Marked Changes in One-Carbon Metabolism on a Low-Carbohydrate High-Fat Diet: A Randomized Controlled Trial (CARBFUNC)

Studying the impact of dietary carbohydrates and fat on markers of one-carbon metabolism could provide important insight into distinct metabolic adaptations and how diets affect disease risk.

This study aimed to determine changes in plasma one-carbon metabolites and B-vitamers on isocaloric diets differing in carbohydrate processing or amount.

A total of 193 individuals with obesity were randomly assigned to equal-energy and protein diets either with more processed, acellular carbohydrates [acellular, high-carbohydrate low-fat (A-HCLF), comparator arm], minimally processed cellular carbohydrates [cellular high-carbohydrate low-fat diet (C-HCLF)] [both designed with 45 energy percent (E%) carbohydrates and 38 E% fat], or a low-carbohydrate, high-fat (LCHF) diet designed with 8 E% carbohydrate and ∼75 E% fat, including 30 E% saturated fat. To address secondary outcomes of the trial, we analyzed changes in plasma one-carbon metabolites and B-vitamers, and dietary B-vitamin intakes after 3, 6, 9, and 12 mo, using constrained linear mixed-effect modeling.

After 3 mo, the relative change score after the LCHF diet was significantly different from the A-HCLF comparator for plasma α-hydroxybutyrate (LCHF/A-HCLF: +41.3%∗/-1.56%, P < 0.001; ∗: significant within-group change), methylmalonic acid (MMA) (-5.69%∗/+8.64%∗, P < 0.001), methionine (-7.81%∗/+1.44%, P = 0.001), nicotinamide (-19.7%∗/+14.0%, P = 0.007), 1-methylnicotinamide (mNAM) (-14.1%∗/+14.1%, P = 0.003), pyridoxal (-2.42%/+18.9%∗), 4-pyridoxic acid (-12.4%∗/+21.4%∗, P < 0.001), and cysteine (-3.00%∗/+1.10%). Between-group differences for most of these metabolites remained significant after 6 and 9 mo. Comparing C-HCLF with A-HCLF, significant differences in relative change scores were found for mNAM after 3 mo (C-HCLF/A-HCLF: -11.5%/+14.1%, P = 0.012) and 9 mo (-14.9%∗/+18.6%, P = 0.001). Plasma changes in one-carbon metabolites and B-vitamers showed weak correlations with dietary B-vitamins.

Compared with A-HCLF, the LCHF diet was followed by significantly different changes in plasma α-hydroxybutyrate, MMA, methionine, nicotinamide, mNAM, pyridoxal, 4-pyridoxic acid, and cysteine. These shifts were largely independent of vitamin consumption and may rather reflect ketoadaptive mechanisms, including enhanced fatty acid oxidation and upregulated antioxidant defense. Carbohydrate quality had less impact on the one-carbon metabolites. This trial was registered at clinicaltrials.gov as NCT03401970 (https://clinicaltrials.gov/study/NCT03401970).