Pathogen seasonality and links with weather in England and Wales: a big data time series analysis

Cherrie, Mark P. C.; Nichols, Gordon; Iacono, Gianni Lo; Sarran, Christophe; Hajat, Shakoor; Fleming, Lora E.

Salmonella serotypes featured heavily with varying strength and pattern of seasonality detected. Salmonella Enteritidis and Salmonella Typhimurium had the strongest associations with meteorological variables. The remaining Salmonella serotypes were split between being weakly correlated (n = 15) and very weakly correlated (n = 8). There is some reason to believe that the epidemiological causes of seasonality in most Salmonellas is similar (24/25; 96% belong to Group 1) and the association with temperature might be linked to growth in prepared foods. In addition, the strength of association in linking the seasonality or temperature to cases will be limited to the number of isolates in each serogroup. Because of this the salmonellas were grouped into four groups (1. Salmonellas causing enteric fever that are usually acquired overseas (S. Typhi/S. Paratyphi); 2. Seasonal salmonellas; 3. Strains showing no evidence of any seasonality and 4. The remaining strains where there are insufficient numbers to determine seasonality). The remaining strains included serotypes that had so few isolates that seasonality could not be determined. When grouped thus, the seasonality of the seasonal salmonellas (2) resembled that of the remaining strains (4), while the overall seasonality of serotypes that individually showed little evidence of seasonality were not obviously seasonal when combined (Fig. 5) . The seasonality of groups 2 and 4 showed a high degree of correlation using data averaged over the 25-year period (r 2 = 0.98; Fig. 5b ).