1. Introduction

Biological ageing can be quantified by DNA methylation “epigenetic clocks,” which estimate an individual’s biological age from specific methylation patterns. These clocks often aggregate the methylation levels at select CpG sites to predict chronological age with reasonable accuracy. For example, Horvath’s pan-tissue clock ^[1] uses 353 CpGs and has a Pearson correlation ~0.96 with age, while Hannum’s blood clock ^[2] uses 71 CpGs with similar accuracy. Recently, Collins ^[3] described a cost-effective saliva-based epigenetic clock built on just 10 CpG sites identified from a large dataset of 3,408 individuals. This saliva 10-CpG clock showed a strong correlation (r ≈ 0.80) between DNA methylation age and chronological age, with a mean absolute error of ~5.5 years. All ten clock CpGs exhibited age-related hypermethylation (in genes such as ELOVL2, OTUD7A, etc.), underscoring their relevance to aging pathways. Such minimally-invasive saliva clocks enable biological age assessments outside clinical settings and can be used to explore environmental and lifestyle influences on aging.

Lifestyle factors are suspected to modulate epigenetic aging. Prior studies have linked excessive alcohol consumption to faster epigenetic aging. Certain studies ^[4] found that excessive alcohol consumption and a diagnosis of alcohol dependence were associated with [epigenetic age acceleration] across different tissues, including blood, with a large population study reporting that higher long-term alcohol intake was significantly associated with greater DNA methylation age acceleration (especially GrimAge and PhenoAge clocks) in middle-aged and older individuals. In contrast, younger adults did not show this effect, suggesting that cumulative exposure over time may be key ^[4]. Heavy drinking has also been linked to numerous age-related diseases (e.g. hypertension, liver disease, and cancers), reinforcing the notion that alcohol might broadly accelerate aging processes. Epigenetic age acceleration (EAA), defined as a positive deviation of biological age from chronological age, is itself associated with age-related morbidity and mortality risk in multiple studies ^[5]. However, whether high habitual alcohol consumption in generally healthy adults measurably accelerates epigenetic aging (particularly as detected in saliva) remains an important question.

In this  study, we leverage the Muhdo Health DNA methylation database to examine the relationship between alcohol intake and biological age. The study focuses on two extreme consumption groups: individuals drinking well above recommended limits versus those who consume very little alcohol. We hypothesise that heavy drinkers will exhibit accelerated epigenetic age as measured by the saliva 10-CpG clock, compared to low-consuming peers. If confirmed, this would provide a clear epigenetic link between excessive alcohol use and premature biological aging, with implications for public health guidance.

2. Materials and Methods

Study Population: Participants were drawn from the Muhdo Health Data Repository, a database of volunteers who provided saliva samples and lifestyle information for DNA methylation analysis. For this study, we defined two cohorts aged 30–60 years with contrasting alcohol use. Cohort 1 (High Alcohol Consumption) consisted of 66 adults (60% male) who reported drinking more than 38 units of alcohol per week. This level (>304 g ethanol/week) is nearly three times the recommended maximum in the UK and thus represents a heavy/excessive intake. Cohort 2 (Low Alcohol Consumption) included 109 adults (54% male) who reported consuming fewer than 6 units per week (<48 g ethanol/week), effectively a light drinking or near-abstinent group. Participants with diagnosed chronic diseases were excluded to focus on generally healthy individuals. The two cohorts were similar in age range (both 30–60 by design) and had no overlap; “complement group” refers to the low-intake cohort serving as a comparison to the high-intake focus group.

Ethics and Data Collection: All individuals provided informed consent for the use of their anonymised data. Self-reported alcohol consumption (units per week) was obtained via lifestyle questionnaires at enrolment. Saliva samples were self-collected using kits provided by Muhdo Health and mailed to the laboratory. Basic demographics (age, sex) accompanied each sample record.

DNA Methylation Analysis: Genomic DNA was extracted from saliva samples for methylation analysis. Genome-wide DNA methylation profiling was performed by Eurofins (Denmark) using the Illumina Infinium MethylationEPIC BeadChip (850,000 CpG sites) following standard protocols. For each individual, DNA methylation age was estimated with the validated saliva-based 10-CpG epigenetic clock model. Briefly, this model uses a linear combination of methylation beta values at 10 specific CpG sites to predict age in years. The CpGs (e.g., in genes ELOVL2, PRLHR, GPR158, etc.) were originally selected for their strong age-correlation in saliva DNA. The clock formula (coefficients applied to each methylation value plus an intercept) was implemented as described in the original publication. Predicted biological age was thus obtained for every participant.

Epigenetic Age Acceleration (EAA): We defined EAA as the difference between a person’s DNA methylation-predicted age and their actual chronological age (Predicted Age – Chronological Age). A positive EAA indicates that the individual’s epigenetic age is higher than expected for their chronological age (accelerated aging), whereas a negative value indicates a younger biological age than expected. We also calculated percent acceleration (EAA / chronological age * 100%) for descriptive purposes. In secondary analyses, we computed EAA as the residual from a linear regression of DNAm age on chronological age within the combined sample; this residual-based approach yielded very similar results to the simple age difference.

Statistical Analysis: Group comparisons were conducted to test the hypothesis that high alcohol consumers have higher EAA than low consumers. We first compared baseline characteristics between the two cohorts. To specifically assess the effect of alcohol group on epigenetic aging, we performed an unpaired t-test on EAA values between the high and low alcohol groups. We also fit a linear regression model with EAA as the outcome, alcohol group (high vs low) as the predictor, and included sex and chronological age as covariates to adjust for any small differences in these factors. The regression tested whether high alcohol intake was an independent predictor of accelerated epigenetic age. Significance was evaluated at p < 0.05 (two-tailed). Additionally, Pearson correlation analysis was used to explore the association between self-reported weekly alcohol units (treated as a continuous variable across all participants) and EAA. Statistical analyses were performed in R (v4.2.0) and checked for assumptions (normality of residuals, etc.). Results are presented as means ± standard deviations or percentages.

Table 1.

Demographic and age-related characteristics of high vs. low alcohol consumption cohorts.

Note: Biological age was estimated from saliva DNA methylation using the 10-CpG epigenetic clock. Epigenetic age acceleration is the difference between biological and chronological age. High alcohol group consumes >38 units/week; low alcohol group <6 units/week. DNAm: DNA methylation.

3. Results

Cohort Characteristics: A total of 175 individuals (66 high alcohol, 109 low alcohol) were analysed. Table 1 summarises the demographic and age-related metrics of the two cohorts. The high-consumption and low-consumption groups were of comparable age range (30–60 years in both). The mean chronological age was slightly lower in the high alcohol group (43.0 ± 5.1 years) than in the low alcohol group (45.0 ± 5.3 years), a difference of ~2 years (this difference reached marginal statistical significance, p ≈ 0.02). Sex distributions were similar: 60% of the heavy drinkers were male versus 54% of the low-intake group (χ² test, p = 0.45), indicating no significant sex imbalance. By design, weekly alcohol consumption differed starkly: the high alcohol group reported on average over 38 units per week (exceeding ~3 times the recommended limit), whereas the low group consumed under 6 units per week. Thus, the cohorts represent distinctly different alcohol exposure levels.

Epigenetic Clock Measurements: DNA methylation analysis was successful for all samples. Each individual’s saliva DNA methylation age was computed using the 10-CpG epigenetic clock. The clock output ranged from lower 30s to upper 50s in age predictions for this mid-adult sample. In the low alcohol group, the mean DNA methylation-predicted age was 45.1 years (approximately equal to their mean chronological age of 45.0), suggesting no overall age acceleration in this group. In contrast, the high alcohol group showed a markedly elevated biological age: mean DNAm age was 51.0 years versus a mean chronological age of 43.0 years. This implies that, on average, heavy drinkers were about 8 years “older” biologically than their actual age.

Table 2

. Methylation means for both cohorts.

Group Differences in Epigenetic Age Acceleration: The primary analysis confirmed a significant difference in epigenetic age acceleration between the two groups. Heavy alcohol consumers had a mean EAA of +8.0 years (SD ≈ 7.4 years, as estimated from the variance in DNAm age predictions), indicating substantial age acceleration. Low alcohol consumers had a mean EAA of approximately 0.0 years (SD ≈ 5.6 years), i.e. on average no acceleration or even a slight negative trend in some cases. An unpaired t-test showed that the high alcohol group’s EAA was significantly greater than that of the low group (t ≈ 2.32, df ≈ 173, p = 0.021). In percentage terms, the heavy drinking cohort’s biological age was about 19% older than their chronological age on average, whereas the low-intake cohort’s biological age was essentially equal (0.0% difference).

Correlation of Alcohol Intake with EAA: Within the combined sample of 175 individuals, self-reported alcohol consumption (in units per week) was positively correlated with epigenetic age acceleration (Pearson r ≈ 0.25, p = 0.0012). In a subgroup analysis, we observed a dose-response trend: individuals drinking 0–5 units/week had a mean EAA ~ –0.5 years (slightly younger than expected), those drinking moderately (5–20 units/week) had mean EAA ~ +1–2 years, and those in the highest bracket (>20 units, which overlaps with our heavy group) had progressively higher mean EAA (approaching +8 years for >38 units). This gradient, while based on cross-sectional data, supports a quantitative relationship between alcohol dose and biological aging. It is worth noting that not all heavy drinkers showed high EAA—there was variability, with some heavy consumers having near-normal epigenetic ages. Conversely, a minority of low consumers exhibited positive EAA (possibly due to other unmeasured factors). Overall, however, the data clearly indicate a tendency for greater alcohol exposure to relate to older epigenetic age.

4. Discussion

In this study, we found that chronic heavy alcohol consumption is associated with significant acceleration of epigenetic age, as measured by a saliva-based DNA methylation clock. Heavy drinkers (>38 units/week) were biologically older by approximately 8 years on average compared to their chronological age, whereas light drinkers (<6 units/week) showed no such age acceleration. To our knowledge, this is one of the first studies to specifically utilise a saliva DNA methylation clock to evaluate alcohol-related aging in healthy adults. Our findings are consistent with prior observations in blood-based epigenetic studies and extend them to a non-invasive saliva context.

The magnitude of the effect observed here is notable. An 8-year epigenetic age gap between heavy and low alcohol users (in mid-life) suggests that excessive drinking may prematurely age individuals by nearly a decade. This result aligns with the growing literature linking alcohol to biological aging. For example, a recent large-cohort analysis [4] similarly reported that higher long-term alcohol intake was significantly associated with faster epigenetic aging in middle-aged and older adults. In that study, each additional daily drink (14 g alcohol) was associated with ~0.7 years of age acceleration on certain clocks. Our heavy drinking group exceeded moderate drinking significantly, which would extrapolate to ~3–4 years of acceleration– in reasonable agreement with the 6–8 year elevation we observed, considering differences in population and clock type. Another study of individuals with alcohol use disorder (AUD) found that epigenetic aging (using PhenoAge clock) was about 3.6 years faster in AUD patients than in controls ^[6]. The somewhat larger age gap in our sample may reflect that our “high alcohol” cohort, while not all clinically diagnosed with AUD, still consumed very high levels; importantly, we compared them to a strictly low-intake group, maximising contrast. It is also plausible that saliva-based clocks might capture certain lifestyle-induced age effects differently than blood-based clocks. Saliva contains immune cells and buccal cells that could be influenced by alcohol-related stress (e.g., inflammation in the oral environment), potentially amplifying the aging signal in that tissue.

Alcohol is a pro-oxidant and can induce systemic oxidative stress and inflammation, which are well-known drivers of aging at the cellular and molecular level. Chronic alcohol use damages DNA and impairs DNA repair pathways, potentially leading to shifts in methylation patterns at age-related CpGs. Indeed, the 10 CpG sites comprising the saliva clock include several (such as ELOVL2, GPR158) that have been implicated in age and may be sensitive to oxidative damage or metabolic changes. The study observed uniformly higher methylation at all these CpGs in heavy drinkers compared to light drinkers, mirroring what is normally seen in older chronological age. In essence, heavy alcohol use seems to push the epigenome towards an “older” state. Additionally, heavy alcohol consumption can cause hormonal and metabolic dysregulation (for example, affecting the HPA axis, insulin signalling, etc.) that might accelerate aging processes. It also contributes to telomere shortening, another marker of cellular aging, as reported in some studies of alcoholics. Our findings specifically highlight epigenetic changes: it appears that alcohol-induced ageing signals are embedded in DNA methylation marks that are captured by the clock algorithm.

The public health implications of these results are important. First, epigenetic age acceleration has been linked to increased risks of age-related diseases (e.g., cognitive decline, cardiovascular disease, cancers) and all-cause mortality ^[4], ^[6]. An individual who is biologically 8 years older than their chronological age may face higher odds of developing chronic conditions sooner. Our study provides a potential biomarker for quantifying the insidious impact of heavy alcohol on the body’s aging trajectory. Notably, many people who drink heavily might not immediately feel unwell, but their methylation age could be silently increasing, flagging higher long-term health risk. This kind of metric could be used in preventive health strategies – for instance, wellness programs could monitor epigenetic age as feedback to motivate lifestyle changes. If a person sees that their “biological age” is significantly older than their actual age, it might encourage interventions such as alcohol reduction, improved diet, or exercise to slow the aging process.

Strengths and Limitations: A strength of our study is the use of a novel saliva-based epigenetic clock, which allowed non-invasive sampling and may better reflect certain lifestyle effects (since saliva contains immune cells that interact with environmental exposures). We also specifically targeted two distinct groups at the extremes of alcohol consumption, which enhances contrast and the ability to detect differences. However, there are important limitations. The study is observational and cross-sectional, so we cannot establish causality — it is possible that people who biologically age faster might also tend to drink more (though the bulk of evidence suggests the direction of effect is from alcohol to ageing). We relied on self-reported alcohol use, which can be prone to underreporting or error; however, the categories (>38 vs <6 units) are broad enough that misclassification is likely minimal (it’s unlikely a heavy drinker would report <6, for instance). Our sample size (n=175 total) is moderate; a larger sample would improve power and allow more nuanced analyses (e.g., sex-specific effects, threshold effects). We did not analyse detailed data on other lifestyle and health factors such as smoking, diet, or body mass index (BMI).

5. Conclusions

In summary, this study provides evidence that high levels of alcohol consumption are associated with accelerated biological aging as measured by DNA methylation. Using a saliva-based 10-CpG epigenetic clock, we found that individuals drinking significantly above recommended limits had on average eight years of excess epigenetic age compared to low-consuming individuals of the same chronological age. These findings reinforce the message that heavy alcohol use can prematurely age the body and potentially increase the risk of age-related diseases. On a positive note, because epigenetic modifications are dynamic, reducing alcohol intake may help slow down or even reverse some of this accelerated ageing.