Contents - Previous - Next

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

6. The role of prolactin on the contraceptive effect of lactation, and the influence of breast-feeding practices and of maternal dietary status

General Summary of Hormonal Changes
Prolactin, Breast-feeding Frequency, and Supplementary Feeding in Scotland
The Effect of Maternal Diet on the Endocrinology of Lactational Infertility in the Gambia


6.1. It is clear that a substantial number of mothers who breast-feed experience a period of infertility and amenorrhoea. It is also clear that many women, in particular in Western societies, resume menstrual cyclicity while still breast-feedingl1). It is of major importance that the factors influencing these responses are clarified.

Past investigations on the changes in reproductive hormones post-partum have been basically of three kinds: la) cross-sectional studies in which single samples of blood have been collected from large numbers of women at various stages post-partum; (b) detailed studies in a few subjects; and (c) cross-sectional studies of pituitary responsiveness to hypothalamic-releasing factor stimulation such as a gonadotrophin-releasing hormone (GnRH). Unfortunately, in all studies of the first two types, the duration of suckling and whether supplements were also given to the baby have not been recorded, and in only a few has some estimate been made of the number of breastfeeding episodes. Thus, only a very general picture of the hormonal pattern associated with lactational infertility has emerged.


General Summary of Hormonal Changes(1)

6.2. Breast-feeding is associated with high plasma concentrations of prolactin, at least at the onset of lactation, the levels correlating to some extent with the number of suckling episodes 12). The prolactin response to suckling declines with time post-partum, but if suckling frequency is maintained at a high level basal levels may well remain above normal for 18 months or more (2, 3) (see fig. 15).

Blood levels of follicle-stimulating hormone IFSH) are necessary for ovarian follicular growth and development, and quickly return to normal menstrual cycle levels within a week or two post-partum. At no stage during lactational amenorrhoea do FSH levels appear to be inadequate for ovarian function. Pituitary levels of luteinizing hormone (LH) are very low immediately postpartum, but by 15 to 20 days blood levels have increased significantly and remain throughout lactation on the lower side of normal.

During lactational amenorrhoea in fully breast-feeding women, the response of LH to GnRH stimulation is diminished, while the FSH response is normal. In the same situation, women fail to show a positive feed-back response, with an increase in LH and FSH to exogenously administered oestrogen, whereas they show an enhanced negative feed-back effect with prolonged suppression of LH levels in contrast to normally cyclic women. In lactational amenorrhoea, ovarian oestrogen and progesterone secretion is below normal, and is equivalent to that seen in post-menopausal women in spite of normal levels of FSH.

Complete weaning results in an immediate drop in the blood levels of prolactin and an increase in blood levels of LH and oestradiol, indicating a prompt resumption of ovarian activity. Actual ovulation usually occurs within 14 to 30 days. These results suggest that a maintained suckling stimulus, and the associated hyperprolactinaemia, suppress LH but not FSH post-partum and lead to both a failure of ovarian follicular development and lactational amenorrhoea. There is also some information that if ovulation does occur it results in a deficient corpus luteum function.


Prolactin, Breast-feeding Frequency, and Supplementary Feeding in Scotland

6.3. These interrelationships have recently been studied on a prospective, longitudinal basis among a small group of mothers living in Edinburgh, Scotland 14, 5). The investigation was planned in such a way that account could be taken of the frequency and duration of breast-feeding and the introduction of supplementary food to the baby. This has produced valuable information on the return of fertility post-partum in relation to changes in the pattern of breastfeeding. The essential results are summarized in fig. 16 (see FIG. 16. Comparision of Suckling Episodes, Suckling Duration, Supplementary Feeds, and Basal Prolactin IMean Concentrations + SE) in Breast-feeding MothersWho Ovulated or Continued to Suppress Ovulation after Introduction of Supplementary Feeds. IPoints are significantly different from * onwards) (Source: ref. 4)). Plasma prolactin concentrations remained significantly above normal levels until the frequency of feeding and the total duration of lactation per day fell. Thereafter all three parameters declined approximately in parallel and the changes were highly correlated. They also showed an inverse correlation with the introduction of supplementary food, which was introduced between three and 24 weeks postpartum. Before this no mother had ovulated, although two out of 27 had some unsustained follicular activity.

Following the introduction of supplementary food there was a progressive increase in the number of mothers showing evidence of ovarian activity, and within 16 weeks 20 out of 27 had follicular activity and 14 out of 27 had ovulated. While there was no difference in the suckling patterns or prolactin levels before the introduction of supplementary food, the non-ovulatary group had subsequently maintained both frequency and the daily duration of feeding at a higher level; they also introduced infant supplements less abruptly than did the ovulating group. Mean basal prolactin levels remained above the nonpregnant range in the non-ovulating group for at least 16 weeks after the onset of infant supplementary feeding. Values were significantly lower in the ovulating women.

6.4. Half of the women studied failed to ovulate throughout breast-feeding, and even in the remainder corpus luteum function, in terms of levels and duration of secretion of progesterone, was deficient. In some cases the increase in plasma progesterone concentrations was so minimal it was not clear whether ovulation had taken place or not, even though menstruation occurred. Four women who were apparently ovulating and wished to become pregnant failed to do so, presumably because of an inadequate luteal phase.

6.5. In all women FSH levels returned to normal within a short period postpartum, while LH levels remained suppressed until just before the increase in ovarian follicular development and oestrogen secretion.

Other studies have indicated that during lactational amenorrhoea, low LH levels are associated with an absence or reduction in the frequency of pulsatile LH discharges which are necessary for ovarian steroid secretion (6) This is related to a maintained suckling frequency of adequate duration and to plasma concentrations of prolactin. When breast-feeding declines with the associated decrease in prolactin levels, LH pulsatility returns to normal and ovarian follicular development resumes. It seems, however, that there are a few cases in which LH levels return to normal but the ovaries do not respond to this stimulation.

6.6. It is obvious that full breast-feeding, even in well-nourished women, inhibits ovarian activity completely for up to six months or more, but in undernourished mothers, who tend to feed their children much more often, this inhibition can be for two years even though infant supplementation has been introduced from three to six months. It is of interest that similar observations have been made on a number of Perth mothers who have breast-fed for two to three years. However, prolonged lactational amenorrhoea is not always assured even in mothers with similar patterns of breast-feeding. One critical factor would appear to be the maintenance of a high suckling stimulus both in terms of frequency and duration. The introduction of additional sources of food inevitably helps to satiate the baby's appetite, setting in motion a cascade of mechanisms that result in a release of the inhibition of ovarian activity. The rapidity of this response would appear to be greater in the industrialized countries, perhaps because "traditional" infant foods less adequately satisfy the child. The mechanism of suppression appears to be a decrease in the normal episodic secretion of LH caused by the neural suckling stimulus and/or the raised levels of prolactin, together with an increased sensitivity to the inhibitory effect of LH release of ovarian steroids, especially oestrogen.

6.7. It is obvious, however, that lactational infertility cannot be explained on the basis of prolactin alone. When ovarian activity resumes during lactation, ovulation is frequently followed by the formation of inadequate corpora lutea which are probably not compatible with pregnancy, thus menses is not a reliable guide to fecundity. The mechanisms linking neural suckling and the brain, the increase in prolactin, and the suppression of LH secretion are not known, and a truly adequate understanding of lactational infertility depends on filling these gaps in out knowledge.


The Effect of Maternal Diet on the Endocrinology of Lactational Infertility in the Gambia

6.8. The key information suggesting that there might be a link between maternal dietary adequacy and the duration of infertility in lactating mothers has already been discussed (section 4.12). The basic hypothesis is that the undernourished mother has more difficulty in synthesizing milk, thus the child has to suck more intensively and usually more frequently and for longer periods of time in order to obtain an adequate amount of milk. This would produce an intense neural suckling stimulus that would inhibit LH release and ovulation either through prolactin, by direct neural action, or both. Recent work from the Gambia (3, 7), in which the diet of a large group of rural mothers was considerably enhanced by the provision of a biscuit-based supplement (see section 3.26), appears to have substantiated this basic hypothesis. As figure 14 shows, plasma prolactin concentrations were significantly lower at practically all stages of lactation in women who received the supplement.

6.9. It is not possible to explain exactly why such a dramatic effect on prolactin was produced; the total volume of milk the child subsequently consumed did not improve, nor did the actual frequency of feeding (7). Unfortunately the time the mother was forced to spend feeding her child was not measured, nor could any objective assessment be made of the intensity of that feeding. it was suggested, however, that the increased levels of prolactin reflected the ease with which the mother was able to produce the milk (3).

6.10. The desperately low levels of dietary-energy and nutrient intake of women in the Third World, including the Gambia, has already been discussed in section 2.2. Under such circumstances it is vital that the nutrients that are absorbed are actively channelled towards the tissues where they are most urgently required. Considering the structural and functional similarities between prolactin and growth hormone, it is not unreasonable to suggest that this could be at least part of the functional role of the elevated prolactin concentrations found in undernourished mothers. If prolactin does act by ensuring that an adequate supply of nutrients is made available to the breast to preferentially maintain milk synthesis, then it is probable that the amount of hormone required would need to be higher when maternal diet is deficient than when it is good.

6.11. There is another potential role for elevated prolactin levels. Malnourished mothers during pregnancy are unable to accumulate large stores of fat as do their well-nourished counterparts. Since this fat is an important source of milk energy, it is not unreasonable to suppose that a stronger stimulus, such as could be provided by persistently high prolactin levels, will be necessary to ensure the mobilization of what little fat there is. Although this idea is hypothetical, there is evidence from rats that prolactin does act directly in this way (8). The proposed mechanism is also very similar to the well-documented human growth-hormone-induced lipolysis and enhancement of milk production that occurs in dairy cattle (9). The basic mechanisms are summarized in figure 17 (see FIG. 17. Nutritionally Relevant Hormonal Interactions Relating to Milk Production (Lunn, unpublished data)).

6.12. In order to test whether or not reduced concentrations of prolactin were leading to a reduction in lactational infertility, plasma concentrations of oestradiol and progesterone were also measured, and the values are shown in figure 18 (see FIG. 18. The Effect of Dietary Supplementation on Plasma Oestradiol and Progesterone Concentrations in Lactating Women (Source: ref. 9))(10). The sharp increase in plasma oestradiol concentrations, indicative of the time when the women were resuming their menstrual cycles, occurred about 14 weeks sooner in the supplemented mothers. Even more dramatic were the differences in plasma progesterone. Mean values for the supplemented women started to show some increase by 54 weeks postpartum and continued to rise after this time towards the value found in nonpregnant, non-lactating women. In contrast, there was little evidence of a general return of fertility in the unsupplemented mothers even at 78 weeks.

These biochemical data were supported by an analysis of the proportion of lactating women becoming pregnant again by 18 months. In the unsupplemented mothers it was 19 per cent, and in the supplemented ones 33 per cent. This was in spite of a taboo against sex while lactating. Clearly this rule was not being strictly adhered to, but the decrease in birth interval might have been even more pronounced in its absence.

6.13. The results of this controlled intervention study are totally compatible with the epidemiological findings of Chavez in Mexico ( 11), of Delgado in Guatemala(12), and Vis's group in Zaire (13, and H.L. Vis, personal communication), and also of Prema in India(14) but are at variance with the experiences of Chowdhury (15) and of Bongaarts (16), whose statistical predictions led them to conclude that the effect of diet on birth-spacing would be insignificant. The size of any effect, however, presumably depends on the degree of difference in dietary intake.

In the Gambian study, dietary energy intake was boosted to a level customarily encountered only in much more affluent women in the Western world, and this could be why the response was more dramatic. The contraceptive effect of prolactin can be looked upon as a protective mechanism, which, when food is less readily available, prevents a new pregnancy occurring too precipitously, which would further exacerbate the nutritional stress in the mother as well as her child. With adequate nutrition, such a barrier to reproduction may not be such a biological necessity.



6.14. The message seems clear that there is reasonably good substantive information indicating that improving maternal diet without concurrently introducing some artificial form of contraception may shorten the birth interval and consequently increase birth-rate. It is of crucial importance that this effect is studied directly in other developing countries in addition to the Gambia, so that its public-health significance can be judged with complete confidence.



  1. A.S. McNeilly, "Effects of Lactation on Fertility," Br. Med. Bull. 35: 151-154 (1979).
  2. P. Delvoye, M. Demaegd, J. Delogne-Desnoeck, and C. Robyn, "The Influence of the Frequency of Nursing and of Previous Lactation Experience on Serum Prolactin in Lactating Mothers," J. Biosoc. Sci., 9: 447-451 (1977).
  3. P.G. Lunn, A.M. Prentice, S. Austin, and R.G. Whitehead, "Influence of Maternal Diet on Plasma-prolactin Levels during Lactation," Lancet, i: 623-625 119801.
  4. P.W. Howie, A.S. McNeilly, M.J. Houston, A. Cook, and H. Boyle, "Effect of Supplementary Food on Suckling Patterns and Ovarian Activity during Lactation," Brit. Med. J., ii: 757-759 (1981).
  5. A.S. McNeilly, P.W. Howie, and M.J. Houston, "Relationship between Feeding Patterns, Prolactin and the Resumption of Ovulation Post-partum," in G.l. Zatuchni, M.H. Labbok, and J.J. Sciarra, eds., Research Frontiers in Fertility Regulation (Harper & Row, Mexico City, 1980), pp. 102-116.
  6. P.T. Baird, A.S. McNeilly, R.S. Sawers, and R.M. Sharpe, "Failure of Oestrogen-induced Discharge of Luteinising Hormone in Lactating Women," J. Clin. Endocrinal Metab., 49: 500-506 (1979).
  7. A.M. Prentice, R.G. Whitehead, S.B. Roberts, A.A. Paul, M. Watkinson, A. Prentice, and A.A. Watkinson, "Dietary Supplementation of Gambian Nursing Mothers and Lactational Performance," Lancet, ii: 886-888 (1980).
  8. D.H. Williamson, "Integration of Metabolism of Tissues in the Lactating Rat," FEBS Letters, 117, Supplement K93K105 (1980).
  9. J.A. Bines and l.C. Hart, "Hormonal Regulation of the Partition of Energy between Milk and Body Tissue in Adult Cattle," Proc. Nutr. Soc., 37; 281-287 (1978).
  10. P.G. Lunn, M. Watkinson, A.M. Prentice, P. Morrell, S. Austin, and R.G. Whitehead, "Maternal Nutrition and Lactational Amenorrhoea," Lancet, i:1428-1429 (1981).
  11. A. Chavez and C. Martinez, "Nutrition and Development of Infants from Poor Rural Areas. III: Maternal Nutrition and its Consequences on Fertility," Nutr. Rep. Internat., 7: 1-8 (1973).
  12. H. Delgado, A. Lechtig. E. Brineman, R. Martorell, C. Yarbrough, and R.E. Klein, "Nutrition and Birth Interval Components: The Guatemalan Experiences," in W.H. Mosley, ea., Nutrition and Human Reproduction (Plenum, New York, 1979), pp. 385-399.
  13. M. Carael, "Relations between Birth Intervals and Nutrition in Three Central African Populations (Zaire)," in W.H. Mosley, ea., Nutrition and Human Reproduction (Plenum, New York, 1979), pp. 365-384.
  14. K. Prema, A. Nadamuni Naidu, S. Neelakumari, and B.A. Ramalakshmi, "Nutrition-Fertility Interaction in Lactating Women of Low Income Groups," B.J. Nutr., 45: 461-467 (1981).
  15. A.K.M.A. Chowdhury, "Effect of Maternal Nutrition on Fertility in Rural Bangladesh," in W.H. Mosley, ea., Nutrition and Human Reproduction (Plenum, New York, 1979), pp. 401-410.
  16. J. Bongaarts, "Does Malnutrition Affect Fertility? A Summary of Evidence," Science, 208: 564-569 (1980).


Contents - Previous - Next