Contents - Previous - Next

This is the old United Nations University website. Visit the new site at


Discussion of paper by McNeilly


Dr. McLaren

Thank you for a very clear exposition of a very complex subject. 1 was particularly interested to learn about that double-assurance system that is nature's way of suppressing fertility, emphasizing the importance of this mechanism for spacing births.

Dr. Pérez

You mention that the key is suckling, but what is the mechanism? How could one reproduce that?

Dr. McNelily

We do not know how to mimic the effect of suckling. If you increase prolactin by drug treatments, that has some effect on LH pulse generation, perhaps by an effect on GnRH rather than by a direct effect of prolactin. The bottom line is still that lactational infertility is driven by a change in GnRH pulse output, and we don't know how to regulate that.

Professor Hanson

Would a mother of a premature baby who is pumping or extracting her milk be able to stimulate her nipples to the extent that she would be able to induce infertility?

Dr. McNelily

Induced lactators make milk, but no one has really looked properly at reproductive performance, so there is no published literature. There are anecdotes that suggest that there is an effect on the resumption of menstruation when they express milk by hand but not with breast pumps, so I do not think it is just emptying the breast. I think it is actually nipple stimulation.

Professor Howie

I don't think that the situation in the pre-term infant has been studied systematically as it has been with term babies. Because pre-term babies will not be suckling with the same regularity or strength, their suckling might not have the same effect as that produced by the full-term baby.

Dr. Garza

There are some very limited data showing that use of the breast pump does not result in the same prolactin surges as direct suckling, probably not only because of the lack of physical stimulation, but also because of the lack of a psychosomatic component. It may be that the neural pathways do not respond in the same way. In the normalterm infant, do you know if anyone has compared the effects of nutritive and non-nutritive suckling on maternal endocrinology? The pattern of suckling is so different that there may be reason to suspect that one or another is more effective in sustaining the hormonal changes that lead to amenorrhoea.

Dr. McNailly

The problem is that when you look at the world literature, nobody uses the same definition. Now that a single definition has been developed, it should be used to describe suckling. In Thailand, with Dr. Sodsai Tovanbutra, we counted the number of sucks per suckling episode to determine if it was related to prolactin release. There was no relationship at all. So the idea that prolactin is one of the monitors of the suckling stimulus and GnRH pulse generation doesn't work. There was one defect. We just counted the number of sucks. We did not have a recorder, so we could not work out the intersuck interval. The problem is that we do not really know how to monitor suckling activity.

Dr. Woolridge

The data on pump use and prolactin are ambiguous. Professor Howie, I think that your study is the only one that shows the complete absence of a prolactin response during the use of an electric breast pump, but there are at least four published studies that show a reasonable prolactin surge. We have just completed a study with women who have established milk production with an electric pump, and they have been doing repeated within-measures studies. On every occasion, 24 women had a prolactin surge every time they used an electric pump. Maybe one of the reasons why non-nutritive sucking is effective is that the endocrine system may be more sensitive to particular types of sucking, such as that associated with low milk release. We have a pump that operates at twice the normal rate, and we are anxious to look at the prolactin surge.

Dr. McNeilly

Is the pump a suction pump, or is it actually a proper milking pump?

Dr. Woolridge

We have just compared single and double pumping with and without one of the Silastic inserts that give it a more physiological style of action. We were unable to show any impact of the Silastic inserts on prolactin release.

Professor Hanson

The effectiveness of suckling is especially important in poor countries, where 20% to 30% of newborns may be of low birthweight. Is the spacing between pre-term, lowbirthweight children shorter than that between full-term children?

Dr. Prentice

We have to remember that many low-birthweight babies in the third world are actually term babies. They are small for dates, but they are not premature, so we should not get confused.

Dr. Pollitt

There are large and significant differences between the sucking patterns of pre-term and term babies on the bottle. I do not know about the breast. But almost any parameter that you can think of or identify in the term baby is going to be different from that of the pre-term baby. In other words, the amplitude, the frequency, the duration of the burst, the length of the period without sucking, and so forth are all different. What is particularly interesting in the pre-term baby is that there is continuous sucking. There are no bursts, but instead there are continuous, varied, small-amplitude sucks.

Dr. McNeilly

It is very difficult to know exactly what part of the suckling stimulus is important. We have seen women who suckled 25 times a day and women who suckled 5 times a day, each with a total suckling duration of 60 minutes per day. They have the same patterns of gonadotrophin secretion and ovarian function, so clearly the brain is very adaptable at reading a signal coming from the nipple. I would not be surprised if there was an interaction between the baby and the mother as well, helping to amplify this stimulus. In Chile, Dr. Diaz* increased the suckling frequency in breastfeeding women in an attempt to switch off ovarian activity after it had resumed. The attempt failed, because although the number of suckling episodes was increased, the babies reduced the duration of each feeding so that the overall apparent suckling input was the same and the increased number of episodes had no effect whatsoever. The problem is that the system is so adaptable. To say you need so much suckling to maintain infertility is very difficult. What we should try to do is to determine the minimal parameters for each particular society, parameters that we can actually measure. In Edinburgh we chose frequency and duration of each feeding to see if that would give a guide to maintaining infertility. It worked in our particular group of women.

Dr. Prentice

Could you bring us up to date on the thinking about the mother's nutritional status? Our supplementation studies in lactating women suggested that feeding the women more resulted in their ovarian activity returning earlier, and we put that down at the time to the lower prolactin levels. Do you think that it is sucking frequency only, or does the mother's nutritional status have some impact?

Dr. McNelily

If you look at situations involving acute weight loss or chronic weight loss, you actually have to lose a lot of weight to switch off ovarian activity. Andrew Loudon, John Milne, and I** compared red deer on a restricted diet and on a full diet. We showed that on a restricted diet, the rate of milk secretion was much lower, so the calves kept going back more often. The supplements in your study may have subtly altered the suckling activity, perhaps the strength of sucking at each episode, without a change in the number of suckling episodes.

Dr. Prentice

In the context of the meeting, what you appear to be saying then is that if a well-nourished mother has a high suckling frequency, she is likely to have just as long a period of fertility as an undernourished mother. The common perception is that for women in the third world, breastfeeding as a contraceptive is all right, but for women in developed countries, it's not a particularly good method of contraception.

Dr. Rasmussen

My question relates to night feeding. We were all very intrigued by your previous night-feeding findings. They seemed to justify feeding our babies at night. Your comments now seem to have wiped out that justification. Can you tell us more about what you now know that makes you think night feeding is not important?

Dr. McNeilly

Our early studies suggested that a woman who was breastfeeding at night was likely to remain amenorrheic longer. As we accumulated more data, it became apparent that it was not necessarily the night feeds that were important. It now seems that if women feed six times a day, it can be between eight in the morning and six at night, and that will maintain the infertility.

Dr. Rasmussen

So it is just the number of times a day or the total minutes of suckling, not when in the 24-hour period it occurs. Earlier, you hypothesized that it was interfering with the LH surge at ovulation.

Dr. McNeilly

Yes, that's right, because at that time (10 years ago), that was our state of knowledge. But I think now, having accumulated more information, that the night feeding is not necessarily as important as we assumed. It is the suckling input during the 24-hour period that matters. In Keith Gordon's*** studies in monkeys, which have patterns similar to the human in resuming ovarian activity (pulsatile LH secretion and so forth), restricting suckling to either daytime or night-time has made no difference to gonadotrophin secretion, provided the suckling frequency remains the same. Furthermore, the anecdotal reports that we have from women who are working during the day and feeding only at night suggest that infertility is maintained, provided suckling is continued.

Dr. Valdés

If a working mother breastfed more frequently at night but kept the same number of daily feedings, would that be enough to maintain amenorrhoea?

Dr. McNeilly

I think it probably would, but I would not like to be held responsible for that practice being adopted and then failing. It requires specific investigation.


Demographic effects of breastfeeding: Fertility, mortality, and population growth

Jane Menken and Randall Kuhn



This article reviews the demographic effects of breastfeeding on fertility and child survival and, ultimately, on population growth. Extended breastfeeding both reduces fertility by prolonging birth intervals and increases child survival through improved nutrition, especially where adequate substitutes are not available. The results presented show, however, that although breastfeeding is a major determinant of fertility in the absence of other means of fertility control, prolonged breastfeeding alone cannot reduce fertility to levels consonant with slow or zero population growth. The benefits, at least for the first year of life, demonstrate the need for policies that promote breastfeeding and encourage compatibility between breastfeeding and other aspects of women's fives. In particular, policies are needed that permit women to breastfeed their children while, at the same time, improving their socio-economic circumstances through participation in the labour force.



Breastfeeding is known to reduce fertility and increase child survival. It affects fertility by prolonging the interval between births. It increases child survival both directly, through the provision of beneficial nutrition to the child, and indirectly, through improved child spacing. In this article we review evidence that leads to these conclusions and consider two questions that relate the findings on breastfeeding to population growth:

We first consider how low fertility must be, when combined with the low mortality even the poorest countries have achieved in recent years, to lead, ultimately, to either very slow or no population growth. We next offer illustrative data on recent levels of fertility and mortality around the world. This section is followed by a discussion of the factors that determine levels of fertility, including breastfeeding, and then by an examination of the evidence relating breastfeeding to infant and child survival. We end with a set of policy recommendations.


Population growth

During the 1960s and 1970s, there were many assessments of the world population situation. Some were by alarmists who wrote of the "population explosion" and credited population growth with responsibility for many, if not all, of the afflictions of the developing world. Others, taking a more judicious view, still concluded that rapid growth, on balance, had consequences sufficiently detrimental to prospects for improving conditions in much of the world to warrant efforts to slow the increase in population by reducing fertility. Indeed, this has been the primary rationale for family planning programmes.

In 1974 the United Nations sponsored an International Conference on Population in Bucharest, Romania. Although some countries urged the adoption of policies aimed at slowing growth, primarily through family planning, out of that conference instead came the dictum that "development is the best contraceptive." However, this dictum was not followed by even its most fervent supporters in the developing world. Instead, developing countries increasingly turned to population policy and family planning programmes aimed at reducing fertility.

One of the major challenges to the received wisdom on population questions came in 1981 with the publication of Julian Simon's book The Ultimate Resource [1]. Simon, while not opposed to the principle that individuals should be able to control the number of children they have if they so choose, contested the view that rapid growth had deleterious effects. He argued that population growth is a significant and effective long-term stimulus to economic development, exerting its influence by, among other means, increasing the tempo of innovation. His views profoundly influenced the US Government's stance at the 1984 United Nations International Conference on Population [2].

Ten years later, the 1994 United Nations Conference on Population and Development was held in Cairo. The US position, like those of many other countries, had changed, with concern for human rights, the status of women, maternal and child health, and the environment being discussed in preparatory meetings around the world as well as at the conference itself.

Any discussion of breastfeeding and population must take place against the backdrop of these concerns. It must also take into account the fact that population growth, continued indefinitely into the future, will ultimately outstrip space on earth. There have been any number of demographers who have amused themselves (if no one else) with calculations of how long it would take for the population to reach numbers so large that each person had only one square foot of earth to occupy, or until the numbers represented mass equal to that of the earth, and so on. At an annual growth rate of 2%, it takes about 35 years for a population to double, at 3% it takes only 23 years, and at 4% it takes about 17 years. (The doubling time is about equal to 69/annual growth rate.) The non-trivial point of these calculations is that ultimately birth and death rates must again come into balance. In the past they were balanced at high levels of both fertility and mortality. The issue is whether they can be brought into balance for the world, as well as the currently developed nations, at low levels of mortality, which require correspondingly low levels of fertility.


Some mathematical relations

If fertility and mortality rates are constant over time, the mathematics of population growth lead to some important relationships that are useful in this discussion [3, 4]. The growth rate r is related to the average number of children a woman bears and to the survival of those children so they can contribute progeny to the subsequent generation. The net reproduction rate (NRR) is a measure that summarizes both fertility and child survival. It is defined as the average number of daughters a girl born today will contribute to the next generation. If birth and death rates are unchanging over time, results from mathematical demography show that r is related to NRR through the average age at which women bear their children. Thus, if that average age is 28 (as is characteristic of a wide range of populations) and r is expressed as a proportion (for example, if growth is 2% annually, then r = 0.02), then a powerful result of the mathematics of population growth [3, 4] is that

In (NRR) ~ 28r

so that

NRR ~ e28r.

NRR itself can be estimated from two other measures related to fertility and survival. The total fertility rate (TFR) is the average number of children borne by a woman who survives her entire reproductive span, in other words, to age 50. The number of daughters is about equal to TFR x .488, where .488 is roughly the fraction of children who are girls. The proportion of girls born today who may contribute children to the next generation can be approximated by the proportion surviving to the average age of child-bearing, for which we are using p(28). With these assumptions,

NRR ~ TFR ´ .488 ´ p(28).

Using this approximation, table 1 shows the NRR needed for growth rates ranging from 0% to 4% per year and, for TFR ranging from two to nine children, the p(28) that would lead to these growth rates. For example, an annual growth rate of 1% can be achieved with TFRs from three to nine, but the corresponding survival probability that would maintain this growth goes down sharply as the average number of children increases. Table 2 shows the p(28) values corresponding to life expectancies at birth that range from 20 to 80 years, taken from the West Model Life tables developed by Coale and Demeny [5]. The values that correspond to life expectancies of 50 years or more are in bold type in table 1. They clearly indicate that TFRs of five or more are associated with growth rates of at least 2.5% per year and doubling times of less than 28 years.


Contents - Previous - Next