Maternal and neonatal outcomes of pregnancies after metabolic bariatric surgery: a retrospective population-based study

Abstract

Background: The incidence of post metabolic bariatric surgery (BS) pregnancies is rising. Previous studies provided conflicting results regarding the risk of prematurity, stillbirth, perinatal death and the optimal time from BS to conception. This study examined maternal and neonatal outcomes of post-BS pregnancies and influencing factors.

Methods: Nationwide retrospective study of all post-BS pregnancies in France from January 1st 2013 to December 31st 2022. We compared 55,941 post-BS pregnancies with 223,712 controls matched on delivery date, parity, age, obesity, hypertension, diabetes, and socio-economic status (1:4 ratio) using generalized estimating equations. We also compared 11,777 post-BS pregnancies with 11,777 pre-BS pregnancies in the same women, using conditional logistic regression. Maternal outcomes included gestational hypertension, preeclampsia, and gestational diabetes. Neonatal outcomes included small-for-gestational-age (SGA), prematurity, stillbirth, and perinatal death. We tested for interactions with BS type, BS to pregnancy time interval and malnutrition.

Findings: Post-BS pregnancies were associated with reduced risk of gestational hypertension (odds ratio [OR] 0.57 [95% CI 0.53-0.62]), preeclampsia (OR 0.59 [0.55-0.64]), and gestational diabetes (OR 0.64 [0.62-0.66]) as compared with control. Similar but stronger risk reductions were observed compared with pre-BS pregnancies. Risk of SGA was increased (OR 1.74 [1.68-1.79]) as compared with controls and pre-BS pregnancies (OR 1.88 [1.64-2.16]). Risk of prematurity was increased (OR 1.27 [1.22-1.31]) as compared with controls but not pre-BS pregnancies (OR 0.95 [0.85-1.06]). Compared with controls, risk of stillbirth was increased (OR 1.2 [1.06-1.35]), mediated by SGA, as was perinatal death (OR 1.5 [1.13-1.99]), mediated by both prematurity and SGA. Increased SGA risk compared to controls was higher with malnutrition (OR: 2.38 [1.96, 2.88], p_interaction <0.0001), with <6 months (OR: 1.95 [1.72, 2.21], p_interaction = 0.01) or 6-12 months between BS and pregnancy (OR: 1.86 [1.70, 2.04], p_interaction = 0.02) and with gastric bypass (OR: 1.88 [1.77-2.00], p_interaction = 0.027). Increased prematurity risk compared to controls was higher with malnutrition (2.45 [1.99, 3.00], p_interaction <0.0001) and gastric bypass (OR: 1.46 [1.36-1.57], p_interaction = 0.0003).

Interpretation: Post-BS compared with pre-BS or control pregnancies were associated with reduced risk of maternal adverse outcomes but increased risk of neonatal adverse events. The risks of SGA and prematurity are higher with shorter intervals between BS and conception, gastric bypass, and malnutrition. Post-BS pregnancies could be considered high risk, requiring close nutritional and obstetrical monitoring.

Funding: Support from INSERM and the French Ministry of Health (Messidore 2022 n°97).

Keywords: Bariatric surgery; Gastric bypass; Gestational diabetes; Gestational hypertension; Malnutrition; Neonatal death; Nutrition; Obesity; Preeclampsia; Pregnancy; Prematurity; Sleeve gastrectomy; Small-for-gestational age.

Fig. 1

Study flow chart. BS: metabolic bariatric surgery; T2D: type 2 diabetes mellitus; T1D: type 1 diabetes; HTA: arterial hypertension; SNDS: French National Health Data System.

Fig. 2

Adverse pregnancy and neonatal outcomes associated with post-bariatric surgery (BS) pregnancies. Occurrence of adverse pregnancy and neonatal outcomes of post-BS vs. matched control pregnancies (a) and pre-BS control pregnancies (b). Odds ratios are calculated from generalized estimated equations adjusted for age (continuous), health coverage assistance, deprivation index quintiles, pregnancy obesity status, sleep apnea syndrome and medically assisted reproduction for that pregnancy in a and from conditional logistic regression with the mother ID as a stratum in b. Confidence intervals are enlarged proportionally to the p-value increase after multiple comparison adjustment by using the false discovery rate method. BS: metabolic bariatric surgery; W: weeks of gestation.

Fig. 3

Surgical and nutritional factors associated with adverse pregnancy and neonatal outcomes. Occurrence of adverse pregnancy and neonatal outcomes of post-BS vs. matched control pregnancies (significant in the whole population) stratified by BS type (a), time from BS to pregnancy (b), obesity class before BS (c) and malnutrition status from BS to delivery (d). Odds ratios are calculated from generalized estimated equations adjusted for age (continuous), health coverage assistance, deprivation index quintiles, pregnancy obesity status, sleep apnea syndrome and medically assisted reproduction for that pregnancy. Confidence intervals are enlarged proportionally to the p-value increase after multiple comparison adjustment by using the false discovery rate method. Interactions for each subgroup were tested with sleeve gastrectomy as a reference in a, 2–5 years post-BS in b, BMI 30–40 kg/m² in c and no malnutrition from BS to pregnancy in d. ∗ p_interaction <0.05; ∗∗ p_interaction <0.01; ∗∗∗ p_interaction <0.001. False discovery rate multiple comparison adjustment was applied for calculating p_interaction values. BS: metabolic bariatric surgery; W: weeks after last menstrual period; ref: reference.

Fig. 4

Main factors associated with small-for-gestational-age (SGA) and prematurity in post-BS pregnancies. Factors selected in a lasso regularization model associated with SGA (a) and prematurity (b), among post-BS pregnancies. Beta coefficients from the regularization model are displayed. A negative coefficient implies a protective association, and a positive coefficient implies a risk association. BS: metabolic bariatric surgery.