Psychometric evaluation of the SPEED questionnaire: Factor structure and measurement invariance across contact lens wearers and non-wearers, sex and age

Purpose: Dry eye disease (DED) is a multifactorial condition in which patient-reported symptoms play a central role in diagnosis. The Standard Patient Evaluation of Eye Dryness (SPEED) questionnaire is widely used; however, its factorial structure and measurement invariance across multiple clinically relevant groups has not been systematically investigated. Methods: A total of 703 participants were included (234 contact lens (CL) wearers and 469 non-wearers). The psychometric properties of the SPEED questionnaire were evaluated within a structural equation modelling framework. Exploratory factor analysis (EFA) was conducted without a priori assumptions, followed by confirmatory factor analysis (CFA) to compare alternative theory driven models. Measurement invariance across CL status, sex, and age groups was tested at configural, metric, and scalar levels. Latent mean comparisons were performed when appropriate. Results: EFA supported a four-factor solution reflecting the four symptoms (dryness, soreness, burning, and eye fatigue), each assessed in terms of frequency and severity. CFA indicated that a second-order model provided the best fit, with a general factor representing DED symptom severity underlying the four symptom-specific factors. Full measurement invariance was established across CL status, sex, and age. Latent mean comparisons showed no significant differences across CL status and age groups, whereas females reported higher levels of the latent construct than males. Conclusions: The SPEED questionnaire shows a robust second-order structure and operates equivalently across clinical and demographic groups. Clinically, females experienced a greater burden of dry eye symptoms than males. These findings support SPEED use for symptom assessment and highlight the importance of psychometric evaluation in DED research.