Although one type of sugar in breast milk from HIV-positive mothers can boost likelihood of transmission, many other sugars protect against disease.
A type of sugar that occurs naturally in breast milk can double the likelihood of a HIV-negative baby acquiring the virus through breast feeding if the mother has HIV.
The molecule, called 3'-sialyllactose (3'-SL), is found in varying concentrations in the milk of different women. In a study in Zambia, HIV-negative newborns breastfed by HIV-positive mothers are twice as likely to catch the virus during their first month of life if the mother's milk has an above-average level of 3'-SL1.
But not all milk sugars are problematic, nutritional scientist Lars Bode of the University of California, San Diego, told a symposium of the International Society for Research in Human Milk and Lactation in Trieste, Italy, over the weekend. The same study in Zambia found that five more of the 150-odd complex sugars in breast milk seem to have a protective effect. HIV-negative infants who consumed these sugars had a better chance of reaching their second birthday than did HIV-negative babies who drank breast milk lacking those sugars irrespective of their mothers' HIV status. (Once a baby had caught HIV, however, breast-milk sugars had no influence on survival.)
The findings by Bode and his team are an example of a quietly accelerating area within lactation research. Several labs are trying to identify how variation in the prevalence of the large sugar molecules in breast milk, collectively known as human milk oligosaccharides (HMOs), influences infant health.
Once clear links are established, clinical trials to test HMOs as health-boosting additives in infant formula milk can be drawn up. And, potentially, those women whose milk is found to contain less favourable biochemical characteristics — such as HIV-positive mothers who make lots of 3'-SL — might consider giving their infants donor breast milk in place of their own.
But it is not clear whether a HIV-positive woman in Zambia who learns that she produces 3'-SL-rich breast milk should switch to formula feeding, which would raise the chance of her baby succumbing to other intestinal and respiratory infections.
"I think the field has underestimated the variation in human milk," says David Newberg, a carbohydrate specialist at Boston College in Massachusetts. "The often confusing literature on breast feeding's impact on disease will be largely explained by this underestimation."
Newberg was part of a team that reported2 an association between a dangerous gut disease in babies called necrotizing enterocolitis (NEC) and the inability of affected infants to secrete a suite of oligosaccharides in their mucus. These babies are considered particularly likely to benefit from drinking the sugars via breast milk, but about 10% of European women cannot make them in their milk either.
At the symposium, Bode described how his lab has pinpointed an HMO called disialyllacto-N-tetraose (DSLNT) that seems to underlie breast milk’s contribution to NEC risk. In rats, they found that upping the levels of that sugar could reduce the severity of NEC on its own3.
A quick milk test for this sugar might be able to tell physicians how much they should worry about infants developing NEC. In a step towards that approach, five US hospitals are set to monitor DSLNT concentration in the breast milk of mothers whose premature babies develop NEC, and those whose premature infants remain NEC-free.
But adding DSLNT to the diets of premature babies is still a long way off. It is longer than any oligosaccharide that has so far been synthesized in the lab — and extracting it from breast milk would be prohibitively expensive.
“DSLNT is probably the longest one [HMO] that has been identified to have an effect,” says Bode, noting that other HMOs can be much longer. “I'm pretty sure that most of the different stuff that HMOs do, we haven't even thought about.”
Bode, L. et al. Am. J. Clin. Nutr. 96, 831–839 (2012).
Morrow, A. L. et al. J. Pediatr. 158, 745-751 (2011).
Jantscher-Krenn, E. et al. Gut 61, 1417-1425 (2012).