Back to Annual Meeting

Back to Annual Meeting

APHA Scientific Session and Event Listing

Recorded presentation

Leslie London, Professor¹, Mary-Lou Thompson, Assoc Professor², and Jonathan E. Myers, Professor². (1) Health and Human Rights Programme, School of Public Health and Family Medicine, University of Cape Town, Anzio Rd, Observatory, Cape Town, 7925, South Africa, + 27 21 406 6524, ll@cormack.uct.ac.za, (2) Occupational and Environmental Health Research Unit, School of Public Health and Family Medicine, University of Cape Town, Anzio Rd, Observatory, Cape Town, 7925, South Africa

Alcohol consumption is an important confounder in occupational epidemiology, particularly in studies examining chronic effects of workplace exposures. A study of neurobehavioural effects of organophosphate exposure amongst 247 workers on deciduous fruit farms in the Western Cape Province of South Africa compared the repeatability and predictive validity of different measures of alcohol consumption. Measures included: questionnaire-based quantification of usual weekly consumption, lifetime consumption, consumption at one sitting, and perceptions of normative consumption; scores on symptom inventories (CAGE and modified MAST), and; serum levels of Gamma Glutamyl Transpeptidase (GGT). Reported cumulative lifetime, usual weekend and past week consumption of alcohol tended to have stronger correlated with each other (Pearsons r2's > 0.57) than with either the CAGE or MAST scores (Pearsons r2's <0.44), and gave very low correlation with GGT levels (r's<0.1). Agreement between dichotomized CAGE scores and GGT levels was extremely poor (Kappa=0.02). Data from a sub-sample of 29 workers who underwent repeat questionnaire 3 months later showed moderately good repeatability for alcohol consumption questions, the highest correlations being found for CAGE score (Spearmans r=0.75; 95% CI 0.53 -0.88) and cumulative lifetime alcohol consumption (Pearsons r = 0.69; 95% CI 0.44-0.85). Of all the alcohol metrics, CAGE score was the only significant predictor of neurological symptoms (OR=1.37; 95% CI 1.07-1.77) but gms alcohol consumed in the preceding week generated the strongest association (OR=2.01; 95% CI 0.98-4.12) in a multivariate logistic regression model. The data offer insights into appropriate choices to measure alcohol consumption in working populations in developing countries.

Learning Objectives:

• Appreciate the importance of measurement of alcohol as a confounder in occupational epidemiology studies.
• Understand the role of different metrics for measuring alcohol consumption
• Select an appropriate metric for measurement of alcohol consumption in occupational epidemiology studies

Presenting author's disclosure statement:

Any relevant financial relationships? No

Recorded presentation