Characterising the alcohol harm paradox: a population-based survey of adults in England

Abstract

BackgroundThe alcohol harm paradox refers to the positive association of socioeconomic status (SES) with alcohol consumption and negative association with alcohol-related harm and dependence. To inform future research and to help elucidate the cause of the alcohol harm paradox, this study aimed to assess how far the paradox extends to a range of measures of SES and whether it varies by demographic characteristics. MethodsBetween March and December, 2014, data were collected on 16 871 participants from the Alcohol Toolkit Study, a monthly population survey of adults aged 16 and older. In this survey, interviews with 1800 individuals in England are conducted each month by the market research company, Ipsos MORI. Participants were asked to complete the Alcohol Use Disorders Identification Test (AUDIT), which consists of three parts: alcohol consumption (AUDIT-C), alcohol dependence (AUDIT-dependence), and alcohol harm (AUDIT-harm). SES was categorised as follows: qualifications after the age of 16 years (yes, no), employed full time (yes, no), owns own house (yes, no), owns own car (yes, no), income of less than £11 499 (yes, no), and a classification based on occupation called social grade (AB, C1, C2, D, E). A composite score was also derived with multiple correspondence analysis. Prevalence data were weighted to match the population in England. Findings11 295 participants (71%) reported that they drank alcohol (95% CI 69·7–71·2). Those who were aged 35–44 years (p=0·0009), 45–54 (p<0·0001), 55–64 (p<0·0001) and 65 years or over (p=0·0107) had higher odds of reporting that they drank alcohol than those aged 16–24. Those of higher socioeconomic status (p<0·0001) and men (p<0·0001), also had higher odds of drinking alcohol. After adjustment, positive associations with AUDIT-C were found between social grade AB relative to C2 (β=–0·26, p=0·0067) and D (β=–0·54, p<0·0001), educational level (β=–0·19, p=0·0242), and the composite score (β=–0·12, p<0·0001). All SES measures, except for car ownership and educational qualifications, were negatively associated with AUDIT-harm and AUDIT-dependence scores. The alcohol harm paradox was moderated by demographic characteristics: AUDIT-dependence was associated with measures of SES in men (β=–0·07, p=0·0110) but not women; and associations between AUDIT-C and SES were strengthened with increasing age, whereas associations between AUDIT-dependence and SES diminished. InterpretationAmong adults in England, the alcohol harm paradox is apparent across a range of measures of SES and seems to be more evident in younger men than in other demographic groups. FundingEB's salary is funded by the National Institute for Health Research (NIHR) School for Public Health Research (SPHR) and Cancer Research UK (CRUK). JB is funded by CRUK and the Society for the Study of Addiction. RW is funded by CRUK. The Alcohol Toolkit Study is funded by the NIHR SPHR.