Optimal Glycemic Control, Pre-eclampsia, and Gestational Hypertension in Women With Type 1 Diabetes in the Diabetes and Pre-eclampsia Intervention Trial

The study population comprised 762 women with type 1 diabetes recruited from 25 joint antenatal-metabolic clinics across Northern Ireland, Scotland, and northwest England between April 2003 and June 2008 into the Diabetes and Pre-eclampsia Intervention Trial (DAPIT) (19). DAPIT was a multicenter, randomized, placebo-controlled intervention trial of vitamin C and E supplementation to prevent pre-eclampsia in pregnant women with type 1 diabetes. Women were enrolled between 8 and 22 weeks’ gestation. Eligibility criteria included type 1 diabetes preceding pregnancy, a singleton pregnancy, and being aged ≥16 years (19). Women with chronic hypertension were included in the trial.

As previously described (19), pre-eclampsia was defined as gestational hypertension with proteinuria for previously normotensive women, according to the International Society for the Study of Hypertension in Pregnancy guidelines (4) and according to National High Blood Pressure Education Program Working Group's guidelines for women with pre-existing hypertension and/or proteinuria (5). Each case of hypertensive pregnancy was confirmed by three senior clinicians, acting independently. Pre-eclampsia and gestational hypertension could be assessed for 749 (98%) women whose pregnancies progressed to at least 20 weeks’ gestational age.

Prepregnancy and first antenatal visit glycosylated hemoglobin (A1C) values (measured locally using Diabetes Control and Complications Trial–aligned methods as part of routine care) were abstracted by trained researchers from the women’s hospital records and recorded on study-specific case report forms. Prepregnancy A1C values were those measured up to 6 months prior to pregnancy, as recorded in the booking history by the attending physician or on recall from the patient. The first antenatal visit A1C values were measured at a median of 9 weeks’ gestation (95% by the end of week 15) at the joint antenatal-metabolic clinic. Study-specific peripheral venous blood samples for A1C were collected at 26 (±2) and 34 (±2) weeks’ gestation and stored immediately at −70°C until analysis. Samples were batch analyzed centrally in the Nutrition and Metabolism Laboratories, Centre for Public Health, Queen’s University Belfast, at the end of the trial. A1C (Diazyme Laboratories, Poway, CA) was measured by spectrophotometry using an automated ILab600 biochemical analyzer. As a National Glycohemoglobin Standardisation Programme– and International Federation for Clinical Chemistry–certified method, the values reported were aligned with the Diabetes Control and Complications Trial system, with intra- and interassay coefficients of variation values <2%. A1C results were arbitrarily categorized as optimal (<6.1%: the referent value), good (6.1–6.9%), moderate (7.0–7.9%), and poor (≥8.0%) glycemic control. The West Midlands Multicentre Research Ethics Committee provided ethical approval (MREC 02/7/016).

Group comparisons were performed using independent-samples t tests and χ² tests. Logistic regression was used to estimate the odds of pre-eclampsia and gestational hypertension in women with poor, moderate, and good control relative to women with optimal control both before and after adjustment for potentially confounding variables, including treatment group, center, BMI, diabetes duration, parity, current smoking, age, aspirin consumption, microalbuminuria before pregnancy, low serum α-tocopherol, and low plasma ascorbate at randomization (or plasma ascorbate level in the 26- and 34-week analyses). Similar analyses were conducted using uncategorized A1C results and were used to derive estimates of the odds of pre-eclampsia and gestational hypertension for each 1% decrement in A1C. All statistical analyses were performed using SPSS software, version 17 (SPSS, Chicago, IL).

Pre-eclampsia and gestational hypertension developed in 17 and 11% of 749 pregnancies, respectively (19). Seventy-two percent of women (n = 542) reported an A1C result measured in the 6 months prior to pregnancy. A1C results were available for 96, 79, and 69% of participants at the first antenatal visit, at 26 weeks’ gestation, and at 34 weeks’ gestation, respectively. Maternal characteristics and glycemic control of women with and without pre-eclampsia and of women with and without gestational hypertension (excluding women with pre-eclampsia) are shown in Table 1. The clinical characteristics of the two groups were similar, with the exception that significantly more women with pre-eclampsia were primiparous compared with women without pre-eclampsia (P < 0.001); primiparity did not differ significantly between those with and without gestational hypertension. Although not significant, more women without pre-eclampsia were current smokers. Women with pre-eclampsia had significantly poorer glycemic control both before and during pregnancy compared with women without pre-eclampsia, whereas glycemic control did not differ between women with or without gestational hypertension (Table 1). In women who developed pre-eclampsia, there was no significant difference in glycemic control between women with and without essential hypertension before pregnancy or between women with and without pre-existing proteinuria (data not shown).

The frequency of subjects by category of glycemic control before pregnancy, at the first antenatal visit, at 26 weeks’ gestation, and at 34 weeks’ gestation is shown in Table 2. Only 7% of women had optimal glycemic control (A1C <6.1%) at their first antenatal visit, increasing to 23 and 25% at 26 and 34 weeks’ gestation, respectively. At 34 weeks’ gestation, 22% of women had an A1C ≥7.0% and 4% had a value of ≥8.0%.

Data on the risk of pre-eclampsia and gestational hypertension in relation to glycemic control throughout pregnancy are shown in Table 2. Glycemic control was not significantly associated with gestational hypertension either before or during pregnancy. There was a progressive increase in the odds of pre-eclampsia with worsening glycemic control before pregnancy relative to the referent A1C <6.1%; although none of the individual odds ratios reached statistical significance, this may have reflected the small number of subjects in the reference category (Table 2). In early pregnancy (median 9 weeks’ gestation), there was a trend for increasing pre-eclampsia with worsening glycemic control, and poor glycemic control (A1C ≥8.0%) was associated with a significantly increased risk of pre-eclampsia (odds ratio 3.68 [95% CI 1.17–11.6]) compared with optimal control. At both 26 and 34 weeks’ gestation, there was a progressive and significant increase in the odds of pre-eclampsia for suboptimal control (A1C ≥6.1%) compared with the referent value (A1C <6.1%), with odds ratios ranging from 2.09 to 3.81 and 1.78 to 8.01, respectively. For each 1% decrement in A1C at the first antenatal visit, at 26 weeks’ gestation, and at 34 weeks’ gestation, the risk of pre-eclampsia was significantly reduced (0.75 [0.64–0.88], 0.57 [0.42–0.78], and 0.47 [0.31–0.70], respectively), with a nonsignificant reduction in risk for the preconception A1C of 0.88 (0.75–1.03) (Table 3). A test for interaction between parity (primiparous or multiparous) and A1C in the logistic model did not show evidence that the change in pre-eclampsia risk per 1% decrement in A1C differed between primiparous and multiparous women. The effect of a 1% decrement in A1C during pregnancy was examined in two separate logistic models containing measurements: 1) at the first antenatal visit and 26 weeks’ gestation and 2) at the first antenatal visit and at 34 weeks’ gestation, fitted simultaneously to predict the risk of pre-eclampsia. Although the numbers of patients with complete data were reduced, in both models A1C at the first antenatal visit remained a significant independent predictor of pre-eclampsia. Controlled for A1C at the first antenatal visit, A1C levels at 34 weeks’ gestation remained significantly predictive of pre-eclampsia with only slightly attenuated odds ratios for a 1% decrement (Table 4), but A1C levels at 26 weeks’ gestation were no longer significant (P = 0.07).

The overall incidence of pre-eclampsia and gestational hypertension reported in this population of women with type 1 diabetes was 17 and 11%, respectively, in agreement with previous reports (1–3). The data indicate that poor glycemic control during pregnancy is associated with the development of pre-eclampsia but not with gestational hypertension. Furthermore, we have shown, for the first time, an association between A1C values before pregnancy and the development of pre-eclampsia. These results are of clinical importance and suggest that optimal glycemic control (A1C <6.1%) during early pregnancy and later in pregnancy may reduce the risk of pre-eclampsia in women with type 1 diabetes.

Our results are consistent with other data in showing a clear association between glycemic control in early pregnancy and pre-eclampsia (10,11). Conflicting findings may possibly be explained by other studies being generally smaller (8,9) and by the timing of the first antenatal visit A1C measurement; A1C values in the study reported here and in that of Hanson et al. (10) were measured at a mean of 9 weeks’ gestation, whereas Temple et al. (8) reported their first-visit A1C levels at 7 weeks’ gestation. However, Hiilesmaa et al. (11) also found a positive association between pre-eclampsia and A1C levels at 7 weeks’ gestation. Our data relating glycemic control in mid- to late-pregnancy with pre-eclampsia concur with some reports (8,9,20), although others (8) found no association between A1C values and pre-eclampsia in the third trimester.

We have shown that higher A1C values during pregnancy are independently predictive of pre-eclampsia. International recommendations for optimal A1C values during pregnancy are inconsistent. Our data show that a reduction of 1% in A1C at any time during pregnancy is associated with a significantly reduced risk of pre-eclampsia, pointing to a benefit of even a modest reduction in A1C at any time for the individual patient. Our data also suggests that optimal A1C (<6.1%) in late pregnancy may be of particular importance in risk prediction; for each 1% decrement in A1C the lowest odds ratios for pre-eclampsia were observed in the third trimester. Contrary to other recommendations (14,15), and consistent with American Diabetes Association guidelines (16), our findings support the use of A1C both to monitor glycemic control and also to assess pre-eclampsia risk throughout pregnancy. However, in the interpretation of A1C, cognizance needs to be taken of the reduction in A1C levels during pregnancy observed in nondiabetic women of similar age (18), which may help explain why in our study, optimal A1C (<6.1%) in late pregnancy was associated with the lowest risk of pre-eclampsia.

The DAPIT population is the largest contemporary prospective dataset of women with type 1 diabetes in the U.K. (19). This cohort consists of a carefully characterized population of women with type 1 diabetes, in whom comprehensive data about pregnancy progress and pregnancy outcomes has been collected. Furthermore, a major strength of this study is the extent to which the International Society for the Study of Hypertension in Pregnancy pre-eclampsia definition criteria have been applied, with each case of hypertensive pregnancy undergoing rigorous review and the diagnosis of pre-eclampsia confirmed by three senior clinicians, thus minimizing potential confounding by chronic hypertension or pre-existing nephropathy. Another strength of this study is the availability of serial A1C values during pregnancy and rigorous control for relevant confounding variables. Furthermore, although prepregnancy A1C values were collected by recall, to the best of our knowledge this is the first study to date to report such data in relation to the risk of pre-eclampsia and adds to the current knowledge on the relevance of prepregnancy planning to pregnancy outcome.

Nevertheless, our study has a number of limitations. First, not all blood samples were centralized for A1C analyses. Although samples for A1C determination at 26 and 34 weeks’ gestation were collected by trained study personnel, according to a strict protocol, and batch analyzed centrally using Diabetes Control and Complications Trial–aligned methodology, A1C values for up to 6 months before pregnancy and at the first antenatal visit were measured in local laboratories where different A1C assays (Diabetes Control and Complications Trial aligned) may have been used. In addition, prepregnancy A1C values were collected from a review of patient notes and patient interviews, which is highly dependent on patient recall. However, frequently this is the best and only prepregnancy clinical information available, and although subject to recall bias, it is conceivable that patients may have recalled more favorable values than the true value, thus tending to diminish the association between A1C and pre-eclampsia. Finally, women in this study were those who consented to participate in DAPIT and thus may not be totally representative of the entire population of women with type 1 diabetes. It should be noted that the DAPIT intervention (vitamins C and E) did not significantly reduce the incidence of pre-eclampsia, and both treatment groups were therefore included in a single analysis, albeit with adjustment for treatment group, low plasma ascorbate, and serum α-tocopherol at randomization (or plasma ascorbate level in the 26- and 34-week analyses) in the logistic regression model.

In line with most other studies, we found no impact of glycemic control on gestational hypertension (10,11), which might lend support to the previously proposed divergent etiologies for pre-eclampsia and gestational hypertension (21,22). Although the pathophysiology of both conditions remains unclear (22), these data add to increasing evidence linking glycemic control with pre-eclampsia in women with type 1 diabetes (8–11) and indeed in women with only minor degrees of hyperglycemia during pregnancy (23,24). As a concept, the role of oxidative stress in the development of pre-eclampsia remains persuasive, but the lack of randomized trial evidence demonstrating any reduction in incidence with specific antioxidant therapy rather points to a multifactorial etiology in which poor glycemic control (possibly compounded by vitamin depletion) (19) remains among the most readily identifiable and treatable risk factors.

The results of this study highlight the importance of assessing glycemic control throughout pregnancy to reduce the incidence of pre-eclampsia. The 26 and 34 weeks’ gestation A1C values were most predictive of pre-eclampsia, with only minor attenuation after controlling for A1C at the first antenatal visit. On the other hand, after controlling for A1C values later in pregnancy, A1C at the first antenatal visit remained an independent predictor of pre-eclampsia, lending additional support to the importance of periconceptual glycemic control for maternal fetal outcome. Although women who went on to develop pre-eclampsia reported higher A1C values in the 6 months prior to pregnancy, additional prospective research is needed to evaluate the impact of glycemic control before pregnancy on pregnancy outcomes such as pre-eclampsia.

This study was funded by grants 067028/Z/02/Z and 083145/Z/07/Z from The Wellcome Trust (registered charity no. 210183).

No potential conflicts of interest relevant to this article were reported.

V.A.H. researched data and wrote, reviewed, and edited the manuscript. I.S.Y., C.C.P., D.W.M.P., J.D.W., M.J.A.M., and D.R.M. researched data, contributed to discussion, and reviewed and edited the manuscript.

Parts of this study were presented in abstract form at the 46th Annual Meeting of the European Association for the Study of Diabetes, Stockholm, Sweden, 20–24 September 2010.

The authors thank Hanne Kruger and Dr. Caroline Mercer from the Centre for Public Health, Queen’s University Belfast, U.K., for their assistance with A1C analysis. The authors are grateful to the patients who took part in DAPIT, the DAPIT Research Midwives who collected the data, and the collaborators at each center.