Employment Transitions and Health: Data from the English Longitudinal Study of Ageing Neil Rice Epidemiology & Public Health Group, Peninsula Medical School, University of Exeter
Briefly today An introduction to the English Longitudinal Study of Ageing My research interests, and some work we have been doing using the ELSA data
ELSA Study Description An interdisciplinary data resource on health, economic position and quality of life as people age beyond 50, and the factors associated with it ELSA data collection began in 2002 Respondents are interviewed every two years, with an additional biomedical assessment (by separate nurse visit) every four years (2004, 2008 etc) Sample members were drawn from respondents to the Health Survey for England (HSE) in years 1998, 1999 & 2001 who were born before 1 st March 1952, and their partners
ELSA Numbers Involved HSE Interview Years (ELSA wave 0) ELSA Interview Years Wave 1 Wave 2 Wave 3 Wave 4 2002-03 2004-05 2006-07 2008-09 HSE 1998 n=4,793 n=3,692 n=3,149? HSE 1999 n=2,288 n=1,684 n=1,406? HSE 2001 n=4,311 n=3,404 n=2,980? ELSA Sample Core Members n=11,392 n=8,780 n=7,535? New or Younger Partners n=636 n=493 n=305? Wave 3 refreshment sample - - n=1,712? Non-responses n=72 n=159 n=219? Total N N=12100 n=9,432 n=9,771?
ELSA Potentially interesting data Work questions current employment activities previous employment activities retirement timing and reasons pensions provision and forecast pension wealth income and wealth (personal and household)
ELSA Potentially interesting data Health questions Self-reported health status and disability Self-reported physician diagnosed diseases Functional limitations - Activities of Daily Living Physical functioning performance tests (Nurse visit) Balance tests, gait speed, chair rises
ELSA Potentially interesting data Wellbeing questions CASP-19 & GHQ12 scales Quality of life measures Social participation measures Social networks / relationships measures Detailed cognitive function section Retrospective early life experiences questionnaire
ELSA Workers Numbers Involved (1) Wave 1 status Wave 2 status Wave 3 status N = 331 Working N = 610 N = 425 N=789 N=3607 N = 33
ELSA Workers Numbers Involved (2) Wave 1 status Wave 2 status Wave 3 status N = 52 N = 5654 N = 4490 N=167 N = 228
Current Work (1) Poorer health of workers predicts early exit from the labour market - findings from the English Longitudinal Study of Ageing Literature review: Premature retirement is complex and multidimensional: individual finance pension and non-pension wealth (e.g. Samwick 1998) partner s pension wealth (Chan 2004) complimentarities in leisure amongst partners (Banks 2008) personal attitudes to retirement and job satisfaction (Mein 2000) job demands and control (Elovainio 2005) poor conditions (Krause 1997) social networks (Van Solinge 2007) childhood adversities (Harkonmaki 2007)
Current Work (2) Poor health as a predictor of retirement is not a new concept Previous studies used subjective measures of self-rated poor health, which may be endogenous in the retirement model. ELSA has much more detailed markers of health Early work exit => Moving from to between baseline and follow-up, in those under the SPA at follow-up: n=308 vs n=1385 who remain in work
Odds Ratios (95% Confidence Intervals) for some health predictors of early work exit (models adjusted for age, gender, wealth, education, BMI, partner employment status) 0 1 2 3 4 5 6 7 Poor self-rated general health Having a longstanding limiting illness ADL problems (any) IADL problems (any) Any difficulty walking 1/4 mile Leg pain causes walking problems Pain (oftened trouble by?) Back pain (any) Hip pain (any) Knee pain (any) Foot pain (any) Depression (symptomatic) CES-D score 3 Diabetes Currently taking diabetes medication Angina Asthma Expectations of poor health p=0.0000006 p=0.0002 p=0.230 p=0.004 p=0.000001 p=0.000007 p=0.028 p=0.208 p=0.069 p=0.002 p=0.005 p=0.001 p=0.034 p=0.035 p=0.017 p=0.072 p=0.011
Previous Work Smoking cessation and transition into retirement: analyses from the English Longitudinal Study of Ageing. Age & Ageing. 2007 Nov; 36(6): 638-43 Lang IA, Rice NE, Wallace R, Guralnik J & Melzer D. 1712 smokers (aged 55 to 70) followed up for 5 to 6 years. WHAT NEXT?... ONE OUT OF THE SMOKING SHORTENS YOUR LIFE PACKET? OR SHALL WE GIVE SMOKING KILLS A GO? Retirees more than twice as likely to quit smoking as those who remained in work. (Odds Ratio = 2.50 (95% CI 1.35-4.62)) Retirees should be targeted with smoking cessation interventions.
Some questions I m looking at using ELSA data To what extent do health behaviours change in the workretirement transition? To what extent are mild mobility problems in middle-age predictive of disability / early work exit in later life? Are genetic variants associated with chronic disease (from GWAs) also associated with early onset disability / early work exit in the general population?
Any questions???