Jaime Angamarca-Iguago, Jaen Cagua-Ordoñez, Juan Marcos Parise-Vasco, Daniel Simancas-Racines
Among 74,601 dengue hospitalisations with a resolvable canton of residence (232 deaths; case fatality 0.31%), dengue was recorded in 68 cantons above 1500 m, including altitudes exceeding 3000 m. The national hospitalisation rate rose from 29.7 to 77.0 per 100,000 and the Andean rate approximately 7-fold (1.9 to 12.8). The minimum hospitalisation temperature was 18.4°C nationally and 18.9°C in Andean cantons. A +1°C anomaly versus zero was associated with a 1.07-fold increase in hospitalisation rates (95% CI 1.02-1.13; p = 0.012). Without canton fixed effects, the same contrast yielded 1.21 (1.07-1.36); the within-canton association fell to 1.04 (1.00-1.08) when the 2024 epidemic year was excluded.
INTRODUCTION: Dengue is expanding beyond its traditional lowland range as rising temperatures may enable Aedes aegypti establishment at progressively higher altitudes. Ecuador, spanning sea level to above 4500 m, offers a natural altitudinal experiment, yet hospital-based evidence on the thermal drivers of Andean dengue is lacking.
METHODS: We conducted a national ecological time-series analysis at the canton level using Ecuador's hospital discharge registry (INEC) for 2011-2024, the period with standardised canton coding, identifying dengue hospitalisations by ICD-10 codes A90, A91 and A97. Monthly land surface temperature, precipitation and vegetation index were derived from MODIS and GPM products. A distributed lag non-linear model estimated the temperature-hospitalisation association; a quasi-Poisson model restricted to Andean cantons (≥1500 m), with canton, calendar-month and year fixed effects, tested whether thermal anomalies above each canton's 2007-2015 baseline were associated with hospitalisation rates.
RESULTS: Among 74,601 dengue hospitalisations with a resolvable canton of residence (232 deaths; case fatality 0.31%), dengue was recorded in 68 cantons above 1500 m, including altitudes exceeding 3000 m. The national hospitalisation rate rose from 29.7 to 77.0 per 100,000 and the Andean rate approximately 7-fold (1.9 to 12.8). The minimum hospitalisation temperature was 18.4°C nationally and 18.9°C in Andean cantons. A +1°C anomaly versus zero was associated with a 1.07-fold increase in hospitalisation rates (95% CI 1.02-1.13; p = 0.012). Without canton fixed effects, the same contrast yielded 1.21 (1.07-1.36); the within-canton association fell to 1.04 (1.00-1.08) when the 2024 epidemic year was excluded.
DISCUSSION: Positive deviations from local historical baselines, rather than absolute temperature, were associated with rising dengue hospitalisation rates among Andean residents, by a modest margin that depends materially on the 2023-2024 epidemic. In-hospital case fatality was nearly twice as high in low Andean cantons (0.62%, 95% CI 0.20-1.45; five deaths) as on the coast (0.33%), a non-significant difference that nonetheless supports strengthening dengue recognition in highland facilities.