STATS ARTICLES 2005

2009 | 2008 | 2007 | 2006 | 2005 | 2004 | 2003


Oxytocin Study Results “Too Good to Check”
November 29, 2005
Maia Szalavitz
Neglected adoptees and hard-wired unhappiness

Coverage of a study of Eastern European orphans published in the November 21 issue of the Proceedings of the National Academy of Sciences was hard to miss : the New York Times, CBS, CNN and dozens of others gave it prominent play.

Virtually all of the reporting focused on claims that children who suffer early neglect have lower levels of Oxytocin, a protein believed to be involved in creating feelings of love, trust and affection and also implicated in orgasm.

The study compared 18 four-and-a-half year-old children raised in the affection-deprived environment of Russian and Romanian orphanages and then adopted to 21 children of the same age who’d spent their whole lives with their biological parents.

The media all explained that such research is important for understanding how physical affection helps wire the brain —and how early neglect may set children up for behavioral and psychiatric problems. But in its rush to confirm a chemical connection between a brain state and behavior that researchers have long sought, the media failed to note serious problems with the data.

The Oxytocin levels of both groups were compared after the children sat on the laps of either their mother (in the case of the former orphans, their adopted mother) or an affectionate female stranger and played a video game which instructed the mothers to hug or gently tickle the toddlers.

If you relied solely on the Times for your news, for example, you would probably believe that the study found rises in Oxytocin in the normal group after the interaction with mom, but not amongst the adoptees.

But what the Times, CBS, Newsday and even Nature left out was that the result was close to, but did not reach statistical significance. The study did find significant differences in baseline levels of vasopressin—a brain chemical linked to a sense of comfortable familiarity with people. But vasopressin isn’t anywhere near as sexy as Oxytocin and wasn’t recently studied in the form of a nasal spray that appeared to make people increase their trust in each other.

Newsday did note that all but three of the adopted subjects showed no rise in Oxytocin at all after playing with their adoptive moms—and those three, in a sample this small, could be responsible for the failure of the results to reach significance.

Only CNN’s doctors’ site, MedPageToday, noted the lack of significance.

And this is critical. Lack of statistical significance isn’t a petty problem: it means that the results could have occurred simply due to chance and may be meaningless. Only more research will tell—and the media should have presented the Oxytocin finding as far more preliminary than it did.

Worse, there was one other gaping hole in the study that the media missed. While it is probably the case that the results are linked primarily to the early neglect, the control group was not really comparable to the experimental group.

The kids who’d experienced early neglect were not only neglected, but neglected and then adopted. They were compared to kids simply raised by their biological parents. Unless they had been compared to children adopted at birth, there’s no way to parse out whether the effects are related to neglect or to adoption itself.

Coverage of the study was full of cautionary quotes aimed at concerned adoptive parents, saying that these results probably don’t apply to their kids if they did not suffer early neglect. And that’s almost certainly the case. But the truth is, the study says nothing about that issue because there wasn't a group of adopted- but-not-neglected kids studied for comparison — and that severely limits its interpretation.