Discussion Paper

THE NATURE OF THE EVIDENCE

The literature on this subject is immense, with well over 1,000 papers published on environmental contaminants and milk production (Berlin and Kacew 1997:72). It is challenging material for a non-specialist to absorb. The scientific evidence used in reviewing toxins and breastmilk comes primarily from three sources: animal studies, clinical research and case reports, and industrial accidents. Animal studies are done primarily on mice, rats, and monkeys but large species exhibit differences in sensitivity. Often evidence of toxicity only occurs when exceptionally large doses are given. However, "... effects such as carcinogenicity, immunosuppression, reproductive and behavioural effects, and enzyme induction have been reported in laboratory animals at dose levels equal to, or only one or two orders of magnitude higher than, the levels normally found in human milk in many industrialized countries" (Jensen and Slorach 1991:246). 

The evidence for many environmental contaminants is easiest to collect from breastmilk. But that does not mean that the means of obtaining and interpreting that evidence is without problems. Sometimes breastmilk samples are pooled, but results are interpreted as if all women who contributed to the pooled sample had equally contaminated breastmilk. The time lag between exposure and outcome is another problem. Most women with contaminants in their breastmilk have been exposed for years to chemicals, including during their pregnancy (Berlin and Kacew 1997:73). Shrader-Frechette points out:

This makes setting acceptable limits on toxins particularly difficult. Setting limits is a different operation from monitoring exposure. Some systems such as the Global Environment Monitoring System (GEMS) assume an infant consumes 120 grams of breastmilk per kilogram body weight. They acknowledge that "routine analyses may not be reliable," and data reported is "from the literature and its reliability is unknown" (GEMS 1998:4-5). They warn that data from testing the breastmilk of individual mothers is easily misinterpreted and slow - around 6 weeks (GEMS 1988:16). However, the amount of fat soluble contaminants is difficult to measure because breastmilk composition varies among women, over the course of a single feeding, during the day and night, and from day to day. Residues are lower in colostrum than mature milk, but residues are higher in colostrum fat than the fat of mature milk. High levels of pesticides have been reported in colostrum fat from Italy (Jensen and Slorach 1991:200). Other factors affecting the levels of residues in breastmilk include the age, weight, parity and dietary habits of the mother, seasonal differences, and smoking (due to pesticide residues in tobacco). It is hard to separate in utero affects from postnatal affects; however, since intrauterine contamination occurs at critical earlier stages of fetal development and is unavoidable, it is an important concern. Comparisons are particularly difficult since testing procedures differ in different countries; special projects such as GEMS are set up internationally to overcome this. But as analytical methods improve statistics and rates become even less comparable over time. There is also the potential that private and government agencies might falsify data on toxicology tests, resulting in approval for insecticides and pesticides in Canada, United States and Sweden based on incorrect data (Jasanoff 1990:4). Even political factors are relevant, as countries vie for evidence to show they have been regulating toxic wastes. For example, New Zealand boasted that pesticides in breastmilk were lower than in European and North American mothers. The vested interests of governments and industry are becoming more apparent in the face of privatisation. Researchers also have to consider the complex relation between free informed consent and socially beneficial research.  

None of the studies reviewed consider the perspective of the breastfeeding mother who must weigh evidence and make a decision about infant feeding at a time when she is narrowly focused on childbirth and establishing a relation with her baby. If she is breastfeeding, she may be very sensitive to external criticism and threats, and unwilling to contemplate the possibility that there is something 'wrong' with her milk. She may thus protect herself from unwelcome information until she has stopped breastfeeding.

Risks, Rights and Regulation: Communicating about Risks 
and Infant Feeding