W. Roodly Archer*,, Benoît Lamarche*,, Olivier Dériaz, Nancy Landry, Louise Corneau*, Jean-Pierre Després,,, Jean Bergeron, Patrick Couture*, and Nathalie Bergeron*,
* Nutraceuticals and Functional Foods Institute, Quebec, Canada;
Department of Food Sciences and Nutrition, Laval University, Quebec, Canada;
Lipid Research Center, Laval University Hospital Research Center, Quebec, Canada; and
Quebec Heart Institute, Laval Hospital Research Center, Quebec, Canada.
Address correspondence to Nathalie Bergeron, Nutraceuticals and Functional Foods Institute, Laval University, Pavillon Paul Comtois, Quebec, G1K 7P4 Canada. E-mail: Nathalie.Bergeron@inaf.ulaval.ca
Objective: To examine the extent to which variations in body composition modulate changes in the lipid profile in response to the ad libitum consumption of a diet rich in carbohydrates (CHOs) (high-CHO diet: 58% of energy as CHOs) or high in fat and in monounsaturated fatty acids (MUFAs) (high-MUFA diet: 40% of energy as fat, 23% as MUFAs).
Research Methods and Procedures: Sixty-three men were randomly assigned to one of the two diets that they consumed for 6 to 7 weeks. Body composition and fasting plasma lipid levels were measured at the beginning and the end of the dietary intervention.
Results: The high-CHO and high-MUFA diets induced significant and comparable reductions in body weight and waist circumference. These changes were accompanied by significant and comparable (p lipoprotein cholesterol levels. However, the high-MUFA diet had more beneficial effects on plasma triglyceride concentrations (p levels (p = 0.02) compared with the high-CHO diet. Diet-induced changes in waist circumference were significantly associated with changes in low-density lipoprotein cholesterol levels in the high-CHO group (r = 0.39, p = 0.03) but not in the high-MUFA group (r = 0.16, p = 0.38).
Discussion: Improvements in plasma lipids induced by the ad libitum consumption of a high-CHO diet seem to be partly mediated by changes in body weight, whereas lipid changes induced by the high-MUFA diet seem to be independent of changes in body weight.
Key Words: dietary components • lipoprotein • body weight • waist circumference • ad libitum
Source: Obesity Research 11:978-986 (2003). © 2003 The North American Association for the Study of Obesity
The use of low fat, high-carbohydrate (CHO)1 diets as a primary strategy to prevent cardiovascular disease has been challenged recently because of their potentially undesirable impact on plasma triglyceride (TG) and high-density lipoprotein cholesterol (HDL-C) concentrations (1) . It must be stressed that most studies that have highlighted these undesirable effects of high-CHO diets on plasma lipids were carried out under isocaloric conditions where caloric intake was imposed to keep body weight constant (2) . Another element that has led to increasing concern that high-CHO diets may not represent the optimal diet to prevent obesity and cardiovascular complications is that over the last 20 years, the reduction in fat intake in North America has been paralleled by a paradoxical increase in the prevalence of obesity (1) (3) . However, it must be noted that CHO quality [simple vs. complex or with a high vs. low glycemic index (GI)] may induce different metabolic responses and should be taken into account when evaluating the impact of high-CHO diets on the health profile. Indeed, the ad libitum consumption of high-complex CHO diets, which generally leads to significant reductions in body weight (4) (5) , has also been associated with highly desirable improvements in the cholesterol profile (6) (7) . What remains unclear is the extent to which these changes associated with high-CHO diets are dependent on concurrent reductions in body weight. The long-term compliance to high-complex CHO diets has also been an issue of concern in the free-living state because of the lower palatability of high-complex CHO diets compared with high-fat foods (8) (9) .
In that context, diets rich in monounsaturated fatty acids (MUFAs) (high-MUFA diets) have been advocated as dietary alternatives with good long-term compliance to prevent cardiovascular disease. In trials conducted under isocaloric conditions, high-MUFA diets have been reported to lower plasma low-density lipoprotein cholesterol (LDL-C) and TG concentrations with no undesirable effect on HDL-C levels (10) . However, due to the high caloric density of lipids, the effects of high-fat diets on weight management remain controversial. Several studies have reported that high-fat diets may promote body fat accumulation, which may in turn perpetuate a deteriorated lipid profile (11) (12) (13) . It must be stressed that very little is known about the impact of ad libitum consumption of high-fat diets rich in MUFA on body weight and on the plasma lipid profile. It is also unclear how changes in the lipid profile induced by such diets are a function of variations in body weight in ad libitum conditions.
Therefore, the objectives of this study were to compare the effects of ad libitum consumption of a diet rich in complex CHO vs. a high-fat diet rich in MUFA on body weight and body fat distribution in men and to examine the extent to which diet-induced variations in body weight and body fat distribution modulate the diet-induced changes in plasma lipid levels.
The study design has been described in detail previously (14) . In brief, 63 sedentary men were recruited in the Quebec metropolitan area. Exclusion criteria included endocrine, cardiovascular, hepatic and renal disorders, use of medication known to affect lipid metabolism, smoking, and significant change in body weight within the year that preceded study onset. Individuals with excessive alcohol intake, unusual dietary habits, and food aversions or allergies were also excluded. Each participant signed a consent form approved by the Clinical Research Ethical Committee of Laval University.
Subjects were randomized to either a low-fat, high-CHO diet or a high-fat diet rich in MUFAs, which they consumed for 6 to 7 weeks. Subjects were instructed to maintain their usual physical activity except for the 3 days preceding the beginning and the end of the study, during which they were required to remain inactive. The baseline characteristics of study subjects are presented in Table 1 . Participants and staff performing laboratory measures were blinded to dietary treatments. Compliance to the experimental diets was assessed using riboflavin incorporated in the foods, and only two subjects were excluded from the analysis due to a low compliance.
On weekdays, subjects came to the metabolic unit daily to consume their 12 PM meal under the supervision of at least one member of the staff, at which time they were also given their evening meals and next day’s packaged breakfast to take home. On weekends, all meals were provided by the research unit but were packaged to take home. To achieve ad libitum conditions, participants were blinded to the fact that they were receiving food in quantities that met 150% of their habitual daily energy intake as assessed by 3-day food records (2 weekdays and 1 weekend day), obtained at baseline. The breakfast meal represented 20% of the daily energy intake, whereas the lunch and dinner meals each provided 40% of daily energy intake. Subjects were instructed to consume their entire breakfast but were asked to eat their lunches and dinners on an ad libitum basis, until satiety was met. To ensure that participants consumed the adequate proportion of macronutrients, they were instructed to consume the same proportion of each component of the meal, which was set out in layers in their plates for that purpose. For example, a participant who consumed 75% of the meal’s meat had to consume 75% of the pasta, 75% of the vegetable, and 75% of the dessert, if any. All leftovers were returned to the laboratory and weighed to calculate actual energy intakes. For participants used to eating between meals, 200-kcal snacks were provided on demand. These high-CHO and high-MUFA snacks were prepared in our kitchen and had the same macronutrient composition as that of the two experimental diets. Caffeine-containing beverages were restricted to two per day, but subjects had free access to water and to diet, caffeine-free beverages.
Anthropometric and Body Composition Measurements
Body weight and waist circumferences were measured according to standardized procedures (17) at the beginning and at the end of the study period. Total, subcutaneous, and visceral adipose tissue (AT) accumulation were assessed by computed tomography as described previously (18) .
Data were analyzed using both SAS software (version 8.2; SAS Institute Inc., Cary, NC) and JMP statistical software (version 4.0.5; SAS Institute Inc.). Differences among and between dietary groups were tested by ANOVA for repeated measurements using relative changes from baseline. Spearman’s correlation coefficients were calculated to test for associations between diet-induced changes in body weight and body composition and changes in plasma lipid levels.
As shown in Table 2 , the averaged daily energy intake in the high-CHO and high-MUFA groups during the experiment was similar (3038 ± 456 vs. 2976 ± 586 kcal respectively, p = 0.65). Energy intake did not vary significantly over the 6- to 7-week intervention period in either group (not shown). Usual energy intakes measured at baseline (high-CHO: 3053 ± 855 kcal, high-MUFA: 2842 ± 798 kcal) were not statistically different from daily energy intakes during the experiment (p > 0.5). As shown in Figure 1 , this apparent lack of change in mean energy intake was not indicative of the considerable interindividual variation in body weight and waist circumference observed in response to the experimental diets. These diet-induced variations in body weight were significantly associated with the variability in energy intake induced by the study protocol (on-diet vs. baseline) in both the high-CHO and the high-MUFA groups (r = 0.43 and 0.44, respectively, p = 0.02).
As shown in Table 3 , ad libitum consumption of the high-CHO diet led to a moderate but significant (p in body weight (-2.4%); in waist circumference (-2.6%); and in total, visceral, and subcutaneous AT areas (-11.1%, -11.3%, and -9.9%, respectively). These changes were accompanied by significant (p LDL-C (-21%), and HDL-C concentrations (-10%), with no effect on plasma TG levels (p = 0.33).
Ad libitum consumption of the high-MUFA diet also induced significant (p (-2.3%), and total, visceral, and subcutaneous AT areas (-8.7%, -11.7%, and -6.6%, respectively). The high-MUFA diet was also associated with significant reductions (p cholesterol (-14%) and LDL-C levels (-16%), but had no effect on plasma HDL-C concentrations (p = 0.26). In contrast to the high-CHO diet, the high-MUFA diet resulted in a significant reduction in plasma TG concentrations (-17%, p the impact of both diets on physical characteristics and on total plasma and LDL-C levels was comparable, the high-MUFA diet had a more favorable impact on plasma TG and HDL-C levels.
As shown in Table 4 , there were significant inverse correlations between baseline body weight and diet-induced changes in body weight (r = -0.49, p r = -0.47, p group. As shown in Figure 2 , diet-induced changes in waist circumference were significantly related to concurrent variations in plasma LDL-C levels in the high-CHO group but not in the high-MUFA group. Although visceral AT area was positively correlated with total plasma cholesterol (r = 0.82, p (r = 0.78, p r = 0.71, p 0.01) at baseline, diet-induced changes in visceral AT did not correlate with concurrent variations in any of the plasma lipid variables in either the high-CHO or the high-MUFA diet (data not shown).
Two approaches were used in an attempt to dissociate the concurrent effects of weight loss and diet intervention on plasma lipid concentrations. First, we arbitrarily divided the study participants into those with no diet-induced reduction in body weight ( weight 0 kg) and those who had a reduction in body weight greater than 0 kg ( weight (Figure 3) . The mean change in body weight associated with the consumption of the high-CHO diet was 0.71 ± 0.69 kg among men who did not lose weight (n = 6) and -2.95 ± 2.29 kg among those who lost weight (n = 25). Using this approach, we found that the reduction in total plasma cholesterol, LDL-C levels, and plasma cholesterol-to-HDL-C ratio was significantly greater among men who lost weight compared with those who did not (p On the other hand, the reduction in plasma HDL-C levels in the high-CHO diet group did not seem to be attenuated by a greater reduction in body weight. Finally, men who did not lose weight while consuming the high-CHO diet showed a greater elevation in plasma TG levels compared with men who lost weight (p 0.01) (Figure 3 , top panel). In the high-MUFA dietary group, the mean body weight change was 1.73 ± 2.05 kg among men who did not lose weight (n = 9) and -3.53 ± 1.88 kg among those who lost weight (n = 23). In general, the reduction in total plasma cholesterol, LDL-C, and TG was greater among men who lost weight compared with those who did not, but none of these differences reached statistical significance (Figure 3 , lower panel).
In a second approach aimed at dissociating the concurrent effects of weight loss and diet on plasma lipids, we adjusted the diet-induced changes in plasma lipid levels for concurrent variations in body weight and visceral AT using multivariate analysis. Consistent with the previous analysis, changes in body weight significantly contributed to the diet-induced changes in LDL-C (p = 0.03) and TG levels (p = 0.01) in the high-CHO dietary group but not in the high-MUFA group. Changes in visceral AT did not modulate the diet-induced changes in plasma lipid levels in either the high-CHO or the high-MUFA groups.
This study investigated the extent to which the effects of ad libitum consumption of a high-CHO diet vs. a high-MUFA diet on the lipid profile are modulated by concurrent changes in anthropometry and body composition in men. We report that both high-CHO and high-MUFA diets consumed ad libitum were associated with a moderate but significant body weight loss and with beneficial reductions in total plasma cholesterol and LDL-C levels. Because our subjects did not apparently modify their physical activities during the study, the diet-induced weight loss is most likely to have resulted from a reduction in daily energy intake. Surprisingly, energy intakes during the experiment did not differ from usual daily energy intakes measured at baseline. We believe that this apparent inconsistency may be largely explained by underreporting of energy intake measured at baseline (19) , although the limitations of 3-day dietary records as a tool to adequately assess usual energy intake may also be evoked (20) . The possibility that subjects entered the study with the objective to lose weight cannot be ruled out, although each participant was strictly instructed not to diet on a voluntary basis.
The present study also emphasized the important interindividual variability in body weight change in response to the high-CHO and high-MUFA diets consumed under controlled but unrestricted caloric intake (ad libitum conditions). Within the high-CHO group only, men with excess body weight at baseline were more likely to experience a greater diet-related reduction in body weight and waist circumference than leaner men (Table 4) , which was, in turn, associated with a more pronounced improvement in the plasma cholesterol profile (Figure 2) . On the other hand, variations in body weight in response to the high-MUFA diet were not related to baseline body weight and did not modulate the diet-induced changes in plasma cholesterol levels of participants.
The present study also concurs with the concept that isocaloric consumption of low-fat high-CHO diets has potentially detrimental cardiovascular effects by inducing undesirable elevations in plasma TG levels (21) (22) . Indeed, the group of men who lost weight while consuming the high-CHO diet showed no change in plasma TG levels, whereas those who had no change or slight elevations in body weight, thereby mimicking isocaloric conditions, experienced concurrent elevations in plasma TG concentrations. Although ad libitum consumption of the high-CHO diet in association with significant reductions in body weight or waist circumference did not prevent a reduction in HDL-C concentrations, the plasma cholesterol-to-HDL-C ratio nevertheless was significantly lowered, suggesting that the magnitude of LDL-C lowering was greater than that of HDL-C. Because the plasma cholesterol-to-HDL-C ratio is considered as a very powerful indicator of cardiovascular risk, the overall benefit of the high-CHO diet, when associated with moderate weight loss, can be considered as positive. Finally, results from multivariate analyses indicated that diet-induced changes in body weight amplified the concurrent change in the plasma cholesterol-to-HDL-C ratio when subjects consumed the high-CHO diet.
Results from the present study are consistent with data from Schaefer et al. (6) , who reported that consumption of a very low-fat (15% of energy), high-CHO (68%) diet under isocaloric conditions was associated with significant reductions in total plasma cholesterol, LDL-C, and HDL-C and with a significant increase in TG levels. However, when the 27 moderately hypercholesterolemic subjects consumed the same very low-fat, high-CHO diet under ad libitum conditions, i.e., without being imposed a specified energy intake, there was a moderate but significant weight loss of 3.6 kg accompanied by a greater reduction in plasma total and LDL-C levels, but not in HDL-C, and a normalization of TG levels when compared with the isocaloric phase (6) . Taken together, these results support the thesis that a low-fat, high-CHO diet exerts its beneficial impact on the plasma lipid profile only when associated with a moderate weight loss (6) (7) (23) .
The weight loss induced by consumption of a high-CHO diet could be of particular benefit to obese men with excess abdominal visceral AT levels. Increased abdominal visceral AT is often associated with an increase in the risk of cardiovascular disease in men and women due to an altered plasma lipid profile, including high plasma TG and low HDL-C levels (24) (25) . Although levels of visceral AT correlated with a deteriorated plasma lipid profile at baseline in both dietary groups (not shown), diet-induced changes in visceral AT area were not associated with concurrent changes in plasma lipid levels. Therefore, future investigations are warranted to clarify why body weight, and not the visceral AT levels, predicted and modulated the response of the lipid profile induced by the high-CHO diet.
Current dietary guidelines that until recently promoted the consumption of low-fat, high-CHO diets have not slowed down the increasing rates of obesity in industrialized countries (26) . Many high-CHO foods common to Western diets have a low fiber content and high simple sugar levels, and, hence, have a high GI (27) . Such foods have been associated with an accelerated rate of development of obesity (28) and have been shown to independently increase the risk of cardiovascular disease (29) . In contrast, it has been demonstrated that fiber-rich, high-CHO diets with a low GI may facilitate weight control by promoting satiety and inducing a negative energy balance (5) (28) . Because results from our study suggest that most of the benefits attributed to low fat, high-complex CHO diets may be dependent on a concurrent reduction in body weight, it may be argued that low-fat diets may not reduce long-term risk of cardiovascular disease because they seem to be inefficacious for weight control over time (30) . Strategies to adequately promote low-fat, high-CHO diets as a means of promoting weight loss will nonetheless have to emphasize the importance of selecting beneficial low-GI foods rich in complex CHO and fiber rather than their undesirable fiber-poor, energy-dense counterparts.
In contrast to the high-CHO diet, changes in the plasma lipid profile after the high-MUFA diet included a lowering of TG with no effect on HDL-C levels, as observed in previous studies (31) (32) . In addition, changes in the lipid profile induced by the high-MUFA diet were independent of changes in body weight and body composition. This is an important observation in the context of the ongoing debate on the impact of dietary fat on obesity. In fact, despite their now well-accepted beneficial impact on plasma lipid levels (10) (31) (32) , diets rich in MUFA have been criticized because of their potential to promote weight gain when consumed ad libitum (11) (23) . Bray and Popkin (11) and Astrup et al. (33) argued that dietary fats play a role in the development of obesity by favoring passive overconsumption. As a consequence, a reduction in total energy intake facilitated by a decrease in dietary fat has been suggested as the most powerful way to prevent obesity and its complications. Other evidence suggests that it may be important to consider the type of fat consumed when evaluating the impact of dietary fat on body weight and composition. In an experiment undertaken to determine whether the type of dietary fat could influence body composition in mice, Bell et al. (34) observed that a diet rich in MUFA derived from canola oil caused a less important body weight gain compared with a diet rich in saturated fatty acids from beef, thus suggesting that the type of fat consumed could also modulate body composition. Larson et al. (35) evaluated the association between dietary fat and AT levels in a cross-sectional analysis of middle-aged men and women. They found that dietary fat was not independently associated with abdominal visceral and subcutaneous AT in a multiple regression model (35) .
A limited number of controlled clinical trials have investigated the effects of ad libitum consumption of a high-fat, high-MUFA diet on body composition and the lipid profile. Golay et al. (36) evaluated the effects of diets equally low in energy but with different amounts of dietary fat and CHOs on body weight in a sample of 43 obese patients hospitalized for a period of 6 weeks. Concordant with our observations, there was no significant difference between the weight loss achieved by the low-fat (26% of energy), high-CHO (45%) diet compared with the high-fat (53%) diet. In addition, the magnitude of diet-induced changes in body composition reported as changes in total body fat and variations in the waist-to-hip ratio were similar between the two groups. However, because the hypocaloric diets were combined with a structured physical activity program, the diet-specific effects on body weight and body composition could not be readily dissociated from the simultaneous impact of the physical activity program. Nevertheless, results of that study also showed that energy intake, not macronutrient composition, was the most significant determinant of body weight loss (36) .
In conclusion, results from the present study suggest that part of the benefits of a high-CHO diet on plasma lipid risk factors may be attributable to concurrent reductions in body weight. The high-MUFA diet, however, induced improvements in total and LDL cholesterol levels that were similar to those observed in the high-CHO group, with more desirable changes in plasma TG levels and no apparent deterioration in plasma HDL-C concentrations. In contrast to the high-CHO diet, the lipid changes induced by the high-MUFA diet were not related, at least statistically, to concurrent changes in body weight or composition. More clinical trials are needed to assess the effects of long-term ad libitum consumption (over 6 months) of high-CHO and high-MUFA diets on body composition and long-term risk of obesity and cardiovascular disease.
This work was financed by Knoll Pharmaceuticals, Human Nutrition ILSI Institute Research Foundation, the Canadian Institute for Health Research, and the International Olive Oil Council. W. R. A. received a scholarship from the Réseau en Santé Cardiovasculaire du Québec and the Fonds de la Recherche en Santé du Québec. B. L. is Chair Professor in Nutrition, Functional Food, and Cardiovascular Health from the Canada Research Chair Program. J. B. and P. C. are Clinical Research Scholars from the Fonds de la Recherche en Santé du Québec.
1 Nonstandard abbreviations: CHO, carbohydrate; TG, triglyceride; HDL-C, high-density lipoprotein cholesterol; GI, glycemic index; MUFA, monounsaturated fatty acid; LDL-C, low-density lipoprotein cholesterol; AT, adipose tissue.
Figure 1. Individual diet-induced changes in body weight and waist circumference.
Figure 2. Correlations between diet-induced changes in waist circumference and changes in plasma LDL-C levels.
Figure 3. Effects of the high-CHO (top panel) and the high-MUFA diet (bottom panel) on the lipid profile of men who did not lose weight ( weight 0 kg) vs. those who lost weight ( weight p
Source: Obesity Research 11:978-986 (2003). © 2003 The North American Association for the Study of Obesity
* Mean ± SD.
Mean ± SD; the percentage of macronutrients was predetermined in each of the experimental diets; SDs for macronutrients, cholesterol, and ratios in experimental diets were virtually equal to zero and, therefore, are not shown.
* Significantly different from the high-CHO experimental diet; p 0.01.
Table 3. Effects of both diets on physical characteristics and the lipid profile*
* Mean ± SD.
p Value related to the repeated-measure analysis (within-diet effect), based on the relative change compared with the baseline value.
p Value related to the percentage change between the two diets (between-diet effect).
Table 4. Correlations between baseline body weight and diet-induced changes in anthropometry and body composition
Source: Obesity Research 11:978-986 (2003)
© 2003 The North American Association for the Study of Obesity