Gestational diabetes mellitus (GDM) is a common complication of pregnancy, affecting 2–5% of all gravidas in the United States.1 It has been defined as carbohydrate intolerance first diagnosed during pregnancy.2 Specifically, GDM is a condition of maternal hyperglycemia that develops subsequent to abnormal insulin production (inadequate pancreatic β-cell response to increasing insulin resistance) and decreased skeletal muscle glucose uptake.3 The placenta has been implicated as the source of hormones that propagate insulin resistance during pregnancy. Progesterone, estrogen, placental lactogen, and human chorionic somatomammotropin have been shown to affect glucose metabolism and increase insulin resistance.4–6 In particular, the diabetogenic properties of progesterone have been reported in several investigations in both pregnant and nonpregnant women.7–14
Based on the findings of a study performed by the Eunice Kennedy Shriver National Institute of Child Health and Human Development Maternal-Fetal Medicine Units Network,15 17-alpha-hydroxyprogesterone caproate therapy has become a common treatment for the prevention of recurrent preterm birth. However, limited data are available regarding potential adverse effects related to17α-hydroxyprogesterone caproate treatment. A recent retrospective study by Rebarber et al16 found that the use of 17α-hydroxyprogesterone caproate for prevention of recurrent preterm birth was associated with an increased risk of developing GDM.
Given the issues raised by the investigation of Rebarber et al, the intent of the current investigation was to evaluate the frequency of abnormal glucose testing and GDM among women receiving 17α-hydroxyprogesterone caproate and compare the findings with those of a group of 17α-hydroxyprogesterone caproate–unexposed controls matched by maternal age and body mass index (BMI). Unlike previous investigations, we sought to estimate the effect of 17α-hydroxyprogesterone caproate on maternal glucose testing within BMI categories.
MATERIALS AND METHODS
This is a retrospective cohort study of women treated with weekly 17α-hydroxyprogesterone caproate at MetroHealth Medical Center, an urban tertiary care institution affiliated with Case Western Reserve University, with institutional review board approval. The primary outcome was the mean venous plasma glucose level (mg/dL) obtained 1 hour after the 50-g oral glucose screen performed as part of routine screening for gestational diabetes. Secondary outcomes included an abnormal 1-hour glucose screen and the diagnosis of GDM. An abnormal 1-hour 50-g screen was defined as a venous plasma glucose level of at least 135 mg/dL 60 minutes after a nonfasting 50-g oral glucose load. Gestational diabetes mellitus was diagnosed in any patient with a 1-hour 50-g glucose screen of at least 200 mg/dL or if two or more abnormal results were identified on a confirmatory 3-hour 100-g oral glucose tolerance test according to the Carpenter-Coustan criteria.17
We queried our electronic medical records to identify all women treated with 17α-hydroxyprogesterone caproate because of a history of a prior preterm birth. Individual chart review was performed to determine maternal prepregnancy weight, height, age, and screening 1-hour glucose test results. Women with a history of diabetes mellitus before pregnancy, those with multiple gestations in the incident pregnancy, and those who did not have a screening 1-hour 50-g glucose test were excluded. A history of GDM in a prior pregnancy was not part of our exclusion criteria. Prepregnancy BMI was calculated using the last recorded maternal weight (kg) before pregnancy and the maternal height (m2) at the time of the first prenatal care visit. Each 17α-hydroxyprogesterone caproate–exposed patient was matched randomly with three unexposed controls by maternal age (±1 year) and prepregnancy BMI (±2). For one woman with extreme obesity (BMI 73), matching was within the same BMI category (obese class III). Controls were excluded if they were receiving 17α-hydroxyprogesterone caproate in the incident pregnancy, had a history of diabetes mellitus before pregnancy, were carrying a multiple gestation, or if they did not have a 1-hour 50-g glucose screen. Random assignment was achieved using electronic random assignment for all women delivering at this institution during the same time period (2001–2008), with the cohort then sorted by maternal BMI and age. The first three controls meeting inclusion criteria for each case were selected. If a control participant did not meet inclusion criteria, the next eligible control was selected.
Mean values of the 1-hour 50-g glucose screen, the frequency of an abnormal screening test, and the frequency of GDM were compared between women exposed to 17α-hydroxyprogesterone caproate and controls. Patients then were classified by BMI category as follows: underweight (lower than 18.5), normal (18.5 to 24.9), overweight (25.0 to 29.9), obese class I (30.0 to 34.9), obese class II (35.0 to 39.9), obese class III (40.0 or higher).18 Mean results of the 1-hour 50-g glucose screen were compared by 17α-hydroxyprogesterone caproate exposure within each BMI category. Conditional multiple logistic regression was performed to evaluate the effect of 17α-hydroxyprogesterone caproate exposure on the diagnosis of an abnormal glucose screen and GDM after controlling for potentially confounding variables including maternal age, BMI, parity, and race.
A review of our records identified 110 eligible 17α-hydroxyprogesterone caproate–exposed patients. A 3:1 control:17α-hydroxyprogesterone caproate–exposed patient ratio was chosen to enhance the power given the limited number of available cases. Assuming 20% of pregnant women would have a positive 50-g glucose screening result with a screen positive cutoff of 135 mg/dL,17 we determined that our sample size was sufficient to evaluate a 70% increase in the incidence of an abnormal glucose screen in the 17α-hydroxyprogesterone caproate group with 80% power and alpha=0.05. Data were evaluated by χ2, analysis of variance, and t-test where appropriate. Conditional multiple logistic regression was performed to control for confounding variables. P<.05 was considered significant. All statistical analyses were performed using Statview 5.01 (Cary, NC), except for the conditional multiple logistic regression analysis, which was performed using StatsDirect 2.7.2 (Cheshire, UK).
We identified 110 women treated with 17α-hydroxyprogesterone caproate and 330 unexposed controls who met the inclusion criteria for our investigation. Table 1 presents the maternal characteristics for the cohort by 17α-hydroxyprogesterone caproate exposure. As expected, there were no differences in maternal age at delivery, prepregnancy BMI, or ethnicity by 17α-hydroxyprogesterone caproate exposure. Women treated with 17α-hydroxyprogesterone caproate were not nulliparous and all had a prior preterm birth. Figure 1 presents the distribution of the entire study cohort by BMI category. Only 39% of patients had normal BMIs; few women were underweight (0.9%). The remaining 60% were overweight or obese. There were no differences in the distribution of BMI categories by 17α-hydroxyprogesterone caproate exposure (P=.99).
The primary outcome data are presented in Table 2. The mean results of the 1-hour 50-g glucose screen were significantly different by 17α-hydroxyprogesterone caproate exposure, with the 17α-hydroxyprogesterone caproate–exposed group having a higher mean result (115.6 compared with 105.4 mg/dL, P<.001). Additionally, the frequency of an abnormal 1-hour 50-g glucose screen (23.6% compared with 11.2%, P=.001) as well as the diagnosis of GDM (10.9% compared with 3.6%, P=.003) were significantly higher in the 17α-hydroxyprogesterone caproate–exposed patients. Evaluation of the mean results of the 1-hour 50-g glucose screen as well as the frequency of an abnormal 1-hour screen within each BMI category are presented in Figures 2 and 3, respectively. Analysis of variance demonstrated a significant difference between 17α-hydroxyprogesterone caproate exposure and control groups and the mean 1-hour 50-g glucose results (mg/dL) regardless of BMI category (P=.01), although significant differences between 17α-hydroxyprogesterone caproate–exposed patients and control participants were not seen within individual subgroups. The 1-hour 50-g glucose screen was more frequently abnormal in 17α-hydroxyprogesterone caproate–exposed women within each BMI category, reaching statistical significance in the obese I (38.1% compared with 15.9%, P=.03) and obese III (36.4% compared with 6.1%, P=.01) subgroups (Fig. 3). The frequency of GDM by 17α-hydroxyprogesterone caproate exposure within each BMI category is presented in Figure 4. Gestational diabetes mellitus was more frequent in the 17α-hydroxyprogesterone caproate–exposed patients within a majority of the BMI categories, reaching statistical significance in the obese III (18.2% compared with 0%, P=.01) and overweight subgroups (15.4% compared with 2.6%, P=.02).
Using conditional multiple logistic regression, we examined the effect of 17α-hydroxyprogesterone caproate exposure on both abnormal 1-hour 50-g glucose screens and the diagnosis of GDM (Table 3). For this analysis, we controlled for maternal prepregnancy BMI and maternal age (years) at delivery as continuous variables and African–American race and parity as categorical variables. 17α-hydroxyprogesterone caproate–unexposed participants served as the reference group. In this analysis, 17α-hydroxyprogesterone caproate exposure was associated with increased odds of an abnormal 1-hour 50-g glucose screen (odds ratio 2.7, 95% confidence interval 1.4–5.1) and GDM (odds ratio 3.3, 95% confidence interval 1.3–8.0).
In this retrospective cohort study of abnormal glucose testing with 17α-hydroxyprogesterone caproate exposure, we found a significant association between 17α-hydroxyprogesterone caproate treatment and elevated mean values of a 1-hour 50-g glucose screen. Additionally, we found a higher frequency of an abnormal 50-g glucose screen as well as an increased frequency of gestational diabetes among 17α-hydroxyprogesterone caproate–exposed women. These findings persisted after controlling for maternal age at delivery, prepregnancy BMI, a history of preterm birth, and maternal race.
Gestational diabetes mellitus is a common complication of pregnancy, affecting approximately 2–5% of all gravid women in the United States.1 It is associated with significant maternal, fetal, neonatal, and infant morbidity. Women with GDM are at higher risk for hypertensive disorders, infections, labor abnormalities, cesarean or operative vaginal delivery, and the development of type-2 diabetes later in life. Increased risks for the fetus and neonate include macrosomia, shoulder dystocia, birth trauma, hypoglycemia, hyperbilirubinemia, poor feeding, and neonatal intensive care unit admission.
Several of the hormones produced by the placenta, including estrogen, placental lactogen, human chorionic somatomammotropin, and progesterone, have been implicated in the development of insulin resistance and subsequent hyperglycemia. In particular, progesterone has been shown to reduce insulin sensitivity through several pathways. Pregnant rats infused with exogenous progesterone demonstrate a decrease in insulin sensitivity7 with subsequent hyperglycemia.8 Additionally, several authors have reported on the association of progestin-only contraception and increased insulin resistance.9,10 These changes could be attributed to progesterone-mediated reductions in GLUT-4 expression in skeletal muscle and adipose tissue11–14 or impairment of insulin release.8 Investigations by Cambell et al demonstrate that progesterone can reduce the expression of GLUT-4 in skeletal muscle and adipose tissue11–14 with subsequent decreases in glucose uptake. Additionally, Picard et al report that exogenous progesterone exposure in mice produced hyperglycemia and a higher susceptibility to frank diabetes. In this same investigation, the authors also report that progesterone receptor knockout mice had improved glucose homeostasis, most likely owing to improvements in insulin release from the pancreas.
The findings of these and other studies are of particular importance given the use of exogenous 17α-hydroxyprogesterone caproate for the prevention of recurrent preterm birth. To date, scant data have shown any adverse effects of therapy with 17α-hydroxyprogesterone caproate. However, one previous retrospective investigation does show a higher risk of GDM in women exposed to 17OPHC.16 Other investigations of nonpregnant women also have demonstrated an association between exogenous progesterone use and both insulin resistance and diabetes mellitus. The findings of the current investigation corroborate these results.
Our study has several limitations. Most importantly, as a retrospective chart review, this study is open to ascertainment bias. Additionally, patient compliance with 17α-hydroxyprogesterone caproate was not documented in the medical record. Therefore, it is unclear what percentage of 17α-hydroxyprogesterone caproate patients were taking their therapy at the time of the 1-hour 50-g glucose screen. However, we assume noncompliance to treatment would have tended to mask evident effects. Finally, the frequency of an abnormal 1-hour glucose screen in the control group was less than anticipated; however, the frequency of GDM in our control participants was consistent with our expectations. Regardless, we conclude that 17α-hydroxyprogesterone caproate treatment is associated with increased odds of developing GDM.
1. American Diabetes Association. Diagnosis and classification of diabetes mellitus. Diabetes Care 2008;31:S55–60.
2. Metzger BE, Buchanan TA, Coustan DR, De Leiva A, Dunger DB, Hadden DR, et al. Summary and recommendations of the Fifth International Workshop-Conference on Gestational Diabetes Mellitus [published erratum appears in Diabetes Care 2007;30:3154]. Diabetes Care 2007;30:S251–60.
3. Gabbe SG, Graves CR. Management of diabetes mellitus complicating pregnancy. Obstet Gynecol 2003;102:857–68.
4. Ryan EA, Enns L. Role of gestational hormones in the induction of insulin resistance. J Clin Endocrinol Metab 1988;67:341–7.
5. Kalkoff RK, Kissebah AH, Rim H-J. Carbohydrate and lipid metabolism during normal pregnancy: relationship to gestational hormone action. In: Merkatz IR, Adam PAJ, editors. The diabetic pregnancy: a perinatal perspective. New York (NY): Grune & Stratton; 1979. p. 3–21.
6. Barbier RL. Endocrine disorders in pregnancy. In: Yan SSC, Jaffe RB, Barbier RL, editors. Reproductive endocrinology. 4th ed. Philadelphia (PA): WB Saunders; 1999.
7. González C, Alonso A, Alvarez N, Díaz F, Martínez M, Fernández S, et al. Role of 17beta-estradiol and/or progesterone on insulin sensitivity in the rat: implications during pregnancy. J Endocrinol 2000;166:283–91.
8. Picard F, Wanatabe M, Schoonjans K, Lydon J, O’Malley BW, Auwerx J. Progesterone receptor knockout mice have an improved glucose homeostasis secondary to beta-cell proliferation. Proc Natl Acad Sci U S A 2002;99 (24):15644–8.
9. Xiang AH, Kawakubo M, Kjos SL, Buchanan TA. Long-acting injectable progestin contraception and risk of type 2 diabetes in Latino women with prior gestational diabetes mellitus. Diabetes Care 2006;29:613–7.
10. Kahn HS, Curtis KM, Marchbanks PA. Effects of injectable or implantable progestin-only contraceptives on insulin-glucose metabolism and diabetes risk. Diabetes Care 2003;26:216–25.
11. Campbell SE, Febbraio MA. Effect of the ovarian hormones on GLUT4 expression and contraction-stimulated glucose uptake. Am J Physiol Endocrinol Metab 2002;282:E1139–46.
12. Campbell SE, Febbraio MA. Effect of ovarian hormones on mitochondrial enzyme activity in the fat oxidation pathway of skeletal muscle. Am J Physiol Endocrinol Metab 2001;281:E803–8.
13. Sugaya A, Sugiyama T, Yanase S, Shen XX, Minoura H, Toyoda N. Expression of glucose transporter 4 mRNA in adipose tissue and skeletal muscle of ovariectomized rats treated with sex steroid hormones. Life Sci 2000;66:641–8.
14. Sutter-Dub MT, Kaaya A, Sfaxi A, Sodoyez-Goffaux F, Sodoyez JC, Sutter BC. Progesterone and synthetic steroids produce insulin resistance at the post-receptor level in adipocytes of female rats. Steroids 1988;52:583–608.
15. Meis PJ, Klebanoff M, Thom E, Dombrowski MP, Sibai B, Moawad AH, et al. Prevention of recurrent preterm delivery by 17 alpha-hydroxyprogesterone caproate [published erratum appears in N Engl J Med 2003;349:1299]. N Engl J Med 2003;348:2379–85.
16. Rebarber A, Istwan NB, Russo-Stieglitz K, Cleary-Goldman J, Rhea DJ, Stanziano GJ, et al. Increased incidence of gestational diabetes in women receiving prophylactic 17alpha-hydroxyprogesterone caproate for prevention of recurrent preterm delivery. Diabetes Care 2007;30:2277–80.
© 2009 by The American College of Obstetricians and Gynecologists.
17. Moore TR, Catalano PM. Diabetes in pregnancy. In: Creasy RK, Resnick R, Iams JD, editors. Creasy & Resnick’s maternal fetal medicine principles and practice. 6th ed. Philadelphia (PA): Saunders; 2009.