Temperature seasonality and spatially structured variation in cutaneous leishmaniasis in Algeria: A source-verified national panel, 2000-2024.

Abstract

Algeria carries one of the highest cutaneous leishmaniasis (CL) burdens worldwide, yet the climatic structure of its provincial distribution has not been quantified nationally with year-specific covariates. We assembled a balanced panel of all 48 wilayas over 2000-2024 (1200 province-years), linking notified CL counts to annual ERA5-Land bioclimatic indicators, and verified the case series province-by-province against the primary national epidemiological bulletins. Counts were strongly overdispersed and modelled by negative binomial regression with a log-population offset and province-clustered standard errors, then extended with year fixed effects, a within-between decomposition, and BYM2 spatially structured area effects on a shared-border contiguity graph. Temperature seasonality was the dominant bioclimatic correlate (incidence rate ratio 4.28 per standard deviation, 95% CI 2.91-6.29) and the only term robust across non-spatial specifications, concentrated in the between-province component (7.64, 2.71-21.55) with no detectable within-province association (0.92, 0.77-1.10). Spatially structured area effects left the point estimate positive but its interval spanning the null (1.07, 0.93-1.22), confirmed by an independently coded full-posterior cross-check. Annual precipitation was inversely associated under naive standard errors but did not survive province clustering (p=0.063); precipitation seasonality was null under province-clustered inference, but showed a modest inverse association in its within-province component (0.90, 0.82-0.99) and under spatial adjustment (0.91, 0.83-0.98), an exploratory finding. Incidence declined 3.4% per year (0.966, 0.945-0.987), consistent with the separately curated World Health Organization series. The climatic signal in these data is a persistent geographical gradient that spatial adjustment substantially reshapes rather than an identified driver of year-to-year change.