Factors influencing infant mortality in western Europe during the eighteenth and nineteenth centurie
“Twenty years ago our knowledge of the decline in western mortality was fairly rudimentary” (Tedebrand, 1988, p7). Since then historical demographers have religiously reconstructed families from pre-1837 parish registers and collected post-1837 data from the census/civil registration to examine population characteristics for a variety of western countries. The results have shown that up until the end of the nineteenth century mortality throughout the whole of Europe was extremely high. They have also highlighted the enormous contribution infant deaths have made to this total. Consequently, much research has been done into the determinants of infant mortality as an important mechanism in demographic theory.
This essay attempts to review some of the influential factors identified in the infant mortality literature. No attempt is made to distinguish between those that explain inter- and intra-regional trends or those that supposedly initiate widespread change. Instead, all are included on the basis that they have had some effect on infant mortality. It is hoped that approaching the subject in this way will allow the diversity and interrelationships that exist between them to shine through and that, in the process, monocausal explanations will be proved deficient.
One of the most influential, and controversial, theories regarding infant mortality was posited by Thomas McKeown, an eminent British historical demographer, in 1976 with his book ‘The Modern Rise of Population’. Central to his text was the thesis that advances in material conditions, especially nutrition due to greater food supplies and possibly housing standards, was the variable that best explained (infant) mortality changes over the last three centuries since infants with a healthy nutritional level acquired a greater resistance to disease (McKeown, 1976, Corsini and Viazzo, 1993). In other words, his theory stressed economic factors at the expense of all others.
Whilst undoubtedly radical, this thesis was influential in the sense that its monocausality prompted a reaction in demographic circles that generated voluminous literature on a multitude of alternative, and other similar, factors that influence infant mortality (Tedebrand, 1988, Corsini and Viazzo, 1993).
Turpeinen (1979) in a study of infant mortality in Finland between 1749 and 1865 essentially provided a critique of McKeown’s thesis. Comparing infant mortality with fluctuations in crop yield for predominantly agrarian Finland represented a fairly accurate gauge of how access to food (and therefore nutrition) affected infant mortality. Some correlation between the two was found; for instance in 1755, 1807 and 1832 infant mortality increased following poor harvests (Turpeinen, 1979). However, in 1769 and 1772, also after poor harvests, infant mortality was found to decrease; conversely, infant mortality sometimes had a tendency to increase in years with good crops (Turpeinen, 1979). The conclusion reached was that,
“fluctuations in crop yield and in infant mortality did not correlate positively, and the relationship between infant mortality and economic factors may not be as self-evident as is often claimed” (Turpeinen, 1979, p5).
Pre-dating both McKeown and Turpeinen’s work was Beaver (1973) who, studying England and Wales around the turn of the century, also refuted the monocausal link between economics and infant mortality since a rise in material living conditions could not explain the sudden and continued reduction in infant mortality they experienced around 1900. The factor he believed to correspond with the decline was an improvement in the safety and availability of substitute infant foods (eg. dried milk etc.) (Beaver, 1973). Most importantly this involved the increased public availability of pasteurised cows milk as prices fell with a wholesale agricultural shift from cattle to dairy farming (Beaver, 1973).
Whilst this theory places great emphasis on progressive public health issues, this factor is undoubtedly intertwined with pure economics (the supply and demand of cows milk and farmers falling rate of profit etc.) to the extent that it is difficult to decipher precisely which is predominant. Had farmers not switched to dairy production, cows milk would not have become available to the populace. Yet without pasteurisation, the same cows milk would have passed on dangerous pathogens to infants and increased the respective mortality rate. Woods et al. (1993) essentially provide a critique of this time-specific analysis of infant mortality decline in England and Wales though, believing the plateau of approximately 150 infant deaths per 1000 maintained throughout the nineteenth century is a composite of complex patterns that require the adoption of a long time perspective to fully understand.
However, the important role milk, in particular a mothers breastmilk, has to play in the determination of infant mortality rates has been recognised in many studies. Woods et al. (1993) acknowledge that the practice of breastfeeding does have a significant impact; its widespread use keeping infant mortality rates in the UK during the nineteenth century comparably low even when levels of urbanisation, in European terms, were unusually high. As does Turpeinen (1979) who concludes that “evidence tentatively suggests that a decline in infant mortality may be related to some extent with an increase in the practice of breastfeeding” (ibid., p14). Also, extensive studies of foundlings, who miss the antibodies passed from mother to child in breastmilk being artificially fed, find exceptionally high infant mortality rates which obviously supports this thesis (Williams, 1995).
Brändström (1993) illustrates the relationship between breastfeeding and infant mortality by a comparison of two maps showing mid-nineteenth century data for districts in Sweden (figures 1 and 2).
Whilst the errors arising from a comparison between deaneries and districts over different time periods are considerable, these maps do serve the purpose of showing several obvious similarities between breastfeeding and infant mortality. For instance, the coastal region of northern Sweden is seen to have been characterised by high infant
Figure 1. Figure 2.
Breastfeeding in Swedish Infant Mortality
Provincial Physicians Districts in Swedish Deaneries
1869-1874. 1855-1856.
Source: Brändström, 1993, p31 and p25.
mortality and low levels of breastfeeding and “the same pattern appears in the county of Västerbotten, in the Lake Mälar region and in southern Sweden” (Brändström, 1993, p30). However, in addition to this Brändström (1993) notes that contradictory patterns can also be found in some areas of Sweden.
A similar pattern is described by Kintner (1987) using 1910 data for Germany. In this instance, as with Brändström (1993), a high degree of correlation was found between low levels of infant mortality and high levels of breastfeeding in some cases. Kintner (1987) cites the examples of Bavaria in southern Germany where infants were never breastfed or weaned early and infant mortality was extremely high (sometimes 276 per 1000) and rural areas of north-west Germany where the prevalence of breastfeeding was high and infant mortality was low (78 per 1000). Once again however, significant exceptions to the pattern existed. Large urban centres such as Hamburg, Bremen and Berlin with substantial proportions of their population never having been breastfed had only low to moderate levels of infant mortality (Kintner, 1987). The existence of such exceptions in both studies thus suggests that other variables also have an important part to play in infant mortality rates.
Fertility is one of the variables generally regarded as having an important effect on infant mortality, though it is widely appreciated that the precise relationship is very difficult to quantify. Hobcraft et al. (1983) in a study of the deleterious effects of rapid childbearing make reference to Jelliffe’s (1966) ‘maternal depletion syndrome’. In it, through repeated pregnancies and short birth intervals, mothers are noted as having little time for physical and nutritional recovery. This results in a greater chance of miscarriage, stillbirth and low birth weight babies that have a lower chance of survival as competition for food resources increases and infants are weaned from the breast too soon (Hobcraft et al., 1983). Yet closely intertwined with this is propensity to breastfeed, since deliberately cutting its duration not only increases the direct risks to the infant but also means that the conception rate increases with the earlier resumption of ovulation (Hobcraft et al., 1985). Thus, once again it is difficult to determine exactly which is the decisive factor.
However, the alternative variable most widely cited in literature pertains to female labour obligations, though differing levels of importance and relevance have been attached to it (Williams, 1995). Lithell (1988) studying Ostrobothnia in Finland in the eighteenth and nineteenth centuries found women’s severe life conditions (hard labour in the fields and at home combined with raising children and looking after her husband etc.) explained the regions extremely high incidence of infant mortality. A women’s busy and demanding schedule meant: breastfeeding patterns were forced to change, infants did not receive adequate care and attention and consequently poor nutritional states prevailed in mother and child (Lithell, 1988). Woods et al. (1993) also emphasised the role female agricultural employment played in keeping infant mortality in rural districts of eastern England at a level comparable to northern manufacturing towns up until the 1870s.
Brändström’s study of Jokkmokk and Nedertorneå in Sweden during the nineteenth century (1988) also partly supported this theory. He linked the high levels of infant mortality amongst the nomadic settlers of Jokkmokk (which sometimes reached 250 per 1000 for neonatal mortality, almost twice that of the settled population) to the extreme seasonality of their work rhythms. Rather than in winter when the nomads moved with their herds of reindeer, infant mortality was concentrated during the labour intensive summer months when women could not properly care for their children (Brändström, 1988, Brändström, 1993);
“despite…the nomadic children [receiving] the breast up until their third or fourth year…the influence of hard and highly seasonal female labour on mortality seems to have been substantial” (Brändström, 1988, p355).
Contrasting with this however was the data for Nedertorneå in northern Sweden, a rural area with an infant mortality rate high enough to concern the mercantilist government of the time (Brändström, 1993). An inquiry was launched in the early 1800s and it was found that the highest rates of infant mortality (indeed those that pushed up the yearly average) occurred during the summer when the regional custom of giving infants cows milk soured, insects laid eggs in the horns used as bottles and the mothers care of her children declined whilst she worked outdoors;
“The raw, unboiled cows’ milk was difficult to keep fresh in the hot summer months from June to August. During this period of the year the workload for women was also at its peak. The cattle had to be looked after…and both hay and barley had to be harvested. Time available to take care of the newly born was probably then at its lowest” (Brändström, 1988, p337).
Subsequently, a breastfeeding campaign was introduced that successfully halved the infant mortality rate by 1880 even though the demand and use of female labour had increased over the period. Thus, Brändström was confronted with a situation of high and low levels of infant mortality coinciding with the extensive use of female labour, hence his conclusion;
“The results from the two parishes should caution us to be sceptical when we use female labour conditions as an explanatory variable for changes in infant mortality” (Brändström, 1988, p355).
So far then, it is evident that by far the two most widely mentioned factors affecting infant mortality, infant feeding practices and female labour, are some way from providing an accurate description of national and regional differences and trends. Consequently, apart from varying fertility rates, a variety of more particular theories have been proposed to explain outstanding discrepancies.
These explanations make reference to a variety of factors that include: problems associated with increased levels of urbanisation such as high population densities, poor housing (provision) and sanitary conditions and the associated rise and spread of infectious diseases (Turpeinen, 1979, Kintner, 1988, Woods et al., 1988, Watterson, 1988, Woods et al., 1993); availability and improvements in medical care and infant welfare centres (Kintner, 1987, Corsini and Viazzo,
Figure 3. A flow diagram of the factors influencing levels
of infant mortality in the nineteenth century.
Source: Williams and Galley, 1995, p23.
1993); the proportion of illegitimate births (Turpeinen, 1979, Woods et al., 1988); the social class of the family (Turpeinen, 1979, Watterson, 1988, Brändström, 1993, Woods et al., 1993) the prevalence of infanticide and abortion (Sauer, 1978); the proportion of the population who is Catholic (Kintner, 1987, Kintner, 1988); society/a community’s psychological attitude to births, deaths and life in general (Imhof see Williams, 1995); and varying climatic conditions affecting the spread of disease and ‘quality of life’ (see for instance Woods’ et al. (1988) ‘urban-sanitary-diarrhoeal-effect’, Kytir and Münz, 1993, Woods et al., 1993). Many of these factors and there interrelationships are shown in figure 3.
In conclusion, the vast array of literature on factors affecting infant mortality is evidence in itself that any monocausal explanation must be an inadequate answer to what is an inherently complex issue. It has been shown that infant feeding practices and female labour obligations, the two most widely quoted causal factors, are interrelated and also dependent on a host of other determining factors that have only been mentioned here. One way to grasp the complex nature of infant mortality rates is via an appreciation of the susceptibility and sensitivity of infants to their new world, something that mothers from around the world realise and try to protect their children from even today (note responsible parents’ obsession with cleaning and sterilisation!). Small wonder then that,
“studies…aimed at monocausal explanations of trends in infant mortality have fast proven unsatisfactory: infant mortality is clearly conditioned by a [complex] combination of social, cultural, biological, economic and psychological factors [that interplay with each other]” (Brändström, 1993, p24).
BIBLIOGRAPHY
Beaver, MW (1973) “Population, infant mortality and milk”, In Population Studies, 27, pp243-54.
Brändström, A (1988) “The Impact of Female Labour Conditions on Infant Mortality: A Case Study of the Parishes of Nedertorneå and Jokkmokk, 1800-96”, In The Society for the Social History of Medicine, 1, 3, pp329-58.
Brändström, A (1993) “Infant Mortality in Sweden, 1750-1950: Past and Present Research into its Decline”, In Corsini, CA and Viazzo, PP (eds) The decline of infant mortality in Europe – 1800-1950: four national case studies, Florence, UNICEF, pp19-34.
Corsini, CA and Viazzo, PP (1993) “The Historical Decline of Infant Mortality: an overview”, In Corsini, CA and Viazzo, PP (eds) The decline of infant mortality in Europe – 1800-1950: four national case studies, Florence, UNICEF, pp9-17.
Hobcraft, JN, McDonald, JW and Rutstein, SO (1983) “Child spacing effects on infant and early child mortality”, In Population Index, 49, pp585-618.
Hobcraft, JN, McDonald, JW and Rutstein, SO (1985) “Demographic Determinants of Infant and Early Child Mortality: A Comparative Analysis”, In Population Studies, 39, pp363-385.
Kintner, HJ (1987) “The impact of breastfeeding patterns on regional differences in infant mortality in Germany, 1910”, In European Journal of Population, 3, pp233-261.
Kintner, HJ (1988) “Determinants of Temporal and Areal Variation in Infant Mortality in Germany, 1871-1933”, In Demography, 25, 4, pp597-610.
Kytir, J and Münz, R (1993) “Infant mortality in Austria – 1820-1950: Trends and regional patterns”, In Corsini, CA and Viazzo, PP (eds) The decline of infant mortality in Europe – 1800-1950: Four national case studies, Florence, UNICEF, pp71-86.
Lee, R (1988) “Infant, Child and Maternal Mortality in Western Europe: A Critique”, In Brändström, A and Tedebrand, L (eds) Society, Health and Population during the Demographic Transition, Stockholm, Umea university, pp9- 22.
Lithell, U (1988) “Childcare – A Mirror of Women’s Living Conditions. A Community Study Representing 18th and 19th Century Ostrobothnia in Finland”, In Brändström, A and Tedebrand, L (eds) Society, Health and Population during the Demographic Transition, Stockholm, Umea university, pp91-108.
McKeown, T (1976) The Modern Rise of Population, London, Arnold.
Sauer, R (1978) “Infanticide and Abortion in Nineteenth- Century Britain”, In Population Studies, 32, 1, pp81-93.
Tedebrand, L (1988) “Introduction”, In Brändström, A and Tedebrand, L (eds) Society, Health and Population during the Demographic Transition, Stockholm, Umea university, pp7-9.
Turpeinen, O (1979) “Infant Mortality in Finland 1749-1865”, In Scandinavian Economic History Review, 27, 1, pp1-21.
Watterson, PA (1988) “Infant mortality by father’s occupation from the 1911 census of England and Wales”, In Demography, 25, 2, pp289-306.
Williams, N (1995) Lecture delivered at the University of Liverpool on the 28th April 1995.
Williams, N and Galley, C (1995) “Urban-rural differentials in infant mortality in Victorian England”, In Population Studies (in press).
Woods, R, Watterson, PA and Woodward, JH (1988) “The causes of rapid infant mortality decline in England and Wales, 1861-1921. Parts I and II”, In Population Studies, 42, pp343-66 and 43, pp113-32.
Woods, R, Williams, N and Galley, C (1993) “Infant Mortality in England, 1550-1950: Problems in the identification of long-term trends and geographical and social variations”, In Corsini, CA and Viazzo, PP (eds) The Decline of Infant Mortality in Europe – 1800-1950 – four national case studies, Florence, UNICEF, pp35-50.