Different associations between body composition and alcohol when assessed by exposure frequency or by quantitative estimates of consumption

Title
Different associations between body composition and alcohol when assessed by exposure frequency or by quantitative estimates of consumption
Publication type
Journal Article
Year of Publication
2018
Journal
Journal of Human Nutrition and Dietetics
Volume
31
Issue
6
Pagination
747-757
Date published
2018
Abstract

Background: Alcohol intake is widely assumed to contribute to excess body fatness, especially among young men; however, the evidence is inconsistent. We have addressed this research question by investigating associations between reported alcohol consumption and body composition from large representative national surveys in a high alcohol-consuming country with a high obesity prevalence. Methods: The present study comprised a secondary analysis of combined cross-sectional nationally representative Scottish Health Surveys (1995-2010). Reported alcohol-drinking frequency was divided into five groups: from 'nonfrequent drinking' (reference) to daily/'almost every day' among 35 837 representative adults [mean (SD) age: 42.7 (12.7) years (range 18-64 years)]. Quantitative alcohol consumption was categorised into seven groups: from '1-7 to ≥50 10 g units per week'. Regression models against measured body mass index (BMI) and waist circumference (WC) were adjusted for age, physical activity, income, smoking, deprivation category and economic status. Results: Among alcohol-consuming men, heavier drinking (21-28 units per week) was associated with a higher BMI by +1.4 kg m-2 [95% confidence interval (CI) = 1.38-1.43] and higher WC by +3.4 cm (95% CI = 3.2-3.6) than drinking 1-7 units per week. However, those who reported daily drinking frequency were associated with a lower BMI by -2.45 kg m-2 (95% CI = -2.4 to -2.5) and lower WC by -3.7 cm (95% CI = -3.3 to -4.0) than those who reported less-frequent drinking. Similar associations were found for women. Most of these associations were restricted to subjects aged >30 years. Unexplained variances in BMI and WC are large. Conclusions: Quantitative alcohol consumption and frequency of consumption were positively and inversely associated, respectively, with both BMI and WC among alcohol-consuming adults. Surveys are needed that evaluate both the quantity and frequency of consumption. The lowest BMI and WC were associated with a 'Mediterranean' drinking style (i.e. relatively little, but more frequently). Journal of Human Nutrition and Dietetics