Systematic under-reporting of drinking frequencies in international surveys of alcohol consumption

It is well known that self-reported alcohol consumption to telephone and in person surveys results in substantial under-reporting. In Canada, we have found this to be up to 80% for some population groups. In recently published analyses of a Canadian national survey (CADUMS), we found evidence that much under-reporting was due to under-reporting of drinking frequencies, especially among people describing themselves as occasional drinkers. Sharing these findings with international colleagues at the Kettil Bruun Society annual symposium in Uganda in 2013 led to an agreement to explore whether similar patterns of results would be obtained in Australian, English and US surveys. A paper is currently being prepared which reports that this pattern is indeed constant across different English speaking countries. The results suggest a methodology for correcting this kind of bias in epidemiological research on alcohol.

Funding body: CARBC Endowment Fund