Wang · European heart journal 2019 · randomized crossover trial · n=113

Impact of chronic dietary red meat, white meat, or non-meat protein on trimethylamine N-oxide metabolism and renal excretion in healthy men and women.

Cited 492 times in the scientific literature.

Level 2 - randomized trial

Individual randomized crossover feeding trial

PubMed 30535398 · doi:10.1093/eurheartj/ehy799 · record verified 2026-08-27

What was done

A randomized 2-arm crossover feeding trial was conducted in 113 healthy volunteers. Across high- or low-saturated fat arms, participants completed three 4-week isocaloric diets (separated by washout periods) with all meals prepared in a metabolic kitchen containing 25% of calories from protein derived from red meat, white meat, or non-meat sources. Plasma and urinary trimethylamine N-oxide (TMAO) and related metabolites were quantified at the end of each diet period. A subset of participants (n = 13) completed oral heavy isotope tracer studies to track TMA/TMAO synthesis from carnitine and choline.

What was found

Chronic red meat consumption increased plasma and urine TMAO by >two-fold compared with white meat or non-meat diets (P < 0.0001). Red meat ingestion reduced fractional renal excretion of TMAO (P < 0.05), while increasing fractional renal excretion of carnitine, gamma-butyrobetaine, and crotonobetaine (P < 0.05). In tracer studies, both red and white meat increased TMA and TMAO production from carnitine compared to non-meat protein (P < 0.05 each), with no difference observed from choline. Saturated fat content did not impact TMAO levels or metabolites. Discontinuation of red meat reduced plasma TMAO levels within 4 weeks.

Why it matters

This study demonstrates that chronic red meat consumption elevates systemic TMAO through a dual mechanism: enhancing gut microbial production from carnitine and reducing renal excretion of TMAO. It also demonstrates that this elevation is reversible within 4 weeks of stopping red meat intake.

Limits

The abstract reports relative changes rather than absolute plasma or urinary concentrations. The mechanistic isotope tracer sub-study had a small sample size (n = 13), and the trial evaluated intermediate circulating and excretory biomarkers rather than clinical cardiovascular events.

Cited by