Aim: We studied fasting plasma glucose (FPG) as a determinant of weight change on high-fat diets in the PREDIMED trial. 

Methods: A total of 3,622 participants were randomized to receive one of two Mediterranean diets (n = 2,616) or a control diet (n = 1,006) for 5 years and had complete data for baseline FPG and body-weight development. Weight change by pre-treatment FPG categories (<100 mg/dL; 100–114.9 mg/dL; 115–125.9 mg/dL; ≥126 mg/dL) were examined. 

Results: The two Mediterranean diets contained 41.5 E% fat, 16.5 E% protein, and 40 E% carbohydrate whereas the control diet contained 37.8 E% fat, 16.8 E% protein and 43.2 E% carbohydrate. In the Mediterranean diet groups, participants with FPG≥115 lost 1.04 kg (95% CI 0.68; 1.41, n = 1115) whereas participants with FPG<115 mg/dL lost 0.30 kg (95% CI -0.01; 0.60, n = 1501) resulting in a between-group difference of 0.75 kg (95% CI 0.20; 1.30, P = 0.007). Similar, although none-significant, between-group difference was observed in the control group [0.70 kg (95% CI -0.25; 1.65, P = 0.15)]. 

Conclusions: Participants with elevated pre-treatment FPG lost more weight than participants with lower FPG on these three moderate to high fat diets. Although the effect size is small, results are important as these individuals represent those in most need of weight loss.