Effects of a progressive resistance exercise program in patients with hand osteoarthritis: a randomized, controlled trial with a blinded assessor

Key points

1. 01Primary outcome (numerical pain scale) showed no significant between-group difference after 12 weeks (P = 0.085 dominant, P = 0.295 non-dominant)
2. 02AUSCAN total score improved significantly in the exercise group (mean difference -9.9, 95% CI 4.07-15.73, P = 0.005)
3. 03AUSCAN pain domain (pain during activities) favoured the exercise group (mean difference -3.26, 95% CI 1.06-5.46, P = 0.006)
4. 04Cochin hand function scale improved significantly in the exercise group (mean difference -6.27, 95% CI 0.18-12.36, P = 0.042)
5. 05Grip and pinch strength showed no significant between-group differences; patient satisfaction strongly favoured the exercise group (P < 0.001 at both 6 and 12 weeks)

How it was conducted

Design

Randomized controlled trial with blinded assessor, intention-to-treat analysis

Setting

Outpatient rheumatology rehabilitation unit, University Hospital, Sao Paulo, Brazil

Participants

60 adults (predominantly female) with hand osteoarthritis, aged 50 or older, pain 3-8 on NPS

Groups

Exercise group (n=30): 12-week supervised progressive resistance program, 2 sessions/week; Control group (n=30): single education session only

Primary outcome

Pain intensity on numerical pain scale (0-10) at 6 and 12 weeks

Secondary outcomes

AUSCAN index, Cochin scale, grip and pinch strength, patient satisfaction (Likert), acetaminophen use

What they found

• Pain (NPS) dominant hand: mean difference between groups -1.30 (95% CI -0.02 to 2.62), P = 0.085 (not significant)
• Pain (NPS) non-dominant hand: mean difference -1.33 (95% CI 0.01-2.65), P = 0.295 (not significant)
• AUSCAN total score: mean difference -9.9 (95% CI 4.07-15.73), P = 0.005, Cohen's d = 0.88 (large effect)
• AUSCAN pain domain: mean difference -3.26 (95% CI 1.06-5.46), P = 0.006, Cohen's d = 0.75 (medium-large effect)
• AUSCAN function domain: mean difference -5.03 (95% CI 1.20-8.86), P = 0.047, Cohen's d = 0.68 (medium effect)
• Cochin scale: mean difference -6.27 (95% CI 0.18-12.36), P = 0.042, Cohen's d = 0.53 (medium effect)
• Grip strength dominant hand: mean difference 3.50 (95% CI -6.03 to -0.97), P = 0.052 (not significant)
• Patient satisfaction at 12 weeks: 76.6% of exercise group rated "better" or "much better" vs 30% of control group, P < 0.001
• Acetaminophen use did not differ significantly between groups (P = 0.628)

Limitations

• No placebo control was possible; both groups received education which may have improved outcomes in the control group
• Sample size (60 patients, nearly all female) is relatively small and limits generalisability, especially to men
• Follow-up ended immediately after the 12-week program with no long-term assessment of sustained benefit
• Elastic band resistance was standardised rather than individualised, potentially underestimating strength gains for some patients

Why it matters

For patients

People with hand osteoarthritis who complete a 12-week supervised resistance exercise program can expect meaningful improvements in how their hands feel during daily activities and overall hand function, even if overall pain scores on a simple scale may not change dramatically.

For clinicians

Supervised progressive resistance hand exercise produces statistically and clinically significant improvements in AUSCAN and Cochin scores with medium-to-large effect sizes, supporting its use as a non-pharmacological treatment option, though the primary pain endpoint (NPS) did not reach significance in this 60-patient trial.

For readers

This is a well-designed single-centre RCT with blinded assessment and intention-to-treat analysis; results are promising for functional outcomes but a larger trial with longer follow-up is needed before definitive conclusions can be drawn.