Unraveling legionellosis dynamics in the Lazio Region: effects of individual, geographical and meteorological determinants

Abstract

The incidence rate of Legionnaires’ disease (LDIR) has increased in many countries over the last decade. However, the combined contribution of demographic, behavioural, geographic, and meteorological factors remain poorly understood. This study assessed LDIR trends in Lazio from 2015 to 2023 and association between LDIR and individual, geographic, and meteorological factors in 2017–2023 years. Individual data were obtained from the regional Legionaires’s disease (LD) surveillance database, also collecting smoking/alcohol consumption and comorbidities individual data. Daily rainfall and temperature data were retrieved from the regional hydrometeorological network. Descriptive and annual LDIR per 100,000 person-years were calculated. Associations between LDIR and 8 predictors, selected based on literature and explorative data analysis, were assessed using a negative binomial multivariate regression model, estimating main effects and two-way interactions. On purpose, individual level values of the 6 categorical predictors were aggregated at strata level, obtained by cross-combining these 6 predictors levels at month level. Rainfall and temperature average for each stratum, obtained cross-combining all month and geographical levels, were calculated and linked to the corresponding model stratum. Model offset was calculated applying smoking/alcohol use and comorbidities regional prevalence estimates to the regional population denominators. Between 2015 and 2023, 1,932 LD cases were reported. LDIR rose from 2017 to 2019, decreased during the COVID-19, and increased back in post-COVID-19. In the final regression model age (≥ 70yo: IRR = 17.1; 95%CI: 11.0–27.2; 50-69yo: IRR = 4.68, 95%CI: 3.03–7.43), male sex (IRR = 5.22; 95%CI: 3.61–7.80) and reported comorbidity (IRR = 4.61; 95%CI: 3.41–6.18) were significantly, strongly associated with LDIR. Smoking/alcohol consumption (IRR = 2.05; 95% CI: 1.53–2.75), domicile/residence in Rome metropolitan area (IRR = 2.03; 95% CI: 1.58–2.61), average monthly precipitation (IRR = 1.15; 95% CI: 1.11–1.20) were also significantly, directly associated with LDIR. COVID-19 period was significantly inversely associated with LDIR (IRR = 0.45, 95% CI: 0.32–0.65). Two-way interaction terms COVID-19*monthly average minimum temperature, 50–69yo*smoking/alcohol consumption, Rome metropolitan area*monthly average rainfall had significant IRR > 1; male*50-69yo, male*≥70yo, ≥ 70yo*smoking/alcohol consumption, 50-69yo*comorbidity, ≥ 70yo*comorbidity, COVID-19*Rome metropolitan area post-COVID-19*Rome metropolitan area had significant IRR < 1. These findings underscore the multifactorial nature of LD occurrence and may help inform targeted surveillance and prevention strategies.

The authors' abstract, as published at the source. BMC Public Health, 2026 · DOI ↗