Multidimensional Associations Between Psychiatric Symptoms, Sleep Disturbance, Biological Variables, and Fibromyalgia Severity: A Classification and Regression Tree and Path Analysis Study

Abstract

Background: Fibromyalgia (FM) is a complex chronic pain disorder characterized by interactions among psychological, biological, and clinical factors. Although psychiatric symptoms and sleep disturbance are recognized contributors to disease burden, their multidimensional relationships with selected biological variables and fibromyalgia severity remain incompletely understood. Objective: This study aimed to investigate multidimensional associations among psychiatric symptoms, sleep disturbance, selected biological variables, and fibromyalgia severity using complementary analytical approaches, including multiple linear regression, Classification and Regression Tree (CART) analysis, and path analysis. Methods: In this cross-sectional study, 102 women with FM and 53 healthy female controls underwent standardized clinical and laboratory assessment. Fibromyalgia severity was assessed using the Fibromyalgia Impact Questionnaire (FIQ), depressive symptoms with the Beck Depression Inventory (BDI), anxiety symptoms with the Beck Anxiety Inventory (BAI), and sleep disturbance with the Jenkins Sleep Scale (JSS). Serum thyroid-stimulating hormone (TSH) and 25-hydroxyvitamin D concentrations were measured as selected biological variables. Multiple linear regression identified independent associations with disease severity, CART analysis evaluated hierarchical classification patterns, and path analysis examined direct and indirect associations within a predefined conceptual framework. Results: Depressive symptoms, sleep disturbance, and anxiety symptoms were independently associated with greater fibromyalgia severity, whereas TSH and vitamin D showed no consistent independent associations. CART identified sleep disturbance as the primary hierarchical classifier, followed by depressive and anxiety symptoms, achieving an overall classification accuracy of 93.5% (AUC = 0.951). Within the proposed conceptual framework, path analysis demonstrated direct associations between psychological symptoms and fibromyalgia severity, whereas TSH and vitamin D exhibited distinct indirect relationships through depressive and anxiety symptoms, respectively. Conclusions: Psychological symptoms were more consistently associated with fibromyalgia severity than the selected biological variables across complementary analytical approaches. The integrated use of multiple linear regression, CART analysis, and path analysis provided a multidimensional perspective on fibromyalgia and highlighted the importance of comprehensive biopsychosocial assessment while emphasizing that the observed findings represent statistical associations rather than causal relationships.