Glucose-lowering effect of whey protein depends upon clinical characteristics of patients with type 2 diabetes

OBJECTIVE Whey protein (WP) intake has been shown to reduce postprandial glycemia. Majority of WP research in type 2 diabetes (T2DM) involved acute challenge or weight loss studies. It is not known if WP supplementation can provide sustained glucose lowering. Our goal was to investigate the effects of WP on glycemia comprehensively by using continuous glucose monitoring (CGM) while avoiding the confounding effects of variable food intake through controlled feeding. RESEARCH DESIGN AND METHODS This double-blinded and placebo (PL)-controlled study included 22 patients with T2DM patients (11 male, 11 female; age 57.1±12.6 years) on diet or metformin monotherapy. First, one serving (21 g) of WP was compared with PL in parallel-armed acute challenge studies. Next, in a crossover design, each patient underwent CGM twice, over 2 consecutive weeks, 3.5 days each week. Identical diets were provided by the study during both CGM periods. During the first CGM, one serving of either WP or PL was consumed before breakfast and another before dinner. During the second CGM, participants switched to the alternate supplement. Order of the supplements was randomized. RESULTS During acute challenge studies, WP stimulated insulin and glucagon-like peptide (GLP)-1 secretion; suppressed ghrelin (all p<0.05), while PL had no effect. During CGM, glucose response to WP varied depending on the baseline characteristics of the patients. When evaluated using linear regression, the most predictive baseline variables were body mass index (BMI) (p=0.0006), triglycerides (p=8.3×10-5) and GLP-1 (p=0.006). Lower BMI, triglyceride and GLP-1 predicted decreased glucose levels on WP. Obesity, hypertriglyceridemia and high fasting GLP-1 concentrations predicted increased glucose levels. CONCLUSIONS Effects of WP supplementation on glycemia in T2DM depend on the baseline characteristics. Lower body weight, normal triglyceride and lower GLP-1 levels predict glucose lowering. In contrast, obesity, hypertriglyceridemia and high baseline GLP-1 predict increased glucose response.