Individual-level determinants of death during the ‘Spanish flu’: the case of Belgium

The 1918-1919 ‘Spanish flu’ pandemic caused very high morbidity and mortality worldwide, with Belgium recording about 50,000 excess deaths (Devos et al. 2021). This paper aims to identify the socio-demographic factors influencing mortality during this period, utilizing individual-level data from various Belgian municipalities. Prior research on this pandemic suggested an uneven impact across different regions, with rural areas in eastern and southern Belgium showing higher excess mortality for women, and industrial regions showing higher male mortality (Bourguignon et al, forthcoming). This paper seeks to understand how individual characteristics and regional differences contributed to these different pandemic mortality risks. Employing a comparative approach with individual-level death certificates from approximately 60 Belgian municipalities, we examine deaths across various residential and military contexts affected by the Spanish flu. These municipalities are representative of the different socioeconomic environments (large towns, medium-sized towns, industrial areas, rural municipalities, etc.) and different military regimes that characterized Belgium in 1918, as well as the differing impact of the Spanish flu (areas of low and high excess mortality). Our analysis spans two timeframes: the peak of the flu pandemic in 1918-1919 and the pre-pandemic years of 1912-1913, serving as a baseline. With around 70,000 deaths in the database, we employ bivariate and multivariate analyses, including regression models, to compare mortality by age, sex, residence, and socio-professional status during these two periods.