Associations of alcohol intake with subclinical carotid atherosclerosis in 22,000 Chinese adults

Abstract

Background and aims: We investigated the causal relevance of alcohol intake with measures of carotid artery thickness and atherosclerosis in Chinese adults.

Methods: The study included 22,384 adults from the China Kadoorie Biobank, with self-reported alcohol use at baseline and resurvey, carotid artery ultrasound measurements, and genotyping data for ALDH2-rs671 and ADH1B-rs1229984. Associations of carotid intima media thickness (cIMT), any carotid plaque, and total plaque burden (derived from plaque number and size) with self-reported (conventional analyses) and genotype-predicted mean alcohol intake (Mendelian randomization) were assessed using linear and logistic regression models.

Results: Overall 34.2% men and 2.1% women drank alcohol regularly at baseline. Mean cIMT was 0.70 mm in men and 0.64 mm in women, with 39.1% and 26.5% having carotid plaque, respectively. Among men, cIMT was not associated with self-reported or genotype-predicted mean alcohol intake. The risk of plaque increased significantly with self-reported intake among current drinkers (odds ratio 1.42 [95% CI 1.14-1.76] per 280 g/week), with directionally consistent findings with genotype-predicted mean intake (1.21 [0.99-1.49]). Higher alcohol intake was significantly associated with higher carotid plaque burden in both conventional (0.19 [0.10-0.28] mm higher per 280 g/week) and genetic analyses (0.09 [0.02-0.17]). Genetic findings in women suggested the association of genotype-predicted alcohol with carotid plaque burden in men was likely to due to alcohol itself, rather than pleiotropic genotypic effects.

Conclusions: Higher alcohol intake was associated with a higher carotid plaque burden, but not with cIMT, providing support for a potential causal association of alcohol intake with carotid atherosclerosis.

Keywords: Alcohol; Alcohol dehydrogenase; Aldehyde dehydrogenase; Atherosclerosis; Carotid intima media thickness; Carotid plaque; Mendelian randomization.

Figure 1. Associations of carotid measurements with self-reported alcohol consumption and with genotype-predicted mean alcohol intake, in men

Conventional epidemiological analyses (A–C) of baseline self-reported drinking patterns with cIMT (A), presence of carotid plaque (B), and carotid plaque burden (C) in men without prior cardiovascular disease. The reference group was non-drinkers and results were adjusted for age, area, education, income, and smoking. The means or odds ratios for current drinkers were plotted against usual alcohol intake, with a fitted line giving the mean change or odds ratio (95% CI) per 280 g intake per week usual alcohol intake. Genetic epidemiological analyses (D–F) of genotype predicted mean alcohol intake with cIMT (D), presence of carotid plaque (E), and carotid plaque burden (F) in all men. Results were adjusted for age, area, and genomic principal components. The means or odds ratios were plotted against genotype-predicted mean alcohol intake, with the mean change or odds ratio (95% CI) per 280 g intake per week genotype-predicted mean alcohol intake calculated within study areas and combined by inverse variance-weighted meta-analysis. The area of each square is inversely proportional to the variance of the least square mean in (A, C, D, F), and the variance of the log odds in (B, E). The group-specific 95% CIs, calculated from this variance, are shown by error bars. cIMT: carotid intima media thickness; CI: confidence interval.