Alcohol consumption and all-cause mortality in older adults in Spain: an analysis accounting for the main methodological issues
Background and aims: Observational evidence that light-to-moderate alcohol consumption lowers mortality is questioned because of potential selection biases and residual confounding. We assess the association between alcohol intake and all-cause death in older adults after accounting for those methodological issues. Methods: Data came from 3045 individuals representative of the non-institutionalized population aged ≥ 60 years in Spain. Participants were recruited in 2008–10, when they reported current and life-time alcohol intake; drinkers were classified as occasional (< 1.43 g/day), light (≥ 1.43 but < 20 g/day for men and ≥ 1.43 but < 10 g/day for women), moderate (≥ 20 but < 40 g/day for men and ≥ 10 but < 20 g/day for women) or heavy (≥ 40 g/day for men and ≥ 24 g/day for women)/binge. Participants were followed-up to 2017 to assess vital status. In analyses, ex-drinkers were removed from the abstainer group and were classified according to their life-time intake to address the ‘abstainer bias’. Moreover, analyses were replicated in individuals without functional limitations, and excluded deaths in the first year of follow-up, to address reverse causation. Also, occasional drinkers were used as reference in some analyses to reduce the ‘healthy drinker/survivor’ bias. Results were adjusted for many covariates to minimize residual confounding. Results: Compared with never-drinkers, the hazard ratio (95% confidence interval) of mortality for light drinkers was 1.05 (0.71–1.56) and 1.20 (0.72–2.02) in those without functional limitations. Corresponding values for moderate drinkers were 1.28 (0.81–2.02) and 1.55 (0.87–2.75) and for heavy/binge drinkers 1.85 (1.07–3.23) and 2.15 (1.09–4.22). Results were consistent when occasional drinkers were used as reference. Among drinkers without functional limitations, the hazard ratio (95% confidence interval) of mortality per 10 g/day of alcohol was 1.12 (1.02–1.23). Conclusion: After accounting for potential biases, light-to-moderate drinking among people 60+ years of age appears to have no statistically significant benefit on mortality compared with abstention from alcohol. By contrast, heavy/binge drinking shows a higher death risk compared with abstention from alcohol. Alcohol intake appears to have a positive dose–response with mortality among drinkers. © 2018 Society for the Study of Addiction
| Main Authors: | , , , , , |
|---|---|
| Other Authors: | |
| Format: | artículo biblioteca |
| Language: | English |
| Published: |
Wiley-Blackwell
2019-01-01
|
| Subjects: | Alcohol, Mortality, Confounding, Reverse causation, Older adults, |
| Online Access: | http://hdl.handle.net/10261/201013 http://dx.doi.org/10.13039/501100000780 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Background and aims: Observational evidence that light-to-moderate alcohol consumption lowers mortality is questioned because of potential selection biases and residual confounding. We assess the association between alcohol intake and all-cause death in older adults after accounting for those methodological issues. Methods: Data came from 3045 individuals representative of the non-institutionalized population aged ≥ 60 years in Spain. Participants were recruited in 2008–10, when they reported current and life-time alcohol intake; drinkers were classified as occasional (< 1.43 g/day), light (≥ 1.43 but < 20 g/day for men and ≥ 1.43 but < 10 g/day for women), moderate (≥ 20 but < 40 g/day for men and ≥ 10 but < 20 g/day for women) or heavy (≥ 40 g/day for men and ≥ 24 g/day for women)/binge. Participants were followed-up to 2017 to assess vital status. In analyses, ex-drinkers were removed from the abstainer group and were classified according to their life-time intake to address the ‘abstainer bias’. Moreover, analyses were replicated in individuals without functional limitations, and excluded deaths in the first year of follow-up, to address reverse causation. Also, occasional drinkers were used as reference in some analyses to reduce the ‘healthy drinker/survivor’ bias. Results were adjusted for many covariates to minimize residual confounding. Results: Compared with never-drinkers, the hazard ratio (95% confidence interval) of mortality for light drinkers was 1.05 (0.71–1.56) and 1.20 (0.72–2.02) in those without functional limitations. Corresponding values for moderate drinkers were 1.28 (0.81–2.02) and 1.55 (0.87–2.75) and for heavy/binge drinkers 1.85 (1.07–3.23) and 2.15 (1.09–4.22). Results were consistent when occasional drinkers were used as reference. Among drinkers without functional limitations, the hazard ratio (95% confidence interval) of mortality per 10 g/day of alcohol was 1.12 (1.02–1.23). Conclusion: After accounting for potential biases, light-to-moderate drinking among people 60+ years of age appears to have no statistically significant benefit on mortality compared with abstention from alcohol. By contrast, heavy/binge drinking shows a higher death risk compared with abstention from alcohol. Alcohol intake appears to have a positive dose–response with mortality among drinkers. © 2018 Society for the Study of Addiction |
|---|