Article Text

Download PDFPDF

Extended report
Individualised aerobic and resistance exercise training improves cardiorespiratory fitness and reduces cardiovascular risk in patients with rheumatoid arthritis
  1. Antonios Stavropoulos-Kalinoglou1,2,3,
  2. Giorgos S Metsios1,4,
  3. Jet JJCS Veldhuijzen van Zanten1,5,
  4. Peter Nightingale6,
  5. George D Kitas1,7,
  6. Yiannis Koutedakis2,3,4
  1. 1Department of Rheumatology, Dudley Group NHS Foundation Trust, Russell's Hall Hospital, Dudley, UK
  2. 2Department of Sport and Exercise Science, University of Thessaly, Trikala, Greece
  3. 3Centre for Research and Technology Thessaly, Institute of Human Performance & Rehabilitation, Trikala, Greece
  4. 4School of Sport, Performing Arts & Leisure, Wolverhampton University, Walsall, UK
  5. 5School of Sports and Exercise, University of Birmingham, Edgbaston, UK
  6. 6Wellcome Trust Clinical Research Facility Birmingham, Queen Elizabeth Hospital Birmingham, Birmingham, UK
  7. 7Arthritis Research UK Epidemiology Unit, University of Manchester, Manchester, UK
  1. Correspondence to Professor George D Kitas, Department of Rheumatology, Dudley Hospitals NHS Foundation Trust, Russell's Hall Hospital, Pensnett Road, Dudley DY1 2HQ, UK;kitas{at}


Background and objectives Low cardiorespiratory fitness (CRF) is a significant predictor of cardiovascular disease (CVD), and interventions aiming at increasing CRF are known to reduce CVD risk. The effects of such interventions on CVD risk have not been studied in patients with rheumatoid arthritis (RA).

Methods 40 age, gender, body mass index (BMI) and disease duration matched RA patients were allocated to either an exercise (receiving 6 months individualised aerobic and resistance high intensity exercise intervention, three times per week), or control (receiving advice on exercise benefits and lifestyle changes) arm. Participants were assessed at baseline, 3 and 6 months for aerobic capacity (VO2max), individual CVD risk factors (blood pressure, lipids, insulin resistance, body composition), 10-year CVD event probability and RA characteristics (C-reactive protein (CRP), Disease Activity Score 28 (DAS28) and Health Assessment Questionnaire (HAQ)).

Results There were no differences between groups at baseline in any of the assessed variables. VO2max (p=0.001), blood pressure (systolic: p<0.001; diastolic: p=0.003), triglycerides (p=0.030), high density lipoprotein (HDL; p=0.042), total cholesterol:HDL ratio (p=0.005), BMI (p=0.001), body fat (p=0.026), 10-year CVD event probability (p=0.012), CRP (p=0.042), DAS28 (p=0.008) and HAQ (p=0.003) were all significantly improved in the exercise versus the control group. The change in VO2max was the strongest predictor for the observed improvements in all of the assessed CVD risk factors and disease characteristics.

Conclusions Individualised aerobic and resistance exercise intervention can lead to significantly improved CRF, individual CVD risk factors, composite CVD risk, and disease activity and severity in RA patients.

  • Rheumatoid Arthritis
  • Cardiovascular Disease
  • Lipids

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


Rheumatoid arthritis (RA) is associated with increased overall mortality compared to the general population, particularly due to cardiovascular disease (CVD).1 Classical risk factors2 ,3 and systemic inflammation4 ,5 contribute towards this adverse CVD profile. RA patients also exhibit relatively low levels of cardiorespiratory fitness (CRF).6 Generally, appropriate CRF protects from CVD mortality,7 ,8 even in the presence of CVD risk factors,9 ,10 and high levels of CRF are associated with low levels of inflammation.11 Even though CRF has a familial component,12 ,13 it can be significantly increased by exercise training, regardless of age, gender, race and initial fitness levels.14 ,15

Despite current scientific evidence showing that exercise has multiple benefits for RA patients, most of them lead a sedentary lifestyle16; low levels of habitual physical activity are associated with increased CVD risk in RA.17 Physically active patients with RA tend to be hospitalised less often and for a shorter duration compared to their less physically active counterparts.18 Therefore, exercise may help improve CVD risk and overall health status of RA patients. However, the potential cardiovascular benefits of engaging in purposeful exercise have not been formally assessed in such patients. This study aimed to evaluate the effects of an individualised exercise intervention programme on CRF and CVD risk in patients with RA.



Consecutive RA patients attending routine clinics of the Department of Rheumatology, Dudley Group NHS Foundation Trust, UK, were approached. Inclusion criteria were: RA fulfilling 1987 revised American College of Rheumatology criteria,19 sedentary lifestyle (no participation in structured exercise for the preceding 6 months), and stable disease (no changes in disease-modifying antirheumatic drugs (DMARDs)—including biologics—or oral steroids and no parenteral steroid administration in the last 3 months). Exclusion criteria were: joint surgery (in the preceding 6 months), amputation, and co-morbidity incompatible with exercise as per American College of Sports Medicine guidelines.20

A case matched design based on age (±3 years), gender, body mass index (BMI) (±1 kg/m2 but within the same BMI category), and disease duration (±2 years) was used to allocate patients to either the exercise or the control arm. Cases (ie, patients included in the intervention group) were identified and approached. Following recruitment of each case, patients from the remaining RA population of the hospital (∼1500 patients) matching their characteristics were also approached. On average, a list of five matches was available for each case. Patients from this list were approached in order of visit to the clinics. A total of 45 potential controls were approached to reach the required number of 20 participating controls.

Participants in the exercise group received a 6-month individualised exercise intervention, while the control group only received verbal advice about the cardiovascular and arthritis-related benefits of exercise for the same period.21 ,22 All participants received relevant information leaflets of the British Heart Foundation and Arthritis Research UK. During the 6-month training period, participants and their managing consultants were asked to avoid changes in DMARDs or steroid administration.

General procedures

The trial was registered with the ISRCTN register (ISRCTN50861407). All assessments were conducted within the clinical research unit, and exercise training took place in ‘Action Heart’, the cardiac rehabilitation centre, of Dudley Group NHS Foundation Trust. On signing the informed consent, an appointment was made to evaluate the participant's cardiorespiratory status with an exercise tolerance test (ETT). Based on each patient's ETT outcome, and functional ability, an individualised exercise training programme was prescribed. Exercise training started within 2 weeks of ETT. Demographic and anthropometric data, physical function, disease activity, lifestyle parameters and CVD risk factors were assessed at baseline and repeated following 3 and 6 months of intervention.

The first three visits were reserved for instruction on the use of relevant exercise equipment and induction to the individualised exercise programme. Thereafter, patients exercised in a semi-supervised manner. Two of the researchers, ASK and GM, are exercise physiologists with experience in RA patients and were responsible for exercise prescription and induction to the programme. Resident exercise supervisors in AH provided the supervision for the remaining duration. Any issues arising during the exercise sessions would be reported to the researchers who would in turn adjust the exercise programme.

Participants in the control group were given information during their first visit to the testing venue. Thereafter, they were contacted monthly over the phone for further advice and support. During their 3- and 6-month assessments, they were also invited for a one-to-one discussion on lifestyle changes. Assessors were blinded to group allocation throughout the study and patients were instructed to refrain from discussing their intervention with them.

Individualised exercise prescription

Based on the ETT results (ie, CRF as measured by maximal oxygen consumption: VO2max), an individualised exercise programme was prescribed for each participant utilising exercise apparatus such as treadmills, and cycle, hand and rowing ergometers. The decision about the type of equipment to be used depended on the patient's preferences and perceived ability, and the exercise physiologist's assessment of their ability to attain the training objectives.

During each exercise session, patients were asked to maintain a given intensity (ie, heart rate corresponding to 70% VO2max) for a total of 30–40 min. Each session consisted of three circuits. In each circuit, participants were asked to perform 3–4 different exercises (ie, walk on treadmill, cycle, row or use the hand ergometer) for no more than 3–4 min each, with a resting interval of about 1 min. Total time for each session was 50–60 min, including 10 min warm-up, 30–40 min main session, and 5–10 min cooling down. This schedule was implemented for the first 3 months of the exercise programme.

Thereafter, some resistance training was also added to the above schedule. Patients were asked to perform four exercises utilising large muscle groups (eg, leg press, shoulder press, chest press and pull ups). Intensity was determined during their first resistance exercise session using an established protocol.20 They were required to complete three sets of 12–15 repetitions during each exercise session. Mode and intensity of exercise were adjusted monthly (box 1).

Box 1

Design of the exercise intervention

Exercise prescription

  • Baseline to 3 months (aerobic training)

  • Frequency: Three times per week; two at a supervised setting (ie, Action Heart), one unsupervised

  • Intensity: 70% of VO2max; as indicated by heart rate corresponding to 70% of VO2max attained during the exercise tolerance test

  • Type: Three circuits of 3–4 aerobic exercises in intervals of 3–4 min each; aerobic exercises included: brisk walking on treadmill, cycling on stationary bicycle, rowing on row-ergometer, hand-cycling on arm-ergometer

  • Time: 60 min; including 10 min warm-up (gentle range of motion exercises), 30–40 min main session (as described in type above), 5–10 min cool down.

  • 3–6 months (aerobic and resistance training)

In addition to the above, resistance exercises were added as follows:

  • Frequency: as above

  • Intensity: as above for aerobic exercise; resistance exercise 70% of 1RM as indicated by a submaximal protocol (4–6RM)

  • Type: In addition to the above at the end of each circuit, patients were asked to complete one set of 12–15 repetitions of the following resistance exercises: leg press, shoulder press, chest press and pull ups

  • Time: 70 min; resistance exercises added 10 min to the main session

Participants executed three exercise sessions per week (with at least one rest-day in between), two of which were in a supervised environment (Action Heart); the third one took place in the patient's home. Heart rate monitors (Polar S610i, Polar Electro Oy, Kempele, Finland) were used during each session to ascertain exercising at the prescribed intensity and monitor adherence to the exercise targets; adherence was calculated as a ratio of exercises where target heart rate was achieved versus total number of exercises in each session. Attendance was recorded by monitoring the number of visits to the exercise venue; percentage attendance was calculated. After the end of the 6-month intervention, participants in both groups were invited to continue exercising as part of the routine activities of Action Heart.


Demographic and anthropometric data

Demographic data were collected using a self-administered questionnaire. Standing height was measured to the nearest 0.5 cm (Seca 214 Road Rod). Weight and body composition (ie, body fat and fat free mass) were evaluated using a Tanita BC-418 MA Segmental Body Composition Analyser (Tanita Corporation, Tokyo, Japan) and BMI was calculated.

Lifestyle assessments

The long form self-administered International Physical Activity Questionnaire was used to assess 7-day physical activity.23 Overall energy expenditure for the preceding week was calculated.

Cardiorespiratory fitness

All participants were subjected to an ETT on a mechanical treadmill (HP Cosmos Mercury, Nussdorf-Traunstein, Germany). CRF was determined on the basis of maximal oxygen uptake (VO2max) using a calibrated breath-by-breath system (Metalyzer 3B, CORTEX, Leipzig, Germany). Heart rate corresponding to 70% VO2max was identified and used to determine optimum exercise intensity.

An individualised ramp test protocol, based on the guidelines of the American Heart Association,24 was used. Such protocols are better suited to clinical settings than protocols using large intervals or steps.25 Testing started at a convenient speed for the participants (in most cases 2.0 mph) and 1% inclination. Speed increased by 0.5 mph every 1 min until 4.0 mph. Thereafter, inclination of the treadmill would increase every 30 s by 1%. Testing was terminated when the participant reached volitional exhaustion or the criteria for terminating a VO2max test had been met. In all cases exhaustion was achieved within 7–12 min.

Individual CVD risk factors

Blood pressure (BP) was assessed following at least 5 min rest, on the right arm with the patient in a seated position.26 Reported value is the mean of three readings taken at 5-min intervals. Blood lipids, glucose and insulin were assessed in venous blood collected in the fasting state. Insulin sensitivity was evaluated with the Homeostasis Model Assessment of insulin resistance (HOMA=(Glucose×Insulin)/22.5)27 and the Quantitative Insulin sensitivity Check Index (QUICKI=1/(logInsulin+logGlucose)).28 Smoking status was recorded by patient self-report. Surrogate 10-year CVD event probability was established using the HeartScore programme ( of the European Society of Cardiology. The ‘Europe High Risk’ option was used.

RA assessments

Contemporary inflammation was evaluated by the erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP). The Disease Activity Score-28 (DAS28) was used to assess clinical disease activity.29 The anglicised version of the Stanford Health Assessment Questionnaire (HAQ)30 was used to assess functional disability. Disease duration and RA medication were obtained from patients’ clinical notes.

Power calculations

CRF was selected as the primary end-point for its importance as a surrogate predictor of CVD events and because it is the only relevant parameter that has been previously assessed pre- and post-exercise in RA patients.31 Assuming a detectable difference of 5 ml/min/kg with 4 ml/min/kg SD32 and 90% power, a sample of 15 patients is required; to allow for 25% drop-out rate we recruited 20 patients per group (nQuery Advisor V.6.0, Statistical Solutions, Massachusetts, USA).

Data management and analyses

Repeated measures analysis of variance (ANOVA) with one between-subjects factor (group: exercise vs control) and one within-subjects factor (time: baseline vs 3 months vs 6 months) were implemented to investigate the changes in CRF and CVD risk factors between the groups at the three different time points. Bonferroni adjustment was used to detect where the differences in the assessment points occurred (baseline vs 3 months vs 6 months). Finally, to investigate the effects of disease characteristics, attendance and adherence on CRF change, and the effects of inflammation (CRP) and CRF (VO2max) on the observed changes in CVD risk factors and disease characteristics on the two groups, we utilised generalised estimating equations (GEE). Like repeated measures ANOVA, GEE take into account the correlation of measurements on the same subjects but, unlike repeated measures ANOVA, they can use time dependent covariates. We used unstructured correlation matrices where possible (DAS, HAQ, CRP, triglycerides, BMI), and M-dependent (M=2) for the remaining variables (VO2max, systolic and diastolic BP, total and high density lipoprotein (HDL) cholesterol and their ratio, body fat (BF), HeartScore). Data were analysed using SPSS V.18.0 and the level of significance was set at p<0.05.


From the 40 patients recruited, four dropped out: two from the exercise group, after the 3-month assessment (one due to an ulcer and one due to arrhythmia), and two from the control group, after the baseline assessment (loss of interest in the study). Attendance was 88%, and adherence 76%. Participants’ baseline characteristics are shown in table 1 (baseline characteristics between the groups were similar).

Table 1

Baseline demographic, anthropometric, RA-related, CRF and CVD characteristics for the total RA population as well as the exercise and control groups

Effects of exercise on cardiorespiratory fitness

Repeated measures ANOVA detected a significant change over time (p<0.001) in the exercise but not the control group in VO2max (figure 1). GEE identified adherence (B=3.2, p<0.001) and attendance (B=4.1, p<0.001) as the strongest predictors for the change in VO2max.

Figure 1

Change in VO2max from baseline to 3- and 6-month time-points for exercise versus control groups. *Significantly different compared with control group (3 months: p=0.023; 6 months: p=0.002).

Effects of exercise on cardiovascular risk factors

Results are summarised in table 2. GEE revealed that VO2max was a strong predictor for the observed changes in systolic (B=1.7, p<0.001) and diastolic (B=1.2, p<0.001) BP. Although VO2max was not a strong predictor for the changes in triglycerides (B=0.02, p=0.5) or total cholesterol (B=0.1, p=0.091), it was the strongest and sole predictor for HDL cholesterol (B=0.41, p<0.001). Among the other factors used in the models (BMI, BF, CRP, disease duration, number of previous medications, biologic medication), BMI was a significant predictor for systolic (B=4.0, p<0.001) and diastolic (B=1.0, p=0.005) BP, while BF was only a significant predictor for diastolic BP (B=0.6, p=0.050). Similarly, VO2max was a strong predictor for the observed reduction in 10-year CVD event probability (B=0.1, p=0.002). Finally, VO2max was again the strongest and sole predictor of changes in BMI (B=0.9, p<0.001) and BF (B=0.4, p=0.001).

Table 2

Change in cardiovascular risk factors and body composition over time

Effects of exercise on disease characteristics

Changes in disease severity (HAQ score) and activity (DAS28) over time were significantly different in the two groups (p=0.003 and 0.008, respectively) with significant reductions in the exercise group but not in the control group. Also, repeated measures ANOVA detected a significant group×time interaction (p=0.042) for CRP (table 2).

VO2max was the strongest predictor for the changes observed in DAS28 (B=2.2, p<0.001) and HAQ (B=0.7; p=0.001), but not for CRP (B=−0.2, p=0.065). Other covariates included in the models were: BMI, BF, disease duration, number of previous medications, and biologic therapy. Of these, only BF was a significant predictor for the change observed in CRP (B=0.3, p=0.044).


To our knowledge, this is the first study to investigate the effects of individualised exercise training on the cardiovascular profile of patients with RA. It shows that the exercise form used in the present study can significantly improve CRF, several CVD risk factors, and physical function in these patients.

Aerobic capacity significantly increased in the exercise group but not in the control group. This is partly due to the fact that individualised exercise interventions result in higher attendance and adherence in several populations, including RA33–35; indeed, in our study, patients who visited the exercise venue more frequently and trained harder experienced the largest increases in VO2. This is in line with data from other disease groups,36 ,37 athletes,38 ,39 and RA patients.40

Based on their baseline VO2max (24.2 ml/min/kg), our participants would be classified as having a poor or very poor (depending on age) aerobic fitness, which associates with an increased risk for ill health.41 Following exercise, however, VO2max improved by >10% and 17% within 3 and 6 months, respectively, compared to baseline; the 3-month VO2max value of 27.1 ml/min/kg classified our participants as average, while their 6-month value of 28.3 ml/min/kg was above average for the younger participants (<60 years of age) and good for those above 60.42

Aerobic capacity is a strong and independent predictor of CVD and overall mortality,43–45 and increasing VO2max levels associate with reduced prevalence and severity of CVD.46 Exercise has major beneficial effects on reducing CVD risk and individual risk factors in healthy47 and disease populations.48 Indeed, both systolic and diastolic BP decreased at 3 months and continued decreasing throughout the duration of the study. The exact mechanism by which this occurs is not completely understood; it is thought that aerobic training decreases blood pressure through a reduction of vascular resistance, in which both the sympathetic nervous system and the renin–angiotensin system are involved.49

Blood lipids also appeared to improve although changes were mainly observed at the 6-month time-point. Especially important is the change in total cholesterol:HDL ratio as this has been shown to be a more valid predictor of future CVD event than cholesterol measures only.50 Previous studies in the general population have indicated that physical exercise drastically improves plasma HDL cholesterol, and to a lesser extent, triglycerides or total cholesterol.51 ,52 However, the duration of the intervention seems to be critical as studies utilising longer interventions (ie, 2 years) show a more decisive effect of exercise on blood lipids.53 It is still not clear whether this is a direct effect of exercise, or mediated via body weight reductions and, consequently, fat.54

Surprisingly, insulin and markers of insulin resistance and sensitivity (HOMA and QUICKI, respectively) were unaffected by the present exercise training regime. Exercise is known to improve insulin action in the muscle and endothelium.55 Perhaps the intervention length of time might not have been sufficient to trigger the necessary processes that would improve insulin sensitivity, given that such changes may require longer periods of adaptation through continuing exercise. Despite this, overall CVD risk was significantly reduced in the exercise but not in the control group, mainly as a result of the observed reductions in BP and the increase in HDL.

Body composition indices were also significantly improved by exercise. This is particularly important due to the prevalence of muscle wasting in RA.56 These patients commonly present with a condition known as rheumatoid cachexia, where they exhibit low fat free mass without any significant weight changes.57 This condition associates with higher disease activity, increased morbidity and risk for CVD.58 Individualised exercise training seems to be effective in reducing weight and BF while maintaining muscle mass.

There are a few limitations in our study. Foremost is the lack of a randomised design. Our approach may leave room for bias since cases were motivated to participate in an exercise intervention while controls might not be. However, we do not believe that this may have affected the results of this study. Moreover, the tight matching of the two groups for age, gender, BMI and disease duration might not have been possible in a randomised design. Generalisability may be another limitation; we conducted our intervention in a large and well equipped rehabilitation centre, within hospital premises, which may not be available in other settings. Due to the supervised setting, we were able to respond to any changes in our patients’ ability to exercise and adjust their training schedule accordingly. However, in terms of the facility, virtually no extra costs were incurred. Resident exercise supervisors were educated on the specific requirements of RA patients during a 1-day course and the patients were incorporated in their normal schedule. Finally, the group of patients that volunteered for the study had relatively well controlled RA and a low CVD risk even at baseline, but the beneficial changes may have been even more impressive in patients with worse baseline measures.

In conclusion, this study demonstrates that individualised aerobic and resistance exercise training associates with significant improvements in CRF and individual CVD risk factors, and reduction of CVD risk in RA patients, and confirms that these patients are able to participate in this form of exercise without adverse effects on their condition. The long-term effects of exercise interventions on disease activity, overall health and survival in RA should be further evaluated.


View Abstract


  • Handling editor Tore K Kvien

  • Contributors ASK: conception, design, patient recruitment, assessments, exercise prescription, data analyses, manuscript preparation. GSM: conception, patient recruitment, exercise prescription, exercise supervision, data analyses, manuscript preparation. JJJCSVvZ: design, patient recruitment, assessments, data management, paper editing. PN: design, data management and analyses, paper editing. GDK: conception, design, patient recruitment, clinical supervision, paper editing. YK: conception, design, exercise prescription, paper editing, final approval.

  • Funding This study was funded by the Dudley Group of Hospitals R&D Directorate cardiovascular programme grant and a Wolverhampton University equipment grant. The Department of Rheumatology has an infrastructure support grant from the Arthritis Research Campaign (number: 17682).

  • Competing interests None.

  • Ethics approval Black Country Ethics Committee.

  • Provenance and peer review Not commissioned; externally peer reviewed.