Feed

Background: Although the concepts of meteoropathy and meteosensitivity are not included in official classifications, such as the ICD-11 or DSM-5, they are increasingly being studied as potential symptom complexes linking weather variability to health status. The METEO-Q questionnaire, originally developed in Italy, has been adapted in Japan and Turkey, where it has demonstrated satisfactory reliability parameters, although the authors emphasized the need for further verification of the tool's temporal stability. The present study aimed to adapt METEO-Q to the Polish language and conduct a critical assessment of its factor structure, measurement invariance, and validity in clinical groups. Methods: This cross-sectional study involved 1128 adults: healthy individuals (n = 711), cardiac outpatients (n = 194), and subclinical group with diagnosed mental disorders (n = 223). Data from healthy participants were divided into a training sample (n = 426) for exploratory factor analysis (EFA) and a test sample (n = 285) for confirmatory factor analysis (CFA). Measurement invariance was assessed in the clinical groups. Validity was verified through correlations with a list of 21 symptoms and measures of anxiety and worry about climate change. Results: A two-factor model (meteoropathy and meteosensitivity) was better fitted to the data than a one-factor model, which is consistent with findings from Italian, Japanese, and Turkish studies. However, absolute fit indices in the test sample indicated significant model misfit [CFA: χ² (43) = 210.192, p < 0.001, RMSEA = 0.120, CFI = 0.927], suggesting the presence of local errors in the tool's structure. The reliability of the subscales was high (α from 0.86 to 0.93). Multi-group analyses suggested metric and scalar invariance. Patients with mental disorders obtained the highest scores, while cardiac outpatients reported a lower level of meteoropathy (M = 6.13) than healthy individuals (M = 7.24). Conclusions: METEO-Q demonstrates a stable two-factor structure and high internal consistency. The obtained RMSEA index (0.12), although indicative of some misfit, is similar to results obtained in other adaptations, such as the Japanese (RMSEA = 0.10) and the Turkish (RMSEA = 0.11), which suggests it is a consistent feature of this tool across different cultural contexts. Accordingly, the instrument is suitable for research purposes; however, its clinical application requires considerable caution and further work to optimize the model.