Table 1

Characteristics of seven cohort studies predicting all-cause mortality through Alcohol Use Disorders Identification Test (AUDIT) scores, 1989–2016

StudyDesignSampleMeasurementsQuality rating
(NOS)
BaselineFollow-upAverage length*LocationSettingSex; age mean; range*Total N
(deaths)
Exposure, categoriesMortalityAssociationAdjustments
Bradley et al 21 1993–19941997–1999†4.5USAAmbulatory Care Quality Improvement Project (ACQUIP) at 3 Veteran Affairs general medicine clinics.M;
65 (median);
18+.
1 930
(324)
AUDIT-10:
1–7 and ≥8.
VA Beneficiary Identification and Record Locator SystemHRAge, smoking, depression, education, diabetes mellitus, myocardial infarction or coronary artery disease, liver disease. Among drinkers only.7
Harris et al 15 2004–20052006–2007†2USAVeteran Affairs ‘Survey of Health Experiences of Patients (SHEP)’ databaseM, W;
65 (median);
18+.
Men: 215 924 (13 040†), women: 9 168 (224†).AUDIT-C:
0, 1–4, 5–8 and ≥9.
(1) VA Beneficiary Identification Records Locator System, (2) Social Security Administration Death Master File, (3) Medicare Vital Status File and (4) VHA Medical SAS Inpatient Datasets.HRAge, race, education, marital status, past year depression and Deyo comorbidity index scores.7
Jeong et al 24 2005–200620093.75South KoreaRandom and representative panel survey: Korean Longitudinal Study on Health and AgingM/W;
76.3;
65–98.
846 (88)AUDIT-10‡:
Lifetime abstainers (ex-drinkers were excluded), 1–7 and ≥8.
Ministry of Public Administration and Security’s National DatabaseHRAge, educational level, smoking, Mini-Mental Status Examination score and Geriatric Depression Scale score.8
Justice et al 22 2008–201220144.8USAVeterans Aging Cohort Study (VACS) based on electronic medical records.M;
54.0;
n.a.
42 228 (1 985)AUDIT-C:
1–3, 4, 5–7 and ≥8.
(1) VA Beneficiary Identification Records Locating System, (2) VHA Patient Treatment File, (3) Medicare Vital Status File and (4) Social Security National Death Index.HRAge, race/ethnicity, smoking status and hepatitis C infection.
Participants with HIV were excluded.
7
Kinder et al 23 1997–199920002.61USAACQUIP at 7 Veteran Affairs internal medicine clinics.M;
65 (median);
18+.
34 292 (2 955)AUDIT-C:
0, 1–3, 4–5, 6–7, 8–9 and 10–12.
(1) VA Beneficiary Identification and Records Locator System, (2) Veterans Health Information Systems and Technology Architecture.HRAge, race, education, marital status, smoking status, smoking intensity, drug use, Mental Health Inventory-5 scores (continuous), post-traumatic stress disorder, cancer, lung disease, congestive heart failure, diabetes, myocardial infarction, pneumonia and and stroke.7
Nadkarni et al 25 2006–20082012–20146IndiaPopulation-based sample of men in urban and rural areas of Goa.M;
32.8;
18 to 49.
1 899
(62)
AUDIT-10:
0, 1–7, 8–15 and 16–40.
Official death certificate (no further information was provided).ORAge, residence, religion, marital status, education, employment status and socioeconomic status.8
Sidorenkov et al 26 1999–2000201010.2RussiaAnnual medical check-up and register-based recruitment in an outpatient clinic in Arkhangelsk.M, W;
45 (median); 18–60.
Men:
1 966 (147).
Women: 1 738
(95).
AUDIT-10:
0, 1–7 and ≥8.
Regional mortality registry.HRAge, education, physical activity, diastolic blood pressure, ApoB/ApoA1 ratio, body mass index and history of cardiovascular disease.
Comparison at-risk drinkers versus abstainers: age only.
8
  • *In years.

  • †Estimated.

  • ‡Korean version.

  • ApoB/ApoA1, ratio between Apolipoproteins B and A1; M, men; M/W, combined sex; n.a., not available in the respective manuscript; NOS, Newcastle-Ottawa Scale; W, women; W, M ,  stratified by sex .