A note on: The joint effect of maternal malnutrition and cold weather on neonatal mortality in nineteenth-century Venice: An assessment of the hypothermia hypothesis, Population Studies 63(3): 233–251 by Renzo Derosas

Abstract

In a paper published in this journal, Renzo Derosas (2009) presents a study on the determinants of infant mortality in nineteenth-century Veneto in which he uses micro-data to analyse the relationship between maternal malnutrition and the risk of dying during the first weeks of life. Emphasizing that the case of Veneto, a large region in north-eastern Italy, is of particular interest, he says 'Scattered evidence from parish-register studies suggests that infant mortality rates rose from around 250 per thousand in the mid seventeenth century to 350 per thousand at the beginning of the nineteenth, levels unsurpassed throughout the rest of Italy and most European countries' (p. 233). Infant mortality played a key role in the popula tion stagnation of Veneto after its recovery and growth following the last plague of 1630. Indeed, the decline in the likelihood of surviving the first phases of life was one of the main factors driving population stagnation during the century 1730-1830, a stagna tion that was in contrast to the upward trend in population size common to many other European regions. Conversely, the decline in infant mortality from 1830 to 1915 (from 350 to 150 per thousand live births) was the main driver of the consistent population growth and mass emigration before the First World War (Rosina 1995; Rossi and Rosina 1998). Various studies have described this trend, demonstrating that the dramatic rise in infant mortality and its subsequent decline was due to variation in winter neonatal mortality (see, among others, Rosina and Zannini 2004; Rossi and Tesolat 2006). 'How should such a trend be explained?' is the main question raised in Derosas's paper (p. 234). In a recent paper (Dalla-Zuanna and Rosina 2008), we pursued a possible explanation that is employed by Derosas as the starting point for his detailed analysis. Thanks to the availability of data on daily temperature (Camuffo and Jones 2002) we analysed, at the individual level, the association between external temperature and the risk of dying in the first days and weeks of life in several Veneto parishes during the first half of the nineteenth century. We found that the individual risk of dying for the unlucky children born during the winter was tremendously susceptible to external temperature?the most critical period occurred dur ing the first week of life. We used the results obtained from this analysis?together with empirical evidence from the literature and information gath ered from historical documents?to propose the following causal chain as a plausible explanation (our 'pauperization hypothesis'): general worsening in living conditions -> malnourished mothers -+ low-weight newborns more susceptible to neonatal hypothermia -> lower possibility of survival in the first days of life during the winter. The explicit aim of Derosas's study was to test this hypothesis 'of a causal pathway linking maternal malnutrition, low birth weight, susceptibility to cold, and neonatal mortality' (p. 236). To this end, he uses a large longitudinal sample of individuals living in several selected parishes within the city of Venice, the capital of the Veneto region, from 1850 to 1869. Because no direct information on nutritional status