Objective
The purpose of this study was to examine associations of fasting C-peptide, body mass index (BMI), and maternal glucose with the risk of preeclampsia in a multicenter multinational study.
Study Design
We conducted a secondary analysis of a blinded observational cohort study. Subjects underwent a 75-g oral glucose tolerance test at 24-32 weeks’ gestation. Associations of preeclampsia with fasting C-peptide, BMI, and maternal glucose were assessed with the use of multiple logistic regression analyses and adjustment for potential confounders.
Results
Of 21,364 women who were included in the analyses, 5.2% had preeclampsia. Adjusted odds ratios for preeclampsia for 1 SD higher fasting C-peptide (0.87 ug/L), BMI (5.1 kg/m 2 ), and fasting (6.9 mg/dL), 1-hour (30.9 mg/dL), and 2-hour plasma glucose (23.5 mg/dL) were 1.28 (95% confidence interval [CI], 1.20–1.36), 1.60 (95% CI, 1.60–1.71), 1.08 (95% CI, 1.00–1.16), 1.19 (95% CI, 1.11–1.28), and 1.21 (95% CI,1.13–1.30), respectively.
Conclusion
Results indicate strong, independent associations of fasting C-peptide and BMI with preeclampsia. Maternal glucose levels (below diabetes mellitus) had weaker associations with preeclampsia, particularly after adjustment for fasting C-peptide and BMI.
Preeclampsia complicates 2–8% of pregnancies worldwide and is associated with increased risk of adverse outcomes for mother and baby. It is a systemic disease that is characterized by increased vascular resistance, endothelial dysfunction, proteinuria, coagulopathy, and hypertension. The pathophysiologic condition is not completely defined but probably includes immune, genetic, and placental abnormalities. Insulin resistance and secretion rise during normal pregnancy; there is growing evidence that preeclampsia is related to increased insulin resistance. There is also strong evidence that women with higher prepregnancy body mass index (BMI) are more likely to experience preeclampsia.
See related editorial, page 207
Most observational studies of gestational diabetes mellitus (GDM) and preeclampsia included an adjustment for maternal BMI, but none of the studies included an adjustment for insulin resistance. Nordin et al showed that lesser degrees of glucose intolerance are associated with the risk of preeclampsia; however, it is not clear whether there was an independent association because this study did not include an adjustment for BMI or other potential confounders. Another study looked at the association with preeclampsia across quartiles of glucose values below those values that are diagnostic of GDM and found a positive association that became nonsignificant after adjustment for confounders. Thus, the nature of the association between lesser degrees of hyperglycemia and preeclampsia remains uncertain.
The objective of the Hyperglycemia and Adverse Pregnancy Outcome (HAPO) study was to assess the risk of adverse pregnancy outcome that is associated with lesser degrees of hyperglycemia than overt diabetes melltius. Participants underwent a 2-hour 75 g oral glucose tolerance test (OGTT). Glucose results were blinded from participants and caregivers, except when results fell outside predetermined ranges.
We previously reported continuous positive associations between maternal glucose values and preeclampsia. In this report, we examine the associations of maternal C-peptide and BMI with preeclampsia and more fully examine the associations of maternal glucose with preeclampsia, including adjustment for fasting C-peptide.
Materials and Methods
The HAPO study was an international, multicenter epidemiologic study. Detailed methods have been reported. The study was approved by the local institutional review board at each center. All participants gave written informed consent, and the study was overseen by an external Data Monitoring Committee.
Participants
All pregnant women at each center were eligible to participate, unless they had ≥1 exclusion criteria. Methods to determine gestational age and expected date of delivery have been described.
OGTT
Participants underwent a standard 75-g OGTT between 24 and 32 weeks’ gestation (as close to 28 weeks as possible). Data concerning smoking and alcohol use, first-degree family history of diabetes mellitus and hypertension, and demographic data were collected with the use of standardized questionnaires. Race/ethnicity was self-identified by participants.
A sample for random plasma glucose was collected at 34-37 weeks’ gestation as a safety measure to identify cases with hyperglycemia that were above a predefined threshold.
Glucose analysis and unblinding
Aliquots of fasting and 2-hour OGTT and random plasma glucose samples were analyzed at field center laboratories. Values were unblinded if the fasting plasma glucose (FPG) level exceeded 105 mg/dL (5.8 mmol/L), if the 2-hour OGTT plasma glucose level exceeded 200 mg/dL (11.1 mmol/L), if the random plasma glucose level was ≥160 mg/dL (8.9 mmol/L), or if any plasma glucose value was <45 mg/dL (2.5 mmol/L). Otherwise, women, caregivers, and the HAPO study staff (except laboratory personnel) remained blinded to glucose values. All OGTT specimens were analyzed at the HAPO Central Laboratory, and the results are used here. Only women whose results remained blinded, with no additional glucose testing outside the HAPO study protocol, are included in these analyses.
Fasting C-peptide
A serum sample for fasting C-peptide was collected at the OGTT visit. Because hemolysis is known to increase insulin degradation but not to affect C-peptide and because C-peptide and insulin are secreted in equimolar amounts, we measured C-peptide rather than insulin. Levels of fasting insulin or C-peptide or derived variables that also include fasting glucose concentration (eg, homeostatic model assessment) are used commonly as an index of insulin sensitivity. We used fasting C-peptide as the index so that we could look at the association of fasting C-peptide and glucose separately. C-peptide measurements were performed on an Autodelfia instrument (Perkin-Elmer Life Sciences, Boston, MA). The functional sensitivity of the assay is 0.02 ng/mL with intra- and interassay coefficients of variation of 3.2-5.0% and 1.9-3.0%, respectively, in samples with high and low C-peptide concentrations.
Maternal height and weight
Maternal height and weight, which were measured at the OGTT visit, were used to calculate maternal BMI. Height was measured twice to the nearest 0.5 cm with a stadiometer or wall-mounted measuring tape with shoes removed and the participant’s head facing forward in the horizontal plane. If results differed by >1.0 cm, measurements were repeated. Weight was measured twice to the nearest 0.1 kg on a calibrated scale with outer garments and shoes removed and repeated if results differed by >0.5 kg. Recalled maternal prepregnancy weight was also recorded but was not the focus of this report because of its inherent subjectivity and the absence of data for 1966 participants (8.4%). No center provided specific interventions to participants based on weight or BMI.
Prenatal care and delivery
Prenatal care and timing of delivery were determined by standard field center practice. No field center arbitrarily delivered patients before full term or routinely performed cesarean delivery at a specified maternal or gestational age. Medical records were abstracted to obtain data regarding maternal and delivery course.
Classification of preeclampsia and other hypertensive disorders of pregnancy
Blood pressure was measured at the OGTT visit by standardized procedures and a calibrated electronic device (Omron 711; Omron Healthcare, UK LTD, Buckinghamshire, UK). Additional blood pressure measurements were abstracted from medical records. Hypertension that was present at <20 weeks gestation that did not progress to preeclampsia was classified as chronic hypertension. After 20 weeks’ gestation, blood pressure measurements were used to classify hypertensive disorders in pregnancy according to the International Society for the Study of Hypertension in Pregnancy guidelines. A woman with a systolic blood pressure of ≥140 mm Hg and/or a diastolic blood pressure ≥90 mm Hg on ≥2 occasions a minimum of 6 hours apart and proteinuria of ≥1+ dipstick or ≥300 mg per 24 hours was classified as having preeclampsia. If the criteria for elevated blood pressure, but not proteinuria, were met, this was classified as gestational hypertension.
Statistical analysis
Descriptive statistics include means and SDs for continuous variables and numbers and percentages for categoric variables. For associations of fasting C-peptide, BMI, and fasting, 1-hour, and 2-hour plasma glucose levels with preeclampsia, each predictor was considered as both a categoric and continuous variable in multiple logistic regression analyses. In categoric analyses, each predictor was divided into 7 categories, with approximately 50% of all values in the 2 lowest categories and 3% and 1% in the 2 highest categories, respectively. Categories for FPG were prespecified in 5-mg/dL increments as <75, 75–79, 80–84, 85–89, 90–94, 95–99, and ≥100 mg/dL, respectively. The highest and second-highest categories, which accounted for 1% and 3% of participants, respectively, were chosen specifically to allow for assessment of whether there were threshold effects. Categories for fasting C-peptide, BMI, and 1- and 2-hour plasma glucose were also selected so that similar proportions fell into each of the 7 categories.
For analyses of fasting C-peptide, BMI, and glucose as continuous variables, odds ratios (ORs) and 95% confidence intervals (CIs) were calculated for each predictor higher by 1 SD. To assess whether the log of the odds of preeclampsia was related linearly to each of these variables, we added squared terms in each predictor. Because of the large sample size in the HAPO study and the large number of women with preeclampsia, squared terms were considered statistically significant only for a probability value of < .001.
For each predictor, 3 logistic models (I, II, and III) were fit. Model I included adjustment only for field center. Model II included adjustment for multiple potential prespecified confounders that included maternal age, height, smoking, alcohol use, family history of diabetes mellitus, family history of high blood pressure, gestational age at the OGTT, baby’s sex, parity (0, 1, 2+), and any maternal urinary tract infection during the pregnancy. For fasting C-peptide and measures of glycemia, model II also included adjustment for BMI and for analyses that focused on BMI as the predictor of interest; model II also included adjustment for FPG. For analyses of fasting C-peptide, model III included further adjustment for fasting glucose. For BMI and measures of glycemia, model III included further adjustment for fasting C-peptide.
All analyses were conducted in statistical software (SAS version 9.1; SAS Institute Inc, Cary, NC; or Stata version 10.0; Stata Corporation, College Station, TX).
Results
There were 1116 cases of preeclampsia among the 23,316 blinded participants. In addition, 582 women had chronic hypertension, and 1370 women experienced gestational hypertension. Only the women who remained normotensive (20,248) or who experienced preeclampsia are included in these analyses. Characteristics of these 21,364 women are shown in Table 1 . Maternal age, BMI, and glucose values at the OGTT and gestational age at delivery are similar to those that have been reported for the entire cohort. Our primary objective was to determine associations of glucose, BMI, and C-peptide with risk of preeclampsia. Because of center-by-center differences in frequency of preeclampsia and other variables (maternal age, BMI, parity), a comparison of characteristics of those women who experienced preeclampsia with those women who remained normotensive optimally would use a case-control design that is outside the scope of this report.
| Characteristic | n | Measure |
|---|---|---|
| Mother b | ||
| Age, y c | 21,364 | 29.2 ± 5.8 |
| Body mass index (kg/m 2 ) c , d | 21,364 | 27.4 ± 4.9 |
| Fasting plasma glucose, mg/dL c , d , e | 21,364 | 80.7 ± 6.8 |
| Plasma glucose, mg/dL c , d , e | ||
| 1-hr | 21,364 | 133.4 ± 30.8 |
| 2-hr | 21,364 | 110.5 ± 23.4 |
| Fasting C-peptide, μg/L c , d | 21,141 | 1.9 ± 0.8 |
| Gestational age, wk c , d | 21,364 | 27.8 ± 1.8 |
| Prenatal smoking (any) | 1429 | 6.7% |
| Prenatal alcohol use (any) | 1476 | 6.9% |
| Family history of diabetes mellitus | 4795 | 22.5% |
| Family history of hypertension | 7427 | 34.8% |
| Parity (previous delivery ≥20 wk) | 11,246 | 53.1% |
| Prenatal urinary tract infection | 1483 | 6.9% |
| Newborn infant | ||
| Gestational age, wk c | 21,364 | 39.4 ± 1.7 |
| Sex: male | 10,974 | 51.4% |
a Hypertension that was present at <20 weeks’ gestation that did not progress to preeclampsia was classified as chronic hypertension. After 20 weeks’ gestation, hypertensive disorders in pregnancy were categorized according to International Society for the Study of Hypertension guidelines. Preeclampsia was defined as systolic blood pressure of ≥140 mm Hg and/or diastolic blood pressure of ≥90 mm Hg on ≥2 occasions a minimum of 6 hours apart and proteinuria of ≥1+ dipstick or ≥300 mg per 24 hours. If the criteria for elevated blood pressure, but not proteinuria, were met, this was classified as gestational hypertension;
b Participants with chronic or gestational hypertension were excluded;
c Measurement data are given as mean ± SD;
d Measured at the oral glucose tolerance test visit;
Table 2 shows associations between maternal fasting C-peptide and preeclampsia and includes ORs and 95% CIs for each category compared with the lowest or referent category. A fasting C-peptide level was available for 21,141 of the 21,364 women with either no hypertension or preeclampsia. The frequency of preeclampsia rose from 1.7% in the lowest category (≤1.2 ug/L) of fasting C-peptide to 16.9% in the highest (≥4.8 μg/L). Odds ratios rose across categories in each of the models and were highest in the second-highest category. With adjustment for confounders that included BMI (model II), there was substantial attenuation. However, there was very little additional attenuation with further adjustment for FPG (model III). With fasting C-peptide considered as a continuous variable, ORs for C-peptide that were higher by 1 SD (0.87 ug/L) were 1.61, 1.32, and 1.28, in the 3 models, respectively. In each model, there was a significant nonlinear association of fasting C-peptide with preeclampsia ( P < .001, when C-peptide squared was added to the model).
