Caroline S. Duchaine, PhD candidate, CHU de Québec-Laval University Research; Chantal Brisson, PhD, CHU de Québec-Laval University Research Center; Denis Talbot, PhD, CHU de Québec-Laval University Research Center; Mahée Gilbert-Ouimet, PhD, Department of Health Sciences, Université du Québec à Rimouski; Xavier Trudel, PhD, CHU de Québec-Laval University Research Center; Michel Vézina, MD, Institut national de santé publique du Québec; Alain Milot, MD, Faculty of Medicine, Laval University; Caroline Diorio, PhD, CHU de Québec-Laval University Research Center; Ruth Ndjaboué, PhD, VITAM, Centre de recherche en santé durable; Yves Giguére, MD, Faculty of Medicine, Laval University Benoît Mâsse, PhDh, School of public health, University of Montreal ;Clermont E. Dionne, PhD, CHU de Québec-Laval University Research Center; Elizabeth Maunsell, PhD, CHU de Québec-Laval University Research Center; Danielle Laurin, PhD, Faculty of Pharmacy, Laval University
The current longitudinal study aimed to evaluate the associations of long-term exposure to psychosocial stressors at work from two recognized theoretical models, namely Karasek’s Demand-Control-Support and Siegrist’s Effort-Reward Imbalance (ERI) models, with serum concentration of two inflammatory biomarkers, namely C-reactive protein (CRP) and interleukin-6 (IL-6), separately and combined into an inflammatory index.
Normal or acute inflammation is part of the natural host-defence mechanism, but prolonged exposure to low-grade inflammation, characterized by an increased concentration of inflammatory biomarkers in the systemic circulation of healthy individuals, may lead to adverse health effects with time. Chronic low-grade inflammation and psychosocial stressors at work have both been associated with high risk of several health problems such as cardiovascular diseases, diabetes, depression, and cognitive decline. Psychosocial stressors at work have been suggested as modifiable risk factors of low-grade inflammation, but few longitudinal studies have evaluated the association between these stressors and inflammatory biomarkers. As low-grade inflammation could be present long before the onset of the disease, identifying modifiable risk factors could allow to act upstream.
Data came from the PROspective Quebec (PROQ) Study on Work and Health, a cohort of 9188 white-collar workers recruited in 1991-1993 (T1) and followed-up after 8 (T2, 1999-2000) and 24 (T3, 2015-2018) years. For this study, 3411 participants were randomly selected from all T1 participants that were still alive and have accepted to be recontacted at T2 follow-up. Among these, 2557 participants were still alive and available for the measure of serum biomarkers at T3 and were included in the present study. Psychosocial stressors at work were assessed at T2 according to the Demand-Control-Support and the ERI models using validated questionnaires. High job strain was defined by an exposure to high psychological demand combined with low job control, and iso-strain was defined by an exposure to high job strain combined with low social support at work. ERI was defined by an imbalance between psychological demand and social, economic, and organisational reward. CRP and IL-6 serum concentrations were measured using standardized protocols. Several covariates were considered including sociodemographic, anthropometric and lifestyle characteristics, and comorbidities. Adjusted prevalence ratios (PRs) and 95% confidence interval (CI) for the highest quartile of CRP, IL-6, and inflammatory index at T3 according to psychosocial stressors at work measured at T2 were calculated using generalized estimating equations. Multiple imputation and inverse probability of censoring weighting were done to consider lost to follow-up between measurement times.
The mean age of participants at T2 was 46 years, 52% were women, and 44% had a university degree. Mean follow-up time from T2 to T3 was 17 years. Associations were observed between exposure to iso-strain and the inflammatory index among men, but not among women (PRs and 95% CIs of 1.42 (1.06-1.90) and 1.04 (0.80-1.34), respectively). Analyses by age group revealed that this association was mainly found among men aged <65 years compared to those aged ? 65 years (PRs and 95% CIs of 2.00 (1.37-2.92) and 0.99 (0.62-1.59), respectively). In the <65 years age group, associations with inflammatory biomarkers were observed among men with exposure to high job strain and ERI, and among women, with exposure to low reward or moderate social support at work.
Few previous longitudinal studies have evaluated the association between psychosocial stressors at work and inflammatory biomarkers. Most of them reported associations between exposure to some of these stressors and high inflammation. The effect modification by sex was not evaluated by all of them and appeared inconsistent across studies. The age and the inflammatory biomarker evaluated seem to matter as associations were stronger in younger participants and with combinations of biomarkers instead of each biomarker separated. This study has several strengths including a large sample size, a long follow-up, validated measures of psychosocial stressors at work, adjustment for many potential confounders, and the use of complex statistical analyses to consider the potential of selection bias due to lost to follow-up. This study has also limitations including: one time-point measure of inflammatory biomarkers that do not allow to take into account the change over time; the potential of the healthy worker effect bias that tends to underestimate the real effect; and the population restricted to white-collar workers that may not be representative of the entire workforce population.
The results of this study suggest that psychosocial stressors at work may increase low-grade inflammation, especially among men aged less than 65 years. However, further studies are needed to clarify the potential differences between men and women. As these stressors are frequent and modifiable, they could play a role in the primary prevention of low-grade inflammation and chronic diseases.