Cardiorespiratory fitness and mortality in rheumatoid arthritis

Abstract

Background Rheumatoid arthritis (RA) is an autoimmune inflammatory disease that affects synovial joints and internal organs. It affects around 1% of the population, and the disease is associated with increased all-cause mortality rates compared with the general population. Through the last decades, medical treatment of RA has improved largely. Treatment strategies aim at early medical intervention to reduce inflammation and disease activity to improve function and inhibit joint destructions. As far as we know, there is still no robust evidence that improved medical treatment with modern drugs has reduced the mortality gap between RA patients and the general population. Cardiorespiratory fitness (CRF) is an important modifiable predictor of all-cause mortality in the general population. Still, there is no robust evidence that this is true for RA patients. Some studies have demonstrated that RA patients have reduced CRF compared to healthy people, which may contribute to their increased mortality rates. The gold standard test to quantify CRF is to measure a person’s maximum oxygen uptake (VO2max) during cardiopulmonary exercise testing (CPET) on a bicycle ergometer a treadmill. Less resource-intensive estimation equations for CRF (eCRF) without the need for a physical test have been developed for the general population, but have not been optimized for persons with RA. There is also a knowledge gap concerning which variables are associated with CRF in RA patients. Aims This study aimed to - investigate which variables are associated with CRF in RA patients - develop eCRF equations suitable for RA patients - investigate differences in eCRF and age-related change in eCRF in RA patients and controls - investigate which variables are associated with the eCRF change - investigate which variables are associated with mortality in RA patients and controls - study possible consequences from low eCRF on mortality in an RA population Methods During 2017, 93 RA patients from St. Olavs Hospital’s rheumatology outpatient clinic were recruited for CPET to measure their VO2max. We also collected RA-specific variables like presence of autoantibodies and various measures of disease activity along with vital measures like blood pressure, pulse, and body mass index (BMI). Multiple linear regression was then used to identify variables that were associated with CRF in RA patients and to develop new eCRF equations suitable for RA patients. The new eCRF equations were used to calculate the eCRF of RA patients attending the second and third waves of the Trøndelag Health Study, HUNT2 (1995-1997) and HUNT3 (2006-2008). eCRF results were then compared with results from the general population in HUNT2 and HUNT3. Multiple linear regression with change in eCRF from HUNT2 to HUNT3 as the dependent variable was used to investigate if CRF in RA patients decreased faster with increasing age compared to the general population in HUNT. Furthermore, multiple linear regression was used to find variables associated with eCRF change. All-cause mortality in RA patients and controls in HUNT2 and HUNT3 was analyzed using Cox proportional hazard regression. The analyses were stratified on sex, and age was the time variable. The date of the first participation in HUNT2 or HUNT3 was the baseline, and participants were followed until they died, or until the December 31st, 2018. To investigate the effect of having RA on mortality and to answer the question “How much of the associations of RA with all-cause mortality is caused by low eCRF?”, a Cox regression-based mediation analysis was performed. Results BMI, physical activity, systolic blood pressure, resting heart rate, and smoking were associated with VO2max in RA patients. The only RA-specific variable associated with VO2max was the patient’s global assessment. Our investigations resulted in five new eCRF equations with some variations regarding variables to allow for the calculation of eCRF according to data availability. One eCRF equation for individual use (RAfitCALC) was published online. eCRF in RA patients was lower and eCRF decreased more rapidly with increasing age in RA patients compared to the general population. In addition to sex and RA status, age, baseline eCRF, smoking, cardiovascular disease, BMI, high-density lipoprotein concentration, asthma, and hypertension were associated with the change in eCRF from HUNT2 to HUNT3. Using Cox regression, low eCRF was associated with mortality both in RA patients and controls from the general population. This was also true after adjustment for hypertension, BMI, smoking, total cholesterol, diabetes, and creatinine concentration. The mediation analysis showed that RA patients had a 28% excess risk of all-cause mortality compared to controls. The direct effect of RA was 5%, the indirect effect of RA via low eCRF was 4%, and the effect from an interaction between RA and low eCRF accounted for 19%. Conclusion and clinical implications Low physical fitness is an underestimated risk factor for premature death in patients with RA, and its contribution to excess all-cause mortality by far exceeded the isolated effect of having RA. Along with medical treatment, measures to improve physical fitness in RA patients should be part of early intervention strategies to reduce the mortality gap between RA patients and the general population. In addition to investigations about eCRF in RA patients attending large population-based studies, the new eCRF equations developed for RA patients make it possible for patients and physicians to follow eCRF improvements after a period of relevant physical training. Furthermore, physicians can easily identify if an RA patient has an eCRF level that needs to be addressed for better health. In this way the new eCRF equations can contribute to improved fitness in RA patients.