Using a Metabotype Framework to Deliver Personalized Nutrition Improves Dietary Quality and Metabolic Health Parameters: A 12-Week Randomized Controlled Trial.

Abstract

Effective strategies for tailoring dietary advice to individuals are urgently needed. The effectiveness of personalized nutrition advice delivered using a metabotype framework in improving dietary quality and metabolic health biomarkers compared to population-level advice is investigated. A 12-week parallel randomized controlled trial is performed with 107 healthy adults. Individuals in the personalized group are classified into metabotypes using four markers (triacylglycerol, high-density lipoprotein [HDL]-cholesterol, total cholesterol [TC], and glucose) and received dietary advice from decision tree algorithms containing metabotypes characteristics and individual traits. Individuals in the control group received generic dietary advice based on national guidelines. The personalized approach results in higher dietary quality assessed by the Alternate Mediterranean Diet Score(effect size [95% confidence interval, CI], 0.77 [0.07, 1.48], 12% versus 3% increase) and significantly lower concentrations of triacylglycerol (-0.17 [-0.28, -0.06] log10mmolL-1 ), TC (-0.42 [-0.74, -0.10]mmolL-1 ), low-density lipoprotein (LDL)-cholesterol (-0.34, [-0.60, -0.09]mmolL-1 ), and lower triacylglycerol-glucose index (-0.40, [-0.67, -0.13]). Sixteen phosphatidylcholines and six lysophosphatidylcholines, predominately with chain lengths of 30-36 carbons, are lower in the personalized group. Personalized nutrition advice delivered using the metabotype framework is effective to improve dietary quality, which could result in reduced CVD risk, and metabolic heath biomarkers.