Individual effect of diet on postprandial glycemic response and its relationship with gut microbiome profile in healthy subjects: protocol for a series of randomized N-of-1 trials

BackgroundType 2 diabetes causes over a million deaths annually, ranking in the top ten causes of death worldwide. Glycemic control through dietary adequacy is essential for treatment success and disease prevention. Recent evidence indicates that the glycemic response to various foods varies from individual to individual. The intestinal microbiome is seen as a potential key player, mediating the effect of foods on glycemic response. By design, however, most published studies cannot separate variation in the individual treatment effects (ITE) of different diets from within-individual variability of glycemic responses. In this context, the present study aims to assess the heterogeneity in the ITE of diet on glycemic response and investigate the relevance of the intestinal microbiome profile as a predictor of this heterogeneity. MethodsThis study is a series of N-of-1 randomized clinical trials. Each participant will undergo five treatment cycles of two prescribed diets (low-carb versus vegan) in one of two randomly chosen treatment sequences (ABBABAABBA or BAABABBAAB). The primary outcome is the positive incremental area-under-the-curve (iAUC) of the postprandial interstitial fluid glucose measured within 2 hours of meal consumption. The trial plans to recruit 80 healthy volunteers with ages between 18 and 60. Fecal samples will be collected at baseline for microbiome analysis by metagenomics shotgun technique. Random effects linear models will be used for the primary analysis. DiscussionWhile significant variation of individual effects warrants personalized interventions, it is well-known that glycemic responses to the same food, in the same individual, vary from occasion to occasion. Yet, most clinical studies are based on designs that are incapable of separating ITE variation from within-individual variability. This is a major limitation since the personalization of dietary interventions is only justified by clinically relevant heterogeneity of individual-level effects. In this study, if significant ITE variation is indeed observed, then we will also be able to estimate the relationship between the intestinal microbiome and the expected diet effects. This is essential to identify predictive biomarkers, which can identify those who intrinsically benefit the most from which diet, going beyond pure associations with glycemic response. Conversely, observing negligible ITE variation in a large series of N-of-1 trials would cast major doubts on the relevance of personalizing dietary interventions for glycemic control. Therefore, the present study represents a major step toward understanding the clinical value of microbiome-driven precision nutrition.