Breast-feeding and child-spacing: importance of information collection for public health policy.

Breast-feeding and child-spacing: importance of information collection for public health policy R. Saadehl & D. Benbouzid2 The presence of lactational amenorrhoea cannot be fully relied upon to protect the individual mother against becoming pregnant. Nevertheless, the use of breast-feeding as a birth-spacing mechanism has important implications for global health policy. This article identifies the information that should be collected and examined as a basis for developing guidelines on how to reduce the dual protection afforded by postpartum lactational amenorrhoea and other family planning methods, and discusses when such methods should be introduced. Introduction Many women have little or no access to modern contraceptive techniques and consequently they make little use of family planning methods of fertility regulation. Thus, in cultures where frequent and prolonged breast-feeding is common, postpartum amenorrhoea and suppressed ovulation are frequent- ly the principal mechanisms that ensure adequately spaced pregnancies. Indeed, in many developing countries, more births are still averted this way than by any other single family planning method. Although for the individual mother, lactational amenorrhoea does not provide completely reliable protection against pregnancy, the effectiveness of breast-feeding as a birth-spacing mechanism has important global health policy implications. An in- creased understanding of the reasons for the consider- able differences in the duration of lactational amenor- rhoea between communities and individuals, and of the factors that control lactational infertility, should facilitate development of guidelines for use by health planners and health-care administrators in maximiz- ing the contraceptive effect of breast-feeding. Practical guidance about when mothers should adopt artificial contraceptive methods and how to prevent the adverse effects of some hormonal contraceptives on lactation are particularly important. To contribute to improved understanding of the effectiveness of lactation as a contraceptive method and its relation to child-spacing, the WHO Breast- feeding Data Bank collects and analyses information ' Consultant, Nutrition unit, World Health Organization, CH-1211 Geneva 27, Switzerland. Requests for reprints should be sent to this author. 2 Medical Officer, Nutrition unit, World Health Organization, Geneva Switzerland. Reprint No. 5117 on breast-feeding and its effect on fertility regulation.8 This article first describes the methods used to assess lactational infertility and how the information obtain- ed is incorporated into the data bank. Relevant information gathered from published sources and from studies commissioned by WHO since 1983 is then summarized. Finally, practical health policy implications that are associated with lactation-asso ciated infertility are briefly discussed. Birth interval, fertility, and lactational amenorrhoea Numerous studies indicate that breast-feeding length- ens the interval between pregnancies and thus decreases natural fertility (Fig. 1) (1). In turn, longer intervals between births significantly enhance infant survival and reduce maternal morbidity and mor- tality, especially in developing countries (2-3). In populations where the prevalence of family planning methods is low, birth intervals are largely determined by the frequency and duration of breast- feeding. Where breast-feeding is both frequent and prolonged, birth intervals are increased by an estimated 15-32%. Studies in rural areas of India (Punjab) and China (Province of Taiwan) indicate that birth or pregnancy intervals for women who breast-fed for an average of 1-2 years, and who used no other form of contraception, were 5-10 months longer than those who miscarried or whose infants died shortly after birth (4-5). Studies among Inuits in Alaska (6) and Indians in Punjab show that rates of conception for women who did not use family planning methods, but who continued to breast-feed, remained low for nearly one ' The WHO Breast-feeding Data Bank is managed by the Nutrition unit, World Health Organization, 1211 Geneva 27, Switzerland. Relevant information and inquiries should be sent to this address. Bulletin of the World Health Organization, U (5) 625-631 (1990) (3 World Health Organization 1990 625 R. Saadeh & D. Benbouzid Fig. 1. Illustration of the relationship between the dura- tion of lactational amenorrhoea and birth Interval. 40- 30' 25- 0 520- 15 o v Jordan Mexico Zaire Philippines vo i India Birth interval f Amenorrhoea year postpartum. Of the Indian women who breast- fed for 7 months or longer, 98% had not conceived by the end of the sixth month; however, the cumu- lative rate of conception increased to 13% after one year and continued to rise rapidly thereafter, even though the women in question were still breast-feed- ing at the time of conception (7). Breast-feeding alone is not always responsible for increasing birth intervals. For example, man- datory sexual abstinence during, or even beyond, the normal period of lactation in some traditional societies (e.g., among the Yoruba in Nigeria) can also play an important role. Although this practice is gradually being eroded, it remains important enough to warrant careful consideration in any study of the relation between breast-feeding practices and fertility. The duration of lactational amenorrhoea, which varies from 2 months to 20 months, has the greatest impact on differences in birth intervals, and is an- other method of assessing lactational infertility (8-9). Based on the results of a number of studies that have attempted to establish a relation between lactation and postpartum amenorrhoea (see Table 1), the median duration of amenorrhoea following child- birth appears to be about 7-10 weeks in the absence of lactation. In contrast, the average duration of amenorrhoea among lactating women can be up to 2 years (Table 1 and Fig. 2). The extent to which fertility is reduced during the months prior to, and immediately following, resumption of menstruation can thus be determined and analysed by collecting data, by month postpartum, on the proportion of mothers who are menstruating. Table 1: Median duration of breast-feeding and postpartum amenorrhoea In selected countries Median duration (in months) of: Country Period Sample size Breast-feeding Postpartum amenorrhoea Mexico 1985-86 1131 5.0 3.0 Colombia 1988 702 8.0 3.0 Jordan 1978 769 13.0 3.0 Zaire 1987-88 1102 9.0 4.0 Fiji 1981-84 2660 9.0 5.0 Thailand 1984 - 18.0 6.0 Yemen 1981-84 2216 11.0 7.0 Syrian Arab Republic 1981-84 4025 11.0 7.0 Turkey 1975 515 13.0 7.0 Tunisia 1981-84 3021 14.0 7.0 Botswana 1984 3064 19.5 7.3 Philippines 1981-84 6667 13.0 8.0 Sri Lanka 1986-87 1350 23.0 8.0 Mauritania 1981-84 2447 16.0 9.0 Egypt 1981-84 5667 16.0 9.0 Kenya 1981-84 5679 17.0 10.0 Cote d'lvoire 1981-84 3804 18.0 10.0 Lesotho 1981-84 2348 19.0 10.0 Indonesia 1982 1254 24.0 10.0 Sudan 1981-84 2242 16.0 11.0 China (Province of Taiwan) 1972 1516 18.0 11.0 India 1982 428 20.0 11.0 Haiti 1981-84 1489 15.0 12.0 Pakistan 1986 1098 17.0 12.0 Ghana 1981-84 3335 18.0 12.0 Cameroon 1981-84 4650 18.0 12.0 Benin 1981-84 2803 19.0 12.0 Republic of Korea 1973 746 25.2 13.6 Bangladesh 1981-84 3836 27.0 15.0 Nepal 1976 5954 25.0 18.0 WHO Bulletin OMS. Vol. 68 1990. - 626 Breast-feeding and child-spacing Fig. 2. illustration of the relationship between the dura- tVon of breast-feeding and that of postartum amenor- rhoea. -a 30 i 24 .! iX 18 IJ 12 =6 u 0 Jordan 0 Colombia * - Mxico 0 6 Rep. of Korea hdonsala * 040 * Egypt* 0 1 * Phillppirnz 12 Duraton of postpartum amenofrhoea (months) The return of menstruation and fertility occur much more rapidly among women in industrialized countries than among those from poor segments of populations in developing countries (Table 2) (10, 11). This may be due to differences in breast-feeding practices, since full or exclusive breast-feeding is more closely associated with longer periods of lac- tational amenorrhoea and infertility than is partial or supplemented breast-feeding. It is very important to make a distinction be- tween full and partial breast-feeding. Women who are partially breast-feeding are at higher risk of conceiving than women who are fully breast-feeding. The consensus statement on the use of breast-feeding as a family planning method made by participants at the Bellagio Consensus Conference on Lactational Infertility in 1988 affirmed that the maximum birth- spacing effect of breast-feeding is achieved when a mother "fully" or "nearly fully" breast-feeds and remains amenorrhoeic (12). When these two condi- tions are fulfilled, breast-feeding provides more than 98% protection from becoming pregnant in the first 6 months postpartum. The data in Table 1 need to be analysed in this light. Thus since in Jordan, for example, the median duration of postpartum amen- orrhoea is only 3 months, but the median duration of breast-feeding is 13 months, the vast majority of these women must be introducing feeding supplements to Table 2: Percentage of mothers who had begun menstruating, by age of the child, and whether or not i was breast-fed" Percentage of mothers, with child aged:b 3-4 months 7-8 months 11-12 months Country Groupc BF NBF BF NBF BF NBF Ethiopia A 56 94 - 100 - C 32 - 31 90 38 83 R 8 - 9 - 16 Nigeria A 52 - 58 82 - B 32 - 54 91 50 90 C 7 - 29 - 25 - R 7 - 28 - 39 - Zaire A 26 - 46 - 64 - C 25 - 32 - 44 - R 4 - 20 - 39 - Chile A 23 85 - 98 - 98 C 53 100 64 96 69 97 R 26 80 71 100 56 97 Guatemala A 40 96 - 99 - 98 C 35 94 43 100 50 94 R 11 - 17 - 50 100 India A 56 84 74 97 86 100 B 33 80 62 100 85 97 C 17 - 25 - 33 - R 1 - 12 - 20 - Philippines A 21 98 - 100 - 100 C 28 94 44 91 69 88 R 23 87 41 94 69 95 Hungary All 52 93 63 98 79 99 Sweden All 17 93 42 97 - - ' Source: Contemporary patterns of breast-feeding: report on the WHO Collaborative Study on Breast-feeding. Geneva, World Health Organization, 1981. ° BF= breast-fed; NBF= not breast-fed. c A=urban, economically advantaged; B=urban, middle income; C=urban poor; R=rural. WHO Bulletin OMS. Vol. 68 1990. 627 R. Saadeh & D. Benbouzid their children at a very early age. Without informa- tion on the quality of breast-feeding (whether it is full or partial), such data are of limited value. There is essentially no difference in the contraceptive effect of breast-feeding between industrialized and developing countries when completeness of breast-feeding is taken into account. Several studies have reported a link between use of dietary supplements and shortened periods of postpartum amenorrhoea, which is presumed to result from a reduction in the vigour and frequency of sucking associated with the introduction of other foods. Information on the proportion of infants receiving supplements, by month postpartum, has therefore also been included in the WHO Breast- feeding Data Bank. Breast-feeding patterns (e.g., scheduled versus on-demand feeding, and the prevalence and duration of night feeds) and practices related to the type of complementary foods and when they are introduced, all have an impact on the duration of postpartum amenorrhoea. On the other hand, some evidence suggests that when frequent breast-feeding is main- tained the introduction of complementary foods, par- ticularly after feeding at the breast, has little effect on the resumption of menstruation and return of fer- tility. For breast-feeding to serve optimally as a means of prolonging intervals between pregnancies, full, on-demand breast-feeding is recommended for 4-6 months. The introduction of complementary foods should be delayed as long as possible, but a child's health or growth should not be jeopardized to achieve this (13). Return of menstruation is frequently used as a convenient indirect indicator of resumed ovulation, especially in studies of large populations. Consensus reached at the 1988 Bellagio Conference defined the return of menses as the first vaginal bleeding after the fifty-sixth day postpartum (13), although this need not be similar to the woman's menstrual episodes before the pregnancy. There is a direct relation between lactational amenorrhoea and the duration of lactational infer- tility. Fertility is substantially reduced during the period of lactational amenorrhoea-only 2-10% of breast-feeding mothers conceive before resuming postpartum menstruation. The risk that an amenor- rhoeic lactating woman conceives is related to the time elapsed since delivery. Breast-feeding women who resume menstruating within 6 months of child- birth usually do so before their first postpartum ovulation, whereas those who remain amenorrhoeic for longer than 6 months are most likely to ovulate before their first menstruation. Thus the return of menstruation can be used as an indicator of when to begin using other methods of contraception only during the early postpartum period. Information on variations in the duration of amenorrhoea in rural and urban areas of selected industrialized and developing countries, showing the proportions of women who began breast-feeding and for how long, and when they resumed their menses at specified intervals after birth, could be useful for identifying those women who are at risk of becoming pregnant and for understanding better the relation between lactation and amenorrhoea. Assessment of the dura- tion of lactational infertility can also be based on the detection of ovulation. Various investigations have attempted to define the timing of ovulation after childbirth using endometrial biopsy and by studying the basal body temperature, cervical mucus, or vaginal cytology as indirect markers of ovulation. Other studies have used more direct indices such as estimations of hormone levels in blood, saliva, or urine, or ultrasonic scans of the ovaries in combina- tion with endocrine measurements. The frequency of ovulation in the cycle before the end of lactational amenorrhoea varies from 14% to 75%, which may be partly due to different suckling patterns. As discussed above, the longer the first menses are delayed during lactation, the more likely the first cycle will be ovulatory (14-17). Nevertheless, the duration of lac- tation has a marked impact on the onset of ovula- tion; the longer lactation continues, the longer the anovulatory period (18). In general, after the sixth month postpartum, there is an increased likelihood that fertility precedes the first vaginal bleeding and initiation of com- plementary feeding (13). At this stage the protection against another pregnancy provided by lactational infertility falls below that afforded by other con- traceptive methods. Other factors that Influence lactational Infertility and breast-feeding behaviour Age and parity have long been regarded as important factors that influence lactational infertility and breast-feeding behaviour, since the period of lac- tational amenorrhoea tends to be longer for older and multiparous than for younger and primiparous women. The two factors are nearly always strongly correlated; "age of mother" has therefore been included in the WHO Breast-feeding Data Bank in order to test this relation. Other possible factors that affect lactational infertility, e.g., maternal nutritional status, are independent of the frequency and duration of breast- feeding. Except in extreme cases of deprivation and famine, malnutrition has only a minor impact on the WHO Bulletin OMS. Vol. 68 1990.628 Breast-feeding and child-spacing quality and quantity of breast milk.b It has been suggested that the return of menstruation during lactation may be delayed in undernourished women. Results from India suggest that, among women of the same socioeconomic status who have similar breast- feeding patterns, lactational amenorrhoea is longer for those of low body weight for any given duration of lactation (19,20). Studies in the Gambia on the effects of food supplements given to lactating women showed that, even though supplementation did not influence the quantity of milk, there was a significant drop in plasma prolactin levels and a more rapid return of menstruation (21, 22). It is difficult, however, to draw definite conclusions from these results because of the potential effects of important confounding factors such as differences in socio- economic status, working conditions, breast-feeding practices, and the quality and number of health and nutrition interventions. As mentioned previously, the type and timing of complementary feeding has an important impact on suckling behaviour, and is generally associated with the following: a reduction in the number, intensity and duration of suckling episodes, resulting in a relaxation of the neural inhibition of the hypo- thalamus; a decline in plasma prolactin levels; and, inevitably, an earlier return of ovulation and men- struation. It has been proposed that if additional foods are introduced into the baby's diet gradually and over an extended period of time while breast- feeding is continued, the latter may still exert a significant antifertility effect for one year or longer; the main point is that breast-feeding should not be reduced and other foods should not replace breast- feeding (12). Duration of breast-feeding and prevalence of contraception The duration of breast-feeding accounts for 96% of the variation in the length of time of postpartum amenorrhoea. Lactation has its greatest impact on postpartum amenorrhoea when breast-feeding is con- tinued for 6 to 18 months (2). Several epidemiological studies have documen- ted the prevalence and duration of breast-feeding, which is usually shorter among young, affluent, urban, educated women than among their older, poorer, rural, and less-educated counterparts (10). A decline in the prevalence and duration of breast- feeding has been reported in several countries in Asia b An annotated bibliography of the effect of maternal nutritional status on the duration of lactation. Unpublished WHO (Nutrition Unit) listing, 1988. and Latin America, although in Malaysia, Singapore, and Thailand the situation appears to be stabilizing (23,24). In developing countries any reduction in the present high incidence, frequency, and duration of breast-feeding is likely to increase fertility, especially if the use of other family planning methods fails to increase correspondingly. In developing countries where contraceptive use is less than 10% and the fertility rate is high, as was the case in, for example, Bangladesh in the late 1970s (25,26), any substantial decline in breast-feeding would require a more than fivefold increase in contraceptive use to prevent the already high fertility rates from rising further. The situation in Kenya is also illustrative; declines in the duration of breast-feeding and of postpartum abstin- ence, combined with infrequent use of artificial meth- ods of contraception, have resulted in one of the world's highest population growth rates. Currently, the population is doubling approximately every 17 years. Health policy implications The interplay between lactational infertility and con- traceptive practice needs to be considered before their ultimate combined effect on fertility can be determined. The effect of breast-feeding on fertility control and child-spacing has to be considered in the context of all the other factors that are relevant both to a particular population and to individuals. Every effort should be made to gather and analyse complete information on all the factors that affect breast-feeding behaviour and lactational in- fertility: the mother's age, the median duration of breast-feeding and of postpartum amenorrhoea, the length of birth intervals, the prevalence of con- traceptive use, the proportion of women menstrua- tion (by month postpartum), and the proportion of infants given complementary foods (by age). Such information could improve understanding about how breast-feeding influences fertility and facilitate the preparation of policy guidelines on breast-feeding and child-spacing. It would be particularly useful, for example, to analyse the interaction between the age and the proportion of lactating women who have the greatest risk of conception by month postpartum, based on data on the median duration of breast- feeding and postpartum amenorrhoea for a given community or group.c It is not sufficient to demonstrate a direct relation c Breast-feeding and fertility: a simplified methodology for com- munity-based calculation of the proportion of mothers at risk of conception by breast-feeding status. Unpublished WHO document MCH/85. 15/Rev.87. WHO Bulletin OMS. Vol. 68 1990. 629 R. Saadeh & D. Benbouzid between breast-feeding and lactational infertility and hence child-spacing, since it is also important to determine when to introduce other family planning methods and how to minimize the incidence of the "dual coverage" provided by lactational amenor- rhoea and other contraceptive measures. By delaying the introduction of such measures, the extent of overlap between the two is reduced and the con- traceptive benefit of exclusive breast-feeding is increased. On the other hand, if other contraceptive measures are introduced too late, there is an increased risk of unwanted pregnancy with all that this implies for the health and nutritional status of mothers and children alike. Breast-feeding has an important untapped potential as a natural family planning method because of its effect on child-spacing. This is par- ticularly true for mothers who lack access to artificial family planning methods or who are dissatisfied with their inconvenience or potential side-effects. Guidelines are therefore needed on the most appropriate timing of the introduction of other family planning methods in specific environments, based on prevailing breast-feeding patterns and the duration of lactational amenorrhoea. Acknowledgements We acknowledge the assistance of Dr P. Van Look in preparing this article, Mr J. Akr6 in editing it, and Mr J. Gorstein in analysing the data. R6sum6 Allaltement maternel et espacement des nalssances: du recuell de l'lnformatlon A la politique de sant6 publique L'amenorrh6e liee & la lactation ne doit pas etre consid6r6e au niveau individuel comme une garantie contre une autre grossesse. N6anmoins, I'allaitement au sein en tant que m6canisme d'es- pacement des naissances a d'importantes implica- tions dans une politique globale de sante. Cet article examine l'information qu'il faut recueillir et analyser pour 6laborer des directives pertinentes, en particulier celles qui indiquent comment r6duire la double couverture apport6e par l'am6norrh6e du post-partum et la contraception artificielle, et quand introduire la seconde. Cet article d6crit les m6thodes utilis6es pour 6valuer l'infecondite liMe a la lactation et analyse les donnees disponibles sur l'allaitement maternel et sur la r6gulation de la fecondite. 630 De nombreuses etudes ont montre que l'allaitement maternel prolonge l'intervalle entre les grossesses et reduit en cons6quence la f6con- dit6 naturelle. L'allongement de l'intervalle entre les naissances va de pair avec I'augmentation des chances de survie de l'enfant et la reduction de la morbidit6 et de la mortalite maternelles, sp6ciale- ment dans les pays en d6veloppement. La duree de l'amenorrhee liee a la lactation, principal facteur responsable des diff6rences dans les intervalles entre les naissances puisqu'elle peut etre de 2 a 20 mois, represente un autre moyen d'appr6cier l'infecondite liee a la lactation. Calculee a partir d'un certain nombre d'etudes ayant essaye d'etablir la relation entre la lactation et l'amenorrh6e du post-partum, la dur6e mediane de l'amenorrhee qui suit I'accouchement semble etre de 7 a 10 semaines en l'absence de lactation. En revanche, sa dur6e moyenne chez les femmes allaitantes peut aller jusqu'a deux ans. La mesure dans laquelle la fertilite est diminuee durant les mois pr6c6dant ou suivant immediatement le retour des regles peut donc etre ainsi determinee et analysee en recueillant les donnees sur la proportion de meres ayant leurs r6gles par mois du post-partum. Pour que l'allaitement maternel puisse servir au mieux de moyen de prolonger l'intervalle entre les naissances, il doit Otre exclusif et donne a la demande pendant les 4 a 6 premiers mois. L'in- troduction d'aliments de complement doit gtre retardee aussi longtemps que possible, sans com- promettre pour autant la sante ou la croissance de l'enfant. Le retour des regles est fr6quemment utilis6 comme indicateur indirect de la reprise de l'ovula- tion, sp6cialement dans les enquetes sur de larges populations. Le consensus adopte lors de la con- f6rence de Bellagio en 1988 d6finit le retour des r6gles comme le premier saignement vaginal apres le cinquante-sixieme jour du post-partum, meme s'il n'est pas n6cessairement identique aux saignements menstruels de la femme avant la grossesse. La duree de l'allaitement maternel explique 98% des variations de la duree de l'amenorrhee du post-partum. La lactation a le plus d'impact sur l'am6norrhee quand la duree de l'allaitement se situe entre 6 et 18 mois. Plusieurs enqu6tes 6pid6miologiques ont documente la pr6valence et la duree de l'allaitement. Cette dur6e est habit- uellement plus courte chez les femmes jeunes, ayant fait des etudes et de milieux aises urbains que chez les femmes plus Ag6es, pauvres et sans instruction des zones rurales. Un declin de la dur6e et de la prevalence de l'allaitement maternel WHO Bulletin OMS. Vol. 68 1990. Breast-feeding and child-spacing a ete constate dans plusieurs pays d'Asie et d'Amerique latine, alors que dans quelques rares pays la situation semble se stabiliser. L'allaitement maternel represente un potentiel non exploit6 en tant que m6thode naturelle de planification familiale grace a son r6le dans l'espacement des naissances. Ceci est par- ticuli6rement vrai pour les meres qui ont peu acc6s A la contraception artificielle ou qui ne sont pas satisfaites du fait des inconvenients et des effets secondaires possibles de la methode. Des directives sont donc necessaires sur le moment le plus approprie pour l'introduction d'autres methodes de planification familiale dans des environnements specifiques, en se basant sur les modes d'allaitement maternel et la duree de l'amenorrhee liee a la lactation qui y prevalent. References 1. van Glnneken, J.K. The impact of prolonged breast- feeding on birth intervals and on postpartum amenor- rhoea. In: Mosley, W.H., ed. Nutrition and human reproduction. New York, Plenum Press, 1978, pp. 179-195. 2. Thapa, S. et al. Breast-feeding, birth-spacing and their effects on child survival. Nature, 335: 679-682 (1988). 3. Short, R. Breast-feeding. Scientific American, 250(4): 35-41 (1984). 4. Jain, A.K. Pregnancy outcome and the time required for next conception. Population studies, 23: 421-433 (1969). 5. Pofter, R.G. et al. Applications of field studies to research on the physiology of human reproduction: lactation and its effects upon birth intervals in eleven Punjab villages, India. In: Sheps, M. & Ridley, J., ed. Public health and population change. Pittsburgh, University of Pittsburgh Press, 1965, pp. 377-399. 6. Berman, M. et al. Effect of breast-feeding on post- partum menstruation, ovulation and pregnancy in Alaskan Eskimos. American journal of obstetrics and gynecology, 114: 524-534 (1972). 7. van Glnneken, J.K. Prolonged breast-feeding as a birth-spacing method. Studies in family planning, 5: 201-206 (1974). 8. Huffman, S. Maternal and child nutritional status: its association with the risk of pregnancy. Social science and medicine, 17: 1529-1540 (1983). 9. Jelliffe, D. & Jelliffe, P. Lactation amenorrhea: an important present-day component of family planning programmes. Journal of tropical pediatrics, 31: 240- 241 (1985). 10. Contemporary patterns of breast-feeding: report on the WHO Collaborative Study on Breast-feeding. Geneva, World Health Organization, 1981. 11. Breast-feeding, fertility and family planning. Popula- tion reports: series J, 24: 525-591 (1981). 12. Kennedy, K. et al. Consensus statement on the use of breast-feeding as a family planning method. Con- traception, 39: 477-496 (1989). 13. Consensus statement. Breast-feeding as a family planning method. Lancet, 2: 1204-1205 (1988). 14. Perez, A. et al. Timing and sequence of resuming ovulation and menstruation after childbirth. Popula- tion studies, 25: 491-503 (1971). 15. Howle, P.W. et al. Fertility after childbirth: effect of breast-feeding on ovulation and menstruation. British medical journal, 283: 757-759 (1981). 16. Howle, P.W. & McNellly, A. Effect of breast-feeding patterns on human birth intervals. Journal of reproduction and fertility, 65: 545-557 (1982). 17. Jain, A. & Bongaarts, J. Breast-feeding patterns, correlates and fertility effects. Studies on family plan- ning, 12(3): 79-99 (1981). 18. Perez, A. et al. First ovulation after childbirth: the effect of breast-feeding. American journal of obstet- rics and gynecology, 14: 1041-1047 (1972). 19. Prema, K. et al. Nutrition fertility interactions in lactating women of low income groups. British jour- nal of nutrition, 45: 461-467 (1981). 20. Bongaarts, J. Does malnutrition affect fecundity? A summary of evidence. Science, 208: 564-569 (1980). 21. Lunn, P.G. et al. Influence of maternal diet on plasma-prolactin levels during lactation. Lancet, 1: 623-625 (1980). 22. Lunn, P.G. et al. Maternal nutrition and lactational amenorrhoea. Lancet, 1: 1428-1429 (1981). 23. World Health OrganizatIon. The prevalence and duration of breast-feeding: a critical review of avail- able information. World health statistics quarterly, 35: 92-116 (1982). 24. World Health Organization. The prevalence and duration of breast-feeding: updated information, 1980-89. Part I. Weekly epidemiological record, 64(42): 321-324 (1989). Part II. Weekly epidemio- logical record, 64(43): 331-335 (1989). 25. Losthaeghe, P. et al. Compensating changes in inter- mediate fertility variables and the onset of marital fertility transition. (International Union for the Scien- tific Study of Population). In: Proceedings of the International Population Conference, Manila, 1981, pp. 71-95. 26. Bongaarts, J. The fertility-inhibiting effects of the intermediate fertility variables. Studies in family planning, 13(6/7): 179-189 (1982). WHO Bulletin OMS. Vol. 68 1990. 631