1. Drinking profiles, social capital and health in later life (2013-2014) Clare Holdsworth (PI)Nicola Shelton Marina MendonçaHynek Pikhart Martin FrisherCesar Oliveira

2. Background Levels of alcohol consumption increased for middle and older age groups in the last 30 years for both men and women (Smith and Foxcroft, 2009). – Cohort effect (i.e. Generation) – Period effect (e.g. more financial resources) – Age (e.g. active ageing and greater life expectancy) Number of elderly population expected to increase 38% in the UK by 2031 (IAS, 2010): Alcohol consumption among the elderly - cause of public concern

3. Background U-shape relationship between alcohol consumption and health: Poorer health associated with abstainers (e.g. ‘sick- quitter’ hypothesis ) or excessive drinkers (e.g. Polen et al, 2010). Relation between drinking and health contradictory in older ages: – Alcohol consumption associated with more physical and mental health problems and higher mortality rates (e.g. Crome, et al. 2011; Dar, 2006; NHS Information Centre, 2008) – But, some studies found no association between levels of alcohol consumption and health (e.g. Lang, et al., 2007).

4. Background Alcohol consumption and health moderated by social capital (Bloomfield et al., 2006; Marmot 2005) – Alcohol Harm Paradox: Higher social position associated with higher alcohol consumption and better health, whereas alcohol-related illness and mortality linked with deprivation. Research on drinking behaviours focused both on quantity and frequency of drinking. – But findings suggest that they are not equally associated with health and social capital Unpacking complexity : Explore relationship between different drinking behaviours/profiles and its association with health and social capital

5. Aims To identify and characterize drinking profiles in older age To analyse how these drinking profiles are associated with social capital, health and socio-demographic characteristics

6. Data ELSA W0 (HSE): Baseline for alcohol variables Present study: – W0: Drinking and health variables – W1: Social capital variables 199819992001 W6 2012/13 W4 2008/9 W5 2010/11 W3 2006/7 W2 2004/5 W1 2002/3 N = 11205 English Longitudinal Study of Ageing (ELSA): Panel study of people living in England aged 50 and older

7. Drinking Variables Drinking Profiles Drinking StatusAmount of AlcoholFrequency of drinking Non-drinkerNon-Drinker Drinker Below Limits (Men ≤21 units; Women≤ 14 units) Occasional (≤ 4 days) Low Drinker Daily (≥ 5days) Steady Drinker Above Limits (Men > 21 units Women > 14 unit) Occasional (≤ 4 days) Focal Drinker Daily (≥ 5days) Heavy Drinker Social Capital & Health Variables Level of EducationWealth QuintilesSESSelf-Rated Health Socio-demographic AgeMarital Status Variables

8. Men (%) (n = 5057) Women (%) (n = 6148) All (%) (N = 11205) Mean age at baseline62.09 (9.97)62.63 (10.49)62.39 (10.26) 45-6460.9059.2259.98 65-7425.7124.6925.15 75 +13.3916.0914.87 Marital Status Single6.394.985.62 Married/remarried76.0361.0467.81 Separated/divorced9.2012.3210.91 Widowed8.3921.6615.67 Educational Qualification No qualification37.0048.0143.04 Intermediate35.5834.7335.11 High27.4217.2521.84 Wealth Bottom quintile17.7721.0719.57 2 nd quintile19.8820.1120.00 3 rd quintile19.7420.1719.97 4 th quintile21.0219.2120.03 Top quintile21.6019.4420.42 SES Manual50.3636.4042.69 Non Manual48.2759.7954.60 Self-Rated Health Poor Health9.818.399.03 Fair Health23.75 25.07 24.48 Good Health66.44 66.5466.49 Descriptives: Socio-demographic, social capital and health variables by gender

9. Descriptives: Percentage distribution of drinking patterns by gender

10. Multinomial Logistic Regression: Odds of being Steady and Heavy/Binge Drinker versus Low Drinker Note: *** p<.001, **p<.01, *p<.05 Steady Heavy/Focal MenWomenMenWomen Age groups (Ref: 45-64) 65-741.54***1.94***.75**.92 75+2.21***3.05***.53***.78 Mar. Stat. (Ref: Married) Single1.38.671.60**.79 Separated/divorced.76.801.46**1.06 Widowed.96.811.57**.93 Self-Rat. Heal. (Ref: Poor) Fair.90.921.24.82 Good.711.031.46**1.03 Educ. Qual. (Ref.:No qual.) Intermediate1.141.251.19*1.27* High1.45**1.77***1.45***1.57*** Wealth (Ref.:Bottom) 2 nd 1.081.121.041.08 3 rd 1.281.08.861.02 4 th 1.77**1.55*1.011.33 Top2.36***2.47***1.34*2.00*** SES (Ref.: Manual) Non Manual1.171.41**.971.63***

11. Summary Gender: Important moderator of drinking behaviours. Men are more likely to be drinkers and to be heavy/focal drinkers than women. Age: Older participants are more likely to be steady drinkers, while those at ‘younger’ older ages are more likely to engage in heavy/focal drinking. Health: Drinking profiles are generally not associated with self-rated health, but male heavy drinkers are more likely to report better health. Social Capital: People in higher social position are more likely to be steady and heavy drinkers. This association is stronger for women. Marital Status: Being married seems to be a protective factor for men against heavy drinking

12. Thank you for your attention! For further information please contact mmendonca@keele.ac.uk

13. Descriptives: Sociodemographic, Health and Drinking Variables by Gender