Bariatric Surgery and Consequences for Mother and Baby in Pregnancy (BAMBI)

The Effect of Gastric Bypass Surgery on Glucose Metabolism, Gestational Weight Gain and Fetal Growth in Subsequent Pregnancy

Background Roux-en-Y gastric bypass (RYGB) is a well-established treatment of obesity, most often performed in women during their reproductive years. Adverse events related to RYGB include hypoglycemia. Though usually attenuated in pregnancy, the incretin response is reinforced in subjects with RYGB and the resulting changes in insulin and glucagon responses together with the resultant weight loss are possible underlying mechanisms for hypoglycemia. The majorities of women who have undergone RYGB conceive shortly after RYGB and have an increased risk for inappropriate gestational weight gain (GWG) and thereby fetal growth restriction. However, studies of hypoglycemia and GWG in pregnant women following RYGB are lacking. Objective In women with previous RYGB we aim to investigate a) glucose level and incretin response during a mixed meal test (MMT) in early and late pregnancy, b) trimester specific incidence of postprandial hypoglycemia and c) fetal growth. Methods 20 women with RYGB and 20 age-, BMI- and parity-matched controls will be studied with a) 2nd and 3rd trimester 4-hour liquid MMTs, b) 6-days Continuous Glucose Monitoring (CGM) once every trimester and post partum and c) maternal and fetal anthropometrics including antenatal ultrasound examinations and neonatal DXA-scans. The primary outcomes are nadir plasma glucose levels during the 4-hour liquid MMT, number of hypoglycemic episodes during CGM and birthweight standard deviation scores. Discussion A better understanding of maternal metabolism and fetal growth in women with RYGB will support risk stratification, patient information and management both before and during pregnancy.

Background Post-prandial hypoglycemia Roux-en-Y Gastric Bypass (RYGB) is a well-established treatment of severe obesity and is performed in women most often during their reproductive years. RYGB is a hormonal, malabsorptive as well as a restrictive surgical procedure, leading to complications such as vitamin deficiencies and postprandial hypoglycemia, which are exaggerated during pregnancy. Hypoglycemia usually does not occur until one year post surgery and evolves gradually with non-specific symptoms, causing delayed diagnosis. Symptoms of hypoglycemia include autonomic symptoms (palpitations, lightheadedness, sweating) and neuroglycopenic symptoms (confusion, decreased attentiveness, seizure, loss of consciousness). While severe hypoglycemia is estimated with a frequency of 0.2 % following RYGB, the frequency of mild to moderate hypoglycemia is unknown due to a lack of recognition of symptoms. Insulin sensitivity increases due to weight loss and the resulting changes in insulin and glucagon response are a possible underlying mechanism for postprandial hypoglycemia. In pregnant women, insulin sensitivity changes from early to late pregnancy. Whilst usually attenuated in late pregnancy, the incretin response is reinforced in subjects with RYGB. These factors make it difficult to predict the risk of hypoglycemia in pregnancies following RYGB. Studies have shown that plasma insulin concentration is inappropriately elevated during the hypoglycemic episodes and that fasting hypoglycemia is uncommon, suggesting that hypoglycemia is primarily associated with postprandial dysregulation of insulin secretion. Beta-cell diameter has been shown to correlate with pre-operative Body Mass Index (BMI). Accordingly, hypertrophy of the beta-cells in the pancreas has previously been proposed as an explanation of hyperinsulinemic hypoglycemia. However, pancreatic resection has not provided a cure for hypoglycemia following RYGB. In a recent study, administration of a Glucagon Like Peptide-1 (GLP-1) receptor antagonist was shown to eliminate postprandial hypoglycemia in patients with symptomatic hypoglycemia following RYGB. In this study altered gastro-intestinal function was suggested to contribute to the altered glucose metabolism in combination with dysregulation of insulin secretion. Hence, there is no general consensus on the subject. No evidence-based guidelines for the diagnosis of postprandial hypoglycemia or investigation of glucose-metabolism in RYGB patients exist. Continuous Glucose Monitoring (CGM) has proven to provide high detection rates for hypoglycemia under real life conditions and can therefore be used as an important tool for screening of hypoglycemia. Currently, the preferred method for diagnosis of postprandial hypoglycemia is the mixed meal test (MMT), as it provides a more physiological stimulus compared to the oral glucose tolerance test (OGTT). While the OGTT consists of ingestion of a liquid containing 75 g glucose, the MMT consists of a liquid meal of 55 % carbohydrate, 25 % protein and 20 % fat. Both tests are performed after fasting with blood samples drawn during the test measuring glucose, insulin and c-peptide. Knowledge of predictive factors for postprandial hypoglycemia in pregnancy will be an important tool for improving dietary management and quality of life in women with RYGB. Fetal growth During the last decades, prevalence's of obesity and diabetes have reached epidemic proportions - even in children. This is only partly explained by sedentary life-style and unfavorable diet. Exposure to obesity and/or a diabetic intrauterine environment is a risk to the fetus and seems to program long-term effects. Studies have consistently shown that offspring of diabetic mothers have an increased risk of macrosomia and obesity and risk of metabolic syndrome in later life. These observations have contributed to the theory of the so-called "vicious intergenerational circle of obesity". Large epidemiological studies such as the Dutch Famine Study have documented a number of adverse metabolic effects in adults with low birthweight. Thus, both intrauterine over- and under-nutrition have severe consequences for fetal growth and future health. However, it is not known whether repeated events of maternal hypoglycemia per se are harmful to the fetus. Fetal growth restriction is a potential risk in pregnancies with limited Gestational Weight Gain (GWG) or even weight loss - i.e. the first period of time following bariatric surgery and certain eating disorders. Data from our own clinic show that 40% of women with previous RYGB fail to fulfill minimum recommendations for GWG. A recent registry study from Sweden found that pregnancies in women with bariatric surgery were associated with risk of Small for Gestational Age (SGA) infants. Supporting this, a study in Hvidovre in 25 term infants reported lower birthweight, fat free mass and fat percentage in offspring of mothers with RYGB compared to controls. Furthermore, inappropriate GWG might be associated with hyperketonaemia which have potential adverse effects on offspring CNS. These adverse effects call for further exploration of the fetal growth and the need of defining optimal post-surgery timing of pregnancy, GWG, diet, vitamin supply etc. Given the harmful effects of both intrauterine over- and under-nutrition, the time around pregnancy is the window of opportunity to change factors with long-term impact on the child. Aims In women with RYGB the investigators aim to study: Glucose and incretin response during a 4-hour liquid MMT in early and late pregnancy Trimester-specific incidence of hypoglycemia Fetal growth Study design Longitudinal, prospective cohort study Subjects Study participants will be enrolled in early pregnancy at Departments of Endocrinology and Departments of Gynecology and Obstetrics at Odense University Hospital and Sydvestjysk Sygehus. 20 pregnant women with RYGB (n = 20) will be matched on age, prepregnancy-BMI and parity to 20 pregnant women without RYGB. Exclusion criteria for GB and non-GB group multiple pregnancy age below 18 and above 45 years ongoing smoking and substance abuse severe psychiatric disorder severe chronic disease diabetes women with overt diabetes at inclusion (HbA1c ≥ 48 mmol/l and/or fasting p-glucose ≥ 7 mmol/l) women with pre-gestational diabetes (type1 or 2) prior to RYGB women with Gestational Diabetes Mellitus (GDM) in a previous pregnancy will not be excluded Methods During a two-year period the investigators aim to include 20 pregnant women with RYGB, 20 age-, BMI- and parity-matched controls and the 40 offspring of these women. A 2-hour 75 g OGTT, as described in the Background section, will be performed in the control group in week 24 to detect GDM. Postprandial hypoglycemia At inclusion (gestational week 12-14) and at week 34 a 4-hour liquid MMT with concomitant mea-surements of glucose-, insulin-, glucagon and GLP-1 levels by blood samples will be performed in all participants. Glucose levels during everyday life will be assessed in each trimester and 4-6 weeks post-partum by 6 days CGM combined with Self-Monitoring of Blood Glucose (SMBG) for calibration of CGM system (24 hours). CGM consists of a sensor inserted under the skin to measure glucose in the interstitial fluid surrounding skin cells. One sensor lasts 6 days. The sensor measures glucose every 5-10 seconds averaging the values every 5 minutes and storing the data in the monitor's memory. The device is equipped with an alarm, which will be turned off, blinding the women. CGM provides information of fluctuations over time whereas SMBG only provides a snapshot. Hypoglycemia will be defined as blood glucose <3.0 mmol/L. The women will report any hypoglycemic symptoms. Fasting blood beta-Hydroxybutyrate will be measured at home each morning before breakfast during 6 days CGM with registration of nightly meals. Maternal clinical information on time for RYGB, weight trajectories (before RYGB, pre-gestational, gestational, post-partum), post-operative and pregnancy related complications, blood pressure etc. will be gathered. Fetal growth Antenatal serial ultrasound measurements will be performed at gestational week 24, 28 and 34. Measurements will include abdominal circumference (AC) and calculation of fetal weight deviation. Postnatally, anthropometric measures including birthweight and length, AC, skinfold measurements and DXA-scans within 72 hours after birth will be performed. DXA-scans will be performed, when the newborn has fallen asleep after breastfeeding/formula feeding. The examination will last for about 5 minutes and is not associated with any pain or discomfort. Perspectives Due to new national guidelines for bariatric surgery the investigators expect an increase in pregnant women having undergone this operation. A better understanding of maternal metabolism and fetal growth in women with previous RYGB surgery will support risk stratification and development of treatments both before and during pregnancy. This study will contribute with valuable knowledge about the body composition of children of women with RYGB. The children will be compared with children of weight-matched, obese and lean controls thereby increasing the understanding of how the mothers' nutritional status affect the children's body composition. Power Calculations To our knowledge, there are no published studies on effect of a MMT in pregnant women after RYGB. Based on MMT's in women before and one year after RYGB a SD of 0.5 in BG is expected. In order to provide a power greater than beta=80% (type 2-error), and a significance of alfa=5% (type 1-error) for observing a difference of (MIREDIF) = 0.5 mmol a sample size of 16 women should be included in each group. Due to the unknown variance parameters regarding fetal growth and body composition, the investigators plan to include 20 women in each group. Ethics All projects described have been approved by the local Ethics Committee (project ID S-20160134) and will be performed in accordance with the principles in Helsinki Declaration II. The DXA-scanner used in this study will be a Hologic 4500A (Waltham, MA, USA) which applies low radiation (approximately 0.02 mSv per scan). The lifetime risk of developing radiation-induced lethal cancer is 0.012% per mSv for children. Thus, the exposure of 0.02 mSv onto 1 million children could lead to 2 deaths by cancer as a consequence of radiation. The radiation is negligible in comparison with the background radiation of 4 mSv a year and the method is classified as category 1, which does not poses a health hazard for the neonate (effective dose < 0.03 mSv). The examination is not associated with any pain or discomfort for the neonate. Handling and archiving data The protocol will be registered on the Region of Southern Denmark's record over processing activities and the documents and materials related to the clinical study will be stored in REDCap and OPEN Analyse according to the Danish Data Protection Legislation and the EU GDPR (General Data Protection Regulation). Finance The study is 100 % non-profit, completely independent from commercial interests. There will be applied for public and private grants to finance the study. The primary investigator and the supervisors have taken the initiative to the study. The primary investigator has no association to the funders. There will not be provided salary for participation in the study. The study participants will be shown gratitude by dispensing printed ultrasound photos at the extra ultrasound examinations.

20 pregnant women with previous gastric bypass surgery. During pregnancy the women will be tested with mixed meal test and continuous glucose monitoring for the diagnosis of hypoglycemia. Ultrasounds of fetal growth will be performed and after birth anthropometrics of the newborn will be meaured including a DXA-scan to estimate bodycomposition.

20 pregnant women matched on age, prepregnancy-BMI and parity (n = 20). During pregnancy the women will be tested with mixed meal test and continuous glucose monitoring for the diagnosis of hypoglycemia. Ultrasounds of fetal growth will be performed and after birth anthropometrics of the newborn will be meaured including a DXA-scan to estimate bodycomposition.

All women will have mixed meal test, continous glucose monitoring and antenatal ultrasounds performed. The children born of these women will have DXA-scans performed after birth.

Inclusion Criteria: GB group: pregnant women with RYGB (n = 20) non-GB group: pregnant women matched on age, prepregnancy-BMI and parity (n = 20) neonates: offspring of abovementioned women (n = 40) Exclusion Criteria: multiple pregnancy age below 18 and above 45 years ongoing smoking and substance abuse severe psychiatric disorder severe chronic disease diabetes women with overt diabetes at inclusion (HbA1c ≥ 48 mmol/l and/or fasting p-glucose ≥ 7 mmol/l) women with pre-gestational diabetes (type1 or 2) prior to RYGB women with Gestational Diabetes Mellitus (GDM) in a previous pregnancy will not be excluded