Principal component analysis of the biopsychosocial features relevant to temporomandibular disorders.

Abstract

This study aimed to explore the dimensionality of three biopsychosocial constructs with multiple subdimensions, including the Symptom Checklist-90-Revised (SCL-90R), pain catastrophizing scale (PCS) and Pittsburgh Sleep Quality Index (PSQI), and to identify the latent dimensions of five biopsychosocial constructs (SCL-90R, PCS, PSQI, pain severity, and pain interference) using the principal component analysis (PCA) in patients with temporomandibular disorder (TMD). A secondary analysis of a previous cross-sectional study comprising 1488 patients with painful TMD was conducted using multiple questionnaires. PCA of the SCL-90R and PCS identified one factor, which explained 60.8% and 80.2% of the total variance, respectively. For the PSQI, three factors explained 61.3% of the variance. PCA resulted in two main orthogonal components: factor 1, which comprised a combination of scores for pain severity, pain interference and global scores of PCS, and PSQI; and factor 2, which comprised one measure of the SCL-90R. Factors 1 (46.5%) and 2 (20.0%) explained 66.5% of the total variance. The findings of this study revealed that five measures can be primarily categorised into two latent constructs of the psychological (affective) and pain-related (sensory-cognitive) dimensions. These core components could be applied in clinical settings and for research purposes.