Skip to content

Adverse Prenatal Exposures Sharply Augments Risk of Psychiatric Symptoms in Childhood

Key findings

  • This analysis of the Adolescent Brain and Cognitive Development Study linked adverse prenatal exposure burden to psychiatric symptoms at age 9 or 10 in 9,290 participants
  • Six prenatal exposures were independently associated with small but significant increases in total symptom scores on the Child Behavior Checklist (CBCL): unplanned pregnancy; maternal use of alcohol, tobacco or marijuana before pregnancy was recognized; and pregnancy or birth complications
  • 44% of children had two or more such exposures, and a greater number of exposures linearly increased the odds of CBCL total score ≥60, the threshold for clinically meaningful symptoms (OR, 1.39 per exposure; P<0.001)
  • 29% of children with to four or more exposures exhibited CBCL total ≥60 compared to 7% of children with no exposures
  • Promoting healthy pregnancies may also promote the offspring's brain health, and future studies might determine that early psychosocial interventions such as enriched learning environments, can temper the risk conferred by adverse prenatal exposures

For decades, evidence has been accumulating that adverse prenatal exposures, such as fetal exposure to maternal substance use, increase a child's subsequent risk of psychiatric disorders. However, the effect of multiple types of adverse exposures is unclear. Furthermore, it's been hard to disentangle adverse prenatal exposures from postnatal exposures that influence symptom risk, such as caregiver income and education.

The Adolescent Brain and Cognitive Development (ABCD) Study, the largest study in the U.S. ever to assess brain development, is designed to address such questions. Supported by the National Institutes of Health, the study has 21 sites and has enrolled 11,875 children, nine or ten years old, and will follow them into young adulthood. The sample is geographically diverse and reflects the nation's socioeconomic, racial and ethnic diversity.

Using baseline data from this study, Joshua L. Roffman, MD, MMSc, director of the Mass General Early Brain Development Initiative, and co-director of Mass General Neuroscience, and Erin C. Dunn, ScD, MPH, an investigator in the Psychiatric and Neurodevelopmental Genetics Unit at Massachusetts General Hospital, and colleagues identified prenatal exposures that increase the risk of childhood psychopathology and found that the greater the number of exposures, the greater the jeopardy. Their report appears in PLoS One.

Study Methods

The researchers reviewed in-person intake data on 9,290 non-adopted participants in the ABCD study who had valid scores on the Child Behavior Checklist (CBCL). The children were categorized as having (n=1,392) or not having (n=7,898) a sibling participating in the study. The non-sibling group was tested initially and the sibling group was used as a validation set.

The CBCL assesses symptoms in 10 domains (anxious/depressed, withdrawn/depressed, somatic complaints, social problems, thought problems, attention problems, aggressive behavior, rule-breaking behavior, internalizing behavior and externalizing behavior). A total score ≥60 indicates clinically meaningfully symptoms. 87% of study participants had a normal score.

Single Exposures

Six prenatal exposures were independently associated with small but significant increases in CBCL total score:

  • Unplanned pregnancy
  • Maternal use of alcohol, tobacco or marijuana before pregnancy was recognized ("early" exposure)
  • Pregnancy complications
  • Birth complications

When these six factors were tested in the validation set of siblings, similar effects were observed.

Two other factors tested were unassociated with CBCL score (preterm birth and caesarian section) and several others were included in the analysis but occurred with insufficient frequency for conclusive findings (early maternal use of cocaine or opiates; maternal use of alcohol, tobacco, marijuana, cocaine or opiates after pregnancy was recognized) preterm birth; and caesarian section.

Multiple Exposures

Considered alone, none of the six adverse factors was statistically important. However, 44% of children had at least two exposures. Exposure to multiple factors increased the odds of CBCL score ≥60 (P<0.001 for all of the following comparisons):

  • 2 exposures vs. none—OR, 1.86
  • 3 exposures—OR, 2.76
  • ≥4 exposures—OR, 3.53
  • Linear effect—OR per exposure, 1.39

These cumulative effects of exposure burden were calculated after adjustment for other exposures known to influence risk for psychiatric symptoms, including (but not limited to) maternal age at birth, caregiver socioeconomic status, neighborhood safety, presence of a caregiver partner and presence of an older sibling.

In absolute terms, 29% of children exposed to four or more factors exhibited CBCL total ≥60 compared to 7% of children with no exposures. Exposure burden had analogous effects across all CBCL domains.

Within some pairs of siblings, the number of exposures was different. The sibling with a higher exposure burden had a higher CBCL score, suggesting that findings in the larger non-sibling group were not confounded by unmeasured family-level factors.

Potential for Prevention

Policies and interventions that promote healthy pregnancies may be able to protect the offspring's brain health in childhood. Identifying, developing and implementing these interventions is the primary goal of the Mass General Early Brain Development Initiative, an ambitious program founded by Dr. Roffman that comprises investigators from the Department of Psychiatry, Department of Obstetrics & GynecologyDepartment of Medicine and the MassGeneral Hospital for Children.

greater risk of psychopathology in children who had three adverse prenatal exposures

greater risk of psychopathology in children who had four or more adverse prenatal exposures

of children exposed to four or more adverse prenatal exposures had psychopathology

Learn more about the Department of Psychiatry

Explore research in the Department of Psychiatry


At least five developmental properties of teeth suggest they could be used to identify individuals at risk of mental health problems following exposure to early-life psychosocial stress.