Illness Coherence, Exercise Adherence, and Patient-Reported Outcomes in Chronic Spinal Pain: A Moderated Mediation Analysis

Abstract

Background/Objectives: Illness coherence has been associated with treatment engagement and adherence, but its relationship with exercise adherence and patient-reported outcomes in chronic spinal pain remains unclear. This study examined whether unsupervised exercise adherence mediates the association between baseline illness coherence and patient-reported outcomes and whether this indirect association differs according to the type of intervention received. Methods: A theory-driven secondary analysis was conducted using data from a randomized controlled trial of adults with chronic spinal pain receiving either exercise-only or combined exercise plus diaphragmatic breathing intervention (initially randomized = 183; n = 57/group). Measures included Revised Illness Perception Questionnaire (IPQ-R), Brief Pain Inventory–Short Form (BPI-SF), Hospital Anxiety and Depression Scale (HADS), World Health Organization Quality of Life Instrument, Brief Version (WHOQOL-BREF), and exercise logs. Moderated mediation analyses examined whether the intervention group moderated the indirect associations between baseline illness coherence and patient-reported outcomes through unsupervised exercise adherence. The parent trial was retrospectively registered at ClinicalTrials.gov (NCT07539116) on 12 April 2026. Results: Mean age was 46.95 ± 14.54 and 46.32 ± 15.97 years, and women comprised 40.4% and 47.4% of the exercise-only and combined groups, respectively. Baseline illness coherence was positively associated with unsupervised exercise adherence in the exercise-only group (b = 71.04, p = 0.018). Significant conditional indirect effects of illness coherence on all patient-reported outcomes through unsupervised exercise adherence were observed in the exercise-only group (b = −1.55 to 1.32, ps < 0.05), but not in the combined intervention group (b = −0.59 to 0.32, ps > 0.05). Indices of moderated mediation were significant across all outcomes (all ps < 0.05). No intervention-related adverse events were reported. Conclusions: Diaphragmatic breathing may reduce the relative contribution of illness-related cognitive representations on patient-reported outcomes by introducing additional self-regulatory pathways through which clinical improvement is achieved.