Within-Family Analysis of Interpregnancy Interval and Adverse Birth Outcomes.

OBJECTIVE

To examine associations among interpregnancy interval, the duration from the preceding birth to the conception of the next-born index child, and adverse birth outcomes using designs that adjust for measured and unmeasured factors.

METHODS

In this prospective cohort study, we used population-based Swedish registries from 1973 to 2009 to estimate the associations between interpregnancy interval (referent 18-23 months) and adverse birth outcomes (ie, preterm birth [less than 37 weeks of gestation], low birth weight [LBW; less than 2,500 g], small for gestational age [SGA; greater than 2 SDs below average weight for gestational age]). Analyses included cousin and sibling comparisons and postbirth intervals (ie, the interval between secondborn and thirdborn offspring predicting secondborn outcomes) to address unmeasured familial confounding.

RESULTS

Traditional cohort-wide analyses showed higher odds of preterm birth (adjusted odds ratio [OR] 1.51, 99% CI 1.39-1.63, 5.99% preterm births]) and LBW (adjusted OR 1.25, 99% CI 1.13-1.39, 3.32% LBW) after a short interpregnancy interval (0-5 months) compared with offspring born after an interpregnancy interval of 18-23 months (3.21% preterm births, 1.92% LBW). Except for preterm birth (adjusted OR 1.72, 99% CI 1.26-2.35), associations were attenuated in cousin comparisons. A small association between a short interpregnancy interval and preterm birth remained in sibling comparisons (adjusted OR 1.22, 99% CI 1.11-1.35), but associations with LBW (adjusted OR 0.83, 99% CI 0.74-0.94) and SGA (adjusted OR 0.74, 99% CI 0.64-0.85) reversed direction. For pregnancy intervals of 60 months or more, odds of preterm birth (adjusted OR 1.51, 99% CI 1.43-1.60, 5.07% preterm births), LBW (adjusted OR 1.61, 99% CI 1.50-1.73, 3.43% low-birth-weight births), and SGA (adjusted OR 1.54, 99% CI 1.42-1.66, 2.49% SGA births) were also higher when compared with the reference interval (1.53% SGA). Associations between long interpregnancy interval and adverse birth outcomes remained through cousin and sibling comparisons. Postbirth interval analyses showed familial confounding is present for short interpregnancy intervals, but supported independent associations for long interpregnancy intervals.

CONCLUSION

Familial confounding explains most of the association between a short interpregnancy interval and adverse birth outcomes, whereas associations with long interpregnancy intervals were independent of measured and unmeasured factors.