The impact of breastfeeding patterns on regional differences in infant mortality in Germany, 1910.

The purpose of this paper is to assess the impact of breast feeding on regional differences in infant mortality in Germany in 1910, when other social, medical, and economic factors are controlled. Around 1910 the infant mortality rate ranged from 78/1000 to 276/1000, and the percentage of infants ever breast fed varied from 25% to 95%. There was a rough correlation of breast feeding with low infant mortality in the northwest and the major cities (Berlin, Hamburg, Bremen) and of the absence of breast feeding with high infant mortality in the southeast. Since breast feeding and marital fertility are interactive variables, 2 equations were built to show the effect of the several variables on each as well as the effects of each on the other. The data for infant nutrition practices and causes of death are taken from statistical records for 54 administrative districts. In 1910, only 44% of the population was rural, and population density was .82 persons/sq km. There were 5 physicians/10,000 people and practically no infant or maternal health centers. Infant mortality and maternal fertility were both high; 9% of births were illegitimate; and 35% of the population was Catholic. Maternal fertility was found to correlate with Catholic religion and rural residence, rather than with breast feeding. In 3 separate equations infant mortality was regressed on: 1) marital fertility, percent Catholic, access to medical care, population density, and level of illegitimacy; 2) these variables plus percent ever breast fed and access to sewerage; and 3) the same variables minus prevalence of breast feeding. The results showed that illegitimacy, access to medical care, marital fertility, and Catholic religion had more to do with the regional differences in infant mortality than did breast feeding. This remains true even though the chief causes of infant death were water-borne diseases. These findings suggest that, although breast feeding generally has a negative effect on infant mortality, its impact may have been exaggerated in earlier studies, many of which did not control for other social, economic and medical factors.