Characteristics of the included studies
| Study and year | Study design | Geographical location | Alcohol ascertainment | Alcohol definition | CAP ascertainment | Confounders adjusted | Effect estimate |
| Almirall et al,45 1999 | Case–control | Europe/Spain | Self-report/questionnaire | Quartiles of alcohol intake >35.3 vs 0 (g/day) | Clinically suspected and chest radiography | Age, sex, municipality | OR |
| Almirall et al,44 2008 | Case–control | Europe/Spain | Self-report/questionnaire | Quartiles of alcohol intake (g/day) Men: >80 vs 0 Women: >40 vs 0 | Clinically suspected and chest radiography | Age, sex, primary care practice | OR |
| Baik et al,10 2000 | Cohort | America/USA | Self-report/questionnaire | Men: >30 vs never Women: >30 vs never (g/day) | Physician diagnosis/medical records | Age, smoking status, BMI, quintile of metabolic equivalent | Relative risk |
| Breitling et al,39 2016 | Cohort | America/USA | Self-report/questionnaire | Quartiles of alcohol intake (g/day) Men: >20 vs ≤20 Women: >10 vs 0 | Self-report questionnaire | Age, sex, smoking, BMI, diabetes mellitus, stroke, congestive heart failure, cancer | Relative risk |
| Clough et al,42 2003 | Case–control | Australia | Self-report/interview | Alcohol yes vs alcohol no | Clinically suspected/X-ray findings | –* | OR |
| Fernández-Solá et al,43 1995 | Case–control | Europe/Spain | Self-report/interview and questionnaire | High intake (men: >100 g, women: >80 g) vs lower intake (g/day 2 years before submission) | Clinically suspected/chest X-ray | Liver cirrhosis, smoking, COPD, diabetes, heart failure, malnutrition | OR |
| Inoue et al,36 2007 | Cohort | Asia/Japan | Self-report/questionnaire | Current vs never drinking | Mortality ICD codes | Age and history of diabetes mellitus | HR |
| Jackson et al,46 2009 | Case–control | America/USA | Medical records | Current alcoholism vs no alcoholism | ICD-9 codes | Age, sex, pneumonia-free person-time | OR |
| Koivula et al,37 1994 | Cohort | Europe/Finland | Medical records | Alcoholism vs no alcoholism | Medical records | Age, sex, chronic conditions | Relative risk |
| Lipsky et al,11 1986 | Case–control | America/USA | Medical records | Heavy vs moderate (drinks/day) | Clinically suspected | –* | Relative risk |
| Loeb et al,48 2009 | Case–control | America/USA | Self-report/questionnaire | Alcohol yes (previous 12 months) vs alcohol no (g/month) | Clinically suspected and chest radiography | Multivitamins, smoking, history of gas and fumes exposure | OR |
| Phung and Wang,38 2013 | Cohort | Australia | Self-report/questionnaire | Alcohol yes vs alcohol no | Hospital records—ICD codes | –* | HR |
| Quraishi et al,49 2013 | Cross-sectional | America/USA | Self-report/interview | Alcohol consumption (≤30 vs >30 drinks per month) | Self-report interview | –* | Relative risk |
| Shen et al,35 2013 | Cohort | Asia/China | Self-report/interview | Excessive vs never drinkers (units/week) | Mortality ICD codes | Age, sex, education, housing, monthly expenditure, smoking, BMI, exercise, health status | HR |
| Watt et al,47 2007 | Case–control | America/USA | Medical records | Alcoholism/alcohol use vs no use of alcohol | Clinically suspected pneumococcal isolation in patient from sterile body fluid | Smoking, BMI, electricity/indoor plumbing in home, living with unvaccinated child, unemployed, wood/coal, smoke | OR |
| Yende et al,40 2013 | Cohort | America/USA | Medical records | Alcohol abuse vs no alcohol abuse | ICD-9 codes | –* | Relative risk |
| Zaridze et al,41 2009 | Case–control | Europe/Russia | Self-report interview | ≥3 bottles(per week) vs ≤0.5 bottles of vodka | ICD codes Death records | Age, city and smoking | Relative risk |
*Crude analysis reported.
BMI, body mass index; CAP, community-acquired pneumonia; COPD, chronic obstructive pulmonary disease; ICD, International Classification of Diseases.