Atherosclerosis and cardiovascular disease (CVD) have an inflammatory origin. Moreover, chronic psychosocial stress is associated both with inflammation and CVD. Fibrinogen, a coagulation factor and CVD risk factor, is also a positive acute-phase reactant protein, its concentration increasing with presence of inflammation. Elevated inflammation measured on a single occasion can reflect either systemic inflammation and be the consequence of chronic psychosocial stress or the acute response to a recent stressful event or infection in an otherwise healthy, non-psychosocially stressed individual. Therefore, the prognostic value for future CVD risk of a single inflammation test may depend on the presence of chronic psychosocial stress. We tested this hypothesis in middle-aged to older adults aged 52 and above participating in a nationally representative cohort study.

A total of 4541 men and women, free of CVD and aged 52 to 92 at baseline (2004–2005), participants in the English Longitudinal Study of Ageing, provided serum levels of fibrinogen during the nurse visit at baseline, alongside with clinical measurements and other blood biomarkers. Psychosocial factors were collected during the computer assisted personal interview and included financial strain, depression, social isolation, loneliness and social support. The outcome of interest was CVD mortality as ascertained up to February 2013. Cox proportional hazards regression models using age as the underlying time variable were fitted to estimate the relationship (hazard ratios [HR] and 95% confidence intervals) between fibrinogen and CVD death. Interactions terms between fibrinogen and each psychosocial factor were tested. Models were stratified by sex and adjusted for age, smoking, body mass index, physical activity, cholesterol, triglycerides and diabetes. Restricted cubic splines were used to assess the shape of the associations and test nonlinearity by using the likelihood ratio test, comparing nested models with a linear or linear and cubic spline terms.

In this sample with a baseline median age of 63 years (interquartile range 57–71), there were 130 CVD deaths after a median follow-up of 8.1 y. The association between fibrinogen and hazard of CVD mortality was linear and positive: HR for the increase in 1g/L of fibrinogen=1.56; 95% CI: 1.26, 1.94, P<0.0001 and P for curvature=0.53 (non-significant value means the relationship is linear). We found a strong interaction (P=0.02) between fibrinogen and financial strain (“shortage of money stops from doing what I want”, classified as often vs. less, binary), but not with the other psychosocial indicators. In the presence of financial strain (n=479, 21 deaths), the HR for 1g/L fibrinogen increase was 4.04; 95% CI: 1.99, 8.19, whereas in the absence of financial strain (n=4062, 109 deaths), HR=1.39; 95% CI: 1.09, 1.79. When further adjusting for C-reactive protein, the association in the group experiencing financial strain got attenuated but remained significant (HR=2.74; 1.10, 6.78), whereas it became non-significant in the low financial strain group (HR=1.24; 0.91, 1.69).

The positive association between fibrinogen and CVD death was much stronger in the presence of financial strain. An implication of our findings is that, when assessing the presence of inflammation with a single test for estimation of CVD risk, it may be necessary to take into account the presence of chronic psychosocial stress.