Historical demography strand abstracts

Historical mortality: Tuesday 10 September, 9.00am

From ague to pyrexia and from smallpox to heart disease: a general overview of causes of death in Scotland 1855-1955
Alice Reid, Cambridge Group for the History of Population and Social Structure, University of Cambridge; Eilidh Garrett, University of Cambridge & University of St Andrews; Chris Dibben, University of St Andrews; Lee Williamson, University of St Andrews

This paper provides a broad overview of mortality in Scotland for the century following the inception of civil registration in 1855. A newly created dataset comprising the detailed returns of deaths by cause of death, age and year as given in the Annual Reports of the Registrar General for Scotland will be used. The role of variations in nosologies, developments in medical knowledge and fashions, differences in age structure and the evolution of the disease environment in the changes in causes of death reported at the national level between 1855 and 1955 will be examined. Numbers of deaths were published annually by the Registrar General for Scotland by age, sex and cause. In 1855 deaths were classified into 109 causes; by 1949 255 separate causes were enumerated. The way individual causes were recorded varied over time, resulting in nearly 1000 distinct descriptors. These have been re-coded and regrouped so as to maintain (as far as possible) consistent causal groupings across time. The resulting time series enables numbers of deaths, as well as crude, age-specific and standardised death rates, to be considered by sex and cause over time. Broad trends will be presented, and comparisons of the various measures allow the roles of changing age structure and the decline in infectious disease in the overall decline of mortality to be assessed. The data will also be used to comment on Flinn’s version of the McKeown thesis regarding the reasons for mortality decline in Scotland.


Was there a causal relationship between early-life living conditions and adult survivorship in the past? Insights on socioeconomic determinants from a height-mortality linkage in Spanish rural settlements (1780-1900)
Antonio D. Cámara , Joana Maria Pujadas, Centre d'Estudis Demografics

The mediation of socioeconomic factors on the health outcomes of past populations has revealed to be complex in that socioeconomic inequalities do not always result in a health gradient or at least this gradient does not hold for all types of health outputs. Hence, it seems convenient to reassess the existence or absence of a solid relationship between certain inputs like income and/or occupation and health outputs on the basis of the potential mediation of uncontrolled factors related to the cumulative impact of living conditions over the life course. The aim of this work is twofold: 1) To test the relationship between health status in early stages of life (as approached by adult height) and survivorship (as approached by age at death). 2) To test whether that potential relationship was somehow mediated by the socioeconomic status of individuals and/or whether a potential attenuation of SES-related effects holds for all causes of death. Height (from recruitment acts) and mortality data (from death registers) were collected, tabulated and linked for male cohorts born between 1780 and 1900 in the municipality of Montefrio (Spain). Based on the valid matches, survival analysis techniques are being used in order to calculate relative hazards in function of age (at enlistment and death), birth cohort, adult height and several combinations from the set of socio-demographic variables available in the sources.


The direct influence from early-life disease exposure on young adult height, southern Sweden 1814-1950
Stefan Öberg, Department of Economy and Society, Unit for Economic History, University of Gothenburg

We use a sibling comparison design to investigate if there is a direct influence from disease exposure early in life, as measured by mortality rates, on young adult height. The data comes from the Scanian Economic Demographic Database, covering five parishes in southern Sweden, with heights linked from universal conscript inspection lists to men born 1814-1950. The longitudinal data structure and detailed information make it possible not only to calculate the community level infant mortality rate for the year of birth but also for each individual’s first year of life and time in utero. Since family relations are known we can use a sibling comparison design to remove all confounding factors constant between the brothers. No statistically significant and consistent negative influence is found. There are reasons to believe that the results represent lower bound estimates of any such influence. We for example find some indications of mortality selection counteracting any negative influence, or scarring effect, from disease on height. We can therefore not conclude that there is no negative influence. We can conclude that both the selection and the scarring effect on height from early life disease exposure, as measured by mortality rates, are likely to be very weak. 


Causes of death and diseases in the past: Tuesday 10 September, 1.30pm

The experience of trachoma, a poverty-related disease, in Italy and Spain between the 19th and the 20th Century 
Lucia Pozzi, University of Sassari; Josep Bernabeu Mestre, University of Alicante; Maria Eugenia Galiana, University of Alicante

Trachoma, an eye disease caused by the Chlamydia trachomatis bacterium, is still today one of the three main causes of avoidable blindness, widespread in most countries in the developing world. The areas most affected by trachoma are characterised by conditions of extreme poverty and strong deficiencies in sanitation and housing. In the early 20th century it was still endemic in many European countries, in particular in the Mediterranean: Italy and Spain were two of the countries most affected by the disease; in the southern regions of the Italian peninsula and along the Mediterranean coasts of Spain only in the 60s it definitively disappeared. A first goal of our work is to analyse the incidence of the disease in these two European countries from the late 19th century until the 1960s, using all available documentation, highlighting its determinants and the causes of its prolonged persistence in the territories longer suffering from trachoma. The comparison between the strategies used to fight the disease in the two countries will allow us to highlight the importance of the health education, inspection activities carried out by dispensaries and rural hygiene centres, intervention and environmental sanitation. The historical experience will help us to reaffirm the validity of those strategies to improve the hygienic conditions and the environmental factors that cause the prevalence of trachoma, through a transversal community. Also, the limitations of an exclusively pharmacological treatment thereby providing valuable reflection also on today's health interventions made in the poorest countries. The analysis will also allow to contextualise the experiences of the fight against trachoma in two European countries, highlighting the crucial role played by international health activities in the fight against trachoma. 


Control of tuberculosis in the Gold Coast, 1900-1960
Joshua Amo-Adjei Department of Population and Health, Faculty of Social Sciences, University of Cape Coast; Frank Bonsu, National TB Control Programme, Accra Ghana; Kofi Awusabo-Asare, Department of Population and Health, Faculty of Social Sciences, University of Cape Coast

The paper explores the control of tuberculosis (TB) in the Gold Coast from 1900 to 1950. Using archival data, it deals with the epidemiology of the disease, spatial dimension and the responses from the government. First recorded cases were in 1904 when the government expressed concern over the health of Africans, especially at mining centres and major towns. Mining areas were the epic-centres of TB. For instance, in 1935, Tarkwa, a mining town with a population of 2,671 recorded 56 cases compared to 92 for Accra with population of 60,000. Reported deaths were highest among migrants from the north working in the mines. The high reported cases could be attributed to poor socio-economic conditions. Although, TB was common, emerging as the second highest reported cause of mortality, responses to address tuberculosis were weak. Among the challenges with control were poor record keeping and local opposition to data collection on illnesses, weak government commitment towards TB, issues which still plague the control of TB in Ghana.


Towards a long-term historical cause of death coding scheme
Alice Reid, Cambridge Group for the History of Population and Social Structure, University of Cambridge; Eilidh Garrett, University of Cambridge & University of St Andrews; Chris Dibben, University of St Andrews; Lee Williamson, University of St Andrews

The comparison of causes of death over time is extremely difficult. Changes in medical knowledge and fashion mean diagnoses of the same condition may change over time: in most cases they become more precise, or two conditions, initially confused, are later recognised to be different. Cause-of-death classifications have changed over time to reflect medical knowledge and the cause of death structure and, even if causes are regrouped, it often proves impossible to follow causes reliably over time. The use of individual level cause of death data would allow the problem of changing nosologies to be addressed, as individual cause of death strings can be reclassified, but a classification system sensitive and flexible enough to reflect changes in the way the causes of death are recorded over time has not been developed. This paper reports on an attempt to develop such a classification, starting with the basic building blocks of the 4 character ICD10 codes but then adding additional historic codes for historical descriptors (HICOD). Of course mis-diagnoses, ‘unknown’ causes and overly vague descriptions of symptoms mean that it will never be possible to know the exact distribution of historical causes of death, and this may hamper analyses of change over time. However some groups of causes can be followed with more consistency than others, and this paper will give examples of successful and more problematic groupings using a variety of nineteenth century data-sets containing individual level causes of death.


Fertility and households in the past: Tuesday 10 September, 4.45pm

Son preference in a sharecropping society: Gender composition of children and reproduction in the pre-transitional Italian community of Casalguidi
M. Manfredini, Department of Life Science, University of Parma; M. Breschi, A. Fornas

According to the evolutionary theory of Trivers and Willard (1973), biased gender compositions of children should be more male-oriented in high-status families, and more female-oriented in low-status ones. It has been then acknowledged that cultural and socioeconomic factors might offset and even reverse such a theoretical assumption. Gender-related reproductive behaviours could be the result of pure discrimination, usually towards females, but could also arise from different economic costs and benefits of sons and daughters. The present study aims at investigating the existence of son-oriented reproductive behaviours in a pre-transitional population of mid-nineteenth century Italy. The community was characterized by a large presence of sharecroppers, whose economic and occupational status was based on the availability of male working force within the household. By analyzing the reproductive life-histories of the villagers, we found evidence that such urgency translated, for sharecroppers, into an element able to affect fertility and reproductive behaviours. 


Innovation-diffusion, geography and clustering of fertility behaviour during the fertility transition: A multi-level analysis of micro census data for Sweden 1880, 1990 and 1900
Martin Dribe Department of Economic History and Centre for Economic Demography, Lund University; Sol Pía Juarez Centre for Economic Demography, Lund University; Francesco Scalone Department of Statistics, University of Bologna

Trying to describe and explain the fertility transition has been one of the main tasks of demographers and economic historians. Large scale projects such as the European Fertility Project have made great contributions to our understanding, although controversy remains about the causes and mechanisms of the fertility decline. There is an ongoing debate about whether the transition was mainly a response to changing socioeconomic structural conditions (adjustment) or a result of new attitudes, norms and behaviour spreading at about the same time in many regions of Europe (innovation diffusion). Although empirical evidence has supported both views, it has proven especially difficult to empirically corroborate the innovation diffusion theory (IDT). By studying fertility behaviour during the early phase of the fertility transition in Sweden, we aim at empirically testing the IDT. Since at least implicitly, IDT is often connected to geographic differences and ideas about clustering of fertility behaviour across space, we apply a multilevel approach on micro-level census data covering the entire population. The idea is to look at measures of clustering of fertility behaviour by geographical units to find empirical evidence that the community where people lived had an independent effect on their fertility. Despite average differences in fertility at geographical level, our findings demonstrate that fertility decisions are explained almost exclusively by family-level characteristics regardless of where the mother lives. Our results suggest that the IDT is based purely on an ecological interpretation of geographical differences (averages) rather than actual individual variability (variance).


Urban and Rural Illegitimacy in Coimbra, Portugal (1885-1910)
Mafalda Moura Pereira, University of Cambridge

Demographic treatments of births in Portugal around the turn of the twentieth century suggest (though partially and inconsistently) that illegitimacy was higher in cities than in the country. My paper tests this assumption through a study of some 16,000 baptisms in Coimbra's urban parishes between 1885 and 1910. It attempts a relatively rigorous analysis of what may be conceived an intractable problem in seeking to break down Coimbra’s supposedly higher incidence of illegitimacy into a genuinely urban and an extra-urban component. Throughout the study period, Coimbra offered not only free medical care for single women but also private clinics and a selection of university-trained, fee-taking midwifes. It is thus plausible that it served as a magnet for pregnant women from the city’s rural hinterland, who for health and other reasons chose to come to Coimbra to lie in.
The paper lays claim to at least some methodological originality in its use of parish records to recategorize illegitimate births. Its procedure is to strip out from a record of the ‘city’s’ baptisms 1) adults, latterly recognised as legitimate by fathers or meeting a precondition of their own marriage, and 2) relevant children, who may have been abandoned in the county and sent to Coimbra's orphanage, born outside but baptised within Coimbra, or born in Coimbra to non-resident mothers. Coimbra's baptismal records enable household reconstitution in stating parents’ occupations, civil status, home town and residence, and grandparents’ and godparents’ names. By subtracting extra-urban births, we arrive at a single figure for cases of ‘resident’ illegitimacy and so derive the city’s illegitimacy rate.
Early findings for the relative incidence of rural and urban illegitimacy tend to bear out the assumption that rural women, following Catholic practices, were less likely to give birth out of wedlock. Over a 26-year period, the rural parish Sta. Clara numbered 179 illegitimate births in 1,452 (12.3%). The urban S. Bartolomeu had 412 such births in 2,176 (18.9%). Rural parents also tended to take their children to church for baptism earlier than their urban counterparts, after an average of 96 days against 110.  


Investigating the immobile
Lyn Boothman, University of Cambridge

My research has examined long-term immobility in one small town in the two hundred year period to 1861.What proportion of couples, and single adults, had ancestors in the town, what proportion spent a large proportion of their lives in this one settlement? Did this change over time and how did it relate to social status, to kinship and to receipt of poor relief? This archive-based study used a population reconstruction which included a very wide range of data. Two types of indicator were used to assess immobility: 1. Family: the proportion of adult residents (singularly, or as a partner in a couple) who had ancestors in the parish; most analysis is based on those who were at least third generation 2. Time: the proportion of adult residents (singularly, or as a partner in a couple) who lived in the town for at least 30 years over the age of 16. These indicators were then related to social status and to kinship links within the settlement. Amongst the most noteworthy results: 1. An increase in immobility on both family and time indicators across the period. 2. A higher rate of change amongst married women than their husbands. 3. A considerable change in the social composition of the immobile across the period. 4. A concentration of immobility in the lower social groups, and amongst some families, by the nineteenth century. 5. Kinship links amongst immobile couples increased and kinship density grew considerably by the nineteenth century.