The role of intervening pregnancy loss in the association between interpregnancy interval and adverse pregnancy outcomes

Objective To investigate whether intervening miscarriages and induced abortions impact the associations between interpregnancy interval after a live birth and adverse pregnancy outcomes. Design Population-based cohort study. Setting Norway. Participants A total of 165 617 births to 143 916 women between 2008 and 2016. Main outcome measures We estimated adjusted relative risks for adverse pregnancy outcomes using log-binomial regression, first ignoring miscarriages and induced abortions in the interpregnancy interval estimation (conventional interpregnancy interval estimates) and subsequently accounting for intervening miscarriages or induced abortions (correct interpregnancy interval estimates). We then calculated the ratio of the two relative risks (ratio of ratios, RoR) as a measure of the difference. Results The proportion of short interpregnancy interval (<6 months) was 4.0% in the conventional interpregnancy interval estimate and slightly increased to 4.6% in the correct interpregnancy interval estimate. For interpregnancy interval <6 months, compared with 18-23 months, the RoR was 0.97 for preterm birth (PTB) (95% confidence interval [CI] 0.83-1.13), 0.97 for spontaneous PTB ( 95% CI 0.80-1.19), 1.00 for small-for-gestational age ( 95% CI 0.86-1.14), 1.00 for large-for-gestational age (95% CI 0.90-1.10) and 0.99 for pre-eclampsia (95% CI 0.71-1.37). Similarly, conventional and correct interpregnancy intervals yielded associations of similar magnitude between long interpregnancy interval (>= 60 months) and the pregnancy outcomes evaluated. Conclusion Not considering intervening pregnancy loss due to miscarriages or induced abortions, results in negligible difference in the associations between short and long interpregnancy intervals and adverse pregnancy outcomes.