Objective
Investigate associations between preâpregnancy participation and performance in a demanding crossâcountry ski race (proxy for exercise volume and fitness) and perinatal outcomes. Preâregistered protocol: osf.io/aywg2.
Design
Prospective cohort study.
Setting
Based on entire overlap between the Vasaloppet registry and the populationâbased Swedish Pregnancy Register.
Sample
All female Vasaloppet participants 1991â2017 with subsequent singleton delivery (skiers), and ageâ and countyâmatched nonâskiers.
Methods
We calculated odds ratios (ORs) for nonâskiers versus skiers (model 1) and, among skiers, by performance (model 2), in Bayesian logistic regressions adjusted for socioâdemographics, lifestyle factors, and comorbidities. We repeated calculations adjusting for early pregnancy body mass index (potential mediator) and explored robustness (selection/exposure settings; multiple comparisons correction).
Main outcome measures
Twentyânine important perinatal outcomes, predefined based on existing expert consensus.
Results
Nonâskiers (n = 194â384) versus skiers (n = 15â377) (and slower versus faster performance, not shown) consistently had higher odds of gestational diabetes mellitus (GDM) (OR 1.70, 95% highest density interval: 1.40â2.09), excessive gestational weight gain (GWG) (1.28, 1.22â1.38), psychiatric morbidity (1.60, 1.49â1.72), any caesarean section (CS) (1.34, 1.28â1.40), elective CS (1.39, 1.29â1.49), and largeâforâgestationalâage babies (>90th percentile, 1.11, 1.04â1.18); lower odds of inadequate GWG (0.83, 0.79â0.88); and no associations with fetal/neonatal complications (e.g. preterm birth [1.09, 0.98â1.20], small for gestational age [SGA] [1.23, 1.05â1.45]). Adjustment for body mass index attenuated associations with excessive (1.20, 1.14â1.30) and inadequate GWG (0.87, 0.83â0.92) and large for gestational age (1.07, 1.00â1.13).
Conclusion
Nonâskiers compared with skiers, and slower versus faster performance, consistently displayed higher odds of GDM, excessive GWG, psychiatric morbidity, CS and largeâforâgestationalâage babies; and lower odds of inadequate GWG, after adjustment for socioâdemographic and lifestyle factors and comorbidities. There were no associations with fetal/neonatal complications.