Allostatic load as a mediator of the association between psychosocial risk factors and cardiovascular diseases. Recent evidence and indications for prevention.

Aims. Some categories of workers are more vulnerable to the detrimental effect of job strain on cardiovascular risk. We investigate allostatic load, the physiological "wear and tear" resulting from adaptation to chronic stress, as a candidate pathway to explain such vulnerability. Methods. We selected 25-64 years old salaried workers participants to three population-based cohorts. We defined allostatic load (AL) as the sum of z-scores of 9 selected biomarkers; occupational classes (OCs) from the Erikson- Goldthorpe-Portocarero schema; and job strain (JS) according to Karasek's demand-control model. We adopted the Oaxaca- Blinder decomposition to disentangle the OC gradient in AL into the differential exposure (attributable to different JS prevalence across OCs) and the differential vulnerability (attributable to a different effect of JS on AL across OCs) components. Results. In the n=2010 workers (62% men, 34% manuals), OCs, but not JS categories, were associated with AL, independently of age and gender (p-value: 0.02). In the overall sample, JS did not have an effect on the OC gradient in AL. Conversely, in workers with sleep impairment, depression, or not engaged into physical activity, JS had a positive differential vulnerability coefficient of 0.63 (95%CI 0.05 to 1.21). Conclusions. In manual workers with impaired capacity of response, job strain is associated with a disproportional allostatic load accumulation.