Meteorological factors and primary care consultation patterns in rheumatoid arthritis and axial spondyloarthritis: a 14-year population-based time-series study.

BACKGROUND: Weather-related symptom fluctuations are frequently reported in patients with inflammatory and rheumatic diseases. However, evidence regarding the influence of meteorological conditions on healthcare utilization remains inconsistent, and long-term population-based studies are scarce. OBJECTIVE: To investigate the association between meteorological variables and primary care consultation frequency for rheumatoid arthritis (RA) and axial spondyloarthritis (axSpA) over a 14-year period and evaluate the influence of demographic factors on healthcare-seeking behavior. METHODS: A retrospective population-based time-series study was conducted using anonymized electronic health records from three primary care centers in the Community of Madrid (Spain) between 2010 and 2023. Weekly consultations for RA and axSpA were identified using the International Classification of Primary Care, Second Edition (ICPC-2) diagnostic codes. Meteorological data obtained from the Spanish State Meteorological Agency included temperature-related variables, precipitation, wind characteristics, sunshine duration, and barometric pressure. Associations were analyzed using generalized additive models (GAMs) with a Poisson distribution, accounting for repeated observations and temporal trends. RESULTS: Among 1,175 unique patients (mean age, 60.55 ± 17.03 years; 73.11% female), 1,677 consultations were analyzed: 1,461 (87.12%) for RA and 216 (12.88%) for axSpA. Greater average rainfall was associated with lower consultation frequency in RA (p = 0.010) and the overall cohort (p = 0.009). Longer sunshine duration was associated with lower consultation frequency in axSpA (p = 0.027) and overall (p = 0.018), whereas higher average temperature was associated with higher consultation frequency in axSpA (p = 0.041). Wind direction showed nonlinear associations in RA (p = 0.004) and overall (p = 0.022), but not in axSpA (p = 0.058). Age and sex were not independently associated with consultation frequency. CONCLUSIONS: Meteorological associations with healthcare utilization were selective and diagnosis-specific. These findings concern consultation patterns, do not establish corresponding changes in disease activity or causality, and indicate that weather alone is insufficient to predict healthcare demand. Key Points • In this 14-year population-based analysis, most meteorological variables were not significantly associated with primary care consultation frequency for rheumatoid arthritis and axial spondyloarthritis, and neither sex nor age was independently associated with consultation frequency after adjustment for meteorological variables, patient-level clustering, and temporal trends. • Greater rainfall was associated with lower consultation frequency in rheumatoid arthritis and the overall cohort, while wind direction showed significant nonlinear associations in these two models. • Longer sunshine duration was associated with lower consultation frequency in axial spondyloarthritis and the overall cohort, whereas higher average temperature was associated with higher consultation frequency only in axial spondyloarthritis. • Meteorological associations with healthcare utilization were selective and diagnosis-specific; therefore, weather conditions alone appear insufficient to explain or predict primary care demand for these diseases.