Insulin resistance in early- and late-onset preeclampsia. A proposal to revise the current classification of the preeclampsia syndrome

Preeclampsia is one of the most serious diseases among hypertensive syndromes in pregnancy and is responsible for a high proportion of maternal deaths [1]. It is commonly defined as hypertension of 140/90 mmHg or higher and proteinuria of 0.3 g/24 h or higher, which appears after 20 weeks’ gestation in a previously normotensive woman [2]. Preeclampsia includes a large range of clinical manifestations, both mild and severe, the degrees of which are not specified in the current classifications. Only the American College of Obstetricians and Gynecologists suggest distinguishing a mild from a severe preeclampsia, even if the criteria for defining severe preeclampsia may be confusing [3]. Preeclampsia is commonly considered a disease of placentation, which results in reduced placental perfusion and subsequent reduced fetal growth later in preg-nancy [4]. Placental damage appears to trigger the maternal syndrome, which is characterized by hypertension and involvement of the kidneys and also, frequently the liver; hypertension and organ damage are mediated by endothelial dysfunction and the endothelium seems to be the target of factors produced by a damaged placenta [5]. The pathogenesis of preeclampsia is thus considered to be an abnormal placentation, which, by inducing endothelial dysfunction, is subsequently responsible for the maternal syndrome.

This theory is suitable for all the preeclampsia cases with abnormal placentation and intrauterine growth restriction, which is usually revealed early in the third trimester, often before 34 weeks’ gestation. But what about the cases that appear after 34 or 37 weeks’ gestation with a normal placentation and normal or ‘large-for-gestational-age’ fetuses? In these cases, what is the cause of the endothelial dysfunction that results in hypertension and proteinuria? One may invoke a placental origin linked to excessive fetal demands, as occurs in multiple pregnancy [6] or in fetal overgrowth [7]. However, in contrast to this theory, cases with normal placentation [8] and normal fetal weight [9] have been reported.

These considerations introduce an important issue that has not been accounted for in the current classifications of hypertensive syndromes in pregnancy, even though it is the most important variable in predicting maternal and fetal outcomes: the gestational age at which preeclampsia occurs [10]. Terms such as ‘early-onset’ and ‘late-onset’ are helpful in distinguishing preeclampsia syndrome in relation to gestational age. Before 32 weeks’ gestation (early-onset), maternal mortality is 20-fold higher than when preeclampsia occurs at term [1], and an adverse infant outcome is influenced predominantly by gestational age in early-onset severe preeclampsia [11]. The occurrence of intrauterine growth restriction in pregnancies complicated by preeclampsia before 37 weeks’ gestation [9] confirms the importance of placentation in the pathogenesis of early-onset preeclampsia. However, an increase in large babies among women with preeclampsia delivering after 37 weeks’ gestation [7] introduces the issue of metabolic features in determining late-onset preeclampsia. Furthermore, very recently, Egbor and colleagues have shown that abnormal placental morphology was associated only with early-onset (<34 weeks) preeclampsia, whereas placentas from late-onset (>34 weeks) preeclampsia were similar to placentas from gestational-age-matched controls [8]. This study supports the hypothesis that late-onset preeclampsia is a maternal disorder and not a placental disease [8]. This is the concept that, in my opinion, should be stressed for a better understanding of the two features of preeclampsia syndrome.

Among maternal metabolic disorders, glucose intolerance is considered a predictor of hypertensive disorders in pregnancy [12] and it is well known that gestational diabetes strongly predisposes women to preeclampsia [13]. Insulin resistance has been evaluated at delivery, with conflicting results [14,15], but is also considered a marker when it has been evaluated at the first [16] or second trimester [17] in women who subsequently developed preeclampsia. Conflicting results were probably caused by the different populations of preeclampsia patients involved in the trials; in the Caruso and colleagues study, insulin sensitivity, assessed with the hyperinsulinemic euglycemic clamp, was similar in women with preeclampsia and the control group; but the women with preeclampsia were likely to suffer from early-onset preeclampsia because their mean gestational age was 32.1 ± 1.4 weeks [14]. An analogous situation was observed in a study by Salamalekis and colleagues, in which insulin resistance in women with preeclampsia (mean gestational age of 29.9 weeks, therefore early-onset) was not different from the control group [15]. By contrast, in a study by Kaaja and colleagues, the preeclampsia patients were predominantly late-onset, and the insulin resistance, measured by the minimal model technique, was significantly higher in preeclamptic women than in the control group of pregnant women [16]. Consequently, the different results in these studies might be owing only to different populations of preeclamptic women, with different levels of insulin sensitivity. In the same way, in a study by Wolf and colleagues, the women who had an increased insulin resistance from the first trimester and subsequently developed preeclampsia were certainly late-onset, because the mean gestational age at delivery was 37.6 ± 3 weeks [17].

Recently, Parretti and colleageues evaluated insulin sensitivity in the second trimester of pregnancy of normally glucose-tolerant women and showed that increased insulin resistance was predictive of a subsequent preeclampsia syndrome [18]. In this study, it was not specified at what gestational age preeclampsia occurred, so it is possible that the population of preeclamptic women comprised a mixture of early- and late-onset cases. Also, in a recent study by the author and colleagues, if we consider the normal glucose-tolerant preeclamptic women in total, there was an increased insulin resistance in the first trimester compared with the control group, as evaluated with the Homeostasis Model Assessment Ratio (HOMA-IR), after adjusting for maternal age, gestational age and prepregnancy body mass index [19]. However, when we dichotomized the preeclampsia patients into two subgroups, early-onset (<37 weeks gestation) and late-onset (≥37 weeks gestation), HOMA-IR mean values selected two completely different populations. In fact, in the first trimester late-onset preeclampsia cases, HOMA-IR mean values were significantly higher than those for early-onset preeclampsia (<37 weeks gestation) and also than those for the control group of pregnant women. By contrast, no difference in insulin resistance was reported between early-onset preeclamptic women and the control group.

Therefore, it seems that in preeclamptic women with a normal glucose tolerance, an increased insulin resistance is associated only with late-onset preeclampsia, and so is predictive of preeclampsia, as found by Parretti and colleagues [18]. However, in my experience, this result is valid only for late-onset preeclampsia cases. In our study, the two subgroups did not differ in mean diastolic and systolic blood pressure, nor in proteinuria mean values. Also, endothelial damage, revealed by serum vascular cell adhesion molecule-1 levels, was similar in the two subgroups [19]. Therefore, we suggest that in early-onset preeclampsia, as is well known, endothelial damage depends on abnormal placentation, and this was the case for our patients, who all showed an intrauterine growth restriction; whereas, in late-onset preeclampsia, endothelial damage depends on elevated insulin resistance. In fact, insulin resistance may impair endothelial function [20] and hyperinsulinemia may predispose to hypertension through increased renal sodium reabsorption and stimulation of the sympathetic nervous system [21]. Moreover, hyperinsulinemia was also associated with impaired lipid metabolism and it has been reported that elevated total cholesterol [22], triglycerides and free fatty acid levels [23] may precede the development of preeclampsia.

We concluded our study by suggesting that “…the syndrome characterized by induced pregnancy hypertension and proteinuria, commonly called preeclampsia, could be the expression of two different pathogeneses. The first, late-onset, which is characterized by first-trimester increased insulin resistance, high pre-pregnancy body mass index with normal placentation and adequate for gestational age fetuses; and the second, early-onset, with normal first trimester insulin resistance and normal pre-pregnancy body mass index, which is characterized by placental impairment and a worse fetal outcome. Both are characterized by endothelial damage, which leads to the clinical expression of the syndrome: hypertension and proteinuria…” [19]. Thus, it may be better to revise the current classification of hypertensive disorders in pregnancy and hope that terms such as early- and late-onset preeclampsia might find a place in a novel classification, which needs to determine the stage of gestation (34 or 37 weeks) at which late-onset preeclampsia begins. This revision might be helpful for a clinical approach to the diagnosis and prognosis of the different features of this syndrome.