Eclampsia is a main component of maternal mortality, accounting for about 50,000 deaths a year worldwide.(1) Its prevalence varies widely according to standards of living. It is as high as 9% in Bangladesh,(2) while it is in the region of 1 in 3 000 in wealthy countries, where the rates of pre-eclampsia can be as low as 0.4%.(3) Let us recall that eclampsia is characterized by convulsions, while the diagnosis of pre-eclampsia is not based on specific symptoms, but on the association of de novo hypertension after midgestation (blood pressure >140 mm Hg systolic and/or >90 diastolic) and new-onset proteinuria (urinary excretion>0.3 g/d).
The strong correlations between standard of living and prevalence of eclampsia suggest the importance of nutritional factors in the genesis of this disease. This is why there is an urgent need for studies of nutrition in pregnancy among at-risk populations with low standards of living. At the present time the medical literature is inundated by studies about pre-eclampsia conducted in developed countries. These studies try to answer questions that are relevant to the practice of medicine in wealthy populations. They investigate mild forms of the disease and ignore nutritional factors. They look in particular at genetic factors and they try to develop prediction tests, searching for measurable manifestations of abnormal placentation that is said to initiate this disorder. Many of these efforts have focused on biochemical markers, primarily those suggesting endothelial dysfunction and activation of coagulation. The current studies cannot help moderating the rates of maternal deaths. They have not lead to significant advances in our understanding of the pathophysiological processes.
Today the priority is to plan studies that are feasible in poor countries. To achieve significant conclusions, large studies are needed among populations where the rates of genuine eclampsia are still high. Eclampsia is an easy diagnosis, since it is associated with convulsions and the mother eventually passes into coma. Thus, such studies are feasible among high-risk populations.
In such a context I find it useful to present our research project. Our hope is that other researchers will be in a position to conduct similar studies among other populations with low standards of living, particularly on the African or Asian continents.
We are in the process of designing a randomised controlled trial of food supplementation at the Alexander Fleming hospital, in a poor district of Rio de Janeiro, Brazil. In that hospital, where the number of births a year is around 6 000, eclampsia is still common. The objective is to evaluate the effects on the incidence of eclampsia of giving canned sardines to a certain number of women, as early as possible in pregnancy. The choice of canned sardines is based on theoretical considerations that we have previously presented in several occasions.(4,5) We’ll just summarize the main points of the hypothesis our study is supposed to test.
There is a widespread belief that reduced uteroplacental perfusion is the central pathophysiological process in preeclampsia. Several puzzling aspects of the disease challenge this belief. For example, a study looking at 97,270 births in 35 hospitals in Alberta, Canada, revealed that there is a significant association between preeclampsia and large-for-gestational-age infants, in addition to the well-known association with small-for-gestational age infants.6 Such findings are more easily interpreted if this multifactorial syndrome is presented as an expression of a maternal/fetal conflict. From this perspective, it is plausible that a large fetus's high demand for nutrients can be the root of conflict. Faulty placentation, inadequate maternal nutrition, and certain combinations of maternal and fetal genotypes are other factors that can independently increase the probability of conflict. Since mother and fetus do not carry identical gene sets, maternal and fetal interests are not always in harmony. The nature and the expression of such conflicts differ according to the species of mammals being considered.
For example, veterinarians use the term eclampsia to refer to a life-threatening disease that occurs in various mammals, including dogs. In this species, the so-called eclampsia is in fact related to hypocalcemia (it is a "puerperal tetany"). Of course, where dogs are concerned, the priority at the end of pregnancy and at the beginning of lactation is development of the bones of the offspring, which are much more developed at birth than the bones of other mammalian species.
Interspecies comparisons encourage us to raise new questions concerning the potential for conflict among humans. The spectacular brain growth spurt during the second half of fetal life is a specifically human trait. A conflict between the demands expressed by the fetus and which of those demands the mother can fulfill without depleting her body leads us to consider first the needs of the developing brain.
It is well known that the developing brain has special needs for long-chain polyunsaturates from the n-6 and n-3 families, particularly arachidonic acid (AA = 20:4 n-6) and docosahexaenoic acid (DHA = 22:6 n-3). At least 50% of the molecules of fatty acids that incorporate into the brain are represented by DHA. One can therefore assume that the most likely reason for a conflict is when the mother cannot keep up with the increased demands for DHA. A disease will be the price the mother's body has to pay to meet the needs of the developing brain.
The concept of maternal-fetal conflicts directs us to establish a new classification of the numerous well-documented biological imbalances associated with preeclampsia among humans. The first step should be to look at the status of maternal fatty acids at the end of normal pregnancy and in preeclampsia. We should look particularly at the group of long-chain n-3 polyunsaturates, which includes DHA and eicosapentaenoic acid (EPA = 20:5 n-3).
It seems that the central imbalance in human preeclampsia is the enormous discrepancy between the maternal plasma levels of DHA and EPA. In preeclampsia, the level of DHA is not significantly decreased, whereas the level of the parent molecule EPA is about 10 times lower than in normal pregnancy.(7) These are exactly the data we are expecting when assuming that brain development is a priority among humans. Such data are confirmed by the "Curacao study,"(8) which looked at the fatty acid compositions of maternal and umbilical cord platelets from preeclamptic women. Whatever the circumstances, the levels of DHA remain stable. This fact is noticeable when keeping in mind the low delta 4 -desaturase activity among humans.(9) The price of a stable DHA is an imbalance inside the family of n-3 fatty acids that is at the root of a long chain of further imbalances.
This is how one can understand the onset of a vicious circle when the demand in long-chain fatty acids is at its greatest: at that stage, if the amount of n-3 polyunsaturates available is low, the priority is to keep the level of DHA as stable as possible.
The use of biochemical markers of dietary intakes of lipids has demonstrated that a diet poor in n-3 fatty acids is a risk factor for preeclampsia. Studies of the erythrocyte fatty acids profile found that women with the lowest levels of n-3 fatty acids were 7.6 times more likely to have had their pregnancies complicated by preeclampsia as compared with those women with the highest levels of n-3.(10) A 15% increase in the ratio of n-3 to n-6 was associated with a 46% reduction in the risk of preeclampsia. Evaluating the fatty acid compositions of maternal platelets is another way to use biological markers of dietary fat intake. According to the Curacao study,8 the ratio of AA to EPA is significantly higher in maternal platelets of preeclamptic women (109.13 vs 78.13; P < .05).
These concordant and significant data suggest that when the amount of n-3 available is low, the first compensatory effect -- in order to maintain an adequate supply of DHA available -- is the collapse of the level of the parent molecule EPA: this precipitating factor explains the well-known imbalances in the system of prostaglandins and particularly the decreased ratio of prostacyclin to thromboxane-2. When the level of EPA has collapsed, there is no production of the physiologically inactive thromboxane-3. This leads to an overproduction of the physiologically active thromboxane-2, through a mechanism of enzymatic competition. Moreover, when the level of EPA is low, there is no production of the physiologically active prostacyclin-3. In normal pregnancy, the ratio of prostacyclin to thromboxane-2 in maternal blood progressively favors prostacyclin.
Our theory of preeclampsia is in a position to address the many intriguing aspects of the disease. One of them is that preeclampsia is principally a disease of first pregnancies. We must recall that the metabolism of n-3 fatty acids is influenced by parity.(11,12) The DHA content of cord blood phospholipids depends on birth order; in other words, the capacity to provide preformed DHA is depleted with repeated pregnancies. It is as if brain development is a higher priority in the case of a first baby.
Our perspective can also establish links between different approaches that have been used in the effort to prevent preeclampsia. Effective preventive action at the very beginning of the chain of events -- at the stage of faulty placental implantation -- cannot be considered.(13) The fact that a previous miscarriage, a previous blood transfusion,(14) or a long sexual cohabitation before conception(15) reduces the risk of preeclampsia confirms the probable importance of the immune response during that phase. It seems more realistic, on the other hand, to try to moderate the effects of the precipitating factors during the second half of pregnancy.
Theoretically, the most direct way to prevent preeclampsia would be to consume sea fish that is rich in n-3 polyunsaturates and also in minerals that are essential nutrients for the brain (eg, iodine, selenium, and zinc). This conforms with the geographical variations in the rates of preeclampsia and with the results of our encouragement of pregnant women to eat sea fish.(16)
Until now, all studies have been conducted in wealthy countries with very low rates of preeclampsia, such as Scandinavian countries. These studies usually involved controlled trials of fish oil supplementation that began during the second half of pregnancy. Based on the results of several studies, preeclampsia has not been dissociated from the framework of pregnancy-induced hypertension. For many reasons, it is therefore not surprising that meta-analyses(17) and systemic reviews (18, 19) have found insufficient evidence of the effects of fish oil on the risk of preeclampsia. In fact, most studies were too small to even address the issue of preeclampsia.
We can make such comments about our own study, conducted during the years 1991-1992 in a London hospital.(16) We randomly selected 499 pregnant women and encouraged them to increase their intake of oily sea fish and to reduce their intake of food rich in trans fatty acids. A hospital- and parity-matched control group included 500 pregnant women. Because of the study's size and the fact that the study population had a low rate of preeclampsia, we did not find it relevant to mention in the abstract or in the conclusion that there were no cases of eclampsia or severe preeclampsia in the study group vs 1 case of eclampsia with convulsions and 2 cases of severe preeclampsia in the control group.
It is remarkable that the only study that demonstrated highly significant effects of fish oil supplementation on the risk of “toxemia” was conducted in London by the People's League of Health during 1938-9, at a time when the rates of severe "toxemia" were in the region of 6%. This controlled trial was saved from oblivion by S.F. Olsen and N.J. Secher.(20) The authors randomized 5644 pregnant women to receive or not receive a dietary supplement containing vitamins, minerals, and halibut liver oil from about week 20 of pregnancy. A significant effect of treatment was seen in primiparae, with a 31.1% reduction in the incidence of “toxemia” (95% CI 5-50%, P = .021). Interestingly, no significant effect of treatment was seen with regard to the incidence of hypertension in the absence of edema and proteinuria.
Our understanding of preeclampsia also suggests that catalysts for the metabolism of unsaturated fatty acids should be preventive agents. Let us recall that only the precursor in the n-3 family (18:3 n-3) is abundantly provided by the land food chain. Magnesium,(21) calcium,(22) and zinc(23) are such catalysts and have been explored as preventive agents.
It also makes sense that, in order to prevent preeclampsia, the level of blocking agents of the metabolic pathways must be reduced as much as possible. Alcohol, pure sugar, and trans fatty acids are such blocking agents. A correlation has been established between the intake of trans fatty acids and the risk of preeclampsia.(24) Hormones such as cortisol are also known blocking agents. This can explain how the emotional state of the pregnant woman influences the risk of preeclampsia.(25) It is also theoretically important to avoid a fast destruction (via peroxidation reactions) of the available long chain fatty acids. The preventive effects of antioxidants are well documented.(26)
In theory, it is easier to meet the specific needs of the developing human brain when the diet includes some food from the sea, because the sea food chain provides preformed and abundant molecules of very long-chain fatty acids. The sea food chain has other characteristics. Any food from the sea is rich in iodine, a major component of thyroid hormones, which are involved in brain development. Imbalances of thyroid hormones (high ratio of thyroxine [T4] to triidothyronine [T3]) are associated with preeclampsia and should not be overlooked. Finally, it appears that pregnant women (and probably Homo sapiens in general) ideally need a certain balance between food from the land and food from the sea. Studies of preeclampsia in the framework of evolutionary medicine are needed.(27) In conclusion, preeclampsia may be understood as the price some human beings must pay for having a large brain when they are more or less separated from the sea food chain.
Those who understand that brain development is a priority among humans, those who are interested in eclampsia as the “disease of theories”, and those who are aware of the high rates of maternal deaths related to eclampsia in low-income populations should be easily convinced that a new generation of research is urgently needed.