Intrauterine cocaine exposure and the risk for sudden infant death syndrome: a meta-analysis

J Perinatol. 1997 May-Jun;17(3):179-82.


Objective: To determine whether an association exists between intrauterine cocaine exposure and sudden infant death syndrome (SIDS).

Study design: A meta-analysis of 10 published studies that reported the incidence of SIDS in infants born to mothers who used cocaine during pregnancy. We computed the rates of SIDS in cocaine-exposed infants for each of the 10 studies and in comparison groups for each of the eight studies that reported the incidence of sudden infant death syndrome in comparison groups. We also computed odds ratios and 95% confidence intervals individually for the eight studies with comparison groups. Then we computed adjusted overall odds ratios (Mantel-Haenszel) by combining the eight studies with comparison groups; combining three of these studies in which the comparison group infants were exposed to illicit drugs other than cocaine (polydrug controls); and combining the other five studies in which the comparison group infants were not exposed to illicit drugs (drug-free controls).

Results: For all 10 studies combined, SIDS occurred in 84 of 12,163 infants with intrauterine cocaine exposure; the variance weighted estimate for incidence was 5.2 per 1000 with a 95% confidence interval of 4 and 7 per 1000. The combined odds ratio for SIDS in cocaine-exposed versus all comparison group infants was 3.9 (95% confidence interval 3 and 5), a significantly increased risk. However, in a subset analysis we found that the odds ratio for SIDS in cocaine-exposed versus the polydrug comparison group infants was 2.7 (95% confidence interval 0.9 and 8.2), not a significantly increased risk. Whereas in a subset analysis comparing risk for SIDS in cocaine-exposed versus the drug-free comparison group infants, the odds ratio was 4.1 (95% confidence interval 3.2 and 5.3), indicating a highly significant effect of cocaine exposure on SIDS risk.

Conclusion: After we controlled for the confounding variable of concurrent use of other drugs, the increased risk for SIDS could not be attributed to intrauterine cocaine exposure alone. The increase in risk for SIDS was found not to be specific to cocaine but to intrauterine exposure to illicit drugs in general.

Publication types

  • Meta-Analysis
  • Research Support, Non-U.S. Gov't

MeSH terms

  • Cocaine*
  • Confidence Intervals
  • Female
  • Humans
  • Infant
  • Infant, Newborn
  • Odds Ratio
  • Pregnancy
  • Pregnancy Complications*
  • Prenatal Exposure Delayed Effects*
  • Substance-Related Disorders / complications*
  • Sudden Infant Death / etiology*


  • Cocaine