Is Breast Best? Depends on Baby's DNA

You can bet that the increase in the percent of newborns who are breastfed, from 68 percent in 1999 to 74 percent in 2004, didn’t happen because more mothers cared about strengthening their child’s immune system (one of many reported benefits of breast over bottle, according to the American Academy of Pediatrics). Instead, this period coincides with more reports that breastfeeding spurs a baby’s brain development, conferring an extra half-dozen or so IQ points by the time he or she enters school. Talk about feeding into the neuroses of middle-class parents.

But not all breastfed babies are little Einsteins, and some parents may well wonder why all the months of milk-stained blouses and balky breast pumps didn’t seem to boost Junior's cognitive development. A remarkable study unveiled Monday evening offers a clue. Researchers are reporting in the Proceedings of the National Academy of Sciences that only babies who carry a particular form of a gene derive an IQ benefit from being breastfed. Without this form of the gene, breastfeeding has no effect on later IQ.

The study, of more than 3,000 children in Britain and New Zealand, cuts through the stultifying debate about whether intelligence reflects nature or nurture. Of course it reflects both, which is not exactly a stop-the-presses statement. More interesting is the finding that intelligence reflects a specific interaction of genes and environment: in children with a particular version of a gene called FADS2, breastfeeding raises intelligence an average of nearly 7 IQ points, find scientists led by Terrie Moffitt and Avshalom Caspi of King's College London and Duke University.

The good news is that this form of the gene is more common: 90 percent of the children in the study had at least one copy of it (it’s called the “C” version of FADS2), while 10 percent had the “G” version and therefore got no IQ advantage (or disadvantage) from breastfeeding.

This makes sense, because the magic brain-boosting compounds are fatty acids present in human but not cow’s milk or most infant formulas. The C version of the FADS2 gene (located on chromosome 11) produces an enzyme that helps convert the fatty acids in breast milk into DHA (docosahexaenoic acid) and AA (arachidonic acid). Both accumulate in the human brain during the months after birth. Both play important roles in making signals zip along brain neurons efficiently and in spurring neurons to sprout lots of connections (which underlie intelligence, memory and, probably, creativity). Without the “smart” form of the gene, you lack the enzyme required to carry out this conversion.

Parents should not rush out to have their child genotyped before deciding whether or not to breastfeed, though. For one thing, there is no commercial test for this gene. For another, breastfeeding confers other health benefits, mostly on the developing immune system. (Or so we think: as any mom who breastfed her child for two years, only to see him develop asthma or another immune-system glitch, wonders right about now, maybe those benefits also depend on the child carrying certain genes.)

This careful, nuanced study stands in contrast to so much other research that claims a connection between a gene, or an environment/experience, and some outcome. Caspi and Moffitt have led the way in showing how to move beyond such crude associations: genes have an effect only in some environments, and environments have effects only on people with specific genes. For instance, a gene associated with depression actually leads to that outcome only in people who suffer stressful life events; if you have the “depression gene” but lots of social support and little stress, this gene won’t do what its name implies, Caspi and Moffitt showed in 2003. Similarly, they showed in 2002 that although children who are maltreated are widely assumed to be at higher risk for growing up to be anti-social, more likely to commit crimes and drop out of school, these terrible experiences lead to this tragic outcome only in children with a particular form of a gene called MAOA.

These studies are extremely hard to do, since they have to rule out long lists of other explanations before they can establish that an environment/experience works through a gene. In the breastfeeding study, for instance, the scientists had to rule out the possibility that mothers with “smart” genes were both more likely to breastfeed and to have children with the form of the FADS2 gene that lets babies derive brain benefits from breastfeeding.

Despite their difficulty, more studies need to enter the 21st century when it comes to gene-environment interactions. Earlier today, for instance, researchers presented a study at the American Heart Association’s Scientific Sessions in Orlando claiming that breastfed babies are less likely to have certain risk factors for cardiovascular disease in adulthood than bottle-fed babies. Specifically, breastfed babies grew up to have a lower average body mass index (BMI) and higher average HDL (“good” cholesterol), said Nisha I. Parikh of Beth Israel Deaconess Medical Center in Boston. But the study (which used data from the Framingham Heart Study) did not find that it was a sure thing. That is, adults who were breastfed as infants were only 55 percent more likely to have high levels of good cholesterol than those who were not breastfed. Assuming that who benefits in this way from having been breastfed is not a complete crapshoot, isn’t it time for all studies like this to see if there is a genetic explanation for the results? That is, is there a genetic variant that allows some people to derive heart benefits from being breastfed, but not others—just as with IQ and being breastfed?

If we are ever going to get serious about the age of personalized medicine, surely it’s time to stop giving one-size-fits-all health advice.