BMC Public Health · Nov 18, 2026

Meat preference–intake discordance and incident cardiometabolic disease in UK Biobank: a prospective cohort study

By Zhu Q, Tan Y, Yang Y et al.

Background Dietary intake and food preference are commonly examined separately in nutritional epidemiology. Whether discordance between meat preference and intake is associated with incident cardiometabolic disease (CMD) remains unclear. Methods We included 168,163 UK Biobank participants without prevalent CMD. Meat preference was assessed using the Food Preference Questionnaire (FPQ), and habitual meat intake was derived from repeated food-frequency assessments. Preference–intake discordance was calculated as the standardised preference score minus the standardised intake score. Cox models, intake-stratified analyses, generalised random forests, counterfactual mediation analyses, and sensitivity analyses were used to examine incident CMD and potential variation in associations. Results During follow-up, 39,428 incident CMD cases occurred. Overall meat preference > intake was not significantly associated with CMD after full adjustment (HR 1.027, 95% CI 0.994–1.061). Preference > intake for beef, lamb, and pork was associated with higher CMD risk (HRs 1.045, 1.041, and 1.052), especially for hypertension and type 2 diabetes. Overall discordance was associated with CMD among low meat consumers (HR 1.078, 95% CI 1.022–1.136). HbA1c was a consistent positive mediator; liver- and bilirubin-related markers showed negative indirect effects suggestive of statistical suppression. Conclusions Meat preference-intake discordance was not uniformly associated with CMD risk; relevance varied by meat subtype and intake context. Preference-aware dietary assessment may refine individualised CMD risk evaluation.

doi.org/10.1186/s12889-026-30314-yOpen accessPopulations

AgingResearch.News