Contents - Previous - Next

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

Vitamin A status of lactating women

Along with ensuring adequate vitamin A levels in breast milk for the infant, improving the vitamin A status of lactating women may benefit maternal health directly. Recent studies in human subjects indicate that an adequate dietary intake of retinol and carotene helps to prevent breast and cervical cancer [32-37]. Studies in animals indicate that vitamin A protects females from infections. For example, cows with mastitis had significantly lower serum retinol levels postpartum than cows without mastitis [38]. Vitamin A supplementation protected rats and mice against experimentally induced uterine and mammary gland infections [39, 40]. Supplementation with vitamin A and -carotene improved mammary health in dairy cows around dry-off (weaning) [41].

Severe deficiency of vitamin A causes infertility or impaired reproduction in all vertebrate species that have been studied [42, 43]. In animals, the oestrus cycle is disrupted and the vagina becomes permanently keratinized. However, rather than ovarian dysfunction or failure of fertilization or implantation, the most characteristic results are foetal resorption, stillbirths, and congenital malformations [42]. Foetal demise apparently results from placental necrosis.

TABLE 5. Weighted average daily dietary intake of vitamin A by unsupplemented lactating women


No. of groups

No. of subjects

Intake (RE/day)

Developed countries




Developing countries




Weighted averages are derived from data summarized in ref.13.


Dietary intake

Few studies since 1975 report the dietary vitamin A intake of lactating women (table 5). The weighted average daily intake by unsupplemented women in developing countries (660 RE per day) is less than half that of women in developed countries (1,540 RE per day). This is lower than the 850 RE per day safe level for lactating women but above the basal requirement of 450 RE per day recommended by the FAD/WHO).

Whereas several studies from developed countries [44-47] reported mean daily vitamin A intakes that were lower than the 1,200-1,300 RE recommended by the US government [16], only one study of Canadian women [48] reported a mean daily intake below the FAD/WHO recommendation of 850 RE. The reason for this particularly low figure is not apparent.

Most of the mean intake figures from developing countries, with a few exceptions [21, 49], were well below 850 RE. Of interest is a study showing a substantial increase in dietary intake of vitamin A over time after delivery [50]. As might be expected, intake during the 24 hours before delivery was low, but it rose progressively until six months postpartum in both groups studied, and continued to increase until nine months in one group. It is unlikely that the increase was related to seasonal availability of fruit and vegetable sources, because of the length of the study. It is possible that the lower vitamin A intakes in the early postpartum period resulted from cultural food restrictions during this time, although this was not discussed in the paper.

TABLE 6 Weighted average biochemical measures of vitamin A nutritional status for unsupplemented lactating women


No. of groups

No. of subjects

Average serum value

Retinol (g/L)

Carotene (g/L)

RBP (mg/L)

Developed countries














Developing countries














Weighted averages are derived from data summarized in ref. 13; values from delivery were omitted.


Biochemical measurements

Studies during the last 40 years reporting maternal serum (or plasma) levels of retinol, carotene, and RBP are summarized in table 6. Overall, the average serum levels of retinol and RBP in unsupplemented lactating women from developing countries were approximately 70% those of their counterparts in developed countries. The lower levels may reflect chronic poor dietary intake in developing countries, as suggested by the FAO [15]. Despite these large differences, the average levels of both components in both developing and developed countries were in the adequate range.

On the other hand, the average serum carotene level of lactating women from developing countries was about 35% above the average of women from developed countries. This may reflect the fact that most of their vitamin A intake was from carotenoids and/or be related to an insufficient intake of protein that is necessary to transform carotene into active vitamin A [51].


Serum retinol

Serum (plasma) retinol levels decrease during pregnancy and increase to non-pregnant levels during the postpartum period [18, 42, 52, 53]. The increase to normal values occurs within 24-48 hours postpartum [50, 52]. The concentration of retinol in the serum of lactating women is not significantly different from that in non-pregnant, non-lactating women but is higher than that at term [52]. After the increase, there appears to be a slow decrease during lactation [54-56].

Mean serum retinol levels in lactating women in developed countries, with or without vitamin A supplements, were in the adequate range (>300 g/L) in all studies reviewed, with the exception of one group evaluated at delivery in Finland [27]. The marginal levels (296 g/L) were probably related to the known decrease during pregnancy. On the other hand, several studies of lactating women in developing countries [50, 53, 57, 58] reported mean serum levels of retinol in the deficient range (<200 g/L) or the marginal range (200-300 g/L).


Serum carotene

The effect of parturition on serum carotene appears from some reports to be opposite to that of retinol. The concentration of carotenoids is reported to increase late in gestation [53] and to fall during the first three to six weeks postpartum [44, 52, 59]. However, this does not appear to be a consistent pattern [60].

In studies of serum carotene levels in lactating women from developed countries, mean values ranged from 275 to 1,405 g/L, and those from developing countries from 718 to 1,870 g/L. Due to the small number of reports, it was not possible to calculate weighted averages at different times after delivery.


Serum retinol-binding protein

All mean serum RBP levels in lactating women from developed countries were in the recommended adequate ranges (>26 mg/L [22] or >=30 mg/L [61]). However, two studies [50, 62] reported mean levels of 26 mg/L or less, but neither observed levels in the deficient range ( <22 mg/L).

Plasma levels in Swedish mothers were significantly higher than in both privileged and non-privileged Ethiopian mothers [62]. In the latter group, 14 of 81 manifested values below 20 mg/L, and levels were significantly subnormal in both groups of Ethiopian mothers. This seems surprising since there was no indication that on average the vitamin A intake of privileged mothers was unsatisfactory, and there was even less reason to suspect other dietary insufficiencies in this group (e.g., protein deficiency) that may exert an influence on the RBP level. On the other hand, the low values in the privileged mothers were consistent with the simultaneously low vitamin A concentration of their milk [13].


Vitamin A in breast milk

Retinol derived from the circulating RBP-retinol complex is transferred from blood to milk. Most of it is re-esterified in the mammary glands and occurs as retinyl esters in milk [42, 63]. Whereas most of the vitamin A activity in mature human breast milk is in the form of retinol (retinyl esters), some is provided by carotene. -carotene is stored in the mammary glands during pregnancy and is rapidly secreted into milk during the first few days of lactation [64]. Thus, carotene provides almost 20% of the retinol equivalents during the first day, but this drops to less than 5% by the end of the first week [13]. Unlike retinol, 0-carotene is a very effective antioxidant and thus provides the infant a defence against oxygen toxicity [26, 65, 66]. This may be particularly important during the first several days of life, as the infant adjusts to its new oxygen-rich environment.

The vitamin A content of human milk is significantly affected by maternal nutrition during pregnancy and lactation [67, 68]. The fat content of the milk, time after birth (postpartum age), gestational age at birth, parity of the mother, and individual variation also have an influence. The use of oral contraceptives is reported to affect the amount of retinol in human milk. Thus, it is important to consider these factors when evaluating milk vitamin A levels.


Effect of maternal status on vitamin A levels in breast milk

Relatively few studies compared breast-milk vitamin A levels with the intake of retinol or retinol equivalents by lactating women, or directly compared the breast-milk concentration in well-nourished and poorly nourished communities. According to several authors, the amount of vitamin A in milk decreases with maternal deficiency of the vitamin and increases with excessive intake [12, 42, 57, 62, 69-72]. Also, data on maternal serum retinol and body mass index (BMI) were positively associated with milk retinol level in a multiple regression analysis of data on 153 rural Indonesian women that controlled for milk fat content [56]. Thus heavier and presumably better-nourished women had higher levels than thinner women who were less well-nourished. However, no association was seen between maternal intake of retinol and carotene and the corresponding values in the milk of well-nourished Canadian women [48], probably indicating adequate liver stores.

Overall, there are considerable differences in the vitamin A (retinol plus -carotene) of the milk of unsupplemented mothers in developed countries and developing countries (table 7). The range of retinol in presumably well-nourished women in developed countries was about 330-1,130 RE/L, and the range of carotene was about 10-65 RE/L. In developing countries, the corresponding ranges were 170-790 RE/L and about 40-47 RE/L. During the first six months of lactation, the retinol content of mature milk from mothers delivering at term averaged approximately 660 RE/L in developed countries and approximately 330 RE/L in developing countries. (It was possible to compare the vitamin A content of human milk only during the first six months of lactation, because no information is currently available after that time.) According to the Institute of Medicine [12], infants who consume human milk that provides 100-151 RE grow well and do not show signs of vitamin A deficiency. Thus, even though the retinol content of milk from mothers in developing countries averages about 50% of that of mothers from developed countries, it is sufficient to meet infant needs, not even counting the equivalents from carotene.

TABLE 7. Weighted average daily concentration of retinol and carotene in mature-term human milk from unsupplemented mothers (1-6 months)


No. of groups

No. of subjects

Vitamin A concentration (RE/L)



Retinol + carotene

Developed countries










Developing countries










Weighted averages are derived from data summarized in ref. 13 and include only mothers who delivered at term. Only studies that specify the number of subjects, whether birth was term or preterm, and the time after delivery are included.

Comparison is possible only for the six months postpartum because there is no information on the retinol and carotene content of human milk in developed countries beyond that time.

The carotene content of human milk appears to be affected by maternal dietary intake and nutrition status. Approximately 5% of the vitamin A of breast milk in developed countries comes from carotene, compared with over 10% in developing countries. The higher proportion of -carotene in the milk of disadvantaged Ethiopian women than in the milk of those of higher socio-economic status may have been due to inadequate esterification in the mammary glands [62]. It could also reflect a higher dietary intake of carotene compared with preformed retinol.


Non-dietary Actors influencing vitamin A levels In human milk

Fat content

The level of vitamin A in human milk is associated with its fat content, as was clearly demonstrated in classic studies conducted in the United Kingdom [59]. The fat content of milk is in turn affected by the time at which the sample is taken during a nursing session, the length of time between sample collections, and probably also the percentage of maternal body fat [12] and parity [73]. Thus collection procedures, as well as maternal nutrition status, are important variables.

Because it is higher in fat content, hind-milk contains more vitamin A than fore-milk [59, 60]. There is also evidence that the longer the time between nursing sessions, the lower the fat content [12]. Therefore, the procedure used to collect samples should ensure that both fore-milk and hind-milk are included, and should take into account the length of time between samples. Thus, to obtain accurate vitamin A levels, it is necessary to apply proper sampling, extraction, handling, and storage procedures as well as a sensitive and selective detection system [12]. Unfortunately, specific methodological information on how the milk samples were obtained was not always available in the studies reviewed, particularly those performed many years ago.

Studies conducted in communities where maternal undernutrition is prevalent furnished evidence indicating that the percentage of maternal body fat may influence the concentration of fat in milk [12]. Milk fat concentrations in the Gambia [74] and Bangladesh [75] were positively correlated with maternal skinfold thickness and decreased over the course of lactation. This positive relationship between milk fat concentration and body fat (as a percentage of ideal body weight) was also noted in women in the United States in late lactation (6-12 months) but not in early lactation [76]. High-parity (>10) Gambian women had a decreased capacity for total milk fat synthesis and thus lower milk fat concentrations [12, 73].

Of interest is a study reporting that the milk of low-income Egyptian mothers with sick infants was inferior in quality with respect to both lipid and retinol content [77]. A significant correlation existed between the milk retinol and growth of healthy infants, which was attributed to the higher lipid and thus higher energy concentration of the milk. Insufficient information was available to determine the average level of retinol per gram of fat in milk from the two groups.


Time after birth

The vitamin A content of breast milk changes over time to meet the changing needs of the growing infant. It is particularly high in colostrum and decreases over the course of lactation, with the rate of decline greatest during early lactation. There is nearly a 50% decrease over the first month, from an average of about 1,650 RE/L to about 700 RE/L in developed countries, and from about 1,250 RE/L to about 550 RE/L in developing countries. After that time, the decrease is much more gradual (table 8). The values in table 8 are somewhat different from the overall averages reported in table 7 because the overall averages included studies conducted during the first six months of lactation that did not specify the particular month that the milk was taken.

The FAO/WHO recommend a minimum of 180 RE per day for infants to meet basal needs. To allow normal liver storage of vitamin A, however, 350-375 RE per day is preferable throughout the first year of life and 400 RE per day during the second and third years. Thus, under conditions of relatively good maternal health and nutrition status in developed countries, mature human milk alone provides considerably more vitamin A than necessary to allow liver storage for at least the first six months of life, and probably for at least the first year; however, because of the lack of data on the vitamin A content of human milk in developed countries after six months postpartum, this cannot be known with certainty.

Even in presumably more poorly nourished populations in developing countries, human milk alone provides sufficient vitamin A activity to prevent clinical deficiency throughout the first 12 months of life. However, the average amount in milk is not sufficient to allow liver storage after about six months of lactation, even when both retinol and carotene are counted.

TABLE 8. Retinol and carotene concentration in term milk of unsupplemented mothers according to time after delivery, and average daily intake vitamin A by breast-fed infants

Time after delivery

Vitamin A concentration (RE/L)a

Breast milk intake

( L/day)b

Vitamin A intake (RE/day)



Retinol + carotene


Retinol + carotene

Developed countries

1-6 days

1,524 (161)

130 (170)


0.43 (47)



7-21 days

1,023 (117)

25 (64)


0.61 (60)



1-2 mo

683 (284)

33 (82)


0.71 (59)



3-4 mo

640 (242)

54 (34)


0.72 (22)



5-6 mo

745 (151)

35 (12)


0.81 (34)



7-12 moc







Developing countries

1-6 days

1,193 (129)

50 (129)


0.43 (47)



7-21 days

866 (182)



0.61 (60)



1-2 mo

495 (347)

46 (17)


0.71 (59)



3-4 mo

480 (328)

41 (19)


0.72 (22)



5-6 mo

459 (221)

43 (25)


0.81 (34)



7-12 mo

347 (254)

35 (20)





13-24 mo

278 (237)

31 (22)





>24 mo

130 (45)






a. Weighted averages on values in ref. 13. Figures in parentheses indicate the total number of samples for each average.
b Weighted averages based on those in ref. 78, except as otherwise indicated. Figures in parentheses indicate the total number of samples for each time period.
c. No information is available on the vitamin A content or human milk from developed countries after six months postpartum.
d. Based on data in ref. 78. Ref. 79 gives a range of 0.60-0.65 L/day; weighted average not available.
e. Ref. 79 gives a range of 0.50-0.55 L/day; weighted average not available.
f. Refs. 79 and 8O; weighted averages not available.

Of course, these are only averages. Some low-income populations may provide insufficient amounts earlier than that time [81] and other populations later [13]. Breast milk continues to be one of the most important sources of vitamin A in the infant's diet in developing countries throughout the second year, supplying a daily average of approximately 140 RE of retinol, or 160 RE of retinol plus carotene, from 13 through 24 months.


Gestational age at birth

Only four studies were located that reported the vitamin A content of preterm milk, and all were conducted in women in developed countries [47, 48, 82, 83]. Two of the studies [47, 82] contained apparent typographical errors in the units used to report the values. These were corrected [13], and the corrected values were included in the calculation of the weighted averages presented in table 9. However, to be sure that these weighted averages were correct, all were recalculated leaving out the originally reported values. As the recalculated averages were similar to those that included the questionable values, all four studies are shown in the table.

Preterm milk contains higher concentrations of vitamin A (retinol plus carotene) than term milk during the first two months of lactation, according to the weighted averages presented in table 9. As in term milk, the concentration decreases over the course of lactation, from an average of about 1,850 RE/L during the first week to about 1,250 RE/L by between one and two months postpartum. However, the rate of decline is not as great as in the milk of mothers who deliver at term.

The higher concentration in preterm milk appears to complement the lower stores of preterm infants. Although the average milk volume ingested is considerably less than that of term infants, preterm infants appear to receive a similar amount of vitamin A daily because of the higher concentration. The average amount ingested by preterm infants per day is adequate to meet basal needs ( > 180 RE per day) during the first week, and well above the recommended safe level ( > 350-375 RE per day) from the second week through two months of age. After that time, it is necessary to refer to information on the vitamin A content of term milk as no data are available beyond two months.

TABLE 9. Retinol and carotene concentration in preterm milk of unsupplemented mothers in developed countries according to time after delivery, and average daily intake of vitamin A by breastfed infants

Time after deliverya

Vitamin A concentration (RE/L)b

Breast milk intake


Vitamin A intake




Retinol + carotene


Retinol + carotene

1-6 days

1,671 (44)

188 (44)





7-13 days

1,707 (37)

89 (8)





14-21 days

1,355 (70)

65 (28)


0.38e (39)



1-2 mo

1,215 (140)

41 (40)


0.42e (52)



a. No data are available beyond two months postpartum
b. Weighted averages based on values in ref. l3. Figures in parentheses indicate the total number of samples for each average.
c. Adapted from ref. 84, table 4 (p. 318); averaged intake recommended for 1,501-1,800 g infant for days l-6.
d. Extrapolated from ref. 82 for day 7 and ref. 84 for days 7-9.
e. Weighted averages based on those in ref. 82. Data are for premature infants all 632 weeks, gestation; average birth weight is not available, so 1,500-2,000 g is assumed on the basis of the 50th percentile for 30-32 weeks, gestation [85] Volume is expressed volume and so is probably somewhat higher than actual intake.


Other non-dietary factors

Parity appears to affect the retinol and carotene levels in human milk. It was negatively associated with milk retinol content in a multivariate analysis of data on 153 rural Indonesian women [56]. This negative association may be related to the lower milk fat concentration associated with high parity. On the other hand, multiparous mothers in the United States had significantly greater mean colostrum carotenoid concentrations (363 RE/L) than did the primiparas (190 RE/L) [64].

Individual patterns in secreting vitamin A into breast milk are retained throughout lactation [60]: 7% of the samples measured had consistently low levels of vitamin A (<490 g/L); 15% had consistently high levels (>1,000 g/L). Consequently, infants with a constantly low vitamin supply over a long period of exclusive breast-feeding are exposed to a high risk of vitamin deficiency. On the other hand, mothers secreting vitamin A during protracted lactation might be themselves at high risk of deficiency if not adequately supplemented [60].

Both heat and light can decrease the vitamin A content of expressed human milk. Heating milk samples, as in banked milk, reportedly decreases the retinol content by approximately 25% [60]. Up to 70% of the vitamin A in human drip-breast-milk samples was destroyed during controlled exposure to daylight, either in translucent plastic bottles or where the milk was pumped through nave-gastric tubing from a syringe to mimic the conditions of enteral feeding [86]. Losses were also observed in milk that was exposed to standard phototherapy illumination under conditions similar to those encountered in the nursery ward. This may be of particular concern for hospitalized newborns who receive phototherapy for jaundice, particularly for those born preterm or with low birthweight, whose carotene stores are low.


Contents - Previous - Next