Associations among cardiorespiratory fitness, C-reactive protein, and all-cause mortality in men and women

Abstract

We examined individual and joint associations among high-sensitivity C-reactive protein (CRP), cardiorespiratory fitness (fitness), and mortality in healthy men and women. Between January 1, 2000 and December 31, 2016, 30,077 adults (31.3% women) received a comprehensive physical examination. Fitness was determined from maximal treadmill exercise test duration. Participants were categorized as unfit (Quintile 1) and fit (Quintiles 2–5), and by normal (<2 mg/L) and elevated (≥2 mg/L) CRP categories. Adjusted hazard ratios (HRs) with 95% confidence interval (CI) for all-cause mortality were computed with Cox regression. During an average of 10.1 years of follow-up, 576 deaths occurred. Following adjustment for age, smoking status, sex, exam year, body mass index, systolic blood pressure, total cholesterol, triglyceride:high-density lipoprotein ratio, and fasting glucose, HR (95% CI) for all-cause mortality were 1.0 (referent) and 1.52 (1.14–2.02) for fit and unfit categories, respectively. Corresponding values for normal and elevated CRP categories were 1.0 and 1.50 (1.20–1.89), respectively. When grouped by fitness and CRP category, there was significantly greater mortality risk in the unfit than the fit category within the elevated CRP category (HR = 1.77 (1.14–2.75)), but not in the normal CRP category (HR = 1.38 (0.96–1.98)). Each 1 metabolic equivalent increment in fitness and 1 mg/L increment in CRP were associated with 10.0% (95% CI: 5.1–14.8%) decreased and 7.3% (95% CI: 2.0%–12.9%) increased mortality hazard, respectively. Compared to the unfit, fit individuals have an attenuated mortality risk within each CRP category. Thus, higher fitness appears to provide some protection against all-cause mortality, particularly among those with elevated levels of inflammation.