点击显示 收起
Of all health indicators, the maternal mortality ratio shows the greatest discrepancy between developed and developing countries. This ratio relates maternal deaths to births, not to the population at risk (the latter would be women who are pregnant or within 6 wk of birth or abortion, a number that is not usually available). The 5 main causes of maternal mortality are induced abortion, puerperal infection, pregnancy-induced hypertension (preeclampsia and eclampsia), hemorrhage (including anemia), and obstructed labor. Nutrition is often associated with the last 3 causal factors. A major component of prenatal care is iron or iron and folate supplementation because this is believed to prevent anemia during pregnancy, which is important for preventing hemorrhage and possibly even death during childbirth. To identify program directions for future efforts, the US Agency for International Development convened a meeting of experts in the field to discuss what is known about the nutritional factors associated with maternal mortality and whether the underlying premises and expectations for reducing maternal mortality through nutrition interventions are based on scientific criteria.
Rush, who has long focused on this issue, was asked to write an article on nutrition and maternal mortality in developing countries. His strategy was to return, wherever possible, to the original data on which current policies are based and to look at them afresh. Many of these data were often incomplete, conflicting, or ambiguous; depending on clues for inferring underlying patterns without clear-cut results was sometimes necessary. Rush (1) presents an overview of nutrition and maternal mortality in which he challenges the public health nutrition community to "assess its ignorance forthrightly." He notes that policymakers and funding agencies must be made aware that lowering maternal mortality and improving maternal and child health significantly will require far better information than is now available. This requires both revisiting extant data and undertaking new research. Rush also points out that nutrition programs must be better integrated with (reproductive) health programs because all women have a right to good access to emergency obstetric care. We need to ensure that such care is available when we develop programs to further reduce maternal mortality.
In his article, Rush concludes that existing data are not sufficient to show that nutrition affects abortion, puerperal infection, and pregnancy-induced hypertension (ie, preeclampsia and eclampsia). Nutrition, however, can be an important determinant of hemorrhage and obstructed labor. Maternal mortality is much higher among the severely anemic, but data are not available to show whether this mortality is lowered by treatment of anemia (except when blood transfusion therapy is used to treat very severe anemia). Only 2 studies related moderate degrees of anemia to mortality; 1 study showed moderately increased maternal mortality and the other showed no increase. In most populations, the proportion of mildly or moderately anemic women far exceed that of severely anemic women. Few, if any, pregnancy anemia control programs in developing countries differentiate levels of anemia (ie, mild, moderate, and severe) or attempt to specify the cause of anemia, and severe anemia is most often due to complex, multiple causes. Rush concludes that the problem of severe anemia, a medical and obstetric emergency, is not being directly or appropriately addressed.
The risk of obstructed labor is closely related to the physical size (ie, growth) of both the mother and the fetus (2–6), but nutritional strategies for increasing adult stature are rare. Supplementary feeding appears to have little benefit after 3 y of age (7), and data show that improved nutrition and environmental conditions for short girls in developing countries have mixed benefits for reproductive health (8, 9). Published research on the association between fetal size and obstructed labor is scanty. Nevertheless, the few existing studies are consistent in showing that the incidence of operative delivery increases sharply with increased fetal size (K Merchant, unpublished observations, 1991). This applies to moderate-size and large newborn infants, in whom there is a higher risk of obstructed labor with a birth weight of 3.0–3.5 kg than with a birth weight of 2.5 kg. The risk for obstructed labor is especially severe among women shorter than 1.5 m (9, 10). The problem is exacerbated among young and primiparous mothers. Rush concludes that the knowledge base necessary to model the competing risks between the benefits and dangers of increased fetal size through supplementary feeding of pregnant women does not exist.
Rush's article is followed by a series of articles that review conventional wisdom. Loudon (11) discusses the relevance and appropriateness of collecting maternal mortality data. Using historical data from Europe and the United States, Loudon shows that the main determinant of maternal mortality before 1937, and its decline since the 1930s, was the overall standard of maternal care provided by birth attendants. Poverty and associated malnutrition played little part in determining the ratio of maternal mortality in developed countries. This means that data on the attributable cause of death are important for developing appropriate interventions. The question of how to collect these data in developing countries, however, remains open.
Although much money is spent on purchasing and distributing prenatal iron or iron and folate supplements through public health systems worldwide, data do not show that these interventions are effective in reducing the prevalence of anemia. This ineffectiveness is generally blamed on poor compliance or inadequate supply of supplements. Whether the underlying assumption—that supplementation can reduce the prevalence of mainly mild to moderate anemia—is correct continues to perplex biologists and policymakers alike. van den Broek and Letsky (12) present new data from a study in Malawi that assess the etiologic factors that could be responsible for pregnancy anemia, namely iron, folate, vitamin B-12, and vitamin A status. The iron status of anemic women was determined only from bone marrow aspirates. Although between one-quarter and one-half of the women studied had at least one micronutrient deficiency, one-fifth were not deficient in any of the micronutrients studied but were anemic. Among the latter, more than one-half of the women had notably increased C-reactive protein concentrations, showing that the prevalence of iron deficiency can easily be overestimated if inflammatory disease is not considered. Because of the lack of a control group, no conclusions can be drawn from this study. Nevertheless, van den Broek and Letsky note that iron supplementation alone is unlikely to provide the optimal solution for controlling anemia because infection is likely to be an important factor. This is clearly an area that warrants further work.
From a policy perspective, it is assumed that positive nutrition (more nutrients) is intrinsically beneficial. The establishment of a tradition and practice of iron supplementation seems to have resulted in insufficient attention being given to the need to recognize the fact that safety precedes efficacy in the development of new approaches. Something can be efficacious but not safe (eg, thalidomide for pregnant women). Rush raises doubt about both the efficacy and the safety of routine iron supplementation. Bothwell (13) and Beaton (14) discuss the evidence for a clinical benefit of iron supplementation in pregnancy. Bothwell reviews the factorial method for determining iron requirements in pregnancy. Iron requirements decrease in the first trimester, after which they rise steadily, and this matches the changes in iron absorption. The total iron requirement in a 55-kg woman during pregnancy is 1000 mg, but the amounts of iron that can be absorbed from even an optimal diet are less than the iron requirements later in pregnancy. A woman, therefore, must enter pregnancy with iron stores of 300 mg if she is to fully meet her requirements. Bothwell contends that any deficit in iron can be met by supplementation therapy and that it is the inadequacies in healthcare delivery systems that limit the effectiveness of large-scale interventions. Bothwell says that, on a long-term basis, iron fortification offers the most cost-effective option for the future.
In contrast, Beaton (14) argues that in the factorial method, the target hemoglobin concentration appears to be based on the maximum concentration that can be achieved with iron supplementation of well-nourished women, a situation that is likely to reduce the absorption efficiency of iron utilization. Beaton suggests that the present approach may seriously overestimate iron need and discourage food-based programs. The present approach may also result in iron supplementation programs being declared as failures when, in fact, many may be successful in preventing functional effects of iron deficiency anemia. The latter is illustrated with data from a comparative study of daily and weekly iron supplementation (15). Beaton concludes that the functional criteria for anemia and the functional goals for hemoglobin concentrations to be achieved in pregnancy need to be defined.
The topic of iron safety is discussed by Ray Yip (16), whose article examines the significance of an abnormally low or high hemoglobin concentration during pregnancy. He notes that a causal relation between moderate anemia and poor perinatal outcomes has not been established. Severe anemia, however, may be a direct factor of poor pregnancy outcomes because of hypoxia and heart failure. Yip also notes that high hemoglobin concentrations have been associated with poor health outcomes, but a causal association between higher maternal hemoglobin concentration and increased risk of adverse birth outcomes has not been shown. He suggests that this observed association could be better attributed to hypertensive disorders of pregnancy and preeclampsia; higher-than-normal concentrations of hemoglobin should be regarded as a risk indicator of pregnancy complications, not necessarily a sign of adequate iron nutrition. Maine (17), however, says that no available data suggest that maternal mortality behaves like a nutritional disease, which would tend to discount this suggestion. Yip's conclusion is that iron supplementation does not increase the hemoglobin concentration higher than the optimal concentration needed for oxygen delivery. The question of whether iron supplementation can cause excess mortality remains open.
With the recent interest in moving from iron and folate supplementation to multiple micronutrient supplementation, Ladipo (18) discusses micronutrients as they relate to pregnancy physiology and biology. He notes that although multiple micronutrients are often consumed by pregnant women in developed countries, the benefits of theses micronutrients are limited, except for prophylactic folic acid consumed during the periconceptional period. Women in developing countries, however, may benefit from selective micronutrient prophylaxis in pregnancy, but there must be a basis of deciding what to include in a generic supplement. Ladipo concludes that improving the general nutritional status of women would be a more cost-effective approach for long-term sustained health than would a policy of providing prophylactic micronutrients, iron and folate or multiple, to all pregnant women.
Turning to the other causes of maternal mortality, Konje and Ladipo (19) discuss obstructed labor. This is one of the most common preventable causes of maternal and perinatal morbidity and mortality in developing countries and has been shown to be associated with maternal stature. Short stature is caused by early childhood malnutrition and can be intergenerational. Specific nutritional deficiencies such as rickets and osteomalacia may also play a role. Recognizing the causes of obstructed labor is important if the complications are to be prevented. Adequate prevention, however, can be achieved only through a multidisciplinary approach involving clinicians, governments, and organizations that deal with social, cultural, and health factors.
Maine (17) discusses the role of nutrition in preventing toxemia in pregnancy. She notes that very little is known about the causes of toxemia. Although many nutritional factors have been suggested as playing a causal role in toxemia, mortality from this disease has not varied over time and between circumstances as one would expect a nutritional disease to do.
During the discussions on programmatic implications and the need for further research, agreement was reached that existing programs do not need to be changed (20, 21). Clearly, a review of the existing data on the efficacy of iron supplementation for moderate anemia is needed as well as new research. More attention also needs to be given to understanding the causes of anemia, especially severe anemia, to develop and implement programs that can have an effect on anemia-associated maternal mortality.
REFERENCES