Socioeconomic Factors, Work and Chronic Diseases

Transcription

1 Winnie, a -years old British women Socioeconomic Factors, Work and Chronic Diseases Jussi Vahtera Professor of Public Health University of Turku Finnish Institute of Occupational Health Finland Davey-Smith Int J Epidemiol Leading Causes of Death: WHO 3 scenario High-income countries Ischaemic heart disease (5.8%) Cerebrovascular disease (9.%) Lung cancer (5.%) Diabetes mellitus (4.8%) COPD (4.%) Low-income countries Ischaemic heart disease (3.4%) HIV/AIDS (3.%) Cerebrovascular disease (8.%) COPD (5.5%) Lower respiratory infections (5.%) Mathers et al. PLoS Med 6 WHO : Leading risk factors globally for mortality All deaths (4%) Raised blood pressure (3%) Tobacco use (9%) Raised blood glucose (6%) Physical inactivity (6%) Overweight and obesity (5%) Deaths from ischaemic heart disease (>75%) Heavy drinking Smoking Raised blood pressure Overweight and obesity High cholesterol Raised blood glucose Low fruit and vegetable intake Physical inactivity Socioeconomic factors??? Work??? Classification of CVD risk (QRISK ): UK Patient age (35-74). Patient gender. Current smoker (yes/no). Family history of heart disease aged <6 (yes/no). Existing treatment with blood pressure agent (yes/no). Body mass index (height and weight). Systolic blood pressure (use current not pre-treatment value). Total and HDL cholesterol. Self-assigned ethnicity. Rheumatoid arthritis. Chronic kidney disease. Atrial fibrillation. Townsend score (postcode-based measure of neighborhood deprivation)

2 Neighbourhood socioeconomic disadvantage: the Grid Database Grid dimension 5 m x 5 m. Median income. Education (basic, %) 3. Unemployment rate Coordinates for the participants' addresses (n=6 964) from Population Register Center Using GPS-coordinates, employees linked to their neighbourhoods (grid database: 8 74 neighbourhoods) Health risk behaviours: Smoking Heavy drinking Physical inactivity Statistics Finland: Population Statistics Neighbourhood socioeconomic effects in clustering of life-style risk factors Halonen, et al. () PLoS ONE Relative inequalities in the rate of death from any cause by occupation Mackenbach et al. N Engl J Med 8 Marmot Review: Fair society, healthy lives

3 OR (99% CI) LOW HIGH Odds Ratio.5 Reported stress during the past year and risk of MI.5 Home Work General.5 Never Some pd Several pd Permanent Work stress the second most common threat posed by the working environment (after musculoskeletal problems) Source: OSH in figures: stress at work facts and figures 9 Odds ratio adjusted for age, sex, region, and smoking (Rosengren et al. Lancet 4) INTERHEART: Risk of AMI with Multiple Risk Factors (Yusuf et al. Lancet 4) Job Strain Model by Karasek and Theorell Psychological demands LOW HIGH Job control - skill discretion - decision authority - participation in decision making low strain passive active high strain 8 4 Smk DM HTN APoB/A ++3 all4 +O +PS All RFs job strain Source: Karasek 979; Karasek & Theorell 99 Effort-Reward Model by Siegrist Organizational injustice theory effort Effort-reward imbalance reward Procedural justice: fairness of procedures used (decision criteria, voice, control of the process; the rules are applied equally for everyone) Relational justice: fairness of the interpersonal treatment received (dignity and respect; deals with employees in a truthful manner) Source: Siegrist Moorman, Greenberg & Cropanzano 3

4 Work stress and IHD: Meta-analysis of prospective cohort studies Long working hours and CHD: meta-analysis Stress model Age- and sexadjusted Multiple adjusted Job strain.45 (.5 to.84). (.9 to.35) Effort-reward imbalance.5 (.63 to 3.9).5 (.58 to 3.98) Organizational injustice.6 (.4 to.3).47 (. to.95) summary estimates risk factors and potential mediators Kivimäki et al. Scand J Work Environ Health 6 Virtanen et al. Am J Epid Overtime and CHD: Whitehall II 64 civil cervents free from CHD and worked full time at baseline CHD: fatal/non-fatal MI or angina pectoris (-year follow-up) Monday April 4, Hard work won't kill you? Well it might actually It is often said that "hard work won't kill you". Long hours at work may boost heart-attack risk By Amanda Gardner, Health.com April 4, -- Updated 43 GMT (543 HKT) April 5, HR (.-.9) AMI:.67 (..76) (.9-.69).4. ( ) > Framingham risk score + Long working hours: UK NEWS Heart risk of long hours Long hours at work increase heart risk Tue Apr 5,.7.5 5% net reclassification improvement Paid hours /work day Adjusted for demographics, diabetes, diastolic blood pressure, LDL and HDL cholesterol, triglycerides, type A behaviour Virtanen et al. Eur Heart J, Kivimäki et al. Ann Intern Med Kivimäki et al. Ann Intern Med Job strain and BMI: The IPD-Work Consortium Work stress Job strain High ERI Injustice Long hours Health-related behaviors Smoking Diet Alcohol Physical activity Direct pathophysiological mechanisms Clinical CHD Pooled analysis of 4 prospective European cohort studies (Belgium, Denmark, Finland, France, Germany, the Netherlands, Sweden, UK) >7, participants (49% men, mean age 44 years). Model Underweight Normal weight (reference) Overweight Obese, class I Obese, class II-III Model Underweight Normal weight (reference) Overweight Obese, class I Obese, class II-III.75.5 Odds Ratio for Job Strain OR 95%CI. (..5)..7 (..).9 (.3.5).3 (.6.46). (..5).. (.96.6).7 (..).4 (..8) Sex and age adjusted Sex, age and SES adjusted Nyberg et al. Journal of Internal Medicine 4

5 Unemployment (%) Job strain and physical inactivity: The IPD-Work Consortium Model, adjusted for sex and age, n=7,6 Low strain (reference) Passive Active High strain Model, adjusted for sex, age, SES and smoking, n=63,4 Low strain (reference) Passive Active High strain.75.5 Odds Ratio OR..34 [.3;.47]. [.93;.6].36 [.5;.48] OR 95%-CI 95%-CI.. [.;.3].6 [.;.].6 [.5;.38] OR Cumulative work stress (phases -) and health risk behaviours (phase 3). Chandola et al. Eur Heart J Exposure.75.5 Odds Ratio Fransson, et al. (submitted) Life-style explained 6 % of the association between work stress and IHD OR Chandola et al. Eur Heart J Cumulative work stress (phases -) and biological risk factors (phase 3) Exposure Efforts to strengthen causal inference:. Using experiments of nature (exposure random in relation to characteristics of individuals) Physical inactivity, poor diet and the metabolic syndrome the most important explanatory factors (3% of the association) Unemployment rate (%) in Finland Finland EU-5 Quasi-experiment Statistics Finland, Eurostat All 6 68 employees from 4 towns Study population (9 had missing data) Organisational downsizing Non-downsized group Downsized group Those lost/left job n = n = 4783 n = 47 Outcome: Health status in 994-5

6 Post-downsizing health problems among "survivors" Poor SRH 993 Poor SRH 997 Musculoskeletal problems Sickness absence Injuries Psychotropics (men) Psychotropics (women) Disability retirement Other deaths CVD mortality,5,5,5 3 Risk ratio Exposure to downsizing (>8%) Yes No Vahtera et al. Lancet (997), BMJ (4), JECH (5); Kivimäki et al. OEM (), BMJ (4), JECH (7) Validation: Downsizing as a proxy measure of work stress Work demands Skill discretion Participation Job insecurity Supervisor support Co-worker support Spouse support Alcohol intake Smoking Physical activity Body mass % Kivimäki et al, BMJ P<.5 Difference in post-downsizing values (adjusted for pre-downsizing values) between groups of major vs no downsizing. Efforts to strengthen causal inference: Employment circumstances and mental health Household, Income and Labour Dynamics in Australia (HILDA), a nationally representative annual household panel survey (7,55 respondents) Seven waves of data Longitudinal random-intercept regression models regressing mental health on time-varying employment circumstance Adjustments for demographics, education, years in employment, physical functioning (SF-36), financial difficulties, neighborhood disadvantage. Using repeated measurements (individuals their own controls) Source: Butterworth et al. Occup Environ Med Change in mental health for transitions from unemployment Efforts to strengthen causal inference: 3. Removal of all work-related stress ( intervention study ) Source: Butterworth et al. Occup Environ Med 6

7 Prevalence (95% CI) Decrease Increase Number of employees retiring in the Gazel cohort by year of retirement Trajectory of prevalence of suboptimal self-rated health ,884 (9%) retired by 7. All participants with repeated measurement on health before and after retirement Employees retire early (mean age 55 yrs) and benefit from good social security (pension 8% of salary) Westerlund et al. Lancet 9 Trajectory of prevalence of sleep disturbances Trajectory of prevalence of fatigue and depression Panel B Mental fatigue Depression CES-D Physical fatigue.77 ( ) 5 Vahtera et al. SLEEP Westerlund et al. BMJ Finnish public sector employees 738 statutory retirees (mean age at retirement 6. yrs). Purchases of antidepressant medication derived from comprehensive national pharmacy records in Oksanen et al. Epidemiology Explanations for the trajectories Suboptimal health Sleep disturbances Low risk profile High risk profile Year in relation to retirement Westerlund et al. Lancet (9) Vahtera et al. SLEEP (9) Poor health (OR) Changes in suboptimal health around retirement by SES 7y before.57 y before Westerlund et al. Lancet 9 Higher Intermediate Lower.7 y before y after y after Time in relation to retirement. 7y after 7

8 5-y risk of CHD death per men Characteristics associated with cardiovascular health Extended working lives? No use of tobacco Adequate physical activity: - at least 3 minutes 5 times a week Healthy eating habits No overweight Blood pressure below 4/9 Blood cholesterol below 5 mmol/l Normal glucose metabolism Avoidance of excessive stress European Heart Health Charter designed to prevent cardiovascular disease (developed by the European Heart Network and the European Society of Cardiology, with the support of the European Commission and WHO. Best-practice interventions Pharmacological and lifestyle interventions have reduced - systolic blood pressure by about mm Hg, - total cholesterol by mmol/l, - blood glucose among pre-diabetic people by mmol/l, and have halved the prevalence of non-insulin-dependent diabetes mellitus (non-iddm) in adults. In principle, all smokers could quit. Best-practice interventions applied universally 7 86 male civil servants aged 4 69 years at baseline (967-97; Whitehall study). Baseline measures: Socioeconomic position based on employment grade, systolic blood pressure, total cholesterol, postload glucose/diabetes, smoking. Main outcome measure: 5-year risk of coronary heart disease SEP group mortality high low What would happen to socioeconomic inequalities if the best-practice risk reductions applied to both high and low socioeconomic groups? Kivimaki et al. Lancet 8;37: None TDM & Syst BP Smoking TC All glucose Intervention target Kivimaki et al. Lancet 8;37: