Effects of Prenatal Cocaine Exposure on IQ
In Obama’s unexciting review of the Bell Curve, he remarked:
no one disputes that children whose mothers smoke crack when they’re pregnant are going to have developmental problems.
The relevant studies reveal a more complex picture, though. The effects of prenatal cocaine exposure on IQ remain heavily contested to this day. However, recent evidence from Bennett et al points to a 3 to 5 IQ point drop, on average. This is the most recent study on this subject that I’m aware of.
Interestingly, in following with a few previous studies, it was found that boys suffer a greater cognitive loss from prenatal cocaine exposure than girls. Also, the study found that 9 year olds had equally fewer IQ points as their 4 year old counterparts, countering to a certain extent the idea that the IQ loss goes away as development progresses.
If Bennett’s numbers are correct, they have small– but significant– implications for the Black-White IQ gap. Unlike tobacco and alcohol, which are used by pregnant white and black women at about equal rates and intensities on average*[1], black women are much more[2] likely than white women to use cocaine or crack while pregnant. This is relevant to behavioral genetic studies– both past and present– which have aimed to understand the relative contributions of genetics and environment to the IQ gaps. There is no way, as far as I know, to extract prenatal factors like cocaine use from measures of heritability without explicitly measuring such inputs. As far as adoption studies in particular, it stands to reason that women who place their babies up for adoption exceed the rest of the US population in pregnant cocaine use. An interesting thing about the Scarr adoption study is that all of the mothers of the half-black kids were white.
[1] Today, that is. 1989 was the earliest year I could find data for, and in that year the pattern is starkly different from today– the black-white ratio in fetal alcohol syndrom for this year has way more alcohol use by pregnant black women than pregnant white women, and also much higher rates of fetal alcohol syndrome among black babies. I’m not sure if the rates were comparable in say the 70′s, when the Scarr adoption study was performed. That would be interesting data if anyone happens to have it.
[2]~12 times more in the second link, from 1994
*Source for the alcohol/tobacco/fetal alcohol syndrome rates is the CDC.
Labels: IQ
Ben,
Where is that long collaborative post you were writing on IQ – I haven’t seen it posted yet?
pconroy,
it’s in the works. you can probably expect it sometime before october 1st.
it’s in the works. you can probably expect it sometime before october 1st.
inshallah.
Responsible behavior while pregnant is correlated with IQ.
It’s one of many ways in which low IQ parents provide an inferior environment for their children (beginning in the wom).
The abstract provided for the Bennett piece doesn’t suggest that they controlled for maternal race (and I daresay it would have been very un-PC to do so). If as you indicate black women are more likely to use cocaine during pregnancy, a race-blind sample of cocaine-exposed and non-cocaine-exposed children would have proportionally more black children in the cocaine-exposed cohort. Though maternal intelligence was accounted for, regression to the mean as well as assortative mating (i.e. fathers more likely to be black) would result in the cocaine-exposed kids having lower IQs on average, even in the absence of real pathology. In short, I’m not sure this effect is real.
Animal models to the rescue.
A couple of remarks. First, Obama didn’t mention IQ. He said “developmental problems.” Second, as Half Sigma commented, “Responsible behavior while pregnant is correlated with IQ.” Indirect causes of developmental difficulties (e.g., maternal neglect of proper pre-natal nutrition by crack moms) are just as attributable to the crack as are direct causes. When “developmental problems” are mentioned not everyone makes the immediate leap to IQ, particularly not special ed teachers like my wife.
bbartlog,
the sample as a whole was heavily (87%) African-American, so it seems unlikely that race could confound the study considering it controlled for maternal IQ.
Obama didn’t mention IQ. He said “developmental problems.”
I’m aware that IQ is just one part of development. I chose to focus on it for this post. Surely Obama was including IQ when he referred to “developmental factors” because he was reviewing the bell curve.
Indirect causes of developmental difficulties (e.g., maternal neglect of proper pre-natal nutrition by crack moms) are just as attributable to the crack as are direct causes.
i agree, this explanation is definitely worth studying.
Among High Andes Indians, chewing coca leaves is legal and general. Maybe that explains their relatively low IQ.
Ben,
Where is the long post you were working on? Did I miss it?
pconroy,
as we read more and more we found that candidate-gene association studies aren’t reliable. genome-wide association studies with large enough sample sizes will be powerful enough to identify the SNPs for IQ variation.
We have been slowly working on a post which explains why this is the case, discussing how future progress will be made in this area, etc.
Ben,
What do you think of the idea – which I think I read on this site – of a multitude of SNP’s defining IQ variations of large effect, which are mostly distributed in narrow geographic or family lineages?
In other words, hard to detect, without studying specific families where gifted individuals are more common than expected. Like the Darwin family…
What do you think?