HIV/AIDS, Infant Nutrition, and Human Rights

By George Kent
Sept. 2000

Abstract

If the Human Immunodeficiency Virus (HIV) can be transmitted through breastfeeding, how should infants of HIV-positive mothers be nourished? Alternatives to breastfeeding may be even more risky than breastfeeding, and some methods of breastfeeding may be less risky than others. Parents need to be able to make informed choices regarding feeding strategies, but this is impossible where information is wholly inadequate. There is uncertainty regarding the likelihood of transmission of the virus through breastfeeding, uncertainty regarding the consequences of HIV infection in infants, lack of knowledge of ways in which the negative consequences might be ameliorated, inadequate understanding of the alternative approaches to feeding, etc.

Guidelines for infant feeding should be based on the human right to adequate nutrition, established in the Universal Declaration of Human Rights, the International Covenant on Economic, Social and Cultural Rights, the Convention on the Rights of the Child, and other international agreements. Application of the basic principles of nutrition rights in relation to HIV/AIDS implies that (1) parents have not only a need but also a right to be well informed, and (2) they have a right to good information not only about breastfeeding but also about a broad range of alternative feeding methods. The right of informed choice implies a right to good information. Means are suggested for realizing that right.

Introduction

It is generally acknowledged that the Human Immunodeficiency Virus (HIV) can be transmitted from mother to infant through breastfeeding. As a result, there is a serious debate now underway regarding the feeding of infants by mothers who are diagnosed as HIV-positive. The major tension arises out of the fact that under some circumstances the dangers of breastmilk substitutes may outweigh the risk of being infected with HIV through breastfeeding.

It is widely agreed that medical and other decisions relating to health care generally should be made on the basis of the principle of informed choice (or informed consent). This principle may sometimes be suspended, as in cases in which the individual is not competent to make informed choices, or where the risks to individuals or to the society as a whole warrant suspension of the principle. The argument here is that there is no scientific or policy basis for suspension of the principle for situations in which mothers diagnosed as HIV-positive are faced with the challenge of deciding how to feed their infants.

In this situation, as in others, there are two types of possible violation of the principle. One is that there may be violations of the "choice" part, through attempts to coerce the mother to take a particular course of action. I have discussed the issue of coercion elsewhere (Kent 1999a; Kent 1999b). There is also the possibility of violation of the "informed" part. That is the central concern of this analysis.

Most observers agree that parents should be free to make informed choices, but that freedom is of little value when the information available is inadequate. Governments have responsibilities relating to the quality of our health care and the quality of our food, and they also have responsibilities regarding the quality of our knowledge about these things. Thus the core question of how to feed infants in the context of HIV/AIDS raises serious human rights issues.

After presenting an overview of the debate, I suggest a framework for structuring our thinking about the issues. Then the human rights considerations relating to nutrition are introduced. I present basic principles on human rights and infant nutrition, and follow this with a brief discussion of possible tensions between the rights of the infant and the rights of the mother. This then leads to recommendations regarding the application of the principles in the context of HIV/AIDS, especially in relation to the right to the information required to make an informed choice regarding feeding methods. A line of reasoning for arriving at a prudent choice in the absence of the necessary information is suggested. And finally, a research plan for obtaining the necessary information is proposed.

The HIV/AIDS and Infant Feeding Debate

In recent years there has been a great deal of concern with the possibility of transmission of HIV from mother to child, in the uterus, during the birth process, or through breastfeeding. Predictably, the suggestion that the virus can be transmitted through breastmilk has raised concern about whether mothers who are HIV-positive should breastfeed their infants. If there is some chance that HIV can be transmitted through breastfeeding, how should mothers who are HIV-positive feed their infants? The experts are having trouble agreeing on appropriate guidelines.

In October 1995, the U.S. Food and Drug Administration’s FDA Consumer magazine published an article that said without qualification, "Women who are HIV positive should not breast-feed (Williams 1995)." Presumably it referred only to the United States context.

In May 1997 the Joint United Nations Programme on HIV/AIDS (known as UNAIDS), the World Health Organization (WHO), and the United Nations Children’s Fund (UNICEF) issued a joint "HIV and Infant Feeding" policy statement. It took an "informed choice" approach, meaning that mothers were free to choose the method of feeding, but should be fully informed of the benefits and risks of each in their particular circumstances.

Similarly, in June 1997 the American Academy of Pediatrics’ Committee on Pediatric AIDS said "in the United States where suitable alternative sources of nutrition are readily available, HIV-infected women must be counseled not to breastfeed or provide expressed breast milk to any infant (CPA 1997)".

In the early 1990s, UNICEF, WHO, and UNESCO collaborated in preparing a small booklet, Facts for Life, to communicate the most important guidelines relating to child care (Adamson undated). The section on AIDS said:

A mother with the AIDS virus should continue to breastfeed her baby. There is a very small risk that the AIDS virus could be passed on to the baby by breastfeeding. But the risks of bottle-feeding a baby are known to be very much greater, especially in a poor community.

In May 1998, the relevant provisions of Facts for Life were updated as follows:

• If a mother is infected with HIV/AIDS then there is a one in seven risk that breastfeeding may give the virus to her uninfected baby. If a mother with HIV/AIDS chooses to feed her infant with infant formula or other breastmilk substitutes there is a risk of her baby getting diarrhoea and other illnesses which could lead to malnutrition and death. The extent of this risk depends on local circumstances.

• Mothers who are infected with HIV/AIDS should consider not breastfeeding only if they are sure of regular supplies of alternatives to breastmilk, as well as clean water, and sufficient time and fuel to boil water for every feed. Feeding cups, not bottles, should be used, even for newborn babies.

• Mothers who have HIV/AIDS need to know the risks to their infant's health of both breastfeeding and not breastfeeding so they, and the infants’ fathers, can make informed decisions about what is best for their baby in their own local and personal circumstances.

• For mothers who do not have HIV/AIDS, breastfeeding is the best way to ensure the survival, protection and healthy development of their babies.

• It is your right to have access to voluntary and confidential AIDS testing and counseling (UNICEF 1998).

In June 1998 UNAIDS announced New Initiatives to Reduce HIV Transmission from Mother-to-Child in Low-Income Countries (UNAIDS 1998a). They centered on joint UNAIDS/ UNICEF/WHO pilot projects aiming "to offer voluntary and confidential HIV counseling and testing to pregnant women, and to provide those who learn they are infected with antiretroviral drugs, better birth care and safe infant feed methods". Regarding feeding:

UNICEF will work with governments and suppliers of milk products to identify practical ways of helping pilot countries to provide alternatives—such as home-prepared and commercially-prepared infant formula—to mothers participating in the projects. WHO, UNICEF and the UNAIDS Secretariat will continue to protect, promote and support breastfeeding as the best feeding method for infants whose mothers are HIV-negative or do not know their HIV status. Among other things, this means ensuring that the methods used for producing and distributing alternatives to breast milk comply with the International Code of Marketing of Breast-milk Substitutes and subsequent resolutions of the World Health Assembly.

Old debates about the merits of formula feeding are being revisited in the context of HIV/AIDS. A journal in South Africa ran a special issue on the question (Special Report 1997). In a letter to the influential British medical journal, The Lancet, Michael Latham and Ted Greiner, experts on breastfeeding, said they were troubled by "the new proposals to conduct large-scale trials in several developing countries to replace breastfeeding with formula feeding in HIV-1 positive mothers". They said:

We are concerned that WHO and UNICEF will invest major resources in formula feeding and few into alternatives, such as modified breastfeeding, heat treatment of expressed breastmilk to kill the virus, wet nursing, donation (or even sales) of breastmilk, and use of animal milks or homemade formulas. These options are preferable to the use of infant formulas in poor communities. None of them are easy, or ideal, but they warrant careful study. Much of the successful work over the years to stem the use of commercial breastmilk substitutes in poor countries is now threatened. The involvement of the commercial infant formula industry, both in deliberations leading to the new policy and also in offering to make their products available, is troubling.

We recommend that the UN agencies assess carefully the economic, social, and health consequences of their new policy, and that they provide adequate support to allow investigations of alternative methods. It is a grotesque reality that all HIV-1 infected mothers cannot have full coverage of antiretroviral therapy, that so many mothers and infants do not have access to adequate health care, and that inequities lead to a high prevalence of malnutrition. Given this unfortunate situation, is it wise to be recommending the costly and risky approach of formula feeding for infants born to poor HIV-1 infected mothers (Latham 1998)?

In June 1998 the Steering Committee of the World Alliance for Breastfeeding Action issued a statement, WABA Position on HIV and Breastfeeding. It said, in part, that

WABA is concerned about what appears to be recent changes in the WHO, UNICEF and UNAIDS policy regarding breastfeeding and HIV. We are especially concerned that these changes appear to put major stress on the use of infant formula and less on alternative feeding methods (WABA 1998).

The statement closed by saying "Extreme caution must be shown in involving the commercial firms that have direct economic interests in the outcome of such policy deliberations."

In June 1998 WHO published three manuals on HIV and Infant Feeding (WHO 1998). They offered a comprehensive overview of the issues, but focused on the objective of preventing HIV transmission through breastfeeding. They did not acknowledge that under some circumstances, it might be better to risk transmission of the virus and try to minimize or ameliorate its consequences. With qualifications, the approach advocated finding ways to provide breast-milk substitutes to infants of HIV positive mothers, possibly with the support of government subsidies. The health risks and the economic plausibility of this approach were not assessed.

On July 26, 1998 the New York Times ran a front-page article on "AIDS Brings Shift in U.N. Message on Breast-Feeding". It began...

Countering decades of promoting "breast is best" for infant nutrition, the United Nations is issuing recommendations intended to discourage women infected with the AIDS virus from breast-feeding.

It added:

In its directive, the United Nations said it was deeply concerned that advising infected mothers not to breast-feed might lead many mothers who are not infected to stop breast-feeding. To reduce that possibility, it is advising governments to consider bulk purchases of formula and other milk substitutes, and to dispense them mainly through prescriptions (Altman 1998).

On October 5, 1998 the United Nations’ Committee on the Rights of the Child held a Day of General Discussion on "Children Living in a World with HIV/AIDS". Its report said:

Participants discussed at length the need for additional research and to look for strategies that minimize the risk of mother-to-child transmission of HIV without automatically promoting the use of bottle-fed formula. Alternatives such as warming mother’s milk to destroy the virus, or establishing breast-milk banks, [using] wet nurses, etc. need to be better explored, and health care workers must be trained on the availability of such alternatives and on the need to support the mother’s decisions, with primary consideration given to the best interests of the child (CRC 1998).

Setting the task as being "to look for strategies that minimize the risk of mother-to-child transmission of HIV" frames the question too narrowly. Careful attention also should be given to health outcomes, and to the many factors other than the virus that can affect those outcomes.

Framework

The debate has not been clearly structured. The core question is, in the context of HIV/AIDS, how should parents be advised to feed their infants? The objective is to formulate clear advice that health workers can give to parents in a variety of different situations. Guidance for the parents should provide not only scientific sorts of information but also suggestions on how to think through the question. Thus health workers’ advice might come in the form: "If you are in this situation and you have these preferences, you should feed your infant by this method." The scientists and policy analysts should be able to back up the advice with clear evidence and argument.

In principle, health workers could be provided with a contingency table to determine which advice is appropriate under particular circumstances. We can imagine a contingency table with three columns. The first column specifies the particular conditions that matter in giving advice, such as information on local circumstances; clinical, social, and economic data on the family in question; and information on their beliefs and preferences. The second column describes the feeding advice appropriate for the set of conditions described in that row. The third column provides the evidence and argument supporting that advice. Each row can be read across, filling in the sentence, "If conditions A prevail, then take action B, for reasons C". For those contingencies in which it is not clear what advice ought to be provided, the researchers and policy analysts have more work to do.

In this debate it is important not to confuse the advice given to parents or health workers with the demands made of scientists and policy analysts. Parents and health workers cannot be expected to undertake the basic research and analysis work that is needed. Most importantly, parents cannot be expected to make responsible informed choices when they do not have the information that is required to make such choices.

In trying to work out appropriate advice, several different kinds of concerns arise:

(a) Likelihood of transmission.  One report says that prior to the widespread use of antiretroviral therapy, the rate of transmission of HIV from HIV-positive mothers to their infants ranged from 14% to 33% in the United States and Western Europe, and in the developing world, rates as high as 43% have been reported (Stoto 1998, Section 4, p. 1). According to another report, in the United States, "The maternal to infant transmission rate is approximately 20% to 30%, with the majority of infants who are born to an infected mother being ultimately uninfected (CPA 1997)". Another study reported a transmission rate without drug treatment in the U.S. of 15% to 30% (Burr 1997).

At the XIII International AIDS Conference held in Durban, South Africa in July 2000 it was reported that "HIV transmission rates have been reduced to less than 3% in most cohorts in the United States." This was attributed primarily to the widespread use of highly active antiretroviral therapy, known as HAART. In Europe it was estimated that overall rate of vertical transmission in Europe declined from 16% in 1994 to 7 % in 1998-99 (Lala 2000).

These figures are estimated rates of transmission through all three pathways—during pregnancy, in the birth process, or through breastfeeding. The rate of transmission through breastfeeding itself is a fraction of this figure. Some have estimated that breastfeeding by HIV-positive mothers increases the risk of HIV infection of the infant by about 14 percent (Dunn 1992). Some reports suggest "the incremental risk of transmitting HIV infection to the breastfeeding infant range from 3% to 12% in various African populations (CPA 1995)."

One study estimated that only about one percent of infected infants are infected through breastfeeding (Burr 1997). If, as this source estimates, the transmission through all three paths is between 15% and 30%, this means that at most about 0.3% of the infants of HIV-positive mothers are at risk of infection through breastfeeding.

For the U.S., it has been estimated that where there is transmission of the virus to the infant, 70% to 75% of the cases occur during delivery, and 25% to 30% occur in utero (Stoto 1998, Section 4, p. 1). This suggests that at most only about 5% of the cases of infection occur as a result of breastfeeding. If, as they estimate, the overall transmission rate is at most about 33%, then the risk of infection of infants by HIV-positive mothers through breastfeeding is at most about 1.65%.

There are many uncertainties about these figures. While the transmission of the virus through breastfeeding has been widely discussed, there is in fact little firm knowledge about how likely it is to happen.

(b) Influences on likelihood of transmission. While there may be some sort of broad average regarding the likelihood of transmission, it is important to keep in mind that the likelihood may differ for different subpopulations or different kinds of circumstances. One must be careful about generalizations. For example, it has been estimated that while the average rate of mother-to-child transmission of the virus is around 25%, rates vary from less than 14% in Europe to 45% in sub-Saharan Africa (Recommendations 1998, p. 313). The advice may need to be different according to whether the family being advised lives in a poor area or rich area, or whether the family itself is of low or high socio-economic status, or whether the mother diagnosed as HIV-positive has started showing signs of AIDS.

These kinds of variability suggest that transmission risks can be meaningfully assessed only with regard to particular kinds of groups or conditions. However, one then encounters the fact that it is difficult to estimate the risk of virus transmission where there are small numbers of cases. In the United States as a whole, it is estimated that there have been 7,943 AIDS cases in children resulting from HIV infection from their mothers from 1988 up through the end of 1999 (CDC 1999, p. 14). Only 232 new cases of AIDS in infants of mothers with HIV were reported in 1999 (CDC 1999, p. 14). If we are concerned with other categories, such as particular states or cities, particular ethnic groups, or asymptomatic mothers, the numbers can become very small, thus making it difficult to produce meaningful estimates of the likelihood of transmission.

The transmission likelihood also may be influenced by different kinds of treatments. Some researchers attribute the sharp reduction of pediatric HIV cases in the United States to widespread use of AZT treatments. There does not appear to be any evidence regarding the extent to which such treatments might affect the likelihood of transmission via breastfeeding in particular, partly because mothers adopting such drug treatments generally are instructed to not breastfeed. Recent trials with nevirapine with breastfeeding mothers showed that with this drug treatment, "reduction in mother-to-infant transmission of HIV was sustained even though the infants were breastfed (Susman 2000)." This suggests that while the drug treatment may limit transmission during pregnancy and during delivery, it has no impact on transmission through breastfeeding.

Maternal micronutrient deficiencies may increase incidence of infections and viral load in the mother's body fluids, including breastmilk. Furthermore, maternal micronutrient deficiencies may influence the micronutrient status of the infant, thereby affecting the infant’s immune functions and susceptibility. Maternal vitamin A deficiency in particular could lead to increased exposure of the child for HIV (Semba 1994; Friis 1998). Some studies suggest that supplementation with multivitamins might be better than vitamin A alone (Semba 1999). However, other studies raise doubts about the potential value of vitamin supplementation for limiting vertical transmission (Fawzi 2000). Some of these studies do not make a sufficiently clear distinction between the breastfeeding route and other paths of mother-to-child transmission of the virus. Further research is needed to identify the impact of vitamin supplementation on HIV transmission via breastfeeding.

There may also be differences depending on methods and timing of breastfeeding. A mother in advanced stages of disease may be more likely to transmit the virus through breastfeeding. In addition, because of her illness, she may be less able to sustain breastfeeding, and less able to care for her infant whether the infant is infected or not. Also, there may be differences in the virus content of colostrum and early human milk compared with later milk.

It now seems clear that the likelihood of transmission is reduced if the infant is fed exclusively through breastfeeding, rather than in combination with other foods (Coutsoudis 1999). It has been suggested that feeding with anything but breastmilk may risk damaging the gut of the infant, thus increasing the ease with which the HIV can pass into the child’s blood.

(c) Likely consequences of infants’ HIV infection via breastfeeding. It has been reported that in Thailand about half the children born with HIV develop AIDS rapidly and die within two years (UNAIDS 1996). However, there is no consistent data, over time and through space, on the likelihood of death or disease resulting from HIV infection. The risks may well depend on a variety of conditions. The pattern for children who are not born with HIV but contract the virus through breastfeeding could be different from the risks encountered under other conditions. The immunological properties of breastmilk could outweigh the effects of the virus, or some other mechanism might intervene.

A physician in Uganda claimed, "If mothers who are infected with the virus do not breastfeed, their children will have a far better chance of survival." How could he be so sure when he also acknowledged that "In rural areas, 85 percent of babies will die from dirty water used in formula (Specter 1998)"?

If an infant contracts the virus through breastfeeding, what are the likely consequences in terms of morbidity and mortality? How frequently does the virus lead to increased morbidity and mortality associated with AIDS? It may be that the course of HIV/AIDS is different in infants than it is in older people.

It is often recommended that mothers who are designated HIV-positive should not breastfeed because that would prevent the transmission of the virus through breastmilk (Williams 1995; Ramanathan 1996). While it is clear that without breastfeeding there would be no possibility of transmission through that route, it is not clear whether the infant would be better off as a result.

In discussions about feeding choices by HIV-positive mothers, there is a preoccupation with the possible transmission of a virus through breastfeeding. There is practically no discussion of the consequences of that transmission. In the absence of explicit information, people tend to assume the worst.

For the purposes of formulating feeding advice, however, it is not necessary to know the likelihood of virus transmission via breastfeeding. To guide policy as to whether an HIV-positive mother should breastfeed or use some other specific feeding procedure, we need to know and compare the consequences, in terms of the infant's health, that are likely from taking each of these courses of action. The feeding strategy is the key independent variable and health outcome is the key dependent variable. HIV transmission via breastfeeding is an intervening variable that need not be visible in the analysis. For policy purposes, the research needs to focus on likely consequences for the infant, not on the postulated intervening mechanisms. Moreover, it is much easier to assess health outcomes than to try to track a poorly identified virus whose role in causing disease is not entirely clear.

Most critically, we need to know how the health outcome prospects for breastfed infants of HIV positive mothers differ from the prospects of those who are not breastfed. Currently, we do not have adequate information on what this difference is for infants of infected mothers who are born virus-free, and we do not know what it is for infants of infected mothers who are born with the virus.

The central question is, what are the likely differences in infants’ health outcomes over the long run for different feeding methods. As a practical matter, it is generally necessary to assess health status at particular points in time, e.g., at monthly or yearly intervals. For this reason, it may be useful to have information about the infant’s HIV status for the purpose of predicting future health status. It is certainly better for an infant to be healthy and HIV-negative than to be healthy and HIV-positive. An infant who is HIV-positive has a higher likelihood of becoming unhealthy in the near future. However, we do not yet have adequate data to make good estimates of this likelihood. Information about the infant’s HIV status may be used as a surrogate marker for future health conditions, but its quality as a marker has not been established.

While the health consequences for the infant may be the central concern, there are other possible impacts of the feeding decision that should be considered. For example, formula feeding may reduce the duration of infertility of the mother, thus leading to a new pregnancy and the birth of another infant. The burden may lead to further deterioration in the health of the mother. The new infant is also placed at risk of infection. Moreover, since formula feeding is very costly, this feeding choice may impoverish the entire family, with dire consequences for the entire family.

(d) Knowledge of Mother's HIV status. The current consensus among international agencies is that advice regarding feeding alternatives should be provided only when the health worker knows for certain, preferably on the basis of laboratory testing, that the mother is HIV-positive. This position is based on several concerns, including the need to protect and promote breastfeeding among women who are not HIV-positive, and the need to respect the privacy of women who may be HIV-positive. Confidentiality is important because in many settings there is a well-justified fear of stigma and discrimination against those who are identified as HIV-positive. In some cases even the women’s spouses may not know be told their diagnoses, thus creating extremely delicate situations. The task of formulating advice is especially difficult because there may be circumstances in which concern for confidentiality must be balanced against concern for the interests of the child.

There are at least four different kinds of concerns about this premise that advice should be offered only when a laboratory test has demonstrated that the mother is HIV-positive.

First, some feel that currently available methods of testing for HIV are unreliable.

Second, because of costs, reluctance to submit to tests, administrative difficulties, and other factors, widespread testing is not feasible, especially in developing countries.

Third, it appears that most people now described as having AIDS have not been tested. What advice, if any, is to be given to those who are presumed to have AIDS but have not been tested?

Fourth, why shouldn't advice be given to all, rather than being given only to those who have been tested and shown by the test to be HIV-positive?

Further thought needs to be given to the ways in which advice ought to depend on the character and quality of different parties’ knowledge about the mother’s HIV/AIDS status. At any given moment, the mother, the health worker, the mother's spouse, and the surrounding family and community may have different beliefs regarding her status. Moreover, those beliefs may be influenced in different ways depending on the feeding practice she adopts.

Similarly, more thought must be given to the question of what advice, if any, should be given when the mother's HIV/AIDS status is not definitely known—not to the health care worker, and perhaps not to the mother herself. Advice should be considered at least for those cases in which there is good reason to believe that the mother is likely to be infected.

Where a health worker does not know a mother’s HIV status, it might be appropriate to give advice in a conditional form: "If you believe you are HIV positive, you should... ." Conditional advice of this form allows the health worker to respect the mother’s concerns for confidentiality. It also makes it possible to advise women in groups.

(e) Knowledge of Infant's HIV Status. There are serious ambiguities regarding the very meaning of "HIV infection". The terminology that emphasizes virus transmission suggests that if a small number of "copies" of the virus pass from the mother to infant, that infant is thereby infected, by definition. However, some analysts emphasize the importance of "seroconversion", the change in the blood stream that can result from the activity of the virus. Transmission is instantaneous, but seroconversion, the process of infection, is said to take weeks or even months. The evidence of seroconversion is the presence of antibodies in the bloodstream. It is important to distinguish infection in the first sense (presence of the virus) from infection in the second sense (seroconversion) because the first does not always lead to the second.

The PCR test attempts to detect the presence of virus particles in the bloodstream, no matter how small their number. In contrast, the Western Blot and ELISA tests are designed to detect antibodies. In some studies all three tests are treated as more or less equivalent indicators of HIV infection, but they in fact measure quite different things.

If we are concerned with the question of whether an infant becomes HIV-positive as a result of breastfeeding, we must first decide whether that refers simply to the presence of the virus or to seroconversion as signaled by the presence of antibodies. To prove that either outcome results from breastfeeding, it would be necessary to show that the infant is HIV-negative at birth and then HIV-positive after a period of breastfeeding.

The seroconversion process takes some time. Also, maternal antibodies are passed to the infant during the birth process. It is not possible to distinguish between maternal antibodies and the infant’s own antibodies in the newborn infant’s bloodstream. Thus, the antibody tests cannot be used to isolate the role of breastfeeding in virus transmission.

The PCR test is the remaining option, but questions have been raised about its reliability and validity. Some claim it frequently yields false positives. An interesting test of the test could be made by applying it to a sample of infants of mothers diagnosed as HIV-negative. If the test detected the virus in these infants, one would have to reconsider the question of what it is that is really being measured.

Those who support the PCR test acknowledge that it cannot reliably detect HIV in the infant until about three weeks after birth. Thus, there may be no good way to determine whether HIV-infected infants are infected as a result of breastfeeding rather than during pregnancy or in the childbirth process. As indicated above, in Subsection (a), it may be that only a small proportion of infected infants is infected as a result of breastfeeding. The proportion is not clearly known, and it may not be knowable.

For practical decision-making, the critical issues are not the ones relating to technical definitions and scientific explanations of how the virus behaves. The key issue of concern here is that there is no clear and systematic information about how effective these tests are for predicting the future health of the infant. The discussions on the issue tend to imply that a positive result on any of these tests is a sure sign of impending death, but there is no solid published evidence to support that assumption.

Research may eventually show that the best method for feeding infants of mothers diagnosed as HIV-positive does not depend on the HIV status of the infant. In that case, tests of the infant would be irrelevant in relation to the feeding decision.

(f) Feeding Alternatives. If one is going to recommend against breastfeeding, what are the alternatives? It is sometimes automatically assumed that the alternative to breastfeeding is using commercial infant formula. However, the options are not simply breast versus formula. There are in fact many options that need to be plainly identified, and their merits and demerits in different circumstances need to be systematically assessed.

Breastmilk can be provided in many different ways, and many of these variations can make a difference in the context of HIV/AIDS. Exclusive breastfeeding is different from breastfeeding combined with other liquids or solids. Breastmilk can be delivered directly from the source, or indirectly. Wet nurses, relatives, or friends can provide direct breastfeeding. Or the mother’s breastmilk can be provided indirectly by being expressed, heat treated to inactivate the virus, and then supplied to the infant with a cup.

Several alternatives to breastmilk have been used, including not only commercial products but also home-made formulas based on fresh or processed animal milks, usually diluted with water and with added sugar and micronutrient supplementation. However, some experts feel that animal milks are never suitable, even with dilution, supplementation, or other manipulation of the product.

The use of commercial formula may itself be managed in a variety of ways. For example, some proposals call upon national governments to pay for the formula and provide it free to HIV-positive mothers. Some hope there will be international subsidies. Some proposals call for using generic labels on formula containers to minimize the promotion of particular brands. It is generally agreed that the use of commercial formula should be in conformity with the International Code of Marketing of Breastmilk Substitutes and subsequent clarifying resolutions of the World Health Assembly.

Considerable effort should be devoted to identifying and creating alternatives, and to designing variations on them. All plausible options should be fairly assessed. For example, while banking of breastmilk may have been deemed impractical in the past, in the context of HIV/AIDS there should be renewed interest in its potential. Even commercial milk banking, with appropriate safeguards, might be feasible (Rao 1977).

(g) Protective Effects of Breastfeeding. Outside the HIV/AIDS context, there is clear and consistent documentation showing that, in every kind of circumstance, compared to the "gold standard" of exclusive breastfeeding, all other methods of feeding yield inferior results in terms of the health of the infant. Even in cases in which breastmilk is tainted by pathogens such as viruses or heavy metals, the infant is almost always better off being breastfed (Lawrence 1998).

Does this pattern hold when mothers are diagnosed as HIV-positive, or does that constitute an exception to the prevailing pattern? The question has not been studied adequately. However, there is fragmentary evidence suggesting that the dominant pattern holds even in these cases.

Breastfeeding is a form of immunization against a broad variety of diseases, including diseases for which the pathogens are transmitted via breastmilk (Goldman 1998). Since breastfeeding transmits immunological properties from the mother to the infant, it can be especially beneficial to breastfeed infants believed to be at risk of immune deficiency. All infants of HIV-positive mothers are exposed to that risk, the risk of getting AIDS. Researchers at the University of California at San Diego are looking into the potential of lactoferrin, a substance that is found naturally in human milk, for inhibiting mother-to-child transmission of HIV. The researchers point out, "lactoferrin can be easily produced and could be orally administered to newborns in formula (Reuters 1998; also see Harmsen 1995 and Huisman 1996)". It is not yet clear whether the health benefits resulting from the immunological properties of breastmilk would outweigh the health risks of virus transmission through breastmilk.

Breastfeeding increases the likelihood of transmission of HIV, but at the same time it also provides important protection against the diseases that can result from HIV. That is, apart from any effect on transmission of the virus, breastmilk may inhibit the virus’s activity:

Some breastmilk factors may be specifically protective against postnatal transmission of HIV. Anti-HIV IgG and IgA antibodies have been identified in colostrums from HIV-positive women, but not from HIV-negative women (Duprat et al 1994). Van de Perre et all (1993 suggest that HIV-1 IgM in breastmilk could be protective against postnatal transmission of the virus... . Human milk also contains a glycoprotein, gp 120, which is able to inhibit the binding of HIV to CD4, which may block the first step that is critical for infection of a target cell (Morrison 1999).

It has been reported that the human milk factors active against HIV include secretory IgA, IgG, IgM, unsaturated fatty acids, ribonuclease, chondroitin, sulphate, sulphatide, secretory leukocyte protease inhibitor, lactoferrin, and lysozyme. Remarkably, most of these factors are not destroyed by pasteurization (May 2000). Thus, since pasteurization is known to destroy HIV, probably the best food for infants of mothers diagnosed as HIV-positive would be pasteurized breastmilk.

For infants that are not infected at birth, the risk of HIV infection that might come from breastfeeding must be weighed against the risks of depriving the infant of the advantages breastfeeding, and at the same time exposing the infant to the risks of alternative feeding methods. For those infants that are already infected, some may feel that since they are already infected, there is no reason to deprive them of the protective and other benefits of breastfeeding. However, it is conceivable that there is a dose effect in virus transmission, such that the delivery of additional copies of the virus through breastfeeding might accelerate the onset of disease. Many people have strong opinions about how these considerations balance out, but in the final analysis these are empirical questions, questions that may have to be answered differently in different settings. The researchers have not given these questions the attention they deserve.

(h) Quality of Analysis and Recommendations. There is a tendency to jump to conclusions based on the untested assumption that breastfeeding by HIV-positive women is bad for their infants. To illustrate, the ACTG 076 research trials showed that a "short-course" ZDV treatment could substantially reduce the rate of mother-to-child transmission of HIV. (ZDV is zidovudine, the generic name for azidothymidine, or AZT.) The research trials were done with HIV-positive women who were counseled to not breastfeed, and instead exclusively bottle-fed their infants with formula. Subsequently, it has become common practice to recommend that women who are given this treatment use formula. A key report says that together with the treatment, "breast-milk substitutes (commercial infant formula or other products for home preparation) must be organized (Recommendations 1998, p. 316)." However, research results on the use of the short-course treatment with breastfeeding women were not yet reported at the time these recommendations were made. The fact that the women in the research trials did not breastfeed does not establish that it is wiser to not breastfeed. There was no evident basis for recommending that women taking the short-course treatment should not breastfeed. Apparently it was grounded in the unchallengeable assumption that the possibility of transmission of any virus through breastfeeding will be eliminated if there is no breastfeeding.

This study argued that, "Thus for maximum reduction of MTCT [mother-to-child transmission], alternatives to breastfeeding should be considered and their acceptability and safe use examined in developing countries where breastfeeding is the norm (Recommendations 1998, p. 317)."  Moreover, "Women must be informed that breastfeeding may reduce the effectiveness of treatment with ZDV ...." This illustrates the preoccupation with minimizing the likelihood of virus transmission, when the issue of concern should be the well-being of the infant.

Consider, as another example, the much-cited Nduati study comparing the impacts of alternative feeding methods (Nduati 2000). The section on Data Analysis simply asserts that death and HIV-1 infection are highly correlated. Where are the scientific data that demonstrate that association? In what context are those data relevant? If we make no assumptions about the association between HIV diagnosis in infants and their health outcomes, the study tells us very little. Its most important actual finding may be that the 2-year mortality rates were similar in the breastfeeding group and the formula feeding group. The study did not demonstrate any statistically significant difference in health outcomes. Nevertheless, the fact that formula-fed infants showed higher mortality during the first six weeks of life (1.0 % breastfed vs. 3.9 % formula-fed) is a cause for concern.

HIV testing of mothers and infants is frequently done not through detection of the virus itself but through the detection of antibodies that are presumed to be associated with HIV. While some take the view that the presence of these antibodies is a cause for alarm, it should perhaps be viewed as just the opposite, a highly desirable finding.

There are remarkable inconsistencies in the discussions. Why is there so much concern for HIV transmission via breastfeeding in developing countries when, in a thorough study of mother-to-child transmission in the United States, the issue was passed over lightly, and the discussion of strategies for preventing transmission of the virus did not even mention feeding options (Stoto 1998)?

Some of the global agencies that discuss the question of feeding strategies by HIV-positive mothers hedge their advice. Instead of providing clear instructions, they say mothers "might want to consider" using formula rather than breastfeeding, and they qualify their positions with numerous cautionary remarks. Despite the agencies' qualifications, their persistent expression of alarm over the risk of virus transmission tends to lead health workers and mothers to only one conclusion: HIV-positive mothers should not breastfeed. To illustrate:

• In a study in Khayelitsha, South Africa, HIV-infected mothers were not given a choice regarding breastfeeding. All mothers were told not to breastfeed and all complied with this advice (Lala 2000).

• In a report out of the AIDS conference in Durban in July 2000, a physician from Philadelphia was quoted as saying, "I came here thinking I knew at least one thing for sure, that HIV-positive women should not breastfeed their babies" The reporter herself thought that "it had been roundly agreed in the past decade that women with HIV, the virus that causes AIDS, should definitely not nurse their babies with mother’s milk (DiManno 2000)."

• In Alberta, Canada, pamphlets tell mothers who are diagnosed as HIV-positive, "Do not breastfeed", with no reservations of any kind.

Surely, if the global agencies interviewed health workers and mothers, they would find that their careful cautions and qualifications have been lost by the time they reach the ground. If their advice is not being heard the way they intend it, perhaps the agencies should change their message.

Instead of hedging their advice, or giving overly definitive instructions, the agencies should be candid, describing the information that is available, and presenting the varying views of specialists. They also should do what they can to make sure that the research that is needed to resolve the ambiguities is carried out.

Mothers are urged to make informed choices, but they are not provided with the means required to do that. The agencies should not evade responsibility for providing clear guidance by saying the choice must finally be made by the mother herself. They should do more to assure that mothers are provided the information they need to make informed decisions.

Nutrition Rights

The articulation of food and nutrition rights in modern international human rights law begins with the Universal Declaration of Human Rights (UDHR 1948). The declaration asserts in article 25(1) that "everyone has the right to a standard of living adequate for the health and well-being of himself and his family, including food... ."

The right was reaffirmed in two major binding international agreements. In the International Covenant on Economic, Social and Cultural Rights, which came into force in 1976, article 11 says that "The States Parties to the present Covenant recognize the right of everyone to an adequate standard of living for himself and his family, including adequate food, clothing, and housing... " and also recognizes "the fundamental right of everyone to be free from hunger... (ICESCR 1976)."

In the Convention on the Rights of the Child, which came into force in 1990, two articles address the issue of nutrition (CRC 1990). Article 24 says that "States Parties recognize the right of the child to the enjoyment of the highest attainable standard of health... (paragraph 1)" and shall take appropriate measures "to combat disease and malnutrition... . through the provision of adequate nutritious foods, clean drinking water, and health care (paragraph 2c)." Article 24 says that States Parties shall take appropriate measures... "To ensure that all segments of society, in particular parents and children, are informed, have access to education and are supported in the use of basic knowledge of child health and nutrition [and] the advantages of breastfeeding... ." Article 27 says in paragraph 3 that States Parties "shall in case of need provide material assistance and support programmes, particularly with regard to nutrition, clothing, and housing."

Thus, the human right to food and nutrition is well established in international law. Even if the right had not been stated directly, it would be strongly implied in other provisions such as those asserting the right to life and health, or the Convention on the Rights of the Child’s requirement (in article 24, paragraph 2a) that States Parties shall "take appropriate measures to diminish infant and child mortality". The human right to food and nutrition has been reaffirmed at the international level in many different settings.

The foundations for the international human right to food and nutrition lie in the binding international human rights instruments in which they are explicitly mentioned, primarily the International Covenant on Economic, Social and Cultural Rights and the Convention on the Rights of the Child. Other binding international human rights agreements such as the Convention on the Elimination of All Forms of Discrimination Against Women contribute to the articulation of relevant rights (CEDAW 1981).

Several non-binding international declarations and resolutions also help to shape the emerging international consensus on the meaning of the human right to food and nutrition. The major initiatives include the following:

• In response to concerns about inappropriate marketing and promotion, the International Code of Marketing of Breastmilk Substitutes was adopted by the World Health Assembly in 1981 (WHO 1997). The WHA has approved a series of resolutions in subsequent years to further clarify and strengthen the code.

• On August 1, 1990 the Innocenti Declaration on the Protection, Promotion and Support of Breastfeeding was adopted by participants at a meeting on Breastfeeding in the 1990s held at the International Child Development Centre in Florence, Italy. The declaration stated a variety of specific global goals, including the goal that "all women should be enabled to practice exclusive breastfeeding and all infants should be fed exclusively on breast-milk from birth to 4-6 months of age (Innocenti 1990)". In 1991 the UNICEF Executive Board passed a resolution (1991/22) saying that the Innocenti Declaration would serve as the "basis for UNICEF policies and actions in support of infant and young child feeding". In May 1996 the World Health Assembly passed a resolution on Infant and Young Child Nutrition (WHA49.15) in which it confirmed its support for the Innocenti Declaration.

• The World Summit for Children held in 1990 called for "Empowerment of all women to breast-feed their children exclusively for four to six months and to continue breastfeeding, with complementary food, well into the second year."

• In 1992 the World Declaration and Plan of Action for Nutrition, agreed upon at the conclusion of the International Conference on Nutrition in Rome, pledged "to reduce substantially within this decade... social and other impediments to optimal breastfeeding". The Plan of Action asserted, in article 30, that "Breastfeeding is the most secure means of assuring the food security of infants and should be promoted and protected through appropriate policies and programmes." Article 33 stated that "Governments, in cooperation with all concerned parties, should... prevent food-borne and water-borne diseases and other infections in infants and young children by encouraging and enabling women to breast-feed exclusively during the first four to six months of their children’s lives." Article 34 provided a detailed call for action on promoting breastfeeding.

• In 1995 the Platform for Action that came out of the Fourth World Conference on Women in Beijing called for promoting public information on the benefits of breastfeeding, implementing the International Code of Marketing of Breastmilk Substitutes, and facilitating breastfeeding by working women.

• In November 1996 the World Food Summit concluded with agreement on the Rome Declaration on World Food Security and World Food Summit Plan of Action. The first paragraph declared: "We, the Heads of State and Government, or our representatives, gathered at the World Food Summit at the invitation of the Food and Agriculture Organization of the United Nations, reaffirm the right of everyone to have access to safe and nutritious food, consistent with the right to adequate food and the fundamental right of everyone to be free from hunger."

On May 12, 1999 the UN's Committee on Economic, Social and Cultural Rights released its General Comment 12 (Twentieth session, 1999): The Right to Adequate Food (Art. 11) (General Comment 12 1999). This statement by the committee constitutes a definitive contribution to international jurisprudence.

There is increasing recognition at the international level that good nutritional status is an outcome that depends not only on good food but also on good health services and good care (Engle 1997; Longhurst 1995). Health services include a broad range of measures for the prevention and control of disease, including the maintenance of a healthy environment. Thus, infant feeding is not simply a matter of the physical transmission of nutrients. There should be a strong component of caring in it, through the closeness and contact that can be provided during feeding. Breastfeeding can be regarded as a kind of health service because of the fact that it immunizes the infant against a broad variety of diseases.

Because of their immediate and direct dependence on their mothers, the nutrition status of infants is determined not only by the quality of the food, health services, and care they receive directly, but also by the food, health service, and care received by the mother herself. The infant’s nutrition status at birth depends on the quality of the mother’s health status and prenatal care, and whether she has had a good diet in general and has been protected from iron deficiency anemia in particular.

Women have their own rights. Mothers should be entitled to particular services not only because of their own rights but also because of their obligations to provide for their children. Mothers should receive good pre-pregnancy and prenatal care, and parents should be well informed about the risks and benefits of all alternative means for feeding their infants because, like everyone else, their infants have a human right to adequate nutrition.

Women’s Rights to Breastfeed vs. Infants’ Rights to Be Breastfed

What is the relationship between the mother’s interest in breastfeeding and the infant’s interest in being breastfed? How do the mother’s rights relate to the infant’s rights?

Many different parties--including the infant, the parents, siblings, the extended family, the community, health professionals, employers, infant formula manufacturers and sellers, local governments, national governments, international agencies, and others--affect infant care and feeding. Each party has its own interests and its own capacities to press for outcomes preferable to itself. At times infants are not nurtured properly because of the pull of others’ interests. They are all concerned, more or less, with the infant’s health, but they also have other interests such as profits, increased leisure time, and having opportunities to do other things. Where these parties do not all have preferred outcomes that are consistent with one another, there is conflict among them.

At times the mother and the infant may have conflicting interests. The conflict is raised in clear relief when it is argued that the infant has a right not only to be well nourished but, more specifically, that the infant has a right to be breastfed. Such a right could clash with the woman’s right to choose how to feed her infant.

Article 3 of the Convention on the Rights of the Child says, "In all actions concerning children... the best interests of the child shall be a primary consideration". Combining this with the observation that breastfeeding is better than alternative methods of feeding, some breastfeeding advocates argue that infants have a right to be breastfed. However, this appears to be a minority view.

While it is true that actions must be based on consideration of the best interests of the child, that is not the only consideration. Moreover, it is assumed that normally the parents judge what is in the child’s best interests. The state should interfere in the parent-child relationship only in extraordinary situations, when there is extremely compelling evidence that the parents are acting contrary to the best interests of the child.

The infant has great interests at stake, but few resources to be used to press for preferred outcomes. Given the infant’s powerlessness, it is sensible to use the law to help assure that the best interests of the infant are served. However, while it is surely appropriate to use the law to protect the infant from outsiders with conflicting interests, it is not reasonable to use the law to compel an unwilling mother to breastfeed, or to prevent a willing mother from breastfeeding. Thus, for the purposes of framing appropriate law, the woman and infant can be viewed as generally having a shared interest in the infant’s well being. From the human rights perspective, the major concern is with protecting the woman-infant unit from outside interference.

Mothers should remain free to feed their infants as they wish, in consultation with other family members. Outsiders are obligated to refrain from doing anything that might interfere with a mother’s freely made, informed decision. Mothers should have appropriate and accurate information available to them so that they can make informed decisions. This is the approach taken in the International Code of Marketing of Breastmilk Substitutes. The code is not designed to prevent the marketing or use of formula, but to assure that parents can make a fully and fairly informed choice on how to feed their infants.

Rather than have the state make decisions for them, citizens in a democracy prefer assurances that nothing impedes them from making their own decisions. To the extent possible we should be free to choose, and that includes being free to some extent to make what others might regard as unwise or sub-optimal decisions.

Principles

After long hard discussion in conferences and in sustained email discussions over the Internet, a number of specialists concerned with the issue formulated the following Consensus Statement Regarding the Nutrition Rights of Infants on the basis of their understanding of international human rights law and principles.

1. Infants have a right to be free from hunger, and to enjoy the highest attainable standard of health.

2. Infants have a right to adequate food, health services, and care.

3. The state and others are obligated to respect, protect, and facilitate the nurturing relationship between mother and child.

4. Women have the right to social, economic, health, and other conditions that are favorable for them to breastfeed or to deliver breastmilk to their infants in other ways. This means that women have the right to:

   1. Good prenatal care.

   2. Basic information on child health and nutrition and the advantages of breastfeeding, and on principles of good breastfeeding and alternative ways of providing breastmilk.

   3. Protection from misinformation on infant feeding.

   4. Support in the practice of breastfeeding.

   5. Maternity legislation to protect and enhance employed women’s opportunities for nurturing their infants.

   6. Baby-friendly health facilities.

5. Women and infants have a right to protection from factors that can hinder or constrain breastfeeding, in accordance with:

   1. The Convention on the Rights of the Child,

   2. The International Code of Marketing of Breastmilk Substitutes and related World Health Assembly resolutions,

   3. The International Labor Organization’s Maternity Protection Convention Number 103 and its subsequent revisions, and

   4. The Innocenti Declaration on the Protection, Promotion and Support of Breastfeeding.

6. States, represented by their governments, have an obligation to:

   1. Protect, maintain, and promote breastfeeding through public educational activities,

   2. Facilitate the conditions of breastfeeding, and

   3. Otherwise assure that infants have safe access to breastmilk.

7. No woman should be prevented from breastfeeding.

A presentation was made on this consultation process at the meeting of the UN’s Sub-Committee on Nutrition in Washington, D.C. in April 2000.

Recommendations

In my view, these principles regarding the human rights of infants in relation to nutrition apply with equal effect in the context of HIV/AIDS; they are not to be suspended. This means, for example, that even HIV-positive mothers have a right to breastfeed. If any government were to prohibit HIV-positive mothers from breastfeeding, that would violate their human rights, and also violate their infants’ human rights.

In particular, the idea that parents should be able to make informed decisions remains valid in the context of HIV/AIDS. Of course, its application depends on mothers being aware of and having real access to a range of feeding alternatives, and it depends on their having good information about these available alternatives. Where commercial interests are represented, the presentation of options and the information about them are likely to be sharply skewed.

For the purposes of the discussion here, particular attention should be given to Principle 4, which focuses on the obligation to assure that the infants’ parents are well informed with regard to their infant feeding choices. This is the major idea underlying the International Code of Marketing of Breastmilk Substitutes. The code does not prohibit marketing or use of formula, but insists that promotion activities for the products must be conducted in ways that are fair rather than being skewed to favor commercial products. Article 24, paragraph 2e of the Convention on the Rights of the Child goes directly to the point. It calls upon States Parties "To ensure that all segments of society, in particular parents and children, are informed, have access to education and are supported in the use of basic knowledge of child health and nutrition, the advantages of breast-feeding, hygiene and environmental sanitation and the prevention of accidents." This is a legally binding obligation on all States Parties to the convention (all countries except the United States and Somalia), and a strong moral obligation on those that are not. From the debate relating to HIV, it is now increasingly clear that the full array of feeding options should be presented to the parents, and better research is needed about the advantages and disadvantages of each option in particular local circumstances.

In addition, there is a need to enable the mother to learn whether she is HIV-positive through confidential, voluntary counseling and testing so that she can make an informed decision regarding the feeding of her infant in relation to her own condition. This counseling should include factual information on the limitations, validity, and meaning of the test. Of course, if the advice to be given to women diagnosed as HIV-positive eventually turns out to be the same as that given to women diagnosed as HIV-negative, or to those who are undiagnosed, then this testing would not be needed to help in making the feeding decision.

Prudent Judgment

Parents need to make decisions and health workers need to give guidance even in the face of uncertainties. As shown above, there are many important questions still unanswered regarding infant feeding choices in the context of HIV/AIDS. What should be done while we wait for these questions to be resolved?

In addressing this dilemma, it is useful to review our current state of knowledge regarding infant feeding. There is one dominant finding. Outside the HIV context, in almost every kind of situation in which infants are able to breastfeed, they are better off breastfeeding than not breastfeeding. There are some exceptions, but they are relatively rare. For example, the use of breastmilk substitutes may be medically indicated when the mother or the child is unable to sustain breastfeeding, or the mother is taking pharmaceutical or recreational drugs that may be dangerous for the infant.

There may be occasions when the negative impact of not breastfeeding on the infant is judged to be small, and the apparent added convenience of not breastfeeding seems important to the mother, at least in the short term, but the use of formula hardly ever works to the infant’s advantage. Outside the HIV/AIDS context, there are few conditions under which infants are better off with infant formula.

The same conclusion applies to the HIV/AIDS context. There is as yet no clear published scientific evidence that infants of mothers who are diagnosed as HIV-positive would be better off if they were not breastfed. There has been a great deal of inference and surmise suggesting this conclusion, but until now there is no hard evidence to support it. The claim that HIV can be transmitted via breastmilk in itself should not determine the feeding decision. As indicated earlier, many other factors intervene to determine the consequences for the infant. Blood tests alone are not sufficient indicators of health outcomes.

There are very few conditions under which breastmilk substitutes are better for the infant than exclusive breastfeeding. Thus, even in the context of HIV/AIDS, so long as there is doubt and uncertainty, prudence suggests that even HIV-positive mothers should exclusively breastfeed their infants. New policies can be considered when solid new evidence has been obtained.

If this line of reasoning is not convincing, and one wishes to minimize the risk as much as possible (while waiting for more thorough research findings), it is clear that the least-risk options are the use of expressed and warmed breastmilk from the mother, or the use of breastmilk from other women obtained from carefully selected wet nurses or milk banks.

The Research Agenda

The concern of this study is captured in the title of a report from the conference on HIV/AIDS held in Durban, South Africa, in July 2000: The Wait for Unequivocal Advice on Breastfeeding Continues (Howse 2000). There is no lack of debate on the question of how infants of HIV-positive mothers ought to be fed. The debate goes on largely because the research that is needed to resolve the question has not been carried out.

We have an abundance of strong opinions and a meager supply of clear evidence. So far, the professional researchers have focused on the likelihood of transmission of the virus through breastfeeding. They have not posed their questions and formulated their findings in a way that is designed to be useful to the most important decision-makers, the mothers who need guidance on how to feed their infants. Mothers are not concerned with esoteric issues of scientific explanation, but with practical policy decisions: will my infant be healthier if I feed her by this method or by that method?

To address this question in a direct way, it might be useful if clinics and hospitals systematically reviewed their records to see what patterns they reveal. Data on the actual experience of infants of mothers diagnosed as HIV-positive in different villages and towns could be collected in a systematic way. What are their health outcomes? How are those outcomes influenced by the ways in which they have been fed? If the data were presented in a way that would make it easy to see patterns, many people would be able to interpret them. In some cases they might be presented in graphic form to show the relationships more clearly.

Consider the hypothetical data table in Figure 1. It could be used to organize data obtained from a community’s clinic or hospital. The column on the left shows the number of infants that were born in a particular year of mothers that had been diagnosed as HIV-positive. For example, for the purposes of this illustration the table says that there were 22 such infants born at this hospital (or in this town, or at this clinic) in 1995. Then, in the second column, those infants are sorted out according to the methods by which they have been fed. The hypothetical chart shows that of those born in 1995, five were exclusively breastfed, seven were fed with formula, and ten had mixed feeding.

Then we want to know what proportion of each of these groups was healthy at the end of each of these years. In the illustration, three of the five of the five exclusively breastfed infants were judged to be healthy at the end of 1995. This is expressed in the table as the fraction, 3/5. Similar records would be made to show the proportion healthy for the groups of infants fed by other means.

Data of this kind could be recorded at the end of each subsequent year, perhaps until the infant reached five years of age. The table could be endlessly expanded, with columns and rows added as time went on.

The data for different clinics and hospitals in a community could be consolidated into one large table. When records are obtained for a large number of infants, the proportions could be expressed as percentages. This would make it easier to see patterns and to make comparisons.

If data like these were collected over many years for many different hospitals, clinics, villages, etc., distinct patterns would be likely to emerge. The trends over time could be plotted on graphs to make the patterns more visible. If clear patterns emerged, people would be able to see them, and they would be able to make their own judgments as to which feeding method is most likely to lead to better health outcomes for infants.

Pilot studies could develop guidelines for collecting the required data. The point here is not to design such studies, but simply to suggest an approach to generating the knowledge that mothers need to make their feeding decisions.

If diagnoses of the mother’s HIV status are difficult to make for any reason, data of the form suggested in Table 1 could be collected for all infants. If data can be obtained for infants of mothers diagnosed HIV-negative as well as for those diagnosed as HIV-positive, separate tables could be constructed for each of these groups, and then compared to assess the degree to which the patterns differ.

Scientists may object that this approach, based on the collection of real-world data in natural settings, is subject to many different influences, and does not have the potential for definitive interpretation in the way that is sometimes possible with controlled, randomized studies of the sort scientists are trained to design. There is truth to that, but at the same time it must be acknowledged that formal scientific studies have defects and disadvantages of their own. For example, they often create quite unnatural settings, and force people into doing things they would not otherwise do. They are also expensive and time consuming, and they are often difficult to interpret, particularly because they are often based on small samples. Moreover, formal scientific research often centers on questions that are not of great concern to local people.

Many treatment recommendations are now based only on limited clinical trials, without corroborating evidence from the field. Even if a treatment appears to be beneficial in clinical trials, it still must be shown to be effective under actual field conditions.

It is true that "the wait for unequivocal advice on breastfeeding continues". However, this lament seems to suggest that someone else, perhaps someone with more authority, should prepare and deliver that advice. Another approach is to operate on the principle that with appropriate support and information, people can make their own wise judgments. Instead of having outsiders tell local people what to do, the outsiders could instead ask, what would be needed to allow local people to make their decisions for themselves?

Conventional scientific research sometimes conveys an implicit message that people cannot assemble knowledge and make decisions for themselves. It can result in removing decision-making authority away from the local setting in which it should be made. Under such conditions, conventional science can be disempowering to local people. Current research on HIV/AIDS often takes this form. It is based on central elites taking data from the periphery for interpretation and decision-making in capital cities, with the results being shipped back out to the periphery, often in the form of instructions.

People in the periphery sometimes are asked to make their own judgments, but they are not provided with the bases for making those judgments. Such requests are disempowering.

The argument here is that there is a rich lode of valuable data in natural settings that is waiting to be tapped, by as well as for people in the community. Coaching people to assemble their own information locally, and to put it into forms that they themselves can interpret, is empowering for them. Even if some of these modes of analysis cannot practically be brought to the grassroots or consumer level, they certainly would be within the grasp of local professionals, such as nutritionists and public health workers, in developing countries.

Data of the sort outlined here can be presented to mothers in ways that are meaningful locally, and they can be asked to participate in the interpretation of those data.

In some ways, information is like food. In a normal, well-functioning society, we do not expect government to feed us. Instead, we expect government to provide enabling conditions, facilitating us so that we can provide for ourselves. Governments should provide food directly only in extraordinary circumstances. Similarly, insofar as possible, we should develop our information, interpretations, and decisions for ourselves, with government supporting us in these efforts. Governments’ providing food or answers for us can be very disempowering. Providing for ourselves is empowering.

Means need to be developed so that local people can participate meaningfully in making decisions that affect their lives. People have a right to the information they need to make wise judgments about the feeding of their children. Supporting people in developing and interpreting information locally to help make informed judgments about infant feeding would be a major step in advancing the realization of these rights.

Table 1: Health status at year end for infants of mothers diagnosed as hiv-positive, by feeding method (proportion healthy at year end)

Bibliography

(Adamson undated). Adamson, Peter. Facts for Life (New York: UNICEF, WHO, and UNESCO, undated).

(AIDS Law 1997). AIDS Law Project, Centre for Applied Legal Studies. Pregnancy and HIV: Recommended Code of Best Practice: Draft (Johannesburg: University of the Witwatersrand, 1997).

(Altman 1998). Altman, Lawrence K. "AIDS Brings Shift in U.N. Messages on Breast-Feeding," New York Times, July 26, 1998, p. 1.

(Barrington-Ward 1997). Barrington-Ward, Simon "Putting Babies Before Business" Progress of Nations 1997 (New York: United Nations Children’s Fund, 1997, pp. 15-19.

(Bellamy 1998). Bellamy, Carol. The State of the World’s Children 1998 (New York: Oxford University Press, 1998), pp.19-21.

(Burr 1997). Burr, C. K.; Taylor, S.; Bartelli, D., Reduction of Perinatal HIV Transmission: A Guide for Providers (Newark, New Jersey: National Pediatric and Family HIV Resource Center, 1997).

(CDC 1999). United States. Centers for Disease Control. HIV/AIDS Surveillance Report: U.S. HIV and AIDS Cases Reported Through December 1999, Year End Editions, Vol. 11, No. 2 (Washington, D.C.: CDC, 1999).

http://www.cdc.gov/hiv/stats/hasr1102.pdf

(CEDAW 1981). Convention on the Elimination of All Forms of Discrimination against Women. www.unhchr.ch/html/menu3/b/e1cedaw.htm

(Coutsoudis 1999). Coutsoudis, Anna; Pillay, Kubendran; Spooner, Elizabeth; Kuhn, Louise; Coovadia, Hoosen M.; "Influence of Infant-Feeding Patterns on Early Mother-to-Child Transmission of HIV-1 in Durban, South Africa: A Prospective Cohort Study", The Lancet, Vol. 354, No. 9177 (7 August 1999).

(CPA 1995). Committee on Pediatric AIDS, American Academy of Pediatrics, "Human Milk, Breastfeeding, and Transmission of Human Immunodeficiency Virus in the United States (RE9542)," Pediatrics, Vol. 96, No. 5 (November 1995), pp. 977-979.

www.aap.org/policy/01094.html

(CPA 1997). Committee on Pediatric Aids, American Academy of Pediatrics, "Evaluation and Medical Treatment of the HIV-Exposed Infant (RE9721)," Pediatrics, Vol. 99, No. 6 (June 1997), pp. 909-917. www.aap.org/policy/970602.html

(Cracking 1997). Cracking the Code: Monitoring the International Code of Marketing of Breast-milk Substitutes (London: Interagency Group on Breastfeeding Monitoring, 1997).

(CRC 1990). Convention on the Rights of the Child.

http://www.unhchr.ch/html/menu3/b/k2crc.htm

(CRC 1998). Committee on the Rights of the Child. "Children Living in a World with HIV/AIDS," General Discussion, Nineteenth Session, 21 September – 9 October 1998.

www.unhchr.ch/html/menu2/6/crcaids.htm

(DiManno 2000). DiManno, Rosie, "Doctor Drops Breastfeeding Bombshell: Push to Allow HIV-Infected Mothers to Nurse Sparks Outcry," Toronto Star, July 12, 2000.

(Dunn 1992). Dunn, D. T.; Newell, M. L.; Ades, A. E.; and Peckham, C. S., "Risk of Human Immunodeficiency Virus Type 1 Transmission through Breastfeeding," The Lancet, Vol. 340 (Sept. 5, 1992), pp. 585-588.

(Duprat 1994). Duprat, C.; Mohammed, Z.; Datta, P., et al., ""Human Immunodeficiency Virus Type 1 IgA Antibody in Breast Milk and Serum," Pediatric Infectious Diseases, Vol. 13, No. 7 (1994), pp. 603-608.

(Engle 1997). Engle, Patrice L.; Menon, Purnima; and Haddad. Lawrence. Care and Nutrition: Concepts and Measurement (Washington, D.C.: International Food Policy Research Institute, 1997).

(Fawzi 2000). Fawzi, W. W.; Msamanga, G.; Hunter, D.; Urassa, E.; Renjifo, B.; Mwakagile, D.; Hetrzmark, E.; Coley, J.; Garland, M.; Kapiga, S.; Antelman, G.; Essex, M.; Spiegelman, D., "Randomized Trial of Vitamin Supplements in Relation to Vertical Transmission of HIV-1 in Tanzania," Acquired Immune Deficiency Syndrome Vol. 23, No. 3 (March 2000), pp. 246-254.

(Fine 1997). Fine, Derrick; Heywood, Mark; and Strode, Ann, eds. HIV/AIDS and the Law: A Resource Manual (Johannesburg, South Africa: AIDS Law Project and Lawyers for Human Rights, University of the Witwatersrand, 1997).

(Food Policy 1996). Food Policy, Vol. 21, No. 1 (March 1996). Special issue on Nutrition and Human Rights, edited by Wenche Barth Eide, Uwe Kracht, and Robert E. Robertson.

(Friis 1998). Friis, H. and Michaelsen, K. F. "Micronutrients and HIV Infection: A Review," European Journal of Clinical Nutrition, Vol. 52 (1998), pp. 157-163.

(General Comment 12 1999) United Nations. Economic and Social Council. Committee on Economic, Social and Cultural Rights. Substantive Issues Arising in the Implementation of the International Covenant on Economic, Social and Cultural Rights: General Comment 12 (Twentieth Session, 1999) The Right to Adequate Food (art. 11) (Geneva: ECOSOC E/C.12/1999/5). http://www.unhchr.ch/tbs/doc.nsf/MasterFrameView/3d02758c707031d58025677f003b73b9?Opendocument

(Goldman 1998). Goldman, Armond S.; Chheda, Sadhana; and Garofalo, Roberto, "Evolution of Immunologic Functions of the Mammary Gland and the Postnatal Development of Immunity," Pediatric Research, Vol. 43, No. 2 (February 1998).

http://lww.com/PDR/0031-39982-98p155.html

(Harmsen 1995). Harmsen, M. C.; Swart, P. J.; de Bethune, M. P.; Pauwels, R.’ De Clercq, E.; The, T. H.; Mijer, D. K., "Antiviral Effects of Plasma and Milk Proteins: Lactoferrin Shows Potent Activity Against Both Human Immunodeficiency Virus and Human Cytomegalovirus Replication in Vitro," Journal of Infectious Diseases, Vol. 172, No. 2 (August 1995), pp. 380-388.

(Holmes 1992). "Breastfeeding and HIV" (Letter), Lancet, Vol. 340 (1992), pp. 1094-1095.

(Howse 2000). The Wait for Unequivocal Advice on Breastfeeding Continues. Report from HIVAIDS Conference in Durban, South Africa, July 12, 2000.

http://64.28.82.59/projects/aidsonline/aidshomepage.nsf/.LookupIndexV/7754929C5CB47E5142256919008338FD?opendocument

(Huisman 1996). Huisman, H.; Kuipers, M. E.; Smit, C.; De Clercq, E.; Meijer, D. K.; and Swart, P. J., "Antiviral Effects of Milk Proteins: Acylation Results in Polyanionic Proteins with Potent Activity Against HIV Type I and II in Vitro," International Conference on AIDS, 1996, July 7-12; 11 (1):66 (abstract no. Mo.A.1070, NLM Citation ID 96921140).

(ICESCR 1976). International Covenant on Economic, Social and Cultural Rights.

http://www.unhchr.ch/html/menu3/b/a_cescr.htm

(IJCR 1997). International Journal of Children’s Rights, Vol. 5, No. 4 (1997). Special issue on Food and Nutrition Rights, edited by George Kent. See, in particular, articles by Michael Latham on "Breastfeeding—A Human Rights Issue?", Leah Margulies on "The International Code of Marketing of Breastmilk Substitutes: A Model for Assuring Children’s Rights Under the Law", and Allison Linnecar, "The International Baby Food Action Network: Defending Every Child’s Birthright".

(Innocenti 1990). Innocenti Declaration on the Protection, Promotion and Support of Breastfeeding. http://www.elogica.com.br/waba/inno.htm

(Innocenti Progress 1998). "The Innocenti Declaration: Progress and Achievements," Parts I and II. Weekly Epidemiological Record, Vol. 73, No. 5 (1998), pp. 25-32 and Vol. 73, No. 13 (1998), pp. 91-94.

(International Code 1981). International Code of Marketing of Breast-milk Substitutes.

http://www.gn.apc.org/ibfan/thecode.html

(Kent 1999a). "HIV and Breastfeeding," Mothering, No. 94 (May-June 1999), pp. 65-67, 70-71.

(Kent 1999b). "Tested in Court: The Right to Breastfeed," SCN News, No. 18 (July 1999), pp. 89-90. http://www2.hawaii.edu/~kent/TysonsforSCN.pdf

(Lala 2000). Lala, Sanjay and Sanne, Ian, Vertical Transmission of HIV: Recent Developments (Medscape website on XIII International AIDS Conference, 2000).

http://www.medscape.com/Medscape/CNO/2000/AIDS/Story.cfm?story_id=1470

(Latham 1998). Latham, Michael C. and Greiner, Ted, "Breastfeeding versus Formula Feeding in HIV Infection," The Lancet, Vol. 352, No. 9129 (29 August 1998).

(Lawrence 1998). Lawrence, Ruth A. and Lawrence, Robert M., "Making and Informed Decision About Infant Feeding," in Breastfeeding: A Guide for the Medical Profession, 5^th Edition (Linn, Missouri: Mosby, 1998), pp. 217-219.

(Le Couer 1996). Le Couer, Sophie and Lallemant, Marc. "Breastfeeding and HIV/AIDS," in Jonathan M. Mann and Daniel J. M. Tarantola, eds., AIDS in the World II: Global Dimensions, Social Roots, and Responses (New York: Oxford University Press, 1996), pp. 273-277.

(Linnecar undated). Linnecar, Alison and Yee, Virginia. Protecting Women’s Right to Breastfeed. http://www.elogica.com.br/waba/acsh6.htm

(Longhurst 1995). Longhurst, Richard and Tomkins, Andrew. "The Role of Care in Nutrition—A Neglected Essential Ingredient," SCN News (UN Administrative Committee on Coordination—Subcommittee on Nutrition), No. 12 (Early 1995), pp. 1-5.

(May 2000). May, John, Molecular Virology: Human Milk - Tables of Antimicrobial Factors and Microbial Contaminants Relevant to Human Milk Banking (with continued updating), based on Proceedings of Breast Milk and Special Care Nurseries: Problems and Opportunities Conference (Victoria, Australia: J.T. May and Australian Lactation Consultants Association, Victorian Branch, 1995).

http://www.latrobe.edu.au/www/microbio/milk.html

(Meier 1997). Meier, Barry. "AIDS Renews Battle Over Baby Formula," New York Times, June 8, 1997.

(Morrison 1999). Morrison, Pamela, "HIV and Infant Feeding: To Breastfeed or Not to Breastfeed: The Dilemma of Competing Risks," Parts I and II, Breastfeeding Review, Vol. 7, No. 2 (1999), pp. 5-13.

(Nduati 2000). Nduati, Ruth; John, Grace; Mbori-Ngacha, Dorothy; Richardson, Barbara; Overbaugh, Julie; Mwatha, Anthony; Ndinya-Achola, Jeckoniah; Bwayo, Job; Onyango, Francis E.; Hughes, James; Kreiss, Joan, "Effect of Breastfeeding and Formula Feeding on Transmission of HIV-1: A Randomized Clinical Trial", JAMA (Journal of the American Medical Association), Vol. 283, No. 8 (March 1, 2000), pp. 1167-1174.

(Oshaug 1994). Oshaug, Arne; Eide, Wenche Barth; and Eide, Asbjørn. "Human Rights: A Normative Basis for Food and Nutrition-Relevant Policies," Food Policy, Vol. 19, No. 6 (1994), pp. 491-516.

(Ramanathan 1996). Ramanathan, R.; Amudhamozhi, R.; Sheela, R.; Ravinder, P. T.; and Jayabaskar, R. T., "Prevention of Mother to Child Transmission through Perinatal Care," International Conference on AIDS, July 7-12, 1996 11:2, 158 (Abstract no. We.C.3580).

(Rao 1977). Rao, A. R. "Human Breast Milk as a Commercial Infant Food," Environmental Child Health (December 1977), pp. 286-288.

(Recommendations 1998). "Recommendations on the Safe and Effective Use of Short-Course ZDV for Prevention of Mother-to-Child Transmission of HIV," Weekly Epidemiological Record, Vol. 73, No. 41 (9 October 1998), pp. 313-320.

(Reuters 1998). Reuters Health Information, Lactoferrin May Help Reduce Vertical HIV-1 Transmission, Reuters Health Information Newsline (site produced by the Journal of the American Medical Association), 1998.

http://www.ama-assn.org/special/hiv/newsline/conferen/icaac98/092598g.htm

(Semba 1994). Semba, R. D., et al., "Maternal Vitamin A Deficiency and Mother-to-Child Transmission of HIV-1," Lancet (1994), pp. 1593-1597.

(Semba 1999). Semba R. D. et al., "Micronutrients and the Pathogenesis of HIV Infection," British Journal of Nutrition, Vol. 81 (1999), pp. 181-189.

(Sokol 1997). Sokol, Ellen. The Code Handbook: A Guide to Implementing the International Code of Marketing of Breastmilk Substitutes (Penang, Malaysia: International Baby Food Action Network, International Code Documentation Centre, 1997).

(Special Report 1997). Special Report on Infant Feeding, SidAfrique, No. 11-12 (Sept.-Oct. 1997).

(Specter 1998). Specter, Michael, "In AIDS-Torn Africa, Desperate Choices: UN Suggestion to Stop Breast-Feeding Leaves Mothers Confused and Alarmed," International Herald Tribune, August 20, 1998, pp. 1, 7. (Also in the New York Times on August 19, 1998, p. A1, under the title ""Breast-Feeding and HIV: Weighing Health Risks".)

(Stoto 1998). Stoto, Michael A; Almario, Donna A.; and McCormick, Marie C., eds., Reducing the Odds: Preventing Perinatal Transmission of HIV in the United States (Washington, D.C.: National Academy Press, 1998).

http://www.nap.edu/readingroom/books/rto/

(Susman 2000). Susman, Ed, "AIDS Drug Protects Babies a Year Later," United Press International, 13 July 2000.

http://www.aegis.com/channel/s/UP0000729.htm

(Tozzi 1990). Tozzi, A. E.; Pazzotti, P.; and Greco, D, "Does Breastfeeding Delay Progression to AIDS in HIV-infected Children?" (Letter), AIDS, Vol. 4 (1990), pp. 1493-1494.

(UNAIDS 1996). UNAIDS. Children and Aids in Thailand: One Country’s Response. Stockholm, August 28, 1996.

(UNAIDS 1998a). New Initiatives to Reduce HIV Transmission from Mother-to-Child in Low-Income Countries (Geneva: UNAIDS, 1998).

(UNAIDS 1998b). UNAIDS. HIV and Infant Feeding: A Policy Statement Developed Collaboratively by UNAIDS, WHO and UNICEF (Geneva: UNAIDS, 1997. revised 1998).

(UNICEF 1998). United Nations Children Fund. Facts for Life. (New York: United Nations Children’s Fund, 1998).

(UNECOSOC 1996). United Nations. Economic and Social Council. Second International Consultation on HIV/AIDS and Human Rights (Geneva, 23-25 September 1996).

(UDHR 1948). Universal Declaration of Human Rights.

Van de Perre, P.; Simonon, A.; Hitimana, D.; et al., "Infective and Anti-Infective Properties of Breastmilk from HIV-1 Infected Women," Lancet, Vol. 341 (1993), pp. 914-918.

(WABA 1998). World Alliance on Breastfeeding Action. WABA Position on HIV and Breastfeeding.

http://www.elogica.com.br/waba/dechiv.htm

(Whelan 1998). Whelan, Daniel. "Human Rights Approaches to an Expanded Response to Address Women’s Vulnerability to HIV/AIDS," Health and Human Rights, Vol. 3, No. 1 (1998), pp. 21-36.

(WHO 1997). World Health Organization. The International Code of Marketing of Breast-milk Substitutes: A Common Review and Evaluation Framework (Geneva: Nutrition Unit, Division of Food and Nutrition, WHO—WHO/NUT96.2, 1997).

(WHO 1998). World Health Organization, HIV and Infant Feeding: Guidelines for Decision-makers; HIV and Infant Feeding: A Guide for Health Care Managers and Supervisors; and HIV and Infant Feeding: A Review of HIV Transmission Through Breastfeeding, 1998.

(WHO Collaborative 2000). WHO Collaborative Study Team on the Role of Breastfeeding on the Prevention of Infant Mortality," "Effect of Breastfeeding on Infant and Child Mortality Due to Infectious Diseases in Less Developed Countries: A Pooled Analysis," The Lancet, Vol. 355, No. 9202 (5 February 2000).

(Williams 1995). Williams, Rebecca D., "Breast-Feeding Best Bet for Babies," FDA Consumer Magazine (October 1995). www.FDA.gov/fdac/features/895_brstfeed.html