Secondary Logo

Journal Logo


Effects of Structured Exercise Training in Children and Adolescents With Juvenile Idiopathic Arthritis

Klepper, Susan PT, MS, PhD; Mano Khong, Taryn Tamiko PT, DPT; Klotz, Rachel PT, DPT; Gregorek, Adrienne Odmark PT, DPT; Chan, Y Chee PT, DPT, CSCS; Sawade, Samantha PT, DPT

Author Information
doi: 10.1097/PEP.0000000000000555


In the article cited above, one author's name was spelled incorrectly in the originally published version. The correct full author name is Taryn Tamiko Mano Khong, DPT. This has been corrected in the online version of the article.

Pediatric Physical Therapy. 31(2):232, April 2019.


Juvenile idiopathic arthritis (JIA) is defined as arthritis of unknown etiology with onset before the age of 16 years and lasting at least 6 weeks1 Reports of worldwide prevalence range from 7 to 401 cases/100 000 and incidence from 0.8 to 22.6 cases/100 000 youth per year.2,3 An estimated 294 000 children in the United States have some type of JIA.4 Gender distribution and onset age differ by disease type, although more females are affected. Diagnosis is one of exclusion, ruling out other possible causes of inflammation.1 The International League of Associations for Rheumatology recognizes 7 disease types (see Supplemental Digital Content 1, available at: An eighth, “undifferentiated,” includes characteristics that do not fit into any single type or overlap more than one.1,5

Despite advances in pharmacologic management of inflammation, many children may experience a chronic course with periods of active disease continuing into adulthood.1 Muscle atrophy and weakness are common and may not resolve completely during disease remission.6 Research indicates individuals with JIA are less physically fit and active than peers without JIA.7 A 2014 systematic review (SR) by Cavallo et al,8 covering years 1995-2011, found lower levels of self-reported moderate-to-vigorous physical activity (MVPA) compared with controls who were healthy. Insufficient daily MVPA may contribute to impaired physical fitness in JIA.9–12 Fitness deficits, although most pronounced in children with severe disease, are reported in those with mild symptoms and may persist after arthritis resolves. Impaired exercise capacity and delayed development of complex motor skills may contribute to limited participation in age-appropriate MVPA, sports, and decreased overall quality of life (QoL).13,14

Significant improvements in management of inflammation have changed the focus of care from emphasizing rest to encouraging an active lifestyle.15 Research suggests children with JIA can participate in exercise testing and structured exercise training (ET) without disease flares.7,9,16,17 Studies in adults with rheumatoid arthritis demonstrate improvements in aerobic capacity, muscle strength, and performance of daily activities following ET.18 Three previous SRs19–21 and 1 general review,22 published between 2008 and 2017, provide evidence for the safety of structured ET in children with JIA whose disease is under good control. A 2008 Cochrane review by Takken et al,19 including 3 randomized controlled trials (RCTs), found no significant between-group (BG) differences, although results favored the ET group compared with a control group (CG) that participated in less intense or no exercise. Two more recent RCTs reported positive outcomes in aerobic and muscular function and improved functional capability following ET. The 2016 Cavallo et al20 Ottawa Panel Evidence-Based Clinical Practice Guidelines (EBCPG) reported favorable outcomes from 5 RCTs to support structured ET in disease management in JIA. Recently, Kuntze et al21 reported evidence for the benefits of physical therapy (PT) based ET in JIA. A general review by Catania et al22 including RCTs and uncontrolled studies reported benefits of structured ET in JIA but provided no effect size (ES) data. The purpose of this SR was to update the evidence for safety and efficacy of structured ET conducted within a variety of settings—hospital PT departments, participant's home, or professional fitness centers—and aimed at improving health-related physical fitness (HRPF), physical function, pain, and QoL in children and adolescents with JIA. Comparison activities included PT, a less intense physical activity (PA) program, wait-list group, or no intervention.


Search Strategy and Eligibility Criteria

This SR was conducted using guidelines from the Cochrane Collaboration for SRs and reported using the format and checklist in the 2009 Preferred Reporting Items for Systematic Reviews and meta-Analyses (PRISMA) statement for reviews of health care interventions.23,24 Inclusion criteria were (1) participants were children and adolescents ages 4 to 21 years with a definite diagnosis of JIA made before age 16 years; (2) interventions were exercise programs focused on land or water-based aerobic or resistance training or general physical conditioning conducted in home or center-based settings; and (3) studies were RCTs published in English. Studies were excluded if they (1) were not RCTs; (2) included multiple interventions administered simultaneously; and (3) included participants younger than 4 years or older than 21 years or with diagnoses in addition to JIA. Primary outcomes were 3 components of HRPF: aerobic fitness/performance, muscle strength, and joint flexibility. Accepted measures of muscle strength were manual muscle testing, hand-held dynamometry (HHD), or isokinetic testing. Secondary outcomes were physical function, QoL, bone density, and pain intensity.

A comprehensive English-language search strategy, developed and validated with the assistance of an expert medical research librarian, was conducted in January 2014 and updated in April 2018 to locate all controlled exercise trials for children and adolescents with JIA conducted between January 1966 and December 2017. Five electronic databases (MEDLINE, PubMed, CINAHL, PEDro, and Web of Science) were included (see Supplemental Digital Content 2, available at:

Selection of Studies

Two reviewers, working individually, conducted the search to identify eligible articles based on title and abstract. Duplicates and articles not meeting selection criteria for study population or design were removed. Five reviewers individually examined the full text of remaining articles to determine eligibility for inclusion in the SR. The Critical Review Form for Quantitative Studies by Law et al25 provided a structured format to assess study purpose, adequacy of literature review, design, sources of potential bias, outcomes, intervention, and results. Based on this information, reviewers reached a majority consensus on eligibility of individual articles for inclusion in this study. Articles were excluded if the full English-language manuscripts were unavailable. Reference lists in all selected articles were hand searched for additional studies that potentially met inclusion criteria.

Data Extraction

Key information (age, gender, JIA type, and geographic location of participants) was extracted from each study using data collection forms designed for this review. Year and source of publication, sample size, inclusion criterion, research design and methods, intervention type, outcomes measured, and results were recorded. Reviewers worked in pairs to check each other for accuracy and completeness of data extracted from each article. Discrepancies within or between individual pairs of reviewers were checked and resolved by the first author who read and analyzed all studies. Trial authors were contacted if necessary information was unclear or missing, including description of study design or statistical analysis.

Data Analysis

Within-group (WG) and between-group (BG) differences and standard deviations were extracted or calculated for relevant outcomes in each study providing continuous data. Statistical significance for each outcome was extracted to compare similar outcomes across studies. Clinical importance was judged by BG ES and 95% confidence interval (CI) based on the standardized mean difference (SMD) (Cohen's d) or simple mean difference calculated by Cochrane RevMan software ( Strength of ES for outcomes was interpreted using Cohen's convention: small (0.20), medium (0.50), or large (0.80). ES less than 0.20 was considered insignificant.26 Heterogeneity of studies was assessed by analysis of data using χ² and I² statistics. A random-effects model was used to determine whether outcomes met criteria to be pooled for meta-analysis in order to account for heterogeneity in measures. When outcomes were reported as median (range), and raw data were unavailable, statistical significance was used to compare these outcomes with other studies.

Assessment of Risk of Bias and Study Quality

All studies were assessed for internal validity and risk of bias with the PEDro scale (scores range from 1-10; higher scores indicate better quality) using scores extracted from the Physiotherapy Evidence Database27 ( The Oxford Centers for Evidence-Based Medicine rating system was used to judge the level of evidence for each study ( Quality of evidence for outcomes was assessed using the Cochrane GRADE system (Grading of Recommendations, Assessment, Development and Evaluation: high, moderate, low, very low) based on level of evidence and biases in 5 domains: study design, indirectness of the evidence, unexplained study heterogeneity or inconsistency, imprecise results, and probability of publication bias.28 Strength of clinical recommendations (extent to which one can be confident benefits of an intervention outweigh potential adverse effects) was judged using the GRADE scale: strong negative (−), weak negative (−) to weak (+), and strong (+).29


Figure 1 depicts the study selection process based on inclusion criteria. After screening, 9 articles, representing 8 RCTs, met all inclusion criteria and were analyzed individually by each review author.10,30–37 Two articles reported separate outcomes from a single RCT.34,35 Studies were published between 2003 and 2016 and conducted in 7 countries: Sweden, the Netherlands, Canada, the United Kingdom, Turkey, Egypt, and Brazil.

Fig. 1.
Fig. 1.:
Study flow diagram.

Level of Evidence and Risk of Bias in Included Studies

All studies were RCTs with a level of evidence38 judged to be 2b (Table 1). The mean PEDro score (see Supplemental Digital Content 3, available at was 6.56/10, with a range of 430 to 8.32,33,36 The least risk of bias was in random allocation, similarity of groups at baseline, reporting BG differences and point estimate and variability. The most prominent risk concerned lack of blinding of participants and therapists to intervention in all studies and lack of blinding of assessors in 4 studies.10,31,34,35 Additional threats to external validity within all studies included failure to include all JIA disease types, inadequate sample size to detect significant effects of ET, and failure to report test-retest reliability of outcome measures in their sample. However, pediatric rheumatology-specific measures including the Pediatric Escola Paulista de Medicina Range of Motion Scale (pEPM-ROM), Child Health Assessment Questionnaire (CHAQ), Pediatric Quality of Life Inventory (PedsQL), and the health-related quality of life (HRQOL) scale as well as the Child Health Questionnaire (CHQ) have demonstrated acceptable reliability.39–42 Reliability in pediatric samples has been reported for the Harvard Step Test,43 6-minute walk test,44 HHD45 and isokinetic testing.46

TABLE 1 - Characteristics of Studies
Study Level of Evidence Experimental Group Control/Alternate Group Outcome (Baseline to 3 mo) Measure
Balance/Proprioceptive Exercise Strengthening Exercise
N (Male) Age, Mean ± SD (Range), y JIA Type/Disease Duration, y N (Male) Age, Mean ± (SD) (Range), y JIA Type/Disease Duration, y
Baydogan et al30(2b) 15 (5) 10.00 ± 3.66 (6-18) PolyRF+ = 2
PolyRF− = 5
Oligo = 6
JPsA = 2
15 (4) 9.27 ± 1.43 (6-18) PolyRF+ = 2
PolyRF− = 7
Oligo = 5
JPsA = 1
Aerobic fitness
Muscle strength
Functional capability
Hydrotherapy + Land-Based PT Land-Based PT Only (Land) Outcome (Baseline to 2 mo) Measure
N (Male) Age, Mean ± SD (Range), y JIA Type/Disease Duration, y N (Male) Age, Mean ± SD (Range), y JIA Type/Disease Duration, y
Epps et al32 (2b) 39 (15) 11 (4-19) Poly = 15
Systemic = 5
Oligo = 3
OligoE = 8
ERA = 8
JPsA = 0
39 (20) 12 (6-19) Poly = 18
Systemic = 5
Oligo = 4
Oligo(E) = 7
ERA = 4
JPsA = 1
Aerobic fitness
Muscle strength (hip ABD; knee extension; shoulder ABD)
Quality of life
Functional capability
Submaximal HR
Time to exhaustion
HHD (baseline to 2 mo)
CHQ (physical)
CHQ (psychological)
Pain (0-10 cm)
CHAQ (0-3)
Stott Pilates Conventional exercise program Outcome (Baseline to 6 mo) Measure
N (Male) Age, Mean ± SD (Range), y JIA Type/Disease Duration, y N (Male) Age, Mean ± SD (Range), y JIA Type/Disease Duration, y
Mendonca et al33 25 (9) 11.8 ± 3.4 (8-18) Poly =7
Oligo = 14
Syst = 4
RF+ = 6
25 (9) 11.0 ± 3.9 (8-18) Poly = 5
Oligo = 18
Systemic = 2
RF+ = 4
Quality of life
Functional capability
PedsQL 4.0 (patients—total)
VAS—joint pain
Physical Exercise Group Usual exercise/activity Outcome (Baseline to 3 mo) Measure
N (Male) Age, Mean ± SD (Range), y JIA Type/Disease Duration, y N (Male) Age, Mean ± SD (Range), y JIA Type/Disease Duration, y
Sandstedt et al34(2b) Bone mineral density DXA
Physical Exercise Group Usual Exercise/Activity Outcome (Baseline – 3 mo) Measure
Sandstedt et al35(2b) 33 (8) 13.3 (8.8-19.9) Poly = 20
Oligo = 7
ERA/JPsA = 1
21 (4) 14.9 (8.8 – 20.6) Poly = 9
Oligo = 8
ERA/JPsA = 3
Aerobic fitness
Muscle strength
Quality of life
Physical capability
Borg RPE Scale
HHD (arms & legs)
High-Intensity Aerobic Training Qi-Gong Outcome (Baseline to 3 mo) Measure
N (Male) Age, Mean ± SD (Range), y JIA Type/Disease Duration, y N (Male) Age, Mean ± SD (Range), y JIA Type/Disease Duration, y
Singh-Grewal et al36 (2b) 41 (6) 11.7 ± 2.5 (8-16) Poly = 19
Oligo(P) = 5
Oligo(E) = 6
Syst = 1
ERA = 7
JPsA = 2
Other = 1
39 (10) 11.5 ± 2.4 (8-16) PolyRF+ = 2
PolyRF− = 13
Oligo(P) = 2
Oligo(E) = 5
Syst = 6
ERA = 4
JPsA = 6
Other = 1
Aerobic fitness
Quality of life
Physical function
Relative O 2submax
Relative O 2peak
Exercise Training in Pool Assessment Only—Control Group Outcome (Baseline to 3 mo) Measure
N (Male) Age, Mean ± SD (Range), y JIA Type/Disease Duration, y N (Male) Age, Mean ± SD (Range), y JIA Type/Disease Duration, y
Takken et al10 (2b) 27 (11) 8.66 ± 2.29 (5-13) Poly = 15
Oligo = 11
Syst = 1
27 (3) 8.88 ± 1.86 (5-13) Poly = 14
Oligo = 12
Syst = 1
Aerobic fitness
Quality of life
Functional Performance
O 2peak
Combined Resistive Underwater Exercises and Interferential Current Therapy Traditional Physical Therapy Program (Full Description Not Provided) Outcome (Baseline to 3 mo) Measure
N (Male) Age, Mean ± SD (Range), y JIA Type/Disease Duration, y N (Male) Age, Mean ± SD (Range), y JIA Type/Disease Duration, y
Elnaggar and Elshafey31(2b) 15 (sex not given) 10.1 ± 1.2 (age range not provided) Poly with bilateral knee arthritis 15
(sex not given)
9.7 ± 1.5 (age range not provided) Poly with bilateral knee arthritis Muscle strengthPain Peak torque (bilateral hamstrings and quadriceps)
Land-Based Home Exercise Program Wait-List Control Group Outcome (Baseline to 3 mo) Measure
Tarakci et al37 (2b) 43 (18) 10.02 ± 3.44 (5-17) Poly = 27
Oligo = 14
Systemic = 1
JPsA = 1
38 (19) 10.82 ± 4.00 (5-17) Poly = 19
Oligo = 16
Systemic = 3
JPsA = 0
Aerobic fitness
Quality of life
Functional capability
Abbreviations: ABD, abductor; CHAQ, Childhood Health Assessment Questionnaire (lower scores indicate better function); CHQ, Child Health Questionnaire; CHQ-PhS, Child Health Questionnaire-Physical Summary; CHQ-PsS, Child Health Questionnaire-Psycho-Social Summary; DXA, dual-energy x-ray absorptiometry; ERA, enthesitis-related arthritis; ESR, erythrocyte sedimentation rate; FBT, Flamingo Balance Test; FRT, Functional Reach Test; HEP, home exercise program; HHD, hand-held dynamometry; HR, heart rate; HRQOL, health-related quality of life; JAFAS, Juvenile Arthritis Functional Assessment Scale; JIA, juvenile idiopathic arthritis; JPsA, juvenile psoriatic arthritis; NRS, numeric rating scale; NS, not significant; Oligo(E), oligoarticular-extended; Oligo(P), oligoarticular-persistent; PedsQL, Pediatric Quality of Life Inventory; pEPM-ROM, Pediatric Escola Paulista de Medicina Range of Motion (lower score indicates improvement); PolyJIA, polyarticular JIA; PolyJIA RF+, PolyJIA rheumatoid factor positive; PolyRF−, PolyJIA rheumatoid factor negative; PSCS, psychosocial; PT, physical therapy; QOL, quality of life; ROM, range of motion; RPE, Rating of Perceived Exertion; SD, standard deviation; 6MWT, 6-minute walk test; 10MWT, 10-m walk test; 10SCT, 10-stair climbing test; VAS, visual analog scale.

Participants and Interventions

This review included 457 participants (31% male) with JIA, ages 4 to 19.9 years, recruited through pediatric rheumatology centers. Although all disease types were represented, the most common was polyarticular JIA (experimental group [EG] = 52.5%; CG = 47%). Individual sample size ranged from 3030,31 to 81.37 Supplemental Digital Content 4 (available at: includes characteristics of interventions, following the TIDieR checklist for intervention reporting.47 Children in the EG participated in structured ET designed by a PT or exercise professional knowledgeable about JIA. Interventions included, Stott Pilates, aquatic exercise alone or combined with land exercise, balance/proprioception exercise, strength training, and high-repetition rope jumping. Intervention settings included hospital-based or local PT practices, community swimming pools, and participants' homes. A PT or trained instructor led all exercise sessions with the exception of 2 home exercise programs (HEPs) supervised by parents.34,35,37 One protocol combined an HEP 3 times/week with a 1 time/week hospital-based session instructed by a PT.37 Across studies participants in the CG completed an alternate exercise regimen,30,32,33,36 received their usual PT,31 were assigned to a wait-list37 or completed assessment only.10,34,35 Total exercise dosage (20-40 hours), minutes/session (45-60), sessions/week (1-4), and program length (10-24 weeks) varied across trials. Several studies reported a detailed description of the exercise program allowing a guide to replication.30,33–35,37


Primary outcomes included components of HRPF: aerobic fitness, muscle strength, and joint range of motion (ROM). Five studies measured aerobic function: 2 using O2peak and O2submax,10,36 1 the modified Harvard Step Test,35 1 the Borg Rating of Perceived Exertion (RPE) Scale,35 and 2 the 6-minute walk distance (6MWD).10,37 Three studies measured muscle strength using a hand-held dynamometer.30,32,35 Two measured joint ROM using goniometry30,35 and 3 using the pEPM-ROM scale for JIA.10,33,36

Secondary outcomes included bone density, pain intensity, physical function, and QoL. One study measured bone density, using dual-energy x-ray absorptiometry (DXA) and DXA laser Calscan, before and after a program of repetitive rope jumping and strength training.34 Five studies reported outcomes on QoL using 1 of 3 measures: the visual analog scale (VAS),36 CHQ,10,32,35 and Pediatric Quality of Life (PedsQL) self- or parent-report.33,37 Seven studies assessed physical function (capability) using the CHAQ.10,30,32,33,35–37 Five measured pain intensity, 4 using a VAS32,33,35,37 and 1 the numeric pain rating scale.30

Analysis of the Evidence

Aerobic Fitness/Performance.Table 2 and Figure 2 show statistical and clinical results by outcome across studies. These confirm the findings of Takken et al,19 who reported no significant BG changes in O2peak after training (n = 123; SMD 0.08, 95% CI −0.27 to 0.44). Three studies32,35,36 that used measures of exercise effort—O2submax, HR submax, and Borg RPE Scale—found no significant BG differences after ET. Of 3 studies reporting effects of ET on aerobic performance (time to exhaustion on a bicycle ergometer32 or 6MWD10,37), one37 found statistically significant improvement favoring the EG following an HEP; however, results were not clinically significant.

Fig. 2.
Fig. 2.:
Measures of aerobic function.
TABLE 2 - Effects of Exercise Training in JIA: Statistical and Clinical Results by Outcome
Outcome Study Intervention/Length/Total Time Outcome Measure (ICF Classification)a Timing of Test BGb Statistical Significance BG Clinical Significance (ES) Cohen's dc η2d GRADE29
Quality of Evidencee (Favorable) Strength of the Recommendationf (For Intervention)
Aerobic fitness/exercise endurance Baydogan et al30g
12 wk; 27 h supervised by PT
10MWT (body function)
10SCT (body function)
Pretest to immediate postintervention (12 wk) NS Results indicate a larger decrease in time indicating improvement in speed/endurance favoring the P/EG for 10MWT and 10SCTh Moderate Weak+
Small to medium nonsignificant ESs for all but one outcome following a variety of physical training protocols or exercise prescription
Epps et al32
Combined therapy group (24-32 h): first 2 wk = 16 h (8-h hydrotherapy + 8-h land therapy) followed by 1 h/wk or 1 h/2 wk of hydrotherapy for 2 mo (range: 8-16 h)
Land only group: (24-32 h): 16 1-h sessions of land therapy in first 2 wk (16 h) followed by 1 h/wk or 1 h/2 wk of land therapy for 2 mo (8-16 h)
Time to exhaustion on bicycle ergometer (body function)
Submaximal HR recorded during second minute of gentle cycling following final minute of highest work level (body function)
Pretest to postintervention (2 mo post end of initial 2 wk) NS
Greater↑ in exercise time in combined land and water EG
Greater↓ in submaximal HR in land and water EG combined
−0.53 (medium) [−0.03, 1.09] NS
−0.46 (small) [−1.01, 0.10] NS
Lacks evidence for specificity of training
No harm reported by any study
Sandstedt et al35
12 wk; 18-27 h
Rating of Perceived Exertion using Borg Scale (body function) Pretest to immediate postintervention (12 wk) NS −0.03 (negligible) [−0.65, 0.58] NS
Singh-Grewal et al36
12 wk; 30 h
Relative O 2submax at 1.5 km/h (body function) Pretest to immediate postintervention (12 wk) NS −0.42 (small) [−0.90, 0.06] NS
Relative O 2submax at 3.0 km/h(body function) −0.28 (small) [−0.75, 0.20] NS
Takken et al10
6 mo; 20 h
O 2peak (body function) Pretest to end of exercise trial (6 mo) NS
O 2peak was stable in EG, ↓ in CG
0.19 (small) [−0.34, 0.73] NS
6MWD (body function) NS 0.21 (small) [−0.32, 0.75] NS
Tarakci et al37
12 wk; 16 − 36 h
6MWD (body function) Pretest to immediate postintervention (12 wk) .000g
Favored EG
0.33 (small) [−0.11, 0.77] NS
Muscle strength Baydogan et al30
g 12 wk; 27 h
Muscle force/peak torque (HHD)
(body function)
Pretest to postintervention (12 wk) Hip flexion:.000
Hip ABD: .001
Favored P/B EG
i Moderate Weak +
Statistically significant improvement in strength for some muscle groups (hip flexors, extensors, and abductors; knee flexors and extensors) with a variety of exercise training: resistance exercise balance/proprioceptive exercise)
Large ESs favoring intervention group were found for knee flexors and extensors after 12-wk underwater exercise followed by IFC to reduce pain
No harm reported by any study
Recommendation is downgraded due to small number of studies. However, given the known effects of JIA on muscle structure and function, most clinicians would include LE strength training in their treatment regimen. Study by Elnaggar and Elshafey31 suggests benefits may be enhanced by postexercise IFC to decrease postexercise soreness
Epps et al32
10 wk: 24-32 h for each group
Muscle force/peakj torque (HHD)
(body function)
Pretest to postintervention (2 mo) NS Hip abductors: 0.05 (negligible) [−0.53, 0.43] NS
Knee extensors: 0.17 (negligible) [−0.31, 0.65] NS
Sandstedt et al35
12 wk: 18-27 h
Muscle force/peak torque (HHD (body function) Pretest to postintervention (12 wk) <.03
Hip extension knee extension,
(L) hip ABD
Favored EG
Hip extensor (R): 0.39 (small) [−0.20, 0.99] NS
Hip extensor (L): 0.36 (small) [−0.24, 0.96] NS
Knee extension (R): 0.41 (small) [−0.18, 1.01] NS
Knee extension (L): 0.25 (small) [−0.35, 0.84] NS
Hip abductor (L): 0.36 (small)[−0.24, 0.95] NS
Elnaggar and Elshafey31
12 wk; 27 h
Combined underwater LE exercise and postexercise IFC
Muscle force/peak torque (isokinetic/Cybex) (body function) Pretest to post intervention (12 wk) .001
(knee extension and flexion)
Changes favored the underwater exercise and IFC EG
Quadriceps (R): 1.62 (large) [0.79, 2.47]
Quadriceps (L): 1.96 (large) [1.07, 2.86]
Hamstrings (R): 2.74 (large) [1.71, 3.77]
Hamstrings (L): 3.07(large) [1.97, 4.17]
All favored EG
Flexibility/joint range of motion Baydogan et al30
g 12 wk; 27 h
ROM-goniometer (body function) Pretest to postintervention (12 wk) NS i Moderate Weak +
Available data, with the exception of one study with a large, significant ES favoring Stott Pilates, does not provide strong evidence that ROM exercise improves joint flexibility in JIA
However, there is no evidence of adverse effects, and daily ROM and active stretching exercises are an integral part of therapy for this condition and recommended by most rheumatology professionals
Sandstedt et al35
12 wk; 18-27 h
ROM-goniometer (body function) Pretest to postintervention (12 wk) NS
No specific data were reported
No data provided
Mendonca et al33 pEPM-ROM (body function) Pretest to immediate postintervention (6 mo) <.01
Favored EG (Pilates)
−1.72 (large) [−2.38, −1.06]
Favored EG (Pilates)
Singh-Grewal et al36
12 wk; 30 h
pEPM-ROM (body function) Pretest to immediate postintervention (12 wk) NS −0.35 (small) [−0.83, 0.12] NS
Takken et al10
6 mo to 20 h
pEPM-ROM (body function) Pretest to end of exercise trial (6 mo) NS −0.15 (negligible)[−0.68, 0.39] NS
Pain Baydogan et al30g
12 wk; 27 h
Proprioceptive/balance exercise vs traditional strengthening exercise
NRS exercise (body function) Pretest to postintervention (12 wk) NS i Moderate Weak +
Data from 2/6 studies reporting outcomes provides favorable but qualified evidence that structured fitness training may reduce pain scores (VAS) in JIA. The findings across studies are not consistent
Mendonca et al33
6 mo: 40 h
100-mm VAS (body function) Pretest to immediate post intervention (6 mo) <.001
Favored EG (Pilates)
−1.19 (large)[−1.79, −0.58]
Favored EG (Pilates)
Sandstedt et al35
12 wk; 18-27 h
100-mm VAS (body function) Pretest to postintervention (12 wk) No statistical analysis reported Insufficient data to compute ES One study found a large ES favoring Pilates and one a small ES favoring underwater LE exercise followed by IFC self-reported pain in JIA; however, available data indicate there is no increase in pain with structured moderate to intense exercise on land or in water
While the majority of available data indicate most children with JIA tolerate increased exercise intensity and frequency, clinicians should continue to monitor joints and overall pain experience
Epps et al32
10 wk: 24-32 h
100-mm VAS (body function) Pretest to end of exercise trial (2 mo) NS 0.24 (small)[−0.21, 0.68] NS
Elnaggar and Elshafey31
12 wk; 9 h
10-cm (100 mm) VAS (body function) Pretest to post intervention (12 wk) .001
Change at 3-m post-BL favored EG
−0.36 (small)(−0.62, −0.09)Favored EG
Tarakci et al37
12 wk; 16-36 h
100-mm VAS (body function) Pretest to immediate postintervention (12 wk) NS −0.10 (negligible)[−0.53, 0.34] NS
Patient-reported outcome measures
Physical function Baydogan et al30g
12 wk; 27 h
Proprioceptive/balance exercise vs traditional strength training
Pretest to postintervention
(12 wk)
Significant ↓ in DI in both groups
exceeded the MICD (−0.13) for improved functional capability
EG: −0.30
CG: −0.16
i Moderate Weak +
Most studies report a small, insignificant effect from exercise training; however 2/6 reported a medium to large ES favoring structured exercise to improve physical function
Outcome measures used to assess physical function are known to have a floor effect and lack changes in the child's ability to perform complex or more strenuous physical activities
Epps et al32
10 wk: 24-32 h
CHAQ DI (activity) Pretest to postintervention (2 mo) NS 0.26 (small)[−0.20, 0.72] NS
Mendonca et al33
6 mo: 40 h
CHAQ (activity) Pretest to immediate post intervention (6 mo) <.0001
Favored EG (Pilates)
−1.57 (large)[−2.21, −0.93]
Favored EG (Pilates)
Sandstedt et al35
12 wk; 18-27 h
CHAQ (activity) Pretest to postintervention (12 wk) NS
BL—12 wk↓ in DI in EG, but not in CG, exceeded the MICD for improvement
Singh-Grewal et al36
12 wk; 30 h
CHAQ (activity) Pretest to immediate postintervention NS
Changes in DI similar in both groups
−0.02 (negligible)[−0.50, 0.45] NS
Takken et al10
6 mo to 20 h
CHAQ (activity) Pretest to end of exercise trial (6 mo) NS −0.20 (small)[−0.74, 0.33] NS
JAFAS (activity) NS −0.08 (negligible)[−0.81, 0.46] NS
Tarakci et al37
12 wk; 16-36 h
CHAQ (activity) Pretest to immediate postintervention (12 wk) .000
Favored EG
−0.65 (medium)[−1.10, −0.20]
Favored EG
Quality of life Epps et al32
10 wk: 24-32 h
CHQ (physical)
CHQ (psychological
Pretest to postintervention (2 mo) NS
−0.09 (negligible)[−0.59, 0.42] NS
0.17 (negligible)[−0.35, 0.69] NS
Moderate Strong+
Although most studies found small and insignificant effects of ET on QOL, no study reported adverse effects
One study by Mendonca et al33 reported a large, clinically significant improvement in QOL following a 6-mo program of Pilates compared with a conventional exercise program
Singh-Grewal et al36 Overall QOL (10-cm VAS)
Health-related QOL
Pretest to immediate postintervention (12 wk) NS
0.17 (negligible)[−0.30, 0.64] NS
−0.05 (negligible)[−0.53, 0.42] NS
Mendonca et al33
6 mo: 40 h
PedsQL 4.0
Patient self-report
Pretest to immediate post intervention (6 mo) <.001
Favored EG (Pilates)
2.75 (large)[1.96, 3.54]
Tarakci et al37
12 wk; 16-36 h
PedsQL 3.0 arthritis modules
Patient self-report or parent report as proxy for child
Pretest to immediate postintervention at (12 wk) .000
Favored EG
0.47 (medium)[−0.07, 1.01] NS Tarakci et al37 found a statistically significant benefit to QOL after a 12-wk program that combined HEP 3×/wk supervised by parents with a 1×/wk hospital-based session supervised by a PT
Based on results of these studies, the potential benefits of structured ET and the absence of adverse effects on QOL, we believe most clinicians would implement an ET program in children with well- controlled JIA
Sandstedt et al35
12 wk; 18-27 h
“Role Physical”
Pretest to postintervention (12 wk) <.017
Favored CG
Takken et al10
6 mo to 20 h
CHQ-“Role Physical”
Pretest to end of exercise trial (6 mo) NS 0.47 (small)[−0.47, 1.01] NS
0.48 (small)[−0.06, 1.02] NS
Abbreviations: A, Activity; BF, body function; BG, between group; CG, control group; CHAQ, Child Health Assessment Questionnaire; DI, CHAQ Disability Index; CHQ-PhS, Child Health Questionnaire Physical Health; CHQ-PsS, Child Health Questionnaire Psychosocial Health; EG, experimental group; ES, effect size; ET, exercise training; GRADE, Grading of Recommendations, Assessment, Development and Evaluation; HHD, hand-held dynamometry; IFC, interferential current; JIA, juvenile idiopathic arthritis; LE, lower extremity; NRS, numerical rating scale; NS, nonsignificant between-group difference; P, Participation; PB, proprioceptive/balance; PedsQL, Pediatric Quality of Life Inventory; pEPM-ROM, Pediatric Escola Paulista de Medicina Range of Motion Scale; PF, physical fitness; PT, physical therapy; QOL, quality of life; 10MWT, 6MWD, 6-minute walk distance; 10-m walk test; 10SCT, 10-stair climbing test;
aCoded with the International Classification of Functioning, Disability and Health (ICF).11
bSignificant (P ≤ .05) between-group difference.
cCohen's d (0.20 = small; 0.50 = medium; 0.85 = large).
dη2 (0.02 = small, 0.13 = medium, 0.26 = large).
eGRADE describes 4 levels of evidence ratings (A = high, B = moderate, C = low, and D = very low).28
fThe GRADE of a recommendation29 is rated on a continuum defined as the extent to which one can be confident that the desirable effects of an intervention outweigh its undesirable effects—from strong (−)/weak (−) = against the intervention (not recommended) to weak (+)/strong (+) = for the intervention; strong + indicates most well-informed people would implement the intervention; weak + means “probably do it.”11,28,29
gBaydogan applied a Bonferroni's correction for multiple (20) univariate comparisons (P < .05/20 = P < .0025).
hStudy provided insufficient data, or data were reported as median (range); unable to calculate ES;
iUnable to compute ES when data were reported as median value or insufficient data provided.
jKettler ergometer.

Muscle Strength.Table 2 and Figure 3 include outcomes from 4 studies measuring muscle strength using an HHD or isokinetic testing.30–32,35 Three had significant improvement favoring the EG. Sandstedt et al35 reported increased strength in hip and knee extension and hip abduction (P < .03) after 12 weeks of repetitive rope jumping and lower extremity strength training.35 Baydogan et al,30 using P < .0025 to adjust for 20 separate comparisons, demonstrated significantly increased strength in hip flexion (P = .000) and abduction (P = .001), favoring the balance/proprioceptive group.30 Elnaggar and Elshafey31 found statistically (P = .001) and clinically significant improvement for knee and hip flexion and extension strength favoring underwater exercise followed by interferential current (IFC). Figure 3 depicts large ESs for improved quadriceps strength: right (n = 15; SMD 1.63, 95% CI 0.79 to 2.47) and left (n = 15; SMD 1.96, 95% CI 1.07 to 2.86)—and hamstrings strength: right: (n = 15; SMD 2.74, 95% CI 1.71 to 3.77) and left (n = 15; SMD 3.07, 95% CI 1.97 to 4.17) after 12 weeks (36 sessions) of underwater resistance exercise.

Fig. 3.
Fig. 3.:
Measures of muscle strength.

Joint ROM/Flexibility. Baydogan et al30 and Sandstedt et al,35 using standard goniometry, reported no significant improvement following training. Data were insufficient to calculate ES. Table 2 and Figure 4 include results for 3 studies using the pEPM-ROM to assess joint ROM.10,33,36 Only Mendonca et al33 reported statistically (P < 0.01) and clinically significant (n = 50; SMD −1.72, 95% CI −2.38 to −1.06) improvement in joint flexibility, favoring the Pilates group.33

Fig. 4.
Fig. 4.:
Measures of pain: visual analog scale.

Bone Density. Only Sandstedt et al34 examined the impact of weight-bearing exercise on bone health. Bone mineral density (BMD), bone mineral content (BMC), and z-score were measured before and after a 12-week, 36-session program of jumping rope (100 2-footed jumps) and general strength training (Table 2). Baseline values in JIA were similar to normative reference values. Postintervention total body BMD, but not z-score, increased significantly (P = .01) in the EG but not in the CG (P = .06). BMC did not change.

Pain Intensity.Table 2 and Figure 5 include outcomes for 4 studies reporting pain intensity using a VAS.31–33,37 Only Mendonca et al33 (n = 25; SMD −1.19, 95% CI −1.79 to −0.58) and Elnaggar and Elshafey31 (n = 15; SMD −2.18, 95% CI −3.11 to −1.25) found statistically (P < .001) and clinically significant decreases in pain favoring the EG. Two studies reported no increase in pain scores with exercise but did not provide data.30,35

Fig. 5.
Fig. 5.:
Measures of flexibility.

Physical Function (Capability). Seven articles measured physical capability over the previous week using the CHAQ Disability Index (DI) and measurement (0-3) scale: higher scores indicate greater disability. Table 2 and Figure 6 include results for 5 studies reporting DI scores. Only Mendonca et al33 and Tarakci et al37 found statistically (P < .000) and clinically significant BG differences favoring the EG: Mendonca et al33 (n = 25; SMD −1.57, 95% CI −2.21 to −0.93); Tarakci et al37 (N = 43; SMD −0.65, 95% CI −1.10 to −0.20).

Fig. 6.
Fig. 6.:
Measures of functional capability.

Quality of Life. Five studies, using 3 different measures, reported QoL (Table 2 and Figure 7). Three found no significant difference following ET.32,35,36 One found significant improvement (P < .02) favoring the CG.35 Tarakci et al37 (P = .000; N = 43; SMD 0.90, 95% CI 0.44 to 1.35) and Mendonca et al33 (P < .001; N = 25; SMD 2.75, 95% CI 0.44 to 1.35), using the Pedi, demonstrated statistically and clinically significant improvement favoring the EG. Mendonca et al33 reported an increase of 38.8 ± 3.9 points in total PedsQL score in the Pilates group compared with a decrease of 3.8 ± 3.9 points in the CG.

Fig. 7.
Fig. 7.:
Measures of quality of life.


There were 34 dropouts across studies, representing 7.4% of the total sample (457). Four studies, Takken et al,10 Elnaggar and Elshafey,31 Baydogan et al,30 and Mendonca et al,33 reported no dropouts. Singh-Grewal et al36 had10 dropouts after randomization: 6 from the EG (4 before; 2 after BL tests) and 4 from the CG (1 before; 3 after BL tests). Epps et al32 reported 6 participants did not complete the 2-month assessment. Sandstedt et al had 6 dropouts after the baseline assessment (EG = 5; CG = 1). Tarakci et al37 reported 4 dropouts from the EG (1 after randomization; 3 before final testing). Eight dropouts in the CG were lost to follow-up testing.


This SR updated the evidence for structured ET to improve HRPF, bone health, physical function, and QoL in JIA. Nine articles (8 RCTS) met selection criteria: all studies included in reviews by Takken et al19 and Cavallo et al20 plus 4 additional articles.30–32,34 Based on the opinion of Maher48 that a PEDro score of 5/10 was acceptable for exercise trials when neither participants nor trainers could be blinded to group allocation, the overall quality of evidence (6.56/10) was high with the exception of 1 trial.30 Similar to previous reports,19,20 this review supports a trend favoring ET with no adverse effects. Trials by Elnaggar31 and Mendonca et al33 supported clinically significant decreases in pain despite intense training, suggesting structured ET is safe and potentially useful in managing chronic pain in JIA. However, beyond these benefits, there were few statistically (22/65) or clinically (10/65) significant outcomes favoring ET. Only 3 of the 8 protocols demonstrated clinically significant improvements in 2 or more outcomes31,33,37 regardless of whether the CG was assigned to a less intense exercise regimen, wait-list, or assessment only.

Cavallo et al20 developed the EBCPG for structured ET in JIA based on 5 RCTs. They defined minimal clinically important difference (MCID) in an outcome as 30% or more improvement as a result of an intervention. Based on this criterion, the clinical practice guidelines provide 2 strong recommendations for ET in JIA—Stott Pilates33 and an HEP37—to improve QoL and physical function and Stott Pilates alone to improve ROM and pain. While the results of these trials are undisputed, there are several factors to consider. The 30% improvement standard used by Cavallo et al is based on the ACR Pediatric 30, a composite measure of 6 outcomes used to judge effectiveness in pharmaceutical trials that require a high standard of safety and efficacy.49 Three of the measures (swollen joint count, limited joint count, and ESR) target disease activity, previously shown to have no significant association with HRPF.11,12 To date, neither the core set nor the 30% improvement criterion has been validated for exercise trials. In this review we determined clinical effectiveness of interventions using Cohen's criteria26 for BG ES, and published MCID values for specific outcomes.

It is also useful to ask whether the clinical benefits of these trials were due to the specific exercise mode or to overall exercise load. To explore this, we compared studies based on exercise frequency, intensity, session duration, and total time as well as exercise mode, setting, amount of expert supervision, and individualized modifications (see Supplemental Digital Content 4, available at: We gave particular attention to 3 trials that had statistically and clinically significant improvement in 2 or more outcomes.31,33,37 Results suggest the minimal effective exercise amount included 45 to 50 minutes/session, at least two times/week for at least 12 weeks. Overall program length in successful trials ranged from 1231,37 to 2433 weeks and total ET time varied from 1837 to 4033 hours. Session length was consistent over the program in the Stott Pilates27 and underwater resistance-training regimens.31,33 In contrast, the HEP of Tarakci et al37 prescribed a gradual increase in exercise repetitions (from 3 to 15) and session duration (from 20 to 45 minutes) in order to minimize adverse effects and ensure compliance early in the program.

Expert supervision, personalized programming, and modification of exercises also appear to positively influence outcomes. Trials by Mendonca et al33 and Elnaggar and Elshafey31 were exclusively center-based and supervised by ET experts. Exercising in small groups of individuals with JIA may provide a supportive atmosphere and incentive to participate fully in the activities. Participants in the HEP37 attended an additional hospital-based session each week supervised by a PT who modified the child's program as needed. The convenience of a personalized HEP supplemented by a supervised session each week, plus the requirement to keep an exercise diary, may have contributed to the success of this protocol. Current evidence supporting the benefits of ET to improve HRPF (aerobic fitness/performance, strength, and joint flexibility), physical function, and pain is rated as moderate, and the strength of recommendation is rated as weak+ indicating “most well-informed people would implement some form of ET for individuals with JIA.”29 (Table 2)

Aerobic Fitness

Of 5 studies reporting outcomes on aerobic fitness/performance,10,32,35–37 only Tarakci et al37 found statistically significant improvement (increased 6MWD); however, the change was not clinically significant (Figure 2). Although the exercise protocol included “walking activities,” it is unclear whether walking speed or distance was encouraged or recorded. Lelieveld et al50 found 6MWD to be a poor predictor of O2peak in JIA, however considered it a useful measure of walking endurance in JIA where fatigue is commonly reported. A 2010 pilot study, [email protected],51 a combined online and in-person education program, provides a potential alternative to lengthy and intensive in-person ET.8,52–54 The program targeted participants' knowledge of JIA and healthy PA habits.55 Each participant identified a personal “smart goal” for PA and completed online assignments related to healthy behaviors. Although only 4% of participants met Centers for Disease Control and Prevention (CDC) guidelines for PA54 at baseline, both amount and intensity of PA increased significantly based on a 7-day retrospective PA diary following the program.56 Maximum treadmill walking time, which correlates highly with O2max,57 also increased significantly. Most participants and parents rated the program as valuable.58 Components of this program relevant to PT are personalized and measureable goals for daily PA, frequent online interactive sessions with program leaders, education about healthy PA, required assignments, and personalized support to help participants achieve their goals.

Muscle Strength

Muscle atrophy and weakness, common in JIA,59,60 negatively impact joint control during biomechanical loading and potentially increase the risk of injury during physical play and sports.13,61 Three studies (142 participants) showed statistically significant improvements in strength in one or more muscle groups after ET: hip flexion, extension and abduction, and knee flexion and extension.30,31,35 However, only Elnaggar and Elshafey31 reported clinically significant improvement. Their protocol included 18 hours of underwater resistive knee flexion and extension stressing movement speed followed by application of IFC. Authors suggest IFC may improve recovery from intense exercise by increasing blood flow to joints, improving clearance of pain-inducing chemicals and releasing endogenous opiates. Although Fuentes et al62 found significant short-term pain reduction from IFC in adults with musculoskeletal conditions, no other pediatric studies were found. Given the inclusion of only children with polyJIA in this study,31 additional research is warranted.

Bone Density

Weight-bearing exercise is recommended in JIA due to evidence of low BMD and bone strength.63 A 2007 SR by Gannotti et al60 examining the impact of exercise on bone strength in healthy prepubescent children found a large ES supporting high-impact repetitive jumping.64 Sandstedt et al34 found children with well-controlled JIA were able to complete an intensive program of strength training and high repetitions of jumping rope without adverse effects during the training period. However, they showed no significant increase in BMD or BMC after the 12- week program, possibly because this sample showed no deficits in bone health prior to training. This may reflect earlier diagnosis, improved medical control of inflammation, and an increased emphasis on weight-bearing exercise.

Range of Motion

Only 1 of 5 trials in this review found significant improvement in joint mobility after ET. Of 2 studies30,35 using goniometry, only Baydogan et al30 reported outcome data. Although BG differences in knee ROM following ET were insignificant, the proprioceptive/balance group showed greater improvement. Both groups experienced the same protocol with the exception of the target intervention, suggesting proprioceptive exercise might provide a slight advantage over a traditional strengthening regimen. Reports of impaired proprioceptive function in JIA also support balance training.65,66

Of 3 trials10,33,36 that used the standardized pEPM-ROM,39 only Mendonca et al33 found statistically and clinically significant improvement favoring the EG. The success of this program is likely due to multiple factors: high volume (40 hours) and long training period, novelty of the Pilates exercises and equipment, personalized instruction and emphasis on movement form, and postural awareness and control of breathing during movement.

Physical Function

In this review, 7 studies reported outcomes of ET on physical function using the original CHAQ30 Disability Index (DI), 1 component of the ACR Core Set of outcomes for clinical trials.49 The DI includes 30 items grouped into 8 categories that assess the child's capability—“over the past week, I was able to”—perform daily activities during the past week. The mean DI at baseline was similar in the EG (0.65) and CG (0.64). However, based on the wide range of scores (0.030 to 1.2132), the level of disability varied between “none” (score of “0”) and “moderate” (score of 1.75).67 Only Mendonca et al33 and Tarakci et al37 demonstrated statistically and clinically significant improvement in the DI favoring the EG. The change in both trials met or exceeded the MCID for improvement (median change score ranging from 0.13 to 0.57) based on initial level of disability.67 In the Baydogan trial, the proprioceptive/balance and strengthening groups met or exceeded the MCID for improvement in the DI. Both were center-based, led by the same PT and identical in all aspects other than the specific type of training. This may suggest any well-constructed, professionally supervised ET regimen designed for children with JIA that meets minimal criteria for frequency, intensity, and duration may improve functional capability. However, clinicians should also be aware of the “floor effect” of the CHAQ30 when used to assess change in children whose disease is under good control. A revised version of the CHAQ (CHAQ38VAS)41 that includes 8 additional items measuring more challenging physical activities may be more responsive to changes from ET. Clinicians should also consider assessing physical performance (activities the child actually performed over a specified period) and capacity (performance under standardized conditions) to gain a better understanding of the impact of ET.

Quality of Life

Current evidence from 5 studies, using 4 different instruments, to support the benefits of ET on QoL in JIA is rated as moderate and the strength of recommendation as strong+, indicating clinicians would definitely include some form of ET for individuals with JIA.29 Only trials by Tarakci et al37 and Mendonca et al,33 using the PedsQL, demonstrated statistically and clinically significant improvement. Both studies also showed success in other outcomes, most likely due to the high exercise load, supervision, and personalization of training. The PedsQL was originally developed for JIA. A generic HRQOL VAS may not capture the complex nature of QoL outcomes in children with JIA, where periods of quiescence and flares are common. Although the CHQ targets multiple aspects of QoL, it is not specific to JIA. Future research would benefit from a consensus on a single QoL measure as part of a core set of outcomes for ET in this population.

Limitations of the Systematic Review

The authors recognize several limitations to this SR including the small number of studies meeting selection criteria; low number of participants in each study; and failure to include all disease types and report outcomes separately for each JIA diagnosis. We also recognize the potential for bias in this review both in study selection and analysis; however, we believe the process to locate studies meeting selection criteria and the requirement for consensus among 6 reviewers regarding study quality support the objectivity of this review.

Implications for Clinical Practice

Current research indicates only a small percentage of individuals with JIA meet CDC guidelines for daily MVPA.8,52–54 Evidence of low physical fitness12 as well as impaired balance and neuromuscular performance, even in those with minimal disease, shows the potential for increased risk of injury during competitive or recreational sports.16,65,66 Studies by Taxter et al61 and Myer et al17 suggest a 3-dimensional motion analysis of a child's movements during sports is useful to identify impaired movement patterns and guide intervention in JIA. The benefits of movement analysis and individualized neuromuscular training include improved joint protection and impact loading during sports.17 Current research suggests physical therapists working with this population should expand their role to include assessment of health and performance-based fitness as well as advising children and parents on sports participation and sports-specific training to ensure safe and successful play.

Implications for Research

Future research would benefit from improvement in 3 areas: (1) development and validation of a core set of outcome measures focused on effects of ET in JIA, including both standardized fitness assessments and patient/caregiver-reported measures sensitive to change in physical function and quality of life; (2) exploration of interactive Internet-based programs combined with periodic in-person sessions to encourage participants to adopt healthy PA habits; and (3) examination of structured ET based on JIA type, disease activity, and duration.


1. Ravelli A, Martin A. Juvenile idiopathic arthritis. Lancet. 2007;369(9563):767–778.
2. Manners PJ, Bowers C. Worldwide prevalence of juvenile arthritis—why does it vary so much? J Rheumatol. 2002;29(7):1520–1530.
3. Helmick GG, Felson D, Lawrence RC, et al. Estimates of the prevalence of arthritis and other rheumatic conditions in the United States. Arthritis Rheum. 2008;58(1):15–25.
4. Sacks JJ, Helmick CG, Luo YH, Ilowite NT, Bowyer S. Prevalence of and annual ambulatory health care visits for pediatric arthritis and other rheumatologic conditions in the United States 2001-2004. Arthritis Rheum. 2007;57(8):1439–1445.
5. Prakken B, Albani S, Martini A. Juvenile idiopathic arthritis. Lancet. 2011;377(9783):2138–2149.
6. Luca NJ, Feldman BM. Health outcomes of pediatric rheumatic diseases. Best Pract Res Clin Rheumatol. 2014;28(2):331–350.
7. Klepper S. Exercise in pediatric rheumatic diseases. Curr Opin Rheumatol. 2008;20(5):619–624.
8. Cavallo S, April AK, Grandpierre V, et al. Leisure in children and adolescents with juvenile idiopathic arthritis: a systematic review. PLoS One. 2014;9(10).
9. Klepper S. Effects of an eight-week physical conditioning program on disease signs and symptoms in children with juvenile arthritis. Arthritis Care Res. 1998;12(1):52–60.
10. Takken T, Van der Net JJ, Kuis W, Helders PJ. Aquatic fitness training for children with juvenile idiopathic arthritis. Rheumatology (Oxford). 2003;42(11):1408–1414.
11. van Brussel M, Lelieveld OT, van der Net JJ, Engelbert RH, Helders PJ, Takken T. Aerobic and anaerobic capacity in children with juvenile idiopathic arthritis. Arthritis Care Res. 2007;57(6):891–897.
12. Lelieveld OT, van Brussel M, Takken T, van Weert E, van Leeuwen MA, Armbrust W. Aerobic and anaerobic capacity in adolescents with juvenile idiopathic arthritis. Arthritis Care Res. 2007;57(6):898–904.
13. van der Net JJ, van der Torre P, Engelbert RH, et al. Motor performance and functional ability in preschool- and early school-aged children with juvenile idiopathic arthritis. Pediatr Rheumtol Online. 2008;6:2.
14. Hulsegge G, Henschker N, McKay D, et al. Fundamental movement skills, physical fitness and physical activity among Australian children with juvenile idiopathic arthritis. J Paediatr Child Health. 2015;51(4):425–432.
15. Khan P. Juvenile idiopathic arthritis—an update on pharmacotherapy. Bul NYU Hosp Jt Dis. 2011;69(3):264–276.
16. Ford K, Myer GD, Melson PG, Darnell SC, Brunner HI, Hewett TE. Land-jump performance in patients with juvenile idiopathic arthritis (JIA): a comparison to matched controls. Int J Rheumatol. 2009;2009:478526.
17. Myer GD, Brunner H, Melson PG, Paterno MV, Ford KR, Hewett TE. Specialized neuromuscular training to improve neuromuscular function and biomechanics in a patient with quiescent juvenile rheumatoid arthritis. Phys Ther. 2005;85(8):791–802.
18. Hurkmans E, van der Giesen FJ, Vliet Vlieland TP, Schoones J, Van den Ende EC. Dynamic exercise programs (aerobic capacity and/or muscle strength training) in patients with rheumatoid arthritis. Cochrane Database Sys Rev. 2009;4:CD006853.
19. Takken T, van Brussel M, Engelbert RH, van der Net JJ, Kuis W, Helders PJ. Exercise therapy in juvenile idiopathic arthritis: a Cochrane Review. Eur J Phys Rehabil Med. 2008;44(3):287–297.
20. Cavallo S, Brosseau L, Toupin-April K, et al. Ottawa Panel Evidence-Based Guidelines for structured physical activity in the management of juvenile idiopathic arthritis. Arch Phys Med Rehabil. 2017;98(5):1018–1041.
21. Kuntze G, Nesbitt C, Whittaker JL, et al. Exercise therapy in juvenile idiopathic arthritis: a systematic review and meta-analysis. Arch Phys Med Rehabil. 2018;99:178–193.
22. Catania H, Fortini V, Cimaz R. Physical exercise and physical activity for children and adolescents with juvenile idiopathic arthritis: a literature review. Pediatr Phys Ther. 2017;29:256–260.
23. Higgins J, Green S. The Cochrane Handbook for Systematic Reviews of Interventions. London, England: The Cochrane Collaboration; 2011.
24. Moher D, Liberati A, Tetzlaff J, et al. Preferred reporting for systematic reviews and meta-analysis: the PRISMA statement. J Clin Epidemiol. 2009;62(10):1006–1012.
25. Law M, Stewart D, Pollock N, et al. Guidelines for Critical Review—Quantitative Studies. Hamilton, Canada: McMaster University; 1998.
26. Cohen J. Statistical Power Analysis for the Behavioral Sciences. 2nd ed. Hillsdale, NJ: Erlbaum; 1988.
27. Maher CG, Sherrington CH, Herbert RD, Moseley AM, Elkins M. Reliability of the PEDro scale for rating quality of randomized controlled trials. Phys Ther. 2003;83(8):713–721.
28. Balshem H, Helfand M, Schunemann M, et al. GRADE guidelines: 3 rating the quality of evidence. J Clin Epidemiol. 2011;64(4):401–406.
29. Andrews J, Guyatt G, Oxman AD, et al. GRADE Guidelines: 14. Going from evidence to recommendations: the significance and presentation of recommendations. J Clin Epidemiol. 2013;66(7):719–725.
30. Baydogan SN, Tarakci E, Kasapcopur O. Effect of strengthening versus balance-proprioceptive exercises on lower extremity function in patients with juvenile idiopathic arthritis. Am J Phys Med Rehabil. 2015;94(6):417–424.
31. Elnaggar RK, Elshafey M. Effects of combined resistive underwater exercise and interferential current therapy in patients with juvenile idiopathic arthritis: a randomized controlled trial. Am J Phys Med Rehabil. 2016;95(2):96–102.
32. Epps H, Ginnelly L, Utley M, et al. Is hydrotherapy cost-effective? A randomised controlled trial of combined hydrotherapy programmes compared with physiotherapy land techniques in children with juvenile idiopathic arthritis. Health Technol. 2005;9:(39).
33. Mendonca TM, Terreri MT, Silva CH, et al. Effects if Pilates exercises on health-related quality of life in individuals with juvenile idiopathic arthritis. Arch Phys Med Rehabil. 2013;94:2093–2102.
34. Sandstedt E, Fasth A, Fors H, Beckung E. Bone health in children and adolescents with juvenile idiopathic arthritis and the influence of short-term physical exercise. Pediatr Phys Ther. 2012;24(2):155–162.
35. Sandstedt E, Fasth A, Eek MN, Beckung E. Muscle strength, physical fitness and well-being in children and adolescents with juvenile idiopathic arthritis and the effect of an exercise programme—a randomized controlled trial. Pediatr Rheumatol. 2013;11:(1):7.
36. Singh-Grewal D, Schneider WJ, Wright V, et al. The effects of vigorous exercise training on physical function in children with arthritis: a randomized, controlled, single-blinded trail. Arthritis Care Res. 2007;57(7):1202–1210.
37. Tarakci E, Yeldon I, Baydogan SN, Olgar S, Kasapcopur O. Efficacy of a land-based home exercise programme for patients with juvenile idiopathic arthritis: A randomized controlled, single blind study. J Rehabil Med. 2012;44(11):962–967.
38. OCEBM Levels of Evidence Working Group. The Oxford Levels of Evidence 2. Oxford, England: Oxford Centre for Evidence-Based Medicine; 2011.
39. Len C, Ferraz MB, Goldenberg J, et al. Pediatric Escola Paulista de Medicina Rang of Motion Scale: a reduced joint count scale for general use in juvenile idiopathic arthritis. J Rheumatol. 1999;26(4):909–913.
40. Klepper SE. Measures of pediatric function. Arthritis Care Res. 2011;63(S11):S371–S382.
41. Lam C, Young N, Marwaha J, McLimont M, Feldman BM. Revised versions of the Childhood Health Assessment Questionnaire (CHAQ) are more sensitive and suffer less from a ceiling effect. Arthritis Care Res. 2004;51(6):881–889.
42. Stevens S, Singh-Grewal D, Bar-Or O, et al. Reliability of exercise testing and functional activity questionnaires in children with juvenile arthritis. Arthritis Care Res. 2007;57(8):1446–1452.
43. Burnstein BD, Steele RJ, Shrier I. Reliability of fitness tests using methods and time periods common in sport and occupational management. J Athl Train. 2011;46(5):505–513.
44. Klepper S, Muir N. Reference values on the 6-minute walk test for children living in the United States. Pediatr Phy Ther. 2011;23:32–40.
45. Wessel J, Kaup C, Fan C, et al. Isometric strength measurements in children with arthritis: reliability and relation to function. Arthritis Care Res. 1999;12(4):238–246.
46. Fagher K, Fritzon A, Drake AM. Test-retest reliability of isokinetic knee strength measurements in children aged 8 to 10 years. Sports Health. 2016;8(3):255–259.
47. Hoffman TC, Glasziou P, Boutron I, et al. Better reporting of interventions: template for intervention description and replication (TIDieR) checklist and guide. BMJ. 2014;348:g1687.
48. Maher CG. A systematic review of workplace interventions to prevent low back pain. Aust J Physiother. 2000;46:259–269.
49. Giannini EH, Ruperto N, Ravelli A, Lovell DJ, Felson DT, Martini A. Preliminary definition of improvement in juvenile arthritis. Arthritis Rheum. 1997;40(7):1202–1209.
50. Lelieveld OT, Takken T, van der Net J, van Weert E. Validity of the 6-minute walking test in juvenile idiopathic arthritis. Arthr Rheum. 2005;53(2):304–307.
51. Lelieveld OT, Armbrust W, Geertzen JHB, et al. Promoting physical activity in children with juvenile idiopathic arthritis through an internet-based program: results of a pilot randomized controlled trial. Arthritis Care Res. 2010;62(5):697–703.
52. Lelieveld OT, Armbrust W, van Leeuwen MA, et al. Physical activity in adolescents with juvenile idiopathic arthritis. Arthritis Rheum. 2008;59(10):1379–1384.
53. Maggio AB, Hofer M, Martin XE, Marchand LM, Beghetti M, Farpour-Lambert NJ. Reduced physical activity level and cardiorespiratory fitness in children with chronic disease. Eur J Pediatr. 2010;169(10):1187–1193.
54. Bos J, Lelieveld O, Armbrust W, Sauer PJ, Geertzen JH, Dijkstra PU. Physical activity in children with juvenile idiopathic arthritis compared to controls. Pediatr Rheumatol. 2016;14(1):42.
55. Pender NJ. Development and testing of the HPM. Cardiovasc Surg. 1988;24(6):41–43.
56. Bouchard C, Tremblay A, Leblanc C, Lortie G, Savard R, Thériault G. A method to assess energy expenditure in children and adults. Am J Clin Nutr. 1983;37(3):461–467.
57. Cumming GR, Everatt D, Hastman L. Bruce treadmill test in children: normal values in a clinic population. Am J Cardiol. 1978;41:69–75.
58. Armbrust W, Bos JJ, Cappon J, et al. Design and acceptance of [email protected], a combined internet-based and in person instruction model, an interactive, educational, and cognitive-behavioral program for children with juvenile idiopathic arthritis. Pediatr Rheumtol Online. 2015;13:31.
59. Burnham JM, Shults J, Sembhi H, et al. The dysfunctional muscle-bone unit in juvenile idiopathic arthritis. J Musculoskelet Neuronal Interact. 2006;6(4):351–352.
60. Gannotti ME, Nahorniak M, Gorton GE, et al. Can exercise influence low bone mineral density in children with juvenile rheumatoid arthritis. Pediatr Phys Ther. 2007;19:128–139.
61. Taxter A, Foss KB, Melson PM, et al. Juvenile idiopathic arthritis and athletic participation: are we adequately preparing for sports integration? Phys Sportsmed. 2012;40(3):49–54.
62. Fuentes JP, Armijo Olivo S, Magee DJ, et al. Effectiveness of interferential current therapy in the management of musculoskeletal pain: a systematic review and meta-analysis. Phys Ter. 90(9):1219–1238.
63. Von Scheven E. Pediatric bone density and fracture. Bone health in pediatric rheumatic disease. Curr Osteoporos Rep. 2007;5(3):128–134.
64. Johannsen N, Binkley T, Englert V, et al. Bone response to jumping is site-specific in children: a randomized trial. Bone. 2003;33:533–539.
65. Patti A, Maggio MC, Corsell G, Messina G, Iovane A, Palma A. Evaluation of fitness and the balance levels of children with a diagnosis of juvenile idiopathic arthritis: a pilot study. Int J Environ Res Public Health. 2017;14(7):806.
66. Houghton KM, Gazman J. Evaluation of static and dynamic balance in children with lower limb involvement due to juvenile idiopathic arthritis. Pediatri Phys Ther. 2013;25(2):150–157.
67. Dempster H, Poreoa M, Young N, Feldman BM. The clinical meaning of functional outcome scores in children with juvenile arthritis. Arthritis Care Res. 2001;44(8):1768–1774.

exercise training; juvenile idiopathic arthritis

© 2019 Academy of Pediatric Physical Therapy of the American Physical Therapy Association