Objective:

The fat content of a diet has been shown to affect total energy intake, but controlled feeding trials have only compared very high (40% of total calories) fat diets with very low (20% of total calories) fat diets. This study was designed to measure accurately the voluntary food and energy intake over a range of typical intake for dietary fat.

Methods and Procedures:

Twenty-two non-obese subjects were studied for 4 days on each of three diets, which included core foods designed to contain 26, 34, and 40% fat, respectively of total calories and ad lib buffet foods of similar fat content. All diets were matched for determinants of energy density except dietary fat. Subjects consumed two meals/day in an inpatient unit and were provided the third meal and snack foods while on each diet. All food provided and not eaten was measured by research staff.

Results:

Voluntary energy intake increased significantly as dietary fat content increased (P = 0.008). On the 26% dietary fat treatment, subjects consumed 23.8% dietary fat (core and ad lib foods combined) and 2,748 741 kcal/day (mean s.d.); at 34% dietary fat, subjects consumed 32.7% fat and 2,983 886 kcal/day; and at 40% dietary fat subjects consumed 38.1% fat and 3,018 963 kcal/day.

Discussion:

These results show that energy intake increases as dietary fat content increases across the usual range of dietary fat consumed in the United States. Even small reductions in dietary fat could help in lowering total energy intake and reducing weight gain in the population.

Introduction

The public is confused about how to eat in order to maintain a healthy body weight and in particular about how much fat vs. carbohydrate to include in a healthy diet. The popularity of low-fat diets in the 1970s and 1980s (1) was replaced by the popularity of low carbohydrate diets following the publication of Atkins' New Diet Revolution (2) in 1992.

While many popular diet books are based on advocating a specific diet composition for weight loss, it makes more sense to base population diet composition recommendations on the ability to help maintain weight and avoid weight gain or, for people who have lost weight, to avoid regaining weight. In this sense, the best diet for weight management would be one that is feasible to maintain over time and that minimizes the chances of positive energy balance.

One of the most significant influences of diet composition is on voluntary energy intake. There is a very consistent finding that when given ad libitum access to food, subjects eat more total energy when the fat content of the diet is 40% vs. 20% (3,4,5). This may be because of the higher energy density of high-fat diets (6,7,8,9).

There are no controlled feeding trials, using weighed food measures, that have examined the systematic impact of dietary fat level on energy intake across the range of dietary fat intake normally consumed by Americans (10,11,12). There are many intervention trials that have varied the dietary fat level in outpatient settings where food intake was measured by self-reporting (13,14,15) that suggest that the level of dietary fat can affect food intake. However, dietary self reports may not provide accurate information about energy intake (16). Further, it is not clear whether the effect of dietary fat on energy intake is due solely to its higher energy density or whether it has impact independent of energy density. This study was undertaken to investigate further the impact of diet composition on ad libitum energy intake, focusing particularly on the typical range of free-living western diets.

Methods and Procedures

Subjects

Twenty-two healthy, non-smoking, men (n = 15) and women (n = 7) with a BMI between 21 and 28 kg/m² were enrolled in the study. Individuals between the ages of 25 and 40 years were recruited by written advertisements which were distributed throughout the University of Colorado Health Sciences Center and the surrounding community.

Protocol

The research study was approved by the Institutional Review Board at the University of Colorado Health Sciences Center and the Scientific Advisory Committee of the General Clinical Research Center. After providing informed consent, interested participants completed a series of questionnaires to ensure that they met eligibility criteria. The questionnaires used were: menstrual cycle questionnaire, Eating Inventory (17), The Center for Epidemiological Studies—Depression Scale (18) and Eating Attitude Test (EATS-26) (19). We eliminated any subjects who displayed high levels of dietary restraint (>10 for women and >8 for men on the restraint subscale of the Eating Inventory), and those who reported high depression scores (>20 on the Center for Epidemiological Studies—Depression Scale questionnaire). All women participated in the feeding protocols during the early follicular phase of the menstrual cycle to control for any potential interaction between appetite and menstrual cycle phase.

Subjects who met the eligibility criteria were scheduled for a screening visit where they were required to return 3-day food diaries and 7-day physical activity diaries. Food records were reviewed by a registered dietitian. Subjects were excluded if their typical dietary fat intake was <25% or >35% of total energy intake. Subjects were excluded if they consumed more than one drink/day of alcohol or had one episode of consuming three or more drinks of alcohol during the 4 days of recording food intake. They were also excluded if they consumed more than 24 oz/day of sugar-containing soda or if they did not routinely consume breakfast (they should have eaten breakfast at least 3 of the 4 days of recording food intake).

Following screening, participants were randomly assigned to one of six cohorts, each containing four subjects. In a randomized, crossover design, each cohort consumed the controlled diets for 4 consecutive days for each of three fat levels: 26, 34, and 40%, respectively of total calories. The order of fat level was randomly determined for each subject. The diets were developed to have similar levels of other determinants of energy density, namely water and fiber. Thus, differences in energy density of the diets were due only to dietary fat content. There was a 3-week washout period between dietary fat levels. Subjects were not aware of the composition of the diets they were consuming.

Each participant was asked to initially consume half of his/her estimated energy needs for each meal in a "core" meal. In preliminary work, we found that this method allowed for achieving the desired level of diet composition while still providing subjects flexibility in total calories consumed. Estimated energy needs were determined using the following formula: Estimated energy needs = 1.5 (372 + 23.9 fat-free mass). This core meal consisted of foods containing 15% protein, the target level of fat and the remainder from carbohydrate. The energy level was held constant for a subject across all three diets. After the participants consumed their core meal, they were offered an individual buffet containing several foods with a fat content within 10% of the target for the diet. The amount and macronutrient composition of ad libitum food consumed at each meal as well as during two optional snacks/day were measured by research staff. The research staff controlled the number of dishes available and ensured that offerings included both sweet and savory foods. In most cases the same foods (with different levels of fat) were used for each diet condition. For example, yogurt with varying fat levels was provided with each diet. Individuals consumed their breakfast and dinner with others in their cohort at the General Clinical Research Center dining room. They consumed lunch and snacks on their own without supervision. Subjects were queried each morning about any food eaten on their own.

Lunch and snacks were pre-weighed and prepared for subjects each day. Subjects received lunch and snacks after breakfast. All wrappers, containers and uneaten food were returned to the General Clinical Research Center to be weighed. At that time, subjects were asked about their consumption of lunch items and whether anyone else ate any of the food. Energy per gram information on the nutrition label of packaged food and ProNutrasoftware (Princeton, NJ) were used to calculate energy and macronutrient intake.

Throughout the study, participants were asked to wear a pedometer (Accusplit AE120; Accusplit, San Jose, CA), to determine physical activity through walking. It is possible that diet composition could affect the usual amount of physical activity. Each subject recorded total steps/day throughout the study. Subjects were given a hand-held computer on which to rate the palatability of the diet following each meal. Visual analogue scales (100 mm) were used for the ratings (21).

Body composition

Body composition was determined for all subjects using a Hologic (Bedford, MA) dual energy X-ray absorptiometry (22). Subjects were tested after voiding and wearing gowns.

Statistical analysis

Data analysis was performed with SAS software (SAS Institute, Cary, NC). Subject characteristics at baseline were compared between men and women using independent sample t -tests. The primary outcome was ad libitum energy intake, averaged over the 4 days on a given dietary condition. Secondary outcomes including dietary fat level, food weight, and energy density were analyzed in the same way. Separate analyses were carried out for core foods, ad libitum buffet foods, and total intake. Repeated measures analysis of variance with dietary fat level (26, 34, and 40%) as a within-subject factor was carried out using linear mixed models with a heterogeneous compound symmetry covariance using SAS PROC MIXED. The effect of treatment order was also considered by including a variable for treatment order in the model. These methods account for the correlation due to repeated measures of subjects, allow different variances for the three dietary fat levels, and provide valid handling of the occasional missing observations for reasons not related to treatments or outcomes (23). Contrasts were used within these methods to estimate dietary fat effects. Statistical significance was set at P < 0.05. Results are presented as means and s.d. unless otherwise noted.

Visual analogue scale ratings of hunger before the meal, fullness after the meal, and average palatability of foods at each meal were averaged over the 4 days on each diet treatment, and were compared using repeated measures analysis of variance as above.

Results

Subject characteristics and completion/drop outs

Table 1 shows characteristics of subjects. Subjects were generally young and of normal weight, with ages ranging from 26 to 38 for women and 25 to 39 for men, and BMI ranging from 20.7 to 24.1 kg/m² for women and 21.6 to 27.6 kg/m² for men. Twenty of the twenty-two subjects completed all three dietary fat levels. Due to scheduling difficulties, one female and one male completed only two of the three levels.

Table 1 - Subject characteristics (mean (s.d.)).

Full table

Core Foods

The top panel of Table 2 shows the characteristics of the core foods in each diet. We were successful within 0.6 percentage points in achieving the levels of dietary fat intended for each condition. We were also successful in achieving equal amounts of energy from core foods across diets, with the mean energy content of core foods varying by only 23 kcal across diets. The statistical significance of this small difference (P < 0.0001) is due to the accuracy and small variability of the energy content across diets for individual subjects. Due to the substitution of fat for carbohydrate, energy density increased across treatments (P < 0.0001) and, in order for energy content to remain constant, food weight decreased with increasing fat level (P < 0.0001). Other determinants of energy density (i.e., non-calorics, mainly water and fiber) were also very similar across diets, varying by only 0.4 percentage points (data not shown). Thus, changes in energy density were due solely to changes in the macronutrient composition (% dietary fat).

Table 2 - Characteristics of core foods, ad lib intake, and total intake (mean (s.d.)).

Full table (46K)

Ad Libitum Intake

The middle panel of Table 2 shows ad libitum intake from the buffet. Dietary fat as a percent of energy paralleled the increase for the core diets, due to the design of the buffets in having similar fat levels as the core foods. The percentage of fat was several points lower than for the core foods, due apparently to subjects' selection of buffet foods. Food weight remained stable (P = 0.51) and ad lib energy intake increased (P = 0.052) across the diets, due to the increase in energy density (P < 0.0001). Variability of all quantities was much greater for ad lib intake than for core foods.

Total Intake

The bottom panel of Table 2 shows the total intake from core foods and buffet combined. Dietary fat levels paralleled the target levels but were slightly lower due to the lower fat content of the subjects' buffet items. By repeated measures analysis of variance, the energy contents of the diets were significantly different (P = 0 031). On the 26% fat diet, the average energy intake was 2,748 741 kcal/day (mean s.d.). This increased by 6% on the 34% fat diet (2,983 886 kcal/day, P = 0.091 compared to the 26% fat diet) and by 10% on the 40% fat diet (3,018 963 kcal/day, P = 0.011 compared to the 26% fat diet). Energy intake did not differ significantly between the 34 and 40% diets (P = 0.35). Figure 1 shows these results. When tested as a linear trend, the average energy intake increased significantly as the % fat in the diet increased (P = 0.008), by 19 7 kcal/day (mean s.e.m.) for each percentage point of dietary fat level. Food weight did not change significantly across the diets (P = 0.33) but energy density increased with dietary fat level (P < 0.0001), accounting for the increase in energy intake.

Figure 1.

The average energy intake (kcal/day) for each level of dietary fat.

Full figure and legend (16K)

Palatability

There were no significant differences in the visual analogue scale palatability ratings among the three diets, either when examining the three meals (breakfast, lunch, and dinner) separately or combining data from all meals (P > 0.06 for all). Table 3 shows the means, standard deviations, and P values for palatability of meals, hunger before meals, and fullness after meals.

Table 3 - VAS palatability of meals, hunger before meals, and fullness after meals (mean (s.d.)).

Full table (43K)

Hunger and fullness

There were no significant differences among the three diets in the visual analogue scale hunger ratings before meals separately or when combining data from all meals (P > 0.08 for all). Fullness ratings after meals did not differ significantly among the three diets when meals were combined (P = 0.28) or when breakfast and lunch were examined separately (P > 0.73 for both). There was a significant difference across diets in fullness after dinner (P = 0.03), with subjects reporting slightly greater fullness on the 40% fat diet (78.7 13.2 versus 74.6 11.9 and 75.0 9.6 on 26 and 34% fat diets), but this may be a result of doing several significance tests in this analysis.

Steps/day

The mean number of steps recorded per day on the 26, 34, and 40% diets, were 8,077 3,014, 8,028 2,593, and 8,832 3,089, respectively (mean s.d.), and these did not differ significantly across diets (P = 0.15). This suggests that the normal physical activity was not significantly affected by the changes in diet composition.

Discussion

These results show that the amount of fat in the diet can affect voluntary energy intake and can be important for body weight regulation. The study extends previous research (7,8,9) by showing that even within the typical range of consumption in western diets, total ad libitum energy intake increases as the fat content of the diet increases. Reducing dietary fat in the population should reduce total energy intake.

While we only evaluated three different levels of dietary fat, there appears to be a linear relationship between the fat in the diet and voluntary energy intake across the range of dietary fat tested, as shown in Figure 1. Since the pattern of ad libitum energy intake across dietary fat levels was as expected, given results in other studies across wider ranges of dietary fat, and was significant across the entire range (26–40%), we believe the lack of significance between individual fat levels (e.g., 26% vs. 34% and 34% vs. 40%) indicates there were an insufficient number of subjects to detect such small differences. These data would suggest that each 1% reduction in dietary fat within this range would, on average, reduce energy intake by 20 kcal/day. For example, reducing dietary fat by 5% would be expected to reduce voluntary energy intake by 100 kcal/day. Hill et al. (24) calculated that this reduction in energy intake should be sufficient to stop most of the gradual weight gain occurring in the population. For those not experiencing gradual weight gain, this would be sufficient to produce some weight loss. Thus, a 5% reduction in the fat intake of the population could be a strategy for at least stopping the gradual weight gain of the population and perhaps lowering the average BMI.

The 2005 Dietary Guidelines for Americans recommends that dietary fat should comprise no more than 35% of total energy intake (25). This is an increase in the upper limit of dietary fat by 5% over the previous guidelines (26), and according to our study results, could lead to an increase in energy intake and weight gain in the population.

Studies of body weights on low-fat diets in free-living individuals eating ad libitum are mixed. Astrup et al. (27) reviewed 19 intervention trials and found that 18 of 19 showed a lower body weight with low vs. higher fat diets. Alternatively, Willett (28) reviewed very long-term trials of low-fat diets and argued that differences in fat consumption within the range of 18–40% have no lasting impact on body fatness in the long term. The mixed results could be because of poor compliance to low-fat diets over the long term.

In contrast to studies in free-living subjects, controlled laboratory studies are consistent in showing that increases in dietary fat promote positive energy balance. These studies, including the present study, are consistent in that they show that as dietary fat increases voluntary energy intake increases (3,4,5). When studied in a whole-room calorimeter, 24-h energy balance becomes more positive with increased dietary fat (3,4). Further, excess energy in the form of dietary fat is stored more efficiently in the body than excess energy from carbohydrate (29). Dietary fat has been suggested to contribute to a small reduction in total energy expenditure, since fat has a lower thermic effect on food than carbohydrate (30). This results of this work, considered together, suggest that high-fat diets increase the likelihood of excessive energy intake and positive energy balance, and that excess fat is stored more efficiently than excess carbohydrate. There is a need for carefully controlled intervention studies, where dietary adherence is high, to examine the impact of alteration in dietary fat levels on body weight and body fat. Donnelly et al. (31) recently conducted one such study in college students and showed that energy intake increased significantly as dietary fat increased from <25% (low fat) to 28–32% (medium fat) to >35% (high fat). Body weight gain also increased with increasing levels of dietary fat. That study, taken together with the present results, strongly supports the conclusion that increasing dietary fat is associated with increasing energy intake and increasing likelihood of weight gain.

Dietary fat level has also been shown to play a role in regaining of weight after sustained weight loss. Individuals in the National Weight Control Registry, a registry of successful weight loss maintainers, report consuming a diet containing 24% of total energy from fat during weight loss maintenance (32). Further, an increase in dietary fat is one of the best predictors of weight regain over time in these individuals (33).

Others have reported that much of the impact of high-fat diets on energy intake could be due to the high energy density of these diets (34). For example, Stubbs et al. (4) found that subjects consumed 285 kcal/day more on a 40% fat diet as compared to a 20% fat diet when the energy density of food was not controlled. When these diets were studied by Stubbs et al. at a similar energy density, there was no difference in voluntary energy intake (6). G.K.G. et al. (35) showed that even small differences in water and fiber content can have large effects on energy density and energy intake in covert feeding trials. In studies of individual foods it has been reported (36,37,38) that high-fat foods tend to be low in moisture, resulting in further increases in energy density. Thus, in the real world, foods higher in fat may be even higher in energy density than those we used in our diets. However, it is difficult to replicate this correlation realistically between fat and moisture in laboratory studies. In the present study, our more conservative strategy was to control for the major determinants of energy density other than dietary fat–fiber and water. Further, the diets did not differ in palatability. In this way we can attribute the differences in energy intake to the differences in the proportion of fat in the diet.

Subjects in this study did not eat alone but rather consumed meals with other subjects. The number of people present at the meals was constant across the study. There is evidence that the number of people present at a meal may influence the amount eaten at that meal (39). We chose this method because we believe that most people consume meals not alone, but with others. Energy intake may have been different if the subjects had eaten alone, but there is no reason to think that the effects of diet composition would not be similar.

Clearly, dietary fat content is not the only factor affecting body weight. An effective strategy to address obesity will also have to focus on increasing physical activity. While this was not the focus of the present study, substantial other data suggest that increases in physical activity are important in preventing weight gain and weight regain following weight loss (40,41,42,43,44). For example, individuals who are most successful in weight loss maintenance report eating a low-fat diet (24%) and engaging in 60 min/day of physical activity (32). Both low-fat diets and increased physical activity would affect energy balance in a way that would make it easier to match energy intake to energy expenditure. Additionally, increases in dietary fiber may be associated with decreased weight gain (45).

These results do not support the benefits gained by a low carbohydrate diet, per se, in long-term weight control. In fact, the results can be interpreted to conclude that total energy intake increases as dietary carbohydrate is lowered. The primary data in support of the benefits of low carbohydrate diets come from studies of weight loss (46,47,48). There are no randomized controlled trials that have found low carbohydrate diets to be superior to low-fat diets in preventing weight gain. Some have suggested that particular types of carbohydrates (e.g., high glycemic index foods) may contribute to excessive energy intake and obesity (49,50). Sloth et al. (51) found no differences in body weight between low and high glycemic diets fed ad libitum. However, no studies have assessed energy intake in a controlled setting with high and low glycemic index diets.

Data from the present study, collected under controlled conditions, suggest that dietary fat contributes to a metabolic state that favors positive energy balance. The observation that body weight is not always reduced in long-term low-fat diet interventions seems to imply poor compliance with low-fat regimes. Thus, the long-term success of fat reduction in mitigating weight gain and supporting weight loss will depend on a food supply that provides choices that are both lower in fat and good tasting. Our results would suggest that even small reductions in dietary fat, which might be possible without dramatic adverse effects on food preferences, could help in preventing weight gain in the population.

In summary, these results are consistent with the notion that decreasing dietary fat even within the ranges typically consumed in Western diets could decrease the risk of consuming excess energy and could be an important factor in countering the gradual weight gain seen in the population. Whatever dietary advice is provided to the public must be accompanied by advice for engaging in regular physical activity. Based on these results, a low-fat diet combined with regular physical activity could be the foundation for a lifestyle that helps prevent excessive weight gain.

DISCLOSURE

The authors declared no conflict of interest.

References

REFERENCES

1. Department of Health and Human Services. The Surgeon General's report on nutrition and health. US Government Printing Office: Washington, DC, 1988. (DHHS publication [PHS] 88-50210)
2. Atkins RC. Dr. Atkins' New Diet Revolution. Avon Books: New York, 1992
3. Stubbs RJ, Ritz P, Coward WA, Prentice AM. Covert manipulation of the ratio of dietary fat to carbohydrate and energy density:effect on food intake and energy balance in free-living men eating ad libitum. Am J Clin Nutr 1995;62:330–337. | PubMed | ISI | ChemPort |
4. Stubbs RJ, Harbron CG, Murgatroyd PR, Prentice AM. Covert manipulation of dietary fat and energy density:effect on substrate flux and food intake in men eating ad libitum. Am J Clin Nutr 1995;62:316–329. | PubMed | ChemPort |
5. Thomas CD, Peters JC, Reed GW et al. Nutrient balance and energy expenditure during ad libitum feeding of high-fat and high-carbohydrate diets in humans. Am J Clin Nutr 1992;55:934–942. | PubMed | ChemPort |
6. Stubbs RJ, Harbron CG, Prentice AM. Covert manipulation of the dietary fat to carbohydrate ratio of isoenergetically dense diets: effect on food intake in feeding men ad libitum. Int J Obes Relat Metab Disord 1996;20:651–660. | PubMed | ChemPort |
7. Rolls BJ, Bell EA. Intake of fat and carbohydrate; role of energy density. Euro J Clin Nutr 1998;52:1–8. | Article |
8. Stubbs J, Ferres S, Horgan G. Energy density of foods: effects on energy intake. Crit Rev Food Sci Nutr 2000;40:481–515. | Article | PubMed | ISI | ChemPort |
9. Kral TVE,Roe LS, Rolls BJ. Combined effects of energy density and portion size on energy intake in women. Am J Clin Nutr 2004;79:962–968. | PubMed | ChemPort |
10. Briefel RR, McDowell MA, Alaimo K et al. Total energy intake of the US population: the third National Health and Nutrition Examination Survey, 1988–1991. Am J Clin Nutr 1995;62:1072S–1080S. | PubMed | ChemPort |
11. Kennedy E, Bowman S. Assessment of the effect of fat-modified foods on diet quality in adults, 19 to 50 years, using data from the Continuing Survey of Food Intake by Individuals. J Am Diet Assoc 2001;101:455–460. | Article | PubMed | ChemPort |
12. Briefel RR, Johnson CL. Secular trends in dietary intake in the United States. Ann Rev Nutr 2004;24:401–431. | Article | ChemPort |
13. Saris WHM,Astrup A, Prentice AM et al. Randomized controlled trial of changes in dietary carbohydrate/fat ratio and simple vs complex carbohydrates on body weight and blood lipids: the CARMEN study. Int J Obes Relat Metab Disord 2000;24:1310–1318. | Article | PubMed | ChemPort |
14. Bray GA, Popkin BM. Dietary fat intake does affect obesity! Am J Clin Nutr 1998;68:1157–1173. | PubMed | ISI | ChemPort |
15. Skov AR, Toubro S, Ronn B et al. Randomized trial on protein vs carbohydrate in ad libitum fat reduced diet for the treatment of obesity. Int J Obes Relat Metab Disord 1999;23:528–536. | Article | PubMed | ChemPort |
16. Trabulsi, J, Schoeller DA. Evaluation of dietary assessment instruments against doubly labeled water, a biomarker of habitual energy intake. Am J Physiol Endocrinol Metab 2001;281:E891–E899. | PubMed | ISI | ChemPort |
17. Stunkard A, Messick S. The three-factor eating questionnaire to measure dietary restraint. J PsychosomRes 1985;29:71–83. | Article | ChemPort |
18. Radloff LS. The CES-D scale: a self-report depression scale for research in the general population. Appl Psychol Meas 1977;1:385–401. | Article |
19. Garner DM, Olmsted MP, Bohr Y, Garfinkel PE. The eating attitudes test: psychometric features and clinical correlates. Psychol Med 1982;12:871–878. | PubMed | ISI | ChemPort |
20. Grunwald GK, Melanson EL, Forster JE et al. Comparison of methods for achieving 24-hour energy balance in a whole-room indirect calorimeter. Obes Res 2003;11:752–759. | PubMed |
21. McCormack HM, Horne DJ, Sheather S. Clinical applications of visual analogue scales: a critical review. Psychol Med 1988;18:1007–1019. | PubMed | ChemPort |
22. Haarbo J, Gotfredson A, Hassager C, Christiansen C. Validation of body composition by dual energy X-ray absorptiometry (DEXA). Clin Physiol 1991;11:331–341. | PubMed | ISI | ChemPort |
23. Verbeke G, Molenberghs G. Linear mixed models for longitudinal data. Springer: New York, 2000
24. Hill JO, Wyatt HR, Reed GW, Peters JC. Obesity and the environment: where do we go from here? Science 2003;299:853–855. | Article | PubMed | ISI | ChemPort |
25. Department of Health and Human Services. Dietary Guidelines 2005 <http://www.healthierus.gov/dietaryguidelines>. Accessed 16 July 2006
26. United States Department of Agriculture and United States Department of Health and Human Services. Dietary Guidelines for Americans 2000, Home and Garden Bulletin No. 232. 5th edn. US Government Printing Office: Washington, DC, 2000
27. Astrup A, Grunwald GK, Melanson E, Saris WHM, Hill JO. The role of low-fat diets in body weight control: a meta-analysis of ad libitum dietary intervention studies. Int J Obes Relat Metab Disord 2000;24:1545–1552.. | Article | PubMed | ChemPort |
28. Willett WC. Is dietary fat a major determinant of body fat? Am J Clin Nutr 1998;67(Suppl):556S–562S.
29. Horton TJ, Drougas H, Brachey A, Reed GW, Peters JC, Hill JO. Fat and carbohydrate overfeeding in humans: different effects on energy storage. Am J Clin Nutr 1995;62:19–29. | PubMed | ISI | ChemPort |
30. Hill JO, Drougas H, Peters JO. Obesity treatment: can diet composition play a role? Ann Int Med 1993;119:694–697. | PubMed | ChemPort |
31. Donnelly JE, Sullivan DK, Smith BK et al. Alteration of dietary fat to prevent weight gain: Jayhawk Observed Eating Study (JOE). Obesity, this issue
32. Klem ML, Wing RR, McGuire MT, Seagle HM, Hill JO. A descriptive study of individuals successful at long-term maintenance of substantial weight loss. Am J Clin Nutr 1997;66:239–246. | PubMed | ISI | ChemPort |
33. McGuire MT, Wing RR, Klem ML, Lang W, Hill JO. What predicts weight regain in a group of successful weight losers? J Consult Clin Psychol 1999;67:177–185. | Article | PubMed | ISI | ChemPort |
34. Rolls BJ. The role of energy density in the overconsumption of fat. J Nutr 2000;130:268S–271S. | PubMed | ISI | ChemPort |
35. Grunwald GK, Seagle HM, Peters JC, Hill JO. Quantifying and separating the effects of macronutrient composition and non-macronutrients on energy density. British J Nutr 2001;86:265–276. | ChemPort |
36. Poppitt SD. Energy density of diets and obesity. Int J Obes 1995;19(Suppl 5):S20–S26.
37. Poppitt SD, Prentice AM. Energy density and its role in the control of food intake: evidence from metabolic and community studies. Appetite 1996;26:153–174. | Article | PubMed | ISI | ChemPort |
38. Drewnowski A. Energy density, palatability, and satiety: implications for weight control. Nutr Rev 1998;56:347–353. | PubMed | ChemPort |
39. de Castro J, Brewer EM. The amount eaten in meals by humans is a power function of the number of people present. Physiol Behav 1991;51:121–125. | Article |
40. Dipietro L, Dziura J, Blair SN. Estimated change in physical activity level (PAL) and prediction of 5-year weight change in men: the Aerobics Center Longitudinal Study. Int J Obes Relat Metab Disord 2004;28:1541–1547. | Article | PubMed | ChemPort |
41. French SA, Jeffery RW, Forster JL et al. Predictors of weight change over two years among a population of working adults: the Healthy Worker Project. Int J Obes Relat Metab Disord 1994;18:145–154 | PubMed | ChemPort |
42. Hill JO, Wyatt HR. Role of physical activity in preventing and treating obesity. J Appl Phyisol 2005;99:765–770. | Article |
43. Williamson DF, Madans J, Anda RF et al. Recreational physical activity and ten-year weight change in a US national cohort. Int J Obes 1993;17:279–286. | ChemPort |
44. Romieu I, Willett WC, Stampfer MJ et al. Energy intake and other determinants of relative weight. Am J Clin Nutr 1988;47:406–412. | PubMed | ChemPort |
45. Koh-Banerjee P, Franz M. Sampson, Liu L et al. Changes in whole-grain, bran, and cereal fiber consumption in relation to 8-y weight gain among men. Am J Clin Nutr 2004;80:1237–1245. | PubMed | ISI | ChemPort |
46. Foster GD, Wyatt HR, Hill JO et al. A multi-center, randomized, controlled clinical trial of the Atkin's diet. N Engl J Med 2003;348:2082–2090. | Article | PubMed | ISI | ChemPort |
47. Stern L, Iqbal N, Seshadri P et al. The effects of low carbohydrate versus conventional weight loss in severely obese adults: one-year follow-up of a randomized trail. Ann Intern Med 2004;140:778–785. | PubMed | ISI |
48. Samaha FF, Iqbal N, Seshadri P et al. A low carbohydrate as compared to a low fat diet in severe obesity. N Engl J Med 2003;348:2074–2081. | Article | PubMed | ISI | ChemPort |
49. Ludwig D. Dietary glycemic index and obesity. J Nutr 2000;130:280S–283S. | PubMed | ISI | ChemPort |
50. Brand-Miller JC, Holt SHA,Pawlak DV, McMillan J. Glycemic index and obesity Am J Clin Nutr 2002;76:281S–285S. | PubMed | ISI | ChemPort |
51. Sloth B, Krog-Mikkelsoen I, Flint A et al. No difference in body weight decrease between a low-glycemic index and a high glycemic-index diet but reduced LDL cholesterol after 10-week ad libitum intake of the low-glycemic-index diet. Am J Clin Nutr 2004;80:337–347. | PubMed | ISI | ChemPort |

Acknowledgments

This research was supported by National Institutes of Health grants R37 DK042549 (to J.O.H.), P30 DK048520 (to J.O.H.), and MO1 RR00051.