Meta-analysis of the association between preterm delivery and intelligence

Abstract

Background: An increasing proportion of infants are born preterm, and their survival has improved. Therefore, their long-term sequelae are of increasing public health importance.

Methods: We conducted a systematic review covering a 30 year period (1980-2009). A random effects meta-analysis provided a pooled estimate of the difference in IQ score between individuals born preterm and term. Small-study bias was examined using a funnel plot and Egger's test, and meta-regression was used to investigate possible causes of heterogeneity. Cumulative meta-analysis was used to determine if the magnitude of the association had changed over time.

Results: The 27 eligible studies covered 7044 individuals; 3504 (50%) delivered preterm and 3540 (50%) at term. They provided 37 estimates of difference in IQ. All demonstrated a reduced IQ among those delivered preterm and all but four reached statistical significance. Overall, IQ score was 11.94 (95% CI: 10.47-13.42, P < 0.001) points lower among children born preterm. There was moderate heterogeneity (overall I(2) 74.2%, P < 0.001), but no significant small-study bias (P = 0.524). The association between preterm delivery and IQ did not change significantly over time. There was a statistically significant, linear association across the gestational age range (adjusted coefficient: -0.91, 95% CI: -1.64, -0.17, P = 0.018).

Conclusions: There is a strong and consistent body of evidence suggesting an association between preterm delivery and reduced IQ, with evidence of a dose-response relationship with gestational age.