In the present retrospective study, the timing of pregnancies after BS varied from 3.2 months to 9.5 years. However, 63% of women (71/113) conceived within three years after surgery. The overall outcome in pregnancies conceived during the first postoperative year was good: These women were significantly younger, and the incidences of GDM and delivery induction were significantly lower. The mean newborn weight was slightly lower, but the incidence of preterm delivery did not increase.
In the present study, women with SCTs of less than or more than 12 months had similar characteristics at their BS. No difference was detected in the surgical technique used. As reported in our previous study,28 the incidence of pre-eclampsia after BS was low, with no difference between groups. The studies by Sheiner et al.23 and Dao et al.29 reported concordant results when comparing pregnancies conceived during and after the first postoperative year.
Compared with the 20% incidence of GDM in Finland,30 the incidence of GDM in Group 1 was extremely low (5.9%). We acknowledge that our small sample size is prone to bias; however, a Swedish study by Johansson et al.9 reported a post-BS GDM incidence of 1.9% among women with a median surgery-to-delivery interval of 1.8 years (interquartile range, 1.4–2.5) and a mean presurgery BMI of 43.7 kg/m2. The exact diagnostic criteria for GDM in the study were not reported. Different results were reported in a study by Froylich et al.,24 which showed no difference in the incidence of GDM (one-hour blood glucose level ≥ 7.8 mmol/L after a 50 g dose of oral glucose) among pregnancies that were conceived less than 12 months, 12–24 months or more than 24 months after laparoscopic SG. Additionally, a study by Sheiner et al.23 indicated a nonsignificant difference in the incidence of GDM among women who conceived during (n = 104) or after (n = 385) the first post-BS year (10.5% vs. 7.3%, p = 0.159). In their study, GDM was assessed as glucose intolerance first recognised in the ongoing pregnancy. The difference in BMI before pregnancy and after delivery was not significant. Differences in the diagnostic criteria for GDM may explain these conflicting results, but they also make comparisons among studies difficult. However, we believe that our protocol, which involved one week of glucose level monitoring, provides reliable information on glucose tolerance during pregnancy.
The study results of Shah et al.31 may explain the significant difference in GDM incidence between our study groups; they reported that the glycaemic control of obese patients significantly improved during the first post-BS year. Rapid weight loss is assumed to play an important role, but associated hormonal factors are also important.32 The long-term data of post-BS patients revealed evidence of hyperglycaemia recurrence years after surgery.31 Additionally, the risk of GDM linearly increases with increasing maternal age.33
An increased risk of preterm delivery after BS has been reported in previous studies.16,28,34 Compared with the national incidence of preterm birth (5%),5 the incidence in both study groups (11.8% and 8.3%) and after both surgery techniques (10.0% and 10.8%) was high. Overall, 10 cases of preterm birth were detected, with most of them (7/10) being late preterm deliveries (> 34 gw), including three cases at 36 gw.
In our previous study, we reported an increased risk of planned and unplanned caesarean delivery after BS.28 However, the present study revealed no association with SCT. Similarly, the study by Sheiner et al.23 detected no difference in pregnancies with an SCT of less (mean 8.0 months) or more than 12 months (mean 56.7 months), but a significantly higher risk associated with an SCT of more than 24 months was reported by Froylich et al.24
An extremely high incidence (59.4%) of delivery inductions was detected in Group 2. The mean prepregnancy weight and BMI of women with induced delivery were significantly higher, and the most common indication for induction (38.6%) was GDM. The study by Sheiner et al.23 reported opposite results and detected no difference in delivery induction between pregnancies conceived during and after the first postoperative year. Neither did they detect any difference in maternal age or GDM incidence, which may explain the difference compared with our results. An increased risk of delivery induction after BS was reported in a study by Abenheim et al.,35 which compared post-BS women to morbidly obese, nonoperated pregnant women.
Although the small sample size in our study is prone to bias, the high incidence of inductions must be considered a sign of increased concern among these pregnancies. Because all pregnancies were followed in the same hospital, this cannot be explained by different guidelines for delivery induction. Even though we did not detect induction-associated differences in the incidence of unplanned caesarean deliveries or newborn outcomes, delivery induction means intervening in the normal course of pregnancy and is not without risk.
Previous studies have reported nutritional deficiencies and foetal growth restriction after BS.36,37 Compared with the national data on mean newborn weight (3498 g) and the incidence of low birth weight (4.2%),5 the mean birth weight was lower and the incidence of low birth weight was higher in both groups. In 75% (6/8) of the cases, low birth weight was associated with preterm delivery. The mean birth weight was significantly lower in Group 1, which included two preterm deliveries (at 25 gw and 32 gw) with low birth weight newborns (730 g and 2010 g). However, no difference was detected in the mean SD of birth weight (-0.37 and − 0.33).
Previous results concerning the association between SCT and foetal growth are conflicting. Norgaard et al.38 reported no correlation between the incidence of SGA and SCT, suggesting that the risk of SGA is associated with a history of BS itself rather than with SCT. In contrast, the study by Parent et al.39 reported a greater risk for SGA up to three years after BS and a greater risk of prematurity and neonatal intensive care admission with an SCT of less than two years. Accordingly, an SCT of at least 24 months was suggested in a study by Carreira et al.40 to reduce the risk of SGA. In the present study, the incidence of SGA (< 10th percentile) was not significantly higher in patients with an SCT less than 12 months (23.5% vs. 14.6%). We were unable to show any association between newborn weight and surgery technique, but previous studies reported lower newborn weights after RYGB bypass than after SG.17,18,41 A larger sample size may be necessary to study the impact of different surgery types on foetal growth.
Although guidelines21 and consensus recommendations22 still advise delaying pregnancy for at least 12–18 months after BS, some studies have reported the safety of conceiving within one year.23,38 However, relatively small study populations have prevented strong conclusions from being drawn.42–44 The optimal timing of pregnancy involves balancing the risks of short SCT and risks associated with advancing age and possible weight regain as time after surgery increases. If contraception is not carefully planned, an unplanned pregnancy is also possible when the menstrual cycle returns.45
We acknowledge that the present study has several limitations. The small number of pregnancies with an SCT of less than 12 months is prone to bias. Because women are advised to wait at least one year before conceiving after BS, it is reasonable to assume that the number of these patients is small. A larger study population would have allowed us to study pregnancy and delivery outcomes and outcome trends more reliably. Missing data on surgical techniques may also have caused some bias in the results, which can be considered another weakness of the current study.
High-quality register-based data from the Finnish Institute of Health and Welfare can be considered a strength of the present study. Additionally, the follow-up of pregnancies and treatment of deliveries in the same hospital and with the same principles can be regarded as another strength of the study.