Dimensionality and Reliability of the Central Sensitization Inventory in a Pooled Multicountry Sample

Abstract

Central sensitization (CS) involves the amplification of neural signaling within the central nervous system, which evokes pain hypersensitivity. The Central Sensitization Inventory (CSI) assesses 25 overlapping health-related symptom dimensions that have been reported to be associated with CS-related disorders. Previous studies have reported satisfactory test-retest reliability and internal consistency, but factor analyses have exhibited conflicting results in different language versions. The purpose of this cross-sectional study was to thoroughly examine the dimensionality and reliability of the CSI, with pooled data from 1,987 individuals, collected in several countries. The principal component analysis suggested that 1 general factor of CS best described the structure. A subsequent confirmatory factor analysis revealed that a bifactor model, which accounted for the covariance among CSI items, with regard to 1 general factor and 4 orthogonal factors, fit the CSI structure better than the unidimensional and the 4-factor models. Additional analyses indicated substantial reliability for the general factor (ie, Cronbach α = .92; ω = .95; and ω hierarchical = .89). Reliability results for the 4 specific factors were considered too low to be used for subscales. The results of this study clearly suggest that only total CSI scores should be used and reported.