Original Communication

European Journal of Clinical Nutrition (2003) 57, 249–259. doi:10.1038/sj.ejcn.1601549

Interaction of body mass index and attempt to lose weight in a national sample of US adults: association with reported food and nutrient intake, and biomarkers

A K Kant1

1Department of Family, Nutrition and Exercise Sciences, Queens College of the City University of New York, Flushing, New York, USA

Correspondence: A K Kant, Department of Family, Nutrition, and Exercise Sciences, Remsen Hall, Room 306E, Queens College of the City University of New York, Flushing, NY 11367, USA. E-mail: ashima_kant@qc.edu

Received 11 October 2001; Revised 29 May 2002; Accepted 4 June 2002.

Top

Abstract

Objective: This study examined the interaction between body mass index (BMI) and attempting to lose weight for reporting of: (1) macro- and micronutrient intake; (2) intake of low-nutrient-density foods; and (3) serum biomarkers of dietary exposure and cardiovascular disease risk.

Methods: Dietary, anthropometric and biochemical data were from the third National Health and Nutrition Examination Survey (1988–1994), n=13 095. Multiple regression methods were used to examine the independent associations of BMI, trying to lose weight, or the interaction of BMI–trying to lose weight with reported intakes of energy, nutrients, percentage energy from low-nutrient-density foods (sweeteners, baked and dairy desserts, visible fats and salty snacks), and serum concentrations of vitamins, carotenoids and lipids.

Results: BMI was an independent positive predictor (P<0.05) of percentage of energy from fat, saturated fat, but a negative predictor of the ratio of reported energy intake to estimated expenditure for basal needs (EI/BEE), percentage of energy from carbohydrate and alcohol (men only), and serum concentrations of folate, vitamin C, vitamin E and most carotenoids in both men and women. Trying to lose weight was a negative predictor (P<0.05) of EI/BEE, intake of energy, and energy density, but not micronutrient intake. Higher mean serum ascorbate, vitamin E, lutein/zeaxanthin, and other carotenoids (men only) concentrations were associated with trying to lose weight (P<0.05) in both men and women. Few adverse BMI-trying to lose weight interaction effects were noted.

Conclusion: There was little evidence of increased nutritional risk in those reportedly trying to lose weight irrespective of weight status.

Sponsorship: NIH research grant (R03 CA81604) and a USDA NRICGP award (NYR-9700611).

Keywords:

weight loss, dieting, junk food intake, low-nutrient-density foods, dietary patterns, serum vitamins, serum carotenoids, serum lipids, NHANES III

Top

Introduction

Increasing prevalence of obesity is a global public health problem, and weight control is recommended to alleviate the health risks associated with adiposity (WHO, 2000; National Institutes of Health, 1998). Not surprisingly, therefore, a large proportion of overweight individuals report attempting weight loss in national surveys (Serdula et al, 1999). However, a substantial number of non-overweight individuals also report attempts at weight control (Neumark-Sztainer et al, 2000). For instance, nearly 45% of subjects classified as non-overweight were attempting weight control, as were about 60% of moderately or very overweight subjects in a recently reported study (Neumark-Sztainer et al, 2000). The NIH Technology Assessment Conference Panel (1992) identified the need for understanding the consequences of attempting weight loss in individuals who are not overweight. It has been suggested that the perception of dieting behaviors and consequently nutrient intake of dieters may differ by weight status (Neumark-Sztainer et al, 1997). However, remarkably little is known about the effect of interaction of relative body weight and attempting weight control on food and nutrient profiles. It is hypothesized that nutritional risk profile of non-overweight respondents attempting weight control may differ from those of overweight respondents attempting weight control.

Examination of dietary and nutrient intake in relation to body weight and weight control efforts is complicated by the acknowledged problem of dietary misreporting resulting in biased estimates of self-reported intakes of energy and foods considered nutritional undesirables (Lissner et al, 2000). Therefore, it is important to include possible biomarkers of dietary exposure to validate self-reported intakes when examining nutritional health in relation to body weight and weight control behaviors. The purpose of this study was to examine, in a nationally representative sample of US adults, the relation of body mass index and attempting to lose weight with: (1) macro- and micronutrient intake; (2) intake of low-nutrient-density foods; and (3) serum biomarkers of dietary exposure and cardiovascular disease risk.

Top

Methods

This study used data from the third National Health and Nutrition Examination Survey (NHANES III), 1988–1994. The NHANES III is a multistage stratified probability sample of the non-institutionalized, civilian US population, aged 2 months and over (National Center for Health Statistics, 1994). The survey was conducted by the National Center for Health Statistics (NCHS), and included administration of a questionnaire at home and a full medical exam along with a battery of tests in a special mobile examination center (MEC). Demographic and medical history information was obtained during the household interview. The MEC exam included a physical and dental exam, dietary interview, body measurements and collection of blood and urine samples. Body weight, height and circumference at various body sites were measured using standardized procedures in the MEC. During the MEC exam respondents were asked a global question related to weight control, 'Are you currently trying to lose weight'?

Dietary assessment method

A 24 h dietary recall was collected by a trained dietary interviewer in a MEC interview using an automated, microcomputer-based interview and coding system (National Center for Health Statistics, 1994). The type and amount of foods consumed were recalled using aids such as abstract food models, special charts, measuring cups and rulers to help in quantifying the amounts consumed. Special probes were used to help recall commonly forgotten items such as condiments, accompaniments, fast foods and alcoholic beverages.

Analytic sample

All adults aged 20 y and over were eligible for inclusion (n=17 030) in this study. A complete and reliable dietary recall (as determined by NCHS) was not available for 1051 respondents, leaving 15 979 eligible for inclusion. We further excluded respondents stating that food intake on recalled day was 'much less' or 'much more' than usual (n=2245), women who were pregnant (n=282) or nursing (n=91), those missing information on body weight (n=33) or height (n=16), and those missing information on whether trying to lose weight at the time of the survey (n=305). Some respondents were in more than one exclusion category. The final analytic sample comprised 13 095 respondents (6298 men and 6797 women).

Assessment of intake of low-nutrient-density foods

Because of the prevailing hypothesis about misreporting of nutritional undesirables in relation to body weight, it was especially interesting to examine the intake of low-nutrient-density foods. To determine the intake of low-nutrient-density foods, it was necessary to identify foods belonging to this category from those reported in the 24 h dietary recall. As a first step, the 4265 foods reported by adult survey respondents were classified as belonging to one or more of the five major food groups (dairy, fruit, grain, meat and vegetable) or the low-nutrient-density foods group using methods we have described previously (Kant et al, 1991; Kant, 2000). Briefly, the assignment of foods into the various groups was dependent on their nutrient content and uses in the diet. The dairy group included milk, yogurt, cheese and buttermilk but excluded butter, cream cheese and dairy desserts. The fruit group included all fresh, frozen, dried and canned fruits and fruit juices but excluded fruit drinks. The grain group included all breads, cereals, pastas and rice, but excluded pastries. The meat group included meat, poultry, fish, eggs and meat alternates such as dried beans, nuts and seeds. The vegetable group included all raw or cooked fresh, frozen and canned vegetables and juices. Mixed dishes containing foods from several groups were grouped into all the relevant groups. Foods excluded from these major food groups were grouped into the low-nutrient-density foods group. The low-nutrient-density foods were further placed into four subgroups as follows: (1) visible fat—butter, oil, dressings, gravies, etc.; (2) sweeteners—sugar, syrup, candy, carbonated and non-carbonated sweetened drinks, etc.; (3) baked and dairy desserts—cookies, cakes, pies, pastries, ice cream, puddings, cheese cakes, etc.; and (4) salted snacks—potato, corn, tortilla chips, etc.

The NHANES III nutrient database for individual foods, which is derived from the US Department of Agriculture's Survey Nutrient Data base (National Center for Health Statistics, 1997), was used to determine energy and nutrient content of all foods. The nutrients examined included: vitamins A, E, B6, folate and C, and the minerals iron and calcium. The intake of each nutrient was compared with the respective age–sex-specific standard available. The standards used were 1989 Recommended Dietary Allowance (RDA) for protein, and Dietary Reference Intakes (DRIs) for vitamins A, E, C, B6 folate, iron, and calcium (Food and Nutrition Board, 1989, 1997, 1998, 2000, 2001).

As an estimate of dietary misreporting, a ratio of reported energy intake (EI) to energy expenditure for basal needs (BEE) was computed. BEE was estimated using age–sex–weight specific equations according to Schofield (1985).

Data on serum concentrations of vitamin C, folate, carotenoids and lipid were obtained from the NCHS public release compact disc (National Center for Health Statistics, 1996, 1998). The methods used to measure these serum analytes and their associated errors have been described in NCHS publications (National Center for Health Statistics, 1996, 1998). Folate, ascorbate, and the carotenoids—alpha-carotene, beta-carotene, beta-cryptoxanthin, lutein/zeaxanthin and lycopene—were chosen because the dietary intake of these nutrients is believed to be a determinant of serum concentration of the respective nutrient and thus can serve as biomarker of dietary exposure (Food and Nutrition Board, 1998, 2000). Serum lipids are known to be related to BMI and one reason for advocating weight loss is to improve serum lipid profiles (National Institutes of Health, 1998).

Statistical analyses

The mean daily energy, percent energy from macronutrients, percentage energy from low-nutrient-density food groups (separately and combined), mean percentage of the population meeting the RDA or DRI of selected nutrients, mean serum concentration of selected vitamers and analytes, adjusted for multiple covariates, were obtained by weighted tertiles of BMI and response to question about whether trying to lose weight at the time of the survey, separately for men and women. All covariates for inclusion in the various multiple regression models were decided a priori based on known relationships of dietary, socio-demographic, lifestyle and biological variables. The estimates of nutrient and food group intake were adjusted for age (continuous), race (non-Hispanic white, non-Hispanic black, Mexican-American, other), education (continuous), smoking status (never, former, current), and level of weekly recreational physical activity (none, 1–2 times/week, >2 times/week), whether changed diet for medical reasons in the past 12 months (yes, no), self-reported history of diabetes, hypertension or heart disease (yes, no). The estimates of mean serum analyte concentrations were adjusted for age (continuous), race (non-Hispanic white, non-Hispanic black, Mexican American, other), hours of fasting before phlebotomy, supplement use in the last 24 h before phlebotomy (yes, no, unknown), supplement use in the past month (yes, no), whether changed diet for medical reasons in the past 12 months (yes, no), self-reported history of diabetes, hypertension or heart disease (yes, no), dietary intake of folate, vitamin C, vitamin E and carotenoids for serum folate, vitamin C, vitamin E and carotenoids (all continuous), respectively, smoking status (vitamin C and carotenoids), and alcohol intake (HDL-cholesterol only), serum triglycerides and cholesterol (carotenoids and serum vitamin E). The procedure used to obtain covariate-adjusted estimates and standard errors from survey data is based on Taylor linearization methods according to Graubard and Korn (1999). All statistical analyses were performed using SAS (SAS Institute Inc., 2000), and software designed for analysis of survey data (SUDAAN; Shah et al, 1997). This software generates variance estimates that are corrected for multi-stage stratified probability design of complex surveys. Sample weights provided by the NCHS to correct for differential probabilities of selection, non-coverage, and non-response were used in all analyses to obtain point estimates (National Center for Health Statistics, 1994).

The independent association of BMI, and attempting weight loss or the interaction of BMI with attempting weight loss, with intake of foods, nutrients and serum analytes was examined using regression procedures to adjust for multiple covariates mentioned above. Linear regression procedures were used when the outcome variables were continuous (eg dietary nutrient intake, or serum nutrient concentration). For categorical outcomes such as whether or not standard of a nutrient intake was met, we used logistic regression procedures. In these regression models, BMI was entered as a continuous variable and attempting to lose weight as a categorical (yes, no) variable, along with a BMI–trying to lose weight interaction term. If the interaction term was non-significant (Pgreater than or equal to0.05), regression models were run again after excluding the interaction term from the model.

Top

Results

Sociodemographic profile

Table 1 lists the percentage of the population in categories of various sociodemographic variables by tertiles of BMI, by response to question about whether trying to lose weight at the time of the survey. Approximately 34% of the population answered yes to trying to lose weight. A higher proportion of less than or equal to50 y olds, and women in each tertile of BMI were trying to lose weight at the time of survey. The proportion of non-whites, those with <12 y of education, and no regular weekly physical activity increased with increasing BMI. A smaller proportion of those answering yes to trying to lose weight reported no regular weekly physical activity. In each BMI tertile, a higher proportion of those trying to lose weight had never smoked. The proportion reporting vitamin/mineral supplement use in the past month decreased with increasing BMI, but within each tertile of BMI, a higher proportion of those attempting weight loss reported supplement use. The proportion with self-reported health as excellent or very good decreased with increasing BMI, which was consistent with increase in self-reported history of diabetes, hypertension, or heart disease with increasing BMI. In each BMI tertile, nearly 75% of those trying to lose weight at the time of the survey had also tried weight control in the 12 months before the survey.


Energy and macronutrient intake

BMI and attempting weight loss interaction
 

In women, the difference in percentage energy from protein between those trying to lose weight and not trying to lose weight increased while differences in alcohol intake declined with increasing BMI (Table 2).


Independent effect of BMI
 

BMI was related positively with percentage energy from fat, saturated fat, but inversely with EI/BEE, percentage energy from carbohydrate and alcohol (men only).

Independent effect of trying to lose weight
 

Men and women trying to lose weight had lower EI/BEE, intake of energy, and energy density of all foods and beverages reported, and dietary fiber than those not trying to lose weight.

Reported intake of low-nutrient-density foods

BMI and attempting weight loss interaction
 

At higher BMI, men trying to lose weight reported lower percentage energy from all low-nutrient-density foods and desserts. Women trying to lose weight reported a higher Percentage energy from visible fats at higher BMI (Table 3).


Independent effect of trying to lose weight
 

Women trying to lose weight reported lower percentage energy from the sweeteners subgroup.

Percentage of the population meeting the standard of nutrient intake

BMI and attempting weight loss interaction
 

At higher BMI, a greater proportion of men attempting weight loss mentioned all five food groups on the survey day, but a smaller proportion of men reported the estimated average requirement (EAR) for iron (Table 4).


Independent effect of BMI
 

With increasing BMI, a smaller proportion of men and women reported <10% energy from saturated fat, RDA for protein, and EAR for vitamin C. Women at higher BMI were less likely to report all five food groups, less than or equal to30% energy from fat or EAR for folate. With increasing BMI, a higher proportion of men met the EAR for vitamin E.

Independent effect of trying to lose weight
 

A lower proportion of men and women attempting weight loss met the RDA for protein. A higher proportion of women trying to lose weight reported less than or equal to30% energy from fat, but a lower proportion met the EAR for vitamin E.

Serum analyte concentrations

Table 5 lists the multiple covariate adjusted meanplusminusSE of serum concentration of folate, vitamin C, vitamin E, alpha-carotene, beta-carotene, beta-cryptoxanthin, lutein/zeaxanthin, lycopene, total cholesterol, low-density-lipoprotein (LDL) cholesterol and high-density-lipoprotein (HDL) cholesterol, by tertiles of BMI, by response to question about whether trying to lose weight at the time of the survey, by sex.


BMI and attempting weight loss interaction
 

In men attempting weight loss, serum total and LDL-cholesterol were lower and HDL-cholesterol was higher at higher BMI.

Independent effect of BMI
 

BMI was a significant negative predictor of serum folate, vitamin C, vitamin E, alpha-carotene, beta-carotene, beta-cryptoxanthin, lutein/zeaxanthin and lycopene (women only) concentrations in both men and women. In women, serum concentrations of total and LDL-cholesterol were related positively and HDL-cholesterol were related inversely with BMI.

Independent effect of trying to lose weight
 

Trying to lose weight was an independent positive predictor of serum ascorbate, vitamin E and lutein/zeaxanthin concentrations in both men and women. In men, trying to lose weight was a positive predictor of RBC folate, serum alpha-carotene, beta-carotene and beta-cryptoxanthin concentrations.

Top

Discussion

The interaction effects of BMI and trying to lose weight were generally small, with a slightly higher proportion of men trying to lose weight at higher BMI reporting all five food groups but a lower mean percentage of energy from low-nutrient-density foods. To my knowledge, other studies where food and nutrient intake profiles of those attempting weight loss in relation to body weight have been examined are not available for comparison. Energy intake is the only variable for which a weight status–trying to lose weight interaction has been reported previously. Neumark-Sztainer reported no differences in the energy intake reported by non-overweight current dieters vs non-dieters, but energy intake of overweight current dieters was lower than that of non-dieters (Neumark-Sztainer et al, 1997). Such a finding suggests that weight status modified the relation of energy intake with dieting status. This interaction was not observed in the present study. Notably, however the question used to elicit weight loss behavior as well as the method of dietary assessment differed in the two studies.

The observations of lower reported energy intake, energy density, and EI/BEE ratio in those trying to lose weight are consistent with the expected changes in food selection with weight loss attempts (Andersson et al, 2000). However, the extent to which these results reflect reporting bias vs actual dietary behaviors cannot be ascertained without independent validation of dietary intakes. With the exception of an increased risk of not meeting the weight-based protein standard (0.8 g/kg body weight), attempting weight loss was not associated with a significantly increased risk of inadequate intake of the nutrients examined despite lower energy intake (except vitamin E in women and iron in men at higher BMI). A trend for a lower proportion of those attempting weight loss meeting the calcium standard was noted but did not reach statistical significance. Ritt et al (1979) also identified iron and calcium as nutrients at risk in diets used for weight control. More recently, vitamin E, calcium, iron and zinc were identified as nutrients that were consumed at lower levels from a well-planned, energy-restricted diet (Benzera et al, 2001). Conversely, Neumark-Sztainer et al (1996) reported higher intakes of many micronutrients estimated from a 60-item food frequency questionnaire in adult dieters relative to non-dieters.

There are few published reports where changes in serum vitamer status of subjects trying to control weight have been investigated to enable comparison with results of the present study. Notably, trying to lose weight was a positive independent predictor of serum ascorbate, vitamin E and most carotenoids in both men and women trying to lose weight, which supports the notion of better dietary selections with attempting weight loss. Also, at higher BMI, there was a suggestion of improvement in serum lipid profiles in men trying to lose weight.

Although BMI was not a significant predictor of total energy intake, it was a strong inverse predictor of EI/BEE independent of trying to lose weight status, confirming that energy reporting relative to requirement for basal expenditure was lower at higher BMI (Lissner et al, 2000). BMI was not an independent predictor of percentage of energy intake from all low-nutrient-density foods combined or most low-nutrient-density subgroups in women; in men low-nutrient-density food reporting showed an interaction of BMI with trying to lose weight (Table 3). These results do not support the suggestion that BMI, especially in women, was associated with differential reporting of nutritionally undesirable low-nutrient-density foods as a proportion of the total energy intake. These results differ from other reports where BMI was an independent negative predictor of intake of sugar, fat or other low-nutrient-density foods (Slattery et al, 1992; Drewnowski et al, 1997a; Macdiarmid et al, 1998). Consistent with other reports, in the present study, BMI was an independent positive predictor of percentage of energy from total fat, but negative predictor of percentage of energy from carbohydrate and alcohol (men only; Miller et al, 1990,1994; Slattery et al, 1992; Ortega et al, 1995; Lissner et al, 1998).

Increased prevalence of low serum folate in overweight relative to non-overweight subjects and the negative association of BMI with plasma ascorbate and beta-carotene noted in this study has been previously reported (Hotzel 1986; Moor de Burgos et al, 1992; Hebert et al, 1994; Drewnowski et al, 1997b; Wallstrom et al, 2001). Published reports on the relation of serum vitamin E and BMI are inconsistent. While the findings of Moor de Burgos et al (1992) are in accord with our results, Wallstrom et al (2001) reported a positive association of cholesterol adjusted BMI with serum tocopherol in men and no relation in women. Little has been published on possible explanations for the BMI effect on circulating vitamers and carotenoids. Putative reasons may include increased utilization, dilution, redistribution, or tissue sequestration of these nutrients (Hebert et al, 1994). Recent data on vitamin D metabolism in association with adiposity suggest that tissue sequestration of fat soluble analytes may partially explain lower circulating levels (Wortsman et al, 2000).

The strengths of this study include a large, nationally representative sample of US adults for whom information on most major confounders of BMI–diet association, along with objective markers of nutrient intake, was available. Limitations that warrant a cautious interpretation of results of this study include the following. The estimates of food and nutrient intake for men in the first tertile of BMI who answered yes to attempting weight loss may be somewhat unreliable due to small numbers. The question 'Are you currently trying to lose weight?' does not specifically probe how the weight loss is being attempted or the duration or efficacy of these attempts. A positive response to the question probably includes individuals using non-dietary strategies which will probably attenuate the relation of this variable with dietary and biochemical outcomes. Although it is notable that consistent with expectation, energy intake and energy density in those attempting weight loss was lower. Also, in a recent survey, over 90% of those reporting attempts at weight control included dietary strategies (Serdula et al, 1999). It is also possible that some individuals, especially at higher BMI, replied affirmatively to trying to lose weight due to their perception of it being the expected or desired response. Thus it is likely that the reliability of this response for classifying individuals as whether or not weight loss was being attempted differs by BMI status. One can nevertheless argue that a positive response to this question may at least be an indicator for concern about body weight, which may in turn relate to dietary behaviors as is evident from the data presented.

All methods of dietary assessment, including the 24 h recall used in NHANES III, have several inherent measurement errors (Bingham, 1987). Estimates from 24 h dietary recall are generally reliable for assessing groups, but not individuals. Accordingly, in the present analyses, the conclusions are limited to groups. Nevertheless, because the estimates presented are not adjusted for within-subject variability in reported food intake, measurement error may have contributed to the inability to find significant differences. The results presented in Table 4 warrant cautious interpretation as prevalence of inadequacy of micronutrients from a single dietary measurement tends to be overestimated. Thus, it is preferable to interpret this table as indicating trends rather than prevalence estimates. In the present study, dietary ascorbate, folate, vitamin E and carotenoid intake was a significant independent predictor of serum concentration of the respective nutrients, providing an independent validation of the estimated dietary intake.

These results suggest that for most examined food, nutrient, and biomarker outcomes nutritional health was not compromised in those reportedly trying to lose weight irrespective of weight status. On the average, reporting of lower intake of energy, and energy density in those concerned about their weight was not associated with lower micronutrient intake irrespective of the degree of adiposity, suggesting a more judicious pattern of food selection. If the differences in dietary behaviors of those concerned about body weight are not largely due to reporting bias, these behaviors may help in decreasing the risk of subsequent major weight gain in groups of varying BMI. To my knowledge, no prospective studies of such groups have been published. These results provide little evidence for discouraging attempts to lose weight on nutritional grounds. However, this study did not focus on psychological or other physical consequences, if any, of adopting reported dietary behaviors by those concerned about body weight. Further work is needed on evaluation of nutritional status using biomarkers in relation to BMI, changes in biomarkers and variation in nutritional risk due to different strategies for weight management and repeated attempts at weight control.

Top

References

  1. Andersson, I, Lennernas, M & Rossner, S ((2000)). Meal pattern and risk factor evaluation in one-year completers of a weight reduction program for obese men—the 'Gustaf' study. J. Intern. Med., 247, 30–38.
  2. Benzera, LM, Nieman, DC & Nieman, CM et al ((2001)). Intakes of most nutrients remain at acceptable levels during a weight management program using the food exchange system. J. Am. Diet. Assoc., 101, 554–558, 561.
  3. Bingham, SA ((1987)). The dietary assessment of individuals: methods, accuracy, new techniques and recommendations. Nutr. Abstr. Rev. (Ser. A), 705–742.
  4. Drewnowski, A, Henderson, SA & Shore, BA et al ((1997a)). The fat–sucrose seesaw in relation to age and dietary variety of French adults. Obes. Res., 5, 511–518.
  5. Drewnowski, A, Rock, CL & Henderson, SA et al ((1997b)). Serum beta carotene and vitamin E as biomarkers of vegetable and fruit intake in a community-based sample of French adults. Am. J. Clin. Nutr., 65, 1796–1802.
  6. Food and Nutrition Board ((1989)). National Research Council Recommended Dietary Allowances, 10th edn.Washington, DC: National Academy Press
  7. Food and Nutrition Board ((1997)). Dietary Reference Intakes for Calcium, Phosphorus, Magnesium, Vitamin D, and Fluoride, Washington, DC: National Academy Press
  8. Food and Nutrition Board ((1998)). Dietary Reference Intakes for Thiamin Riboflavin, Vitamin B-6, Folate, Vitamin B-12, Pantothenic acid, Biotin, and Choline, Washington, DC: National Academy Press
  9. Food and Nutrition Board ((2000)). Dietary Reference Intakes for Vitamin C, Vitamin E, Selenium, and Carotenoids, Washington, DC: National Academy Press
  10. Food and Nutrition Board ((2001)). Dietary Reference Intakes for Vitamin A, Vitamin K, Arsenic, Boron, Chromium, Copper, Iodine, Iron, Molybdenum, Nickel, Silicon, Vanadium, and Zinc, Washington, DC: National Academy Press
  11. Graubard, BI & Korn, EL ((1999)). Predictive margins with survey data. Biometrics, 55, 652–659.
  12. Hebert, JR, Hurley, TG & Hsieh, J et al ((1994)). Determinants of plasma vitamins and lipids: the working well study. Am. J. Epidemiol., 140, 132–147.
  13. Hotzel, D ((1986)). Suboptimal nutritional status in obesity (selected nutrients). Biblithica Nutr. Dieta, 37, 36–41.
  14. Kant, AK ((2000)). Consumption of energy-dense, nutrient-poor foods in the US population: effect on nutrient profiles. Am. J. Clin. Nutr., 72, 929–936.
  15. Kant, AK, Schatzkin, A, Block, G, Ziegler, RG & Nestle, M ((1991)). Food group intake patterns and associated nutrient profiles of the US population. J. Am. Diet. Assoc., 91, 1532–1537.
  16. Lissner, L, Lindroos, AK & Sjostrom, L ((1998)). Swedish subjects (SOS): an obesity intervention study with a nutritional perspective. Eur. J. Clin. Nutr., 52, 316–322.
  17. Lissner, L, Heitmann, BL & Bengtsson, C ((2000)). Population studies of diet and obesity. Br. J. Nutr., 83, S21–S24.
  18. Macdiarmid, JI, Vail, A & Blundell, JE ((1998)). The sugar–fat relationship revisited: differences in consumption between men and women of varying BMI. Int. J. Obes. Res., 22, 1053–1061.
  19. Miller, WC, Lindeman, AK, Wallace, J & Niederpruem, M ((1990)). Diet composition, energy intake, and exercise in relation to body fat in men and women. Am. J. Clin. Nutr., 52, 426–430.
  20. Miller, WC, Niederpruem, MG, Wallace, JP & Lindeman, AK ((1994)). Dietary fat, sugar, and fiber predict body fat content. J. Am. Diet. Assoc., 94, 612–615.
  21. Moor de Burgos, A, Wartanowicz, M & Ziemlanski, S ((1992)). Blood vitamin and lipid levels in overweight and obese women. Eur. J. Clin. Nutr., 46, 803–808.
  22. National Center for Health Statistics ((1994)). Plan and operation of the Third National Health and Nutrition Examination Survey, 1988–94. Vital Health Stat., 1, (32)
  23. National Center for Health Statistics ((1996)). Third National Health and Nutrition Examination Survey, 1988–1994. NHANES III Laboratory Data File, (CD-ROM, Series 11, no. 1A).Hyattsville, MD: Centers for Disease Control and Prevention
  24. National Center for Health Statistics ((1997)). Third National Health and Nutrition Examination Survey, l988–1994. NHANES III, (CD-ROM, Series 11, no. 1A).Hyattsville, MD: Centers for Disease Control and Prevention
  25. National Center for Health Statistics ((1998)). Third National Health and Nutrition Examination Survey, 1988–1994. NHANES III Second Laboratory Data File, (CD-ROM. Series 11, no. 2A).Hyattsville, MD: Centers for Control and Prevention
  26. National Institutes of Health ((1998)). Clinical guidelines on the identification, evaluation, and treatment of overweight and obesity in adults: the evidence reports. Obes. Res., 6, (Suppl 2)51S–209S.
  27. Neumark-Sztainer, D, French, SA & Jeffrey, RW ((1996)). Dieting for weight loss: Associations with nutrient intake among women. J. Am. Diet. Assoc., 96, 1172–1175.
  28. Neumark-Sztainer, D, Jeffery, RW & French, SA ((1997)). Self-reported dieting: How should we ask? What does it mean? Associations between dieting and reported energy intake. Int. J. Eat. Disord., 22, 437–449.
  29. Neumark-Sztainer, D, Rock, CL & Thornquist, MD et al ((2000)). Weight-control behaviors among adults and adolescents: associations with dietary intake. Prev. Med., 30, 381–391.
  30. NIH Technology Assessment Conference Panel ((1992)). Methods for voluntary weight loss and control. Ann. Intern. Med., 116, 942–949.
  31. Ortega, RM, Redondo, MR, Zamora, MJ, Lopez-Sobaler, AM & Andres, P ((1995)). Eating behavior and energy and nutrient intake in overweight/obese and normal weight Spanish elderly. Ann. Nutr. Metab., 39, 371–378.
  32. Ritt, RS, Jordan, HA & Levitz, LS ((1979)). Changes in nutrient intake during a behavioral weight control program. J. Am. Diet. Assoc., 74, 326–330.
  33. SAS Institute Inc. ((2000)). SAS User's Guide. Release 8, Cary, NC: SAS Institute Inc
  34. Schofield, WN ((1985)). Predicting basal metabolic rate, new standards and review of previous work. Hum. Nutr. Clin. Nutr., 39C, 5S–41S.
  35. Serdula, MK, Mokdad, AH, Williamson, DF, Galuska, DA, Mendelein, JM & Heath, GW ((1999)). Prevalence of attempting weight loss and strategies for controlling weight. JAMA, 282, 1353–1358.
  36. Shah, BV, Barnwell, BG & Bieler, GS ((1997)). SUDAAN User's Manual. Release 7.5, Research Triangle Park, NC: Research Triangle Institute
  37. Slattery, ML, McDonald, A & Bild, DE et al ((1992)). Associations of body fat and its distribution with dietary intake, physical activity, alcohol, and smoking in blacks and whites. Am. J. Clin. Nutr., 55, 943–994.
  38. Wallstrom, P, Wirfalt, E, Lahmann, PH, Gullberg, B, Janzon, L & Berglund, G ((2001)). Serum concentrations of beta-carotene and alpha-tocopherol are associated with diet, smoking, and general and central adiposity. Am. J. Clin. Nutr., 73, 777–785.
  39. WHO ((2000)). Obesity: Preventing and Managing the Global Epidemic, WHO Technical Report Series 10.894Geneva: WHO
  40. Wortsman, J, Matsuoko, LY, Chen, TC, Lu, Z & Holick, MF ((2000)). Decreased bioavailability of vitamin D in obesity. Am. J. Clin. Nutr., 72, 690–693.
Top

Acknowledgements

I thank Daniel F Heitjan, Columbia University, School of Public Health, New York, for statistical consultation, and Lisa L Kahle for expert programming assistance.

Extra navigation

.

natureevents

ADVERTISEMENT