A new

study of behavior problems in children prenatally exposed
to methamphetamine has prompted predictably alarming
headlines that recall the panic about “crack babies,” who
supposedly were handicapped for life by their mothers’ drug use.
Those concerns turned out to be wildly overblown, and for years
scientists who regret that episode have been warning
their colleagues not to make the same mistake with “meth babies.”
But such admonitions may be no match for researchers’ natural
tendency to exaggerate the importance of their findings, combined
with journalists’ constant thirst for a good scare story. 

In this case, researchers led by Brown University psychologist
Linda LaGasse compared the Child Behavior Checklist (CBCL) scores
of 166 children whose mothers had consumed meth during pregnancy to
those of 164 children whose mothers had not. The two groups, drawn
from four cities (Los Angeles, Honolulu, Des Moines, and Tulsa),
were matched for “race, birth weight, public health insurance,
and education,” and the kids were assessed at ages 3 and 5 based on
an oral questionnaire answered by their caregivers. “This first
report of behavior problems in patients as young as 3 years
associated with MA [methamphetamine] exposure identifies an
important public health problem,” LaGasse and her colleagues

write in the April issue of Pediatrics. “These effects
on behavior problems are quite robust and may have substantial
public health implications because problems as noted on the CBCL
tend to persist over time and predict later psychopathology and
criminal behavior that place tremendous burdens on society.”

Sounds serious. If you look at the overall CBCL scores, however,
you will see that the exposed and nonexposed kids were virtually
indistinguishable. At age 3, the “total problems” score was 52.2
for the exposed children and 51.1 for the comparison group; at 5,
the scores were 52.9 and 50.2, respectively. In neither case was
the difference statistically significant. The two groups’
“internalizing” scores (which include four subscales) also were
essentially the same. “There were no effects of MA on the
internalizing or total behavior problems scales,” LaGasse et al.
report.

Instead of concluding that the two groups were basically the
same, the researchers focus on a few subscales where there were
statistically significant differences. “MA exposure was associated
with increased emotional reactivity and anxious/depressed problems
at both ages and externalizing and attention-deficit/hyperactivity
disorder problems by age 5 years,” they write. The differences
range from slightly less than a point to slightly more than a
point. “Because CBCL findings are based on caregiver report,”
the authors concede, “there could be reporting bias.” In other
words, the mothers or other caregivers who answered the
questionnaires may have been more inclined to perceive and report
problems precisely because they worried about the potential damage
caused by meth exposure. There was a similar weakness in much of
the research on children exposed to cocaine in the womb.

Assuming the score gaps reflect real differences, how do we know
they are caused by prenatal exposure to methamphetamine? To their
credit, LaGasse et al. take into account a bunch of other
variables: “prenatal exposure to alcohol, tobacco, and
marijuana; birth weight; gender; SES [socioeconomic status];
maternal age; single (no partner); caregiver change; domestic
violence; postnatal use of MA; tobacco, alcohol, and marijuana
exposure; caregiver psychological symptoms; the quality of the
home; child abuse; and study site.” But they concede that
“our measure of child abuse through caregiver report of
Child Protective Services likely underestimates abuse.”
Likewise, their measure of “caregiver psychological
symptoms” (which were associated with children’s behavior
problems) may not have captured all of the relevant differences
between women who smoke meth when they’re pregnant and women who
don’t.

In short, it is not clear from these data that “meth babies” (as
several headlines called them) are different from other children in
practically significant ways, that they represent “an important
public health problem,” that they they will “place tremendous
burdens on society,” or even that meth itself is the cause of any
differences that actually exist. Given the cautionary example of
“crack babies”—who were stigmatized as dramatically and permanently
damaged, while their mothers were singled out for special
condemnation and punishment—researchers and the journalists who
trumpet their findings should try to be a little more careful this
time around.