HIV and home made formula
Home made infant formulae and HIV Michael H.N. Golden 24.06.97
home made formulas and wet nurses André Briend 24.06.97
Home made formulae and HIV Michael H.N. Golden 24.06.97
Re: hiv et al David Alnwick 02.07.97
Letters to Dr. Alnwich about HIV/breastfeeding/nutrition Henrik Friis 04.07.97
feeding aids patients André Briend 28.04.97

Tue, 24 Jun 1997 15:04:40 +0100 (BST)

Subject: Home made infant formulae and HIV

From: "Michael H.N. Golden" <>


Fellow NGONUTs,

The following message has been sent, in a personal capacity, by an individual form an international agency, who wishes to strictly seperate his personal views and requests from those that might be construed as coming from that agency itself. This is sometimes necessary and allows those in "official" positions to contribute when they might otherwise be reticent. Potential contributers are reminded that I am happy to "anonymise" any proper contribution and post it to the list as anonymous.

Best wishes,


Prof. Michael H.N.Golden --------------------------------------------------------------------------

Dear all,

This is a follow-up to the question raised by Lola Nathanail and Fiona O'Reilly in April regarding 'home made' infant formula for use in emergency situations where infants could not be breastfed.

There was a lively exchange of views but I am not sure if there was any onclusion reached.

Perhaps additional exchanges took place 'off the air' - if they did, and if anyone would like to offer an idea about what the conclusion of the debate was, I would be very interested in hearing it.

The issue of alternatives to breastmilk for use in developing countries is of considerable concern as we try to develop ways to help people tackle the issue of how best to advise mothers who know they are HIV+ to feed their newborn infants.

There are two questions that it would be good to have clear answers to - if they exist.

One of these questions is clearly related to the earlier correspondence.

A) Is there any evidence, experience or other source of wisdom to suggest that poor HIV+ mothers living in a poor environment, who cannot afford an adequate supply of commercial infant formula might in fact be counselled to prepare a 'home made formula' from cow's milk, boiled water, sugar etc, which would be of approximately equal effectiveness in ensuring the growth and survival of their infant as a modern 'humanised' commercial formula? If the risks of morbidity and mortality from the 'home made' formula would be expected to be greater than from the commercial formula, is there anyway of making an educated guess about how much greater such risks would be?

[Sub question - before the days of Henri Nestle and his ilk, just how well, or poorly, were infants fed on cow's milk? I have heard that wet nurses were the best chance for survival, but does anyone know of any numbers suggesting risk from cow's milk based mixtures?]

B) As far as HIV negative women are concerned, nearly all authorites regard the evidence on breastfeeding for optimal child health and growth to be quite clear - exclusive breastfeeding for about 6 months, followed by continued breastfeeding for two years and beyond. However, to ensure child survival and growth in circumstances in which an HIV+ woman has decided not to breastfeed, can the same 'consensus' be applied? If, in a situation of very limited resources (and perhaps rapidly failing health) an HIV+ mother cannot hope to purchase commercial infant formula for the whole of the first 6 months of life, should she be advised that it is 'all or nothing', and that she should not start to purchase formula if she cannot continue for the first XX months - or is there any evidence to suggest that a shorter period of formula use, followed by diluted or whole cow's milk, might be less risky than breastfeeding (the current consensus is that the risk of HIV transmission from a positive mother through breastmilk to an infant who was truly [not antibody] HIV negative at birth is about 1 in 7)

I would certainly appreciate informed views on these questions. In addition, I would appreciate ideas about other knowledgeable people or sources of unbiased information which could help in developing some kind of consensus in this difficult area.



Date:Tue, 24 Jun 1997 11:45:03 +0200 (METDST)

From: (Andre' BRIEND)

Subject: home made formulas and wet nurses


I recently attended a paediatric meeting where a talk on history of infant feeding in France was given by a paediatrician from Le Havre in Normandy (Dr. Bernard Le Luyer, Centre Hospitalier du Havre, France).

Parts of his talks may be relevant to the question raised by David on what happened before invention of modern formulas and what can be done to children of HIV+ mothers.

In the 18th century, there was no choice, when the mother could not breast feed but to givethe kids to wet nurses. Mortality figures were incredibly high. Below are data from a local charity :

  1760-1761 1761-1762 1762-1763
Nb of children 226 183 149
Nb of deaths 202 152 125

Attempts to use cow's milk in 18th century Normandy gave even higher mortality figures.

Most of the nurses were paid nurses, with little interest in the survival of the foster child who competed with their own child for whatever milk they had. These figures should not be extrapolated to present situation, but should warn us that 'wet nurses' are not always a good solution. In many countries, there are taboos around this idea.

At the end of the 19th century, a new programme was introduced 'La goutte de lait' (the milk drop). The idea was to give boiled milk with added water + sugar. Mothers received ready to use bottles every day to give to their child. Milk was prepared in a special kitchen by well trained staff. This programme was associated with a progressive drop in mortality from 12% to 6.5% from 1885 to 1925. At that time, it was by far the best choice to feed children who could not be breast fed. It was seen as a major breakthrough.

Dried powdered milk was developped in the first years of 20th century. It became widely used only in the 20's. Distribution of liquid milk continued for some time, but was progressively phased out. In terms of convenience, and in vitamins and minerals composition, new formula were clearly superior.

In 1993, I was surprised to see in Moldova a 'milk kitchen' which actually had the same approach as 'la goutte de lait'. I was told these programmes were common in former USSR.

Milk powder industry was almost non existant before the collapse of the communist regime, and this provided a substitute (?)/supplement to breast feeding for many kids. I heard since that most 'milk kitchen' are now closed: the system worked when manpower was very cheap, and no accountant checked the real price of the programme. Economically, and in termes of quality, these programmes cannot compete with formulas prepared at home with modern powdered formulas if hygienic conditions are met at home. The formula used in these milk kitchen looked very much outdated. I wonder though if the approach could not be updated and adapted to developing countries for HIV+ mothers.


Dr. André Briend

Subject: Home made formulae and HIV

Date: Tue, 24 Jun 1997 16:25:09 +0100 (BST)

From: "Michael H.N. Golden" <>


Dear Anonymous,

I think that the question posed by Lola and Fiona, was slightly different - what to do in an emergency situation with the <6 month old - here with NGO and donor involvement in an acute situation there should be proper (generic formula) provision made for orphins and abandoned children; in such an emergency the HIV status of the mothers will not be known.

1) This is clearly a very emotive issue and there seems to me to be excess of strongly-held opinion and a dearth of real data. The debate I do not believe to be properly or fully informed. Clearly, there should be both a comprehensive literature review and a research agenda set, by those without preformed opinions, so that future debate can be a properly informed. I am hesitant to be forceful at the present time.

3) One ctitical factor in your question is the dichotomous division into the "mother knows she is HIV+" and the mother whp knows that she is HIV-. In nearly all the Afican and Caribbean countries I have been in this knowlege is extremely uncommon, and most do not want to know whether they are positive or not. Seroprevalence surveys show the locally massive extent of the problem, but there seem to be few places where screening is done. Thus, the answer to your question will affect very few mothers, at present.

If there was shown to be a reasonably safe and effective alternative, for very poor mothers, to breast-feeding, that can actually be used on the ground, then the programatic implications have to include the cost and acceptance of HIV screening as well as the realisation that such a stratagy will be difficult in rural areas.

One factor that we do not seem to know is the frequency with which staff "assume" the diagnosis of HIV on clinical ground when this is wrong - we could end up with a situation where all sick mothers are assumed to have HIV and advised not the breast-feed. We do know that many positive women are not suspected, they presumably will continue to be urged to breast feed.

Although about 1/7 of negative children will become positive during breast feeding, this is not necessarily the ADDITIONAL risk of breastfeeding. Mothers that are likly to give their children HIV through breast milk (high viral load) may also be more likely to have infected their children during gestation/birth. The data that gives rise to the 1/7 figure is I understand from a very highly selected population of women.

4) The problems with all formulae that are used by the poor are not just contamination but also systematic under-feeding. In Jamaica when we asked how a mother made up the formula we usually were given the answer written on the tin. When we asked how long a tin lasted the answer was completely different - the modal answer was 2 weeks with no one saying less than one week. - In fact for these children to be properly fed a tin of the size used in Jamaica should last about 3.5 days.

In hospitals in Guinea and Uganda, where mothers were purchasing milk powder outside for their children and were responible for feeding their children in the hospital, I took diet histories for the malnourished children. Their "supervised" intake was between 50 and 80 kcal/kg/d!

I am not at all sure that good-expensive formulae are better than cheap not-so-good home-made formulae in the hands of the very poor (whereas I am sure that the good-expensive formula is better in the hands of those running programs for refugee children).

5) I am sure that there is a lot that can be done to improve home-made formulae that are based upon ruminant milk. Parenthetically, DSM and fresh milk are not the same in terms of micronutrients.

6) I would very much like to know of anyone doing research on simple ways to make expressed-breast-milk non infective for HIV (without destruction of nutrients). If a way was found then the advice to mothers who where HIV+ would be to express breast milk, treat it and feed the child - by far the best solution in theory if not in practice - such advice could then be extended to those who were unsure of their HIV status.

At the moment I would not advise changing to formula for an HIV-unsure mother.

Best wishes,


Prof. Michael H.N.Golden

Date: Wed, 2 Jul 1997 17:10:20 -0400

From: (David Alnwick)

Subject: Re: hiv et al


Dear Dr Friis,

I am most happy to make your acquaintance through Dr Golden. As the Chief of the UNICEF Nutrition Section I am very intererested in networking with workers who are struggling with the issue of breastfeeding and HIV infection in developing countries.

Some of my questions are contained in the 'Anonymous' note that Mike circulated on the NGONUT network - sorry for the mystery, but I was a little wary of making what could be conceived to be 'public' statements in this sensitive area.

If you want to contribute your thoughts on this subject through NGONUT, fine, - if you would prefer to find a way of dialoguing with a smaller group of concerned people including UNICEF, we would welcome this as well.


David Alnwick, Chief Nutrition

Programme Division, UNICEF New York

Fax 1 212 824 6465


Reply Separator

Subject: hiv et al Date: 7/2/97 3:26 PM

Dear David,

I have been making some enquires about who would be able to fill in the unknown parameters that are needed to decide about breast feeding advice. The person is Henrik Friis in Copehnagen <<>>. Henrik has a large study going on at the moment with HIV infected mothers in Tanzania and is fully up with all this literature - I have just persuaded him to come onto NGONUT - and copied the correspondence to him.

However, you might like to contact him directly.

Best wishes,


Prof. Michael H.N.Golden

Fri, 8 Aug 1997

Subject: Letters to Dr. Alnwich about HIV/breastfeeding/nutrition

From: Henrik Friis <>



Below are my letters to Dr. Alnwick in which I discuss HIV and breastfeeding and describe the micronutrient intervention study we are currently doing:


Dear Dr. Alnwick,

I´m so sorry for my late reply - which is partly due to holiday and email problems. As you may have heard from Dr. Golden, we are currently doing a study in Zimbabwe on the effect of daily multi-micronutrient supplementation to pregnant and lactating women on mother-to-child HIV-transmission. As such, I am interested in what is going on in this field.

While there is no doubt that breast-feeding is a strong risk factor for vertical transmission of HIV, there are insufficient data on the timing of post-natal transmission This is why I am concerned about the latest interim statement on breastfeeding - where attention is drawn to the fact that breastfeeding may cause HIV. While this is definitely true, it may be a dangerous message for women in developing countries who are rarely aware of their HIV status.

Firstly, women may be inclined to stop breastfeeding irrespective of their HIV-status. For example, if 1/3 of pregnant women are HIV+ (34% in our study) and 1/3 of these will have a HIV+ child of which 1/3 will be infected post-natally, then if 100 women stopped breastfeeding we would avoid 3.7 (1/27) children getting HIV. This has to compared to the increased morbidity and mortality of breastfeeding deprivation among the 67 (2/3) infants of the uninfected mothers, as well as among the 22 children (2/3*1/3) HIV+ women that would not have had an infected child anyway, but also among the 7.4 (1/9 -1/27) infants that had been infected in utero or intrapartum would still get HIV - who would be inparticular need of breastmilk.

Secondly, we have not identified other risk factors and effect modifiers. One such potentially important effect modifier is maternal nutritional status - there is observational data to suggest that low maternal vitamin A status may increase the risk of having HIV in breast milk (RR=20). Thus, it may be that the risk of post-natal transmission is not absolute, but rather conditional upon maternal nutrition status. If this can be confirmed, then we may have an affordable public health measure, that can be given to all women irrespective of HIV-status - so that we maintain the benefits of breastfeeding and reduce its risks.

Several studies are in progress that deals with these issues. I hope that some results will be presented at the forthcoming Conference on Global Strategies at NIH in Washington in September. I will be happy to keep you informed about emerging studies.

Yours sincerely,


Dear Dr. Alnwick,

The supplements we are giving contain vitamin A in a relatively high dose: 10.000 IU, and then the following micronutrients in US RDA doses for pregnant or lactating women: vit E, B1,B2, B6, B12, niacin, C, D, zinc copper and selenium, and then 3.5 mg betacarotene. Please note that we are not giving iron and folate, since we - wrongly - assumed that women got this routinely. You may be aware that 3 other studies are ongoing: In Tanzania (Fawzi, NIH) they use vit A and/or 4 vitamins (two-by-two factorial), in Malawi (Semba, NIH) they only give vitamin A, in South Africa (Coutsoudis, U. Natal+ MoH) they give vit A + betacarotene.

Thus, minerals are only given in our study. My arguments for a multi- rather than a single nutrient intervention are: 1. multiple deficiencies co-exist, 2. micronutrients interact, 3. to get the most cost-effective intervention. I agree that we are not able to pin out which micronutrient - if any - that works, or to preclude that some have antagonistic effects.

However, the effect of a single nutrient like vitamin A may also depend on co-existing deficiencies. Hence. the effect of vitamin A can not be isolated. However, I agree that our intervention is not ideal - I would have like to have extra arms that should then receive all the micronutrients EXCEPT one. That one could be vitamin A, but it could also be any of the other micronutrients. But 3 - 15 arms was not possible, since we are strugling with the study we are conducting, having budget less than 10% of the two NIH-funded.

We are much critizied for using a multi-micronutrient intervention - it is considered unscientific. In my opinion because people with a medical background think of nutrients as they think of drugs. They fail to realise that even if we only give a single nutrient intervention, the participants will still have a daily intake - unknown, but probably inadequate - of all the other nutrients, and that these deficiencies may interact. We have just had an application to World AIDS Foundation rejected - primarily on these grounds but after an all together incompetent review at Institut Pasteur. Which is unfortunate, because nutrition is really promising in relation to HIV infection.

It is very interesting to hear that there are considerations to introduce multi-micronutrients as part of ANC. I agree that we probably know to little about requirements, bioavailability and interactions. So, I cannot say that what we give is right - but at least it most be right to give more than iron and folate! In fact, we have found out that a Danish company sell a product similar to ours - also in Zimbabwe - called Pregnatal, which contains all our micronutrients - including minerals - in US RDA doses. We realized this after we had commenced our study.

In contrast to the other 3 studies, we recruit all women - irrespective of HIV-status - attending ANC at Mbare, a high density area in Harare. For 2 reasons: 1. to avoid using a positive test as inclusion criteria, but first and foremost because 2. we have a more general interest in the effect of micronutrients in pregnant/lactating women. We are looking at effects on

1. Infant HIV-status at 3 (using PCR) and 15 months,

2. Mortality among HIV+ mothers and their infants.

3. Seroconversion among women initially found HIV-. And (also) among uninfected:

4. Pregnancy outcome: stillbirth, prematurity, birth weight, etc.

5. Infant growth (including knemometry),

6. CD4/8 and other hematological parameters.

Then we also take serum and breast milk samples which we may analyse if we get the money.

Unfortunately, we are not able to follow-up the women-infants regularly as for example the Tanzanian study, which is thus in a better position to get data on effects on timing of postnatal transmission.

Thanks for informing me about the paper by Kuhn and Stein, which I look forward to reading.

Could you please enlighten me on the vitamin A/carotene study in Nepal - what are doses and outcome parameters?

Yours sincerely,


Dear Dr. Alnwick

True, adding an extra micronutrient to a tablet is almost free of charge - possible with the exception of betacarotene? So, as a public health intervention, a multi-micronutrient tablet can definitely be extremely cost-effective, and a single- or oligo-micronutrient tablet probably not justified.

But the question remains, which one to chose as a STUDY intervention. The factorial design is ideal, but in most cases not feasible. But both the single and multi-nutrient intervention is a compromise. However, I tend to believe that there are even scientific arguments for the multi-micronutrient intervention.

And thanks for your suggestion that I send a copy of my e-mail to Keith West. I will do that, and hope to see him later in Cairo.

I'm presenting our ongoing study at the "Nutritional Interventions" workshop after the conference on Global Strategies for Prevention of Mother-to-Child HIV-transmission. Also in that is it useful to know that UNICEF is considering multi-micronutrient supplementation as part of ANC


Henrik Friis, MD, PhD

Danish Bilharziasis Laboratory, Jaegersborg Allé 1D

DK-2920 Charlottenlund, Denmark

Date: Mon, 28 Apr 1997 15:14:34 +0200 (METDST)

From: (Andre' BRIEND)

Subject: feeding aids patients


Feeding aids patients : comparison of AICF formula with WFP CSB

I attendend a meeting on Nutrition in AIDS patients last week in Cannes (France). I was surprised to read an abstract where two blended foods, normally used in refugee settings, were compared to feed AIDS patients in Uganda. This may be of interest to this group.

"Mathias PM et al. Weight gain and immunomoduolation through early nutritional intervention in adults with HIV infection in Uganda". Abstract 2nd international conference on Nutrition - HIV infection To be plublished in abstract form in Nutrition, March 1997 vol 13 No 3.

This study compared two blended foods:

One made according to WFP specs

One test food made in Ireland according to AICF specifications (see Golden, Briend, Grellety, Eur J Clin Nutr 1995; 49: 137-45.). Brand name : Nutrifil. Contained milk among its ingredients. Apparently, it is a non extruded product.

The results show:

a) weight gain was higher on test food than with CSB mean: 2.14 se 0.6 vs 0.98 sd 0.5 p < 0.05

b) Decreased CD4 count in CSB group, no decrease in the test group.

These results suggest that changing specification of blended food may have a favorable functional impact. Whether presence of milk in the tested blended food, different cooking technique, use of different raw infredients or presence of a different mineral vitamin supplementation made the difference remains to be tested.


Dr. André Briend