The Prenatal Impacts of Violence

A new study of 500,000 homicides in Brazil finds more evidence that exposure to violence has a negative effect on unborn children.


A recent study published in the Journal of Development Economics provides added weight to existing research suggesting that the stress of living in areas with high rates of homicides can adversely affect the health of an unborn child. The body of work on stress and pregnancy, of course, is well established, but the new research attempted to break new ground by hunting for correlations between data culled from a decade’s worth of birth records, beginning in 2000, and Brazil’s high homicide rate.

The study pins the latter at about 21 per 100,000 people for the span in question, though as the graphic below suggests, rates fluctuate from year to year. (Countries with highest global homicide rates are shown, along with the U.S. for comparison.)

The authors — Martin Foureaux Koppensteiner and Marco Manacorda, both economists — worked from the assumption, supported by the psychological literature, that one doesn’t need to be a victim of crime and violence, or even to merely witness it, for stress and fear levels to be affected. Simply living in a violent milieu, where awareness of nearby, ambient violence is shared among friends, neighbors and coworkers, the authors note, are also known to raise stress levels. Embarking from this premise, the pair analyzed eleven years of Brazilian census data, detailed hospital records, and the precise timing and location of homicides in the country (500,000 over the study timeframe) in an attempt to find links between birth outcomes and violence.

They also made a fine parsing of the available data — selecting, for example, only those homicides that occurred publicly, in the streets (about one-third of the total homicides for the period), as opposed to those that unfolded behind closed doors — in an attempt to increase the likelihood that pregnant women living in that area will have known about it.

By comparing the birth outcomes for those women with those of expecting mothers in the same area who happened to be pregnant when no homicides occurred, the authors say, provided clear evidence that simply being around homicidal violence can impact birth outcomes. Specifically, they found that those who were exposed to a homicide during the first trimester of pregnancy had lower birth weights. On average across the thousands of Brazilian municipalities surveyed, for every additional homicide, women saw a reduction in infant birth weight of around 17 grams — an effect ascribed to the release of stress-related hormones which can negatively impact the fetus.

“We find something very consistent with the medical literature that has looked at stress exposure and birth outcomes,” said Koppensteiner, a lecturer at the University of Leicester, who co-authored the study. As a response to the stress of consistent exposure to violence, he said, not only did birth weight decrease but “these women seem to have shorter gestation and that also leads to an effect on their baby’s health.”

This work adds to a growing body of evidence suggesting that exposure to violence can have a significant impact on unborn children. Studies of birth weights and landmines in Colombia and the Israeli-Palestinian conflict both reached similar conclusions, while research out of Mexico is mixed. The difference, Koppensteiner points out, is that this new study looks at the prenatal effects of consistent, day-to-day violence as opposed to rare, extremely violent events like landmines and terrorist attacks.

It’s important to note, of course, that there are numerous other environmental, biological, and sociological factors that affect maternal and prenatal health, and while the new study’s neighborhood-level investigation brings a high degree of resolution to the correlation between homicide and birth outcomes, more work would be needed to demonstrate a clear causal relationship. There were some curiosities in the data here, too. In looking, for example, at Fortaleza — a large and notoriously violent city in northern Brazil where homicides occur much more regularly than in smaller Brazilian municipalities — the authors found that the effect of homicides on birth weight was significantly smaller. They offered two possible explanations: On the one hand, in areas where murder is more commonplace, pregnant women might simply be less sensitive to ambient violence. Or it may be that in cities with large populations, the authors wrote, “the effect of one homicide gets more easily diluted.”

In any event, Koppensteiner suggests that the study — part of a larger effort organized a few years ago by the Inter-American Development Bank to look into the costs of violence and crime in Latin America and the Caribbean — has implications well outside of Brazil.  “These results extend to settings elsewhere,” he said, “certainly across Latin America where there’s lots of violence.”

See What Others Are Saying

3 comments / Join the Discussion

    Northeastern Brazil corresponds more than half murder in Brazil, while South, Southeast, Midwest has reduced more reasonable levels and other casualties, but northeastern Brazil where it concentrates all.

    I haven’t been able to read the entire study yet, but I was wondering whether the authors were able to parse out factors such as cesarean vs vaginal birthing? I’m sure they did but I wanted to check. There’s a large study called Birth in Brazil that has been looking at a lot of these factors and of course, cesarean birthing in Brazil is extremely prevalent and is related to shorter gestation, baby’s size and health, etc.

    Hey Brena, here’s what Koppensteiner told me: “We have used specification where we control for type of birth (vaginal vs. c-section) and this makes no difference to the estimates. C-section in Brazil indeed deserves an entire research agenda on its own.”

Comments are closed.


& Tipsters

Corruption in science?
Academic discrimination?
Research censorship?
Government cover-ups?

Undark wants to hear about it.

Email us at, or visit our contact page for more secure options.