点击显示 收起
1 From the Center for Population and Family Health, School of Public Health, Columbia University, New York.
2 Presented at the meeting Iron and Maternal Mortality in the Developing World, held in Washington, DC, July 6–7, 1998. 3 Address reprint requests to D Maine, Center for Population and Family Health, School of Public Health, Columbia University, 60 Haven Avenue, New York, NY 10032. E-mail: dpm1{at}columbia.edu.
ABSTRACT
Toxemia of pregnancy is called the disease of theories because, over decades of research, numerous causes have been proposed but none proved. Although many nutritional factors have been suggested as playing a causal role in the etiology of toxemia, mortality from this disease has not varied over time or between circumstances as one would expect a nutritional disease to do. This does not mean that there is no nutritional influence, but it does mean that the available evidence does not show that nutrition makes a major difference in maternal mortality from toxemia of pregnancy.
Key Words: Toxemia • maternal mortality • nutrition • preeclampsia • eclampsia • diet • pregnancy
INTRODUCTION
The terminology for this subject is a bit confusing. What is now called hypertensive diseases of pregnancy has also been called toxemia of pregnancy (1). Toxemia, in turn, encompasses both preeclampsia and eclampsia. Preeclampsia is characterized by high blood pressure, headaches, protein in the urine, and swelling of the face, hands, and feet. It can progress to eclampsia, which is marked by convulsions and can lead to coma, brain damage, and death (2).
CAUSES OF TOXEMIA
The literature on eclampsia has long been unsatisfying because even after reviewing dozens of articles, it is not possible to say very much about causation, which is why toxemia has been called the disease of theories (3). Even now this remains a topic full of questions and theories, so much so that the World Health Organization (WHO) Reproductive Health Library (4) made the following statement about anticonvulsants, which were used for >80 y to prevent preeclampsia from progressing to eclampsia: "At the present time, there is no clear evidence either in favor of or against the use of anticonvulsants..."
Hypotheses about toxemia have proposed causal roles for placental abnormalities, immunologic dysfunction, coagulation abnormalities, endothelial damage, endocrine abnormalities, genetic factors, and dietary factors (5, 6). Narrowing the focus to dietary factors, there are many alleged culprits, including the consumption of too little protein, magnesium, calcium, or zinc, and the consumption of too much sodium or fatty acids (7, 8).
Although each theory has, or has had, its proponents, there is contradictory evidence on the causal roles of these dietary components. Moreover, because toxemia is a complex disease— involving a variety of organ systems and functions—it is not clear whether the abnormalities that occur represent causes or effects.
Presently, one of the more popular hypotheses is the belief that a lack of calcium is a precipitating factor in the incidence of toxemia and that calcium supplementation can lower the incidence of this condition. Again, there is conflicting evidence. Several clinical trials have suggested that calcium supplementation may prevent toxemia (9), whereas other trials have not shown positive results (10). Moreover, researchers at the National Institute of Child Health and Human Development in Bethesda, MD, found that methodologic problems and differences in study designs made the positive studies inconclusive (11).
If the exact causes of a condition are not known, how is it possible to evaluate the role of a class of possible causal agents such as nutritional factors? One technique in epidemiology is to observe the way a disease "behaves." Whom does it strike? How has it changed over time? By observing its behavior, it is possible to tell a great deal about the nature of the disease. If a disease has a strong nutritional component, its incidence should vary markedly among social classes because, in every society, nutrition varies among social classes. Such a disease might also vary substantially during periods of time when other diseases influenced by nutrition increase or decrease.
In his invaluable and scholarly review of the history of maternal mortality, Death in Childbirth, Loudon (12) points out several of informative characteristics of toxemia. For example, he notes that, historically, "Eclampsia tended to be spread evenly throughout the population of childbearing women." This is not the description of a disease with a nutritional basis. A second characteristic of toxemia that points toward nonnutritional causes is that it is reported more often in women having their first child and in older women giving birth. It is difficult to imagine what dietary patterns would be shared by these 2 groups.
A third observation that weighs against a strong nutritional role in toxemia is that mortality from toxemia in England and Wales remained high between 1840 and 1933 (12). This was a period when infant mortality and adult mortality from diseases such as tuberculosis were declining markedly (13). Although other aspects of development (such as better housing and sanitation) may have played a role in these declines, better nutrition is thought to have been an important factor. So the historical evidence challenges the idea that nutrition plays a strong role in maternal deaths. Once methods for treating specific obstetric complications became available (eg, antibiotics, banked blood, and safer surgical procedures), maternal mortality plummeted. Between 1934 and 1950, maternal mortality in England and Wales fell from 441 deaths/100000 live births to 87 deaths/100000 births (12). Similar declines took place in the other developed countries. As Loudon (14) notes in his paper in this volume, "There is overwhelming evidence that social and economic conditions were very weak determinants of the levels of maternal mortality, whereas the standard of obstetric care was a very strong determinant."
If nutrition played an important role in maternal mortality, there would be differences in the causes of maternal deaths between developed and developing countries, but this is not so. The WHO estimates that the causes of maternal mortality worldwide reflect the pattern in developing countries because 99% of maternal deaths take place there (1). A comparison of those causes with recent data from the United States shows striking similarities: hemorrhage, 25% worldwide, 29% in the United States; sepsis, 15% and 13%, respectively; and hypertensive disorders of pregnancy (mainly toxemia), 12% and 18%, respectively (15). The relative importance of these conditions is not much different despite the huge difference in maternal mortality ratios: 430 for the world and 12 for the United States.
The big differences in the causes of maternal death between developed and developing countries are related to the availability of medical care (12). For example, deaths due to unsafe abortion account for an estimated 13% of maternal deaths in the world but are extremely rare in developed countries, where safe abortion services are widely available (1). Globally, obstructed labor accounts for an estimated 8% of maternal deaths in the world, but such deaths are almost unknown in countries where cesarean delivery is available (15).
CONCLUSION
Maternal mortality in general and toxemia in particular do not behave like nutritional diseases. This does not mean that there is no nutritional influence, but it does mean that the available evidence does not show that nutrition makes a major difference in maternal mortality.
DISCUSSION
Dr Maine: I would like to introduce a couple of concepts that I think will help clarify the discussions we have been having. Maternal mortality, morbidity, and health are not always on the same continuum. The concept that bad health gets worse and worse and leads to mortality is a model that is really based on child health. We could almost make it a mathematical model for children based on what we know about child health: for example, 1 case of measles, 2 cases of diarrhea, and an upper respiratory infection add up to death. Adults, it seems, can get around in quite poor health for a long time and, unless something goes wrong or there is a complication, lots of malnourished women and lots of anemic women give birth just fine. That is not to say that we should not be concerned about their health, but I think that when we are talking about these things we should be clear as to the endpoint about which we are talking.
If we want to reduce maternal mortality, the place to start is probably at the district hospital. If we are talking about women's health, we meet the very important problem of why we would not just deal with pregnant women. I find it offensive that we just deal with anemia in women who are pregnant, and I would like to reintroduce the concept of attributable risk. It is very possible, especially with these fascinating nutritional complexities, to go far into the details and to lose sight of the big picture. For example, the historical picture is that everything else related to nutrition changed while maternal mortality did not change. This ought to enter our thinking occasionally.
Dr Belizan: When the calcium-toxemia hypothesis was first raised in 1980 [Belizan JM, Villar J. The relationship between calcium intake and edema-, proteinuria-, and hypertension-getosis: an hypothesis. Am J Clin Nutr 1980;33:2202–10], we focused on that part of the population with lower calcium intakes who had an increased incidence of pregnancy-induced hypertension. There are many trials from developing countries on calcium supplementation. The latest trial carried out in the United States and published in the New England Journal of Medicine [Levine RJ, Hauth JC, Curet LB, et al. Trial of calcium to prevent preeclampsia. N Engl J Med 1997;337:69–76] did not find any effect of calcium supplementation on preeclampsia.
The latest information from the Cochrane Controlled Trials Register [University of Adelaide, South Australia, 1998] on calcium supplementation for eclampsia had 67 women assigned to receive calcium or a placebo at 0.1–2.0 g/d and made an issue about women having adequate or low calcium intakes. Among women with adequate calcium intakes, there were no effects of calcium supplementation. Yet among women having a low-calcium diet, there was quite a large reduction of the incidence of preeclampsia with calcium supplementation.
Dr Scholl: Was the calcium intake in the trial reported in the New England Journal of Medicine quite high?
Dr Belizan: It was quite high, 1.1 g Ca/d, and the authors raised the issue of > or < 1.0 g/d. There are 6 trials from developing countries showing a reduction in the incidence of eclampsia [Carroli G, Duley L, Belizan JM, Villar J. Calcium supplementation during pregnancy: a systematic review of randomized controlled trials. Br J Obstet Gynaecol 1994;101: 753–8]. Our current position is that we need a huge trial in developing countries to replicate these studies, but among women with low calcium intakes.
Dr Scholl: As I recall, the Levine [Levine RJ, Hauth JC, Curet LB, et al. Trial of calcium to prevent preeclampsia. N Engl J Med 1997;337:69–76; Levine RJ, Esterlitz JR, Raymond EG, et al. Trial of calcium for preeclampsia prevention (CPEP): rationale, design, and methods. Control Clin Trials 1996;17: 442–69] trial was not completely negative. Was there not one-cell gestational hypertension?
Dr Belizan: Yes, and the women were also divided into quartiles of calcium intakes. The lowest quartile of calcium intake showed a reduction in eclampsia.
We are also trying to relate calcium intake to quality of life, which is related to the vicious cycle of hypertension. We recently published, in the British Medical Journal, a follow-up study on children aged 5–9 y of women who took 5 mg Ca/d during pregnancy [Belizán JM, Villar J, Bergel E, et al. Long-term effect of calcium supplementation during pregnancy on the blood pressure of offspring: follow-up of a randomized controlled trial. BMJ 1997;315:281–5.] We were quite surprised when we saw that the odds ratio for high blood pressure in children whose mothers received calcium during pregnancy was 0.59. Because of the effect of calcium supplementation during pregnancy on the quality of the life of the children—and the effects can be seen mainly in children—we were willing to do a study related to quality of life. We need to test what is going on in populations with a low-calcium diet to give a final answer and to propose better calcium intakes for women during their pregnancy.
Dr Loudon: This chart [Table 3 in Loudon I. Maternal mortality in the past and its relevance to developing countries today. Am J Clin Nutr 2000;72(suppl):241S–6S] shows that over a period of 14 y in the 1870s, 12% of maternal mortality was due to toxemia. Move forward 100 y to the 1980s and toxemia has moved up from third to first place and is, at 20%, ahead of all the other causes—by a long shot. The contributions from puerperal fever and sepsis have decreased, as has hemorrhage, but the contribution that has increased and has now become the major complication in obstetrics is hypertension. So, it is something in which we are interested, both in the developed world as much as in the developing one.
Dr Jacobs: The ranking on the left looks a lot like that in many underdeveloped countries today. So, perhaps that would be a useful comparison.
Dr Loudon: Puerperal sepsis does not seem to be nearly as large a part of maternal mortality in any of the developing world today as it was in the developed world in the past. If this is so, I suspect it is due to a marked decline in the prevalence of hemolytic streptococcus infections, almost certainly because there is other evidence to show this. Could anybody enlighten me on that?
Dr Maine: In the 1920s and 1930s in the now developed countries, deliveries were concentrated at a few places, whereas deliveries are still very much dispersed in developing countries. Even if a traditional birth attendant has a bad technique, she only delivers 1, 2, maybe 3 women a year. So, the chance of infection being spread around I think is much less.
Participant: That is the important point. In most developing countries, we still do not know the causes of death because the best technique is based on verbal autopsy—interviews with family members—and it is very difficult to accurately tease out causes of death from that process. There may have been better access to cause-of-death information in the 1930s in the United States and in the United Kingdom.
Dr Maine: Mistaking puerperal sepsis for anything else is unlikely. You can talk about switches between abortion deaths and hemorrhage deaths. How somebody records an obstructed labor death is a much chancier thing. Sepsis would have quite a different profile. For example, the time of onset would usually be several days after delivery.
Dr van den Broek: Sepsis is the number one killer in Blantyre right now and it is probably related to human immunodeficiency virus positivity. In a district hospital, where we have just done an audit, sepsis was the number one killer until we got an autoclave and got the system sorted out. Now it is the number 2 killer.
Mr Alnwick: I would like a point of clarification on Dr Loudon's statement that preeclampsia has gone up in England and Wales between the 2 time periods he mentioned. I seem to recall hearing that it was going down and continues to go down in rich countries.
Dr Rush: Dr Loudon referred to the proportion of deaths due to toxemia. It remains the leading cause of death, but the number of deaths has fallen sharply. In other words, the rate of death from toxemia has not been falling as steeply as have the rates of death from other causes.
Dr Ladipo: In the United Kingdom no maternal death has been attributable to eclampsia since 1964, and I think the reasons for that are better obstetric care, timely intervention, and early diagnosis of pregnancy-induced hypertension.
Dr Loudon: I am quite convinced that the reason deaths from hypertensive disease of pregnancy, which is really the correct term, have gone down is because of intervention and care. The only effective treatment is delivery in a hospital, early delivery. Nothing else has really worked, and I am sure that is the only reason. Relatively, deaths from hypertension have gone up because everything else has gone down faster.
REFERENCES