Associations between meteorological conditions and Legionnaires’ disease reported incidence in Greece: a 17-year time-series analysis

The reported incidence of Legionnaires’ disease (LD) has been increasing globally, and meteorological conditions have been proposed as important contributors. Evidence from Greece remains limited. This study aimed to investigate the associations between meteorological exposures and LD incidence at national and regional levels. We conducted a weekly time-series analysis of sporadic LD cases exposed in Greece between 2008 and 2024, using data from the National Mandatory Notification System and the European Legionnaires’ Disease Surveillance Network. We excluded healthcare-associated cases, those exposed abroad, and cases exposed in multiple regions. We obtained temperature, precipitation, relative humidity, and wind speed data from the Multi-Source Weather dataset and resident and tourism population data from the Hellenic Statistical Authority. We grouped regions into 4 climatic zones. We used negative binomial regression models to assess associations at national and regional levels, and distributed lag non-linear models (DLNMs) with a maximum lag of 6 weeks to explore potential non-linear and delayed associations. The analysis included 1180 cases. LD incidence increased significantly nationwide over time (IRR per year = 1.08, 95%CI:1.06–1.10, p < 0.001) and showed strong summer seasonality. In multivariable analyses, each 1-mm increase in weekly precipitation was associated with increased LD incidence (IRR = 1.10, 95%CI:1.04–1.15, p < 0.001), and a significant temperature-relative humidity interaction was observed (IRR = 1.005, 95%CI: 1.003–1.007, p < 0.001). The temperature-LD incidence association was strongest at moderate relative humidity (65–73%), with increased risk at 17–24 °C. DLNMs showed a non-linear association of temperature with on LD incidence, with significant temporal delay, peaking at 6 weeks before exposure (30 °C, IRR:1.80, 95%CI:1.12–2.87). Regionally, the temperature-humidity interaction was stronger in zones 2 (West) and 4 (South). Rainy weather and specific temperature-humidity combinations were associated with LD incidence in Greece, with regional variation in the temperature-humidity association. Warm temperatures exhibited delayed associations for up to 6 weeks. In the context of a changing climate, these findings highlight the value of sub-national analyses and may inform targeted prevention during high-risk periods.