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.
9 Consumer Behavior Although it is clear that people do markedly alter their diets during the course of their lives, we know very little about either the environmental factors that are responsible for such changes or the temporal pattern, if any, of their occurrence. There has been essentially no long-ter~ research on patterns of food selection by humans and the factors that affect them. Thus, even when knowledge about diet and disease relationships is certain enough to make interventions prudent and appropriate, there are no proven measures that either educators or policymakers can take to help the public modify its long-term food intake. Because we lack knowledge about the causes and patterns of long- tenm dietary behavior, it is difficult to reach even post hoc conclu- sions about their possible relationships to various (often desirable) effects. It is now clear, for example, that the rates of death from coronary heart disease are declining in the United States and that they have been doing so for at least 14 years (Stamler, 1982; Walker, 1977~. It is also clear from marketing research, from public opinion surveys, and from other research that several other changes have taken place over this same period: (1) some unknown portion of the population has adopted a lowliest or essentially vegetarian lifestyle; (2) other persons have modified their diets in different ways for some specific health-related reason; and (3) a certain segment of the population has taken up some form of regular exercise, begun treatment for high blood pressure, and given up smoking (Jones and Weiner, 1981; Louis Harris and Associates, Inc., 1978, 1979; Mark Clements Research, Inc., 1980; Rowland and Roberts, 1982; Stabler, 1978; Yankelovich, Skelly and White, Inc., 1978, 1979, 1980~. The extent to which any of these changes is related to the reduction in cardiovascular disease and-- more to the point here--to intentionally undertaken educational programs or other interventions is unclear. Indeed, as elucidated in a recent report on the Multiple Risk Factor Intervention Trial (MRFIT), "spontaneous"--that is, unex- plained--behavior change ran reduce the observed differences between experimental and control groups in even carefully designed prospective studies. Enrolled in this 7-year trial were almost 13,000 men at high risk of heart attack, one-half of whom were given "special interven- tion" to help them stop smoking, reduce their blood pressure, and modify their diets to lower their serum cholesterol. As it turned out, the control group (which only had physician contact) also modified its risk factors and had lower mortality from cardiovascular disease over the period of the trial, as did the entire U.S. population (Multiple Risk Factor Intervention Trial Research Group, 1982~. As yet, there 52
Consumer Behavior 53 are no parallel data from "cancer reduction trials." The MRFIT data make it clear that failure to study "spontaneous" dietary change, even as we are mounting intervention trials, may entirely becloud the results of those trials. The results from other intervention programs, even very intensive and expensive ones, would not have encouraged one to predict a wide- spread effect as great as that implied by a 30% "spontaneous" drop in mortality from coronary heart disease (Farquhar et al., 1977; Maccoby et al., 1977; Ringen, 1983; Smith et al., 1982; Stamler, 1982~. The results from several such studies are especially instructive in that they involve the use of media and of governmental intervention. These are often regarded as potentially highly effective channels through which desirable dietary change can be promoted in the United States, even though they are often also regarded as economically and politically unfeasible. In the Stanford Three Community Study, a sophisticated media cam- paign used in conjunction with intensive face-to-face instruction had relatively modest, although statistically significant, effects on dietary behaviors associated with heart disease (Cohen and Cohen, 1978; Farquhar et al., 1977; Fortmann et al., 1981; Leventhal et al., 1980; Maccoby et al., 1977; Meyer et al., 1980a,b; Stern et al., 1976~. More recently, Smith _ al. (1982) conducted a pilot intervention study using mass media plus school, store, community, and parent involvement. They attempted to increase children's consumption of snack foods that were low in caloric content or high i n nutrient density and to decrease their consumption of snack foods containing high levels of sugar and fat. Despite the fact that the campaign materials placed a major emphasis on increasing the use of fruits as snacks, the only measurable effect of the intervention was a modest reduction in the consumption of soft drinks and other sweets not specifically promoted. There was no increase in the consumption of heavily promoted snacks. Schucker and his colleagues (1982) studied the effects of the saccharin warning label on sales of diet soft drinks. Although there was a drop in the growth rate of sales for these products following ex- tensive publicity about the finding that saccharin caused cancer in animals, the investigators found it difficult to disentangle the effects of the warning label from the effects of price increases and advertising decreases during the same period. In summer 1980, when advertising returned to its prepublicity level, sales of diet sodas reached an all- time high. In Norway, an entire nation committed itself to a policy intended to modify its agricultural, marketing, and educational efforts in a manner that would encourage the consumption of a healthier and more "responsible" diet (Ringer, 1983; Royal Norwegian Ministry of Agricul- ture, 1975~. Although there were some changes in food intake in the
54 DIET, NUTRITION, AND CANCER: DIRECTIONS FOR RESEARCH desired direction, it was concluded after 5 years that "where consump- tion patterns so far have agreed with the goal s of the nutrition policy, the credit should be given to chance more than to conscious decision making (R~ngen, 1983~. Extensive data have also been published from two other kinds of studies whose relevance to directed long-term dietary change is un- clear: (1) successful programs that result in the modification of behaviors other than eating and (2) short-term interventions involving populations who volunteer to be educated nutritionally (or who have no choice, e.g., schoolchildren). In regard to the first of these, it is obvious that those concerned with long-term modification of food habits need to become acquainted with the extensive body of research into the psychological factors affecting participation in such health-related behaviors as smoking cessation, compliance with a medical regimen, or participation in immunization or health screening programs (Enelow and Henderson, 1975; Olson and Gillespie, 1981~. How- ever, the extent to which the findings from these studies are trans- ferable to spontaneous dietary modification is not clear. The kind of lifetime dietary change that is usually prescribed for risk reduc- tion--and implied by the m30% decline in heart disease--involves a complex series of choices made 3 or more times a day from an ever- changing mix of many thousands of food products of increasingly unfa- m~liar composition. Therefore, where food is concerned, the desired behaviors are more complex, i.e., determining how to modify one's diet meal by meal every day is a much more complex activity than simply remembering to take a pill or to go to the doctor. Moreover, monitor- ing compliance is also considerably more complicated, involving at the least the elicitation of several dietary reports from the subject. Most studies of the effects of nutrition education on food intake way also not be directly relevant to the question at hand, since they involve measuring short-term changes in the behavior of specific captive populations exposed to a carefully organized series of lessons for a relatively short period. It is not clear how much light such experiments can shed on the changes that actually occur over time in free-living populations exposed to the multimedia message stream in the United States. Moreover, the results from such studies are quite unimpressive where actual behavior change is concerned--when such change i s one of the outcome measures. The data are best, perhaps, for weight loss programs, simply because there are so many of them and the outcome is so readily measurable. A number of studies have now documented the fact that educational programs to encourage weight loss have a generally dismal record (Brightwell and Sloan, 1977; Stunkard and McLaren-llume, 1959; Wing and Jeffery, 1979~. Thus, there is evidence that people have changed their food habits, but there is little evidence that they have done so in large numbers on the basis of nutrition education programs aimed at producing such
Consumer Behavior 55 change. A recent study by Schacter (1982) suggests one possible explanation for this apparent inconsistency. Noting that most studies on the effects of weight-loss regimens show a very low rate of success, especially over the longer term, Schacter undertook a study of weight histories and weight-loss attempts in two arbitrarily selected samples, one consisting of all persons associated with a university psychology department and the other consisting of persons working in establish- ments that are open all year in a resort community. Contrary to ex- pectation, he found that 62.5% of all those with a history of obesity who had attempted to lose weight had succeeded and had kept the weight off. Schacter's suggested explanation of this phenomenon is that weight-loss programs, like those aimed at smoking or narcotic addic- tion, measure only one attempt of any one person to cure himself. Successful dieters, he speculates, probably try many times, finally succeeding on the nth try. Whether or not Schacter's data are confirmed by subsequent studies, they illuminate the reality that dietary change need not be--indeed, probably rarely is--a single dramatic change, but, rather, it may be accomplished through a series of modifications that take place over time in response to a variety of stimuli (Kolasa, 1981~. Meanwhile, intentional educational interventions cannot be shown to have had a substantial measurable effect, even though health statistics (including those on general life expectancy) suggest that substantial positive changes are taking place. It is possible, of course, to speculate that food choices are being beneficially altered by public awareness of a growing professional consensus that lifestyles associated with afflu- ence are related to a whole spectrum of diseases. However, there is no assurance that a consensus on dietary change is being communicated clearly, if at all (Skelly, 1982~. And it i s also not at all clear what people actually do--as opposed to what they say they will do--when they are informed. A survey taken shortly after the release of this committee's first report showed that 53% of those contacted were aware of the report (Anonymous, 1982~. However, 36% of the surveyed popu- lation said they would make practically no change in their diets, 16% said they would make a moderate change, and 2% indicated they would make a major change as a consequence of knowing about the report. All dietary change, whether "directed" or "spontaneous," may have a major impact on one or another sector of the f cod industry, with the result that there is often conflict between the private sector's interest in maximizing production and what public health officials re- gard as the public's interest, where health is concerned. The food industry has already made changes in response to recommendations from various groups in the past, for example, lowering the sodium content or modifying the fat content of foods. Attention should continue to be directed toward finding ways in which recommended changes can be phased into the food supply to minimize the damage to the industry without compromising on matters that affect public health.
56 DIET, NUTRITION, AND CANCER: DIRECTIONS FOR RESEARCH Taking all these factors into account, what research directions would be most useful? We need to know considerably more about the motives that lead people to change their food habits. How could this be studied? An example from another field might be illuminating. In the more than 20 years since the relationship between economic and social "disadvantage" and high rates of school failure was brought to public attention, researchers have looked valiantly for the factors in poverty that might account for its association with failure in school and in subsequent employment. Only recently has it been recognized that a more interesting question is not why so many individuals from disadvantaged environments fail, but why some individuals succeed. Research on consumer education in relation to diet might well profit from a similar emphasis on success, since there is evidence that many people have, in fact, changed what they eat in a direction that most health professionals would consider potentially beneficial. The public is anxious to be given dietary guidelines to avoid cancer, but scientists are anxious not to make promises prematurely. In considering this dilemma, one must recognize that (1) the public must make dietary choices every day; (2) these choices are being made from a food supply that Is changing rapidly; and (3) the best judgment of scientists at any point is certainly as good as--and very likely to be better than--what the public can get anywhere else. To do nothing in the way of helping the public translate the present state of knowl- edge into behavior is to do something, namely, to imply that nothing has been learned from several decades of research. Although the issues raised above are pertinent to many other areas of nutrition education research, the committee has limited its speci- fic recommendations to those that seem especially relevant to its charge to disseminate the findings resulting from this study. Within the past few years, several major conferences and workshops have addressed the topic of nutrition education and needed research in the field. Interested readers are advised to consult the publications reporting the work of those conferences for a fuller understanding of overall research issues (Brun, 1980; Dwyer, 1980; Olson and Gillespie, 1981; Sims and Light, 1982~. The following areas show particular promise as subjects for fruitful research. Many of them have also been recommended by other groups in other reports. RESEARCH RECOMMENDATIONS 0 We urgently need research that would begin to examine the be- haviors and motivations of persons who have already changed their diets--for example, in the directions suggested in the Dietary Guide- lines for Americans (U.S. Department of Agriculture and U.S. Department of Health, Education, and Welfare, 1980) or in this committee's first report. If we simultaneously study those who have not changed their diets, we may begin to understand the obstacles to change. Although
Consumer Behavior 57 individuals who undertake "spontaneous" change may differ from the population as a whole, finding out who they are, and when, why, how, and during what periods in their lives they changed their eating patterns, would help us begin to understand the factors that lead to long-term dietary change (Kasl, 1980~. · Research is needed to learn the "natural history" of diets con- sumed by humans. We know essentially nothing about which sorts of life occurrences--leaving home, getting married, having children, getting a new job, or losing a job--lead to substantial and lasting dietary change in the presence of a rapidly changing food supply and the pro- liferation of information in the media. Over the lifespans of most individuals, are there periods of vulnerability to change in food con- sumption when dietary education might be most effective? Moore et al. (1982) have recently suggested one potentially useful technique for such an investigation. · Suggestions are made in Chapter 4 of this report that more (and more frequent) monitoring of food intake and of patterns of food con- sumption in different groups will add to our understanding of the effects of different diets and of long-term dietary change. We need to collect data on a national basis about what people eat and how this varies in relation to such factors as geographical location, lifestyle, social affiliation, and other social factors as well as such convene tionally considered variables as ethnic and socioeconomic groups. The data should be collected in such a manner that they can also be analyzed for information about the history of an individual's lifetime food habits, including changes in those habits. o Existing bodies of longitudinal dietary data, such as those from the Harvard (Burke et al., 1959) and Colorado (Bear, 1967) growth stud- ies, should be examined for what they might reveal about patterns of food consumption over time and points at which diets change. Some longitudinal data have been examined in terns of nutrient intake (e.g., Valadian _ al., 1981) but not, as far as the committee has been able to determine, in terns of food intake patterns. · Almost nothing is known regarding the acquisition of food habits by children, with the exception of the findings from one 50-year-old study (Davis, 1928, 1934, 1939~. Consideration should be given to supporting additional long-term studies by researchers and/or observers to learn the factors affecting the acquisition of childhood eating patterns (see, e.g., Thomas et al., 1963~.
