Abstract P142: Mediation of Effects of a Low-Carbohydrate Dietary Intervention on Hemoglobin A1c Through Weight: A Randomized Clinical Trial

Introduction: Low-carbohydrate diets lead to weight loss and improvement in hemoglobin A1c and other cardiometabolic outcomes. However, there is limited research on how much of the glycemic effects of low-carbohydrate diets are mediated by weight loss. Hypothesis: The effect of a low-carbohydrate dietary intervention on hemoglobin A1c would be mediated by weight loss. Methods: We conducted a 6-month randomized clinical trial on the effects of a low-carbohydrate diet intervention (target <40 net grams of carbohydrates during first 3 months; <60 net grams for months 3-6) compared with usual diet on hemoglobin A1c. Participants were 40-70 years with untreated HbA1c of 6.0-6.9%. In post hoc mediation analyses, we examined whether the effect of the intervention on 6-month change in hemoglobin A1c was mediated by 3-month weight, after adjustment for age, sex, race, education, and baseline levels of behavioral and cardiometabolic risk factors. We used the causal framework approach described by and the SAS macro developed by Valeri and VanderWeele. Results: 150 participants (mean [SD] age 59 [8] years; 108 [72%] female; 88 [59%] Black) were randomized to either the low-carbohydrate diet intervention (75 participants) or usual diet (75 participants) arm. Mean (SD) HbA1c was 6.16% (0.30%) at baseline. Six-month data was collected on 142 participants (95%). Compared with the usual diet arm, the low-carbohydrate diet intervention arm had significantly greater 6-month reduction in HbA1c (net difference, -0.23% [95% CI, -0.32 to -0.14%]) and body weight (-5.9 kg [95% CI, -7.4 to -4.4 kg]). In mediation analyses, 59% of the total effect of intervention on 6-month change in hemoglobin A1c was mediated by weight measured at 3 months (Table). Conclusions: In mediation analyses, a substantial proportion (59%) of the total effect of intervention on 6-month change in hemoglobin A1c was mediated through weight measured at 3 months. Future research should further explore contributions of weight loss to reductions in glycemia in other populations.