Perreault · British journal of sports medicine 2017 · pooled prospective cohort study · n=36,370

Does physical activity moderate the association between alcohol drinking and all-cause, cancer and cardiovascular diseases mortality? A pooled analysis of eight British population cohorts.

Cited 50 times in the scientific literature.

Level 3 - non-randomized controlled study

Non-randomized prospective cohort study with follow-up linked to mortality data.

PubMed 27581162 · doi:10.1136/bjsports-2016-096194 · record verified 2026-08-29

What was done

A pooled prospective analysis of 8 British population-based surveys (Health Survey for England 1994, 1998, 1999, 2003, 2004, 2006; Scottish Health Survey 1998, 2003) linked to cause-specific mortality. The study evaluated 36,370 men and women aged 40 years and older (353,049 person-years of follow-up, 5,735 deaths). Alcohol intake was categorized into 6 sex-specific groups: never drunk, ex-drinkers, occasional drinkers, within guidelines (<14 UK units/week for women, <21 for men), hazardous (14–35 for women, 21–49 for men), and harmful (>35 for women, >49 for men). Physical activity (PA) was classified as inactive (≤7.5 MET-hour/week), active lower recommendation (>7.5 MET-hour/week), and active upper recommendation (>15 MET-hour/week). Cox proportional-hazard models and stratified analyses evaluated all-cause, cancer, and cardiovascular mortality risks after adjusting for confounders.

What was found

Alcohol consumption had a direct association with cancer mortality risk starting within guidelines, with an HR of 1.40 (95% CI: 1.11 to 1.78) for hazardous drinking. Stratified analyses showed that the association between alcohol intake and mortality risk was attenuated for all-cause mortality and nearly nullified for cancer mortality among participants meeting PA recommendations (exact stratified numerical effect sizes and confidence intervals were omitted from the abstract).

Why it matters

Physical activity may offset some of the cancer and all-cause mortality risks linked to alcohol consumption, offering a potential modifiable target for risk reduction.

Limits

Observational design prevents causal inference and remains susceptible to residual confounding. Exposure measures relied on self-reported survey data. Specific numerical HRs and confidence intervals for the stratified physical activity analyses were not provided in the abstract. Findings are derived from British adults aged 40 and older and may not generalize globally.

Cited by