Discussion
In the entire cohort of pregnant women whose labour was induced, obesity was associated with a higher proportion of caesarean sections than BMI < 25 kg/m2 (reference group): 22 % versus 15 %. There was a significant association between obesity and risk of caesarean section in unadjusted analyses, but this association was no longer statistically significant after adjusting for parity and previous mode of delivery. The rationale for this adjustment is that pregnant women without previous vaginal delivery have an increased risk of caesarean section. Being nulliparous or multiparous without a previous vaginal delivery was more strongly associated with a risk of caesarean section than obesity. Seven out of ten pregnant women with obesity whose labour was induced had a spontaneous vaginal delivery, while six out of ten multiparous women without previous vaginal delivery had a repeat caesarean section after induction of labour.
Pregnant women with obesity whose labour was induced had a higher risk of caesarean section in unadjusted analyses. However, our adjusted analyses showed that the association between obesity and risk of caesarean section was no longer statistically significant, as other researchers have previously reported (13, 14) . Since the change in the estimate of effect was small (OR 1.6–1.5), we cannot rule out pre-pregnancy BMI having an impact or the increased risk being partly explained by one or more confounding variables. Other research has shown that the risk of emergency caesarean section was doubled in pregnant women with obesity compared with those of normal weight, irrespective of whether their labour had been induced or they had previously undergone a caesarean section (12) . Our findings showed that the risk of caesarean section after induction of labour was associated with being nulliparous or multiparous without previous vaginal delivery, irrespective of pre-pregnancy BMI (13, 15) . It is generally important to avoid caesarean section in nulliparous pregnant women, and our findings support this. Several earlier studies excluded pregnant women with a previous caesarean section, and it has not therefore been possible to study the effect of previous mode of delivery on the association between obesity and risk of caesarean section in induced labour (13, 16) .
Pregnant women with overweight and obesity have longer pregnancies and longer labour, in particular a longer latent phase until 6 cm cervical dilation (16–19) . Awareness of this is important in clinical practice to avoid unnecessary interventions.
Prolonged labour in women with obesity can be due to endocrinological and inflammatory changes (18) . An increased percentage of body fat and higher volume of distribution can reduce the efficacy of drugs used to induce labour. It has been shown that higher doses are required for successful induction of labour (19) . Nevertheless, there is no established practice regarding dose adjustments for obesity when inducing labour (8) .
The most common indications for induction of labour (prolonged rupture of membranes, overdue pregnancy and hypertensive pregnancy complications), as well as the methods that were used for induction of labour at the University Hospital of North Norway, Tromsø, were the same as those in a national registration study at 21 Norwegian hospitals in 2018 (8, 11) . All three of these main indications for induction of labour were unevenly distributed between the BMI groups, but only prolonged rupture of membranes was associated with caesarean section. The association between obesity and caesarean section was only marginally affected after adjusting for prolonged rupture of membranes when we tested for confounding by the indication for induction of labour. High BMI alone is not an indication for induction of labour, and the management of pregnant women with a high BMI is therefore no different in the absence of the three main indications mentioned above (8) .
A lack of association for obesity and overweight in the subanalyses is likely due to low statistical power. The data collection was based on a student assignment in 2020, and this limited the scope of the study. Future studies would be strengthened by using national figures. However, our data were obtained from medical records, and this probably resulted in improved recording of BMI compared with the Norwegian Medical Birth Registry, which was missing 9 % of BMI entries in the study period (2) . More recent figures from the Norwegian Medical Birth Registry also provide evidence that our data from 2016–18 are still representative: in 2018 and 2023, the proportion of caesarean deliveries was 16.0 % and 16.2 % respectively nationwide in Norway, and 19.9 % and 19.3 % respectively at the University Hospital of North Norway, Tromsø (2) .
We did not study weight over the course of the pregnancy, only pre-pregnancy BMI. However, overweight and obesity in pregnant women have been reported to be associated with caesarean section (13, 17) . In our analyses, we chose to adjust for height, even though it is part of the BMI calculation, because lower maternal height is associated with an increased risk of caesarean section (20) .
From 2018 to 2023, data from the Norwegian Medical Birth Registry showed a 6 % increase in induced labour both nationally and in Tromsø, as well as a 3–5 % increase in obesity, a one-year increase in the average maternal age, and a 1–2 % increase in nulliparous pregnancies (2) . The increase in induced labour is multifactorial, with possible contributing factors being revised definitions of gestational diabetes (2017) and pre-eclampsia (2020) (8) . Changes in the definitions have resulted in more women being diagnosed and thus an increased chance of being induced. Other relevant factors include changes in protocols for evaluating overdue pregnancies and an increased awareness of the benefits of inducing labour. It is not known to what extent these changes would have affected the results if the study had a more recent and/or larger data set.
Pre-pregnancy BMI recorded on the antenatal health card is based on self-reported pre-pregnancy height and weight, which is cost-effective and practical but entails an increased risk of reporting bias (21) . Nevertheless, earlier studies have shown that self-reported data from the health card is an acceptable and reliable reference for weight (22) . Data on BMI or method of induction of labour were missing from 6 % of pregnant women's antenatal health cards, and these women were excluded from the analyses. We consider the proportion of excluded women to be acceptable, but it may have led to underestimation of the risk of complications associated with overweight and obesity. BMI is a useful tool for estimating weight status in a population, but it does not distinguish between muscle mass and fat mass, which can lead to inaccurate assessments of an individual's weight status in women with high muscle mass. We did not have a control group with spontaneous onset of labour, and it was not therefore possible to study differences in mode of delivery between women whose labour was induced and those whose labour started spontaneously.
Conclusion
Conclusion
Pregnant women with obesity had a higher proportion of caesarean sections than pregnant women without overweight or obesity. This risk of caesarean section following induction of labour was predominantly associated with nulliparous or multiparous women without previous vaginal delivery. A clinical implication of our findings is that particular attention should perhaps be paid to women with a previous caesarean delivery when inducing labour, irrespective of their pre-pregnancy BMI.