Vincent · The American journal of clinical nutrition 2019 · Bayesian meta-regression of randomized controlled trials · n=55 studies (2,652 subjects)

Meta-regression analysis of the effects of dietary cholesterol intake on LDL and HDL cholesterol.

Cited 106 times in the scientific literature.

Level 1 - systematic review of randomized trials

Meta-regression analysis of 55 randomized dietary intervention trials.

PubMed 30596814 · doi:10.1093/ajcn/nqy273 · record verified 2026-08-30

What was done

Bayesian meta-regression using Markov chain Monte Carlo methods evaluated the dose-response relationship between dietary cholesterol intake (0–1500 mg/day) and circulating LDL and HDL cholesterol. The analysis included 55 randomized dietary intervention trials comprising 2,652 subjects and adjusted for concurrent changes in dietary fatty acid intake using linear, Michaelis-Menten (MM), and Hill nonlinear models.

What was found

Nonlinear MM and Hill models best described the relationship across 0–1500 mg/day of dietary cholesterol. For every 100 mg/day increase in dietary cholesterol, the predicted mean change in LDL cholesterol was +1.90 mg/dL in the linear model, +4.46 mg/dL in the MM model, and +4.58 mg/dL in the Hill model. Dietary cholesterol change was modestly inversely associated with HDL cholesterol in men, but positively associated with HDL cholesterol in women.

Why it matters

This study provides an adjusted dose-response quantification showing that dietary cholesterol increases LDL cholesterol independently of dietary fatty acids, fitting best to nonlinear saturation curves.

Limits

The abstract reports only surrogate lipid markers rather than hard cardiovascular events. Specific confidence intervals and numerical effect sizes for the sex-specific HDL findings are not reported in the abstract, and the clinical significance of the observed HDL divergence remains unresolved.

Cited by