Assessment of body composition in relation to the habitual diet and after a six month dietary intervention.After a baseline measurement subjects were randomly assigned to either a group consuming reduced-fat products or a group consuming full-fat products for six months.108 women and 109 men, equally distributed over the age range 19-35 with BMI between 21 and 28, and the age range 36-55 with BMI between 24 and 30.Food intake was measured by three day dietary record, body composition by deuterium dilution.At baseline, explained variance of %body fat on age and fat-carbohydrate ratio in the diet together were 17% (P < 0.0001) and 36% (P < 0.0001) for women and men, respectively, and on diet alone 7-8% (P < 0.01) independent of gender. The diet intervention caused on average a change in fat intake and body fat mass in subjects of the reduced-fat group of -5 +/- 29 g/d (P < 0.05) and -0.1 +/- 2.1 kg (ns), respectively, and of +23 +/- 31 g/d (P < 0.0001) and +0.5 +/- 2.3 kg (P < 0.05) in subjects of the full-fat group. The change in the fat content of the diet was positively related to a change in energy intake (fat-carbohydrate ratio: R2 = 0.15, P < 0.0001; g fat: R2 = 0.70, P < 0.0001), the latter explaining 5% of the variation in the change in body fat mass (P < 0.001). Subjects changing the fat content of the diet showed a consequent change in body fat mass only when energy intake changed as well.The fat content of the diet has an effect on body fat as a function of the effect of dietary fat on energy intake.