Reformulating cereal bars: high resistant starch reduces in vitro digestibility but not in vivo glucose or insulin response; whey protein reduces glucose but disproportionately increases insulin

Background: Resistant starch (RS) and whey protein are thought to be effective nutrients for reducing glycemic responses. Objective: We aimed to determine the effect of varying the sucrose, RS, and whey protein content of cereal bars on glucose and insulin responses. Design: Twelve healthy subjects [mean ± SD age: 36 ± 12 y; mean ± SD body mass index (in kg/m²): 24.9 ± 2.7] consumed 40 g available-carbohydrate (avCHO) portions of 5 whole-grain cereal bars that contained varying amounts of RS and whey protein concentrate [WPC; 70% protein; RS:WPC, %wt:wt: 15:0 (Bar15/0); 15:0, low in sucrose (Bar15/0LS); 15:5 (Bar15/5); 10:5 (Bar10/5); and 10:10 (Bar10/10)] and 2 portion sizes of a control bar low in whole grains, protein, and RS [control 1 contained 40 g avCHO (Control1); control 2 contained total carbohydrate equal to Bar15/0LS (Control2)] on separate days by using a randomized crossover design. Glucose and insulin responses in vivo and carbohydrate digestibility in vitro were measured over 3 h. Results: Incremental area under the curve (iAUC) over 0–3 h for glucose (min × mmol/L) differed significantly between treatments (P < 0.001) [Bar15/0LS (mean ± SEM), 169 ± 14; Control2, 164 ± 20; Bar15/0, 144 ± 15; Control1, 140 ± 17; Bar10/5, 117 ± 12; Bar15/5, 116 ± 9; and Bar10/10, 100 ± 9; Tukey’s least significant difference = 42, P < 0.05], but insulin iAUC did not differ significantly. Higher protein content was associated with a lower glucose iAUC (P = 0.028) and a higher insulin-to-glucose iAUC ratio (P = 0.002) All 5 RS-containing bars were digested in vitro ∼30% more slowly than the control bars (P < 0.05); however, in vivo responses were not related to digestibility in vitro. Glucose and insulin responses elicited by high-RS, whey protein–free bars were similar to those elicited from control bars. Conclusions: The inclusion of RS in cereal bar formulations did not reduce glycemic responses despite slower starch digestion in vitro. Thus, caution is required when extrapolating in vitro starch digestibility to in vivo glycemic response. The inclusion of whey protein in cereal bar formulations to reduce glycemic response requires caution because this may be associated with a disproportionate increase in insulin as judged by an increased insulin-to-glucose iAUC ratio. This trial was registered at clinicaltrials.gov as NCT02537587.