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.
The nutritional composition of the experimental diets was calculated with the Canadian Nutrient File database [Health Canada, Ottawa, 1997 (http://www.hc-sc.gc.ca/food-aliment/ns-sc/nr-rn/surveillance/cnf-fcen/e_index.html)] and the Nutrition Data System for Research software (database version 4.03_30, 1999; Nutrition Coordinating Center, Minneapolis, MN). The experimental diets consisted of usual solid foods that were prepared daily in our metabolic kitchen and weighed in individual portions. Both experimental diets were formulated to have a similar food composition and differed mainly with respect to macronutrients (Table 2)
. The diets were composed of nonhydrogenated unsaturated fats, mostly olive oil, with whole grains and vegetables as the main forms of CHOs. Simple sugars were used only in the preparation of muffins and some desserts.
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.
Plasma lipid levels were measured as described previously (15) (16) on blood samples collected after a 12-hour fasting period at the beginning and at the end of the study.
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.