and one should not be convinced too easily and rush to trials while angels fail to get their grants funded. Our limited resources should focus more on basic research, with human studies stressing prediction markers com bined with ancillary studies to help defi ne mechanisms (and, of course, the safety and effi cacy of therapy). It appears surprising that administration of vitamins in amounts that did not exceed maximum daily tolerable allowances should be associated with adverse eff ects, but the VIP data are not the fi rst time that vitamin or nutrient supplements administered to populations already consuming considerable amounts of the supple ment’s content in their diets have had adverse eff ects. This brings to mind some of the problems when vitamins A and D are ingested by well-nourished populations, as well as one of the earliest supplementation trials in which adverse results were assoc iated with high proteinsupplementation in women who were not protein defi cient. 9 There is also the reverse side of the coin when investigators showed a small but important eff ect of calcium supplementation—ie, a decrease in the severity of adverse outcomes—in a population with defi cient intake of calcium. 2 Thus it would not be surprising if two ongoing trials involving antioxidant supplementation in areas of developing nations where nutrient defi c iencies are rampant produced results diff ering from those here (ie, a multicentre WHO trial and one in Brazil).

OBJECTIVE To compare the rates and perinatal outcome in women who experienced preeclampsia in a previous pregnancy to those in women who developed preeclampsia as nulliparas. STUDY DESIGN This is a secondary analysis of data from 2 separate multi-center trials of aspirin for prevention of preeclampsia. Women who had preeclampsia in a previous pregnancy (n = 598) were compared with nulliparous women (n = 2934). Outcome variables were rates of preeclampsia, preterm delivery at <37 and <35 weeks of gestation, small-for-gestational-age infant, abruptio placentae, and perinatal death. Data were compared by using chi-square analysis and Wilcoxon rank sum test. RESULTS The rates of preeclampsia and of severe preeclampsia were significantly higher in the previous preeclamptic group as compared to the nulliparous group (17.9% vs 5.3%, P <.0001, and 7.5% vs. 2.4%, P <.0001, respectively). Women who had recurrent preeclampsia experienced more preterm deliveries before 37 and 35 weeks of gestation than nulliparous women who developed preeclampsia. In addition, among women who developed severe preeclampsia, those with recurrent preeclampsia had higher rates of preterm delivery both before 37 weeks (67% vs 33%, P =.0004) and before 35 weeks of gestation (36% vs 19%, P =.041), and higher rates of abruptio placentae (6.7% vs 1.5%) and fetal death (6.7% vs 1.4%) than did nulliparous women. CONCLUSION Compared to nulliparous women, women with preeclampsia in a previous pregnancy had significantly higher rates of preeclampsia and adverse perinatal outcomes associated with preterm delivery as a result of preeclampsia.

OBJECTIVE:This study was undertaken to compare rates and severity of gestational hypertension and preeclampsia, as well as perinatal outcomes when these complications develop, between women with twin gestations and those with singleton gestations.STUDY DESIGN:This was a secondary analysis of prospective data from women with twin (n = 684) and singleton (n = 2946) gestations enrolled in two separate multicenter trials of low-dose aspirin for prevention of preeclampsia. End points were rates of gestational hypertension, rates of preeclampsia, and perinatal outcomes among women with hypertensive disorders.RESULTS:Women with twin gestations had higher rates of gestational hypertension (relative risk, 2.04; 95% confidence interval, 1.60-2.59) and preeclampsia (relative risk, 2. 62; 95% confidence interval, 2.03-3.38). In addition, women with gestational hypertension during twin gestations had higher rates of preterm delivery at both <37 weeks' gestation (51.1% vs 5.9%; P <. 0001) and <35 weeks' gestation (18.2% vs 1.6%; P <.0001) and also had higher rates of small-for-gestational-age infants (14.8% vs 7. 0%; P =.04). Moreover, when outcomes associated with preeclampsia were compared, women with twin gestations had significantly higher rates of preterm delivery at <37 weeks' gestation (66.7% vs 19.6%; P <.0001), preterm delivery at <35 weeks' gestation (34.5% vs 6.3%; P <.0001), and abruptio placentae (4.7% vs 0.7%; P =.07). In contrast, among women with twin pregnancies, those who remained normotensive had more adverse neonatal outcomes than did those in whom hypertensive complications developed.CONCLUSIONS:Rates for both gestational hypertension and preeclampsia are significantly higher among women with twin gestations than among those with singleton gestations. Moreover, women with twin pregnancies and hypertensive complications have higher rates of adverse neonatal outcomes than do those with singleton pregnancies.

A 30-year-old woman with a living related six-antigen-matched kidney allograft conceived 10 years posttransplantation. She had discontinued her immunosuppression medications 3 years previously. The allograft functioned well throughout gestation, which was complicated by preeclampsia, leading to induction at 35 weeks and delivery of a 2,175- g male. (Am J Kidney Dis 1998 Oct;32(4):661-3)

BACKGROUND:Whether low-dose aspirin prevents preeclampsia is unclear. It is not recommended as prophylaxis in women at low risk for preeclampsia but may reduce the incidence of the disease in women at high risk.METHODS:We conducted a double-blind, randomized, placebo-controlled trial in four groups of pregnant women at high risk for preeclampsia, including 471 women with pregestational insulin-treated diabetes mellitus, 774 women with chronic hypertension, 688 women with multifetal gestations, and 606 women who had had preeclampsia during a previous pregnancy. The women were enrolled between gestational weeks 13 and 26 and received either 60 mg of aspirin or placebo daily.RESULTS:Outcome data were obtained on all but 36 of the 2539 women who entered the study. The incidence of preeclampsia was similar in the 1254 women in the aspirin group and the 1249 women in the placebo group (aspirin, 18 percent; placebo, 20 percent; P=0.23). The incidences in the aspirin and placebo groups for each of the four high-risk categories were also similar: for women with pregestational diabetes mellitus, the incidence was 18 percent in the aspirin group and 22 percent in the placebo group (P=0.38); for women with chronic hypertension, 26 percent and 25 percent (P= 0.66); for those with multifetal gestations, 12 percent and 16 percent (P=0.10); and for those with preeclampsia during a previous pregnancy, 17 percent and 19 percent (P=0.47). In addition, the incidences of perinatal death, preterm birth, and infants small for gestational age were similar in the aspirin and placebo groups.CONCLUSIONS:In our study, low-dose aspirin did not reduce the incidence of preeclampsia significantly or improve perinatal outcomes in pregnant women at high risk for preeclampsia.

Background Whether low-dose aspirin prevents preeclampsia is unclear. It is not recommended as prophylaxis in women at low risk for preeclampsia but may reduce the incidence of the disease in women at high risk.Methods We conducted a double-blind, randomized, placebo-controlled trial in four groups of pregnant women at high risk for preeclampsia, including 471 women with pregestational insulin-treated diabetes mellitus, 774 women with chronic hypertension, 688 women with multifetal gestations, and 606 women who had had preeclampsia during a previous pregnancy. The women were enrolled between gestational weeks 13 and 26 and received either 60 mg of aspirin or placebo daily.Results Outcome data were obtained on all but 36 of the 2539 women who entered the study. The incidence of preeclampsia was similar in the 1254 women in the aspirin group and the 1249 women in the placebo group (aspirin, 18 percent; placebo, 20 percent; P=0.23). The incidences in the aspirin and placebo groups for each of the four high-risk categories were also similar: for women with pregestational diabetes mellitus, the incidence was 18 percent in the aspirin group and 22 percent in the placebo group (P=0.38); for women with chronic hypertension, 26 percent and 25 percent (P=0.66); for those with multifetal gestations, 12 percent and 16 percent (P=0.10); and for those with preeclampsia during a previous pregnancy, 17 percent and 19 percent (P=0.47). In addition, the incidences of perinatal death, preterm birth, and infants small for gestational age were similar in the aspirin and placebo groups.Conclusions In our study, low-dose aspirin did not reduce the incidence of preeclampsia significantly or improve perinatal outcomes in pregnant women at high risk for preeclampsia. (C) 1998, Massachusetts Medical Society.

This review stresses changes in osmoregulation as well as the secretion and metabolism of arginine vasopressin during pregnancy, focusing on human gestation. Pregnant women experience a decrease in body tonicity, plasma osmolality decreasing immediately after conception to a nadir approximately 10 mosmol/kg below non-pregnant levels early in pregnancy, after which a new steady state is maintained until term. Data from both human and rodent gestation have led to a formation of how these changes occur. The osmotic thresholds for thirst and antidiuretic hormone release decrease in parallel. Lowering the threshold to drink stimulates increased water intake and dilution of body fluids. Because arginine vasopressin (AVP) release is not suppressed at the usual level of body tonicity, the hormone continues to circulate and the ingested water is retained. Plasma osmolality declines until it is below the osmotic thirst threshold, and a new steady state with little change in water turnover is established. Pregnancy is characterized by increments in intravascular volume, but volume-sensing AVP release mechanisms appear to adjust as gestation progresses so that each new volume status is "sensed" as normal. The metabolic clearance of AVP increases fourfold, the rise paralleling that of circulating cystine aminopeptidase (vasopressinase), and enzyme produced by the placenta. Furthermore, the disposal rate of 1-deamino-8-D-AVP, and AVP analogue resistant to inactivation by vasopressinase, is unaltered in pregnancy. Thus, the increase in AVP's metabolism and the high circulating aminopeptidase levels have been implicated in certain forms of transient diabetes insipidus that occur in late pregnancy. Finally, mechanisms responsible for the altered osmoregulation in pregnancy are obscure, but chorionic gonadotropin and relaxin may be implicated in the changes.

Glomerular filtration rate (GFR) and ERPF increase approximately 50% in human pregnancy. To determine if pregnant women have additional “renal reserve,” inulin and p-aminohippurate clearances (Cin, CPAH) were measured in maximally hydrated women before and after a 300-g steak meal, once during late gestation, and again > or = 3 mo postpartum. Protein loading increased Cin [106 +/- 5 (SE) to 119 +/- 4 ml/min, P < 0.003], but not CPAH (587 +/- 35 to 624 +/- 32 ml/min, NS) in the nonpregnant state, but neither clearance was altered during gestation (Cin: 156 +/- 7 to 160 +/- 9.6 ml/min, NS; CPAH: 831 +/- 36 to 899 +/- 37 ml/min, NS). A natriuresis occurred only postpartum (+142 mu eq/min, P < 0.02), which could be explained by the increased GFR alone, since indexes of filtrate delivery and reabsorption (V/GFR, CH2O/GFR, CH2O/V) and fractional sodium excretion changed little. Dopamine excretion, uninfluenced by protein, did not correlate with increments in GFR. A carbohydrate meal (time controls) had no effect on the above described parameters. We make the following conclusions. If protein and pregnancy achieve hyperfiltration by similar mechanisms, these pathways appear “exhausted” in gestation. Also, oral protein loading does not measure maximal renal reserve, since basal GFR in late gestation was substantially greater than that measured after protein feeding in nonpregnant subjects.

OBJECTIVE:We tested the hypothesis that gestational changes in reflex neural control of the heart and vasculature contribute to altered cardiovascular responses to vasopressin during pregnancy. STUDY DESIGN:Changes in mean arterial pressure, cardiac output, total peripheral resistance, and heart rate were measured in response to constant infusion of arginine vasopressin (0.15 to 2.5 mU/kg/min) in conscious pregnant and virgin rats (n = 9) with total autonomic blockade plus restoration of baseline hemodynamics by norepinephrine infusion. RESULTS:Resting cardiac output was 40% higher and total peripheral resistance 30% lower in pregnant animals (p < 0.01). Constant infusion of arginine vasopressin evoked equivalent changes in mean arterial pressure in both groups, but the respective contributions of cardiac output and total peripheral resistance to mean arterial pressure differed between groups. Cardiac output was unchanged and the increase in total peripheral resistance was significantly blunted in pregnant vs virgin rats during arginine vasopressin infusion. Control data in nonblocked revealed similar pressor responses to arginine vasopressin in gravid compared with virgin rats but no differences in the contributions of cardiac output and total peripheral resistance to the change in mean arterial pressure. CONCLUSION:These findings suggest that neural modulation of arginine vasopressin-induced hypertension is altered during pregnancy and are consistent with a reduction in intrinsic vascular sensitivity to arginine vasopressin during gestation.

Pressor responses to angiotensin II (ANG II) are markedly attenuated in reflex-intact pregnant animals, a phenomenon widely attributed to intrinsic changes in vascular reactivity. To test the hypothesis that gestational augmentation of neural reflex activity contributes importantly to this phenomenon, changes in mean arterial pressure (MAP), cardiac output (CO), and total peripheral resistance (TPR) were compared during constant infusion (25-400 ng.kg-1.min-1) of ANG II in conscious virgin and pregnant rats, using a model of total autonomic blockade (chlorisondamine chloride and methscopolamine bromide), with restoration of baseline hemodynamics by infusion of norepinephrine. Basal CO was higher and TPR lower in pregnant (CO 121.8 +/- 3.8 ml/min; TPR 0.78 +/- 0.04 mmHg.ml-1.min) compared with virgin (CO 95.9 +/- 3.9 ml/min; TPR 1.05 +/- 0.08 mmHg.ml-1.min) rats (P < 0.005). Pressor responses to ANG II were similar in both groups of reflex-blocked animals due to comparable changes in TPR and CO (not significant by repeated-measures analysis of variance). Other experiments demonstrated that changes in MAP after bolus administration of ANG II did not differ in areflexic virgin and gravid rats. Thus in the absence of autonomic control ANG II has similar effects on systemic resistance in pregnant and nonpregnant rats, suggesting that reflex neural mechanisms contribute significantly to gestational changes in pressor responsiveness. These data further suggest that pregnancy is not accompanied by a generalized decrease in vascular reactivity to all pressor agents.

I N THIS ISSUE of American Journal of Kidney Diseases, Sturgis and Davison compare, in case control fashion, the long-term renal prognosis of 18 women who underwent 34 pregnancies after kidney transplantation with 18 allograft recipients who never conceived. Posttransplant follow-up in each group averaged 12 years, and the mean time that elapsed from the first pregnancy was 5 years. The authors found no statistical differences in glomerular filtration rate (GFR) (measured by inulin infusion) or prevalence of hypertension between those who conceived and those who did not, concluding, cautiously, that pregnancy had no remote adverse effect on either renal function or blood pressure in these allograft recipients. One might argue that their pairing was a bit imperfect, the number of patients was limited, and the failure to find differences may reflect type II errors, in that studying more patients might show a small but significant adverse effect of childbearing on renal function. However, to our knowledge these data are currently unique and are encouraging to allograft recipients who want to bear children. As noted by Sturgis and Davison and reviewed elsewhere,I,2 renal transplantation is accompanied by a striking improvement of reproductive function, including a reversal of the relative infertility that accompanies end-stage renal disease, so that conception rates estimated at less than 1 in 200 for dialysis patients increase to 1 in 50. Regretfully, however, many of these pregnancies were not planned, the allograft recipients being unaware that they could conceive, and therefore some sought prenatal care for such high risk pregnancies quite late in gestation. Thus, we begin this editorial by underscoring the importance of contraceptive counseling for all renal patients of childbearing age, including transplant recipients. This is a responsibility nephrologists too often neglect. How should we counsel these women? First, physicians should be aware that many aspects of kidney disease and pregnancy are controversial, and transplantation is no exception. For example, in a 1985 editorial in this journal we served as arbitrators of two discordant views concerning the renal and pregnancy prognosis in women with primary glomerular disease who conceive, and these controversies surfaced again in later issues of the Journal. 4,5 Some saw nefarious results when gestation and kidney disease coexisted, especially in the case of certain specific disorders; whereas others, including ourselves, argue that it is the functional status and presence or absence of hypertension that determine outcome. That is, if GFR is well preserved (serum creatinine < 130 ~mol/L [1.5 mg/dLD and the patient normotensive, over 95% ofthe gestations will be successful, and only a minority will experience acceleration of their disease. Our conclusions were based on information from over 1,000 patients with biopsy-diagnosed renal diseases, but the data are primarily retrospective and correlative in nature. The circumstances are similar for transplantation, some suggesting guarded results both for fetus and allograft,6 whereas others are substantially more optimistic. 1,2 Again, the principle that renal function and presence or absence of hypertension determine outcome appears to prevail in this category as well. Substantiating this view, Dr Davison, coauthor of the article in this issue, has periodically provided us with surveys of the literature, which unfortunately consists primarily of case reports and small retrospective series. In his most recent review, which appeared in the February 1991 issue ofthisjournaV he surveyed 2,309 gestations in 1,594 allograft recipients and noted that 40% of the pregnancies did not go beyond the first trimester. The number of spontaneous abortions was similar to that in the population at large (approximately 16%); the remaining terminations (therapeutic) were

This article, a review of factors controlling vasopressin (AVP) release in pregnancy, extends our contribution to a symposium in this journal published in 1987 (vol X, pp 270-275). Body tonicity decreases (approximately 10 mOsm/kg) very early in pregnancy due to decrements in the osmotic thresholds for AVP release and thirst. In addition, the metabolic clearance rate (MCR) of AVP markedly increases between gestational week 10 and midpregnancy, and is paralleled by the appearance and increase of circulating cystine aminopeptidase (vasopressinase), while the MCR of 1-deamino-8-D-AVP (DDAVP), an analogue resistant to inactivation by the enzyme, changes little in pregnancy. These increases (MCR of AVP and plasma vasopressinase) may explain certain syndromes of transient diabetes insipidus (DI) that complicate gestation. Finally, mechanisms responsible for the altered osmoregulation in pregnancy are obscure, but chorionic gonadotropin may be involved in the changes during human gestation.

Advances in prenatal care have resulted in a substantial decline in the number of serious complications associated with high blood pressure during pregnancy. Nevertheless, the hypertensive disorders remain a significant cause of maternal and fetal morbidity, and even mortality [Rochat et al. 1988]. In addition, hypertension in pregnancy is a topic that generates considerable controversy, ranging from the correct manner to measure blood pressure in gravidas, to major disputes concerning management considerations. The pervasiveness of such controversies [Cunningham and Pritchard 1984, Ferris 1984, Disdale 1988, Kaplan et al. 1988] led the United States' National Institutes of Health, through its National High Blood Pressure Education Program, to convince a working group whose "consensus" report has just been published (1990), and is recommended reading for physicians who manage gravidas. This review focuses on preeclampsia, and includes observations concerning the pathogenesis of this disorder as well as strategies to prevent its occurrence. Conflicting opinions regarding the management of preeclamptic women will also be discussed, highlighting the recommendations made by the NIH consensus group (NHBPEP 1990). Space considerations limit the references cited, and the reader is referred to a chapter [Barron 1991] and several texts [Chesley 1978, Rubin 1988] for a more complete survey of the literature.

Contributions of the autonomic nervous system (ANS), renin-angiotensin system (RAS), and arginine vasopressin (AVP) to basal mean arterial pressure (MAP) were evaluated in near-term pregnant and virgin rats as follows. MAP and heart rate (HR) were measured before and after ganglionic, alpha-adrenoreceptor, RAS, and/or AVP blockade. In addition, pressor responses to angiotensin II (ANG II), norepinephrine, phenylephrine, or AVP were determined in ganglionic-blocked animals. In both groups decrements in MAP were greatest after ganglionic or alpha-blockade, intermediate after RAS blockade, and negligible after AVP-V1 antagonism ([d(CH2)5Tyr(Me)]AVP). Recovery of MAP was also similar in the two groups except after phentolamine when MAP and HR remained lower in gravid rats. Superimposition of RAS or AVP blockade during phentolamine infusion suggested that ANG II and AVP were less effective in supporting MAP during alpha-blockade in pregnancy. Pressor responses to ANG II and norepinephrine during ganglionic blockade were markedly blunted during pregnancy; however, those to phenylephrine and AVP were unchanged. We conclude that contributions of ANS, RAS, and AVP to basal MAP are similar in pregnant and virgin rats; neural mechanisms dominating in both groups. However, recovery during alpha-blockade is impaired during gestation, apparently due to blunted HR responses and decreased pressor contributions of ANG II and AVP. This may be explained, in part, by decreased vascular reactivity to ANG II, although a similar mechanism cannot be invoked for AVP.