You are on page 1of 14

British Medical Bulletin, 2016, 119:49–62

doi: 10.1093/bmb/ldw024
Advance Access Publication Date: 30 June 2016

Invited Review

Downloaded from by guest on 27 October 2018

The effectiveness of home hand exercise
programmes in rheumatoid arthritis: a systematic
Alison Hammond* and Yeliz Prior
Centre for Health Sciences Research, University of Salford, Salford, UK
*Correspondence address. Centre for Health Sciences Research, L701 Allerton (OT), University of Salford, Frederick Road,
Salford M6 6PU, UK.
Accepted 3 May 2016

Introduction: Rheumatoid arthritis (RA) commonly reduces hand function.
We systematically reviewed trials to investigate effects of home hand exer-
cise programmes on hand symptoms and function in RA.
Sources of Data: We searched: Medline (1946–), AMED, CINAHL,
Physiotherapy Evidence Database, OT Seeker, the Cochrane Library, ISI
Web of Science from inception to January 2016.
Areas of Agreement: Nineteen trials were evaluated. Only three were rando-
mized controlled trials with a low risk of bias (n = 665). Significant short-term
improvements occurred in hand function, pain and grip strength, with long-
term improvements in hand and upper limb function and pinch strength.
Areas of Controversy: Heterogeneity of outcome measures meant meta-
analysis was not possible.
Growing Points: Evaluation of low and moderate risk of bias trials indicated
high-intensity home hand exercise programmes led to better short-term out-
comes than low-intensity programmes. Such programmes are cost-effective.
Areas Timely for Developing Research: Further research is required to
evaluate methods of helping people with RA maintain long-term home
hand exercise.
Key words: rheumatoid arthritis, hand, upper limb, exercise, rehabilitation

© The Author 2016. Published by Oxford University Press. All rights reserved. For permissions, please e-mail:
50 A. Hammond and Y. Prior, 2016, Vol. 119

Introduction 2 months, with home exercises between twice-weekly

therapy.12 Most people with RA do not receive such
Most people with rheumatoid arthritis (RA)
high levels of supervised hand exercise therapy in the
develop hand symptoms early in their disease. Pain,
United Kingdom, Western Europe and North
stiffness and joint swelling reduce range of move-
America, due to service constraints. Therefore, this
ment (RoM), muscle strength and hand function.1,2
review specifically investigates home hand exercise
Rheumatoid cachexia (loss of muscle mass from
programmes as commonly taught by therapists, i.e.
inflammatory changes) further contributes to mus-
over one or several sessions. The aims are to identify:

Downloaded from by guest on 27 October 2018

cle wasting.3 Within six months of diagnosis,
women’s grip strength reduced to, on average, 12 kg (i) What are the short- and long-term effects of home
force, compared to the norm for a similar age range hand exercise programmes (plus usual care) on
of 28 kg force (i.e. only 40% of normal), and hand function and hand symptoms in adults with
remained lower.1,4 Activity limitations and participa- RA compared to those receiving usual care only?
tion restrictions are common and hand problems (ii) Are home hand exercise programmes cost-
contribute to work disability. Within 10 years, 59% effective?
developed hand deformities (ulnar deviation, button (iii) Which type of home hand exercise regimen is
hole and/or swan-neck finger deformities),5 further most effective?
reducing grip strength.6 (iv) What strategies can facilitate people with RA
Although medication improves hand symptoms, adhering to performing home hand exercises?
and tighter disease control may limit deformities in (v) Are there any safety concerns arising from per-
future, muscle function is not regained when dis- forming home hand exercises?
ease activity is controlled. Disease-modifying drugs
do not reverse muscle wasting and even those with
well-controlled RA have less muscle mass than
healthy age-matched controls.3 Hand exercise pro- Search strategy
grammes are therefore a standard component of We searched electronic databases from inception
RA management, to improve range of motion, until January 2016: MEDLINE; CINAHL
muscle strength, sensorimotor control, hand func- (Cumulative Index to Nursing and Allied Health
tion, activities, and participation.7 Once an exercise Literature); AMED (Allied and Complementary
programme stops, gains can be progressively lost. Medicine); PEDro (Physiotherapy Evidence data-
Accordingly, many with RA need to continue hand base), OT Seeker; Cochrane Library; and ISI Web of
exercises at home as part of self-management. To Science. The search terms included were rheumat$ or
this end, occupational therapists and physiothera- reumat$; arthrit$ or diseas$ or condition$; hand$,
pists often teach people with RA a home hand exer- wrist$, finger$, thumb$, joint$, upper limb; occupa-
cise programme. However, clinical practice varies in tional therapy; rehabilitation; therapy; Physical
the number of sessions to teach this and the number, Therapy Modalities, Exercise Therapy, exercis$;
type and intensity of RoM and resistance exercises Physical therap$; isometric, resistance or strength$,
included. muscle$, stretch$, manipulat$, range of mo$; rando-
A systematic review of eight studies (published mized controlled trial; controlled clinical trial; cost-
between 2000 and 2014) concluded hand exercise effectiveness or cost effectiveness or cost analysis. We
improved grip strength but the effects on hand also hand searched reference lists of reviews.8,13–15
RoM are unclear.8 However, only four were rando-
mized controlled trials (RCTs).9–12 Three of these
were of short-term intensive hand rehabilitation Eligibility criteria
programmes (i.e. 10–15 therapy sessions evaluated Studies had to be trials describing the effects of
over 2 or 3 weeks9,10); or 20 therapy sessions over home hand exercise programmes provided by health
Home hand exercises in RA, 2016, Vol. 119 51

professionals as part of conservative management; to 11 criteria (see Table 1).32,33 The first criterion, par-
adults with RA diagnosed by a physician, recruited ticipant eligibility, assessing external validity, is not
from either in- or out-patient or community settings; included in the total score, which is a maximum of
and at least one of the following outcomes were 10 if all criteria are met. As it is difficult to blind
measured: hand function, pain, grip strength and/ therapists and/or participants in most rehabilitation
or RoM. Studies were excluded if not published in trials, most cannot obtain the maximum score:
English; evaluated post-surgery hand exercise; high-quality trials with low risk of bias score 7 or
reported only in abstracts, poster presentations or more and low-quality trials with high risk of bias

Downloaded from by guest on 27 October 2018

conference proceedings; or were case reports, score 3 or less.33
descriptive articles, commentaries, letters or litera- To evaluate home hand exercise effects on hand
ture reviews. symptoms and function, trials were included if these
had low risk of bias; were randomized; had an RA
control group receiving no treatment or usual care;
Study selection and assessment of study and the home hand exercise programme was repro-
quality ducible. Trials were excluded if they had a moder-
After removing duplicates, we independently screened ate or high risk of bias; were not randomized; had a
titles and abstracts using these criteria. If met, we control group of people without arthritis; compared
retrieved full-text articles, re-checked for eligibility hand exercise regimens without a control group; or
and assessed methodological quality using the did not include home hand exercises.
PEDro scale, resolving disagreements when neces- To investigate different home hand exercise regi-
sary.16 This is a reliable, valid scale assessing mens effects, adherence strategies and safety

Table 1 Quality ratings of evaluated studies according to the PEDro methodology scoring system

1* 2 3 4 5 6 7 8 9 10 11 PEDro score Risk of bias

Manning et al.17 ✓ ✓ ✓ ✓ ✗ ✗ ✓ ✓ ✓ ✓ ✓ 8 Low

Lamb et al.18 ✓ ✓ ✓ ✓ ✗ ✗ ✓ ✓ ✓ ✓ ✓ 8 Low
O’Brien et al.11 ✓ ✓ ✓ ✓ ✗ ✗ ✓ ✗ ✓ ✓ ✓ 7 Low
Dogu et al.19 ✓ ✓ ✗ ✓ ✗ ✗ ✓ ✓ ✗ ✓ ✓ 6 Moderate
Hoenig et al.20 ✓ ✓ ✗ ✗ ✗ ✓ ✓ ✗ ✗ ✓ ✓ 5 Moderate
Brorssen et al.21 ✓ ✗ ✗ ✓ ✗ ✗ ✗ ✓ ✓ ✓ ✓ 5 Moderate
Ellegaard et al.22 ✓ ✗ ✗ ✓ ✗ ✗ ✗ ✓ ✓ ✓ ✓ 5 Moderate
Piga et al.23 ✓ ✓ ✗ ✓ ✗ ✗ ✗ ✗ ✓ ✓ ✓ 5 Moderate
Buljina et al.9 ✓ ✓ ✗ ✓ ✗ ✗ ✗ ✗ ✗ ✓ ✓ 4 Moderate
Ronningen and Kjeken24 ✓ ✗ ✗ ✓ ✗ ✗ ✗ ✗ ✓ ✓ ✓ 4 Moderate
Speed and Campbell25 ✓ ✗ ✗ ✗ ✗ ✗ ✗ ✓ ✓ ✓ ✓ 4 Moderate
Brighton et al.26 ✓ ✓ ✗ ✓ ✗ ✗ ✓ ✗ ✗ ✗ ✗ 3 High
Byers27 ✓ ✗ ✗ ✓ ✗ ✗ ✗ ✓ ✗ ✗ ✓ 3 High
Hawkes et al.28 ✓ ✓ ✗ ✗ ✗ ✗ ✓ ✗ ✗ ✓ ✗ 3 High
Dellhag et al.29 ✓ ✓ ✗ ✗ ✗ ✗ ✗ ✓ ✗ ✓ ✗ 3 High
Cima et al.12 ✗ ✓ ✗ ✗ ✗ ✗ ✗ ✓ ✗ ✗ ✓ 3 High
Rapoliene et al.10 ✗ ✗ ✗ ✓ ✗ ✗ ✗ ✗ ✗ ✗ ✓ 2 High
McLaughlin and Reynolds30 ✓ ✗ ✗ ✓ ✗ ✗ ✗ ✗ ✗ ✗ ✗ 1 High
Bromley et al.31 ✗ ✗ ✗ ✗ ✗ ✗ ✗ ✗ ✗ ✗ ✗ 0 High

Key: 1 = PEDro Scale criteria; External validity: 1 = eligibility criteria were specified*; Internal validity: 2 = random allocation; 3 = concealed
allocation; 4 = similarity at baseline; 5 = blinding of participants; 6 = blinding of therapists; 7 = blinding of assessors; 8 = measures of at least
one key outcome from at least 85% of participants initially allocated to groups; 9 = intention to treat principle; 10 = results of between group
comparisons; 11 = point measures and measures of variability reported. Maximum score = 10 (*criterion 1 is not included in scoring).
52 A. Hammond and Y. Prior, 2016, Vol. 119

additional trials were reviewed if these had a com- eight studies were excluded because of high risk of
parator group of people with RA (receiving usual bias,10,12,26–31 six of which did not include home
care or an alternate exercise regimen) and had mod- programmes.10,27–31 In addition, a cost-
erate risk of bias. effectiveness study of the trial led by Manning
et al.17 was included, as well as cost-effectiveness
Data extraction and analysis reported within the study by Lamb et al.18.

A predefined data extraction form was devised

Evidence for the effects of hand exercise on

Downloaded from by guest on 27 October 2018

including participant characteristics, intervention
hand function and hand symptoms
groups, exercise regimens (type, intensity and dur-
ation), outcome measures and results. Effects were Study features and methodological quality
summarized descriptively. Three prospective RCTs were included, with PEDro
scores of 7 or 8.11,17,18 All had external validity
and their methodological limitations were due to
Results the lack of blinding of therapists and participants
Study selection (see Table 1).

The search resulted in 3456 articles after duplicates

were removed (see Fig. 1). Following title/abstract Demographic data
review, 3433 articles were removed as either not spe- The total number of participants was 665 (range
cifically about hand exercises in RA; protocol articles; 67–490), including 163 men and 502 women, with
commentaries on hand exercise trials; or systematic a mean age of 59 years and disease duration of
or narrative reviews of hand exercises.8,13–15 Twenty- 8.39 years. Two included mainly people with estab-
three articles were selected for full-text review, of lished RA11,18 and one early RA (i.e. less than 2
which three were excluded because they evaluated years since diagnosis).17
general exercise programmes34,35 or a combined ther-
apy intervention,36 from which hand exercise effects Control groups
specifically could not be identified. Control participants continued to receive usual care
Nineteen articles were assessed using the PEDRo from their rheumatology team11,17,18 In two stud-
scale, designed to assess rehabilitation trials (see ies, this also included provision of an Arthritis
Table 1). Three RCTs with low risk of bias met cri- Research UK booklet about joint protection11,17
teria and were included in all evidence synthe- and in the third up to 1.5 hours of joint protection
ses.11,17,18 Four moderate risk of bias trials were education and, if applicable, functional splinting.18
excluded from the effects of the home hand exercise
synthesis but included in the types of exercise, Intervention groups
adherence and safety syntheses.19,20,22,24 All three trials evaluated usual care (as above) plus
Four moderate risk of bias trials were excluded RoM and resistance hand exercises (see
from all evidence syntheses. Buljina et al.9 evaluated Table 2),11,17,18 Participants attended one,11 four,17
immediate effects of 15 therapy sessions of daily or five18 sessions of therapist-supervised exercise
exercise, thermotherapy, faradic, radon and wax therapy, followed by a daily home exercise pro-
baths (a regimen rarely used in Western countries) gramme. O’Brien et al.11 evaluated five RoM and
without home exercise afterwards. Piga et al.23 eval- three light resistance hand exercises and Lamb
uated a purpose-built device, with integrated exer- et al.18 five RoM and four medium resistance hand
cise tools and remote telemonitoring, which is not exercises plus two upper limb RoM exercises.
commercially available and not reproducible as a Manning et al.17 evaluated six medium resistance
home programme. Two further trials were excluded exercises selected from 16 (6 hand and 10 arm)
as they had healthy controls.21,25 The remaining standardized exercises to meet individual’s needs.
Home hand exercises in RA, 2016, Vol. 119 53

Records identified through Additional records identified through hand
database searching searching review article references
(n = 3992) (n = 1)

Downloaded from by guest on 27 October 2018

Records after duplicates removed [n = 537]
(n = 3456)

Records screened Records excluded after screening title/abstract

(n = 3456) (n = 3433)

Full-text articles excluded, with reasons:

Full-text articles reviewed
General exercise trial (inc. hand exercises) x
(n = 23)
2; hand therapy trial x 1.

Full-text articles excluded, with reasons:

Studies evaluated high risk of bias x 8; moderate risk of bias x 8
(n = 20) (including No control group x 2; Control
groups of healthy people x 2).

Studies included in
evidence synthesis of home
hand exercise effectiveness
(n = 3 + 1 economic

Fig. 1 Flow diagram of the results of the study selection procedure, in accordance with Preferred Reporting Items for
Systematic Reviews and Meta-analyses (PRISMA) guidelines.

The number of hand exercises thus varied. Outcome measures and outcomes
However, most arm exercises involved gripping a Pooling of data was not possible as outcome mea-
resistance band whilst pulling, involving finger and sures usually differed between studies (see Table 2).
wrist flexor muscles.17 Self-reported hand function (the primary out-
come in all three studies): this was evaluated using
Follow-up the AIMS2 Hand and Upper Limb Function scales11;
Follow-up also varied with short-term assessments the Disabilities of the Hand, Arm and Shoulder
at 2.5,17 3,11 and 4 months18 and long-term assess- questionnaire (DASH)17; and the Michigan Hand
ments at 6,11 8,17 and 12 months.18 Outcomes Questionnaire (MHQ).18 In the short

Table 2 Summary of low to moderate bias trials of home hand exercise programmes in rheumatoid arthritis

Authors Participants Intervention Groups Exercise regimen; delivery method, Outcome measures and results: (significant results in italics;
intensity and duration % score changes or actual changes for strength and RoM)

RCTs: low risk of bias

O’Brien et al.11 N = 67; mean age E1 Hand and wrist 1 × 30 minute individual training At 6 months: in favour of E1 (RoM + LR)
59.6 y; mean disease RoM + LR (pinching session (+15-minute review at UL/Hand Function: AIMS 2 Upper Limb Function:
duration 13.5 y; towel/resistance bands) 2 weeks) followed by home 3 m NS; 6 m p = 0.002 (E1 +17%; E2 −4%; C −6%);
M:F = 21:46 E2 Hand and wrist exercise programme: AIMS 2 Hand Function: 3 m and 6 m NS
RoM E1: 5× RoM + 3× LR exercises Jebsen Hand Function Test: 3 m and 6 m NS
C Usual care E2: 8× RoM exercises Dominant grip (Jamar dynamometer: lbs): 3 m and 6 m NS
Both groups 1 and 2: initially 5 reps; Key pinch (B&L gauge):3 m NS; 6 m p = 0.01 (E1 + 1 kg;
×10 reps at 1 month; ×20 reps at E2 + 0.3 kg; C −1 kg).
3 months. 10–20 minutes 2×/day; Finger flexion (goniometer): 3 m p = 0.05 (E1 + 20°;
6 months. E2 + 15°; C −3°); 6 m NS
Tender and swollen joint count: 3 m (not reported);
6 m NS
Manning et al.17 N = 108; mean E Hand and arm LR Group programme: 4 × 1 hour 2×/ At 2.5 (12 weeks) and 8 months (36 weeks) in favour of E
age = 55 (SD 15)y; and MR week for 2 weeks followed by daily (LR + MR):
mean disease C Usual care home exercise: UL/Hand function: DASH: 12 w p = 0.02 (E −12%;
duration = 1.67 (SD 6 LR and MR exercises (therapist- C +4%; ES = 0.5); 36 w NS
1.58)y; selected from 16 options), ×10 reps, Grip Ability Test: 12 w p = 0.01 (E −12%; C +2.5%);
M:F = 26:82 increasing to 3 × 10 reps; increasing 36 w NS
resistance; 15 minutes daily; 12 weeks Dominant Grip (Jamar: Newtons): 12 w and 36 w NS
Non-dominant Grip: 12 w p = 0.005 (E +22.4 N; C −9 N);
36 w NS
Overall Pain (VAS): 12 w p = 0.01 (E −26%; C +4%);
36 w p = 0.05 (E −16%; C+9%)
Overall fatigue (VAS): 12 w and 36 w NS
Morning stiffness (min): 12 w and 36 w NS
DAS28: 12 w p = 0.05 (E −15%; C −2%); 36 w NS
Self-efficacy (pain): 12 w p = 0.02 (E +8%; C −10%);
36 w p = 0.05 (E −16%; C +9%)
Self-efficacy (symptoms): 12 w p = 0.04 (E + 8%; C −8%);
36 w NS
RAQoL: 12 w and 36 w NS
A. Hammond and Y. Prior, 2016, Vol. 119

Downloaded from by guest on 27 October 2018

Lamb et al.18 N = 490; mean E Hand, wrist, elbow, 5× individual supervised exercise At 4 and 12 months in favour of E (RoM + LR + MR):
age = 62.4 (SD 11.5)y; shoulder RoM + Hand/ sessions followed by daily home UL/Hand Function: MHQ (overall hand function):4 m
median disease wrist LR and MR; exercise: 7 RoM (×10 reps) and p = 0.0001 (E +17%; C +8%); 12 m p = 0.003 (E +15%;
duration = 10 C Usual care 4LR/MR exercises (×10 reps C +7%; ES = 0.3)
(IQR 4,22)y. increasing to 3 × 10 reps), Dexterity (9 hole peg test): 4 m NS; 12 m p = 0.02
M:F = 116: 374 increasing resistance; 15+ minutes (E −5%; C −0.3%)
daily; 52 weeks Dominant Grip (MIE analyser): 4 m p = 0.01 (E +15.5 N;
C +7.4 N); 12 m NS
Pinch grip (MIE); 4 m NS; 12 m p = 0.04 (E +5.3 N; C +2.4 N)
Finger flexion (mm): 4 m and 12 m NS
Finger extension (mm):4 m p = 0.05 (E +4mm;
C +1.5mm); 12 m p = 0.007 (E +4.8mm; C +1.4 mm)
MHQ (pain): 4m p = 0.04 (E −15%; C −10%); 12 m NS
Home hand exercises in RA, 2016, Vol. 119

Tender joint count: 4 m p = 0.007 (E −25%; C −8%); 12 m NS

Self-efficacy pain: 4 m p = 0.02 (E +8%; C +3%); 12 m NS
EQ5D health status: 4 m and 12 m NS
RCTs and CCT with moderate risk of bias
Hoenig et al.20: N = 57; mean age = 57; E1 RoM (tendon 1 × individual exercise training At 12 weeks, in favour of Groups E2/3, i.e. LR with/
RCT mean disease gliding) session followed by home exercise: without RoM compared to controls:
duration = 11.3y; E2 LR E1 and 3: 1 × RoM; Hand function: Group E3(RoM + LR) Dexterity (9 hole
M:F = ? E3 RoM + LR E2 and 3: 3 × LR exercises. Each ×10 peg test):p < 0.05 left hand only (E3 −17%; C +3%)
C Usual care reps, twice daily 10–20 minutes home Grip (aneroid manometer): Groups E1,2,3 combined v C)
exercise: 12 weeks p < 0.05 (larger increase in Groups E2 and 3 of +22% vs
Group E1 + 6%)
RoM: Finger (PIP) extension: Group E2 (LR) p < 0.05 left
hand only; MCP flexion: NS; Ulnar deviation: NS
Swelling: PIP circumference: NS
Joint count (pain): Group E1 (RoM) p < 0.05 right hand
only (E1 −15%; C +80%)
Ellegaard et al.22: N = 36; mean E MR (Theraband 1 × individual exercise training At 8 weeks: in favour E (MR)
CCT age = 61y; mean Hand Exerciser: session followed by home exercise Between group: Ultrasound: NS synovial perfusion (i.e.
disease duration = 9 therapist choice from E: 1 × MR wrist exercise only; exercise not detrimental as no increase in inflammation)
y; M;F 0:36 four resistance levels) squeeze Exerciser (selected at 60% Within group:
C RA age-matched max. grip strength) for 10 s; 2 × 5 Grip (dynamometer): NS
controls. reps daily; 8 weeks Wrist joint pain on motion (VAS): p = 0.04 (E −36%)
(C: data not reported)


Downloaded from by guest on 27 October 2018


Table 2 Continued

Authors Participants Intervention Groups Exercise regimen; delivery method, Outcome measures and results: (significant results in italics;
intensity and duration % score changes or actual changes for strength and RoM)

Trials with moderate risk of bias comparing exercise regimens (no control group)
Ronningen and N = 60; mean age E1 RoM + LR (variable 7 × daily group in-patient exercise At 2 weeks, in favour of Group E2 (intensive RoM + LR);
Kjeken24: 46.5y; mean disease intensity) sessions, followed by home at 14 weeks, more in favour E2:
(not duration = 10.5y; E2 RoM + LR (high exercise: Hand function: Grip Ability Test: 2 w p = 0.02 (E1 −7%;
randomized) M:F = 10:50 intensity) 1: 4 × RoM + 7 × LR × 3 reps; 10 E2 −21%); 14 w NS
minutes. Home exercise as many days Dominant Grip (Grippit): 2 w and 14 w NS
as patient wishes. Non-dominant grip (Grippit): 2 w p = 0.04 (E +18 N;
2: 3 × RoM + 6 × LR × 10 reps, 10– C −3N); 14 w p = 0.04 (E +28 N; C −6N)
20 minutes; minimum 5 ×/week home Dominant Pinch (Grippit): 2 w p = 0.01 (E +9 N; C −1N);
exercise. 14 w NS
14 weeks Non-dominant Pinch (Grippit):2 w p = 0.05 (E +8 N;
C 0 N); 14 w NS
Hand pain resisted grip (VAS):2 w p = 0.04 (E −24%;
C +24%); 14 w NS
Dogu et al.19 N = 47; mean age E1 RoM 10 × supervised individual exercise At 6 weeks: NS differences between; both groups improved
52.6 y; mean disease E2 LR sessions + wax baths followed by (within- group E1 and E2 differences shown).
duration 9.41 y, M: home exercise: Hand function: DHI: p = 0.002 (E1 −27%; E2 −27%);
F = 0:47 1: 6 × RoM Dexterity (9 hole peg test): p < 0.005) (E1 −7%; E2
2: 6 × LR exercises; −22%).
Both groups: 10 reps × 5 days/week Dominant Grip (dynamometer): Group E2 only p = 0.03
for 2 weeks; followed by ×10 reps (E1 +1 kg; E2 +3.5 kg)
15–20 minutes daily home exercise Non-dominant Grip (dynamometer): Group E1 only
for 4 weeks p = 0.01 (E1 +1 kg; E2 −0.5 kg)
Hand pain (VAS); p < 0.02 (E1 −40%; E2 −14%)
Disease activity (DAS28): p < 0.002 (E1 −20%;
E2 –25%)
RAQoL: p < 0.003 (E1 −21%; E2 −20%)

Key: E = Experimental group (1, 2,3: if more than 1 group); C = control group; ES = effect size; RoM = Range of movement exercises; LR = light resistance exercises (soft therapeutic putty, salt
dough or soft rubber objects, unless otherwise stated); MR = moderate resistance (medium to firm therapeutic putty and elastic resistance bands, unless otherwise stated); reps = repetitions;
UL = upper limb; RCT = randomized controlled trial; CCT = case controlled trial; AIMS2 (Arthritis Impact Measurement Scale 2; MHQ = Michigan Hand Outcomes Questionnaire:
DASH = Disabilities of the Arm, Shoulder and Hand questionnaire; DHI = Duruoz Hand Index questionnaire; RAQoL = RA Quality of Life questionnaire; VAS = visual analogue scale.
A. Hammond and Y. Prior, 2016, Vol. 119

Downloaded from by guest on 27 October 2018

Home hand exercises in RA, 2016, Vol. 119 57

term, compared to the control groups, Manning Disease activity: two studies measured disease
et al.17 and Lamb et al.18 showed significant activity, although differently: DAS2817 and C-reactive
improvements. In the long term, only Lamb et al.18 protein levels.18 Both showed significant improve-
showed significantly improved hand function (effect ments in the short term but not in the long
size 0.3) and O’Brien et al.11 significantly improved term.17,18
upper limb function. Self-efficacy for managing pain: in the short
Objective hand function: this was measured term, this significantly improved17,18 and remained
using the Jebsen Hand Function Test,11 the Grip higher in the long term in one study17 and almost

Downloaded from by guest on 27 October 2018

Ability Test (GAT)17 (both timed tests of hand grip significantly improved in the other.18
function) and the Nine Hole Peg Test, a timed fin- Health: one study measured health-related qual-
ger dexterity test.18 In the short term, significant ity of life17 and another health status18 with no sig-
improvements were found in only one study,17 but nificant changes in either.
Lamb et al.18 showed long-term significant
improvements in dexterity. Cost-effectiveness of home hand exercise
Hand grip strength: two studies used the Jamar programmes
dynamometer but with different methods and units:
Cost-effectiveness was evaluated in only two stud-
mean grip force in pounds11 and peak force in
ies. The exercise programmes, plus other health
Newtons.17 The third measured maximum grip force
care use during follow-up, were identified as £8237
in Newtons using the MIE Digital Grip Analyser.18
and £10318 more expensive than usual care but led
In the short term, one study showed significantly
to an increase in 0.0337 and 0.0118 QALYs, respect-
improved dominant18 and one non-dominant hand
ively. Both concluded the exercise programmes
grip strength.17 In the long term, grip strength
were cost-effective.18,37
improvements were maintained and better than the
control groups but not significantly so.17,18
Effectiveness of different home hand
Pinch strength: two studies used the B&L pinch
exercise regimens
gauge (lbs)11 and the MIE Digital Analyser
(Newtons).18 In the short term, there were no differ- A wider range of trials were included to review the
ences in pinch strength. In the long term, it was sig- effectiveness of different types of home exercise
nificantly improved in both studies.11,18 regimens. Seven studies were therefore included, i.e.
Finger RoM: two studies measured this using a the three above plus the four moderate level of bias
goniometer11 and composite finger flexion and studies meeting this evidence synthesis’ entry cri-
extension with a ruler.18 In the short and long term, teria. These four additional studies included people
significant improvements were found in finger with established RA and had short-term follow-ups
extension,18 whilst finger flexion did not of between 4 and 14 weeks19,20,22,24 (see Table 2).
improve.11,18 Exercise regimens can be considered in terms of
Hand pain: this was measured in two studies, their content, intensity and provision method.
using a 100 mm VAS17 and the MHQ Pain scale.18
Significant improvements were identified in both Exercise programme content
studies in the short term17,18 and in one in the long All seven studies included resistance exer-
term.17 cises.11,17,18,19,20,22,24 Five included between two
Other symptoms: these were also measured dif- and six resistance exercises using therapeutic putty
ferently and not in all three studies: tender and and resistance bands: squeezing putty in a full fist, a
swollen joint counts,11,18 morning stiffness17 and hook fist and/or between fingers; pinching putty;
fatigue.17 In the short term, joint counts improved stretching the fingers out against putty; rolling
only in one18 and not in the long term. Morning putty; and wrist extension exercises using resistance
stiffness and fatigue did not differ significantly.17 bands.15,17–20,24 One used a Theraband hand exercise
58 A. Hammond and Y. Prior, 2016, Vol. 119

ball22 and one used rolling/pinching a towel and putty and resistance bands were initially graded to
resistance bands.11 Three specified exercises were suit individuals’ abilities, then resistance increased
held at the position of maximum effort for 2–5 sec- to medium.17–19,24 Two studies asked participants
onds.15,17–19 Only Manning et al.17 also included to set their effort level during exercise at moderate
arm resistance exercises. progressing to hard on the Borg Scale of Perceived
Five studies also included RoM exercises (median Exertion.17,18 Three included RoM exercises.18,19,24
six RoM exercises; range one to nine)11,18–20,24 The All four were therapist-supervised programmes with
hand RoM exercises commonly included were wrist a median of 6 (range 4–10) sessions.17–19,24 Two

Downloaded from by guest on 27 October 2018

flexion, extension and circumduction, pronation and stated the time taken was 4–5 hours.17,18
supination; finger tendon gliding, radial walking and All resulted in short-term significant improve-
abduction; touching the tip of each finger to the ments in hand function, dominant and/or non-
thumb; and thumb extension, opposition and thumb dominant hand grip strength and pain.17–19,24 Only
interphalangeal joint flexion.11,15,18–20,24 Only Lamb Lamb et al.18 evaluated RoM and no change was
et al.18 included arm RoM exercises (i.e. internal and identified. One study compared an intensive regi-
external shoulder rotation).15 men (a total of 60 repetitions of resistance exercises
Three studies had no explanation for choice of most days) to a conservative regimen (a total of 21
exercise content,19,20,22 although frequency met ‘gen- repetitions of resistance exercises several days a
eral recommendations for resistance programmes.’22 week), identifying better short-term outcomes with
One was based on ‘new research.’24 Manning et al.17 the intensive programme.24 Only two included
adapted a knee pain self-management and exercise long-term follow-ups, with continued improvements
programme, with upper limb exercise content deter- in self-reported hand function,18 pinch strength18
mined collaboratively with clinicians. O’Brien et al.11 and pain.17
surveyed 60 hand therapists about their exercise pro-
vision, although how the exercise programme was
defined from this was not explained. Finally, Lamb Low-intensity single training session exercise
et al. identified potential exercises through systematic regimens
review and a consensus meeting of hand therapists. Three studies were considered low intensity. One
Exercises were then selected by the research team evaluated one medium resistance exercise (squeez-
based on including all functionally relevant move- ing a Theraband Hand Exerciser22) performed ten
ments/muscle actions at the hand and wrist, avoiding times for 5 minutes daily. Two included 2 × 10
replication and ensuring convenience.15,18 repetitions of three light resistance exercises (using
either medium soft putty20 or rolling/pinching a
towel and resistance bands11), i.e. 60 light resistance
Intensity and provision method of exercise repetitions daily. Both were performed for 15–20
regimens minutes daily. Two included RoM exercises.11,20 All
High-intensity, therapist-supervised exercise three programmes were taught through one individ-
regimens ual training session of up to 1 hour,11,20,22 two of
Four exercise programmes were considered high which also added either a 15-minute review11 or a
intensity. Two included 10 repetitions of four to six reminder.22
light progressing to medium resistance exercises for Two of these low-intensity studies resulted in
10–20 minutes daily19 or most days24 (i.e. between short-term significant improvements in finger
40 and 60 medium resistance exercise repetitions). RoM,11,20 one in hand function (dexterity)20 and
Two included 10 progressively up to 3 × 10 repeti- one in pain.22 None showed improvements in grip or
tions of four to six light progressing to medium pinch strength. Only O’Brien et al.11 had a long-term
resistance exercises (i.e. potentially up to 120 or follow-up, demonstrating improved self-reported
180 repetitions) performed daily.17,18 Therapeutic upper limb (not hand) function at six months.
Home hand exercises in RA, 2016, Vol. 119 59

RoM-only exercise regimens repetitions or days exercising and most were then
Three studies included RoM-only exercise pro- able to continue.24 In the other, six withdrew due
grammes as comparator groups.11,19,20 Two studies to pain from exercise or flare-ups and were not
identified no changes, both of which were taught in included in analyses.22 Ronningen and Kjeken24
single therapy sessions.11,20 However, similar compared high- and low-intensity resistance exer-
improvements in hand function, pain and grip to cise regimens, identifying better short-term out-
the RoM and resistance exercise programme were comes from the high-intensity programme, with no
found in one high-intensity 10 session therapist- negative effects. Of the three studies evaluating dis-

Downloaded from by guest on 27 October 2018

supervised programme.19 ease activity, all reported short-term improve-
ments.17–19 Ellegaard et al.22 evaluated blood flow
Strategies to enhance adherence with hand in the wrist joints using Doppler ultrasound, identi-
and upper limb home exercises fying there was no significant increase in synovial
perfusion (reflecting disease activity) following 8
Six studies included self-report daily exercise diaries
weeks of low-intensity daily exercise.
to monitor progress.11,17,18,20,22,24 Four reported
adherence achieved: three that exercises were com-
pleted by between 73 and 95% of participants who Discussion
returned diaries17,20,22; and one that there was
This is the first systematic review to evaluate the
greater compliance with exercise at 4 months com-
effects of home hand exercise programmes in RA.
pared to the control group.18 Exercise intensity
In the short term, resistance exercises, with or with-
(daily repetitions or time) was not reported.
out RoM exercises, improve self-reported hand
Six studies provided a booklet containing exer-
function,17,18 pain,17,18 self-efficacy,17,18 grip
cise instructions with photographs or draw-
strength18 and RoM.11 However, in the long term
ings.11,17–20,24 Three included either a review
(i.e. 6–12 months), benefits were less consistent but
appointment11 or a telephone call(s) to remind par-
still found in hand18 or upper limb function,11
ticipants to perform exercises.19,22
pain,17 pinch strength11,18 and self-efficacy17 but
The two studies with long-term follow-ups
not RoM.11,18 Hand exercises were also cost-
included behaviour change strategies to facilitate
effective.17,18,37 Whilst there are only three high-
long-term adherence.17,18 These included enhancing
quality RCTs published to date,11,17,18 the evidence
self-efficacy to perform exercises, discussing exercise
is that home hand exercise programmes are effective
barriers, problem-solving, using exercise diaries,
in people with RA.
goal-setting, verbal and written contracting.
In clinical practice, therapists need to know
Therapists were trained in using these cognitive-
what exercise regimens to provide and how. By
behavioural approaches for either two17 or four
including both low and moderate risk of bias stud-
hours.18 Both trials used therapist manuals to sup-
ies, we identified that in the short term, the most
port standardizing exercise programme deliv-
effective programmes were all high intensity and
ery.17,18 Therapists also taught participants to use
therapist-supervised for four or more sessions.
the Borg Perceived Rate of Exertion scale to moni-
These consistently resulted in improved hand func-
tor and progress the resistance applied during their
tion, grip strength and pain, with minimal or no
home exercise programmes.17,18
adverse effects.17–19,24 In contrast, low-intensity
programmes taught in one session (with or without
Safety concerns reminders) were less consistent in demonstrating
No adverse effects occurred in four studies.17–19,24 short-term improvements in either RoM,11,20 hand
One did not include adverse effect reporting.11 Two function20 or pain22 and did not result in improved
reported some problems with hand pain. In one, grip strength.11,20,22 RoM-only programmes were
participants temporarily reduced the number of found to have no effect apart from in one trial with
60 A. Hammond and Y. Prior, 2016, Vol. 119

a high level of initial therapy (10 sessions), which is exercises as a rest from other activities and resist-
not usual practice in the United Kingdom.19 ance exercises when watching television.
These findings indicate home hand exercise regi- Potentially, Manning et al.’s programme required
mens should include at least four and up to six light too much time away from other activities.
progressing to medium resistance hand exercises Further research is needed to identify the opti-
using therapeutic putty and resistance bands per- mum type and number of resistance hand exercises,
formed at high intensity (i.e. 10 repetitions of each repetitions and days performed to be effective but
exercise most days/daily, repeated twice daily as still achievable for people with RA. Given that

Downloaded from by guest on 27 October 2018

hands improve). Including some hand and upper home programmes seem to have less effect on hand
limb RoM exercises may also be helpful. Home RoM, research is needed to determine whether or
programmes should be taught initially with therap- how many RoM exercises to include. Programmes
ist support over at least four sessions, in either with too many exercises, taking up too much time,
groups or individually. It is unclear if a high- are less likely to be followed. No studies success-
intensity programme taught over one session (with fully monitored how much and which types of exer-
or without review) would be as effective, as no trial cises were being performed and thus what level
has evaluated this. It is more likely that several ses- provides optimum results. Further consensus study
sions are more effective than only one, because with clinicians is needed to determine which resist-
many with RA have concerns about hand exercise ance exercises, based on systematic review and
increasing pain. Repeated supervision enables peo- practice surveys, are best included. Qualitative
ple with RA to develop the skills and confidence to research is needed to investigate people with RA’s
perform exercises correctly and progress resistance views of home hand exercise programmes, which
and intensity. Including education about RA and resistance exercises they can most easily continue to
hand exercises also helps allay fears and improves perform, how much exercise they can realistically
self-efficacy.17,18 These intensive programmes are integrate into their lives and the best methods to
cost-effective.18,37 support them in continuing to exercise, as only one
Helping people long-term continue performing small study has been conducted to date.24 This will
hand exercises at effective levels is a major chal- help determine an optimum resistance exercise pro-
lenge. The most successful, and largest, study has gramme. Given that regular follow-up therapy
trained therapists in using cognitive-behavioural appointments are now less common in practice,
approaches and integrated these in programme other methods to remind people to do hand exer-
delivery.15,18 As 17 sites were involved, this sug- cises are needed. The development and evaluation
gests such training can be disseminated into prac- of an App, including demonstrations of exercises,
tice. Research evaluating self-management and joint with self-monitoring of frequency and intensity of
protection programmes also demonstrates such hand exercises and providing positive messages
approaches are significantly more effective in when goals are achieved, may have the potential to
improving outcomes than brief therapy.38,39 At pre- support people long term. Home hand exercises are
sent, these approaches are not commonly used in relevant for other conditions, such as hand osteo-
practice and there is a need to increase therapists’ arthritis, meaning a hand exercise App could have a
skills. Interestingly, Manning et al.17 also used simi- wider client base and also enable remote data col-
lar approaches but with less effect. This may have lection in trials. Only one study recruited people
been due to the different exercises included. Most with early RA17 and thus further research is needed
arm exercises were performed whilst standing, into the effectiveness of home programmes in this
requiring full attention and higher time demands. group. Therapists also have concerns about the use
People successfully continuing with home hand of hand exercises when people have inflamed or
exercises explained they integrated these into their deformed joints. Very few adverse effects were
daily routines,24 for example, performing RoM reported in these studies but most people had
Home hand exercises in RA, 2016, Vol. 119 61

established RA and were on stable drug therapy December 2015.

regimens, meaning further evaluation of the effects chapter/Recommendations#the-multidisciplinary-team
of home programmes in these groups is also needed. (2 February 2016, date last accessed).
8. Bergstra SA, Murgia A, Te Velde AF, et al. A systematic
Comparisons between trials would also be easier of
review into the effectiveness of hand exercise therapy in
a core set of outcome measures is agreed. Little
the treatment of rheumatoid arthritis. Clin Rheumatol
attention to date has been focused on the impact of 2014;33:1539–48.
hand exercises on participation and this also should 9. Buljina AI, Taljanovic MS, Avdic DM, et al. Physical
be considered. and exercise therapy for treatment of the rheumatoid

Downloaded from by guest on 27 October 2018

hand. Arthr Rheum 2001;45:392–7.
10. Rapoliene K, Krisciunas A. The effectiveness of occupa-
Conclusion tional therapy in restoring the functional state of hands
Home hand exercise programmes are effective at in rheumatoid arthritis patients. Medicina (Kaunas,
Lithuania) 2006;42:823–8.
improving hand function, grip strength and pain in
11. O’Brien A, Mullis R, Mulherin D, et al. Conservative
RA. High-intensity resistance exercise programmes
hand therapy treatments in rheumatoid arthritis—a ran-
taught by therapists over at least several sessions domized controlled trial. Rheumatology 2006;45:577–83.
including strategies to promote long-term adherence 12. Cima SR, Barone A, Porto JM, et al. Strengthening
seem to be most effective and are cost-effective. exercises to improve hand strength and functionality in
rheumatoid arthritis with hand deformities: a rando-
mized, controlled trial. Rheumatol Internat 2013;33:
Acknowledgements 725–32.
Dr Roy Vickers, Academic Support Librarian (Health 13. Wessel J. The effectiveness of hand exercises for persons
Sciences), University of Salford: for assistance with develop- with rheumatoid arthritis. J Hand Ther 2004;17:174–80.
ing the search strategy and conducting the literature search. 14. Stewart M. Researches into the effectiveness of physiother-
apy in rheumatoid arthritis of the hand. Physiotherapy
References 15. Heine PJ, Williams MA, Williamson E, et al. on behalf
1. Adams J, Burridge J, Mulleee M, et al. Correlation of the SARAH team. Development and delivery of an
between upper limb fuctional ability and structural exercise intervention in rheumatoid arthritis: strengthen-
hand impairment in an early rheumatoid population. ing and stretching for rheumatoid arthritis of the hand.
Clin Rehabil 2004;18:405–13. Physiotherapy 2012;98:121–30.
2. Horsten NC, Ursum J, Roorda LJ, et al. Prevalence of 16. Maher CG, Sherrington C, Mosely AM, et al.
hand symptoms, impairment and activity limitations in Reliability of the PEDro scale for rating quality of ran-
rheumatoid arthritis in relation to disease duration. domised controlled trials. Phys Ther 2003;83:713.
J Rehabil Med 2010;42:916–21. 17. Manning VL, Hurley MV, Scott DL, et al. Education,
3. Roubenoff R. Exercise and inflammatory arthritis. self-management, and upper extremity exercise training
Arthritis Care Res 2003;49:263–6. in people with rheumatoid arthritis: a randomized con-
4. Hammond A, Kidao R, Young A. Hand impairment trolled trial. Arthritis Care Res 2014;66:217–27.
and function in early rheumatoid arthritis. Arthritis and 18. Lamb SE, Williamson EM, Heine PJ, et al. Exercises to
Rheumatism 2000;43(Suppl. 9):S285. improve function of the rheumatoid hand (SARAH): a
5. Johnsson PM, Eberhardt K. Hand deformities are randomised controlled trial. Lancet 2015;385:421–9.
important signs of disease severity in patients with early 19. Dogu B, Sirzai H, Yilmaz F, et al. Effects of isotonic
rheumatoid arthritis. Rheumatology 2009;48:1398–401. and isometric hand exercises on pain, hand functions,
6. Dias JJ, Singh HP, Taub N, et al. Grip strength charac- dexterity and quality of life in women with rheumatoid
teristics using force-time curves in rheumatoid hands. arthritis. Rheumatol Internat 2013;33:2625–30.
J Hand Surg (European) 2013;38:170–7. 20. Hoenig H,Groff G, Pratt K, et al. A randomized con-
7. National Collaborative Centre for Chronic Conditions. trolled trial of home exercise on the rheumatoid hand.
Rheumatoid arthritis: national clinical guidelines for man- J Rheumatol 1993;20:785–9.
agement and treatment in adults. London: National 21. Brorsson S, Hilliges M, Sollerman C, et al. A six-week
Institute for Health and Clinical Excellence, 2009. Updated hand exercise programme improves strength and hand
62 A. Hammond and Y. Prior, 2016, Vol. 119

function in patients with rheumatoid arthritis. J Rehab standard physiotherapeutic techniques. Br J Rheumatol
Med 2009;41:338–42. 1994;33:555–61.
22. Ellegaard K, Torp-Pedersen S, Lund H, et al. The effect of 32. Macedo LG, Elkins M, Maher CG, et al. There was evi-
isometric exercise of the hand on the synovial blood flow dence of convergent and construct validity of
in patients with rheumatoid arthritis measured by colour Physiotherapy Evidence Database quality scale for
Doppler ultrasound. Rheumatol Internat 2013;33:65–70. physiotherapy trials. J Clin Epidemiol 2010;63:920–5.
23. Piga M, Tradori I, Pani D, et al. Telemedicine applied 33. Elkins MR, Herbert RD, Moseley AM, et al. Rating the
to kinesiotherapy for hand dysfunction in patients with quality of trials in systematic reviews of physical therapy
systemic sclerosis and rheumatoid arthritis: recovery of interventions. Cardiopul Phys Ther J 2010;21:20–6.

Downloaded from by guest on 27 October 2018

movement and telemonitoring technology. J Rheumatol 34. Bell MJ, Lineker SC, Wilkins AL, et al. A randomized
2014;41:1324–33. controlled trial to evaluate the efficacy of community
24. Ronningen A, Kjeken I. Effect of an intensive hand based physical therapy in the treatment of people with
exercise programme in patients with rheumatoid arth- rheumatoid arthritis. J Rheumatol 1998;25:231–7.
ritis. Scand J Occup Ther 2008;15:173–83. 35. Baillet A, Payraud E, Niderprim VE, et al. A dynamic
25. Speed CA, Campbell R. Mechanisms of strength gain in exercise programme to improve patients’ disability in
a handgrip exercise programme in rheumatoid arthritis. rheumatoid arthritis: a prospective randomized con-
Rheumatol Internat 2012;32:159–63. trolled trial. Rheumatology 2009;48:410–5.
26. Brighton SW, Lubbe JE, van der Merwe CA. The effect 36. Mathieux R, Marotte H, Battistini L, et al. Early occu-
of a long-term exercise programme on the rheumatoid pational therapy programme increases hand grip
hand. Br J Rheumatol 1993;32:392–5. strength at 3 months: results from a randomised, blind,
27. Byers PH. Effect of exercise on morning stiffness and controlled study in early rheumatoid arthritis. Ann
mobility in patients with rheumatoid arthritis. Res Nurs Rheum Dis 2009;68:400–3.
Health 1985;8:275–81. 37. Manning L, Kaambwa B, Ratclifee J, et al. Economic
28. Hawkes J, Care G, Dixon JS, et al. Comparison of three evaluation of a brief education, self-management and
physiotherapy regimens for hands with rheumatoid upper limb exercise training in people with rheumatoid
arthritis. Physiother Practice 1985;2:155–60. arthritis (EXTRA) programme: a trial-based analysis.
29. Dellhag B, Wollersjo I, Bjelle A. Effect of active hand Rheumatology 2015;54:302–9.
exercise and wax bath treatment in rheumatoid arthritis 38. Iversen M, Hammond A, Betteridge N. Self-
patients. Arthritis Care Res 1992;5:87–92. management of rheumatic diseases: state of the art and
30. McLaughlin JE, Reynolds WJ. An evaluation of thera- future directions. Ann Rheum Dis 2010;69:955–63.
peutic exercise on hand function in rheumatoid arth- 39. Hammond A, Freeman K. One year outcomes of a ran-
ritis. Physiother Can 1973;25:71–6. domised controlled trial of an educational-behavioural
31. Bromley J, Unsworth A, Haslock I. Changes in stiff- joint protection programme for people with rheumatoid
ness following short- and long-term application of arthritis. Rheumatology 2001;40:1044–51.