Associations between alcohol use and accelerated biological ageing

Abstract

Harmful alcohol use is a leading cause of premature death and is associated with age-related disease. Biological ageing is highly variable between individuals and may deviate from chronological ageing, suggesting that biomarkers of biological ageing (derived from DNA methylation or brain structural measures) may be clinically relevant. Here, we investigated the relationships between alcohol phenotypes and both brain and DNA methylation age estimates. First, using data from UK Biobank and Generation Scotland, we tested the association between alcohol consumption (units/week) or hazardous use (Alcohol Use Disorders Identification Test [AUDIT] scores) and accelerated brain and epigenetic ageing in 20,258 and 8051 individuals, respectively. Second, we used Mendelian randomisation (MR) to test for a causal effect of alcohol consumption levels and alcohol use disorder (AUD) on biological ageing. Alcohol use showed a consistent positive association with higher predicted brain age (AUDIT-C: β = 0.053, p = 3.16 × 10^-13 ; AUDIT-P: β = 0.052, p = 1.6 × 10^-13 ; total AUDIT score: β = 0.062, p = 5.52 × 10^-16 ; units/week: β = 0.078, p = 2.20 × 10^-16 ), and two DNA methylation-based estimates of ageing, GrimAge (units/week: β = 0.053, p = 1.48 × 10^-7 ) and PhenoAge (units/week: β = 0.077, p = 2.18x10^-10 ). MR analyses revealed limited evidence for a causal effect of AUD on accelerated brain ageing (β = 0.118, p = 0.044). However, this result should be interpreted cautiously as the significant effect was driven by a single genetic variant. We found no evidence for a causal effect of alcohol consumption levels on accelerated biological ageing. Future studies investigating the mechanisms associating alcohol use with accelerated biological ageing are warranted.

Keywords: Generation Scotland; Mendelian randomisation; UK Biobank; alcohol use; brain ageing; epigenetic ageing.

Figure 1

Alcohol use is associated with advanced brain age. Linear regression models predicting residual brain age from AUDIT-C, AUDIT-P, AUDIT-T and alcohol units, in current drinkers adjusted for smoking status. Plot shows standardised β coefficients with 95% confidence intervals. CI, confidence interval

Figure 2. Alcohol consumption is associated with two measures of advanced epigenetic age.

Effects of alcohol consumption (units/week) on (A) EEAA, (B) IEAA, (C) AgeAccelGrim and (D) AgeAccelPheno in fully adjusted models. Values on forest plot indicate standardised β with 95% confidence intervals. Models are adjusted for sex, BMI and pack-years in Sets 1 and 2 and relatedness in Set 1 by fitting pedigree information as a random effect in general linear mixed models using advanced restricted maximum likelihood (ASReml) method. Fixed-effects inverse variance-weighted meta-analysis was applied using R package meta to combine the standardised coefficient estimates in Sets 1 and 2. FDR correction was applied across all models in Sets 1 and 2 and all meta-analysis models (12 models in total). Sample size: n = 4260 in Set 1, n = 3791 in Set 2 (n = 8051 included in meta-analyses). CI, confidence interval; EEAA, extrinsic epigenetic age acceleration; FDR, false discovery rate; IEAA, intrinsic epigenetic age acceleration; SE, standard error

Figure 3. Alcohol consumption is associated with advanced GrimAge and PhenoAge in non-smokers.

Effects of alcohol consumption (units/week) on (A) AgeAccelGrim and (B) AgeAccelPheno in non-smoking participants. Values on forest plot indicate standardised β with 95% confidence intervals. Models are adjusted for sex and BMI in Sets 1 and 2 and relatedness in Set 1 by fitting pedigree information as a random effect in general linear mixed models using advanced restricted maximum likelihood (ASReml) method. Fixed-effects inverse variance-weighted meta-analysis was applied using R package meta to combine the standardised coefficient estimates in Sets 1 and 2. FDR correction was applied across all smoking sensitivity models in Sets 1 and 2 and all meta-analysis models (six models in total). Sample size: n = 2207 in Set 1, n =1998 in Set 2 (n = 4205 included in meta-analyses). CI, confidence interval; EEAA, extrinsic epigenetic age acceleration; FDR, false discovery rate; IEAA, intrinsic epigenetic age acceleration; SE, standard error

Figure 4. Two-sample Mendelian randomisation analysis provides weak evidence for a causal effect of AUD on brain age acceleration.

(A) Two sample Mendelian randomisation of AUDIT-C on brain age. (B) Two-sample Mendelian randomisation of AUD on brain age. Data on the genetic association with AUDIT-C and AUD were extracted from Kranzler et al. Summary statistics for these SNPs were extracted from a novel GWAS of brain age (see Section 2). AUD, alcohol use disorder; CI, confidence interval; N SNP, number of SNPs included in the MR analysis

Figure 5. Two-sample Mendelian randomisation analysis shows no evidence for causal effect of alcohol consumption on measures of epigenetic age acceleration

(A) AgeAccelGrim and (B) AgeAccelPheno. Data on the genetic association with alcohol use (AUDIT-C) were extracted from Kranzler et al. Summary statistics for these SNPs were extracted from GWASs for AgeAccelGrim and AgeAccelPheno conducted by McCartney et al. rs185177474 was not available in AgeAccelGrim and AgeAccelPheno summary statistics; thus, rs151242810 was used as a proxy (R² = 0.976; see Section 2). CI, confidence interval; N SNP, number of SNPs included in the MR analysis