Benefits of intensive resistance training in patients with chronic polymyositis or dermatomyositis

Authors


Abstract

Objective

To investigate the benefits and safety of an intensive muscular training program in patients with chronic polymyositis (PM) and dermatomyositis (DM).

Methods

Nine patients with chronic PM or DM (median age 53 years, range 44–61) were included. Assessments of impairment (10–15 voluntary repetition maximum [VRM], the Functional Index 2 [FI-2], the Grippit, and pain rated on the Borg CR-10 scale), activity limitation (Myositis Activities Profile), and participation restriction (patients' disease impact on well-being) were performed 4 weeks prior to baseline, at baseline, and after 7 weeks of exercise. A 6-item core set of disease activity measures was administered and muscle biopsy samples of vastus lateralis were obtained at baseline and after 7 weeks of exercise. Response criteria at an individual level were set for disability and disease activity. The patients exercised 3 days per week for 7 weeks on loads allowing 10 VRM.

Results

On a group level there were no significant differences between assessments at 4 weeks before baseline compared with baseline. The group improved significantly regarding 10–15 VRM and FI-2 at 7 weeks compared with baseline (P < 0.05). All patients were responders with respect to impairment and 2 were activity limitation responders whereas participation restriction remained unchanged in all. Two patients were responders with reduced disease activity and no patient had signs of increased muscle inflammation in the muscle biopsy sample after 7 weeks of exercise.

Conclusion

Patients with chronic, stable PM and DM can perform this intensive resistive exercise program with beneficial effects on impairment and activity limitation without increased muscle inflammation.

INTRODUCTION

Polymyositis (PM) and dermatomyositis (DM) are chronic, idiopathic inflammatory myopathies that are characterized by impairment (reduced muscle strength and endurance as well as fatigue and myalgia) (1, 2). Despite the initial favorable effects of aggressive immunosuppressive treatment in most patients with PM or DM, a majority develop sustained disability (3). The mechanisms for this effect are not fully understood and may vary during different phases of the disease. The inflammation is likely to cause muscle impairment in the early phase. Secondary metabolic disturbances including reduced ATP levels in muscle tissue could contribute to impairment (4–6). In the late chronic phase, other factors, such as muscle atrophy, steroid myopathy, and disuse of muscle due to physical inactivity, could also play a role (7).

Physical exercise could be one way to counteract several of the mechanisms that are believed to have a negative effect on muscle function in patients with PM or DM. However, this group of patients has been previously discouraged from active exercise due to fear of exacerbation of disease activity. This cautious attitude is based on previous studies revealing increased creatine phosphokinase levels (CPK) and signs of inflammation in muscle tissue after excessive exercise in healthy individuals (8, 9). Several studies demonstrating the safety of easy-to-moderate exercise programs and efficacy in reducing disability in patients with PM and DM have been published during the last decade (10–18), and resistive exercise is now recommended in the rehabilitation of these patients (19, 20).

In healthy individuals, improved muscle function during strengthening exercise is due to improved neuromuscular adaptation (21), increased cross-sectional muscle fiber area (22, 23), changes in fiber type composition (24), and metabolic changes (25–27). Healthy sedentary individuals usually respond to any type of exercise (28, 29) and can attain a 40% increased muscle strength in 4 weeks, whereas longer periods of exercise are necessary to obtain the same effects in more trained individuals (23). An exercise intensity of 40–50% of 1 repetition maximum (1 RM) is sufficient to obtain increased strength in untrained individuals, whereas 80% of 1 RM is needed for more trained persons (30, 31). According to the literature, it appears that untrained individuals benefit from both single- and multiple-set exercise (32) and that it is necessary to exercise 2–3 times per week to increase muscle strength (33). In healthy muscle, overlap effects of strength exercise to improve muscular endurance to some extent have been described. The resting period between sets should be 1–2 minutes for isolated muscle groups and 2–3 minutes when exercising several joints (23).

It has not yet been determined how different types of training affect muscle strength and muscle endurance in patients with PM or DM. Moreover, it is not known whether these patients tolerate and benefit from more intensive muscular resistance exercise. The goal of this study was to evaluate the potential benefit of an intensive resistive muscular training program in patients with chronic PM and DM regarding different aspects of disability and to ensure the safety of the program with regards to effect on disease activity.

PATIENTS AND METHODS

Patients.

All patients meeting the inclusion criteria for PM or DM with regular followup at the rheumatology clinic at Karolinska University Hospital (Stockholm, Sweden) were invited to participate (n = 29). Inclusion criteria were as follows: diagnosis of definite or probable PM or DM according to the Bohan and Peter criteria (34), disease duration >12 months, and unchanged disease activity and medication for 3 months. Exclusion criteria were diagnosis of inclusion body myositis, comorbidities preventing vigorous exercise, severe osteoporosis defined as multiple vertebral fractures, current malignant disease, and exercising more than once per week. Exercise was defined as a physical activity with the goal to improve fitness (35).

Eleven patients accepted the invitation and gave informed consent. In the weeks before study entry, 2 patients were excluded due to flares of their disease. Therefore, 9 patients (5 women and 4 men) were included in the study. Five of the patients had a DM diagnosis and 4 had PM. One of the patients with PM had also been diagnosed with Sjögren's syndrome. The median age was 53 years (range 44–61 years), the median disease duration since diagnosis was 4.5 years (range 2.7–29 years), and the median prednisone dosage was 2.5 mg/day (range 0–7.5). In addition, 6 patients were treated with azathioprine, 1 in combination with cyclosporin A, and 1 patient was treated with methotrexate. Three patients were working full time, 4 were working part time, and 2 were retired or on sick leave. Three patients exercised once per week, either aerobics or aquatic exercise. The other 6 patients did not perform any physical activity other than activities of daily living and ordinary walks.

Assessments of impairment.

Patients performed 10–15 voluntary repetition maximum (10–15 VRM) for 5 muscle groups: deltoids and quadriceps muscles separately for the right and left sides, biceps/latissimus dorsi and the gastrocnemius muscles bilaterally, and the abdominal muscles. To perform 10–15 VRM, an individual weight load was selected that allowed the performance of 10–15 correct repetitions, but not more. This load was ∼70% of 1 VRM.

The Functional Index 2 (FI-2) measures muscle endurance and comprises 7 functional tasks: shoulder flexion and shoulder abduction separately on the right and left sides, head lifts in supine position, hip flexion and step test separately on the right and left sides, and bilateral heel lifts and toe lifts. Each task is scored individually as the number of correctly performed repetitions. The maximum number of repetitions is 60 for all except the 2 latter tasks, for which the maximum number of repetitions is 120 (36).

The Borg CR-10 scale, a category scale with ratio properties in which numbers are anchored to verbal expressions, was used to assess pain. The Borg CR-10 scale ranges from 0 (no pain) to 10 (very severe pain) (37). Grip strength was evaluated using the Grippit, which is an electronic force instrument measuring both maximum force and mean force in newtons over a 10-second period (38).

Activity limitation assessments.

The Myositis Activities Profile (MAP) is a disease-specific questionnaire assessing activity limitation in patients with PM or DM. The MAP is divided into 4 subscales (movement activities, activities of moving around, self-care activities, and domestic activities) and 4 single items (social activities, avoiding overexertion, work/school work, and leisure activities). Each item is scored on a 7-grade scale from 1 (no problem to do) to 7 (impossible to do). Each subscale is scored as the median of all items included and the single items are scored as the actual item score (39).

Participation restriction assessment.

Patients' global assessment of the overall impact of disease on well-being was rated on a 0–10 visual analog scale (VAS), where 0 indicates no impact of disease and 10 indicates very severe impact of disease. Assessments of impairment and activity limitation/participation restriction were performed by an independent observer on 3 occasions: 4 weeks prior to baseline, at baseline, and after 7 weeks of exercise. These assessments were always performed in the same order and at the same time of day.

Disease activity assessments.

The 6-item core set of disease activity measures developed by the International Myositis Assessment and Clinical Studies Group (IMACS) includes physician's and patient's global assessment of disease activity rated on a 0–10 VAS, the Manual Muscle Test (MMT), the Health Assessment Questionnaire (HAQ), serum levels of CPK (normal values <2.5 and <3.0 μcat/liter for women and men, respectively), and extraskeletal muscle activity (Myositis Intention to Treat Activity Index [MITAX]) (40). The MMT includes assessment of isometric strength in 8 muscle groups: neck flexors, gluteus medius and maximus, deltoids, biceps brachii, wrist extensors, quadriceps, and dorsiflexors of the ankle. Each muscle group is scored from 0 to 10, with a total score varying from 0 to 80 (where 80 = full strength). The HAQ comprises 20 questions divided into 8 categories: dressing and grooming, arising, eating, walking, hygiene, reach, grip, and other activity. The HAQ is scored from 0 to 3 (where 3 = unable to do) (41). The MITAX includes 7 organ systems: constitutional, cutaneous, skeletal, gastrointestinal, pulmonary, cardiovascular, and muscle. The total score is the sum of all organ systems varying from 0 to 63 (where 0 = no disease activity).

Muscle biopsy samples from the vastus lateralis were obtained using a conchotome under local anesthesia (42, 43). The repeated biopsy sample was obtained from the contralateral side to that obtained before training. The biopsy samples were analyzed by a neuropathologist and the grade of inflammation was registered as the degree of mononuclear inflammatory cell infiltrates according to a 4-grade scale from 0 (no inflammatory cells) to 3+ (pronounced cellular infiltration). All disease activity measures were performed 1 week prior to baseline and 1 week after completing the exercise period.

The exercise program.

Patients exercised 3 days per week for 7 weeks at the physical therapy department at Karolinska University Hospital. A physical therapist supervised all patients individually on all exercise occasions. The patients warmed up on either an ergometer cycle or a treadmill for 10 minutes at 50% of their individual estimated maximal heart rate. After warming up, patients performed a 45-minute exercise program with a load of their individual 10 VRM in 5 muscle groups: deltoid muscles by shoulder flexion/abduction with free weights, the quadriceps by knee extension in sitting position with weight cuffs around the ankles, the latissimus dorsi/biceps muscles and the gastrocnemius in a training apparatus, and the trunk muscles by performing situps with/without weight cuffs placed on the chest. These exercises were performed in 3 sets separated by 90-second rests. The program ended with 5 minutes of stretching. During the first week all patients exercised at 50% of their 10 VRM performing 15 repetitions of each exercise in 3 sets. During the following 2 weeks, the resistance was gradually increased to represent 100% of their individual 10 VRM performed in 10 repetitions per set. After 3 and 5 weeks of exercise, respectively, new 10 VRM tests for each muscle group were performed and the exercise loads were adjusted.

Ethical scrutiny.

The local ethical committee at the Karolinska Hospital, Stockholm, approved this study and all patients gave their informed consent to participate.

Statistical analysis.

For analysis on a group level, repeated-measures analysis of variance (ANOVA) was used to compare ratio data for the 3 evaluation occasions and the Friedman ANOVA was used for ordinal data. The level of significance was set to P less than 0.05.

All results were also analyzed individually for each patient, and improvement criteria according to Paulus et al (44) for measures of muscle impairment, activity limitation, and participation restriction were applied. Thus, changes of at least 20% in ≥1 test/task of the 10 VRM, the FI-2, and the Grippit, respectively, were considered as clinically relevant. For variables producing ordinal data, changes of 2 scale points were considered clinically important for the MAP (≥1 subscale/item) and the Borg CR-10 scale, and a change of 20 mm was considered clinically important for the VAS. Patients were considered responders if they improved in 2 of 4 impairment measures (10–15 VRM, FI-2, Grippit, Borg CR-10), in 1 activity limitation measure (MAP), or in 1 participation restriction measure (patients' overall assessment). Regarding disease activity, improvement criteria developed by the IMACS for patients with myositis were used (45), in which a responder should improve by ≥20% in at least 3 of the 6 parameters in the core set with no more than 2 parameters worsened by ≥25% (which cannot include the MMT assessed muscle strength).

RESULTS

After 3 weeks, 1 patient with PM reported influenza-like symptoms and was excluded from the exercise program. The results are therefore based on the 8 remaining patients. The program was well tolerated. Of the possible 21 exercise occasions, the patients exercised on a median of 19 occasions (range 14–21). One patient went on a 1-week holiday during the study, and therefore only exercised for 6 weeks. Medication remained unchanged in all patients and no patient changed any lifestyle factors such as physical activity, smoking, or food intake.

All patients experienced mild delayed-onset muscle soreness for 1 or 2 days after some exercise occasions. One patient experienced increased joint swelling and tenderness of the metacarpophalangeal joints and was therefore unable to perform the latissimus dorsi/biceps exercise during the last 2 weeks. This patient also reported pain in a sternocostal joint and was not able to perform the situp tests of the 10–15 VRM and FI-2 at 7 weeks. Another patient experienced diffuse tendinitis pain of the shoulders and performed the deltoid exercises at ∼80% of the 10 VRM during the entire exercise period.

Impairment changes.

No significant differences occurred in the group during the 4 weeks prior to baseline. The group improved significantly at 7 weeks compared with baseline (P < 0.05) regarding the 10–15 VRM for all muscle groups except for the latissimus dorsi/biceps muscles and in muscle endurance of shoulder flexion bilaterally as assessed by the FI-2 (P < 0.05) (Table 1).

Table 1. Impairment measures at 4 weeks prior to baseline, at baseline, and after 7 weeks of exercise in 8 patients*
Variables4 weeks prior (n = 8)Baseline (n = 8)7 weeks (n = 8)P value baseline vs. 7 weeks
  • *

    Values are the mean ± SD unless otherwise indicated. VRM = voluntary repetition maximum; NS = nonsignificant; FI-2 = Functional Index 2; reps = repetitions.

10–15 VRM, muscle strength    
 Deltoid, right, kg4.5 ± 3.14.7 ± 3.16.9 ± 4.0< 0.01
 Deltoid, left, kg4.5 ± 3.14.6 ± 2.96.3 ± 3.7< 0.05
 Quadriceps, right, kg14.4 ± 3.613.1 ± 2.918.9 ± 5.7< 0.05
 Quadriceps, left, kg14.1 ± 3.813.1 ± 2.919.0 ± 5.7< 0.01
 Latissimus dorsi/biceps, kg51.3 ± 15.353.1 ± 16.057.5 ± 30.6NS
 Gastrocnemius, kg101.3 ± 20.3103.8 ± 19.8138.7 ± 26.7< 0.01
 Abdominal muscles, kg0.6 ± 1.71.2 ± 1.66.5 ± 7.1< 0.05
FI-2, muscle endurance    
 Shoulder flexion, right (0–60 reps)31.6 ± 20.633.8 ± 18.343.6 ± 23.0< 0.05
 Shoulder flexion, left (0–60 reps)28.3 ± 18.329.3 ± 15.843.6 ± 22.0< 0.05
 Shoulder abduction, right (0–60 reps)31.1 ± 20.537.9 ± 21.138.1 ± 21.8NS
 Shoulder abduction, left (0–60 reps)30.0 ± 19.134.4 ± 20.045.6 ± 22.4NS
 Head lift (0–60 reps)14.4 ± 13.812.8 ± 11.115.5 ± 13.4NS
 Hip flexion, right (0–60 reps)23.8 ± 20.427.0 ± 15.730.2 ± 19.0NS
 Hip flexion, left (0–60 reps)22.1 ± 18.126.4 ± 16.128.4 ± 20.6NS
 Step test, right (0–60 reps)37.1 ± 23.347.4 ± 18.943.4 ± 23.4NS
 Step test, left (0–60 reps)38.9 ± 22.946.9 ± 18.949.3 ± 19.9NS
 Heel lift (0–120 reps)74.6 ± 46.571.6 ± 44.976.1 ± 40.4NS
 Toe lift (0–120 reps)49.3 ± 41.150.3 ± 44.769.1 ± 42.6NS
Grippit, grip strength    
 Max, right, newtons284.0 ± 121.9263.5 ± 108.7283.5 ± 103.6NS
 Max, left, newtons273.0 ± 101.0268.5 ± 105.0261.1 ± 87.6NS
 Mean, right, newtons245.3 ± 101.9240.3 ± 99.0254.0 ± 85.5NS
 Mean, right, newtons237.1 ± 85.0238.5 ± 93.4232.6 ± 74.1NS
Muscular pain    
 Borg CR-10 scale, median (range) (0–10)2.5 (0.0–3.0)1.3 (0.0–3.0)1.3 (0.0–3.0)NS

All 8 patients were responders according to our criteria for impairment reduction. All patients improved by ≥20% of the 10–15 VRM compared with baseline (Figures 1 and 2) with improvements for all muscle groups ranging from 20% to 900% (Figure 2). No patient had deterioration in any of the muscle groups.

Figure 1.

Changes in 10 voluntary repetition maximum (VRM) after 7 weeks of exercise compared with 4 weeks before baseline and baseline for A, musculus deltoids, right side, B, musculus deltoids, left side, C, musculus quadriceps, right side, D, musculus quadriceps, left side, E, musculus latissimus dorsi/biceps (patient H not assessed), F, musculus gastrocnemius, and G, trunk muscles. * >20% improvement at 7 weeks compared with baseline. ** >40% improvement at 7 weeks compared with baseline.

Figure 2.

Number of responders at >20%, 40%, 60%, and 100% level of 10–15 VRM of deltoids and quadriceps on the right side, latissimus (lat.) dorsi/biceps, gastrocnemius, and trunk muscles after 7 weeks of exercise compared with baseline.

All patients responded at the ≥20% level set for the FI-2; however, 5 patients worsened in a few tasks (Table 2). Only 1 patient responded according to the criteria set for maximum and mean score of the Grippit on the right side, with 34% and 45% improvements, respectively. No patient improved by ≥2 scale points in pain assessed by the Borg CR-10. No patient experienced ≥20% deterioration in grip strength or pain.

Table 2. Number of tasks improving or worsening according to the Functional Index 2*
Patient, sexDiagnosisTasks improving ≥20%, no.Tasks worsening ≥20%, no. (task)
  • *

    DM = dermatomyositis; PM = polymyositis.

A, maleDM21 (heel lift)
B, femaleDM34 (shoulder flexion, right; hip flexion, right; step test, right; heel lift)
C, femalePM100
D, maleDM20
E, femaleDM10
F, maleDM54 (hip flexion, right and left; step test, right and left)
G, malePM71 (shoulder abduction, right; step test, right)
H, femalePM52 (shoulder abduction, right; hip flexion, right)

Changes in activity limitation and participation restriction.

The group did not improve significantly in activity limitation or participation restriction (Table 3). Two patients were responders according to our criteria for reduced activity limitation in the MAP subscales movement and activities of moving around and the single item social activities. One patient had a decrease of >2 scale points in the MAP single item leisure activities. No changes of >20 mm on the VAS for overall disease impact occurred in any patient.

Table 3. Activity limitation and participation restriction measures at 4 weeks prior to baseline, at baseline, and after 7 weeks of exercise*
Variables4 weeks prior (n = 8)Baseline (n = 8)7 weeks (n = 8)P value baseline vs 7 weeks
  • *

    Values are the median (range) unless otherwise indicated. MAP = Myositis Activities Profile; NS = nonsignificant; VAS = visual analog scale; NA = not assessed.

MAP subscales    
 Movement, 1–73.0 (1.0–4.0)3.0 (1.0–5.0)3.0 (1.0–4.0)NS
 Moving around, 1–74.0 (1.0–5.0)4.0 (1.0–5.0)3.0 (1.0–4.0)NS
 Self care, 1–73.0 (1.0–4.0)2.5 (1.0–4.0)2.5 (1.0–3.0)NS
 Domestic, 1–73.5 (1.0–4.0)3.0 (1.0–4.0)3.0 (1.0–4.0)NS
MAP single items    
 Social, 1–71.0 (1.0–5.0)1.5 (1.0–5.0)1.0 (1.0–5.0)NS
 Avoid overexertion, 1–72.5 (1.0–4.0)3.0 (1.0–6.0)3.5 (1.0–6.0)NS
 Work, 1–74.0 (1.0–5.0)4.0 (1.0–6.0)4.0 (1.0–6.0)NS
 Leisure, 1–74.0 (1.0–5.0)4.0 (1.0–7.0)4.0 (1.0–7.0)NS
Patient's global assessment of disease impact on well-being    
 VAS, 0–10NA2.9 (0.7–5.5)3.3 (1.0–4.8)NS

Disease activity.

On a group level, there was a statistically significant improved MITAX score at 7 weeks compared with baseline (P < 0.05), whereas all other disease activity measures remained unchanged. Two patients were considered as responders fulfilling the IMACS improvement criteria for disease activity (Table 4). One patient improved in 2 variables without worsening in any, 1 patient improved in 2 variables and deteriorated in 1, 2 patients improved and deteriorated in 2 variables, respectively, and 2 patients improved in 1 variable but worsened in 2 variables (Table 4). No patient improved or worsened ≥25% in muscle strength assessed by the MMT.

Table 4. Results of muscle biopsies and 6-item core set to measure disease activity in 8 patients before and after 7 weeks of exercise*
Patient, sexDiagnosisDegree of inflammationMMT, 0–80HAQ, 0–3CPK, μcat/literPhysician's global assessment of disease activity, 0–10 VASPatient's global assessment of disease activity, 0–10 VASMITAX total score, 0–63
First biopsySecond biopsyBaseline7 wksBaseline7 wksBaseline7 wksBaseline7 wksBaseline7 wksBaseline7 wks
  • *

    MMT = Manual Muscle Test; wks = weeks; HAQ = Health Assessment Questionnaire; CPK = creatine phosphokinase (normal values <2.5 μcat/liter in women, <3.0 μcat/liter in men); VAS = visual analog scale; MITAX = Myositis Intention to Treat Activity Index; DM = dermatomyositis; PM = polymyositis; 0 = negative staining; (+) = scattered mononuclear inflammatory cells; + = a few infiltrates with mononuclear inflammatory cells; ++ = several infiltrates with mononuclear inflammatory cells; NA = not analyzed due to end-stage muscle.

  • ≥20% improvement compared with baseline.

  • ≥25% worsening compared with baseline.

A, maleDM0071730.630.753.51.61.31.43.03.753
B, femaleDM0074770.880.501.11.00.71.30.81.0126
C, femalePM(+)(+)68760.630.252.12.10.10.01.71.043
D, maleDMNANA63580.750.6324.615.63.02.74.54.82418
E, femaleDM+073730.000.000.71.10.10.00.71.452
F, maleDM0074760.881.132.01.00.40.13.13.944
G, malePM0073791.131.384.52.01.00.45.54.243
H, femalePM++++67710.500.753.01.90.00.82.72.899

At study start, 3 patients had small infiltrates or scattered mononuclear inflammatory cells in their biopsy samples. After 7 weeks of exercise, there were no signs of increased inflammatory infiltrates (Table 4). In 1 patient the biopsy sample could not be analyzed due to end-stage muscle (patient D).

DISCUSSION

After this intensive resistance training program, a major improvement with reduced impairment was achieved along with limited benefits regarding activity limitation and disease activity. The exercise program was generally well tolerated by the patients in that no one experienced increased muscle pain other than short-term delayed-onset muscle soreness.

The 20% improvement criteria were set to represent clinically relevant changes in the outcome measures of impairment (44). This level is concurrent with the disease activity criteria proposed by the IMACS group (45). The improvement criterion for activity limitation (MAP) was defined as a change of 2 scale points. Because the MAP is scored as the median value of items within a subscale, a calculation of percent change would not have been appropriate. This definition did not seem to confer less sensitivity to change than the application of a 20% change in HAQ score, proposed by the IMACS group, because limited changes were recorded for both measures. For the disease impact on well-being, a scale change of 20 mm was chosen as a significant level of change. Depending where on the VAS the initial value was located, the 20% change proposed by the IMACS group (45) for patient's and physician's global assessment might be within the error of measurement.

All patients achieved clinically significant improvement with decreased impairment in several muscle groups of the 10–15 VRM and tasks of the FI-2. A limitation of this study is the uncontrolled design; however, the fact that the group improved significantly in the 10–15 VRM and FI-2 after 7 weeks compared with baseline while no significant differences were observed during the preceding 4-week nontraining period confirms the benefits of this exercise program. The marked improvement in the 10–15 VRM test is comparable with the effects of improved muscle strength that can be achieved in healthy sedentary individuals with comparable workloads after 4 weeks of exercise (23). Our results are also concurrent with others reporting statistically significant and clinically relevant improvements in muscle function after a short duration of similar exercise programs (10, 11, 14–18). However, because no study has evaluated such an intensive muscular training program and because different outcome measures were used, the degree of improvement is hard to establish. Five patients worsened in various unilateral muscle groups of the FI-2, mainly in the lower limbs, which might raise some caution. The step test of the FI-2, which worsened in 3 patients, was the single task that put the highest demands on fitness and it is likely that variations between test conditions, e.g., in indoor climate or generalized fatigue, might have played a role in the outcome. The deterioration in the other tasks of the FI-2 occurred only on one side of the body without any consistent pattern or obvious explanation. The improved muscle strength and the absence of significant worsening of the patients' assessment of disease impact on well-being in all patients argue against a significant impact of the deteriorated FI-2 variables. No patients rated increased muscle pain after 7 weeks. Their initial rating of pain varied from no pain to moderate, indicating that muscle pain was not a major problem in this small group of patients. Patients with arthralgias and arthritis did not tolerate parts of the exercise program as well as the patients without joint pain, indicating that the loads should be slightly reduced in patients with myositis with additional arthritis.

Two patients were responders with reduced activity limitation assessed by the MAP. The lack of significant improvements in activity limitation on a group level assessed by both the MAP and the HAQ in our study is inconsistent with other studies reporting improvements after 6 weeks of cardiovascular training (12, 13) and only 3 weeks of muscular training using the HAQ (16). This discrepancy might be explained by differences in exercise regimens and study power. A longer exercise period would probably have resulted in larger changes in activity limitation. Furthermore, 2 patients initially scored too low on the MAP (indicating no or slight activity limitation) to allow for a 2-point improvement. The deterioration of the single MAP item leisure activities reported by one patient (patient D) was not associated with worsening in muscle strength, endurance, or disease activity and might be ascribed to the extensive time and energy required to participate in this study.

The exercise program did not result in increased inflammation as assessed by analysis of muscle biopsy samples and CPK levels, which agrees with previously reported results following exercise programs of similar intensity in patients with PM, DM, or inclusion body myositis (12, 13, 46). Although 4 patients worsened >25% in 1–3 parameters of the 6-item core set, these changes were not considered as clinically relevant because they ranged within the error of measurement. The 57% increase of CPK levels in 1 patient was also within range of the normal value of 2.5 μcat/liter. Two patients worsened >25% in the HAQ, which could be defined as a clinically relevant change. One of these patients had increased joint pain in the hands, which could explain the increased HAQ score. There is no obvious explanation for increased HAQ score in the other patient.

Definitions of different exercise protocols for healthy individuals have previously been described in the literature as follows: low repetition/high intensity varying between 3–6 and 6–8 repetitions, moderate repetition/moderate intensity varying between 9–11, 15–20, and 30–40 repetitions, and high repetition/low intensity varying between 20–28, 30–40, and 100–150 repetitions (47–49). According to these definitions, our exercise program could be defined as intermediate with more strength-training properties than endurance training. It was therefore expected that the exercise program would mainly result in improvements of the 10–15 VRM tests. However, because the majority of the 8 patients were untrained and had impaired muscle function, larger overlap benefits of muscular endurance, as assessed by the FI-2, were expected (23). One explanation for the predominant 10–15 VRM improvements could be that the tests and the exercise program involved the same muscle groups in the same range of motion and at the same velocity, which could have resulted in a significant neuromuscular adaptation effect. The FI-2 also evaluated the exercised muscle groups but in other starting positions and velocities. Because the 10–15 VRM evaluations were performed before the FI-2, the patients' possible lack of motivation to perform their maximum number of repetitions in the FI-2 might have played a role. Another interpretation is that during inflammatory disease, muscle might not respond to exercise with the same beneficial overlap effects as in healthy individuals. Therefore, it is possible that the exercise program was too strength-training oriented.

The 10–15 VRM was measured rather than 1 VRM, mainly for safety reasons. Such maximal exercise bouts are hard to perform correctly and could increase the risk for injuries because these patients were not accustomed to exercising at this intensity. It would have been preferable to measure exactly 10 VRM to minimize the error of measurement. However, the 10–15 VRM was used to minimize the number of tests before reaching the correct load to avoid overuse because these measures were followed by the rather energy-requiring FI-2 and the grip strength measure. However, the number of repetitions did not vary >2 repetitions in most patients in most muscle groups.

Because PM and DM are rare diseases, a limited number of patients were included. No prestudy power calculation was conducted and it is possible that this study was underpowered. However, statistically significant and clinically relevant improvements in muscle function and disease activity were detected. A larger number of patients could have also resulted in statistically significant changes in activity limitation.

In conclusion, this intensive exercise program can be safely used with beneficial effects leading to reduced muscle impairment and, to some extent, reduced activity limitation and disease activity in patients with chronic PM or DM. The exercise program may have to be adjusted due to pain levels in joints and tendons for patients with arthritis-associated myositis. Further studies exploring disease mechanisms causing impairment and evaluating long-term effects of exercise need to be conducted.

AUTHOR CONTRIBUTIONS

Dr. Alexanderson had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study design. Alexanderson, Esbjörnsson-Liljedahl, Opava, Lundberg.

Acquisition of data. Alexanderson, Dastmalchi, Lundberg.

Analysis and interpretation of data. Alexanderson, Dastmalchi, Esbjörnsson-Liljedahl, Opava, Lundberg.

Manuscript preparation. Alexanderson, Dastmalchi, Opava, Lundberg.

Statistical analysis. Alexanderson.

Acknowledgements

We thank all patients for participating in this study; Helen Langeland for performing all disability measures; Christina Ottosson for coordinating the study, assisting at muscle biopsies, and obtaining blood samples; Eva Lindroos and Sevim Barbasso Helmers for handling the muscle biopsy samples; Inger Nennesmo for analyzing the muscle biopsy samples; and associate professor Ronald van Vollenhoven for linguistic reading of the manuscript.

Ancillary