Pregnancy and essential hypertension are both associated with insulin resistance1,2. Recent research has suggested that hypertension arising in pregnancy is associated with a reduction in insulin sensitivity below that of normotensive pregnancy3–6. Most of these studies have failed to divide pregnancy-induced hypertension into pre-eclampsia and gestational hypertension according to the presence or absence of proteinuria. There is evidence that pre-eclampsia and gestational hypertension may have different patho-physiological mechanisms7, so the failure to make this distinction may have affected the conclusions of these studies. This may explain the apparent conflict between these conclusions and our own. We have reported that in two groups of pregnant women who were studied in the third trimester and matched for age, body mass index and gestation, those with pre-eclampsia had higher insulin sensitivity than those with normal pregnancy8. We postulated that while pre-eclampsia is not associated with insulin resistance, gestational hypertension may be, and this would fit with the concept of gestational hypertension being a latent form of essential hypertension. In the quest for more evidence on this issue we analysed an existing database of glycosylated haemoglobin in nondiabetic pregnancy to investigate the effect of glucose metabolism in the first and second trimesters on subsequent hypertensive complications of pregnancy.
We analysed a database of glycosylated haemoglobin (HbA1) in nondiabetic pregnant women to investigate the relation between glucose metabolism in the first and second trimesters and hypertensive complications of pregnancy. From a total of 1334 women, 13 had pre-existing hypertension, 225 developed gestational hypertension and 51 developed pre-eclampsia. At 28 weeks of gestation, the women who susequently developed gestational hypertension had a significantly higher mean HbAl than those who remained normotensive (6.33 vs 6.17%, P < 0.02). This difference remained significant after correcting for the effects of age and body mass index (regression coefficient 0.11, SE 0.06, P = 0.05). In contrast, there were no significant differences in HbAl between the women with pre-eclampsia and their normotensive counterparts. This provides indirect evidence to support our hypothesis that gestational hypertension is associated with insulin resistance but pre-eclampsia is not.
Data were analysed from 1334 nondiabetic women with a singleton pregnancy who had attended the Royal Maternity Hospital, Belfast during a one year period for antenatal care. Venous blood samples were taken at the initial visit for random plasma glucose and glycosylated haemoglobin (HbA1), and at 28 weeks for HbA1. Plasma glucose was measured by the glucose oxidase method, and HbAl was measured by the Corning electroendosmosis method.
Several aspects of maternal history and measures of maternal and fetal outcome were recorded in the database, but for the purposes of this study we were primarily concerned with a history of pre-existing hypertension and the presence of pregnancy-induced hypertension, subdivided by the presence or absence of proteinuria into pre-eclampsia and gestational hypertension, respectively.
The diagnosis of pre-existing hypertension was accepted when the woman had been treated for hypertension before the pregnancy. The criteria for a diagnosis of pregnancy-induced hypertension were a systolic blood pressure ≥ 140 mmHg or a diastolic blood pressure ≥ 90 mmHg in at least two readings more than 6 hours apart. The diagnosis of pre-eclampsia was made when pregnancy-induced hypertension was associated with at least 0.3 g/L protein on dipstick testing of urine.
A one-way analysis of variance with Dunnett's t test was used to assess whether there was any significant difference in mean plasma glucose or HbAl between the women who remained normotensive throughout and those with pre-existing hypertension or those who developed hypertension during the pregnancy.
The mean age of the women was 26.3 years (range 1545 years). Parity ranged from 0–9, with a mode of 0 and a median of 1. The mean gestational age at the initial visit was 12 weeks; 50% of the women had attended by 12 weeks, and 95% by 16 weeks. Mean maternal weight at the initial visit was 59.3 kg (range 36.5–120 kg) and mean body mass index (BMI) was 23.5 kg/m2 (range 156 46.9 kg/m2). Mean plasma glucose at the initial visit was 5.7 mmoYL (SD 0.99). Mean initial HbAl was 6.5% (SD 0.84) and mean HbAl at 28 weeks was 6.2% (SD 0.81).
Only 13 women had pre-existing hypertension. These women had a significantly higher BMI than the normotensive women (26.0 vs 23.2 kg/m2, P c 0.05) and a significantly higher random plasma glucose at the initial visit (6.6 vs 5.7 mmol/L; P < 0.01), but there was no significant difference in HbAl either at the initial visit or at the 28 week visit. Of the remaining 1321 women, 276 (21%) developed pregnancy-induced hypertension. At 28 weeks the mean HbA1 of these women was significantly higher than that of the normotensive women (6.28 vs 6.17%; P < 0.05); there was no significant difference in either random plasma glucose or HbA1 at the initial visit. The 276 women with pregnancy-induced hypertension were subdivided into 225 who had gestational hypertension and 51 who developed pre-eclampsia. Data for the four groups of women are summarised in Table 1.
|Age||26.8 (5.7)||26.1 (5.2)||27.6 (6.0)||25.2 (4.9)|
|≤ 12 wks||10||604||139||30|
|≥ 13 wks||3||441||86||21|
|BMI (kg/m2)||26.0 (7.2)||23–2 (3.7)||24.7 (4.1)||24.4 (4.1)|
|Glucose (mmol/L)||6.6 (2.0)||5.7 (1.0)||5.8 (1.0)||5.6 (1.1)|
|HbAl (%)||6.49 (0.58)||6.44 (0.86)||6.52 (0.78)||6.56 (0.89)|
|HbAl (%)||6.43 (0.85)||6.17 (0.81)||6.33 (0.81)||6.08 (0.83)|
The women with gestational hypertension were significantly older and had a significantly higher mean BMI than the women who remained normotensive (27.6 vs 26–1 years, P < 0.01; 24–7 vs 23–2 kg/m2, P < 0.01, respectively). There were no significant differences in age or BMI between those who developed pre-eclampsia and the normotensive group. Measures of glucose metabolism were compared between the normotensive women and both the women with gestational hypertension and those who became pre-eclamptic. The women who developed gestational hypertension had a significantly higher HbA1 at 28 weeks of gestation than those who remained normotensive (6.33 vs 6–17%, P < 0.02). This difference remained significant after correcting for the effects of age and BMI by multiple linear regression (regression coefficient 0.11, SE 0.06, P= 0.05). In contrast, there were no significant differences in any of the measures of glucose metabolism between the women with pre-eclampsia and their normotensive counterparts.
Early studies of women with hypertension during pregnancy suggested an association with insulin resistance; women with pregnancy-induced hypertension had higher glucose levels on intravenous glucose tolerance testing9, attenuated glucose response to intravenous insulin10, and higher insulin concentrations in response to an intravenous glucose load than normotensive controls11. However, these studies did not exclude women with pre-existing hypertension and they did not distinguish between gestational hypertension and pre-eclampsia. Furthermore, they did not measure insulin sensitivity. In our previous study we showed that insulin sensitivity was not reduced in a group of women with pre-eclampsia compared with normotensive pregnant women8, indicating that pre-eclampsia is not associated with insulin resistance. The apparent conflict between these results and the earlier reports might be explained if gestational hypertension is associated with insulin resistance, but we do not yet have information on insulin sensitivity in women with gestational hypertension.
A number of investigators have examined glucose metabolism in women who subsequently developed hypertension in pregnancy. A study of unselected pregnancies showed elevated insulin concentrations during an oral glucose tolerance test at 19 weeks of gestation in women who subsequently developed hypertension3. Another study reported that women who developed hypertension in the third trimester had higher mean fasting insulin concentrations at 2628 weeks than those who remained normotentive3. These studies suggest that hypertension in pregnancy is associated with, and predated by, a reduction in insulin sensitivity. As with the earlier studies, the women with pregnancy-induced hypertension included both those with pre-eclampsia and those with gestational hypertension and, where stated, those with gestational hypertension were in the great majority. It is therefore likely that this condition, and not pre-eclampsia, had the predominant influence on the overall results.
Sowers et al. reported that hyperinsulinaemia and insulin resistance were associated with pre-eclampsia in African American women. They distinguished between pre-eclampsia and gestational hypertension and found that those who subsequently developed pre-eclampsia had higher fasting insulin concentrations at 18–25 weeks than the women acting as controls. It is possible that there are metabolic differences between African American and Caucasian European women. While an association between pre-eclampsia and insulin resistance has been reported in a predominantly European group of women, this was largely determined by the degree of obesity12.
Solomon et al.13 showed that women who developed pregnancy-induced hypertension had significantly higher one hour glucose concentrations after taking 50 g oral glucose between 24 and 32 weeks than those who remained normotensive. The women who became hypertensive also had a signdkantly higher frequency of abnormal glucose loading tests (one hour glucose ≥ 7.8 mmol/L). When these women were subdivided into those who developed gestational hypertension and those who developed pre-eclampsia, only the women with gestational hypertension had a sigdicantly higher incidence of abnormal loading tests (P < 0.01) than the normotensive controls. This is in keeping with our findings of a significantly higher mean HbA1 at 28 weeks in women who developed gestational hypertension and not in those who developed pre-eclampsia, when compared with normotensive pregnant women. Together these studies provide indirect evidence that insulin resistance may play a role in the pathogenesis of gestational hypertension, but not pre-eclampsia. This is in agreement with our previous studf and is also in keeping with the hypothesis that different pathophysio-logical mechanisms are responsible for gestational hypertension and pre-eclampsia.
The information used in this study was collected with the aid of a grant from the British Diabetic Association.