Supplementary material from "Regional differences in historical diphtheria and scarlet fever notification rates in the Netherlands, 1905–1925: a spatial-temporal analysis"

Background. We describe how rates of two frequently occurring notifiable diseases—diphtheria and scarlet fever—varied between regions of the Netherlands in the early twentieth century, and identify potential factors underlying this variation. Methods. Digitized weekly mandatory notification data for 1905–1925, municipality level, were aggregated into 27 ‘spatial units’ defined by unique combinations of province and population density category (high: >4500; mid : 1250–4500; low:less than 1250 inhabitants per km²). Generalized additive regression models were fitted to estimate the associations between notification rates and population density, infant mortality rate and household income, while adjusting for temporal trends per spatial unit. Results. Annual per capita notification rates for both diphtheria and scarlet fever tended to rise from the beginning of the period 1905–1925 until peaking around 1918/19. Adjusted diphtheria notification rates were higher for high- and mid- compared with low-density municipalities (by 71.6 cases per 100 000, 95% CI : 52.7–90.5; 39.0/100 k, 95% CI : 24.7–53.3, respectively). Scarlet fever showed similar associations with population density (35.7/100 k, 95% CI : 9.4–62.0; 21.4/100 k, 95% CI: 1.5–41.3). Conclusions. There was considerable spatial variation in notification rates for both diseases in early twentieth century Netherlands, which could partly be explained by factors capturing variation in living conditions and socio-economic circumstances. These findings aid understanding of contemporary respiratory infection transmission.