Aims:
Australia has a relatively high level of alcohol consumption. Alcohol consumption is known to increase the risk of cancer and numerous causes of death internationally, but local evidence for Australia and evidence regarding pattern of drinking and mortality is limited. We aimed to evaluate the relationship between alcohol consumption, pattern of drinking and cause-specific mortality in a New South Wales (NSW) prospective cohort study.
Methods:
Cox proportional hazards regressions were used to calculate hazard ratios (HR) and 95% confidence intervals (CI) for cause-specific mortality risk in relation to overall alcohol consumption (drinks/week) and pattern of drinking among 180,575 of 267,357 participants aged ≥45 years (2005-2009) in the NSW 45 and Up Study. Pattern of drinking differentiated between participants who concentrated consumption on 1-3 days of the week from those who consumed alcohol on 4-7 days of the week. Cause of death was ascertained to 2019 by linkage to NSW Registry of Births Deaths and Marriages and Australian Bureau of Statistics Cause of Death Unit Record File by the Centre for Health Record Linkage (CHeReL). Secure data access was provided by the Sax Institute’s Secure Unified Research Environment (SURE). To limit bias from reverse causation, participants with pre-existing disease at baseline were excluded.
Results:
Over a median 11.4 years, 18,025 deaths were captured to 2019. Every additional seven drinks per week increased risk of death from any cause by 6% (HR=1.06; 95%CI=1.04-1.08). Increased risk of death was observed for alcohol-related cancers combined (HR=1.11; 95%CI=1.05-1.18), mouth, pharynx and larynx cancer (HR=1.31; 95%CI=1.10-1.56), liver cancer (HR=1.23; 95%CI=1.07-1.40), digestive system disease (HR=1.33; 95%CI=1.22-1.45), liver disease (HR=1.64; 95%CI=1.44-1.88), and falls (HR=1.23; 95%CI=1.03-1.46). Participants consuming as low as >3.5 to ≤10 drinks/week had significantly elevated risk of death from liver disease (HR=2.54; 95%CI=1.05-6.15) compared to light drinkers. Significant interactions with sex and smoking status were observed for certain causes of death, however variation in mortality risk by drinking pattern was not statistically significant (p-interaction=0.08). Results did not change materially when excluding the first three years of follow-up.
Discussion and Conclusion:
Excess risk of death from alcohol consumption in Australia is significant. As a nation with a relatively high intake of alcohol and a high proportion of drinkers, interventions aimed at reducing overall consumption may translate into significant public health gains.