There is no question that nutritionally, breastmilk is the best food for a baby, but can it also lower the risk of that baby developing type 1 diabetes? The majority (but not all) of retrospective studies that have looked at breastfeeding say yes, it can. Yet retrospective studies rely on women's memories of how their children were fed as babies. The majority (but not all) of more powerful prospective studies (that follow events as they occur), found that no, it can't. For example:
An analysis of data from 17 earlier studies found moderately increased incidence of type 1 diabetes in infants given formula before 3 months of age. Yet they also found that studies that depended on parental recall showed breastfeeding to be more protective than the other studies. These and other methodological differences may explain the differing results (Norris and Scott 1996).
A recent retrospective study found that longer exclusive and total breastfeeding was protective against type 1 diabetes in a group of Swedish and Lithuanian children, regions of high and low incidence of type 1 respectively (Sadauskaite-Kuehne et al. 2004). Another recent retrospective study in Germany also found breastfeeding protective (Rosenbauer et al. 2007).
Three prospective studies have not found breastfeeding to be protective:
A recent prospective study of the general population, however, has found breastfeeding to be protective: a study from Sweden found that longer term breastfeeding was protective against type 1-related autoimmunity, even in 5 year olds, in children who were not necessarily at high genetic risk of developing the disease (Holmberg et al. 2007). Perhaps genetic background can explain some of the different results of these studies.
A prospective study of a group of Finnish children, followed from birth, found that infants with genetic risk of developing type 1 diabetes who had been exclusively breastfed for at least 4 months had a lower risk of developing type 1-associated autoantibodies than did infants exclusively breastfed for less than two months (Kimpimäki et al. 2001). Yet a more recent study by the same authors found that breastfeeding was not associated with the development of beta cell autoimmunity (Virtanen et al. 2006).
So what is going on here? Breastmilk contains numerous protective immunological and anti-inflammatory factors (see the inflammation page), may support the maturation of the intestine (see the diet and the gut page), and may enhance thymus size (see the autoimmunity page for more on the thymus), and thereby potentially help a child's immune system develop properly (Jackson and Nazar 2006). Breastfeeding also has been shown to protect against enterovirus infections during infancy, and these infections are a likely trigger of type 1-related autoimmunity (see the viruses page) (Knip et al. 2010). Breastfeeding also decreases the likelihood of obesity in the offspring of mothers with diabetes of any type (Feig et al. 2010). It would seem to make sense, then, that breastfeeding might be protective against type 1 diabetes.
So why, then, are there inconsistent results to these studies on breastfeeding and type 1 diabetes? Some of the reasons that might help explain the inconsistent results of these studies follow. One, exclusively breastfed infants are at higher risk of vitamin D deficiency than are formula-fed infants. This deficiency can be remedied. During pregnancy, the fetus will maintain sufficient vitamin D levels if the mother has sufficient levels of vitamin D, which may require supplementation. After birth, the mother may need up to 4000 IU per day to maintain sufficient levels in her and her exclusively breastfed baby (Kovacs 2008). Vitamin D deficiency may contribute to the development of type 1 diabetes (see the vitamin D page). Two, just because breastmilk contains a number of immune system factors that can protect against infection in early life does not necessarily mean that these factors would be able to influence the development of disease later in life (Jackson and Nazar 2006). In other words, we might be expecting too much. A third, more troubling possibility, is that environmental contaminants may play a role in compromising the protective effects of breastmilk. Read on for some evidence supporting this possibility.
One possible reason for the conflicting results in the studies on type 1 diabetes and breastfeeding is the method of delivery of milk or formula. None of these studies have looked at whether the babies receive breastmilk directly from the breast, from a glass bottle, or from a plastic baby bottle. Was pumped breastmilk stored in plastic? Was the milk or formula heated in plastic? Toxic chemicals such as bisphenol A can leach out of plastic baby bottles, especially when heated (De Coensel et al. 2009). Some of the chemicals found in plastics affect animals in ways that imply they could potentially contribute to the development of type 1 diabetes in humans (see the bisphenol A and phthalates pages, for example). Researchers conducting studies on breastmilk vs. formula might consider controlling for the method of milk delivery, which bottles were used, how often, and how the milk was stored and heated. Mothers could use glass bottles; in my experience they are sturdy and do not break easily.
And then there's the milk itself. Breastfed children are exposed to higher levels of environmental contaminants than those who are not breastfed. In a study of German 10 year olds, for example, breastfeeding was one of the strongest factors influencing the concentration of persistent organic pollutants (POPs) in their bodies. Breastfeeding was associated with about 30% higher median concentrations of various POPs (DDE, HCB, PCBs), and breastfeeding for 6 months doubled the concentrations of DDE and PCBs compared to children who were not breastfed. These findings are consistent with other studies as well (Link et al. 2005).
Weisglas-Kuperus et al. (2004) found that some of the expected protective effects of breastfeeding on the immune system were counteracted by the negative effects of higher perinatal PCB exposure in Dutch children exposed to normal, background levels of contamination.
In another study of Dutch children, breastfeeding for six months was found to make up 12-14% of the children's cumulative exposure to PCBs and dioxins until age 25. Yet since in utero exposures are likely to have more severe effects than those via breastfeeding, and since the benefits of breastfeeding still outweigh the negatives, this finding should not be a reason to reduce breastfeeding (Patandin et al. 1999).
Breastfeeding has been found to be protective against weight gain later in life, in the children of women who had diabetes when pregnant (Crume et al. 2011c). Breastfeeding also can reduce the risk of metabolic syndrome (a cluster of conditions that is common in people with type 2 and type 1 diabetes) in the mother many years later, whether or not they had gestational diabetes. Previous studies have also sometimes found that breastfeeding is associated with a lower risk of type 2 diabetes in the mother (Gunderson et al. 2010).
World Health Organization (WHO) (http://www.unicef.org/) infant feeding guidelines call for exclusive breastfeeding for a full 6 months, the introduction of safe and complementary foods from the sixth month of life while breastfeeding continues, and then continued breastfeeding for up to 2 years of age or beyond.
Only 5% of families of children with first degree relatives who have type 1 diabetes follow the WHO recommendations for infant feeding (Pflüger et al. 2010), and mothers with type 1 diabetes tend to breastfeed their children less than other mothers, and less than WHO recommends (Hummel et al. 2007).
While breastfeeding may or may not protect against the development of type 1 diabetes (and there is some evidence that it does), it has many other benefits to both mother and child, and should be encouraged. Possible benefits for the mother include a lower risk of metabolic syndrome and type 2 diabetes, and a lower risk of weight gain in the child.
If brreastfeeding is not possible, then parents should know that using hydrolyzed infant formula instead of regular formula has been shown to significantly reduce the risk of developing type 1-related autoantibodies in children genetically at risk of type 1 (Knip et al. 2010, described further on the wheat and dairy page).