Alcohol Consumption and Risk of Chronic Obstructive Pulmonary Disease

A Prospective Cohort Study of Men

Joanna Kaluza; Holly R. Harris; Anders Linden; Alicja Wolk


Am J Epidemiol. 2019;188(5):907-916. 

In This Article


The average age of men at baseline was 60.2 (SD, 9.7) years; 24.4% were current smokers, 38.5% were ex-smokers, and 35.8% were never smokers. The average frequency of total alcohol consumption among drinkers was 8.9 (SD, 12.7) drinks/week (median, 6.0 drinks/week), corresponding to approximately 15.2 (SD, 21.8) g of ethanol per day. Correlation coefficients for correlations between total alcohol consumption and consumption of beer, wine, and liquor were 0.30, 0.38, and 0.49, respectively. In comparison with never drinkers (0 drinks/week; 5.3%), men in the highest category of total alcohol consumption (>20.0 drinks/week; 7.6%) were younger, more likely to have a university education, and more likely to be current or ex-smokers (Table 1).

Total Alcohol Consumption and Risk of COPD

During 14.5 years of follow-up of (1998–2014; 643,572 person-years), we ascertained 2,177 incident cases of COPD. Compared with men with moderate total alcohol consumption (7.0–14.0 drinks/week), never drinkers (0 drinks/week) and those with high alcohol consumption (>20.0 drinks/week) had 21% higher (95% confidence interval (CI): −6, 57) and 34% higher (95% CI: 14, 57) risks of COPD, respectively (P for nonlinearity = 0.06) (Table 2). The risk of COPD was also significantly higher in past drinkers (hazard ratio (HR) = 1.38, 95% CI: 1.14, 1.68). When we conducted the analysis with inpatient-diagnosed COPD cases only, the results did not change substantially (see Web Table 1, available at When we analyzed categories of ethanol consumption in grams per day, results were similar (Table 2); the lowest risk of COPD was found in men who consumed 18.0–23.9 g of ethanol per day (vs. consumption of <6.0 g/day; HR = 0.81, 95% CI: 0.68, 0.96). A J-shaped association was observed between ethanol consumption and COPD incidence (P for nonlinearity = 0.003) (Figure 1).

Figure 1.

Multivariable-adjusted hazard ratios (solid line) for incident chronic obstructive pulmonary disease as a function of ethanol consumption (n = 44,254), Cohort of Swedish Men, 1998–2014. The dashed lines represent 95% confidence intervals. P for nonlinearity < 0.003.

To examine the impact of potential underdiagnosis during the early years of follow-up, we calculated hazard ratios according to time of diagnosis in 5- to 6-year periods (Web Table 2). The observed results suggested that during the first 5 years of follow-up (1998–2002), among men who consumed 14.1–20.0 drinks/week and >20.0 drinks/week, COPD might have been underdiagnosed; however, when we excluded the first 5 years of follow-up, the results did not change substantially (Web Appendix).

Specific Alcoholic Beverages and Risk of COPD

Age-adjusted characteristics of men by categories of consumption of specific alcoholic beverages are presented in Web Table 3. More men who consumed ≥1.0 drink/week of wine versus <1.0 drink/week had a university education, while those who consumed ≥1.0 drink/week of liquor versus <1.0 drink/week were more likely to be current smokers. We observed a J-shaped association of COPD incidence with beer consumption (P for nonlinearity < 0.001) and a U-shaped association with wine consumption (P for nonlinearity < 0.001; Web Figures 1A and 1B). Compared with never/occasional drinkers (<1.0 drink/week), men with consumption of 2.0–4.0 or 4.1–6.0 drinks/week of beer or 1.0–1.9 or 2.0–4.0 drinks/week of wine had a decreased risk of COPD (Table 3). No J-shaped association between liquor consumption and COPD risk was observed (Web Figure 1C, Table 3). A nonsignificantly increased risk of COPD with increasing liquor consumption was observed.

Web Figure 1.

Multivariable-adjusted hazard ratio (solid line) of chronic obstructive pulmonary disease (COPD) incidence as a function of A) beer consumption (P for non-linearity <0.001), B) wine consumption (P for non-linearity <0.001) and C) liquor consumption (P for non-linearity <0.028) in the Cohort of Swedish Men (n = 44 254, follow-up 1998–2014). The long dashed lines represent 95% confidence intervals. The distribution of alcoholic beverages consumption is presented at the bottom of the figure as a histogram. Standard drink (~12 g alcohol) = 80 ml strong wine, 150 ml wine, 330 ml beer >3.5% alcohol, 500 ml beer 2.8% to 3.5%, or 660 ml beer <2.25%.

Alcohol Consumption and COPD Risk by Smoking Status and Other Covariates

Although we did not observe a statistically significant interaction between alcohol consumption and smoking status in relation to COPD incidence (P for interaction = 0.55), we stratified results by smoking status, since smoking is the primary risk factor for COPD development (Table 4). We observed that the risk of COPD in current smokers with high total alcohol consumption (>20.0 drinks/week) was 32% (95% CI: 8, 62) higher than that in men with moderate consumption (7.0–14.0 drinks/week). In ex-smokers and never smokers, both never/occasional drinkers and heavier drinkers had a tendency towards higher risks of COPD. Hazard ratios for COPD incidence by age, education, body mass index, and physical activity are presented in Web Table 4, Web Table 5, Web Table 6 and Web Table 7.