be able to follow. (2) Questionnaire-guided topical interviews may be combined with brief periods of socializing with respondents (e.g., Jelliffe et al., 1962~. (3) Living near respondents and making frequent visits may be combined with depth interviews and/or the use of questionnaires on the problem-topic (e.g., Bennett et al., 1942; Jerome, 1969b, 1975~. (4) Living with respondents may be combined with use of depth interviews on the problem-topic and the collection of quantified data (e.g., McArthur, 1962; Cassidy and Stavrakis fieldwork, 1973-74~. In each of these variations, researchers select population subgroups from whom to elicit details on the problem-topic. Selection occurs on many bases, from the decision to use a small number of "key informants" who are willing (self-selection) and able (time and language are common barriers) to spend extended periods in contact with the researcher to random samples of special-interest groups (e.g., female heads-of-household) who then further self-select according to willingness to be interviewed or to permit the an- thropologist to observe them in their homes. Interviews may be partially or wholly structured; anthropologists typi- cally prefer the "open-ended" style, because it gives more leeway for the inclusion of respondent opinion and belief and decreases the risk of missing significant information, because of unrecognized biases introduced by the questionnaire-maker. Bennett et al. (1942) completed a now-classic study of a southern Illinois rural community in which they used partially structured interviews and free-association techniques to elicit information on food habits. Late in their study, they employed paired-comparison to cross-check accuracy of knowl- edge derived from interview. They spent 3 months on the project, visited all over the study area, but lived with only one of the subgroups they identified ("lower hills" people). They introduced the concepts of core (staple, regu- larly used foods), secondary core (recently introduced store-bought foods), and peripheral (very recent additions stimulated by special economic con- ditions) diets as means to order the mass of descriptive data they collected. The result was excellent, but possibly limited in two ways: There was no measure of nutritional adequacy of diets; the choice to live with one group produced better understanding of that group and possibly altered relation- ships with other and rival groups in the area.

Collecting Data on American Food Consumption Patterns 147 Jerome (1969b, 1975) studied how the food habits and conceptualizations of food altered for rural black Americans as they migrated north to major urban centers and how black Americans in a mid-Western urban center adapt their diets to supermarket variety and personal food preference. Both studies employed multiple visits to respondent households and partially structured depth interviews, and both required several months of research time with subsequent computer work-up. Jerome borrows Bennett et al. 's terminology but alters and quantifies their definitions. Thus core includes the essentials of the diet, "consumed 2-3 times per day by ~25% of respondent house- holds"; secondary core includes the food "consumed about once a week by ~25% of respondent households"; peripheral includes the foods "con- sumed 1-2 times per month by ~25% of households"; and ceremonial/ marginal includes "foods and beverages on the borderline of either being completely excluded from the diet or only included on very few occasions by ~75% of respondent households" (Jerome, 1975:92-931. This study approach combines several characteristics that make it possibly of greatest utility for the collection of food consumption data on Americans. McArthur, to quote Wilson (1978:142), "combined participant observa- tion and detailed budgetary studies to several households during one period with "dropping in" at meal preparation times to see what was to be served and on occasion to share a meal with that family." The resultant data was semiquantified. Cassidy and Stavrakis, working with Lowland Maya in Belize, lived 12 weeks (1973-74) in one village, ate regularly in one household, participated fully in accessible village activities, made a census of households, and interviewed all female heads of household in multiple-depth interviews that covered the topics of obstetrical and health history, child health histories, food habits and beliefs, agricultural production, and sources of other food. We then selected eight households for a quantified dietary survey and, using a team of trained undergraduate aides, collected recipes and weighed food and plate waste for all family members in each of the selected households for two 24-hour days. The work included following children to observe snacking behavior and using recall to elicit snacking behavior and the con- tent of meals away from home from adults. Nutritional status was estimated from Guatemalan food tables prepared by INCAP. This approach provided quite complete information on the food beliefs and habits of a whole village and on the food intake of approximately 1/4 of the village. The participant observation aspect of this research worked well, but the in-home weighing was not wholly successful. This was mainly because, though we took precautions to minimize our impact and to minimize dietary change coincident with our visits, we knew our presence was disturbing, hence food use on survey days altered (mostly by protein enrichment). An

148 CLAIRE MONOD CASSIDY adjustment of the weighing portion of the approach would make it less obtrusive yet provide equally (if not more) useful data. In the households selected for intensive study, a week's food use could be collected by arriv- ing unannounced (having previously obtained permission) to observe and weigh, let us say, three morning meals, four noon meals, and three evening meals in a week, collecting at the same time recall information on the nonobserved meals for that day. Snacking behavior could be collected by recall, or by some combination of recall and following. Collection and analysis of resultant data requires use of the computer. Such an approach would be useful where very detailed and precise information was wanted, especially to establish "normality" guidelines, or to answer questions about, e.g., family food distribution patterns (sex, age, preference), the nutritional adequacy of different diets, or the amount of variation in the diet. The technique should be equally applicable to analysis of any well-defined population group. Another route to knowledge of food use is through the analysis of house- hold refuse. Adapting their method and theory from archaeology, the Gar- bage Project of Tucson, Ariz. (Harrison et al., 1975; Rathje and Harrison, 1978) has shown that food waste behavior varies by socioeconomic status and ethnicity and can serve as a measure of responsiveness to business cycles. The method has several advantages. It is a "nonreactive measure of behavior" that is, it measures "what people did, not what they think they did" (Harrison et al., 1975:13~. It is inexpensive and demands no time or cooperation from subjects. It permits household food waste to be estimated quantitatively. However, it provides no information on food beliefs or fam- ily food distribution patterns. Chassy et al. (1967) used the Guttman Scale of increasingly less inclusive steps to weight the importance of various foods in the diet and the change in weighting with urbanization in a model industrial town in Mexico. This scale can also be used to measure the effect of contact with different cui- sines. Me technique has been used by Sanjur et al. (1970) to assess child- feeding practices in a Mexican village, and DeWalt and Pelto (1977) mod- ified it too consider nutritional adequacy indirectly by assessing the degree of complexity of the diet, also in rural Mexico. This approach requires the use of a structured questionnaire on food consumption. It produces information on dietary patterning but no quantified data. However, it requires no highly trained intereviewers and permits information on large numbers of subjects to be collected. Table 2 summarizes these techniques (necessarily somewhat subjectively) according to criteria deemed important to the development of methodology by the Food and Nutrition Board's Committee on Food Consumption Pat- terns (letter of July 24, 1978~.

Collecting Data on American Food Consumption Patterns 149 TABLE 2 Preliminary Assessment of Utility of Some Food Consumption Collection Techniques Techniquesa 1 2 3 4 5 6 Sample size handled large moderate large small moderate moderate Individuals emphasized? no no no yes potentially yes Reaches specific populations yes yes yes yes yes yes Identifies those et risk moderate moderate low high high high Ease of collection high high high moderate moderate moderate Flexibility of collection lows lows uncertain low high high Quantified data? no no yes yes yes yes Categorical data? yes yes yes yes yes yes Subjective data? no some no yes yes yes Personnelb - - - +/_ + + Reliability moderated moderated highs high high high Predictive value moderate moderate highs moderate e high high Rapid handling ease high high high moderatef moderates moderates Cost-effectiveness moderate moderate high moderate fairly high fairly high a Technique 1 = Structured Questionnaires and Guttman Scale; 2 = Structured Questionnaires and Socializing; 3 = Household Refuse; 4 = Person Following; 5 = Depth Interviews and Food Use Observation or Quantification; 6 = Participant Observation and Food Use Quantification. b A " + " indicates highly trained personnel are necessary for some part of the research (usually the unstructured interviews); a '' - " indicates that personnel trained only briefly can administer questionnaires or make observations. Flexibility is considered low where structured questionnaires are used. Thigh or moderate for households or groups, not for individuals. Predictive value is dependent on sample size and other items. fSubjective data requires transformation to objective categories, hence computerization may get complex. _ A Possible Approach to the Collection of Food Consumption Data Table 3 summarizes, in a very preliminary manner, an approach to the collection of food consumption data on the American population. It is an- thropological (specifically ethnological) in outlook as it emphasizes partici- pant observation and the collection of "subjective" data. It is suggested as a platform for discussion. SUMMARY Because American society is large, heterogeneous, and mobile, we set our- selves a peculiarly difficult problem in deciding to design a methodology for

150 an as o ._ Ct Ct EM Ct C~ C~ C) .= o Ct C~ o ._ o V o o o o ._ C~ - o o C~ Ct o s~ P~ a' .= o C) ~ O o ._ U. ,~ ._ m ~ E~ ~ 1 1 ~ V' C) C) s _ , ~ 8, = ~ ~ a. y c,. O ~ ,c ~ c ~ ° ~e u ~ ~ ~ c ~ ~ - .= t_ 0 0 ~ .= ~ 0 C~ O.4 0 ~ ~ 0 3 ~ ~ o o 1 ~ ~° ~,= C~ 1 3 ~ ~ x IU. 1 o 1 ·— 1 3 1 U. 1 ~ 1 Ce 1 ~ ._ Ct Ct ~ _ ._ ca c~ C~ ~ — CL, ~ U. — Ct o C) ~_ C~ - Ct o

151 - c~ _ cat ~ C Hi- ~ ~ ._ 4,, ~ 1 ta E 3 to coo ~ 04 v ~ E = c <, ', j _ A, e 1 To 1 ·_ 1 0 ~ . _ . _ so: ~ 1 ~ ~ c 3 0 i, 0 e .`v ~ g .~: ~ ~ 0 0 E ·°~ ~ E ~ 0 c, 3 o ~ s~ 1 ~ o 1 ~ 1 ~3 1 o 1 c · ~ ~ ~ ~ ~) o ~ no A E ° E con ° ° Den O ~ ~ = ' - <~, V) ~ Cot ~ ~ c ~ o ._ ;^ - ct cat ._ ._ c' cat .. an: on ._ - ._ - c~ ._ - ._ cat on ._ cat cat - c) 3 c' 3 c~ - . r ~ . — _ c~ c~ O — ~ . c~ ,= ~ ._ ~ O c, c ~ m ~ Ct 4, ~_

152 CLAIRE MONOD CASSIDY the collection of food consumption data. The anthropological perspective is valuable as it emphasizes the existence and importance of intracultural di- versity, the constancy of change, the special characters of the Americans and the American scientist that must be adjusted for in research design, and a small number of techniques, mostly nonquantitative, that provide irre- placeable detail on subjective aspects of culture. REFERENCES Arensberg, C. M., and A. N. Niehoff. 1975. American cultural values. Pages 363-378 in The Nacirema, readings in American culture. Little Brown Co., Boston. Arnott, M. L. 1975. Gastronomy, the anthropology of food and food habits. Mouton, The Hague. Bennett, J. W., H. L. Smith, and H. Passin. 1942. Food and culture in southern Illinois, a preliminary report. Am. Soc. Rev. 7:645-660. Bunce, G. G. 1969. Milk and blindness in Brazil. Nat. Hist. 78:52-53. Cancian, F. 1977. Can anthropology help agricultural development? Culture Agric. 1(2): 1 -8. Cassidy, C. M. In press. The clinical significance of imaging body fat as beautiful. Soc. Sci. Med. Chang, K., ed. 1977. Food in Chinese culture: Anthropological and historical perspectives. Yale University Press, New Haven. Chassy, J. P., A. G. van Veen, and F. W. Young. 1967. The application of social science research methods to the study of food habits and food consumption in an industrializing area. Am J. Clin. Nutr. 20:56-64. Christakis, G., ed. 1963. Nutritional assessment in health programs. Am J. Public Health 63(Suppl.), November 1963. Cummings, R. O. 1940. The American and his food. University of Chicago Press, Chicago. Cussler, M., and M. L. DeGive. 1952. Twixt the cup and the lip. Twayne Publications, Inc., New York. Davis, A. E., and T. D. Bolin. 1969. Milk intolerance in Southeast Asia. Nat. Hist. 78:53-55. DeWalt, D. M., and G. H. Pelto. 1977. Food use and household ecology in a Mexican community. In T. K. Fitzgerald, ed. Nutrition and anthropology in action. Van Gorcum, Assen, The Netherlands. DuBois, C. 1944. The people of Alcor, a social-psychological study of an East Indian island. University of Minnesota Press, Minneapolis. Eggan, F., and M. Pijcan. 1943. Some problems in the study of food and nutrition. Am. Indigena 3:9-22. Firth, R. 1966. Housekeeping among Malay peasants, 2d ed. Athlone Press, London. Fischer, R. 1967. Genetics and gustatory chemoreception in man and other primates. Pages 61-81 in M. R. Kare and O. Maller, eds. The chemical senses and nutrition. Johns Hopkins University Press, Baltimore. Fitzgerald, T. K., ed. 1977. Nutrition and anthropology in action. Van Gorcum, Assen, The Netherlands. Freedman, R. L. 1977. Nutritional anthropology: An overview. In T. K. Fitzgerald, ed. Nutrition and anthropology in action. Van Gorcum, Assen, The Netherlands. Haas, J. D., and G. G. Harrison. 1977. Nutritional anthropology and biological adaptation. Ann. Rev. Anthropol. 6:69-182.

Collecting Data on American Food Consumption Patterns 153 Harrison, G. G., W. L. Rathje, and W. W. Hughes. 1975. Food waste behavior in an urban population. J. Nutr. Educ. 7:13-16. , Honigmann, J. J. 1961. Foodways in a Muskey community: An anthropological report on the Attawapiskat Indians. Department of Northern Affairs and Natural Resources, Ottawa, Canada. Jelliffe, D. B., J. J. Bennett, R. H. R. White, T. R. Cullinan, and E. F. P. Jelliffe. 1962. The children of the Lugbara, a study in the technique of paediatric field surveys in tropical Africa. Trop. Geogr. Med. 14:33-50. Jerome, N. W. 1969a. American culture and food habits, communication through food in the U.S.A. In J. Dupont, ed. Dimensions of nutrition. Colorado Associated University Press, Boulder. Jerome, N. W. 1969b. Northern urbanization and food consumption patterns of Southern-born Negroes. Am. J. Clin. Nutr. 22:1667-1669. Jerome, N. W. 1975. Flavor preference and food patterns of selected U.S. and Caribbean blacks. Food Technol. 29(6):46-51. Joffe, N. 1943. Food habits of selected sub-cultures. U.S. Bull. Natl. Res. Counc. No. 108. Johnston, F. W., ed. 1977. Anthropological aspects of human nutrition. School of American Research, Santa Fe. Lee, D. 1957. Cultural factors in dietary choice. Am. J. Clin. Nutr. 5:166-170. Leininger, M. 1969. Some cross-cultural universal and non-universal functions, beliefs, and practices of food. In J. Dupont, ed. Dimensions of nutrition. Colorado Associated University Press, Boulder. Lewin, K. 1943. Forces behind food habits and methods of change. Pages 35-65 in Natl. Acad. Sci./Natl. Res. Counc. Bull. No. 108. Washington, D.C. Lindenbaum, S. 1977. The "Lastcourse": Nutrition and anthropologyin Asia. Pages 141-155 in T. K. Fitzgerald, ed. Nutrition and anthropology in action. Van Gorcum, Assen, The Netherlands. Lowenberg, M. E., E. N. Todhunter, E. D. Wilson, J. R. Savage, and J. L. Lubawski. 1974. Food and man, 2d. ed. John Wiley & Sons, New York. McArthur, A. M. 1962. Malaya 12. Assignment report, June 1958-Nov. 1959. World Health Organization, Regional Office for the Western Pacific, Manila, The Philippines. McCracken, R. D. 1971. Lactase deficiency: An example of dietary evolution. Curr. An- thropol. 12:479-517. Manual for the study of food habits. 1945. Natl. Acad. Sci./Natl. Res. Counc. Bull. No. 111. Washington, D.C. Mead, M. 1943. The problem of changing food habits. Pages 20-34 in Natl. Acad. Sci./Natl. Res. Counc. Bull. No. 108 Mead, M. 1949. Cultural patterning of nutrition related behavior. J. Am. Diet. Assoc. 25:677- 680. Mead, M., ed. 1955. Cultural patterns and technical change. Mentor, New York. Mead, M. 1964. Food habits research: Problems of the 1960's. Natl. Acad. Sci./Natl.Res. Counc. Bull. No. 1255. Washington, D.C. Mitchell, H. S., H. J. Rynbergen, L. Anderson, and M. V. Dibble, eds. 1968. Cooper's nutrition in health and disease, 15th ed. J. B. Lippincott, Philadelphia, pp. 144-154. Montgomery, E. 1978. Anthropological contributions to the study of food-related cultural variability. Pages 42-56 in S. Margen and R. A. Ogar, eds. Progress in human nutrition, Vol. II. Avi Publishing Co., Inc., Westport, Conn. Montgomery, E., and J. W. Bennett. 1978. Anthropological studies of food and nutrition: The 1940's and the 1970's. In W. Goldschmidt, ed. The uses of anthropology. American Anthropological Association Special Publication No. 11. Netting, R. M. 1974. Agrarian ecology. Ann. Rev. Anthropol. 3:21-56.

154 CLAIRE MONOD CASSIDY Pangborn, R. M. 1975. Cross-cultural aspects of flavor preference. Food Technol. 29(6):34-36. Pelto, P. J., and G. H. Pelto. 1975. Intracultural diversity: Some theoretical issues. Am. Ethnol. 2: 1 - 19. The problem of changing food habits. 1943. Committee on Food Habits, Natl. Acad. Sci. /Natl. Res. Counc. Bull. No. 108. Washington, D.C. 177 pp. Rappaport, R. A. 1968. Pigs for the ancestors, ritual in the ecology of a New Guinea people. Yale University Press, New Haven. Rathje, W. L., and G. G. Harrison. 1978. Monitoring trends in food utilization: Applications of an archaeological method. Fed. Proc. 37:49-54. Richards, A. I. 1932. Hunger and work in a savage tribe, a functional study of nutrition among the southern Bantu. G. Routledge and Sons, London. Richards, A. I. 1939. Land, labour and diet in northern Rhodesia: An economic study of the Bema tribe. Oxford University Press, London. Rosenberg, E. 1973. Ecological effects of sex-differential nutrition. Unpublished oral presen- tation to American Anthropological Association, New Orleans. Sanjur, D., J. Craviota, and A. G. van Veen. 1970. Infant nutrition and socio-cultural influ- ence in a village in central Mexico. Trop. Geogr. Med. 20:443-451. Simoons, F. J. 1967. Eat not this flesh, food avoidances in the Old World. University of Wisconsin Press, Madison. Singer, E. A. 1978. Guest editorial: Thoughts on the new foodways. In The Digest, a news- letter for the interdisciplinary study of food 1(2):2-4. Stini, W. A. 1971. Evolutionary implications of changing nutritional patterns in human popula- tions. Am. Anthropol. 73:1018-1030. Wilson, C. S. 1973. Food habits: A selected annotated bibliography. J. Nutr. Educ. 5(Suppl. 1) ( 1):38-72. Wilson, C. S. 1974. Child following: A technic for learning of food and nutrient intakes. J. Trop. Pediat. Environ. Child Health 20:9-14. Wilson, C. S .1977. Research methods in nutritional anthropology: Approaches and techniques. Pages 62-68 in T. K. Fitzgerald, ed. Nutrition and anthropology in action. Van Gorcum, Assen, The Netherlands. Wilson, C. S. 1978. Developing methods for studying diet ethnographically. Pages 141-148 in E. E. Bauwens, ed. The anthropology of health. C. V. Mosby Co., St. Louis.

Individual emanation in Intake of Nubients by Day, Month, and Season and Relation to Meal Patterns: Implications for Dietary Survey Methodology HAROLD B. MOUSER and HELEN T. BEBB Common methods for collection of dietary intake of individuals include the food record or diary kept by the individual or a close family member. This method has been used by us extensively in a variety of settings including yearlong studies of food intake (Houser et al., 1969; Witschi et al., 1970a,b; B ebb et al., 1971; and Bebb et al., 1972~. Our goal in these latter studies has been to establish for each individual a group of data that would be represen- tative of a subject's usual or average or "typical" diet. Our observations indicate the marked variation of an individual's intake over time, as well as the variation between individuals. Because of this variation, it becomes difficult to assess the frequency and duration of data collection needed for a representative picture of an individual's usual or average food intake. The dietary intake of 127 adults who kept 3-consecutive-days diary records of their food intake at 12 intervals for a year are examined in this paper from the standpoints of individual variability and the numbers and frequency of records that are representative of the year's intake. In addition, from the data we will evaluate the ability to predict intake by selected food items, the relation of food-intake history to recorded intake, and the utility of modal patterns of intake. SUBlECTS Study Group I consisted of 32 men and 41 women who were chronically ill with multiple sclerosis and living at home. The subjects represented various stages of the disease from mild to severe. Ages ranged from 29 to 62 years; mean age for women was 43 years; mean for men, 44 years. 155

156 HAROLD B. MOUSER and HELEN T. BEBB Study Group II consisted of 54 healthy males who were employed in executive or management positions in a single corporation. They ranged in age from 23 to 62 years; mean age, 44 years. Study Group III consisted of 28 males, all professionals and executives, who were participating in a physical fitness program. They ranged in age from 27 to 58, with a mean age of 45.1 years. METHODS General Data Collection A diary record was kept by each subject for a period of 3 consecutive days at approximately monthly intervals (Houser et al., 19691. For Group I, the recording days were selected to ensure a sample of weekdays, weekend days, and holidays throughout the year. For Groups II and III, recording days were selected at random. All foods and beverages were recorded by the subjects in common mea- sures, i.e., cups or number of items by count. Special attention was given to notation of brand names, or unusual ingredients. Date, day of the week, and time and place of the meal or snack were recorded. The diary method demands a great deal of cooperation and motivation from subjects, particularly with 36 days of recording. Subjects in Group 1 showed 33 out of 106 subjects who did not complete records for 1 year or who had records considered unsuitable illness (3), death (1), withdrawal from study (4), and moving (1) accounted for 9 subjects. Ten did not submit satisfactory records and 14 did not return or complete records on schedule. Group II had a much higher success rate. Of 55 subjects who began the study, only 1 withdrew and the remaining 54 completed satisfac- tory records. One Group II subject submitted only 11 records (33 days). Six Group III subjects withdrew (Table 11. Of those who finished, the mean number of days to return records to nutritionist remained consistent over the year. For all groups the average number of days ranged from 3 to 4 for all sets of records. Subjects were called, if necessary, about returning records or if specific food items on the record need clarification. The fact that 94 out of 127 subjects received from 0 to 3 phone calls over the entire year also helps establish the records as reliable and helps demonstrate the feasibility of the study method. When records were processed for computer analysis, each food or bever- age item consumed was filed so that it could be identified by subject number, the day of record, date of record, day of the week, item number for the day, meal period code, actual time of intake, food table identity number used for nutrient calculations, usage code, which explained, for example, if the margarine was used as a spread or used over a vegetable, and a code for

individual Variation in intake of Nutrients TABLE 1 Number of Subjects in Dietary Intake Studies with Diary Recording for 1 Year (161 Subjects) or 6 Months (36 Subjects) by Completion of Record Status 157 Subjects MS Executives Fitness Failure to Complete No. No. No. No. % Unsatisfactory records 10 0 0 10 5.0 Nonadherence to schedule 14 0 5 19 9.0 Subject withdrew 9 1 3 13 7.0 Successful completion 73 54 28 155 79.0 TOTAL 106 55 36 197 100.0 quantity and type of measure. This complete identification of our food items permitted flexibility in computer calculation of results. For each subject, daily nutrient intake was averaged for the 36-day study period. Calories and grams of protein and fat from animal and plant sources, of carbohydrate from refined sugar and from natural sources, and alcohol were calculated, as were average daily amounts of calcium, phosphorus, iron, vitamin A activity from beta-carotene and preformed A, thiamin, riboflavin, niacin, and ascorbic acid. METHODS USED TO EVALUATE METHODOLOGY AND VARIATION To Examine Variation Over Time Difference in means For each subject, the 36 days of daily intake were reviewed 5 different ways, using Student's t-test to show difference in means. T-values were evaluated for significance at the 0.05 and 0.01 levels. In addition to the nutrients listed above, total protein, total fat, and total carbohydrate were calculated and tested as well, giving a total of 21 vari- ables tested for each subject. If a significant difference in means was found, both the level of significance and the direction of the difference were noted. Test No. 1: To evaluate a possible effect of differences in eating during summer and winter, study days falling during May through October were averaged and compared to the average of study days occurring during November through April. Test No. 2: To evaluate whether we could have collected data for 3 days a month at intervals of every other month instead of every month, we num- bered each 3-day set from 1 to 12, and tested the means of sets numbered 1, 3, 5, 7, 9, and 11 (odd numbered sets) with the means from the 3-day sets numbered 2, 4, 6, 8, 10, and 12 (even numbered sets).

158 HAROLD B. MOUSER and HELEN T. BEBB Test No. 3: To evaluate whether subjects eat differently on weekdays than they do on weekends, the average for all weekdays (Monday through Fri- day) was compared to the average for all weekend days (Saturday and Sunday). Test No. 4: To evaluate the possibility of collecting only 1 day a month instead of 3 days a month for a period of 1 year, we compared the mean obtained from all first days of each 3-day set with the mean from the remaining second and third days of each 3-day set. Analysis of Variance Analysis of variance was used to measure individual reporting consistency over the twelve 3-day sets. Variation within the 3-day set was compared to the variation among the sets. Only 19 of the above 21 variables were tested. Alcohol calories were included with calories from other carbohydrate, and total vitamin A activity was omitted. To Examine Variation in Meal Patterns Number of meals and snacks reported per day were totaled and the mode assigned to the meal pattern observed most often for each individual. Other calculations included amount and percent of calories consumed during meals and snacks. Methods to Develop Multiple Regression Equations to Predict Dietary Intake In order to form our food groups for the equations all items in our food table (HVH-CWRU Nutrient Data Base), approximately 1,600, were considered as possible predictors. Because the computer program for determining the re- gression equations considered a maximum of 79 independent variables at a time, it was necessary to either eliminate items or group them on the basis of nutrient composition. First, single food items were combined into 495 " specific item categories." Combinations were made on the basis of similar composition; for example, several forms of canned pineapple, i.e., slices, chunks, tidbits, and crushed, were all grouped as canned pineapple and three sizes of whole oranges were grouped as "fresh whole oranges." Then, a frequency count was made of these food categories consumed by the group of multiple sclerosis men and women to determine: (1) the number of subjects reporting the 495 categories and (2) the number of days each category was reported. Two hundred forty-one categories (about 700 single food items) were elimi- nated on the basis that less than 10 percent of the group reported the cate- gory at least once during the 36 days. Additional elimination of 94 groups was accomplished following the criteria of: ( 1 ) an item must have been consumed

Individual Variation in Intake of Nutrients 159 by about 30 percent of the subjects over the 36 days or (2) if fewer than 30 percent, then the item had to have been reported on more than 50 occasions over the 2,628 days observed. Five exceptions were later made to include the specific item categories of lamb chops, haddock, pike, salmon, and wheat thins to give a more complete representation to the lamb, fish-shellfish, and cracker groups. The final single food item count was 510 and these made up 160 specific item categories. A final count of the 510 single items showed that 92 percent of the total food items reported by the Group I subjects were included in the 510 items. A comparison with Group II showed these items accounted for 90 percent of reported items. To further facilitate calculations, another group- ing was made to condense the 160 categories into 69 food groups. All items in the food table are in common household measure, and this is how subjects were requested to report quantities of food consumed. For most items in the food table, it is possible for subjects to report quantitities in more than one type of measure, including (1) weight, (2) volume by measure from a teaspoon to a quart, or (3) by count (where appropriate) of the item itself, e.g., three pineapple rings, four brussel sprouts, two medium pork chops trimmed of fat, or two scoops of ice cream. Each of the 69 food groups was assigned a basic quantity unit and for each subject, amounts of each of the 510 items in the 69 groups were converted from the subject's reported quantity to a fraction of the basic quantity unit. For each subject, an average daily intake for each food group was calculated. METHODOLOGY FOR REGRESSION EQUATIONS The dependent variables included the average daily intakes of calories, animal protein, plant protein, animal fat, plant fat, refined carbohydrate, other carbohydrate, and alcohol. Input data from the 113 men (32 Group I men, 54 Group II men, and 28 Group III men) were used, so that any differences in intake between men and women would not bias the results. To arrive at the final regression equation, the following procedures were followed for each dependent variable: 1. Initial determination of the number of independent variables necessary in each regression formula (N): Through stepwise regression analysis, 15 variables out of the 71 were entered into the equation, F-level for inclusion, 0.01; F-level for deletion, 0.001. The variables were listed in the order as they entered the equation together with the multiple correlation coefficient (R) and the multiple it-square (RSQ); variable number 1 explaining the greatest amount of variance. Then starting with variable number 1, the number of variables was increased by one until: (a) the RSQ reached 70 or greater (ideally 75 or greater); (b) the increase in RSQ showed a reasonable

160 HAROLD B. MOUSER and HELEN T. BEBB break point; and, (c) at this time, for reasons of later computational restric- tions, the total number of variables did not exceed 8. 2. Selection of the best variables to be considered for each equation: Stepwise regression may or may not select the best combination of variables to predict the dependent variable, or there may be another set which would give nearly the same Rand RSQ, which would be just as acceptable. In order to test this, for each dependent variable, up to 15 independent variables were selected using these criteria: (a) when the number of independent variables for each equation (N) was determined, all variables from the first through the Nth were automatically included as 1 of the possible 15 total to be selected; (b) the simple correlation between each dependent variable with all independent variables was calculated. Additional variables were selected on the basis of a relatively high correlation with the dependent variable (ap- proximatelyR = >0.315~; and (c) reasonableness of obtaining intake data about the variable from the subject. 3. Determination of all possible combinations (up to N variables at a time) of the 15 independent variables, and calculation of the residual mean square (RMSQ). Variables were taken one at a time, two at a time, etc. until N number of variables at a time was reached. 4. Combinations of variables having the lowest RMSQ were compared with stepwise regression results and the best combinations of variables were selected for each dependent variable. Methods for Collecting Diet History At the end of the study year, the 73 subjects in Group I completed a dietary history through a structured interview with the study dietitian. The history was directed toward the frequency of consumption of specific food items on a usual basis such as never, once a week, or daily (Table 2~. For some items quantity was also obtained. For each frequency code, 95 percent confidence limits of a Poisson distribution were calculated based on 36 sample days and the actual frequency determined from the 36 days of diary recording were tested against the probability assigned to the frequency code. Fifty-one specific foods were analyzed in this manner and the observed frequency was determined to be less than, the same as, or more than the response indicated by history. Meal Patterns Food intake on the diary recording forms was recorded by the subject as taken at meals or between meals. We accepted the subject's recorded state-

Individual Variation in Intake of Nutrients TABLE 2 Subject Choices for Response to Questions About Usual Frequency of Consumption of Specific Foods Code 2 3 4 s 8 161 Frequency never once a month or less 2-3 times a month once a week 2 - times per week 5-7 times per week once a day 2-3 times a day 4-6 times a day >6 times a day ment even though recorded intake at a snack might exceed that at a preced- ing or following meal. The meal and snack periods were coded using seven code numbers: Code Number 2 3 4 s 6 Indicates Food Eaten: Before breakfast Breakfast Between breakfast and lunch snack Lunch Between lunch and dinner snack Dinner After dinner snack If food was consumed at several different times between meals, only one between meal snack was coded. RELIABILITY AND VALIDITY OF DATA Subjective assessment (by the study dietitian) of the recorded data and of the subjects' reliability, the compliance with schedules and instructions, and the reasonableness of the analyses of intake as previously reported (Witschi et al., 1970; B ebb et al., 1972) have led us to accept the diary records as representative of each individual's usual intake during the recording period. The relationship between the intake of the 73 subjects with multiple sclerosis and the fatty acid composition of their subcutaneous fat has also been examined and lends support to the validity of the data (Table 3~.

162 HAROLD B. MOUSER and HELEN T. BEBB TABLE 3 Rank Order Correlations (rS) of Total Diet-Iodine-Index (Range 50.5-86. l) with Serum Cholesterol and Adipose Tissue Fatty Acid Proportions Cholesterol Palmitic Oleic Linoleic Males (32) - 0.07 - 0.19 - 0.08 +o. 57a Females(41) +0.14 -0.12 -0.18 +0.33b All subjects (73) +0.07 - 0.21 - 0.11 +0.53c al p <0.05. b2 P <0.005- c 3 p <o.ooos Representative sampling of weekdays and weekend days is suggested by the data in Figure l, which show that the proportion of weekend days in each subject's daily records falls within a 95 percent confidence interval for the true proportion of weekend days. 50 40 o in ~ 20 cr UJ 30 p: True Proportion of l Weekend Days in a Year it -I ,71 ' ' ~ 10 o p + 1.96 Standard Error I (95% Confidence Limit) I Group I - W I Group I - M - Group I I - M \~A /1 I _ _ _ _/ ~ 1 1 1 1 1 1 ~ 1 i l 0 16.7 19.4 22.0 25.0 27.8 30.6 33.3 36.1 38.9 PE RCENT O F DAYS AS WE E KE N D DAYS FIGURE 1 The distribution of subjects by the percent of weekend days in the 36 days of diary recording.

Individual Variation in intake of Nutrients RESULTS 163 The number of multiple sclerosis and executive subjects showing significant variation by analysis of variance among their 12 sets of records for each of 18 nutrients and calories is shown in Table 4. While the relative rank of nutrients varied between two groups, Spearman's rank order correlation coefficient is significant; on the strength of this the two groups are combined to show the percent of individuals showing significant variation for each of the nutrients. Table 5 shows the distribution of subjects by the number of items per subject showing significant differences among the 12 sets. The median number of items for the groups is three. The significant differences in mean values for 20 nutrients and calories by the several comparisons are shown in Table 6. The direction of the differ- ences, that is a significant increase or decrease, is also shown in this table. The least number of differences occurred when the first recording days of each set were compared to the second and third days combined. Odd sets compared to even sets, May-October compared to November-April, the TABLE 4 Significant Variation Among 12 Sets of 3-Day Records for Individual Nutrients in 127 Adults Significant Differences Msa Executivea Total Nutrient No. No. No. % Vitamin A 8 7 15 11.8 Riboflavin 8 7 15 11.8 Niacin 9 5 14 11.0 Animal protein 10 4 14 11.0 Total protein 12 4 16 12.6 Animal fat 14 10 24 18.9 Carotene 14 2 16 12.6 Iron 15 8 23 18.1 Thiamin 16 7 23 18.1 Phosphorus 20 10 30 23.6 Total fat 20 8 28 22.0 Plant protein 21 12 33 26.0 Plant fat 23 12 35 27.6 Carbohydrate, refined 24 6 30 23.6 Ascorbic acid 24 11 35 27.6 Total carbohydrate 25 15 40 31.5 Calcium 25 12 37 29.1 Carbohydrate, natural 26 12 38 29.9 Calories 29 11 40 31.5 a Rank order: rs = 0.74 P <0.01.

164 HAROLD B. MOUSER and HELEN T. BEBB TABLE 5 Number of Subjects with Significant Variation in 19 Nutrients Among 12 Sets of 3-Day Food Diaries Subjects Number of MS Executive Total Items Varying per Subject No. Cum. % No. Cum. % No. Cum. % ~0 8 1 1 8 15 16 13 1 6 19 1 1 35 17 26 2 7 29 9 52 16 39 3 10 42 7 65 17 52 4 8 53 4 72 12 61 5 12 70 7 85 19 76 6 7 79 1 87 8 83 7 4 85 2 91 6 87 8 3 89 3 96 6 92 9 1 90 1 98 2 94 ~10 7 100 1 100 8 100 TOTALS 73 54 127 first 6 months of records compared to the last 6 months, * and weekdays compared to weekends, respectively, showed increasing numbers of differ- ences. The individual nutrients, with the exception of carotene and ascorbic acid, followed patterns similar to the overall patterns. The cumulative per- cent of subjects by the number of items showing significant differences is shown in Figure 2. Sixty-four percent of the subjects had no nutrients with significant differences between the first and second and third days of records and 80 percent had one or less difference; all subjects, except one, had four or less significant differences. The other curves show increasingly higher percentages of subjects with larger numbers of differences in the same direction as shown in Table 6. A comparison between the analysis of variance results and the differences in means testing is shown in Table 7. By the goodness of fit chi square test, the patterns in each of the three groups for day one versus days two and three and for weekday versus weekend days show no significant differences. The food items used in the final multiple regression equations are shown in Table 8. Fifteen of the 25 variables entered into more than one equation; COOKIE appears in five of the equations for the eight dependent variables. The number of variables could have been one less for several equations by our original criteria but the necessary presence of the variable in another equation and a slight increase in the correlation coefficient, R. by its inclu- *Most subjects completed their first set of records in late spring or early summer. Therefore there is considerable overlap of the seasonal and early and late records.

Individual Variation in Intake of Nutrients loo 80 '_ 70 Cal m in o 60 so Z 40 30 20 10 o / / .' _ ~ / / / /: 7, .— ... , . 1 1 1 1 1 .. - First Day vs Days 2 and 3 Odd vs Even Playact. vs Nov.-April Weekdays vs Weekend Days 0 1 2 3 4 5 6 7 8 9 10or more NUMBER OF ITEMS PER SUBJECT SHOWING SIGNIFICANT DIFFERENCES FIGURE 2 Cumulative percent of subjects by number of individual nutrients with significant difference in means in 12 sets of 3-day food diaries. 165 sion led us to include it in the equation. With the exception of one variable, animal and plant protein showed different sets of independent variables. This was also true for animal and plant fat and refined and other carbohy- drate. For alcohol, food groups other than the three types of alcoholic beverages explained extremely small amounts of the variance so the total number of variables was limited to two. For each of the 25 food variables in the equation, Table 9 shows the mean units of food consumed over 36 days. Some differences are apparent. The greatest consumption of beer and whiskey was shown by the executive group. There was a greater use of margarine and less butter by the executive men and the physical fitness group (a possible time trend, as the food consumption data for the general U.S. population showed a similar shift during the 1960's). Sandwiches were frequently reported as lunch items for the multiple sclerosis group and this explains the slightly higher group means for bread and the hot dog-lunchmeat category. The multiple correlation coefficients (R) and the percent of variance accounted for by the equation (R2) are presented in column one of Table 10. Multiple R 's ranged from 0.89 to 0.99; the lowest R2 was 0.79 (plant fat and

166 U. ~ _ ;` ~ · r _, Ct ;^ ~ _ Ct : I U. ~ 1 ~ ~ ~ ., 1 ~ \e, · ~ Ci~ ~ o ~ I ~ U) ° 'D m E~ C~ C~ ~ C~ ~ ~ _ ~ ~ O ~ ~ ~ — - ] ~ ~ ~ _ _ '] 00 ~ ~ ~ _ _ _ 00 _ ~ ~ _ _ _ _ O ~ 00 ~ =\ _ _ _ _ ~ ~ 1 _ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ 00 ~ ~ 00 E~ 1 c`, oo ~ _ ~ o 1 — — _ _ _ 1 — — — — ~ ~ u~ ~ _ _ _ _ _ _ _ _ _ _ ~ 00 00 ~ ~ ~ ~ 1 ~ 00 ~ ~ O O C~ O _ _ _ C~ c~~ _ _ ~ 1 o O _ t~ 1 — _ _ ~ O ~ - _ 00 _ ~ _ ~ ~ ~ O ~ 00 ~ I [~ ~ ~) ~ 00 ~ ~ _ O ~ l ~ ~ ~ ~ ~ ~ O _ ~ ') 1 1 — — — — — — — ~ ~ ~ 1 — ·~- O o ~—~ '< ~.,

167 _ ~ ~ ~o o C~ O ~ — ~ r~ 0 c~ ~ ~ — ~ ~o 1 C~ ~ _ ~ ~ oo ~ ~ ~ ~ ~ - , _ oo _ o ~ (~~ _ ~ ~ c~ ~ ~ ~ ~ \0 ~ ~ 1 ~ _ ~ oo _ ~ ~ ~o ~ C~ U~ ~ U~ ~ o ~ ~ U) ___ _ ___ ~oooo ~1 _ ~ 1 ~1 ~— <) t— 00 0 oO ~— ~ oo ~ ~— 1 _ ~ ~ _ _ ~ _ oo C~ ~ o — ~ ON ~~ ~ c<~ c~ [~ —) ~ — ~ ~t cr. V) ~ ~ ~ o _ c~ ~ oo ~ ~ cr~ oo ~ ~ ~ ~ c~ ~ _ ~ ~ oo X ~ ~ oo U~ o _ ~t (~~ oo ___ ___ oo',1_ _ oo ,~\ _ ~ o ~ ~ ~ ~ oo o _ _ ~ ~ U~ cr~ — (', ~ o ~ ~o oo 0 u~ ~ ~ u~ ~ ~ 1 C~ C-~ C~ _ _ _ ~ _ _ — o ~ ~ 1 _ _ — ~ 1 ~ o o ~ oo ~ oo ~ ~ ~ o o C~ V) ~ _ _ ~ ~ ~ ~ _ oo ~ ~ o ~ ~ ~ ~ _ ~ ~ ~ _ o oo oo r~ ~ oo In — 0 ~ oo \0 ~ — ~ ~ _ _ _ _ ~ _ _ ON ~ ~ oo l , 99 e _ v ,= .S ~ S o ~ ~ ~ ~ ~ O O O ~ G u ~

168 ~4 ._ o ._ .c .s o U. 4— U. Al ._ o Cal Cal ._ 3 C) Ct ._ Ct :, o U' ._ U. - Cal ~ a: 44-O1 ce Cal ~ ,= - U. ·_ o ~ o ~ ·= fi Cal I o o Cal ~ E- 1 ~ C,0 Em via o ~ ~ _ _ — U. En C) ~ , _ PI ~ .e Cal ·C~ ~ ;~ ~ ; - ~ ~ s~ ~ ~ ~ o <: ~ . s ^~ zo.3 o ') _ — C~ ~ ~ , _ ~ ~ .= U. U. ~ ,,E, _ Ce V~ o C~ o o ._ m ~o Z 3 ~ _ ~ 0 c) ~ fi.= - C) ·m g :~_ Ct .= ~ o e ~o o C~ o s ~ m -0 Z 3 _ ~ 0N ~D ~ O oo . . . . . . 00 ~ 0 0 v~ ~ 00 0\ ~ u~ ~ ~ u~ 0N ~ u~ ___ ___ ___ ~ ~ 0 ~ ~ c~ . . . . . . u~ ~ ~ 00 _ - , ~ c~ _ 0 0\ 00 00 ~ ~ ~ 0 . . . . . . . . . ~o ~ ~ ~ u~ ~ 00 d- ~ ~ - , ~ _ ~ 0 0 ~ o~ c~ ~ _ 0 . . . . . . . . . — ', a~ 0 U~ ~ _ x ~ c~ O ~ v~ ___ ___ ___ ___ ___ ~ ~ ~ ~ 00 0 u~ O _ ~ ~ ~ cs~ 0 0 . . . . . . . . . . . . . . . ~ ~ ~ ~ ~ ~ ~ 0 ~ - , ~ _ ~ ,3\ _ - , _ ~ — ~ ~ — ~ ~ — d ,` 00 ~o 0 ~ ~ ~ O _ ~ ~ _ ~ _ ~ ~ _ ~ _ ~ <5 q: O~ ~ cq ~ ~ cq ~ `, O 00 ~o 1 1 ~ =° 1 1 ~ ~ 1 1 ~ ~ ~___ ___ _~_~ ~_~'_ ___ ~ 0 0 0 ~ 0 0 0 ~ 0 0 0 ~ 0 0 0 ~ 0 0 0 c ~ ~ ~ .L ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~VVV ~VVV ooov ~ VVV ~VVV

o o o o o o' I - 1 I o o - 'C O ! 1 1 ~ 1 I o I ~ I o ~ ~ -I t ~ 1 ._ ~ al ~ I < Cl ~ o o o ~ o o I o .E E a s 31 < ~ I ~ . o '@1 ~ ! ~1 .; e I o o o ~ ~ o o o < = I ~ o 10 g ~ ~ O < ~ O O 8 ~ ~ o C ~ J ~ ~ o ^ = ~ ~ o o o o o - 'E > o o T o ~ ~ o ~a ~ = ~ O o o

170 HAROLD B. MOUSER and HELEN T. BEBB TABLE 9 Group Mean of Total Intake Over 36 Days for 25 Independent Variables Group Mean for Total Number of Units Physical MS Executive Fitness All Independent Variable Unit Men Men Men Men Bread and Cereal Bread 1 Slice 98.2 86.4 64.7 84.3 Buns hot dog-hamburger 1 Bun 5.1 11.9 13.3 10.4 Cake 1 Piece 4.6 3.3 2.9 3.6 Cereal—dry 1 Cup 6.5 8.6 11.2 8.7 Cookie 1 Cookie 25.7 18.8 21.6 21.4 Doughnut 1 Roll/Doughnut 9.7 13.1 12.2 12.0 Pie 1 Piece 4.5 4.5 2.8 4.1 Fruits and Vegetables Potatoes 1 Cup 16.1 15.2 12.6 14.8 Vegetables cooked 1 Cup 11.2 12.6 9.9 11.5 Meat Beef 1 Ounce 19.9 38.5 24.2 29.8 Chicken 1 Ounce 22.8 28.9 31.3 27.8 Eggs 1 Egg 25.1 24.9 19.2 23.5 Fish—shellfish 1 Ounce 19.9 18.2 24.6 20.1 Hotdog—lunchmeat 1 Ounce 45.6 35.5 35.1 38.2 Pork ham 1 Ounce 24.2 24.9 15.1 22.3 Dairy Butter 1 Teaspoon 74.5 52.3 34.6 54.0 Margarine 1 Teaspoon 17.4 33.6 35.1 29.4 Milk whole 1 Cup 25.4 16.0 10.2 17.1 Other Beer 1 12-Ounce Bottle 11.8 49.7 8.9 29.2 Coke 1 Cup 14.3 12.9 9.8 12.5 French dressing—oil 1 Tablespoon 10.6 19.1 30.2 19.6 Maple syrup 1 Tablespoon 4.3 7.6 5.2 6.1 Soup 1 Cup 8.2 9.8 9.6 9.3 Sugar 1 Teaspoon 94.0 77.2 25.6 69.0 Whiskey 1 Ounce 12.5 53.9 23.6 35.1

Individual Variation in Intake of Nutrients 171 TABLE 10 Correlation Coefficient (R) and R2a for Predicted Intake Versus Actual Intake for All 113 Men Plus Four Individual Groups Using Formula Developed for 113 Men to Predict Variable Values Variable Calories Protein Animal Plant Fat Animal Plant Carbohydrate All Men (113) R R2 0.90 0.81 MS Men (31) R R2 Executive Men (54) R 0.71 0.50 0.77 Physical MS Fitness Men Women (28) (40) R2 R R2 R__________ ___ 0.59 0.68 0.46 0.78 0.61 0.90 0.81 0.82 0.67 0.80 0.64 0.43 0.18 0.69 0.48 0.91 0.83 0.81 0.66 0.81 0.66 0.62 0.38 0.81 0.66 0.93 0.86 0.90 0.81 0.89 0.79 0.67 0.45 0.63 0.40 0.89 0.79 0.79 0.62 0.82 0.67 0.81 0.66 0.88 0.77 Refined sucrose 0.94 0.88 0.84 0.71 0.87 0.76 0.77 0.59 0.68 0.46 Other 0.89 0.~79 0.64 0.41 0.80 0.64 0.65 0.42 0.79 0.62 Alcohol 0.99 0.98 0.97 0.94 0.98 0.96 0.86 0.74 0.90 0.81 aR 2 iS the percent of the total variance explained by the regression equation. other carbohydrate). Multiple R 's and R2's are lower when the 113 men are divided into their respective study groups and the correlation coefficients were calculated for actual versus predicted intake in each separate group. As a test, the data from the 40 women were used with the regression equations for the 113 men. Predicted intake was correlated with actual intake; multiple R's range from 0.63 to 0.90 for the eight dependent vari- ables. Except for animal fat, all are higher than those for the physical fitness group, but for calories, plant protein, plant fat, and other carbohydrate, they compare well with the other two groups of men. The regression equations for each of the eight variables are shown in Table 1 1. In Table 12, the actual group means and the group mean of predicted values are shown with their standard deviations for the four study groups. With few exceptions, the group means from predicted values are either approximately equal to or less than means from actual values and standard deviations are generally of the same magnitude. The modal patterns of meals and snacks and the number of subjects with each pattern are shown in Table 13. The most frequent modal pattern, three meals with three snacks, was present in 63 (50 percent) of our subjects. The distribution of all subjects by the percent of days each subject consumed food by his modal pattern is shown in Table 14. Only 40 percent of subjects had 50 percent or more of their days as modal pattern days and only 15 percent had three-fourths or more days at their modal pattern.

172 HAROLD B. MOUSER and HELEN T. BEBB TABLE 11 Regression Equations Based on Dietary Intake for 113 Men Dietary Factor Regression Equationsa __ , Calories 3.8 (BREAD) + 6.3 (BEER) + 16.2 (DOUNUT) + 25.5 (POTATO) + 7.7 (COOKIE) + 3.7 (WHISKEY) + 7.9 (CHICKN) + 14.1 (VEGCKD) + 513.4 Animal protein 0.488 (BEEF) + 0.145 (HDLCHM) + 0.328 (CHICKN) + 0.228 (MKWB) + 0.291 (FISHSH) + 0.211 (EGGS) + 0.468 (BUNHDH) + 17.717 Plant protein 0.062 (BREAD) + 0.195 (SOUP) + 0.063 (BEER) + 0.426 (CAKE) + 0.245 (VEGCKD) + 0.230 (BUNHDH) + 0.060 (COOKIE) + 5.391 Animal fat 0.191 (BUTTER) + 0.348 (PORKHM) + 0.289 (MKWB) + 0.167 (HDLCHM) + 0.341 (BEEF) + 0.536 (BUNHDH) + 16.465 Plant fat 0.188 (DOUNUT) + 0.680 (PIE) + 0.303 (FRDROL) + 0.547 (BUNHDH) + 0.153 (COOKIE) + 0.102 (MARG) + 0.289 (POTATO) + 0.696 (CAKE) + 5. 598 Refined carbohydrate 0.759 (COKE) + 0.382 (COOKIE) + 0.1~ (SUGAR) + 1.441 (PIE) + 1.875 (CAKE) + 11.484 Other carbohydrate 0.39 (BREAD) + 1.51 (VEGCKD) + 1.25 (MPESYP) + 0.51 (BEER) + 0.61 (COOKIE) + 1.32 (CERDRY) + 1.50 (POTATO) + 0.73 (soup) + 47.18 Alcohol 0.326 (WHISKEY) + 0.356 (BEER) + 0.413 a Variables are listed in the order of entry into equations. The relationship between number of days at modal pattern and (1) number of nutrients showing significant variation among the 12 sets (Figure 3) and (2) number of nutrients showing significant differences betwen odd and even sets of records (Figure 4) suggests independence of modal pattern and the variation in nutrients over time. Similar patterns were observed for the other comparisons of sets of records. The results of the history interview are shown in Table 15 with the food items listed in order from the least to the highest number of differences between the recorded frequency and the history reports of frequency. Those food items also tested in the multiple regression analysis are indicated on this table. For almost all foods, subjects reported more frequent use than their records showed. The degree of consistency of history for individual items and their eventual use in the regression equations does not appear to be related. The percent distribution of subjects by the number of significant differ- ences between history and recorded frequency of consumption of foods is shown in Figure 5. The bimodal distribution suggests two populations of historians, one more valid than the other. A comparison of the disagreements between history and recording both to the analysis of variance data and to the distributions of percent of days with modal patterns of eating is shown in Table 16. None of the correlation coefficients is significant.

173 _4 .= o C) ._ au Ct 3 ~: o C~ o s~ C~ ~: ._ Ct ~ 1 ~ ~ 1 ° ~ 1~ 1 I ~ Ct 1 c~ 1 C~ m cn C: ==; ·C~ C) O O . . 00 ~ ~ O tl +i o o . . ~ o C~ ~ tl +1 =1 V' ~ ~ 1 ~ ~ ~ o 3 ~ ~ ~ ~ ~ ~ ~ ~ ~ oo _ ~ ~ __^ o o ~ ° U~ ~, _ _ ~ ~r _ _ tl tl tl tl tl tl C~ ~i _ _ ~ ~ tl tl tl tl o ~ . . _ _ _ _ tl tl C~ ~ — ~ . . . . _ _ _ _ tl tl tl tl ~ U~ ~ oo oo U~ o ~ — ~ C~ ~ ~ ~ ~ o oo r~ ~ ~ ~ _ u~ _ — j v, ~ ~ - - tl tl tl tl i j Ct 1 1 ~i ~ C~ "C' C~ _ ~ ~ oo tl tl tl tl tl tl ~ ~o oo _ ~ ~ . . . . . . ~ ~ ~ o ~ ~ _ _ _ oo vo ~ V) V) . _ _ - , - , ~ _ tl tl tl tl tl tl o oo ~ ~ ~ ~ o _ O~ o _ ~ o o ') ~ C~ — ~ — ~ ~ d tl tl tl tl +I tl +I tl +1 ~ ~ ~ 1 I. . . U) ~ ~ 1 tl tl tl I ~ ~— 00 0 t— _ ~ — ~ ~ _ ~ r—] t_ 1 . . . . . . . . . . . . . . o ~ ~ ~ U~ oo oo o _ o ~o o ~ ~ ~ ~ ~ ~ ~ C~ U~ ~ Ch oo U~ ~ _ _ o o . . o V~ tl tl 1 oo ~ U~ C~ o o 1 ~ — ~o ~ 1 ~ — _ _ ._ _ Ct p., C ~ ~ C o ~ - C) . ~ Ct ~ ~ C) ~ C ~ - oo ~n ~ c~ · <~\ _ ~ ~ · · — — d. u~ tl tl tl +l tl tl . . . . _ _ d- ~ C~ ~D \0 C) C) . _ _ C.> ~ — Ct C) C) . ~ ~ . _ ~ Ce ~ ~ ~ ~ · ~ ~ . . . . o o crX 't - - ~ c~ tl tl tl tl ~ V) ~ ~ . . . . Ch ~ ~ ~ ~_ ~ . C - ° ~V C) ._ CD ~ C~ . . ~_

174 HAROLD B. MOUSER and HELEN T. BEBB TABLE 13 Modal Pattern for Number of Meals and/or Snacks per Day Number of Subjects Modal Pattern Group I—F Group I M Group II M (41 ) (32) (54) 3 Mealsa 3Meals+Snack3+5+7 17 3 Meals + Snack 7 3 Meals + Snack 5 + 7 3 Meals + Snack 3 + 5 3 Meals + Snack 3 + 7 Other modal patterns 2 4 10 o 1 4 12 34 4 4 3 1 o 8 3 3 4 4 2 aBreakfast, lunch and dinner Snack 3—between breakfast and lunch Snack 5 between lunch and dinner Snack 7 after dinner bOther modal patterns include any 2 or 3 meals in combination with various snacks. TABLE 14 The Distribution of Subjects by the Percent of 36 Days That the Modal Meal Pattern Occurred Percent of Days at Mode 100 75-99 50-74 25-49 <25 TOTAL Subjects No. 2 17 32 67 Percent 1.6 13.4 25.2 52.8 9 7.1 127 100.0

Individual Variation in Intake of Nutrients 20 18 ~ 16 UJ ~ Z LL U.] 3 ~ In ~ — 1— z ILL ~ z 0 ~ G <~, m to Z ~— a o I 4 14 12 10 8 2 . . _ ·~e ~ . · ·. · · . . . · · - · - ~ · ·. ~ · ·... · ~ ·. . . · · · ~ · · ~— . 6 12 18 24 30 36 NUMBER OF DAYS AT MODAL PATTERN FIGURE 3 Comparison of frequency of significant differences among 12 three-day records for 19 nutrients and modal pattern. 20 18 C.' ~ Z am 11 — 1— ~ Z LL C-) ~ — — LL ~ — O ~ G <~, m Z ~ — at o I en 16 14 12 10 8 6 4 2 ·— . · · _ · ~ ~— ·- ~ · · - - .~. · - ~—- ·~. · · a----- ---~e .~. .~-~--. O 1 1 1 1 1 1 1 1 1 1 1 1 6 12 18 24 30 36 NUMBE R OF DAYS AT MODAL PATTERN FIGURE 4 Comparison of frequency of significant differences between odd sets and even sets of records and frequency of days at modal pattern. 175

176 HAROLD B. MOUSER and HELEN T. BEBB TABLE 15 The Occurrence of Significant Differences Between the Frequency of Consumption of Food Determined by Interview and the Re- corded Frequency from 36 Days of Food Diaries in 73 Adults and the Use of the Items in Multiple Regression Analysis for Prediction of Nutrient Intake Observed Food Less More Total Tested for Prediction Yes - Used Not Used Sardines, anchovies 0 0 0 x Pizza 0 0 0 x Waffles, pancakes 0 0 0 x Rice 1 0 1 x Dry fruit 1 1 2 x Vegetable salad 2 0 2 x Cottage cheese 0 3 3 x Sherbert 1 2 3 x Bacon 1 3 4 x Pudding, custard 1 3 4 x Wine 4 0 4 x Syrup 5 0 5 x Beer 5 1 6 x Ice cream 6 1 7 x Doughnuts 7 1 8 x Pickles, olives 2 6 8 x Whiskey, gin 5 3 8 x Spaghetti, macaroni 6 3 9 Citrus juice 7 2 9 Cream soup 6 3 9 x Popcorn, Fritos 5 4 9 x Hotdog, hamburger buns 9 0 9 x Eggs 7 3 10 x Fish, shrimp 10 0 10 x Cooked vegetables (green, yellow) 4 6 10 x Cereal 7 4 11 x Steak, beef 3 8 11 x Chicken 1 0 1 11 x Potatoes 8 3 11 x Canned fruit 9 2 11 x Soup, clear 11 0 11 x Kool Ade, fruit ade 11 0 11 x Pie 1 1 1 12 x Chocolate candy 6 7 13 Jelly, honey 11 3 14 x Citrus fruit 12 2 14 x Other raw fruit 10 4 14 x Soft drinks 10 5 15 x Sweet rolls 10 5 15 Crackers 15 1 16 x Cheese 9 8 17 x No x

individual Variation in Intake of Nutrients TABLE 15 Continued 177 Tested for Protection Observed Yes No Food Less More Total Used Not Used _ Cream 15 2 17 x Gravy 1 1 7 18 x Noncitrus juices 8 10 18 x > 1/2 Pound meat 15 4 19 x Lunchmeat 3 17 20 x Cake, brownies 20 0 20 x Cookies 10 10 20 x x Nonchocolate candy 11 9 20 x Other cooked vegetables 20 1 21 x Tea 22 1 23 x 16 14 12 m cn LL o at UJ cot in: cat 10 8 6 4 o A \/ . / 1. ~ .1 1 1 1 1 1 1 1 1 2 4 6 8 10 A 12 14 16 18 NUMBER OF FOODS FIGURE 5 The percent distribution of 70 subjects by the number of significant differences between stated and recorded usual frequency of consumption of specific foods. TABLE 16 Correlation Coefficients Between History and Recording Disagreements and (1) Frequency of Nutrients with Significant Variation Among 12 Sets of Records and (2) Percent of Days at Modal Pattern of Meals MS Men (31) MS Women (39) Significant Differences Percent of Significant differences Percent of 12 sets History Modal Days 12 Sets History Modal Days Mean 4.77 7.87 49.3 4.15 7.67 55.8 SD 3.34 3.51 22.2 3.88 3.48 20.2 r 0.15 - 0.23 - 0.15 0.05

178 DISCUSSION HAROLD B. MOUSER and HELEN T. BEBB The present studies have been directed to evaluation of methods for deter- mining the usual nutrient intake of an individual rather than the mean intake for groups of individuals. The relatively close prediction of mean intake for groups of subjects by the multiple regression equations and the relatively poor prediction by these same equations for individuals in each group, highlight the problem of determining with sufficient precision the indi- vidual intake in contrast to a group of individuals. Of course, precision is relative to the purpose for which data are collected. The analyses here have concentrated on significant difference within individuals over time and con- sidered the relative frequency of these differences by individual nutrients or by number of differences per individual. When one considers the total pos- sible differences, that is 21 x 127 or 2,667, for the difference in means one is struck by the relatively low percent, ranging from 3.7 to 13.4, of signifi- cant differences for the groups as a whole. However, the proportion of individuals showing mean differences is much higher when 3-day sets of records are compared by season, by alternate months, and particularly by weekdays versus weekend days. Twelve to 32 percent of subjects show significant variation for individual nutrients among the 12 sets of records at approximately monthly intervals. It is apparent from our studies that some nutrients vary more than others and that some individuals vary more than others. Thus, the frequency of obtaining diet information is determined both by the nutrient or nutrients of interest and the proportion of individuals about whom reasonably accurate information is desired. Our studies suggest, as have others, that a represen- tative food intake must include weekend days and weekdays, probably in their true proportion of all days. The data also suggest that a single day's information will yield data equivalent to 3 consecutive days. The data also suggest that infrequent sampling during the year, such as every other month, will provide representative information. They also suggest that sampling over a long term, such as a year, will provide more representative data than sampling over a shorter term. The use of a limited number of food items to predict nutrient intake does not appear to be a good estimator of total intake from our data. The selection of the appropriate items is one problem. It is extremely unlikely that a priori selection of items by us would have included those indicated by our multiple regression equations. Translation of our items to another group is unlikely to provide good individual estimates in that group. If information about individual foods is obtained by history, our data suggest that estimation of usual intake would be poor. Our subjects tended to overestimate their consumption of the foods used in our regression equa- tions. Our history data also suggest that some foods are better remembered

Individual Variation in Intake of Nutrients 179 than others and that this is not related to the frequency of consumption of the foods. As with the variability of 3-day intakes, some foods are good histori- cal items and some subjects are good historians. How to identify each is a problem. The use of modal patterns of eating as an aid in obtaining diet histories, a technique that has proved useful in determining alcohol intake, does not appear from our data to be helpful. The proportion of days in which the modal pattern is followed is relatively low overall and those subjects with a high proportion of modal days do no better in history nor do they show less variation among their sets of records. In summary, our data suggest that the diary recording of a single day's intake, randomly selected to represent weekdays and weekend days, with the recording done over a long enough interval to detect cyclic changes during the year resulted in a valid estimate of the usual intake of the indi- viduals in our study. The subjects were not a probability sample of any defined population and any translation of our findings to any population would be conjectural. Further, within our study population, historical in- formation about frequency of consumption of specific foods, knowledge of modal patterns of eating, or the use of predictive formulas for estimating nutrient intake were found wanting in their use for determination of indi- vidual dietary intake. REFERENCES Bebb, H. T., H. B. Houser, J. C. Witschi, A. S. Littell, and R. K. Fuller. 1971. Calorie and nutrient contribution of alcoholic beverages to the usual diets of 155 adults. Am. J. Clin. Nutr. 24:1042. Bebb, H. T., H. B. Houser, J. C. Witschi, and A. S. Littell. 1972. Nutritive content of the usual diets of eighty-two men. J. Am. Dietet. Assoc. 61:407. Houser, H. B., W. Insult, Jr., J. C. Vandervort, A. S. Littell, and A. I. Sorensen. 1966. The relationships of habitual diets to the fatty acid composition of adipose tissue and serum cholesterol concentration in man. Circulation 34-4: Suppl. 111 - 130. (Abstract) Houser, H. B., A. I. Sorensen, A. S. Littell, and J. D. Vandervort. 1969. Dietary intake of non-hospitalized persons with multiple sclerosis. 1. Food diary and coding methods. J. Am. Dietet. Assoc. 54:391. HVH-CWRU Nutrient Data Base. Department of Biometry, School of Medicine, Case Western Reserve University, Cleveland, Ohio. Witschi, J. C., H. B. Houser, and A. S. Littell. 1970a. Performed vitamin A, carotene, and total vitamin A activity in usual adult diets. J. Am. Dietet. Assoc. 57:13. Witschi, J. C., A. S. Littell, H. B. Houser, and A. I. Sorensen. 1970b. Dietary intake of non-hospitalized persons with multiple sclerosis. 2. Nutrient intake for one year compared with recommended allowances. J. Am. Dietet. Assoc. 56:203.

Diary Interview Technique to Assess Food Consumption Patterns of Individual Military Personnel D. D. SCHNAKENBERG, T. M. HILL, M. J. KRETSCH, and B. S. MORRIS A diary-interview technique has been developed at Letterman Army Insti- tute of Research (LAIR) for collecting and evaluating food consumption data from individual military personnel. Under the direction of Col. John E. Canham, M.D., the technique was initially developed by Dr. Terrell M. Hill and first used at Ent Air Force Base, Colo., to measure nutrient consumption of airmen away from the dining hall (Table 1~. More recently, we have used the technique to evaluate and predict the nutritional impact of new concepts of food service that are being tested at a number of Department of Defense installations. Our diary-interview technique combines selected aspects of conventional dietary recall, food diary, and dietary interview metholologies so as to TABLE 1 Recent Military Nutrition Studies That Have Utilized a Diary-Interview Technique to Assess the Nutrient Consumption of Individuals Location Date No. of Subjects Duration Air Force Ent AFB, Colo. Oct. 1973 149 14 days Navy Alameda, Calif. Mar. 1975 133 17 days Jun. 1976 154 14 days Aug. 1976 158 17 days Marines—Twentynine Palms, Calif. Mar. 1977 315a 14 days Oct. 1978 Ongoing Navy USS Saratoga Jul. 1977 203 17days Nov. 1978 Ongoing a Includes 36 female marines. 180

Diary-Interview Technique 181 hopefully optimize those unique opportunities and requirements for nutrition studies available in a military setting. Subjects who have been selected to participate in the study are briefed as to the purpose of the study, the measures used to maintain the confidentiality of individual data, what will be expected of them as participants, and where and when to report for the series of 20-minute meetings with one of our interviewers. At the first meeting, the subject is asked to recall all foods (except spices) and bever- ages (except water) consumed during the previous day. The subject is in- structed to itemize the information on a pocket-sized card (Figure 1) and to also record the time the item was consumed, the quantity consumed, and the source of each item (e.g., military dining hall, home, restaurant, or vendor). The participant is provided this card as a guide to record food consumption on identical diary cards for the next 3 to 4 days. At the subsequent twice- DA I LY FOOD CONSUMPTION NAME SSAN TIME FOOD ITEM DATE SUBJ. NO. AMOUI`IT SC LAIR FORM 43 (4 Feb 75 - Bioenergetics) FIGURE 1 Daily food consumption.

182 D. D. SCHNAKENBERG et. al. weekly interviews, the diary cards are returned to the interviewer for re- view, assistance in estimating portion size, clarification of any unusual food items consumed, and assignment of each food item as a component of either a meal or a between-meal snack. In most of our studies, we have hired female dietitians and public health nutritionists, on a temporary basis, to serve as interviewers and to code and verify the data for subsequent computer processing. We did, however, train and use male military and civil service biological laboratory technicians as interviewers for the study conducted aboard the aircraft carrier uss Saratoga, while on station in the Mediterranean Sea. The March 1975 study at Naval Air Station, Alameda, Calif., provided an opportunity to assess the validity of our diary-interview technique to esti- mate nutrient intake per dining hall meal (Table 2~. As a separate aspect of the survey, we stationed additional dietitians at the end of each serving line in the dining hall who recorded, by observation, the food items selected by each and every patron of the dining hall. Portion size was determined by weighing the total quantity served and dividing it by the number of portions taken. The food waste on each tray was weighed and subtracted from serving size to obtain the amount consumed. The last four digits of the social security numbers were used to track the trays of all individual patrons and to identify the diary-interview participants. Upon examination of both the diary-interview and observer data bases, we found 423 man-days where the number of reported and observed meals agreed and TABLE 2 Nutrition Survey, Naval Air Station, Alameda, Calif., March 1975 Technique Daily Diary- Dietary Interview Observer Purpose Total daily intake Intake per dining hall meal Length of study 17 days 17 days Population studied 133 All Weight of food item consumed As reported Average portion size minus individual plate waste Composition of foods consumed Nutrient factorfile Nutrient factor file Man-days test population utilized dining hall 534 458 Man-days number of reported end observed meals agreed 423 423 Number of subjects 62 62

Diary-Interview Technique 183 these man-days of data were derived from 62 of the 133 subjects in the test population. The same nutrient factor file of food composition tables was used in the computation of both data bases. One variable in the comparison of the two techniques was the relative agreement in food item selection (Table 31. Food items selected agreed between the two techniques 80 percent of the time. Fourteen percent of the total food items in question were observed but not reported to have been consumed. Only 6 percent of items were reported but not observed. The relative incidence of disagreements by food type are also shown in Table 3. Vegetables, fruits, nuts, and legumes were included in the other food type category. The average nutrient intake per dining hall meal was computed for each of the 62 subjects utilizing the data derived from the two techniques and the comparisons are shown on Table 4. With the exception of protein, fat, and niacin, which did not differ, the diary-interview method significantly un- derestimated nutrient intake per meal compared to the observer technique. However, the magnitude of the underestimation was usually 10 percent or less. The lower values for the interview technique reflect, in part, the previ- ously mentioned greater tendency for an individual to have been observed to consume a food but not report it on his diary card. The larger discrepancy in ascorbic acid intake (41.2 mg vs. 58.0 mg/meal) is partially artifactitious. If an TABLE 3 Comparison of Diary-Interview and Observer Techniques to Assess Food Item Consumption per Dining Hall Meal, Naval Air Station, Alameda, Calif., March 1975 Number of Food Items per Meal . Mean SD Range Foods in questions 6.51b 1.66 2-12 Foods in agreement 5.20 1.73 0-12 Foods reported but not observed 0.42 0.73 0-5 Foods observed but not reported 0.89 0.95 0-4 Disagreements in entrees 0.20 0.47 0-3 Disagreements in salads 0.11 0.29 0-2 Disagreements in desserts 0.15 0.40 0-2 Disagreements in beverages 0.20 0.41 0-2 Disagreements in other food types 0.65 0.81 0-4 Ratio of agreements to total foods in question Ratio of reported but not observed to total foods in question Ratio of observed but not reported to total foods in question aCondiments excluded. en = 423 man-days. 0.80 0.06 0.14

184 D. D. SCHNAKENBERG et. al. TABLE 4 Comparison of Diary-Interview and Observer Techniques to Estimate Nutrient Intake per Dining Hall Meal, Naval Air Station, Alameda, Calif., March 1975 Techniquea Nutrient Diary-Interview Observer P Valueb Quantity, g 917 + 253c 1,042 + 288C <0.001 Energy, g 1,237 + 354 1,373 + 362 <0.001 Protein, g 55 + 16 57 + 15 NS Fat,g 62+19 65+18 NS Carbohydrate, g 117 + 42 144 + 46 <0.001 Calcium, mg 673 + 311 751 + 315 0.006 Phosphorus, mg 930 + 294 986 + 296 0.064 Iron, mg 7.9 + 2.1 8.8 + 2.5 <0.001 Vitamin A, flu 2,933 + 2,122 3,669 + 2,385 <0.001 Thiamin, mg 0.75 + 0.22 0.82 + 0.22 0.022 Riboflavin, mg 1.39 + 0.47 1.52 + 0.50 0.010 Niacin, mg 8.67 + 2.9 9.00 + 3.4 NS Ascorbic acid, mg 41.2 + 33 58.0 + 47 <0.001 a Values computed from the average nutrient intake per dining hall meal of each subject (n = 62) who reported and was observed to have utilized the dining hall one or more times during the survey. bPaired l-test. C Mean + SD. individual selected an orange but did not eat it until after leaving the dining hall, the observer technique considered it to be consumed with the meal, whereas the diary-interview technique considered it to be part of an after-meal snack. To further examine the data, we computed a comparison ratio for each subject where: Nutrient intake/meal from diary-interview technique Compar~so a 0 Nutrient intake/meal from observer technique As shown in Table 5, the mean comparison ratio for energy was 0.92, indicating that compared to the observer technique the diary-interview technique underestimated energy intake of the population by an average of only 8 percent. This compares favorably to the 16 percent underestimation of actual (unobtrusively weighed) energy intake per meal obtained from elderly subjects by a 7-day dietary record technique as reported by Gersovitz et al. (1978~. However, it should also be noted that our diary-interview underestimated the energy intake of 72.6 percent of the individuals at least to some degree. The histogram of comparison ratios for energy intake is shown in Figure 2. The distribution is skewed to the left of 1.0 with 34

Diary-lnterview Technique TABLE 5 Evaluation of Daily Diary-Interview Technique, Naval Air Station, Alameda, Calif., March 1975 185 Companson Ratios a Nutrient Mean + SD Percent < l .ob Percent > 1.Oc Quantity 0.92 + 0.30~ 71.0 29.0 Energy 0.92 + 0.26 72.6 27.4 Protein 1.01 + 0.32 54.8 45.2 Fat 0.98 + 0.28 61.3 . 38.7 Carbohydrate 0.85 + 0.32 82.3 17.7 Calcium 0.96 + 0.56 72.6 27.4 Phosphorus 1.00 + 0.44 62.9 37.1 Iron 0.93 + 0.26 67.7 32.3 Vitamin A 0.85 + 0.33 75.8 24.2 Thiamin 0.96 + 0.35 61.3 38.7 Riboflavin 0.98 + 0.48 69.4 30.6 Niacin 1.05 + 0.52 56.5 43.5 Ascorbic acid 0.78 + 0.42 79.0 21.0 All variables (x) 0~94 68.2 31.8 Diary-interview nutrient intake . observed nutrient intake. b Comparison ratio less than 1.0 (diary-interview < observed). c Comparison ratio greater than 1.0 (diary-interview ~ observed). ~ n = 62 subjects. percent of population in the range of 10-20 percent underestimation (com- parison ratio, 0.8-0.9) of energy intake. Furthermore, 13 percent of the population underestimated by more than 30 percent and 5 percent overesti- mated energy intake by more than 30 percent. Projection of the results of these dining hall meals to the evaluation of total daily nutrient intake data obtained by the same diary-interview technique is tenuous. The average total daily energy intake measured by diary-interview was 2,650 kcal, a value approximately 12 percent less than a reasonable estimate of 3,000 kcal energy expenditure for a light activity male population with an average age of 26 years and weighing an average of 76 kg. This suggests that the underestimation of energy intake away from the dining hall was greater than 10 percent. We were concerned that the apparent greater underestimation of energy intake away from the dining hall might be due to the greater uncertainties regarding portion and unit size than in the dining hall. We have always obtained serving size information from the military dining halls and have made it available to the interviewers to use as a guide in assisting their subjects in estimating amounts consumed. At Twentynine Palms, Calif., the number of commercial food outlets readily available to the marines on and off the base is very limited. Therefore, the interviewers were able to visit

186 D. D. SCHNAKENBERG et. al. 35 30 25 - o o— - >~ 20 By 11 10 5 o [IL EN ERGY ~1 ~ LO LO · . . 1 1 Cal ~ A . . o LO (D ~ 00 1 1 1 1 1 1 V ~ (D ~ 00 ~ O . . . . . — COMPARISON RATIOS FIGURE 2 almost all of the eating establishments frequented by the marines as well as the local grocery stores and obtain portion and unit size information. The average daily energy intakes over the 14-day study period are shown in Table 6. At this time, we are unsure if we shoud attribute the nearly 500 kcal/day difference in reported energy intake between single and married male marines to group differences in reliability of data collection, to real differences in energy expenditures, or both. Since the married males consumed 82.6 percent of their calories at home compared to only 18. 1 percent for the single males, even a modest bias towards greater underestimation of intake at home compared to dining halls, restaurants, and vendors would significantly

Diary-Interview Technique 187 TABLE 6 Energy Intakes and Selected Demographic and Anthropometric Characteristics of Three Populations of Marines, Twentynine Palms, Calif., 1977 Single Males Married Males Single Females Number of subjects 57 59 36 Daily energy intake, kcal/day 3, 105 + 688a 2,637 + 684 1 ,991 + 659 Age, yr 20.5 + 1.6 24.4 + 5.1 20.9 + 2.3 Height, cm 178.4 + 7.2 176.7 + 6.1 165.6 + 6.4 Weight, kg 73.1 + 8.7 76.6 + 11.4 58.5 + 6.9 Percent average daily energy from: Dining halls 52.1 + 23.0 3.3 + 7.7 22.1 + 24.3 Home 18.1 + 19.1 82.6 + 14.2 39.1 + 29.8 Restaurants 19.1 + 16.1 8.0 + 9.8 23.8 + 17.0 Vendors 10.7 + 9.2 6.1 + 7.2 15.0 + 12.3 a Mean + SD. contribute to the reported differences in total daily energy intake. During the current study at Twentynine Palms, we are asking some of the diary-interview subjects to wear heart-rate monitors during 24 to 72 hours of the data collection period. Energy expenditure will be estimated from the average daily heart rate and an individualized heart rate versus oxygen consumption curve obtained during a graded exercise test. We are hopeful that the energy expenditure measurements, combined with weight change data, will greatly assist in our estimation of error in measuring energy intake of male and female military populations. Our rationale for asking our subjects to maintain their food diaries for 1~17 consecutive days was to look for possible effects of weekends and nav~av on total daily nutrient intake. However. as shown in Figure 3, our r—~——~ ~ -, analyses of these effects have been confounded by a definite trend, espe- cially in the male subjects, to report consuming fewer calories as the dura- tion of the study was extended. For example, at Twentynine Palms the single males reported consuming almost 3,400 kcal/day during the first 3 days, compared to approximately 2,900 kcal/day during the last 3 days of the 14-day study (Figure 3~. The magnitude of the decline was approxi- mately 200 kcal/day for the female marines. The decline in reported intake may reflect a change in attitude of many of the subjects as the novelty of the study wears off and they begin to tire of maintaining the diary record. Dr. Kretsch is currently analyzing these data to try to select the appropriate duration for our studies and is also comparing these diary-interview data to the dietary recall data we obtained during the first visit with the subjects. During the uss Saratoga study, we obtained 21-day body-weight change data in addition to age, weight, height, skinfold thicknesses, and perceived

88 3,600 3,400 3,200 3,000 2,800 - 2,600 UJ ye 6 2,400 no 2,200 of LLJ 2,000 1 ,800 1 ,600 D. D. SCHNAKENBERG et. al. Single Females n= 36 \ ~ Married Males n= 59 ~ / \ / \ . T F S 1 1 1 0 ~ \: Si ng~e Males V , ~  , . S M T W T F S S M T W 1 1 1 1 1 1 1 1 1 1 1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 DURATION OF STUDY (days) FIGURE 3 level of physical activity (light, moderate, heavy) information. These data were used to compute a predicted daily energy expenditure value for each subject. A reliability index value was computed for each subject by the following: Reliability Index Reported daily energy intake redacted dally energy expenditure For the entire population, the reported daily energy intake over the 17-day period was 2,004 kcal/day, compared to a predicted daily energy expendi- ture of 3,040 kcal/day. To try to determine why the reported energy intakes were so much lower than in previous studies, Dr. Schnakenberg used the Reliability Index to divide the population into "more reliable subjects" (index ~0.8) and "less reliable subjects" (index <0.81. As shown in Figure 4, the reported total daily energy intakes dropped precipitously during the first 7 days

Diary-lnterview Technique 3,200 3,000 2,800 ~ 2,400 y 2,600 ~ 2,200 o 2,000 1,800 1,600 1,200 UJ = 1,000 6 By 800 400 200 o 2.5 On J UJ 2.0 1.5 189 \ \ \ _ \ More Reliable Subjects n = 52 - Less Reliable Subjects n = 151 ~?~ - - - 1 1 1 1 1 1 1 1 3 5 7 9 11 13 15 17 19 21 23 25 DU RATI ON OF STU DY (days) FIGURE 4

190 D . D . S CHNA KEN BE RG e t al . Of the study. Following an 8-day in-port period when data were not collected, reported energy intake stabilized at a low level. The 50-100 kcal/day drop in reported calories from snacks contributed only slightly to the 600-700 kcal decline in total daily energy intake. Therefore, the major factors were the decline in kcal/meal and the reduced number of meals reported being consumed per day. It is interesting to note that the curves of the "more reliable" and "less reliable" subjects are essentially parallel. We conducted this study at the beginning of the 7-month long cruise, and the subjects' adjustments to the rigors of sea duty and separation from family and friends may have contributed to the problems in completeness of data collection. The Saratoga study posed a serious problem for data evaluation and in- terpretation. Despite the conscientious efforts of all members of the survey team, the data collection procedures resulted in the average energy intake of the population being underestimated by approximately 30-35 percent. Therefore, should the data of only those subjects whose energy intakes appear to be reasonably complete be used in the evaluation of the food service system or should all the subjects be included? If an estimate of the number of lunch meals consumed in the forward galley was needed by food service delivery system planners, Dr. Schnakenberg used data from only the more reliable subjects. However, he used data from all subjects to evaluate the nutritional adequacy of meals and total diet consumed on the ship. This was because of concern that the more reliable subjects might have a greater nutritional awareness and better food selection patterns than their less reliable counterparts. He has assumed however, that although incomplete, the reported nutrient intakes of the less reliable subjects are a representative sample of their food selection patterns. With this assumption, he expressed the nutrient intake data on essentially a nutrient density basis by using the concept of Nutrient Ratio (NR), where: Intake expressed per 1,000 kcal consumed NR = - Nutritional standard expressed per 1,000 kcal He further tried to simplify the data evaluation process by categorizing the nutrient intakes as either "low, " "marginal," or "adequate" according to the following arbitrarily selected criteria: "Low" (NR <0.7) intake less than 70 percent of standard "Marginal" (NR 0.7 <1.0) intake between 70 percent and 100 percent of standard "Adequate" (NR~l.O) intake greater than 100 percent of standard The military dietary allowances (Departments of the Army, the Navy, and the Air Force, 1976) were used as the nutritional standards for these calcu- lations.

Diary-Interview Technique The nutrient ratio concept can be used not only to evaluate the adequacy of average daily nutrient intakes (Table 7), but also to compare the nutri- tional adequacy of meals consumed from various sources such as the aft galley (Table 8), forward galley (Table 9), and nongalley nutrition (Table 10) and to monitor the eating habits (Table 11) and food type consumption patterns (Table 12) of individuals with low, marginal, or adequate intakes of specific nutrients. This type of data presentation is useful in evaluating the reasons for a high incidence of low intakes of a specific nutrient, such as 191 TABLE 7 Evaluation of Average Daily Nutrient Intakes, uss Saratoga, July-August 1977 Percentage of Population a Nutrient Lowb MarginalC Adequate Protein 0 3.9 96.1 Calcium 1.5 21.2 77.3 Iron 3.0 52.7 44.3 Vitamin A 20.2 31.0 48.8 Thiamin 3.0 52.7 44.3 Riboflavin 0.5 24.1 75.4 Niacin 0.5 23.2 76.4 Vitamin C 8.4 18.2 73.4 a203 subjects. Nutrient ratio <0.7. c Nutrient ratio 0.7 to <1.0. Nutrient ratio ~ 1.0. TABLE 8 Evaluation of Aft Galley Meals, uss Saratoga, July-August 1977 Percentage of Population a Nutrient Lowb MarginalC Adequate Protein O O 100 Calcium 5.4 19.8 74.8 Iron 0.5 28.2 71.3 Vitamin A 10.4 22.8 66.8 Thiamin 0.5 33.2 66.3 Riboflavin 0 10.9 89.1 Niacin O 9.9 90.1 Vitamin C 5.0 9.9 85.1 a202 subjects who reported eating at least one meal in aft galley. Nutrient ratio <0.7. CNutrient ratio 0.7 to < 1.0. Nutrient ratio ~1.0.

192 D. D. SCHNAKENBERG et. al. TABLE 9 Evaluation of Forward Galley Meals, uss Saratoga, July-August 1977 Percentage of Population a Nutrient Lows MarginalC Adequate Protein 0.5 13.0 86.4 Calcium 7.1 17.9 75.0 Iron 8.7 43.5 47.8 Vitamin A 61.4 26.1 12.5 Thiamin 2.2 39.1 58.7 Riboflavin 1.1 32.6 66.3 Niacin 8.2 31 .5 60.3 Vitamin C 50.0 14.1 35.9 a 184 subjects who reported eating at least one meal in forward galley. Nutrient ratio <0.7. CNutrient ratio 0.7 to <1.0. Nutrient ratio ~1.0. TABLE 10 Evaluation of Nongalley Nutrition, uss Saratoga, July-August 1977 Nutrient Percentage of Populationa Lowb MarginalC Adequate Protein 78.6 14.3 7.1 Calcium 19.4 38.3 42.3 Iron 76.0 10.7 13.3 Vitamin A 91.3 3.6 5.1 Thiamin 89.3 6.1 4.6 Riboflavin 73 .5 8.2 18.4 Niacin 68.4 6.6 25.0 Vitamin C 77.0 3.6 19.4 a 196 subjects who reported consuming foods and beverages from sources other than from aft and forward galley meals. Nutrient ratio <0.7. CNutrient ratio 0.7 to c 1.0. Nutrient ratio ~1.0. vitamin A, and in designing and predicting the impact of intervention pro- grams. For example, 61 percent of the meals consumed in the forward galley (Table 9) were low in vitamin A, indicating that a salad bar should be added to the predominantly short order type of menu. However, because subjects with low vitamin A intakes (Table 12) avoid tomatoes, carrots, leafy and green vegetables, it is unlikely that simply adding a salad bar will markedly increase the vitamin A intakes of this group. As shown in Table 11, the low vitamin A intake group selected meals from the aft galley that were low in vitamin A (NR =0.73), even though vitamin A was abundantly available.

193 Cq U. au Ct .~ 5 ~ 1 _ 1 ct 1 ~ 1 Ct 1 ~ 1 s~ o ~; '' "] 't ~ ~ ~ ~ oo ~ ~ == I tl tl tl tl tl tl tl tl tl tl I c' 1 cr oo — ~ ~ ~ os O 1 ~ ~ ~ ~ ~ ~ o 0 ~ ~ ~ _o_~___ _oo . . 3 ~ 1 1 ·Z ~; . ;^ . 1 4, 1 u' 1~ C) ~ ..~ C~ ~ ~oo~ ~ (~\ <, ~ · · o _ ~ u~ 1 ', (-, ~ ~ V) o ,~N tl tl tl tl tl tl tl tl tl tl 1 ~ ~ ~ o ~ _ ~ oo ~ c~ ~ 1 . ~ o ~ — ° ~ ~ 1 ! - ° - ~ ~ - - - o o 1 _ ~ C~, ~ 1 · · ~ — _ _ I tl tl tl tl tl tl tl 1 _ o ~ ~ ~ ~ u~ 1 ~ — V) ~o c~\ 1 _ o _ ~ C`l c~] _ 1 1 l | 1 3 1 ~ ~ ~ ~ ~ 1 ~ ~ ~ ~ ~ ~ ~ ~ ~ o oo ~ o ~ 4) | 4' I tl tl tl tl tl tl tl tl tl tl | ,= 1 ,~ 1 ~ o 0 ~ 0 0 0 ~ ~ 0 oo 1 ~ ~ ~ U) - ° ~ o _ ~s' °° ~ ~ —o o ~ C~ ~ · _ | | . _ o . ~ o ~ 1 ~ _ c: ~ ~ 3 V o A\ ~ _ - . ;=, ~ ~ o ~ E ~ . ! m ~ 10cx5cJ_~ ~ <=o3 ~ctC~Ct ~ ~ 1 1 o 0 0 0 ° ° ° ° ce -5 ~ Z ._ .~.~ -' ~ ~zzzz~> c ?>c

194 D. D. SCHNAKENBERG et. al. At sea, there is insufficient refrigerated storage space to keep fresh milk available for periods of more than =7 days. This problem is described in Figure 5a, where the ship ran out of fresh milk on July 15, was resupplied TABLE 12 Food-Type Consumption of Subjects with Low, Marginal, or Adequate Vitamin A Intakes, uss Saratoga, July-August 1977 Quantity (g/day) Consumed by Subjects Whose Daily Vitamin A Intakes Were: Lowa Marginalb AdequateC No. of subjects 41 63 99 Milk 137 236 178 Cheeses and ice cream 19 29 31 Tomatoes and tomato products 9 25 38 Carrots, raw and cooked 0.2 1.5 8 Sweet potatoes 0.3 1.8 2.0 Liver 0 0 1 .3 Leafy and green vegetables 6 14 31 Melons, peaches, plums 11 12 28 Eggs 24 41 42 Potatoes, french fries 8 11 11 a Nutrient ratio <0.7. b Nutrient ratio 0.7 to < 1 . 0. c Nutrient ratio ~1 . 0. DAIRY PRODUCTS AND BEVERAGE CONSUMPTION DAILY TOTALS 2000- TRANSIT | PORT , 1 500 - 6 CD ~ 1 000 - - z C' 500- | F Ll G HT OPS —DAIRY PRODUCTS —OTHER BEVERAGES O- ~ / / ~ 3 5 7 2 14 ·6 18 20 22 24 26 28 30 ~ 1977 AUGUST USS SARATOGA FIGURE Sa

Diary-Interview Technique with filled milk while in port, and ran out again on July 31. The consump- tion of carbonated and noncarbonated beverages markedly increased when milk was not available. The percentage of the population with low and marginal intakes of calcium (Figure Sb) and riboflavin (Figure 5c) substan- tially increased on days when milk was not available. At our urging, a low-fat (25-30 percent fat calories) vitamin A-fortified (30 percent RDA per 195 CALCI U M DAILY TOTALS TRANSIT | PORT | FLIGHT OPS Rn ~ Z O O ~ ,_ 60 Z ~ LLI J UJ o ~ to 40 o ~ 1 2 1 4 1 6 1 8 20 22 24 26 28 30 1 JULY 1977 USS SARATOGA AUGUST O—ADEaUATE MARGINAL LOW FIGURE Sb RIBOFLAVIN DAILY TOTALS 100 - 80 - ~ Z O O 60- Z ~ 111 A O ~ to ~ O Q ~ 40 - 20 12 14 16 18 20 22 24 26 28 30 1 JULY 1977 USS SARATOGA . TRANSIT | PORT | FLIGHT OPS 3 6 7 AUGUST D—ADEQUATE ~ MARGINAL ~ LOW FIGURE Sc

196 D. D. SCHNAKENBERG ET. AL. serving) milk shake, which is reconstituted aboard ship from a dry shelf stable base, is being tested on the uss Saratoga. Based upon the food-type selection patterns previously shown in Table 12, we have projected that this product will effectively reduce the incidence of low and marginal intakes of calcium, riboflavin, and vitamin A as shown in Table 13. One of the pri- mary purposes of our current follow-up study on the uss Saratoga is to assess the validity of our projections. As a result of our recent experiences from seven dietary assessment studies in a variety of military settings and our attempts to validate our methodology, we have made some modifications to our dietary-interview technique. Most significantly, we have reduced the time the participants are asked to maintain their food diaries from 14-17 days to no more than 7-8 consecutive days. We believe that 7-8 days is long enough to obtain a reliable sample of an individual's eating patterns and food selection habits, but short enough to maintain a cooperative attitude in most of the partici- pants. It is still essential though that a diary review session be scheduled 3~ days after the initial interview. We are incorporating some sort of validation procedures in each of our studies, most of which focus on obtaining an estimate of the energy expenditure level of each subject. However, we are keenly aware that the magnitude of the error in estimation of energy expen- TABLE 13 Projected Nutritional Impact of Offering Vitamin A-Enriched Milk Shakes,a uss Saratoga Percentage of Population Nutrient Calcium intake Jul.-Aug. 1977 Projectedb Riboflavin intake Jul.-Aug. 1977 Projectedb Vitamin A intake Jul.-Aug. 1977 Projectedb c Low Marginal Adequate 2 21 77 2 10 88 1 24 75 1 11 88 20 3 1 49 5 13 82 a Shake mix enriched to provide 1,800 flu vitamin A per 12.5 fl. oz. serving. b projections based upon Jul.-Aug. 1977 data selecting only the days when milk was available. c Based upon presumption that individuals who consumed at least one glass of milk per day when available will consume one enriched milk shake per day. NOTE: These projections assumed that milk shakes will be available at both forward and aft galleys.

Diary-lnterview Technique 197 diture and short-term caloric deficit or surfeit of individuals may be as large as the error in estimating energy consumption. Therefore, we are beginning to concentrate more on utilizing a nutrient ratio, nutrient balance, or, if you prefer, nutrient density approach to the analysis and interpretation of the data rather than limiting our interpretations to reported average daily quan- tities of nutrients consumed. We are rapidly expanding our computer pro- gramming capabilities so that we can more fully utilize the information contained in our data bank. Mr. Morris has developed the necessary software to retrieve individual nutrient intake data by demographic and anthropometric characteristics, by date, by time (breakfast, lunch, supper, etc.), by source (home, restaurant, vendor, dining hall, etc.), or by combi- nation of several variables. We can also retrieve, by individual or group, the quantity of any food item (limited to the ~2,500 different food items currently on our Nutrient Factor File), or food type consumed. Dr. Kretsch has recently overseen the addition of cholesterol, animal/plant protein, and animal/plant/marine fat values for nearly all food items on the file and a significant number of vitamin E, vitamin B6, vitamin Bit, folio acid, mag- nesium, copper, zinc, manganese, pantothenic acid, and fatty acids values have been added. These improvements in our data-processing and retrieval system should significantly contribute to the knowledge base of nutrient intake patterns of not only military personnel but our young adult population in general. REFEREN CES Departments of the Army, the Navy, and the Air Force. 1976. Army Regulation 40-25, BUMED Instruction 10110.3E, and Air Force Regulation 160-95. Medical services nutritional stan- dards. Departments of the Army, the Navy, and the Air Force, Washington, D. C., 30 August 1976 (as corrected). Gersovitz, M., J. P. Madden, and H. Smiciklas-Wright. 1978. Validity of the 24-hr. dietary recall and seven-day record for group comparisons. J. Am. Diet. Assoc. 73:48.

Possible Alternative Me~ods for Data Collection on Food Consumption and Expenditures ROBERT B. PEARL There are few behavioral patterns in our society that are more extensively documented than the purchase and consumption of food products. Countless practitioners, both inside and outside of government, are engaged in pro- ducing statistics on this subject either on a continuous or intermittent basis. Moreover, there are probably almost as many different methodologies and techniques employed in gathering these data as there are participants in the field. In view of these facts, it would be surprising if almost every possible means of developing these statistics had not already been attempted or at least considered. This likelihood poses a problem for anyone charged with presenting challenging and innovative ideas along these lines. The tempta- tion exists to opt for exotic possibilities, such as using spy satellites to chart the eating behavior of the population. The writer intends to resist this temp- tation in favor of a few notions that, if not entirely novel, at least do not appear to have entered the mainstream of statistical endeavors in this sector. In the main, these will be aimed at some of the principal deficiencies that are believed to exist in nutritional and dietary data. INTERMITTENT PANELS One of the limitations of the methodologies customarily employed in studies of food consumption is the relatively short period of observation. Because of cooperation problems and concern about reporting errors, comprehensive family consumption data have usually been restricted to a period of 1 week or less and individual intake data to a span of a very few days. Even food 198

Possible Alternative Methods for Data Collection 199 purchase or expenditure information has rarely been collected for more than a 2- to 4-week period. Assuming samples of sufficient size that are adequately distributed over time, these restrictions should not impair such measures as average family or individual consumption. The limitation arises when an effort is made to provide distributions of families and persons by nutritional status. Since studies have indicated that dietary inadequacies exist in every socioeconomic group, the focus may be more on distributions and individual differences than on group averages. In view of the variations that are likely to occur in food consumption from day to day, week to week, or season to season, it is difficult to see how observation for only a brief period can validly depict the nutritional status of any one family or person. The preferable course would probably be to observe each subject for an extensive, continuous period, perhaps up to a year. Unfortunately, experi- ence has indicated that samples used for continuous expenditure surveys become highly biased because of attrition (Quackenbush and Shaffer, 1960), and there would be even less hope for success in the case of indi- vidual food consumption measurement. One alternative would be to attempt an intermittent panel procedure, whereby the same subjects would be can- vassed at various points of the year, but for a relatively brief period each time. The data from the various observations would be accumulated to provide the measure of nutritional status. Although some further ex- perimentation might be needed, a canvass once each quarter could be adequate for a particular case. Some analyses have indicated that a 3-week period of observation each time for family consumption might be the op- timum in this type of design, balancing the need for stable measurements against that of avoiding undue sample attrition or fatigue (Ferber and Sud- man, 1971; Report and Recommendations, 1976~. For a specific family, the observation period could be randomly scheduled to occur at different times each quarter in order to cover a greater variety of situations. The designation of a 3-week observation period would probably permit the use of a food purchase approach in data collection, in lieu of the more complex food consumption procedure, since the two are likely to balance out for periods of that length (especially where there would be four quarterly observations to aggregate). A food purchase approach, in turn, would likely entail the use of diary keeping, which has been found to be relatively successful for that purpose (Pearl, 1977, 19781. For individuals as opposed to families, the procedure would probably call for maintaining records of consumption for possibly a 3-day period randomized within each 3-week cycle; the family data could then be used as a control (for home consump- tion) over the individual data. The use of a panel always raises problems of securing adequate coopera-

200 ROBERT B. PEARL lion over time. An appropriate incentive system might well be necessary, with a modest reward for completing any one quarterly effort and a sizeable bonus for persisting through the entire cycle. Since panel members could move from quarter to quarter, some follow-ups would be required each time, but these are not likely to exceed 5 percent of the total at any one point. UNI VERS AL PRODUCT CODES Another deficiency in much of the information that has been collected in this field, whether through diaries or interviews, has been inadequate and incon- sistent descriptions of products. This problem is especially acute where conversion to a nutrient basis is the objective. The proliferation of prepared foods, containing a variety of ingredients, obviously adds to the difficulty. One possible way of reducing this problem would be to make use of the Universal Product Codes (uPc) that now appear on nearly all canned and packaged food products and even on some fresh items that are repackaged in stores. Experimentation would be needed to determine whether survey re- spondents could be induced to enter these codes in diaries, in addition to a brief product description, and, moreover, whether they could do so accu- rately. If this could be achieved, and assuming that the necessary data on ingredients could be obtained from the respective manufacturers and stored in computers, conversion to a nutrient basis could become a routine matter. Moreover, consistency and reliability in reporting would be materially ad- vanced and clerical coding and processing costs could be minimized. AUTOMATED CHECKOUT INFORMATION Another technological development of major interest, still limited in scope but offering promise for the future, is the appearance of automated checkout systems in food stores based on the Universal Product Codes. As these become more prevalent, they could serve as the basis for obtaining current food sales data in a level of detail never before contemplated. Although sales data are of only limited value for purposes of measuring family or individual food consumption, they could supply control totals for usually deficient survey data or at least provide a standard for evaluating survey results. Moreover, if all purchases are made by credit card in some imagina- ble future era, the combination of the sales data from mechanized systems and the personal characteristics of credit card holders in the possession of various financial agencies could produce the best of all possible statistical worlds. (This target may sound like " 1984," but is not likely to occur by that date.) Perhaps a more realistic application would be the use for survey purposes of the individual cash register tapes produced by the automated systems,

Possible Alternative Methods for Data Collection 201 which generally provide a far more adequate description of the products purchased than has heretofore been available. These tapes could serve as a basic reminder to survey respondents in filling diaries or responding in interviews and, in some cases, could even serve as a substitute for complet- ing parts of diaries. PORTABLE TAPE RECORDERS No matter how carefully surveys of this kind are planned, serious reporting errors are likely to occur because of memory biases, tedium in filling diaries, and other reasons. A possible aid in overcoming these deficiencies would be to provide respondents with small, pocket-size tape recorders to use during the survey period. Respondents would then be asked to record purchases as they are made in the store or are being unpacked at home (possibly including uPc numbers) or, alternatively (if the consumption ap- proach is being used), as various items are being used in food preparation. Moreover, individuals within families might be given such recorders in which to report meals and snacks purchased outside the home. The use of recorders could be especially useful in the case of less literate respondents. DIARY CHECKING PROCEDURE Some recent studies have indicated that there may be considerable dis- parities in the relative accuracy and completeness of reporting various food products in diaries. For example, comparatively costly items, such as meat and poultry, and perishable and frequently used products, such as milk, eggs, and bread, seem to be rather well reported, whereas food staples such as flour, sugar, and shortening tend to be seriously understated (Pearl, 19784. There appears to be a need, therefore, for special checking proce- dures to correct for these imbalances. Some general checking procedures are specified in most diary operations, but what may be needed are specific sets of check questions aimed at those product categories most commonly underreported. In fact, the entire process may become much more of a combination of diaries and interviews than has usually been the case. A companion approach might be to take a brief shelf inventory of these types of staple products, at the time the diaries are being collected, and to inquire about the date of purchase of any that are found in order to determine whether the information was properly recorded. WAS T A GE It is difficult to review the literature without finding a good deal of hand wringing about the inability to measure waste in food use (Pearl, 1977~.

202 ROBERT B. PEARL This writer has no inspired solutions to offer for this long-standing problem. Instead of the rather futile attempts to measure this element as part of a regular consumption survey, however, it might be more useful to carry out a series of small-scale studies devoted specifically to finding out how much of various food items are ordinarily wasted. Some spot visits to small house- hold samples might be made right after the various meals to obtain quick information of this kind. Respondents might also be asked to record wastage on portable tape recorders, etc. In any event, the objective would be to develop adjustment factors that could be used to discount the usual purchase or consumption data prior to conversion to nutrients. MEALS EATEN OUT Another matter of some discouragement has been the impracticality of ob- taining reliable information from survey respondents on the contents of meals eaten out. Since one out of every three food dollars is now spent in this manner, and the ratio is expected to rise to one in two, this problem is rather central in any effort to measure dietary adequacy. One possibility, in addition to obtaining whatever descriptions can be provided by respondents of meals eaten out, would be to record the names and addresses of the eating places. The objective (and this again might be attempted only for small samples) would be to solicit the recipes and lists of ingredients directly from the eating places (hopefully avoiding trade secrets). From a succession of such investigations, a glossary could possibly be developed that—at least in nutrient form could describe a large proportion of meals eaten out (the preponderance of meals eaten in syndicated fast food establishments could help in this respect). Although obviously not without limitations, this ap- proach could possibly improve what is one of the weakest links in the chain of dietary measurements. INDICATOR FOODS One of the problems in many complex surveys is that so vast an array of information is solicited that respondents experience memory failures or fail- ures of execution, and survey personnel become so lost in the details that they are unable to institute adequate checking procedures. An apparent solution would be to reduce the amount of information to be requested from any one respondent, that is, in our immediate field of interest, to ask only about selected foods instead of the entire gamut. Since we are interested in the overall dietary adequacy of individual families and persons, the question is how to accomplish a reduction in detail without defeating the main objec- tives of the inquiry.

Possible Alternative Methods for Data Collection 203 One possibility would be to attempt to develop a set of "indicator" foods, that is, a selection of foods that when translated into nutrient form would approximate the nutrient value of the complete diet of a given family or individual. This approach would represent a form of sampling similar to that used in estimating the overall price index from only a "market basket" of items. The data for the indicator foods would, of course, have to be appro- pnately weighted to arrive at a total nutrient value. In fact, regression equations might be developed to aid the estimation. Various means might be utilized in the development of indicator lists. The knowledge and experience of dietary specialists would represent an impor- tant input. Experiments could be conducted with complete data sets (such as those from the ARS Food Consumption Surveys (see Murray, 1975), whereby venous successive samples of items could be drawn and compared in nutrient value with the total consumption of the family or individual. For every item on an indicator list, there could be two or three options, in case the family did not use a particular product. Moreover, it would probably be necessary to develop different indicator lists for the various geographic regions, socioeconomic groups, and season of the year. REFERENCES Ferber, R., and S. Sudman. 1971. Experiments in obtaining consumer expenditure data by diary methods, J. Am. Stat. Assoc. 66:336, December. Murray, J. 1975. Household food consumption surveys of the U.S. Department of Agriculture: A historical review of data collection methods, 1894-1969. Agricultural Research Service. Pearl, R. B. 1977. Data systems on food demand and consumption: Properties, uses, future prospects. Pages 116-139 in Food demand and consumption behavior. Workshop spon- sored by the S-119 Southern Regional Research Committee, State Agricultural Experiment Stations, and the Farm Foundation, Atlanta, March. Pearl, R. B. 1978. The 1972-73 U.S. Consumer Expenditure Survey: A preliminary evalua- tion. U.S. Bureau of the Census Technical Paper Series. In press. Quackenbush, G. G., and J. D. Shaffer. 1960. Collecting food purchase data by consumer panel. Mich. State Univ. Agric. Exp. Stn. Tech. Bull. 279. East Lansing, Mich. Report and recommendations to the food and nutrition service on a proposed national food stamp program consumer panel survey. April 1976. Unpublished.

Next: Appendix C: Background Papers for Workshop on Nutrition and Health Status Indicators »
Assessing Changing Food Consumption Patterns Get This Book
×
Buy Paperback | $80.00
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF
  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  6. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  7. ×

    View our suggested citation for this chapter.

    « Back Next »
  8. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!