Previous - Next
This is the old United Nations University website. Visit the new site at http://unu.edu
Dr. McLaren
I have two questions. The first is for Professor Hanson. I was very interested in your data on type I diabetes. You showed that the genetic component was inherited from the father two or three times more often than from the mother, and that the risk was twofold more in bottle-fed than in breastfed babies. It would be very interesting to know whether the risk was equally from the father and the mother in the bottle-fed babies. In other words, it would be helpful if you could separate out the bottle-fed and the breastfed and look at the inheritance from the father and the mother. My guess is that this hasn't been done, or you would have told us. If it turned out to be equally inherited from the father and the mother, then it would support your interpretation. I also wonder whether other autoimmune factors have been looked at, because more and more diseases today seem to be autoimmune in character.
Professor Hanson
You are right. That analysis has not been done. The origin of the interest in this study was that the inheritance from the father and the mother differed, and that the difference may be at least partially attributed to the mother's breastfeeding. There is a large Finnish-Canadian study examining the possibility that bovine milk proteins may be part of the initiation of the autoimmune process, but this is still an open question. As for other autoimmune diseases, some have been discussed, but there are no data yet. The results of experimental studies in newborn mice administered, via the milk, cells from the mother with the same transplantation antigen as the father showed a modifying effect on the reactivity of the offspring against the paternal transplantation antigen. Obviously, maternal cells in the milk can influence normal as well as abnormal immune responses in the offspring. I very briefly mentioned
Crohn's disease, whose origin is unknown, but we know that cytokines are involved in the inflammation in the gut. Obviously, the motherinfant relationship through breastfeeding is not just passive, as we used to think; it is also active. The fact that vaccine responses are enhanced and transplant rejection is diminished is quite significant.
Dr. McLaren
My second question is more general and is relevant also to Dr. Victora. It arises from something that Professor Hanson said about the greater risk of gastrointestinal infection among the babies in Pakistan who had extra water. I had always thought that during very hot weather, both in Pakistan and in England, babies actually needed water as well as breastmilk. Perhaps that is wrong, but if it isn't wrong, then did babies in Pakistan who did not get extra water have some other deleterious consequence? In other words, is extra water needed when it is very hot?
Professor Hanson
There have been a number of studies showing that the extra water is not necessary. We have investigated this in Pakistan during the hot and dry season, and considering the concentration capacity of the kidney, extra water is totally unnecessary. It just brings the risk I suggested; if you give the baby water during the hot season, it will suck less, and less milk is produced. I would strongly discourage this, as it brings increased risk of infection. Last summer was very warm in Sweden, up to 30° to 35°C, and the newspapers erroneously stated that one must give extra water to breastfed infants. This is a very common belief, and in many traditional societies the mothers say, "Oh, yes, you have to give extra water." This has very unfortunate consequences in poor countries with a high risk of infection, especially during the hot season, as we have shown an increased number of episodes of diarrhoea and impairment of short-term linear growth, even of head circumference.
Dr. Victora
I have nothing to add to Professor Hanson's comment.
Dr. Garza
I would like to add a slight modification to what Professor Hanson has said, which I think you would agree with. Although what Professor Hanson has said is certainly true of breastfed infants, because the renal challenge of human milk is so low, that is not true of children who are fed substitutes, because the renal challenge is high enough that those children do need extra water. It is very important that we do not leave here thinking that all infants regardless of how they are fed don't need extra water. I realize that does not contradict what you said earlier.
Dr. Prentice
I just wanted to add to that the case in support of prolonged breastfeeding, because that also seems to provide water for children for long periods. Studies in Bangladesh and elsewhere have shown that the severity of dehydrating diarrhoea is reduced in children who are still receiving breastmilk in the second year of life, showing that it is a very important source of clean water.
Dr. Rasmussen
I wish to put a question to Dr. Victora. In your presentation you seemed to indicate that there was a period after the end of breastfeeding when the infant was particularly vulnerable. Yet we have heard elsewhere, particularly from Professor Howie, that the positive benefits of breastfeeding seem to go on for some period of time, and there are others who have reported the same finding. Those reports don't seem to be in accord.
Dr. Victora
They don't seem to me to be in accord, either. I think we always run the risk of lumping together all infections and different parts of the world in our breastfeeding studies. There are so many variables that may affect how breastfeeding may or may not protect under certain circumstances. Therefore I don't think it is a surprise that we find apparently contradictory results. We found particularly strong evidence of a vulnerable period for dehydrating diarrhoea for three months after weaning. If, however, we look at diarrhoea mortality as a whole, there is no evidence of a vulnerable period by the other case-control study I presented. Overall, 62% of those deaths were due to persistent diarrhoea, so that we are talking about different syndromes with different pathophysiologies. It may well be that this vulnerable period is present for some infections but not for others. I would like to see more studies separating the different clinical types of diarrhoea to see whether this is confirmed. You may remember that my findings regarding otitis media were quite different, showing a long carry-over effect of breastfeeding, so that we are talking about entirely different things.
Dr. Colombo
Dr. Victora, if I did not understand well, it is because my English is limited. You spoke of a survey in which you started from the death of the infant in the hospital, and then you went back to get information and calculated rates of mortality. What are the numerator and the denominator of these rates?
Dr. Victora
The study was not only hospital-based. Eighty-five per cent of the deaths occurred in the hospital, but the data on the other 15% came from cemetery registries or from other registration offices. I have calculated the odds ratios from the case-control study: for example, the cross-product ratio in a two-by-two table of exposure against disease. These odds ratios correspond mathematically to the incidencedensity ratios. We could actually have shown the population rates, because the entire population was under surveillance, but in that particular presentation only the relative risks or odds ratios are changed.
Dr. Garza
You concentrated most of your comments on the impact of early weaning. Would you care to comment on the effects of delayed weaning? I know you have studied this in the past, but do you have any recent data that would help us evaluate how long breastfeeding is protective?
Dr. Victora
No, I don't. My studies go up to two years of age, and I can find a protective influence on infection up to that age.
Dr. Garza
What about the nutritional status?
Dr. Victora
The nutritional status studies we have carried out tend to show that at some stage around one year of age, children who are no longer breastfed tend to grow faster than those who are breastfed, showing, therefore, an inverse association between growth and breastfeeding. Again, there are many caveats because this varies from society to society. Dr. Prentice recently reviewed work from China showing the opposite trend. I guess my reading of the literature on breastfeeding during the second year is that most studies show this inverse association, that is, breastfed babies grow less rapidly than babies who are not breastfed. Nevertheless, there may be a number of confounding variables, self-selection and other factors that may affect the association. On the other hand, all the studies on mortality or infection in relation to breastfeeding in the second year of life tend to show a positive effect, although not all of them are significant, as Dr. Menken mentioned yesterday. So, we have this dilemma in the second year, whether this poorer growth is enough to justify any policy change. My feeling is that the positive effects of breastfeeding are greater than the negative ones.
Dr. Garza
When you say growth, do you mean weight-for-age, lengthfor-age, or both?
Dr. Victora
Different studies have looked at different things. My own studies have shown a greater effect on weight-for-length, but there are studies from other countries showing effects on length-for-age as well.
Professor Hanson
You say they grow more, but do you know what is optimal growth? Another question concerns your mentioning the lasting effects of breastfeeding. I do not think I have made myself clear on that point. The newborn would have relatively few Iymphocytes with which to respond against different microorganisms. Breastfeeding obviously enhances the response, and the result is that there will be many more lymphocytes in the baby to respond to each micro-organism, which means that there will be a lasting effect, because next time the immune response will have many more cells to respond with and will be more efficient. This would be a lasting effect of the immune enhancement from breastfeeding.
Dr. Prentice
I want to get some clarification from Dr. Victora about the effects he sees up to two years. Do you mean you see effects in children who are still receiving milk for two years or in those who were exclusively breastfed for three months and showed effects for two years afterwards?
Dr. Victora
No, I was referring to children who were still receiving breastmilk in the second year of life.
Professor Howie
May I ask Professor Hanson a question? I was very impressed by your data that showed lower responses to immunization in formula-fed than in bottle-fed babies. My question is whether this leads to any suggestion that bottlefed babies with lower antibody responses are more prone to the relevant diseases, or are the levels of response still sufficient to mount protection? This could obviously be important for diseases such as tuberculosis in Africa, where the HIV epidemic is expressing itself very strongly as an increased incidence of tuberculosis.
Professor Hanson
That is a very good question. Let me first say that the control groups were really two groups on different formulas, feeding at different protein levels. There was no difference between those two, but the breastfed group did better. Now, your question, of course, requires that we know the protective level of the antibodies we measure. That is known for just a few infections, like rabies and Haemophilus infuenzae-type infections. Actually, we also know the protective level for diphtheria toxin antibodies, so there we could find an answer for you, but, otherwise, I think we have to rely on the general principle that more antibodies give a better chance of protection. On the other hand, I should add another observation we made with Dr. Cruz of INCAP (Institute of Nutrition of Central America and Panama) in Guatemala. If the mother is infected with diarrhoeal agents during pregnancy, the antibodies she has in her milk against certain pathogens come down during the period of infection. Then they return again. This may mean that it is important to prevent maternal infection during lactation, so that the protection of the breastfed infant does not decrease. We know the protection of the breastmilk relates to the amount of secretory IgA antibody to cholera bacteria and toxin and to Shigella, Campylobacter, and so forth.
Dr. Garza
Professor Howie's question related to the quantity. I know that you have also looked at the quality of the antibody in mother's milk in relation to the mother's nutritional status. Are there any differences in the quality of the antibodies that are made by breastfed and bottle-fed infants, in terms of avidity and other properties?
Professor Hanson
What a nice question. I don't know the answer, but avidity is the capacity of the antibody to bind. The more effectively it binds, the better it neutralizes toxins, and the better it may protect. I don't know the answer, because nobody has done the study, which needs to be done. Nevertheless, we have shown that moderate undernutrition of the mother does not impair the avidity of the milk antibodies.
Dr. Rasmussen
I have a question for Professor Hanson that is off the subject of lactation. What you said this morning would cause me to revise my teaching in maternal health, so I wish to be clear about what you said. You contended that the immune system controls pregnancy, if I am paraphrasing you correctly. I would like you to confirm exactly what you said, and ask you what mechanisms may be involved.
Professor Hanson
Two years ago I would not have believed this new concept. There are a number of animal studies, and now more and more human studies, to support the conclusion that the cytokines from the maternal-antipaternal immune response are a driving force during pregnancy. Thus, the trophoblasts are stimulated to grow primarily by interleukin-3 (IL-3) and granulocyte-colony stimulating factor (G-CSF). They stimulate placental formation. The decidua is prepared by cytokines to receive the blastocyst. Cytokines induce the production of human chorionic gonadotrophin (hCG) and progesterone, as well as placental prolactin. Cytokines finally induce the production of the prostaglandins that result in delivery. IL-8 comes in both at menstruation and at delivery, so that a lot of neutrophils come in, surely to defend the bleeding surfaces open to infections, but also to provide the collagenase that opens up the cervix. IL-6 induces the baby to produce surfactant protein A. It all comes together in a most remarkable way. This is the positive side, with an overweight of TH2 cells driving and protecting pregnancy by production of IL-3, IL-4, IL-5, IL6, and IL-10. If there is a predominance of the TH1 cells stimulated by IL-12, then there may be too much interferony on tumour necrosis factor-a (TNF-a), resulting in intrauterine growth retardation or even abortion. IL-10 can down-regulate the TH1 cells to produce less interferon-y. So the question is obvious: can we demonstrate intrauterine growth retardation that can be prevented by IL-10? Of the 20% of children with low birthweight in Pakistan, 75% have intrauterine growth retardation, and in India the figures are even higher.