Methyl-related nutrient quality and stroke: prevalence and stroke-specific mortality in U.S. adults.

Abstract

One-carbon (methyl-donor) nutrients support homocysteine remethylation and endothelial function, suggesting a role in cerebrovascular protection. The Methyl-Related Nutrient Quality Index (MNQI) summarizes adequacy of seven one-carbon-relevant nutrients. Its relationship with stroke outcomes in the general population is not well defined. We analyzed adults ≥ 20 years from NHANES 2005-2014 (n = 22,374) with two 24-hour dietary recalls. MNQI (0-10; higher = better) incorporated intake adequacy of protein, folate, choline, riboflavin (B2), vitamins B6 and B12, and zinc. Prior stroke was self-reported (MCQ160F). Stroke-specific mortality (primary outcome) used National Death Index linkage through December 31, 2019 (ICD-10 I60-I69); all-cause mortality was examined secondarily. Analyses were survey-weighted with multiple imputation. Prevalence ratios (PRs) were estimated using modified Poisson regression; hazard ratios (HRs) for mortality were from Cox models. Models adjusted for demographics, socioeconomic status, lifestyle, adiposity, comorbidities, and energy intake. Dose-response was assessed with restricted cubic splines; effect modification was explored in prespecified subgroups. Participants with higher MNQI (≥6 vs < 6) had a lower crude prevalence of prior stroke (2.6% vs 6.0%). In fully adjusted models, each 1-point higher MNQI was associated with 6% lower prevalence of prior stroke (PR 0.94, 95% CI 0.90-0.98; P = 0.004); MNQI ≥ 6 vs < 6 yielded PR 0.75 (95% CI 0.60-0.95). Over a median 7.8 years of follow-up, 3,054 all-cause and 173 S deaths occurred. MNQI was inversely associated with stroke-specific mortality (per 1-point HR 0.91, 95% CI 0.84-0.99; P = 0.022) but not with all-cause mortality (per 1-point HR 0.99, 95% CI 0.95-1.02; P = 0.48). Splines suggested an approximately linear association for stroke prevalence and a non-lin