Seasonal components of infant mortality at the onset of the transition reveal the role of water-borne and air-borne diseases: the case of the Don Army Territory (Southern Russia), 1872–1915

Seasonal components of infant probabilities of dying are disentangled from monthly death statistics by age and birth by articulating demographic equations and stochastic optimization. In the Don Army Territory, for the period 1872–1915, these components reflect respiratory diseases in autumn and spring, dehydration and waterborne diseases in summer, and cold stress in winter. During the warmer months, they were lower in cities. Summer heat had a lethal effect, mitigated by precipitation; illegitimate infants were exposed to the change of season; high winds increased mortality in late winter for 0–5 month infants and in summer for 6–11 month infants in cities. By the turn of the century, mortality had decreased, thanks to the gradual purification of water supplies and improved health practices.