It is well known that artificial feeding in infancy increased the risk of
infant mortality in historical populations through the nutritional unsuitability of breast milk
substitutes, the absence of adequate sterilization and refrigeration
facilities, and the lack of the mother's antibodies for babies who did not receive their mother's
milk. Even in the late twentieth century, children who were not breast fed have been found to
have been more likely to succumb to respiratory illness during infancy and
childhood. Rates of artificial feeding have been linked to high infant mortality at a population
level, but in England breast feeding has been viewed as practically universal until into the twentieth
century. In reality, however, there is little data available to substantiate this claim
and allow examination of feeding and weaning practices according to the age
of child, let alone other possible influences such as social position or
legitimacy.
Data sets which link feeding practices among individual infants with both
mortality outcomes and other confounding influences on health and mortality
are even harder to find. This paper uses a rare and valuable data set generated by health visitors who visited every child born in the county of
Derbyshire (England) between the years 1917 and 1922. The data set follows
30,488 infants from birth over the first few years of their life, recording
not only the name, address, sex and date of birth, but also the occupations
of both parents, the number of rooms in the house, the number of previous
births, child deaths, stillbirths and miscarriages to the mother, the type
of feeding at each visit and illnesses of the children among other information. The date and cause of death are reported for those who
died. Feeding patterns are examined according to age of child, and for factors
such as social position, legitimacy, prematurity and multiplicity of
birth. Multivariate hazards modeling is used to analyse the relative influences
on mortality of social factors (such as parental occupations, social
class, and house size) environmental factors (such as density of population,
mining community, and community waste disposal systems) and feeding practices over the post neonatal period and during
childhood, both for mortality overall and by broad causal groups. The analysis indicates that
environmental factors are more important determinants of post neonatal and
childhood mortality than social position, but that artificial feeding from
an early age is linked to an increased risk of death in both infancy and
childhood, particularly from both gastro-intestinal and respiratory
diseases.
Additional project in progress :
'Defective
vitality': Cause of death reporting in England in the early
twentieth century