Relationship between tendon tissue and shoulder disability change during an 8-week exercise intervention for rotator cuff tendinopathy: an observational study

Key points

1. 01Shoulder disability (Penn Score) improved significantly at 2, 4, and 8 weeks, reaching a clinically meaningful difference only at week 8 (mean +17.4 points).
2. 02Tendon thickness decreased significantly at all time points but changes were smaller than the minimal detectable change of 0.3 mm.
3. 03A 1-mm decrease in tendon thickness at week 2 was associated with a 14.4-point improvement in the Penn Score (p=0.015), but this relationship was not present at weeks 4 or 8.
4. 04Internal tendon architecture (peak spatial frequency radius) did not change significantly at any time point over 8 weeks.
5. 05Lower baseline Penn Score was the strongest predictor of greater improvement, consistent with regression-to-the-mean effects.

How it was conducted

Design

Prospective longitudinal observational cohort study

Participants

47 adults (age 18-55) with unilateral rotator cuff tendinopathy confirmed by 3 of 5 positive clinical tests and symptoms for at least 3 months

Intervention

8-week progressive resistive exercise program, 5 sessions per week, with twice-weekly supervision by a licensed physical therapist

Primary outcomes

Pennsylvania Shoulder Score (patient-reported disability), supraspinatus tendon thickness, and internal tendon architecture (peak spatial frequency radius) via ultrasound at baseline, 2, 4, and 8 weeks

Analysis

Linear mixed-effects models with participant-level random intercept; leave-one-out cross-validation for predictive accuracy

What they found

• Penn Score improved from baseline: +6.3 points (SE=1.2, p<0.001) at week 2, +11.2 points (SE=1.2, p<0.001) at week 4, and +17.4 points (SE=1.2, p<0.001) at week 8; minimum clinically important difference of 11.4 points was reached only at week 8.
• Tendon thickness decreased significantly: -0.09 mm (SE=0.03, p<0.001) at week 2, -0.16 mm (SE=0.03, p<0.001) at week 4, and -0.20 mm (SE=0.03, p<0.001) at week 8; all changes were below the MDC of 0.3 mm.
• Internal tendon architecture (PSFR) did not change significantly at any time point.
• For every 1-mm decrease in tendon thickness at week 2, the Penn Score increased by 14.44 points (SE=5.88, p=0.015); RMSE=5.6 points.
• Tendon thickness and PSFR changes were not significantly associated with Penn Score changes at weeks 4 and 8 (week 4 thickness: estimate=12.09, p=0.11; week 8 thickness: estimate=2.17, p=0.73).
• Baseline Penn Score was a significant predictor of change: lower baseline scores predicted greater improvement (estimate=-0.17, SE=0.06, p=0.004).
• Model R for the involved-side model was 0.17, indicating it explained a small proportion of variance in Penn Score changes.
• Bilateral tendon thickness difference models showed no significant interactions or main effects for tendon variables.
• Exercise adherence was high: mean 3 missed sessions (SD=3.3) out of 40 total sessions; 40 of 47 participants completed all visits.

Limitations

• Single-cohort design with no control group limits ability to determine whether tendon changes drive clinical improvement or are an unrelated consequence of exercise.
• Only one intervention (resistive exercise) was studied, so findings may not generalise to other treatments such as corticosteroid injection or shockwave therapy.
• The sensitivity of the peak spatial frequency radius to detect small architectural changes over 8 weeks is not well established, so real tissue changes may have been missed.
• No measures of systemic inflammatory markers (e.g., interleukin-6, TNF-alpha) were collected, leaving open the role of systemic inflammation in outcomes.

Why it matters

For patients

Most people with rotator cuff tendinopathy can expect meaningful shoulder pain and function improvement from an 8-week exercise program, though the tendon itself changes only modestly during this time.

For clinicians

Tendon thickness reduction at 2 weeks may be a useful early signal of response, but tendon morphology alone explains very little of the overall disability change and should not be the primary treatment target.

For readers

This study fills a gap by linking ultrasound tendon changes to patient outcomes over time, and points to the need for studies that also measure neuromuscular, sensorimotor, and systemic inflammatory factors.