- Distinct weight loss dietary strategies are associated with changes in specific food groups. To address the effect of changes in specific weight of food groups on weight loss in a 2-year low-fat, Mediterranean, low-carbohydrate intervention trial (DIRECT). We assessed changes in the intake of 12 food groups among 322 participants (body mass index [BMI] = 31 kg/m(2); age = 52 years; 86% men), using a validated electronic food frequency questionnaire. The weight of the 3592.9 ± 1558 (g/d ± SD) of baseline food consumed consisted mainly of liquids, excluding water (32.6% of total weight of food); vegetables (18.8%), fruits (17.7%), dairy (9.0%), meat (7.7%), and bread/cereal/pasta/potatoes (7.1%). Participants significantly reduced food intake by 283.73 ± 1342 (g/d ± SD) at 6 months and by 963.36 ± 1869 (g/d ± SD) at 24 months (p < 0.05 as compared with baseline). Food weight changes were similar across diet groups (p = 0.366), whereas 6-month body weight loss was -4.6 ± 4.4 kg, -4.7 ± 4.9 kg, and -6.4 ± 6.6 kg for low-fat, Mediterranean, and low-carbohydrate groups, respectively; p < 0.026). In multivariate regression models, adjusted for age, sex, baseline body weight, and changes in weight intake of 12 food groups (g/d), independent dietary predictors (standardized-β) at 6 months (rapid weight loss phase) were as follows: decreased consumption of sweets and cakes (β = 0.493; p = 0.008) in the low-fat group, tendency toward increased crude legumes (β = -0.196; p = 0.061) in the Mediterranean group, and increased vegetable intake (β = -0.249; p = 0.018) in the low-carbohydrate diet group. In the entire group, in models further adjusted for diet type, leading predictors for rapid weight loss phase were as follows: increased vegetables by ~140 g/d (β = -0.116; p = 0.045) and decreased intake of sweets and cakes by ~30 g/d (β = 0.162; p = 0.010). Universal predictors for 2-year successful weight loss in the entire group were as follows: increased intake of vegetables (β = -0.192; p = 0.007) and meat (β = -0.146; p = 0.026) and decreased intake of eggs (β = 0.187; p = 0.003), processed legumes (β = 0.195; p = 0.002), and beverages (β = 0.135; p = 0.032). Two-year weight loss is associated with a decrease of ~1 kg of total food consumed and may be achieved by a variety of changes in specific food groups within different diet strategies. Universal predictors of successful weight loss in the rapid weight loss phase across all diet strategies are increasing the weight of intake of vegetables and decreasing the weight of intake of sweets and cakes.