Sex differences in cardiac troponin trajectories over the life course

Abstract

Abstract Background Cardiac troponin is a promising biomarker that may improve cardiovascular risk prediction. Previously, important differences have been demonstrated between cardiac troponin concentration in women and men. At this point, it is unknown whether changes in cardiac troponin concentration over the life course differ in women and men and whether this impacts cardiovascular risk prediction. Purpose Insights into sex-specific cardiac troponin trajectories and how these are influenced by risk factors are necessary to ensure approaches to the prevention of cardiovascular disease in women and men are equitable. Therefore, we examined whether age- and risk factor-related changes in cardiac troponin over the life course differ by sex and if the trajectory of cardiac troponin was informative in respect of cardiovascular outcomes in women and men in the general population. Methods In the Whitehall II cohort, high-sensitivity cardiac troponin I concentrations were measured on three occasions over a 15-year period. Using linear mixed-effects models, the sex-specific trajectories of cardiac troponin were evaluated, and the relationship with conventional cardiovascular risk factors determined. Using multistate joint models, the association between sex-specific trajectories of cardiac troponin and a composite outcome of non-fatal myocardial infarction, non-fatal stroke, or cardiovascular death was evaluated. Results In 2,142 women and 5,151 men (mean 58±7 and 57±7 years, respectively) there were 177 (8.3%) and 520 (10.1%) outcome events, respectively, during a median follow-up of 20.9 [25th to 75th percentile, 15.8 to 21.3] years. Cardiac troponin concentrations were persistently lower in women than in men (median baseline concentration: 2.4 [25th to 75th percentile, 1.7 to 3.6] ng/L versus 3.7 [25th to 75th percentile, 2.6 to 5.8] ng/L, respectively, P<0.001, Figure 1A), with women exhibiting a relatively larger increase with advancing age as compared with men (Pinteraction<0.001, Figure 1B). Apart from age, a significant and divergent interaction with sex was found for the association between cardiac troponin and body mass index (Pinteraction=0.008) and diabetes (Pinteraction=0.003). During follow-up, cardiac troponin concentrations were associated to the outcome in both women and men (adjusted HR per two-fold difference [95% CI], 1.34 [1.17-1.52] and 1.30 [1.21-1.40], respectively, Pinteraction=0.752). The slope of cardiac troponin was significantly associated with the outcome in women, but not in men (adjusted HR [95% CI], 2.70 [1.01-7.33] and 1.31 [0.62-2.75], respectively, Pinteraction=0.250). Conclusion Trajectories of cardiac troponin differ between women and men in the general population, with differing associations to conventional risk factors and cardiovascular outcomes. Our findings highlight the importance of a sex-specific approach when serial cardiac troponin testing is applied for cardiovascular risk prediction.