Pain-related beliefs are associated with arm function in persons with frozen shoulder

Key points

1. 01Pain self-efficacy had the strongest association with arm function (r = -0.69, p < .0001), meaning lower confidence in doing tasks despite pain linked to worse disability.
2. 02Pain-related fear (kinesiophobia) also showed a significant moderate association with arm disability (r = 0.51, p < .0001).
3. 03Pain catastrophizing was significantly correlated with arm function (r = 0.45, p < .0001) but did not independently explain variance in the regression model.
4. 04Pain-related beliefs explained an additional 26% of variance in arm function beyond what pain intensity alone explained.
5. 05Average pain catastrophizing scores were low (mean 13.8 out of 52), suggesting this sample did not generally have extreme catastrophic thinking about pain.

How it was conducted

Design

Cross-sectional observational study

Participants

85 adults with clinically diagnosed frozen shoulder (69% women; mean age 55.2 years)

Recruitment

Belgium via orthopedic surgeons, GPs, and physiotherapy practices, February 2017 to December 2018

Primary outcome

Perceived arm function measured by the Disabilities of the Arm, Shoulder and Hand Questionnaire (DASH, 0-100)

Predictors assessed

Pain intensity (NRS), pain-related fear (TSK-11), pain catastrophizing (PCS), pain self-efficacy (PSEQ)

Analysis

Pearson/Spearman correlations plus hierarchical multiple linear regression

What they found

• Pain self-efficacy (PSEQ) correlated with DASH at r = -0.69, p < .0001 (moderate association).
• Pain-related fear (TSK-11) correlated with DASH at r = 0.51, p < .0001 (moderate association).
• Pain catastrophizing (PCS) correlated with DASH at r = 0.45, p < .0001 (weak association).
• Control variables alone (sex, age, pain intensity, duration) explained 31% of variance in arm function (R2 adj = 0.31); pain intensity was the only significant control variable.
• Adding TSK-11, PCS and PSEQ scores increased explained variance to 57% (R2 adj = 0.57), an additional 26% beyond controls.
• In the full model, significant independent contributors were pain intensity (std Beta = 0.29, p = 0.0005), TSK-11 (std Beta = 0.23, p = 0.01), and PSEQ (std Beta = -0.50, p < .0001); PCS was not significant (p = 0.35).
• Sensitivity check removing overlapping PSEQ items reduced R2 to 55% with the same standard Beta for PSEQ, indicating limited item-overlap inflation.
• Mean DASH score was 41.1 (SD 18.7); mean pain intensity 5.6/10 (SD 2.2); mean TSK-11 24.2 (SD 6.1); mean PSEQ 46.9 (SD 11.9); mean PCS 13.8 (SD 9.8).

Limitations

• Cross-sectional design prevents causal conclusions about whether pain-related beliefs drive disability or vice versa.
• Arm function was measured only by self-report (DASH); perceived and actual physical function can diverge, and the absence of physical performance measures may overestimate the role of pain beliefs.
• No gold-standard diagnostic test for frozen shoulder exists, so misclassification of some participants cannot be excluded, although strict clinical criteria were applied.
• The sample was a convenience sample of Caucasian Belgians, limiting generalizability to other populations.

Why it matters

For patients

People with frozen shoulder who feel fearful of movement or lack confidence in using their arm despite pain may experience more disability, suggesting that addressing these beliefs through education or psychological support could help recovery.

For clinicians

Routine assessment of frozen shoulder should include validated screening for pain-related fear and pain self-efficacy, as these independently predict perceived arm disability and may be targets for biopsychosocial management strategies.

For readers

This is the first study to quantify the role of psychological pain beliefs in frozen shoulder specifically, filling a gap left by prior shoulder-pain research that excluded this diagnosis.