Weight Control Practices of U.S. Adults Trying to Lose Weight

Methods

Data for the WLPS were collected in the continental United States from September to mid-November 1991. Respondents were noninstitutionalized, civilian adults, 18 years or older, who were contacted and interviewed by telephone. A two-stage sampling procedure was used to select the respondents to be interviewed [6]. The first stage involved the selection of a sample of telephone numbers through random-digit dialing procedures. In the second stage, a designated respondent was selected randomly from all adults in the household. The designated respondent was asked, Are you right now trying to lose weight? Respondents who answered yes continued the interview.

Of the 10 840 households contacted, 72% completed the interview through the weight-loss screening question. Of those designated respondents who said that they were attempting to lose weight, 95% completed the WLPS interview. The total sample of 1431 persons trying to lose weight consisted of 1030 women and 401 men.

Our analysis estimated the level or prevalence of weight-loss variables within each of 12 population subgroups defined by sex and three classifications of BMI (weight [kg]/height [m²]) and three classifications of regimen characteristics.

Self-reported height and weight were used to calculate BMI. Respondents were assigned to one of three categories (BMI < 26, 26 to < 30, and 30). We classified persons with a BMI between 26 and 30 as overweight and those with a BMI of 30 or more as obese [7]. Because BMI was based on self-reported weights, overweight and obesity were probably underestimated. Persons consistently under-report their true weights by approximately 1 kg; however, correlations between reported weights and measured weights are typically quite high, ranging between r = 0.96 and r = 0.99 [8].

The regimen characteristics used for classification were as follows: whether respondents had combined dieting (defined as eating differently from the way you usually eat for the sake of losing weight) and exercise (as called for by most weight-loss guidelines) [7, 9]; whether respondents had been on their current weight plan for at least 1 year (an indication of a long-term commitment); and whether respondents engaged in one or more practices that might be considered questionable in some circumstances. Practices were classified as questionable based on the possibility of abuse. Questionable practices included fasting for 24 hours or longer, taking diet supplements such as high-protein or fiber supplements, using any type of diet-aid pill (for example, appetite suppressants, diuretics, or thyroid pills), vomiting after eating, taking laxatives, or using special weight-loss devices. Such practices have been associated with health problems [10] but could be considered acceptable elements of a weight-loss plan if used judiciously or while under professional supervision. For example, although it is acceptable to use diet supplements to supplement a healthy diet, it is unacceptable to use such supplements as a substitute for food to lose weight. The regimen classifications described here are not mutually exclusive and any possible combination of the three characteristics could describe an individual weight-loss plan.

All survey estimates were weighted to the 1990 census distribution on sex, race, age, and education. Because of item nonresponse, the number of respondents varied slightly between analyses. Respondents were excluded only from those analyses for which they failed to answer the relevant items.

Results

Characteristics of Weight-Loss Practitioners

The population prevalence of active weight-loss attempts was calculated by dividing the number of persons who reported attempting to lose weight by the total number of designated respondents who were contacted. Overall, 33% of women and 20% of men reported that they were currently trying to lose weight. The relative prevalence of weight-loss plans by BMI group and regimen classification variables are shown in Table 1. More than two thirds of weight-loss practitioners combined diet and exercise in their weight-loss plan, 30% had been on the current plan for 1 year or more, and 20% engaged in some kind of questionable weight-loss practice. Table 1 also shows the distribution of weight-loss plans by demographic characteristics and regimen features. Compared with the general population, persons trying to lose weight were more likely to be women, young to middle-aged, and better educated. Most men were in one of the two higher BMI categories. In contrast, about one half of women were in the lowest BMI category; that is, a substantial percentage of women who are trying to lose weight may not be overweight.

Younger persons were more likely to include both diet and exercise in their regimens but were less likely than older respondents to be on long-term regimens. More educated practitioners were more likely to be nonoverweight as determined by BMI. Weight-loss practitioners with more education were also more likely to combine diet and exercise in their weight-loss plans and were less likely to engage in questionable practices. Blacks who reported trying to lose weight were much more likely to be overweight and to engage in questionable practices than were whites.

Previous Weight-Loss Experience Characteristics

Respondents reported they had spent an average of 6 months of the past year trying to lose weight, had made an average of more than two weight-loss attempts during the past 2 years, and had tried their current weight-loss plan once before. Women reported more previous weight-loss attempts and more time spent trying to lose weight than did men. Tables 2 and 3 show the means for these and other survey variables by BMI groups and regimen features for women and men, respectively.

Duration of Current Weight-Loss Plan

The average duration of the current weight-loss attempt was between 5 and 6 months for both women and men. Duration of the current weight-loss attempt decreased dramatically as BMI level increased for both women and men. Weight-loss practitioners with higher BMI levels engaged in more frequent weight-loss attempts of shorter duration than did those with lower BMI levels, who tended to remain in a given plan for longer periods.

Weight-Loss Goals

Respondents reported how much weight they had lost with their current plan and how much more weight they hoped to lose. The sum of the two responses defined their weight-loss goal. Weight-loss goals were strongly related to weight status. For both women and men, as BMI levels increased, so did the amount of weight they hoped to lose. Weight-loss goals ranged from around 10 kg for those with a BMI of less than 26 to an average of almost 30 kg for those with a BMI of 30 or more. In contrast to the threefold variation in weight-loss goals, the variation in achieved weight-loss between BMI groups was modest, averaging less than 2 kg for women and less than 3 kg for men. The weight-loss goals of persons at the higher BMI levels appeared unrealistic, given the amounts of weight that they had lost so far.

Information-seeking Behavior

Approximately 40% of those engaged in a weight-loss attempt said that they had checked with their doctor before trying to lose weight. For both women and men, the likelihood of checking with a physician increased with BMI. Less than one third of respondents reported looking for information about how to lose weight from expert sources such as physicians, nutritionists, other health professionals, or specialized reading. Most respondents obtained information from informal sources such as friends, family members, television, newspapers, or popular magazines. For both women and men, the likelihood of seeking expert information increased with BMI.

Reasons for Trying to Lose Weight

Respondents were asked about their reasons for trying to lose weight; if they mentioned more than one, they were asked which reason was the most important. Respondents were asked about current health problems, health problems they wanted to avoid, general fitness, and appearance. In addition, they were asked about two possible reasons for weight gain that may have precipitated their weight-loss attempt, such as a recent pregnancy or recent smoking cessation. Tables 2 and 3 show the motivation results. Men were more likely than women to say that future health and general fitness were the most important reasons for trying to lose weight. Women were more likely than men to report that appearance was the most important reason for trying to lose weight.

Levels of BMI were strongly related to the type of reason given by respondents for trying to lose weight. Among respondents with BMIs of less than 26, health reasons (present and future) were cited as most important by 22% of women and by 32% of men. For respondents with BMIs of 30 or more, health reasons were cited as most important by more than 50% of women and by 60% of men. Appearance and fitness were complementary to health reasons and showed the opposite relation to BMI.

Smoking cessation was cited as the reason for trying to lose weight by 6% of women and by 10% of men. Approximately 10% of women cited a recent pregnancy as their reason for trying to lose weight.

Prevalence of Specific Practices

Tables 2 and 3 give the prevalence by sex for 12 weight-loss practices. Dieting and exercise were easily the most prevalent types of practices used by persons trying to lose weight. The use of vitamins or minerals as part of a weight-loss plan was the next most common practice, followed by the use of meal replacements, weight-loss pills (such as appetite suppressants, diuretics, or thyroid pills), and formal weight-loss programs (such as those available commercially or sponsored by hospitals, clinics, or individual physicians). Relatively infrequent weight-loss practices were combined into a single category that included fasting for 24 hours or more, taking laxatives, using weight-loss devices such as body wraps, having surgery, or vomiting after eating. One or more of these practices was used by 10% of both women and men, with fasting accounting for two thirds of the instances. The use of diet supplements, such as fiber or protein supplements, as part of a weight-loss regimen was also an infrequent practice. Women were more likely than men to engage in almost all of the practices listed.

Exercise was the only practice that decreased in prevalence as BMI increased. Participating in organized weight-loss programs, using weight-loss pills or diet supplements, fasting, and taking laxatives all tended to increase in prevalence with increasing BMI. The prevalence of taking vitamins or minerals and using meal replacements was relatively similar across BMI groups.

Prevalence of Specific Dieting Behaviors

The most common behavior associated with dieting undertaken by both women and men was regular self-weighing, that is, at least once a week. Counting calories and skipping meals were the most common food-related behaviors. Fewer than 15% of weight-loss practitioners used menus, ate more frequent meals, or kept a record of food intake.

The use of special diet food products was widespread. On average, weight-loss practitioners reported using three or four such products. The most widely used diet food products were diet soft drinks, low-calorie dressings, and low-calorie sweeteners. Women were more likely than men to use each kind of diet food, with two exceptions. Men were more likely to use light alcohol beverages, and men and women were equally likely to use low-fat ice cream products. A greater variety of diet food products was used by persons in high BMI groups than by those in low BMI groups and by persons who engaged in both diet and exercise behaviors than by those who did not.

Prevalence of Exercise Behaviors

Persons attempting weight-loss reported spending an average of approximately 3 hours per week exercising. Walking was by far the most common type of exercise behavior (58% of women and 44% of men). Women were more likely than men to engage in both walking and aerobic exercise, whereas men were more likely than women to engage in weight training. Approximately 13% of weight-loss practitioners attended some form of exercise class. Among men, a higher BMI was associated with less time spent exercising. Among women, however, a curvilinear relation existed; those in the highest BMI group reported exercising more per week than those in the middle BMI group.

Discussion

The WLPS was designed to provide an empiric description of the distribution of weight-loss regimens in the population that might help to identify opportunities for improved consumer education and consumer protection strategies directed toward persons who are trying to lose weight. The results showed a great range of behavioral variation in individual approaches to weight-loss. A significant number of persons who reported trying to lose weight had the least need to do so in terms of BMI. Two thirds of those trying to lose weight reported combining dieting and exercise in their weight-loss regimens, with the average person reporting multiple diet and exercise behaviors. This finding represents a substantial increase over those of the 1985 National Health Interview Study (NHIS), which reported that fewer than one half of those persons trying to lose weight combined dieting and exercise in their weight-loss regimens [3]. This higher estimate may be due in part to differing study methods. It appears that weight-loss practitioners are also translating their awareness that dieting and exercise are the two best ways to reduce weight into action (an awareness shown by the 1985 NHIS). In addition, most of those trying to lose weight reported engaging in at least one other practice. Of these, the use of vitamins or minerals and the use of commercially available meal replacements were the next two most commonly reported practices. However, approximately 1 in every 10 persons trying to lose weight was taking weight-loss pills such as appetite suppressants or diuretics, and 1 in 10 was engaging in behaviors such as fasting for at least 24 hours, taking laxatives, vomiting after eating, or taking specialized protein or fiber supplements. The persons most likely to engage in these less frequent behaviors were the least educated and the most overweight. Although published estimates of the prevalence of these questionable practices among adults are scarce, recent evidence indicates that they are disturbingly common among adolescents [11].

A major theme of the WLPS findings is the recurrent nature of weight-loss attempts and the amount of time devoted to such attempts. The average weight-loss practitioner had been working on a current attempt for almost 6 months, had tried a similar kind of plan at least once before, and had averaged one attempt per year for the past 2 years. In addition, approximately 30% of persons trying to lose weight were chronic dieters and had been on their current weight-loss plan for at least 1 year. The 1989 Behavioral Risk Factor Surveillance Survey found a similar prevalence of chronic dieters among weight-loss practitioners and, like the WLPS, found the prevalence of chronic dieters to increase with age [4]. Given the quantity of experience available to persons who are trying to lose weight, it is disturbing that the likelihood of combining diet and exercise in one's weight-loss regimen decreased with age, whereas the likelihood of engaging in questionable practices increased with age. The weight-loss practice most strongly associated with increasing age, however, is the likelihood of engaging in a weight-loss regimen that is effectively a permanent part of one's lifestyle.

The WLPS estimates of prevalence of weight-loss attempts are somewhat lower than those reported in some recent surveys [3, 4]. At least two possible factors can account for this finding. First, the WLPS is probably more sensitive to seasonal variations in weight-loss behavior than are other weight-loss inventories because the data collection period was shorter. The September-to-November period used to collect the WLPS data is probably not a peak season for weight-loss attempts [12]. In addition, the WLPS was introduced to potential respondents as a weight-loss survey, which may have led respondents who were reluctant to discuss weight-loss attempts to refuse to participate at the screening stage of the interview.