Underestimating the alcohol content of a glass of wine: the implications for estimates of mortality risk

$\textbf{Aims}$: Increases in glass sizes and wine strength over the last 25 years in the UK are likely to have led to an underestimation of alcohol intake in population studies. We explore whether this probable misclassification affects the association between average alcohol intake and risk of mortality from all causes, cardiovascular disease and cancer. $\textbf{Methods}$: Self-reported alcohol consumption in 1997–1999 among 7010 men and women in the Whitehall II cohort of British civil servants was linked to the risk of mortality until mid-2015. A conversion factor of 8 g of alcohol per wine glass (1 unit) was compared with a conversion of 16 g per wine glass (2 units). $\textbf{Results}$: When applying a higher alcohol content conversion for wine consumption, the proportion of heavy/very heavy drinkers increased from 28% to 41% for men and 15% to 28% for women. There was a significantly increased risk of very heavy drinking compared with moderate drinking for deaths from all causes and cancer before and after change in wine conversion; however, the hazard ratios were reduced when a higher wine conversion was used. $\textbf{Conclusions}$: In this population-based study, assuming higher alcohol content in wine glasses changed the estimates of mortality risk. We propose that investigator-led cohorts need to revisit conversion factors based on more accurate estimates of alcohol content in wine glasses. Prospectively, researchers need to collect more detailed information on alcohol including serving sizes and strength.Both A.B. and S.B. are supported by grants from the European Research Council (ERC-StG-2012- 309337_AlcoholLifecourse, PI: Britton, http://www.ucl.ac.uk/alcohol-lifecourse) and UK Medical Research Council/Alcohol Research UK (MR/M006638/1)—D.O.N. is also supported by the latter. The Whitehall II study is supported by grants from the Medical Research Council (K013351), British Heart Foundation (RG/07/008/23674), Stroke Association, National Heart Lung and Blood Institute (HL036310) and National Institute on Aging (AG13196 and AG034454)