Comparing diary and summary measures of alcohol use in a sample of heavy drinking men in Goa, India

Careful assessment of heavy drinking is particularly important in India as about half male drinkers meet criteria for heavy use (Benegal, 2005; Nayak et al, 2008). Survey consumption results have previously not been validated against rapid-recall measures such as daily diaries. As part of a larger population survey, male hazardous drinkers from Goa, India (n=56) completed detailed surveys assessing past-year alcohol use and 28-day daily drinking diaries. Survey measures included usual drinking quantity and frequency (QF) and a detailed beverage-specific Graduated Frequencies (Proportional GF) measure maximum quantity (accounting for beverage types, strength and sizes) and its frequency, then frequencies of "about 3/4", 1/2, and 1/4 this amount. Survey and diary drinks per drinking day correlated 0.67, with diaries giving 1/2 standard drink larger volumes (5.6 grams/drinking day, NS). Correlation between the GF- and diary-based volumes was 0.57), with GF volume slightly higher (3.7 grams/day, NS). Proportion of drinking days from the GF, usual frequency, and diary measure had lower correlations (rho = .30, .39, respectively). The GF yielded the largest drinking frequency (proportion of days drinking), then diary and usual frequency (GF 6.2% more; usual 1.6% less). GF yielded a slightly lower maximum (2.5 grams less, NS; rho = .54), with diary amounts largest. Overall, the Proportional GF measure is efficient and reasonably validated by the intensive daily diaries. Those meeting AUDIT Hazardous Drinking criteria gave higher GF maximums and quantities than in the shorter diaries, suggesting GF may specially well assess 12-month alcohol consumption among heaviest, more problematic drinkers.