Abstract
OBJECTIVES: Uncertainty exists as to what extent common risk factors are involved in the associations of unemployment with major health outcomes and mortality. DESIGN: A retrospective and prospective observational study. SETTING: A large population-based French cohort (CONSTANCES). PARTICIPANTS: 99 430 adults at baseline who have been exposed to unemployment during their lifetime and 54 679 of them who were followed for 7 years after baseline. PRIMARY OUTCOME MEASURES: Testing the mediating roles of several risk factors at baseline in the associations of lifetime unemployment exposure with cardiovascular disease, cancer and mortality rates during a 7-year follow-up. Direct and indirect effects were calculated for each risk factor and all together using logistic regression models adjusted for major confounders including sex, age, parental histories of cardiovascular disease and cancer, social position and working conditions. RESULTS: Estimates (95% CIs) of the direct and indirect effects for smoking are 0.0083 (0.0044 to 0.0122), p<0.0001 and 0.0010 (0.0007 to 0.0014), p<0.0001 on cardiovascular disease rate; 0.0059 (0.0028 to 0.0089), p=0.0002 and 0.0007 (0.0004 to 0.0010), p<0.0001 on cancer rate; 0.0105 (0.0058 to 0.0151), p<0.0001 and 0.0010 (0.0005 to 0.0014), p<0.0001 on all-cause mortality. The figures for alcohol consumption are, respectively, 0.0076 (0.0034 to 0.0118), p=0.0004 and 0.0004 (0.0002 to 0.0005), p=0.0006; 0.0067 (0.0035 to 0.0100), p<0.0001 and 0.0004 (0.0002 to 0.0005), p<0.0001; 0.0114 (0.0064 to 0.0164), p<0.0001 and 0.0004 (0.0001 to 0.0006), p=0.0009. For depressive symptoms, 0.0084 (0.0040to 0.0128), p=0.0002 and 0.0007 (0.0002 to 0.0011), p=0.005; 0.0053 (0.0017 to 0.0089), p=0.004 and 0.0001 (-0.0002 to 0.0005), p=0.51; 0.0088 (0.0031 to 0.0144), p=0.002 and 0.0010 (0.0004 to 0.0015), p=0.0005. For leisure-time physical inactivity, 0.0083 (0.0044 to 0.0122), p<0.0001 and 0.0003 (0.0001 to 0.0005), p=0.0006; 0.0057 (0.0026 to 0.0088), p=0.0004 and 0.0002 (0.0001 to 0.0003), p=0.002; 0.0105 (0.0058 to 0.0152), p<0.0001 and 0.0004 (0.0002 to 0.0007), p<0.0001. For blood triglycerides, 0.0080 (0.0042 to 0.0119), p<0.0001 and 0.0005 (0.0004 to 0.0007), p<0.0001; 0.0057 (0.0026 to 0.0087), p=0.0003 and 0.0001 (-0.0001 to 0.0002), p=0.32; 0.0103 (0.0057 to 0.0149), p<0.0001 and 0.0002 (0.0000 to 0.0004), p=0.06. The figures for all risk factors when tested together were 0.0075 (0.0022 to 0.0128), p=0.005 and 0.0020 (0.0011 to 0.0027), p<0.0001; 0.0052 (0.0011 to 0.0093), p=0.01 and 0.015 (0.0009 to 0.0020), p<0.0001; 0.0102 (0.0035 to 0.0169), p=0.003 and 0.0022 (0.0011 to 0.0031), p<0.0001. CONCLUSIONS: These analyses show that common risk factors such as smoking, alcohol consumption, depressive symptoms, leisure-time physical inactivity and blood triglycerides mediate up to 10% of the associations of lifetime unemployment exposure with cardiovascular disease, cancer and mortality rates when tested separately and approximately 20% when tested all together. This highlights the existence of other major mediating pathways that have yet to be identified.