Contents - Previous - Next

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

The proximate determinants framework

To understand the causes of these common patterns and of variation in natural fertility and in controlled fertility, and the magnitude of the effects of various factors on reproduction, demographers usually consider the proximate determinants of fertility. The proximate determinants can be divided, as shown in figure 1, into those that affect the reproductive span and those that influence the rate of births within that span [11].


Reproductive span

A woman's potential reproductive span begins when she first ovulates, an event that is usually roughly measured by first menses. It ends when she becomes sterile or dies, whichever comes first. Yet rarely, if ever, is all of this potential reproductive span devoted to reproduction. Most societies have practices that limit the effective reproductive span to less than its maximum, by postponing marriage (or unions) until some time after menarche or through permanent abstinence or dissolution of marriages. In many traditional societies in the past, marriage took place very close to menarche, so the mean age at first marriage was often under 20 years for women. In contrast, the mean age at marriage in a number of European countries is in the mid-twenties or later, thus substantially reducing the time available for reproduction. Therefore, the effective reproductive span starts at menarche or marriage, whichever is later, and ends at marital dissolution, sterility, the start of permanent abstinence, or death, whichever is earliest. The effective reproductive span can be thought of as the period when a woman is exposed to the risk of child-bearing. Of course, in many societies there are also intermittent "time-outs" when the woman is not exposed to this risk, either because the marriage has dissolved or because of temporary spousal separation.

FIG. 1. Reproductive span and birth intervals


Birth interval dynamics

Within the effective reproductive span, the pace of childbearing is determined by the lengths of the intervals between successive births. These intervals themselves can be divided into their component parts: (1) the post-partum period, which follows each birth and is the time from the birth until both ovulation and sexual relations are resumed; (2) the time to next live-birth conception, which is made up of the time to conception and, if there are any foetal losses, an interval for each foetal loss comprising the pregnancy and post-partum period associated with that foetal loss, plus another time-to-conception interval; and (3) the pregnancy that leads to the next live birth.

The first birth interval (between the start of the effective reproductive span and the first birth) differs from later birth intervals only in that it does not contain the post-partum period that starts the later intervals.

The time to next live birth is determined by fecundability, the monthly probability that a woman will conceive, and by the likelihood that fertilization of the ovum will not lead to a live birth (and the length of the pregnancy and the postpartum period associated with the end of that pregnancy). Fecundability itself is determined by the frequency and pattern of intercourse, the length of the ovarian cycle, whether or not the cycle is ovulatory, and the probability that a single act of intercourse will lead to a live birth (see references 12 and 13 for the most recent modelling of fecundability).

Table 4 lists a number of factors affecting the effective reproductive span and the pace of childbearing within that span and, for each factor, an indication of whether it is primarily biologically or socially determined. It can be argued that all of these factors are determined both biologically and socially. For example, menarche occurs as part of normal biological development. It is well known, however, that it is strongly related to nutrition, and the feeding of girls is determined both by social norms regarding appropriate food distribution within the family and by the availability of food. The latter is a result of the ability to obtain food, which is itself dependent on the supply of food in markets or through ownership of land and access to that supply.

Breastfeeding exerts its effect on fertility in two ways. The first and by far the stronger effect comes through prolongation of post-partum amenorrhea, the time from the birth until a woman resumes ovulation or menstruation. The second effect occurs after the end of the post-partum period, when breastfeeding can reduce the probability that the next ovulation will result in conception. In the first case the frequency and extent of breastfeeding are primarily socially determined-and that very social determination has led to this conference. Its effect on amenorrhoea is, of course, biological. Women who do not breastfeed resume ovulation within a very short time (about six weeks) after a live birth [14]. By contrast, the very long breastfeeding periods in rural Bangladesh around 1970 led to a median time until resumption of menses of about 18 months [15]. Although this mode of breastfeeding action has been well known for many years, more recently it has been found that breastfeeding beyond the resump

TABLE 4. Factors affecting the effective reproductive span and birth intervals

Features that are affected Affecting factors
Effective reproductive span
menarche biological and social
marriage or regular
sexual union
sterility (which may be
earlier than menopause)
biological and social
widowhood biological and social
divorce social
death biological and social
Birth intervals
Post-partum period
amenorrhoea biological
breastfeeding social
resumption of
sexual relations
frequency and pattern
of intercourse
proportions of cycles
that are ovulatory
breastteeding social
duration of the fertility
probability that conception
will follow a single act
of intercourse
contraception social
Probability of foetal loss
spontaneous social
induced social
Length of non-susceptible period associated with foetal loss
spontaneous biological
induced biological and social
Gestation leading to live birth biological and social


An aggregate fertility model

Bongaarts [20-22] developed a model that took into account marriage, post-partum infecundity, and fertility limitation through contraceptive use and through induced abortion as determinants of TFR. It takes as its starting point total fecundity (TF), the number of children women would have if they reproduced at

the maximum, which would occur if (1) no woman ever breastfed, (2) no woman used contraception or had an induced abortion, and (3) all women were married throughout the biological reproductive span, from ages 15 to 50. Bongaarts estimated that TF is just over 15 children. He then developed four indices to be estimated for each population of interest:

Ci=index of post-partum infecundity. The index varies from 0 to 1 and represents the proportion of potential fertility, TF, remaining when the average post-partum period of the population of interest has been taken into account. Therefore, Ci = 1 if the population does not breastfeed at all. Because of our interest in breastfeeding, the equation for estimating Ci is given in the next section. The fertility-reducing effect of post-partum infecundity is (1- Ci).

CA =index of abortion = proportion of TF, after postpartum infecundity has been first taken into account, remaining when the effect of induced abortion in reducing live births has been taken into account. Spontaneous abortions are included in the original estimate of TF, since they are treated as a purely biological occurrence. Note that few countries have sufficient information available on abortion to make reasonable estimates of CA [23].

Cc = index of contraception = proportion of TF, after the effects of post-partum infecundity and induced abortion have been taken into account, remaining after contraceptive use has been considered.

Cm = index of marriage = proportion of TF, after the first three factors have been considered, remaining when the particular marriage pattern has been taken into account. Cm actually represents the number of years a woman is married, relative to the maximum possible, but it is a weighted number, in that a year of marriage during the part of the reproductive span when women are most fecund is counted more than a year of marriage towards the end of the reproductive span, when many women are already sterile or subfecund.

Thus, in the Bongaarts decomposition,


Because the emphasis in this article is on breastfeeding, calculation of the indices other than Ci will not be discussed.


Estimating the index Ci

The estimation procedure is based on the prolongation of the birth interval when the post-partum period extends beyond the biological minimum, and there is no use of contraception or induced abortion. If there is no breastfeeding, a rough estimate of the birth interval is about 20 months, based on 1.5 months of anovulation, 7.5 months waiting time to conception, 2 months added by spontaneous abortion, and 9 months of pregnancy leading to a full-term live birth. The birth interval, divided into the effective reproductive span, gives an estimate of the number of children a woman could bear. The index C' asks how much this birth interval is increased by a longer post-partum period before both ovulation and sexual relations are resumed. Without either lactation or delayed sexual relations, this birth interval averages 18.5 months. It lasts 18.5+i, where i is the average duration of the post-partum period, extended either by breastfeeding or by post-partum abstinence. The index becomes the ratio of the birth interval without breastfeeding or post-partum abstinence (20 months) to the birth interval with either or both:

Ci = 20/(18.5 + i).

The duration of post-partum abstinence is usually estimated from data collected for that purpose. But frequently the average duration of post-partum amenorrhoea is estimated from the time until first menses post-partum, which is itself estimated from women's reports on the length of breastfeeding. Bongaarts [22] developed an equation from an analysis of all data available at the time on observations of both the duration of breastfeeding and the time to first postpartum menses that then permits rough estimation of the duration of amenorrhoea from data on how long women breastfeed. Usually i, the post-partum period, is assumed to be equal to the estimated period of post-partum amenorrhoea. The exception is for those populations in which couples abstain even after amenorrhoea ends. In those cases i is estimated as the length of post-partum abstinence.

TABLE 5. Estimates of total fertility rate and Bongaart's proximate determinants indicesa

Country and date TFR Ci Cm Cc
Developing countries
Bangladesh, 1975 6.34 .54 .85 .90
Colombia, 1976 4.57 .84 .58 .61
Costa Rica, 1976 3.69 .90 .57 .47
Dominican Republic, 1975 5.85 .61 .60 1.0
Guatemala, 1972 7.05 .61 .72 1.0
Hong Kong, 1978 2.26 .93 .88 .18
Indonesia, 1976 4.69 .58 .71 .75
Jamaica, 1976 4.32 .88 .54 .59
Jordan, 1976 7.41 .80 .74 .81
Kenya, 1976 8.02 .67 .77 1.0
Korea, 1970 3.97 .66 .58 .68
Lebanon, 1976 4.77 .78 .58 .69
Malaysia, 1974 4.76 .90 .61 .57
Mexico, 1976 5.73 .84 .61 .73
Nepal, 1976 6.37 .55 .85 .89
Pakistan, 1975 7.02 .64 .79 .91
Panama, 1976 4.57 .88 .62 .55
Peru, 1977 5.11 .76 .57 .77
Philippines, 1976 5.01 .76 .61 .70
Sri Lanka, 1975 3.53 .61 .51 .74
Syria, 1973 7.00 .73 .73 .86
Thailand, 1975 4.70 .66 .63 .74
Turkey, 1968 5.60 .73 .76 .66
Denmark, 1970 1.78 .93 .55 .23
Finland, 1971 1.61 .93 .51 .22
France, 1972 2.21 .93 .52 .30
Hungary, 1966 1.80 .93 .62 .21
Poland, 1972 2.09 .93 .44 .34
United Kingdom, 1967 2.38 .93 .61 .27
United States, 1967 2.34 .93 .63 .26
Yugoslavia, 1970 2.11 .93 .57 .26
Historical populations
Bavarian villages, 17001850 4.45 .85 .37 .91
Crulai, 1674 1742 5.60 .67 .57 .96
Grafenhausen, 1700- 1850 4.74 .67 .44 1.0
Hutterites 9.50 .82 .73 1.0
Ile de France, 1740-1779 6.10 .71 .50 1.0
Oschelbron, 1700-1850 5.06 .73 .48 .95
Quebec, 1700 1730 8.00 .81 .63 1.0
Tourouvre-le-Perche, 1665-1714 6.00 .75 .59 .89
Waldeck villages, 1700-1850 4.41 .68 .44 .96
Werdum, 1700-1850 3.78 .64 .40 .96

The impact of breastfeeding on fertility

Table 5 offers Bongaarts' estimates [22] of these indices, except for CA, which is assumed to be one, for a number of populations around 19.70 and for several historical populations. The major impact of breastfeeding can be seen through the values of the index Ci in countries that, at the period in question, did not limit their fertility through contraception or abortion. Whereas in Western Europe the demographic transition to low levels of fertility was caused, to a great extent, by the adoption of a pattern of very late marriage and a relatively high degree of non-marriage, much of the developing world is characterized by high proportions of the population being married throughout the reproductive span. In these cases the overall level of fertility has been determined, to a great extent, by post-partum infecundity. Although in some societies post-partum taboos set the length of the post-partum period, in most it is breastfeeding that determines when a woman again becomes capable of conceiving after the birth of an infant. Yet the maximum reduction in TF, the potential fertility, is about half; if breastfeeding is the only way fertility is reduced, the numbers of children remain quite high.

Breastfeeding, as already mentioned, has another substantial effect on fecundability. When continued beyond the post-partum anovulatory period, breastfeeding reduces the likelihood of conception. This effect does not enter into the Bongaarts model but can be thought of as reducing the Ci levels slightly below those presented in table 5.

Three conclusions can be reached thus far:

This last statement, although true, must be discussed in the context of adoption of means of fertility control. Although there have been fears, over the years, that populations would experience massive fertility increases as, over the course of development, women reduced both the intensity and the duration of their breastfeeding, these fears have not been realized. In every country for which there are data, the most educated women, who breastfeed the least, also have the lowest fertility (see table 3). Although our models say that if breastfeeding declines, holding all other factors constant, fertility will rise, in the real world it is rare that all those other factors remain constant. The same processes of social change that affect women's attitudes and practices concerning breastfeeding also affect their attitudes towards and practice of fertility control. There may be lags, so that over the short term following a decline in breastfeeding, fertility might increase, but the general pattern has been one of sufficient compensation, through adoption of family limitation, that the expected rise in fertility has not taken place.

There is a countervailing effect, however. To the extent that reduced breastfeeding increases infant and child mortality, the effect on growth rates is muted. Growth rates depend on NRR, whereas our discussion thus far has been limited to TFR. We therefore turn next to a consideration of the effect of breastfeeding on survival.



Although the proximate determinants of fertility have been systematized and to some extent quantified, we are not yet able to set out the determinants of child mortality in such a neat manner. Mosley and Chen [24], in one attempt at developing such a framework, attributed child mortality to five sets of factors: (1) maternal factors, (2) environmental contamination, (3) nutrient deficiency, (4) injury, and (5) personal illness control. They also refer to a negative synergy by which sickness and malnutrition amplify each other's effects and lead to childhood mortality; sickness leads to malnutrition, malnutrition leads to sickness. Therefore, relatively few child deaths are attributed directly to malnutrition, because most malnourished children are recorded as dying of an illness to which their maluourishment has left them susceptible.

Of these five groups of determinants of child mortality, three are profoundly influenced by breastfeeding: maternal factors, environmental contamination, and nutrient deficiency.

The maternal factors include child spacing, which is directly influenced by the effects of breastfeeding. Longer breastfeeding lengthens the birth interval when there is little or no use of contraception. This longer birth interval generally leads to decreased mortality, mostly due to lessened competition for maternal and household resources. With longer birth intervals, there will be fewer children in the family, and perhaps more importantly, it is less likely that the family will have two or more very young children at the same time. Therefore, parental efforts to obtain medical care and provide adequate nutrition and an environment that limits disease transmission can be focused on one child at a time in their period of greatest vulnerability, rather than spreading these efforts thinly across several children simultaneously. This effect may be especially important for daughters, who may be disadvantaged when they must compete with their brothers for familial resources.

Breastfeeding directly influences infant and child health by providing a better food supply when the child is quite young. Until quite recently we would have had no potential negative effects to report. The spectre of AIDS and HIV has cast a shadow over this issue, sometimes invalidating this conclusion. Recent studies have shown that HIV-infected mothers may transmit the infection to their children through their breastmilk, especially in the colostrum in the first few days following the birth [25]. The risk of transmission is difficult to assess, since few largescale studies have been conducted, and it is difficult to distinguish transmission of HIV through breastmilk from earlier transmission. Dunn et al. [26] estimated the risk of transmitting HIV-1 through breastmilk to be 14% to 29%, depending upon whether the mother was infected before or after the birth of the child. The risk of transmission of HIV-2 seems to be much lower [27]. In most developing world environments, the risk of HIV is not as great as the threat of mortality owing to infections from water used in infant formula and from poor nutrition [28]. Therefore, the World Health Organization [29, see also 30] recommends that breastfeeding be promoted in all developing countries, even where there is a high prevalence of HIV infection. It further recommends that "Where infectious diseases and malnutrition are the main causes of infant deaths and the infant mortality rate is high, breastfeeding should be the usual advice to pregnant women, including those who are HIV-infected." Only where mortality is low and there are safe alternatives to breastfeeding should mothers infected with HIV be advised to use these alternatives [29].

Vanlandingham et al. [31] reviewed studies that attempted to quantify the effect of breastfeeding on child survival. They report Holland's [32] finding that in Malaysia breastfeeding was significantly related to infant mortality in the first six months. Infants who never breastfed were 12 times more likely to die in the first two months of life than those who had at least some breastfeeding. Holland divided infants who survived the first two months into those breastfed at least one month, those breastfed less than one month, and those never breastfed. There was a clear gradient in survival over the next two months; those breastfed longest had the greatest chance of surviving to the age of four months. Similarly, among those who survived the first four months, those breastfed longest again had the highest chance of surviving the next two months (until age six months). According to Vanlandingham et al. [31]:

Holland's work is characterized by a careful attempt to eliminate reverse causation-infants may be too sick to suckle, so that death 'causes' weaning rather than the reverse-but this thorny problem probably cannot be entirely overcome. The problem is minimized in the case of infants in age groups after one to two months, because the analysis is conditioned on breastfeeding behavior prior to the start of the interval.

In the demographic research by Montgomery et al. [33], a variety of methods were used to control for the child's health status at birth and to make corrections for selection bias; the same conclusion was reached: "The direct influence of breastfeeding on survival remains of overwhelming importance."

Research in many countries, including those in Latin America [34], has detected improvement in child survival in the last half of the first year for breastfed infants. A number of authors have argued that breastfeeding should be most important in the countries (or the subgroups within a country) with the least favourable health conditions [31, 35]. Palloni and Millman's [36] evidence supports this hypothesis. Using World Fertility Study data from 12 Latin American countries, they found that consistent breastfeeding had a positive effect on child survival through the first year, which was stronger in countries with high levels of infant mortality and for children of less educated mothers. But when older ages (up to five years) were considered, although trends were in the direction of improved survival the longer an infant was breastfed, the differences were not statistically significant [34].

Breastfeeding also has an indirect effect on infant survival by lengthening birth intervals. Hobcraft et al. [37] found that children born after a short interval (less than two years) had a probability of dying in the first five years of life that was 52% to 161% higher, depending on whether the older sibling survived or died. Hobcraft's analysis did not control for breastfeeding of either the older or the index child. According to Vanlandingham et al. [31], others have argued that the effect of a short previous interval on child survival is not the result of a short duration of breastfeeding per se but is caused by other factors (probably maternal depletion, heightened competition for resources among siblings and greater risk of infection among children who are of similar age), because controlling for the duration of breastfeeding of the index child somewhat attenuates but does not eliminate the effect of the previous interval.

The role of maternal depletion has been suggested in research from Bangladesh and the Philippines. Shorter subsequent birth intervals also are associated with higher mortality of the index child, although the causal mechanisms have not been conclusively delineated.


Effects of reducing breastfeeding

We have seen that breastfeeding affects both the number of children borne by a woman who reaches age 50 and the number of her daughters who survive to join the next generation of parents. Reducing breastfeeding may, therefore, affect NRR in different directions, by increasing fertility and thus causing a rise in the number of daughters, and by increasing mortality and thus causing a decline in the survival of those daughters. Do these effects totally counterbalance, cancelling one another? Or does one outweigh the other?

Palloni and Kephart [38] addressed this issue through analysis of data for Latin America from the World Fertility Surveys and through the development of an analytic and simulation model. Considering data for three countries, Colombia, Ecuador, and Peru, they found that if all women stopped breastfeeding without using contraception, infant mortality would increase by 3% to 10%. In two of the countries, the direct effect of stopping breastfeeding would account for about 95% of this effect and the indirect effect through birth intervals would account for the remainder. In the third country, the direct effect would still be about 70% of the total.

They also considered two groups of women, one using ineffective contraception and the other using effective contraception, both of whom stopped breastfeeding. In both situations, the effect on infant mortality was minimal, because these women were already spacing their children. Hence, the direct effect of giving infants foods other than breastmilk increased mortality only slightly. They found, under these and a wide range of other circumstances, with and without family planning, that a reduction in breastfeeding caused fertility to rise more than it caused child survival to fall. The net result of reduced breastfeeding is increased overall fertility and population growth. They also concluded that increased family planning can compensate for this fertility-increasing effect.



The evidence presented here overwhelmingly supports the desirability of promoting breastfeeding. Life chances for children are improved if women have the opportunity to breastfeed them for at least the first year of life, especially in harsh disease environments, where safe water for formula is not available and diseases that can be prevented by breastfeeding are more prevalent. Even in the best circumstances, the benefits of breastfeeding are demonstrable. Therefore, policies that promote breastfeeding at least for the first year, are strongly recommended.

However, we must be judicious in our claims for the benefits of breastfeeding. We suggest two important cautions:

There is great concern that breastfeeding duration tends to decrease as a country's economy develops. This seems logical, as growing economies usually have greater female workforce participation and thus may offer women less opportunity to breastfeed. A drop in breastfeeding under these circumstances may lead to conflicting effects of socio-economic development on mortality. Mortality falls as economic circumstances improve, because of better access to health care and nutrition and better standards of public health and hygiene. These mechanisms work at both the individual and the societal levels. Meanwhile, at the individual level, if these same populations breastfeed their children less, they expose children to greater risks of childhood diseases, and unless they use contraception that compensates for shorter post-partum periods, their fertility goes up. In most countries the positive effects of development have clearly outweighed the negative effects through less frequent breastfeeding. Yet there is room for policies and programmes to alleviate these negative effects.

We recommend policies that ensure the compatibility of a woman's participation in the labour force and breastfeeding for at least the first year, so that she is not forced to choose between two activities that both provide benefits to her children. The scientific evidence of the benefits to children overwhelmingly supports efforts to promote breastfeeding and to maintain and improve its compatibility with all aspects of mothers' lives.



The first draft of this article was prepared at the University of Pennsylvania with support from the Mellon Grant for Research on Developing Countries and a training grant from the National Institutes of Health. It was completed while the senior author was a fellow at the Center for Advanced Study in the Behavioral Sciences, with support from the University of Pennsylvania and grant SES-9022192 from the National Science Foundation to the Center.



1. Simon, J. The ultimate resource. Princeton, NJ, USA: Princeton University Press, 1981.

2. Menken J. Demographic-economics relationships and development. In: Graham-Smith F. ed. Population- the complex reality. London: The Royal Society, 1994: 59-70.

3. Keyfitz N. Applied mathematical demography. 2nd ed. New York: Springer-Verlag, 1985.

4. Coale AJ. The growth and structure of human populations: a mathematical investigation. Princeton, NJ, USA: Princeton University Press, 1972.

5. Coale AJ, Demeny P. Vaughan B. Regional model life tables and stable populations. 2nd ed. New York: Academic Press, 1983.

6. Population Reference Bureau. The world population data sheet. Washington, DC: The Population Reference Bureau, 1992.

7. Muhuri PK, Blanc AK, Rutstein SO. Socioeconomic differentials in fertility. Demographic and Health Surveys Comparative Studies no. 13. Calverton, Md, USA: Macro International, 1994.

8. Wood JW. Dynamics of human reproduction: biology, biometry, demography. Hawthorne, NY, USA: Aldine de Gruyter, 1994.

9. Henry L. Some data on natural fertility. Eugenics Q 1961;2:8191.

10. Coale AJ, Trussell TJ. Model fertility schedules: variations in the age structure of childbearing in human populations. Pop Index 1974;40:185-258.

11. Sheps MC, Menken J. Mathematical models of conception and birth. Chicago, 111, USA: University of Chicago Press, 1973.

12. Wood J. Weinstein M. A model of age-specific fecundability. Pop Stud 1988;42:85-113.

13. Wood J. Weinstein M. Heterogeneity in fecundability: the effect of fetal loss. In: Adams J. Lam DA, Hermalin Al, Smouse PE, eds. Convergent issues in genetics and demography. New York: Oxford University Press, 1990:171 188.

14. Salber EJ, Feinleib M, MacMahon B. The duration of postpartum amenorrhea. Am J Epidemiol 1966;82:34758.

15. Chen L, Ahmed S. Gesche M, Mosley WH. A prospective study of birth interval dynamics in rural Bangladesh. Pop Stud 1974;28:277-97.

16. Jain A, Hermalin A, Sun T-H. Lactation and natural fertility. In: Leridon H. Menken J. eds. Natural fertility. Liege, Belgium: Ordina Editions, 1970:149-94.

17. John AM, Menken JA, Chowdhury AKMA. The effects of breastfeeding and nutrition on fecundability in rural Bangladesh: a hazards model analysis. Pop Stud 1987,41:4436.

18. McNeilly AS. Breastfeeding and fertility. In: Gray RH, Leridon H. Spira A, eds. Biomedical and demographic

determinants of reproduction. Oxford: Clarendon Press, 1993:391412.

19. McNeilly AS. Suckling and the control of gonadotropin secretion. In: Kobil E, Neill JD, eds. The physiology of reproduction. New York: Raven Press, 1994:11791212.

20. Bongaarts 1. A framework for analyzing the proximate determinants of fertility. Pop Dev Rev 1978;4:105-32.

21. Bongaarts J. The fertility inhibiting effects of the intermediate fertility variables. Stud Fam Plan 1982;13:17989.

22. Bongaarts J. Potter R Jr. Fertility, biology, and behavior: an analysis of the proximate determinants. New York: Academic Press, 1983.

23. Reinis K. The impact of the proximate determinants of fertility: evaluating Bongearts's and Hobcraft and Little's methods of estimation. Pop Stud 1992;46: 309-26.

24. Mosley WH, Chen LC. An analytical framework for the study of child survival in developing countries. In: Mosley WH, Chen LC, eds. Child survival: strategies for research. Pop Dev Rev 10(suppl);1984:25-48.

25. Van de Perra P. Lepage P. Homsy J. Dabis F. Motherto-infant transmission of human immunodeficiency virus by breast milk: presumed innocent or presumed guilty? Clin Infect Dis 1992;15:502-7.

26. Dunn DT, Newell ML, Ades AK, Peckham CS. Risk of human immunodeficiency virus type I transmission through breastfeeding. Lancet 1992;340:585-8.

27. Adjorlolo-Johnson G. De Cock KM, Ekpini E, Vetter KM, Sibailly T. Brattegaard K, Yavo D, Doorly R. Whitaker JP, Kestens L, Ou C-Y, George JR, Gayle HD. Prospective comparison of mother-to-child transmission of HIV-I and HIV2 in Abidjan, Ivory Coast. JAMA 1994;272:462-6.

28. Nicoll A, Killewo JZJ, Mgone C. HIV and infant feeding practices: epidemiological implications for sub-Saharan African countries. AIDS 1990;4:661-5.

29. World Health Organization. Consensus statement from the WHO/UNICEF consultation on HIV transmission and breastfeeding. Wkl Epidemiol Rec 1992;67(24): 177-9.

30. Cohen B, Trussell J, eds. National Research Council. Preventing and mitigating AIDS in sub-Saharan Africa. Washington, DC: National Academy Press, 1996.

31. Vanlandingham M, Trussell J, Grummer-Strawn L. Contraceptive and health benefits of breastfeeding: a review of recent evidence. Int Fam Plan Perspect 1991; 17:131-6.

32. Holland B. Breastieeding, social variables, and infant mortality: a hazards model analysis of the case of Malaysia. Social Biol 1987;34:78 84.

33. Montgomery M, Richards T, Braun Hl. Child health, breastfeeding, and survival in Malaysia: a random effects logit approach. J Am Stat Assoc 1986;81:297308.

34. Palloni A, Tienda M. The effects of breastfeeding and pace of childbearing on mortality at early ages. Demography 1986;23:31-44.

35. Huffman S. Lamphere BB. Breastfeeding performance and child survival. In: Mosley WH, Chen LC, eds. Child survival: strategies for research. Pop Dev Rev 10(suppl);1984:93-104.

36. Palloni A, Millman S. Effects of inter-birth intervals and breastfeeding on infant and early childhood mortality. Pop Stud 1986;40:215-27.

37. Hobcraft J. McDonald J. Rutstein S. Demographic determinants of infant and early child mortality: a comparative analysis. Pop Stud 1985;39:363-80.

38. Palloni A, Kephart G. The effects of breastfeeding and contraception on the natural rate of increase: are there compensating effects? Pop Stud 1989;43:235-57.

Contents - Previous - Next