Can be quite substantial. Jump off the Empire State Building and see for yourself. But, beyond that, the question remains how much of the variation in health outcomes and longevity can be explained by behavioral variation? Well, we don’t quite know. But we do have evidence which indicates that – at least in the developed world – that fraction is quite small.
From Ian J. Deary & Geoff Der (2005) Reaction Time Explains IQ’s Association With Death:
After AH4 scores and each of the reaction time measures were adjusted for sex, smoking status, social class, and years of education, all effects remained significant, and the hazard ratios were only slightly attenuated (Table 1). Thus, the relation between IQ and mortality in this sample was not substantially mediated by social class, education, or smoking. Nor was the relation between reaction times and mortality substantially mediated by these variables.
we adjusted for a range of physiological, behavioural, psychological and social risk factors that can be considered as mediating variables in the IQ–mortality relation. The influence of controlling for these factors can be broadly divided into three strata. In the first, despite being associated with both IQ and mortality, adjusting individually for marital status, alcohol consumption, systolic and diastolic blood pressure, pulse rate, blood glucose, body mass index and psychiatric and somatic illness at medical examination had very little, if any, impact on the age-adjusted IQ–death relation (<10% attenuation in risk)
From Hagger-Johnson, Deary, Batty, et al (2014) Reaction Time and Mortality from the Major Causes of Death: The NHANES-III Study:
In fully adjusted models which also adjusted for educational attainment, occupational grade, poverty/income ratio, health behaviors and CVD risk factors, the association was attenuated but remained statistically significant for all-cause mortality (HR = 1.15, 95% CI 1.02,1.29; 37% attenuation), and CVD mortality (HR = 1.22, 95% CI 1.15,1.29; 36% attenuation).
The 2014 study found some attenuation when “health behaviors,” among other things, were factored in. Nonetheless, the association remained. A key problem, however, is that “health behaviors” were gauged via self-report. This has been demonstrated to be highly (though by no means completely) inaccurate. The effect of the measurement error in assessing behaviors in the study is unknown; it could bias the attenuation either upwards or downwards. Another similar study by Deary et al (2008) found that the association between simple reaction time and deaths from cardiovascular disease (and stroke in particular) unaffected by adjusting for covariants (they did however find that the association between reaction time and IQ and deaths from coronary artery disease in particular become non-significant when “health behaviors” were factored in). However, these health behaviors were also assessed by self-report. Yet another study by Shipley, Der, & Deary et al (2006) looked at a British sample (N ~7,400) found the association with simple reaction time and all cause mortality. The effect was also mediated by variables. These results call for a meta-analysis.
I also threw out this idea today:The idea is that some ailments do appear to be heritable (e.g., heart disease), however, may have pathogenic involvement. A (heritable) weaker immune response or otherwise compromised defensive capacity might then at least partly explain observed heritability of these diseases.