The definition of a homeless individual used by the WIC program is broad:

Estimates of the number of homeless people vary widely because of ambiguities in the definition of homelessness and lack of a reliable counting method. During the 1980s, point-prevalence estimates of homelessness in the United States ranged from 350,000 to more than 3 million (USDA, 1991). Owing to inevitable design problems, surveys that actually try to count the number of homeless individuals generally provide smaller estimates than do other methods of estimating prevalence (Link et al., 1994). This recent study reports lifetime (any time) and 5-year prevalence estimates of homelessness that are significantly higher than previous point-prevalence estimates: (1) a lifetime combined prevalence of all types of homelessness of 14.0 percent (26 million individuals); (2) a 5-year (1985 to 1990) prevalence of all types of homelessness of 4.6 percent (8.5 million individuals); and (3) a lifetime prevalence of literal homelessness (sleeping in shelters, abandoned buildings, bus and train stations, etc.) of 7.4 percent (13.5 million individuals). Comparison of these 5-year and lifetime prevalence estimates with the earlier point-prevalence estimates of homelessness suggests a fluid process with people flowing in and out of homelessness.

Despite controversy over the absolute numbers, there is consensus that the number of homeless individuals increased steadily during the 1980s (Rossi et al., 1987). During this period several surveys of local areas found that between 30 and 50 percent of the homeless were families with children, usually headed by single women with at least one child under 18 years (Bassuk et al., 1986; Miller and Lin, 1988; New York Coalition for the Homeless, 1986; U.S. Conference of Mayors, 1984, 1986, 1987).

A variety of risk factors are associated with homelessness. Weitzman (1989) suggests that pregnancy and recent birth are precipitating factors for homelessness. Homeless women, especially homeless pregnant women, are more likely than housed low-income women to have alcohol problems, to use drugs, and to smoke (Becker et al., 1992; Fischer, 1991; Jaffee et al., 1992). Homeless women are also likely to be victims of domestic violence; 33 to 89

FRONT MATTER (R1-R14)

SUMMARY (1-22)

1 OVERVIEW (23-40)

2 POVERTY AND NUTRITION RISK (41-52)

3 PRINCIPLES UNDERLYING THE NUTRITION RISK CRITERIA FOR WIC ELIGIBILITY (53-66)

4 ANTHROPOMETRIC RISK CRITERIA (67-148)

5 BIOCHEMICAL AND OTHER MEDICAL RISK CRITERIA (149-250)

6 DIETARY RISK CRITERIA (251-294)

7 PREDISPOSING NUTRITION RISK CRITERIA (295-334)

8 CONCLUSIONS AND RECOMMENDATIONS (335-352)

APPENDIXES (353-352)

A PARTICIPANTS AT THE FIRST PUBLIC MEETING, MAY 19, 1994 (353-354)

B PARTICANTS AT THE SECOND PUBLIC MEETING, SEPTEMBER 19-20, 1994 (355-356)

C WIC PROGRAM: COMMON NUTRITIONAL RISK CRITERIA (357-358)

D DEFINITIONS OF YIELD AND SENSITIVITY OF CUTOFF POINTS FOR NUTRITION RISK (359-360)

E BIOGRAPHICAL SKETCHES (361-370)

ACRONYMS (371-372)

INDEX OF RISK CRITERIA (373-377)