Harmonization of alcohol use data and mortality across a multi-national HIV cohort collaboration
So far, no standardized approach to assess alcohol use has been established universally, which complicates comparability between different settings and studies. In some settings, including the SHCS since 2013, the Alcohol Use Disorders Identification Test-Consumption (AUDIT-C) questionnaire is used, whereas other settings use other measures like number of drinks within a specified time frame. The AUDIT-C score asks about the frequency of alcohol consumption, the typical number of drinks per occasion and how often 6 or more drinks are consumed on one occasion. To add an additional layer of complexity, the size of a standard drink differs from country to country: In Switzerland, for example, a standard drink is equivalent to 10g of alcohol, whereas in many other countries one drink is equivalent to 14g of alcohol.
In their study, Ingle et al evaluated data on alcohol use and its contribution to morbidity and mortality from 14 European and North American HIV cohort studies. They calculated the equivalent in grams of alcohol per day using the data from over 94,000 participants of the different cohorts. For cohorts who did not provide an exact number of drinks, they estimated grams/day based on the data provided. They then assessed the association between alcohol use and mortality using adjusted Cox models.
As expected, grams/day increased with increasing AUDIT-C scores despite some overlaps. Interestingly cohorts using the AUDIT-C questionnaire reported a lower proportion of people with HIV not using alcohol and were less likely to identify people with HIV with heavier alcohol use than those cohorts using other measures of alcohol use.
In line with previous findings within the SHCS, the authors reported higher mortality among participants not using alcohol and among participants with heavier alcohol use in comparison to moderate users. However, in the subset of participants from three cohorts with available data on former alcohol use, mortality was elevated among those reporting former use of alcohol compared to current moderate alcohol users, whereas mortality did not differ between individuals reporting lifetime abstinence and current moderate alcohol users.
In conclusion, the study shows a useful approach to harmonize different measures of alcohol use. Additionally, it supports existing evidence regarding the effect of alcohol use on overall mortality n persons with HIV.