Abstract
The self-reported Pain Sensitivity Questionnaire (PSQ) is a valid supplement to experimental pain testing. However, the latent constructs determining the originally proposed 1 general score (PSQ-total) and 2 subscores (PSQ-moderate and PSQ-minor) have not been consistently investigated in population-based studies or between genders. Based on a single construct hypothesized by expert knowledge or alternative constructs upon empirical evidence, PSQ structures were explored and confirmed among 4,820 participants aged 18 to 93 years of the Cooperative Health Research In South Tyrol (CHRIS) study. By exploratory factor analysis, we identified 3 alternative sets of PSQ imagined painful situations comprising 14, 10, and 9 items, which displayed simple structures of the rotated factor loadings of direct interpretation. In confirmatory analysis (CFA) of 1 latent factor, the 10-item set yielded acceptable goodness-of-fit overall, better fit than the alternative sets and consistent structural properties between genders. Separate analyses based on 14- and 9-item sets returned considerable correlations between 2 latent constructs. In higher-order CFA with each set, 1 first-order general factor explained a large part of the variances of 2 second-order factors. One dominant construct consistently describes the factorial structure of the PSQ. Averaging across the 10-item set, the PSQ-short score represents a structurally robust, gender-consistent, and practical measure of general pain sensitivity.
PERSPECTIVE:
One dominant latent construct of general pain sensitivity consistently determines responses to the self-reported PSQ. The PSQ-short score maintains similar psychometric properties to the PSQ-total and between genders. This measure is attractive for large-scale research and clinical screening of pain sensitivity