Bone Stable Isotope Evidence for Infant Feeding in Mediaeval England

Article excerpt


In most cultures, babies are breastfed initially, but at some point the infant is introduced to other foods and breastfeeding declines and eventually ceases, a process known as weaning. Infant feeding practices in past populations are becoming an increasingly important field of study. To some extent this can be viewed in terms of the rise of the archaeology of gender, which has resulted in an increased focus upon women's roles and activities in earlier societies. However, breastfeeding practices also have wider implications for population dynamics in earlier human groups. There is evidence that lactation suppresses ovulation, and so breastfeeding is a major determinant of fecundity and birth spacing in societies lacking reliable artificial contraception (Vitzthum 1994). This also has implications for maternal health, as lengthening the birth interval helps avoid the draining of maternal nutritional reserves associated with short-spaced, repeated pregnancies (Vitzthum 1994). Breastfeeding also promotes infant health because of the milk's immunological content, and because it enables early avoidance of potential infection from food and water (Katzenberg et al. 1996).

In an attempt to derive a biologically based `hominid blueprint' for duration of breastfeeding, Dettwyler (1995) suggested, taking into account variables including length of gestation, birth weight and dental eruption patterns, that a natural age for cessation of human breastfeeding might lie in the range 2.5-7 years. However, cultural rather than physiological factors in influencing infant feeding practices in human societies are of great importance (e.g. Stuart-Macadam 1995).

Nitrogen stable isotope analysis of skeletal remains may be used to study duration of breastfeeding in palaeopopulations. There are two stable isotopes of nitrogen, [sup.14]N and [sup.15]N, and the ratio of the two, the nitrogen stable isotope ratio, is measured by [delta][sup.15]N, expressed in parts per thousand ([per thousand]). [delta][sup.15]N increases as one ascends a food chain, the magnitude of the trophic level effect being approximately 3-4[per thousand](Schwarcz & Schoeninger 1991). A foetal or newborn infant has a [delta][sup.15]N similar to that of its mother. Breastfeeding infants are in effect consuming their mother's tissues, so that they are at a higher trophic level. [delta][sup.15]N rises during breastfeeding to give a collagen [delta][sup.15]N about 3-4[per thousand] greater than maternal collagen. During the weaning process, as mother's milk is replaced by other foods, [delta][sup.15]N normally declines (Mays 2000: 429).

In recent years, a number of studies using nitrogen isotopes to investigate infant feeding in palaeopopulations have appeared (refs. in Mays 2000). However, little work has been done on populations outside North America (although see Dupras et al. 2001). The present work represents a first application of nitrogen isotope analysis to investigate infant feeding in a British palaeopopulation.

This study

The study material comes from the churchyard at the deserted Mediaeval village of Wharram Percy, North Yorkshire, England (Beresford & Hurst 1990). The skeletal material dates primarily from the 10th-16th century AD, and represents interments of ordinary peasants; the assemblage was chosen because it provides a large sample of infant and child skeletons for study. The existence of Mediaeval documentary sources concerning infant feeding provides an opportunity to compare archaeological and written evidence.

Collagen was extracted and analysed, following protocols outlined elsewhere (Richards & Hedges 1999), from ribs of 99 individuals, of which 70 were infants or juveniles and 29 were adult (aged 18+ years). Age was determined using dental development (Schour & Massler 1941). Mean error in age estimation using this technique is likely to be about 0.1 years in those under 5 years (Liversage 1994). …