Early in my mother's pregnancy, she became aware her father had a terminal illness. He died two months before I was born. I'm almost 70 years old but have always wondered: What do we know, if anything, about the long-term impact of a traumatic event on human development?
FIRST let's define trauma.
Back in the days when medical insight consisted largely of old wives' tales, it was regularly claimed a pregnant woman shouldn't spend too much time around fish lest her child be born scaly.
Cut to the present: nobody doubts there are some things a pregnant woman can experience or do—alcohol or drug use, smoking—that can mess up the eventual kid. The question is about traumas that fall between these two poles. Combing through the scientific literature—we didn't go trolling on the crackpot sites for this stuff—we find the following remarkable assertions:
• The prevalence of autism among children in Louisiana increased with the severity of prenatal exposure to hurricanes.
• A higher-than-expected number of craniofacial malformations and heart defects were found in babies born to women whose older child died unexpectedly during the pregnancy.
• Being a crime victim or experiencing the death of a relative was associated with increased risk of—get ready for this, LoDub—spontaneous abortion. OK, you seem to have dodged that particular bullet. But you did say you wondered about the long term.
Does your columnist sound skeptical? It's a little more complicated than that.
To be clear: maternal stress can be bad for a fetus. When life gets rough, what are collectively known as stress hormones circulate in the bloodstream. In a pregnant woman, these can be shared with the fetus, possibly affecting brain and body development.
Three areas of the developing brain seem especially sensitive to stress hormones: the hippocampus, which plays a role in memory; the amygdalae, involved in mood and emotional responses; and the frontal cortex, implicated in decision-making and attention.
Sure enough, most gestational stress-related problems reported to date involve intellectual and emotional development. One study found a link between maternal stress during the first trimester and poor attention span and concentration in the resultant offspring. Another found children of highly stressed mothers exhibited more crying, irritability, and temper tantrums as well as ADHD, schizophrenia, and depression.
Some claims are harder to swallow. Several researchers have looked into the consequences of maternal stress due to weather disasters and other unpredictable events, on the theory that these "natural experiments" offer a more objective demonstration of stress. I listed a few such findings above; here are a couple more:
• One study of children of mothers who had experienced high stress while pregnant during a 1998 Quebec ice storm found they had lower IQs and language scores than kids of low-stress moms.
• Another study found children of women who'd lived through a major earthquake during pregnancy had a higher incidence of depression.
At first glance the problems attributed to prenatal stress in these cases seem implausibly diverse: Hurricanes produce autism. Earthquakes lead to depression. Ice storms reduce intelligence. Conceivably each type of natural disaster has a signature outcome, but a simpler explanation is that the results are happenstance and non-confirming findings weren't written up.
One research team (Kinney et al, 2008) suggests the bad things supposedly produced by stress aren't as random as they seem. Rather, they argue, reduced intelligence, poor language skills, depression, and so on are part of a cluster of conditions associated with autism. Among other things, they point to that Louisiana hurricane research, conducted by another Kinney-led team, which found autism diagnoses were significantly higher among children whose mothers had had the severest exposure to storms during several critical months of pregnancy.
But the number of autism cases was tiny—167 children out of more than 300,000 born during hurricanes. (The research focused on storms from 1980 to 1995, before relaxed criteria created a boom in autism diagnoses.) Sure, maybe natural disasters triggered autism in a handful of vulnerable babies, but that just meant a then-rare condition got slightly less so.
This points to a larger problem. Even if all the claimed effects of stress are genuine, so what? No one disputes the general proposition that prenatal trauma can be harmful. Most expectant women already know they shouldn't expose their babies to avoidable everyday stress, and natural disasters and such are usually unpredictable.
Even if we acknowledge that calamities merely highlight the dangers of lesser traumas, telling pregnant women they should avoid having anything go wrong in their lives during gestational months 5, 6, 9, and 10 would surely take the prize for stupid advice. Most children exposed during such times develop normally; any intimation to the contrary would create maternal stress galore, magnifying the problem you were trying to reduce.
We thus find ourselves toying with an odd suggestion for the world of science: By all means find out what you can about the impact of prenatal stress on postnatal development. But if you establish what it looks like you're going to establish, please keep it to yourself. cs