Factor structure and psychometric validation of the antenatal exercise knowledge, attitudes, and practices questionnaire

Assessing pregnant women’s knowledge, attitudes, and practices (KAP) vis-à-vis antenatal exercise is essential for designing targeted interventions that improve maternal and fetal outcomes. The previously published KAP questionnaire has been used in Saudi Arabia, but its latent structures and psychometric properties require thorough examination. The aim of this study was to perform an extended exploratory factor analysis and detailed psychometric evaluation of the antenatal exercise KAP questionnaire, including assessment of factorability, dimensionality, reliability, and item-level psychometric performance. We conducted a secondary analysis on item-level data from 118 pregnant women. In addition to principal-axis factoring with Varimax rotation, we applied parallel analysis for factor retention and assessed item communalities, cross-loadings, and measurement invariance across primiparous and multiparous subgroups. Internal consistency was evaluated using Cronbach’s α, item-total correlations, and McDonald ω. All scales met adequacy criteria (knowledge Kaiser–Meyer–Olkin [KMO] = 0.756; attitude KMO = 0.674; practice KMO = 0.631; Bartlett P < .001). Parallel analyses and scree plots suggested unidimensionality for each domain. The first factor explained 78.6% (knowledge), 94.1% (attitude), and 95.8% (practice) of the variance. Communalities were uniformly high (>0.50), and no substantial cross-loadings (>0.30) were observed. Cronbach’s α (knowledge = 0.991; attitude = 0.996; practice = 0.990) and McDonald ω (0.989–0.995) corroborated high reliability. Measurement invariance testing indicated configural and metric invariance across parity groups. The antenatal exercise KAP questionnaire showed preliminary evidence of a unidimensional structure and high internal consistency in this sample. Given the modest sample size relative to the number of items, particularly in the knowledge domain, confirmation in larger independent samples is warranted.