A and Lactation: Safe Vitamin Dosage During Pregnancy

WHO/NUT/98.
4
Orig~nal:English
Distr: General
Safe vitamin A dosage

during pregnancy
and lactation
Recommendations and report of a consultation
World Health Organization
Micronutrient
Initiative
WHO/NUT/98.4
Original: English
Distr: General
Safe vitamin A dosage

during pregnancy
and lactation
Recommendations and report of a consultation
World Health Organization
Micronutrient
Initiative
0 World Health Organization 1998
This document is not a formal publication of the World Health Organization,

and all rights are reserved by the Organization. The document may, however,
be freely reviewed, abstracted, quoted, reproduced or translated, in part or
in whole, but not for use in conjunction with commercial purposes.
The views expressed in documents by named authors are solely

the responsibility of those authors.
Table of contents
Tableofcontents .......................................................i
Abbreviations ........................................................v
Acknowledgements ...................................................vii
Introduction ..........................................................1
Recommendations for preformed vitamin A supplements
for mothers during pregnancy and the first six months
postpartum. and/or for their infants .......................................3
1. Maternal supplementation during pregnancy . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3
2. Supplementation for mothers in the first six months postpartum . . . . . . . . . . . . . 4
Atthepopulationlevel ................................................ 4
For individuals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4
3. Direct supplementation of infants before six months of age . . . . . . . . . . . . . . . . . . 4
4 . Supplementation both for mothers during the "safe" infertile
postparturn period and for infants under six months of age . . . . . . . . . . . . . . . . . . 5
Report of the consultation: Safe vitamin A dosage during

pregnancy and the first 6 months postpartum.
Geneva. World Health Organization. 19-21 June 1996 ........................7
Objectives of the consultation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .7
Generalobjective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
Specificobjectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
Vitamin A status during pregnancy and lactation ............................. 7
Vitamin A needs during pregnancy and lactation for the health of the mother
and her fetus or infant (Dr Kathleen Rasmussen) ............................ 8
The relation of maternal vitamin A status to other conditions ...................9
Breastmilkcomposition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Iron-deficiency anaemia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
HIV-linfection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
Countryexperiences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
Bangladesh . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .11
India. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
Gambia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .12
Morocco . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .12
Table of contents conf.
................. 12
Countries where vitamin A intakes habitually exceed the RDA
Summary of discussions ................................................. 13
Teratogenicity of vitamin A in humans ..................................... 14
Vitamin A and human developmental toxicity (Dr Edward Lammer) ........... 14
Mechanism of teratogenic action .......................................... 17
Threshold levels for teratogenicity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Evidence from regional registries of birth defects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
Summary of discussions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
Determinants of the return of fertility ...................................... 20
Physiological determinants of the return of fertility postpartum
and the duration of lactational infertility in developing countries
(Dr Kathy I. Kemedy and Dr Alan S. McNeilly) ............................ 20
HRP-sponsored multicentre study ......................................... 23
DHS data on risk of conception ........................................... 23
Countryexperience ..................................................... 24
Summary of discussions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
Policy and programme implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
Recommendations on doses and timing .................................... 25
1. Maternal supplementation during pregnancy ............................ 25
2. Supplementation for mothers in the first six months postpartum . . . . . . . . . . . . 26
Atthepopulationlevel ............................................... 26
Forindividuals ..................................................... 26
3. Direct supplementation of infants before six months of age . . . . . . . . . . . . . . . . 27
4 . Supplementation both for mothers during the "safe" infertile postpartum
period and for infants under six months of age ........................... 27
Researchneeds ..................................................... 27
Contact points with the health services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
Generic scenarios for populations where supplementation
programmesareneeded ................................................. 29
References ................................................................. 31
Table 1 Nutrients in breast milk most affected and least affected
by maternal diet . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Table 2 Ranges of percentage of women still breast-feeding, amenorrhoeic,
abstaining and insusceptible to pregnancy, by region and
timesincedelivery ............................................. 22
Table 3 Ranges of median durations of breast-feeding, amenorrhoea,
abstinence and insusceptibility to pregnancy by region . . . . . . . . . . . . . . . 23
Table 4 Global estimate of the percentage of births receiving antenatal care,
the place of birth, and the level of training of home birth attendants . . . . 29
Table 5 Estimates from 27 countries, by region, of contacts with the
health services from birth and up to 8 weeks. Generic scenarios
for populations where supplementation programmes are needed . . . . . . . 29
Table 6 Suggested situation analysis guideline for selecting appropriate
country-specific vitamin A supplementation programmes . . . . . . . . . . . . . 30
Annexes
b.
Annex l List of participants ............................................. 33

Abbreviations
CNS central nervous system

DHS Demographic and Health Surveys1
ENTIS European Network of Teratology Information Services
GnRH gonadotropin-releasing hormone
HIV human irnrnunodeficiency virus
HRP Special Programme of Research, Development and Research
Training in Human Reproduction
IU international units
LAM lactational amenorrhoea method
LH luteinizing hormone
RBP-R retinol-binding protein-bound retinol
RDA recommended dietary allowance
RE retinol equivalents
MSG monosodium glutamate
UNDP United Nations Development Programme
UNFPA United Nations Population Fund
UNICEF United Nations Children's Fund
VAD vitamin A deficiency
WHO World Health Organization
~p
l DHS Demographic and Health Surveys Program, Columbia, Maryland 21045, USA.
Acknowledgements
The World Health Organization gratefully acknowledges the financial support provided by the
Canadian International Development Agency, through the Micronutrient Initiative, which
enabled WHO to convene the June 1996 consultation on safe vitamin A dosage during
pregnancy and the first 6 months postpartum, and to publish its report.
The Micronutrient Initiative (P.O. Box 8500, Ottawa, Canada K1G 3H9, tel. (613) 236.6163, fax
(613) 236.9579) was established in 1992 by its principal sponsors - the Canadian International
Development Agency, the International Development Research Centre, the United States
Agency for International Development, the World Bank, the United Nations Children's Fund,
and the United Nations Development Programme - to contribute to the sustainable control and
elimination of micronutrient malnutrition.
Thanks are also due to Dr Barbara Underwood, formerly responsible for micronutrients in
WHO'S Programme of Nutrition, for organizing the consultation and participating in the
preparation of its report, to Dr Bruno de Benoist for reviewing the manuscript, to
Mr Jarnes Akre for editorial and organizational inputs, and to Mrs Eileen Brown for secretarial
support.
Introduction
v itamin A is essential for normal maintenance and functioning of body tissues, and for
growth and development, including during pregnancy when the fetus makes demands
on the mother's vitamin A stores, and during the postpartum period when the newborn
is growing rapidly. Although the increased requirement during pregnancy is relatively small,
in many countries where vitamin A deficiency (VAD) is endemic, women often experience
deficiency symptoms such as night blindness that continue during the early period of lactation.
In most cultures, the young infant depends on breast milk to obtain adequate amounts of the
vitamin. Breast milk from deficient mothers is likely to contain insufficient vitamin A to
build -or even to maintain-vitamin A stores in nursing infants. The fact that VAD is known
to be associated with increased child mortality after six months of age is a convincing argument
for improving maternal vitamin A nutrition as a child survival strategy. There may also be an
additional direct benefit to maternal health, though this is not well documented.
Both severe VAD and vitamin A overload are teratogenic in animals and are associated with
adverse reproductive outcomes. Although similar outcomes in deficient human populations are
not documented, the possibility of risk must be considered. Providing a diet adequate in
vitamin A-neither too little nor too much-is the safest solution to meeting needs during
pregnancy and lactation. However, this is not easily accomplished in situations of poverty and
where food with appropriate vitamin A content is in short supply and/or expensive. In such
situations the recommended approach is to provide a vitamin A supplement during pregnancy
at a dosage and frequency that will safely meet the needs of growing maternal and fetal tissue
and will potentially build maternal body stores in anticipation of lactation. However, using
high-dose vitamin A supplements to build maternal stores during pregnancy creates a dilemma
because of the vitamin's potential teratogenicity during the early stages of pregnancy. The
known teratogenic risk from excessive vitamin A taken in early pregnancy way extend to risk
from excessively high doses taken during later periods of pregnancy, with possible non-
teratogenic consequences on the subsequent behavioural and neuropsychological performance
of the offspring. However, non-teratogenic developmental toxicity has not been confirmed. To
avoid risk, an alternative to maternal high-dose supplementation during pregnancy is provision
of a direct high-dose supplement to the mother during the early postpartum period, assuming
that both approaches will improve vitamin A levels in breast milk and reduce mortality among
nursing infants in the first years of life.
Concern about potential teratogenesis has been raised because of recent data from a
retrospective study of vitamin A intake among women in the USA. The study involved
telephone interviews of women who presented for prenatal diagnosis and later gave birth to
babies that had defects of structures purported to have an embryologic contribution from
cranial neural crest cells (tissues associated with 13-cis-retinoic acid teratogenicity). These data
suggest that total dietary intake above 15 000 IU, or above 10 000 IU if from supplements alone,
increased the risk of these defects. These intakes are not uncommon in affluent populations that
habitually consume more than the recommended dietary allowance (RDA) of vitamin A and
also frequently consume vitamin supplements and/or rich sources of preformed vitamin A,
such as animal liver. The study called into question the safety of giving supplements to fertile
women whose habitual diets are either adequate or deficient in vitamin A.
WHO has received requests from its Member States, from UNICEF and from nongovernrnental
organizations (NGOs)for programmatic guidance on the safe use of vitamin A supplements by
women of reproductive age. The Organization currently recommends that the relatively small
increased need for vitamin A during pregnancy should be met through diet, or through a
supplement not exceeding 10 000 IU daily throughout pregnancy. Because of the logistical
difficulties associated with daily supplementation, some advocate a weekly or monthly
supplement. There is, therefore, a need to review the risks and benefits to mother and fetus of
frequencies and levels of vitamin A intake during pregnancy, and to assess their programmatic
implications, keeping in mind the known long-term potential benefit of mortality reduction in
the offspring.
Meanwhile, WHO, UNICEF and the International Vitamin A-Consultative Group (IVACG)
continue to recommend that, in areas of VAD endemicity, high doses of supplemental vitamin
A (200 000 IU) be given to breast-feeding women during the infertile postpartum period which
until recently was interpreted as lasting 4-6 weeks. It is well documented that following
parturition the infertility period is influenced by breast-feeding practices. New data have
become available from a multicentre study of the UNDP/UNFPA/WHO/World Bank Special
Programme of Research, Development and Research Training in Human Reproduction (HRP)
and from the Demographic and Health Surveys (DHS) concerning the return of menses relative
to breast-feeding practices and country-specific contacts with the health system. There is a need
to review these data and additional available information for their programmatic relevance for
identifying opportunities for the safe provision of high-dose vitamin A supplements during the
postpartum period.
In the light of the above considerations, WHO convened a consultation to consider both the safe
dosage of vitamin A during pregnancy and the first six months postpartum, and the relevant
policy and programme implications. This document presents the recommendations of
participating experts in nutrition, teratology, reproductive physiology and population-based
surveys, who have experience in both basic research and its public health applications.
The recommendations of the consultation appear first, and the scientific and programmatic
considerations leading to these recommendations follow in the report of the meeting.
Recommendations for preformed vitamin A
supplements for mothers during pregnancy
and the first six months postpartum, andlor for their infants
Four scenarios were identified in which vitamin A supplements could be given through public
health programmes and where safe dosage and frequency of administration need to be
considered. These scenarios were:
1. Maternal supplementation during pregnancy.

2. Supplementation for mothers in the first six months postpartum.
3. Direct supplementation of infants before six months of age.
4. Supplementation both for mothers during the "safe" infertile postpartum
period and for infants under six months of age.
1. Maternal supplementafion during pregnancy

(Either during thefirst 60 days following conception whez there is a teratogenic risk or after thefirst 60
days following conception,for wo~izenwhose habitual intakes are above the RDA or below the RDA)
For fertile women, independent of their vitamin A status, 10 000 IU (3000 pg RE) is the
maximum daily supplement to be recommended at any time during pregnancy.
Where VAD is endemic among children under school age and maternal diets are low in
vitamin A, health benefits are expected for the mother and her developing fetus with little risk
of detriment to either, from:
either a daily supplement not exceeding 10 000 IU vitamin A (3000 pg RE) at any
time during pregnancy;
or a weekly supplement not exceeding 25 000 IU vitamin A (8500 pg RE). In this
regard:
a single dose > 25 000 IU is not advisable, particularly between day 15 and day 60
following conception (day 0);
beyond 60 days after conception, the advisability of providing a single dose of
> 25 000 IU is uncertain; any risk for non-teratogenic developmental toxicity is likely
to diminish as pregnancy advances. In the case of a pregnant woman who may be
reached only once during pregnancy, health workers should balance possible
benefits from improved vitamin A status against potential risk of adverse
consequences from receiving a supplement.
Where habitual vitamin A intakes exceed at least three times the RDA (about 8000 IU or 2400
pg RE), there is no demonstrated benefit from taking a supplement. On the contrary, the
potential risk of adverse effects increases with higher intakes-above about 10 000 IU-if
supplements are routinely ingested.
2. Supplementation for mothers in the first six months postpartum
(Single high-dose supplenzazt above 25 000 IU, and usually at a level of200 000 IU, during the safe
period ofpostpartum infertility for mothers in vitamin-A-dtficient areas)
At the population level
Mothers who are not breast-feeding will benefit from a high-dose supplement given safely
during the first 28 days (4 weeks or 1 month) postpartum. Although the risk of conception
beyond this point is poorly documented, normal fertility does not usually return for 5-10 weeks.
Beyond 6 weeks, therefore, non-lactating mothers should be given no more than 10 000 IU daily.
Direct supplementation of the non-breast-fed infant < 6 months of age, who is not given a
fortified breast-milk substitute, with as much as 50 000 IU (15 000 pg RE) is the recommended
safe intervention to meet the infant's need for vitamin A.
Mothers who are breast-feeding will benefit from a high-dose supplement given up to 60 days
(8 weeks or 2 months) postpartum, as will their nursing infants-through higher levels of
vitamin A in breast milk. The risk of pregnancy is related to menstrual status of the breast-
feeding mother. If she has resumed menstruation she is regarded as fertile. If she is
arnenorrhoeic, the risk of pregnancy rises after 60 days, in some circumstances reaching 1%-2%
by 6 months postparturn. In some very high-risk areas (e.g. where there is a high prevalence of
clinical symptoms of VAD in mothers) where a high percentage of contact with the health
system occurs within 8-12 weeks, the risk of extending the period for giving a high-dose
supplement from week 8 to week 12 (estimated 2.8% conceptions) might be offset by important
benefits to the mother (relief of symptoms) and the nursing infant (increased breast-milk
vitamin A consumption and consequent increased likelihood of decreased risk of mortality).
For individuals
Mothers who are not breast-feeding, provide a supplement within 28 days of delivery;
otherwise give a supplement directly to the infant.
Mothers who are breast-feeding:

if using reliable contraception, give supplement anytime;
if practising postpartum abstinence, give supplement anytime;
if amenorrhoeic, give supplement up to 6 months postpartum;
if not amenorrhoeic, give supplement to mother at time of next menstruation (an
indication that conception has not occurred) or give supplement to child.
3. Direct supplementation of infants before six months of age

(In areas of endemic vitamin A deficiency)
Firm evidence of benefits to breast-feeding infants of direct supplementation before six months
of age is insufficient. Studies are in progress to clarify the benefits/risks of single
Safe vitamin A dosage during pregnancy and lactation
.. -..-. , YYY*p.-.-&-- ,.
supplementation at 50 000 IU (15 000 pg RE) at birth or thereafter, or multiple supplementation

at 25 000 IU (7500 pg RE).
Infants who are not breast-fed and who are not given fortified breast-milk substitutes should
receive a 50 000 IU supplement, preferably by about 2 months of age -otherwise at any time
within the first 6 months of life. As an alternative, two doses of 25 000 IU can be given with an
interval of a month or more in between.
4. Supplementation both for mothers during the "safe" inferfile

postpartum period and for infants under six months of age
There is currently insufficient information to make firm programmatic recommendations

regarding risks and benefits to infants on the basis of this supplementation strategy. A WHO-
sponsored trial is being conducted in three countries to clarify this issue and results should be
available within the next year.
Report of the consultation
Safe vitamin A dosage during pregnancy and
the first 6 months postpartum,
Geneva, World Healih Organizatton, 19-2 1 June 1 996
Obieciives of the consultatton
Genera l o biective
To review the available information and reach consensus on dosage and contact points during
pregnancy and lactation when vitamin A supplements can be safely administered.
Specific o biectives
1. To review evidence of VAD during pregnancy and/or lactation from countries

WHO identifies as having a public health problem among children of pre-school age
and to summarize information on programmes currently being implemented for this
physiological stage.
2. To review available data on the length of postpartum infertility and how this is
influenced by current and past breast-feeding practices, as well as other
maternal/infant factors.
3. To review available data linking vitamin A to teratogenicity in humans from both
deficient and excess intakes.
4. To review the demonstrated benefits of improved maternal vitamin A status during
pregnancy and lactation compared with the benefits of direct supplementation of the
infant during the first six post-gestational months.
5. To consider the programmatic implications of recommendations in the light of
variations in women's contacts with health services during pregnancy and the first
six months postpartum.
Vitamin A status during pregnancy and lactatton

To provide guidance on safe and effective strategies for improving the vitamin A status of
mothers and their infants, a review was presented of what is known about needs for vitamin
A among pregnant women and their fetuses and among lactating women and their nursing
infants. Although a small amount of vitamin A is essential for successful reproduction, too
much-as well as severe deficit-is teratogenic. A brief summary of the information from the
background review is presented as follows.
Vitamin A needs during pregnancy and
lactation for the health of the mother and her
fetus or infant (Dr Kathleen Rasmussen)
Among pregnant women, signs of VAD (night blindness) have been reported, particularly
among those living on the Indian subcontinent. This problem is associated with poor
socioeconomic status, high parity at young or old reproductive ages, and low intake of foods
containingvitamin A. Vitamin A and related compounds are transferred from mother to fetus
via the placenta. Studies in animals reveal that this transfer is sensitive to maternal vitamin A
intake, both chronic and acute.Low doses (0.2-8.4 pmol, equivalent to about 200-8000 IU') of
vitamin A given to pregnant women increase their plasma retinol values. No birth defects have
been reported with these doses, although none would be expected because in all studies they
were given after embryogenesis was complete. The safe daily or weekly dose that can be given
to pregnant women is not precisely known; it may depend in part on the woman's hepatic
reserve of vitamin A. Massive doses of preformed vitamin A have not been routinely given to
pregnant women so the effect of such a supplementation strategy on maternal and fetal health
is unknown. However, it is clear that massive dosing with vitamin A should be avoided during
the periconceptional period.
Among lactating women on the Indian subcontinent and in south-east Asia, signs of poor
vitamin A status (night blindness) are reported, and low breast milk retinol values are found
among Indian and Indonesian women. The problem is most likely to occur if the prior
pregnancy was characterized by symptoms of night blindness. Vitamin A is transferred to milk
from both retinol-binding protein-bound and chylomicra-associated vitamin A carriers, i.e.
retinol and retinyl ester respectively. The chylomicra-associated vitamin A may be a particularly
important route of transfer among women who receive chronic low doses of the vitamin (e.g.
from food fortification programmes) as well as periodic acute high doses.
Both low daily doses and periodic massive doses of vitamin A increase the concentration of
retinol in human milk. Massive doses, i.e. up to 312 pm01 (300 000 IU), have been given to
lactating women in the first few weeks postparturn with beneficial effects on both the mother
and her nursing infant for at least the subsequent 6 months. The safe upper limit of the amount
of vitamin that can be given at this time is not known. It could be higher than the 312 pm01 that
has been used successfully but additional studies, including monitoring of retinoid metabolites,
are warranted before higher single dosages are recommended. If the infant is also to receive a
l Various units are used in the literature to refer to concentrations of preformed vitamin A. In
this report, equivalent vitamin A (retinol equivalents [RE]) concentrations are defined as:
1RE = l p g all-trans retinol = 0.0035 pm01 = 3.33 IU

1pm01 = 286 pg retinol = 951 IU.
In addition, most supplements are available as retinyl palmitate or retinyl acetate which have
weight equivalence 1.83 and 1.15 respectively, greater than retinol (e.g. 60 mg retinol = 110 mg
retinyl palmitate = 69 mg retinyl acetate, all with a potency of 200 000 IU).
- - - - A . - -
supplement of vitamin A directly, this should be taken into consideration when calculating the
amount to be given to the mother.
Infants are born with low hepatic reserves of vitamin A. These reserves normally increase 60-
fold in the first 6 months of life in response to the vitamin A delivered in breast milk which is
the primary -if not the only -source of vitamin A in their diets. It is safe to supplement infants
by supplementing their mothers. It is also safe to supplement infants directly. Massive doses
of up to 52 pm01 (50 000 IU) have been given to newborns; in one study this dose reduced infant
mortality. Other studies of the same or lower doses, given two or more times somewhat later
after birth but before 6 months of age, failed to show a reduction in mortality or morbidity
among populations where breast-feeding predominates.
Insufficient information is currently available to evaluate the safety and effectiveness of dosing
both a mother and her nursing infant concurrently.
The relation of maternal vitamin A status to other conditions
Breast milk composition
Maternal diet influences the level of several vitamins

and minerals in breast milk, including vitamin A
Table 1 . Nutrients in breast milk
(Table 1). Epiderniologic studies of vitamin-A- most affected and least affected bv
fortified food products, such as monosodium maternal diet
glutamate (MSC) and sugar, and intervention trials Nutrients most Nutrients least
using high-dose supplementation for mothers soon affected affected
after birth, demonstrate the association between diet Fatty acids Protein
and breast milk. Transfer of vitamin A from maternal Vitamin A Lipid
blood to breast milk is regulated primarily by levels Vitamin D Carbohydrate
of maternal retinol-binding protein-bound retinol Vitamin E (lactose)
(RBP-R), and possibly also chylomicra-solubilized Vitamin K Calcium
retinyl esters soon after supplementation with high Thiamin Magnesium
doses of vitamin A. Most retinol in breast milk is in Riboflavin Folate
the form of retinyl esters and is related to the fat Vitamin B, Iron
content of breast milk. Low maternal zinc status, Vitamin B,, Zinc
protein-energy malnutrition (PEM), and situations Selenium Copper
that activate an acute-phase response (e.g. infectious Iodine Manganese
diseases) all reduce levels of RBP and as a con-
sequence cause reduced serum levels of vitamin A, i.e.
RBP-R, which could limit the retinol transferred to
milk.
Iron-deficiency anaemia
Haemoglobin response to iron supplementation is suppressed among anaemic subjects who are
also deficient in vitamin A. Data from the WHO global data bank on anaemia among women
of reproductive age, including those pregnant and lactating, showed considerable overlap with
countries where VAD occurs among children of pre-school age. Presumably these are countries
where women also are likely to have inadequate vitamin A status. For example, the highest
prevalence of anaemia in pregnant women- up to 78%-occurs in south-east Asia where the
prevalence of VAD is also highest. This relationship was recently demonstrated among iron-
deficient, lactating women in Indonesia; VAD depressed their haemoglobin response to iron
supplementation by about 30%.
Participants at the consultation noted that in many parts of the South-East Asia Region,
particularly in India, the reason for the high prevalence of anaemia is the low bioavailability of
food-iron from habitual diets and from programmes that limit supplemental iron doses. Under
these circumstances -which are also likely in some other regions with a high prevalence of
anaemia -bioavailable iron is essential for correcting anaemia. Providing vitamin A alone,
therefore, will not contribute to resolving the anaemia problem unless there is concurrent
improvement in the quantity of bioavailable iron.
HIV- l infection
There is evidence among HIV-infected pregnant women in Africa that suggests poor vitamin
A status (serum retinolc1.05 pmol/l [30 pg/dl]). In Kenya, Malawi and South Africa, where
vitamin A status has been assessed in these women, it is estimated that 52-63% are in poor
status. In Malawi, poor status is associated with increased vertical transmission of HIV-1,
higher infant mortality, low birth weight and increased maternal mortality. In Kenya, HIV-
infected lactating women with severe VAD (serum retinol <0.70 pmol/l [20 ,ug/dl]) are at
increased risk of having HIV-infected cells in their breast milk. However, these findings do not
establish a causal role for VAD in HIV transmission because low serum vitamin A could be a
marker of infection severity or other risk factors.
To establish causality, clinical trials are needed where vitamin A supplements are given to HIV-
infected pregnant women. At least three such trials are known to be in progress-one each in
Malawi, South Africa and the United Republic of Tanzania. In all three trials, a daily dose of
vitamin A or placebo is given during the third trimester of pregnancy. The Malawi study
provides 10 000 IU retinyl palmitate (5500 pg RE) while the South African and Tanzanian
studies provide 5000 IU (2750 pg RE) retinyl palmitate and 30 mg beta-carotene. At delivery,
women in all three studies will receive 200 000 IU (110 mg RE) of vitamin A or a placebo.
Concern has been expressed about the safety of vitamin A supplementation in HIV-1 infected
women. One in vitro study showed that retinoic acid increased replication of HIV-1, while
another in vitro study, using different cell-culture conditions, showed decreased replication. The
question as to whether vitamin A increases the HIV-1 viral burden is important because many
women in vitamin-A-deficient countries who may be targeted to receive supplementation could
be infected with HIV. A preliminary study from South Africa showed no increase in HIV-1 viral
burden when pregnant women were given a low daily dose of vitamin A in the third trimester
of pregnancy and 200 000 IU at delivery. These findings require verification.
-
Participants identified concerns that need additional research because of important pro-
grammatic implications. These included:
how the transfer of vitamin A to breast milk can be optimized;
what maternal conditions (e.g. other nutrient deficits or diseases) can interfere with
transfer;
whether the vitamin A concentration in the mammary gland affects the content of
vitamin A in breast milk;
whether retinyl-ester, or retinol-metabolite, spikes in breast milk correspond to
spikes in blood chylomicra-solubilizedretinyl esters and, if so, how long the levels
remain elevated.
The programmatic concern of the latter question is whether there should be a period after
maternal high dosing during which breast-feeding should be postponed to minimize any
unknown, but potential, risk to the infant from unphysiologically high concentrations of
retinoid metabolites.
Country experiences
Brief presentations were made of programme experiences in Bangladesh, Gambia, India and
Morocco. The presentations and subsequent discussions illustrated the wide variation in
situations where supplementation programmes are, or may need to be, implemented on behalf
of pregnant and lactating women.
Bangladesh
In Bangladesh, there is currently no national vitamin A supplementation programme that

includes women routinely. Nearly 80% of deliveries in rural areas are in the home with no
contact between health workers and mothers before 8 weeks postparturn. Only about one-third
of the home deliveries are attended by trained birth attendants. Under these conditions, it is
programmatically difficult to supplement pregnant and lactating mothers safely through a
nationwide programme implemented through the routine health services. It could be possible
programmatically to add a vitamin A capsule supplement to a "safe delivery kit" or "cord kit"
where these are used. Limited experience using this approach in the Matlab area in Bangladesh
has been positive. However, before this approach can be recommended on a broad scale, cost-
effective studies should be carried out.
India
Data from India indicate that, unlike anaemia, vitamin A deficiency in pregnancy is not
associated with increased maternal morbidity and mortality. There is at this time, therefore, no
prograrnrne for universal supplementation to pregnant women. Most deliveries in rural areas
take place at home and there is no specific maternal health contact point in the first 8 weeks
after delivery when large-dose vitamin A supplementation can be given safely. India does not,
therefore, have a policy encouraging targeted supplementation to lactating women who contact
health services prior to 8 weeks postpartum. Currently infants' vitamin A needs are directly
addressed by providing 100 000 IU at 9 months at the same time as the measles immunization.
Immunization-linked vitamin A supplementation programmes during infancy need to be

reinforced by joining them with growth monitoring or similar programmes that provide
periodic contact when the supplement can be delivered. Contacts for supplement delivery are
necessary where diets remain inadequate to ensure the potential benefit of mortality reduction
throughout the vulnerable pre-school years. Maternal needs, of course, will not be met safely
through infant immunization contacts unless coupled to a BCG or maternal tetanus toxoid
booster within 8-12 weeks of delivery. However, opportunity for targeted supplementation
should be sought for those lactating women who do come into early contact with the health
system, i.e. by 8 weeks from delivery.
Gambia
In Gambia, in spite of a low and seasonably variable dietary intake of vitamin A, almost none
of which is preformed, there is little biochemical evidence (e.g. moderately low serum retinol
values) or functional evidence (e.g. night blindness) of deficiency among pregnant and lactating
women. Both serum levels and breast-milk retinol levels, however, responded modestly to a
small, i.e. less than 1RDA, supervised daily dose of vitamin A over a period of several months.
In this scenario, the data suggest that a dietary increment in vitamin A activity, rather than
supplements, could control any subnormal vitamin A status. One potential diet-based
programme considered appropriate to the area is improved year-round availability of vitamin
A though increased supply of sun-dried mangoes in rural communities.
Morocco
Morocco represents yet another scenario. No recent information is available on the vitamin A
status of different potentially vulnerable groups in the country. Before considering a
programme, therefore, a systematic review (a situation analysis) is required to identify the
target populations for which programmes may be needed (children, pregnant and lactating
women, or both), and to find out the health service coverage or access to services of the
identified target group(s) (e.g. the percentage of women followed at antenatal clinics, the
distribution of home or assisted deliveries, and antigen-specific immunization coverage of
children). On the basis of this information, programmes can be designed that efficiently utilize
health institutions and other health service contact points that achieve the broadest coverage
of the target group during the period of its vulnerability. Only then will it be possible to tailor
the dose and frequency of administration to the existing infrastructure for safe delivery of
vitamin A to vulnerable groups.
Countries where vitamin A intakes habitually

exceed the RDA
Several countries in Europe, as well as Japan and the USA, have expressed concern about the
possible adverse effects of too much vitamin A because large numbers of their populations
regularly consume the vitamin in excess of the RDA. The consumption of liver during
- -U
- . -W. - . - ,,P
pregnancy has been of particular concern. A recent study in the United Kingdom (1)was cited
that compared equivalent amounts of vitamin A (i.e. 50 and 150 mg, about 165 000 and 500 000
W) given as a supplement or as cooked calf liver resulting in markedly different metabolism.
Peak serum values of potentially teratogenic metabolites from liver were several-fold lower and
occurred later than when supplements were given. The consultation reviewed a report from
France which is thought to exemplify this type of situation.
Nutrition surveys in France indicate that vitamin A intakes, especially among women of
reproductive age, are likely to be above adequate levels with the possible exception of a
growing n~unberof people in precarious economic situations. About 60% of dietary vitamin A
is from carotenoids, the remaining coming from preformed sources. Some foods containing
preformed vitamin A (e.g. animal livers), although not widely consumed, have potential for
raising total intakes to levels that could be teratogenic for some pregnant women who consume
liver frequently. Concentrations as high as 100 000 IU/100 g liver (30 000 ,ug RE/100 g), or
occasionally even higher, are reported in France. This is the consequence of supplements added
to the diet of these animals. Recently the situation prompted France and the European Union
to reduce the amount of vitamin A provided in animal feed and, as a consequence, vitamin A
concentrationsin animal liver have been effectively reduced. This control measure is, therefore,
being continued.
Vitamin A supplements are also a source of concern in France because their consumption
appears to be increasing. To further protect against teratogenic risk and hypervitaminosis A,
a recent decision was made to limit the potency of vitamin A supplements available to the
public to a maximum of one RDA, i.e. 1000 pg/unit. This is a precautionary measure since there
has been no increase in teratogenic events reported in France that could be related to an
excessive vitamin A intake for a period when an increase in the vitamin A food supply may
have occurred. This conclusion was reached following a study of 1200 pregnant women among
whom no statistically significant increased risk of malformation was found to be associated
with an increased intake of vitamin A (dietary plus supplements). Because these events are rare,
however, cautious interpretation of this study's conclusion is advised as its design and/or
statistical power may have been inadequate.
France has thus considered the potential harmful effect of too much vitamin A, especially for
women of reproductive age, and has instituted regulatory measures to minimize risks. In
addition to decreasing the vitamin A content of animal feed and limiting the vitamin A
concentration of supplements to one RDA/unit, measures include the establishment of scientific
groups to keep a coordinated watch on the situation. Similar measures should be considered
in other countries where vitamin A intakes habitually exceed the RDA.
Summary of discussions
Adequate maternal vitamin A status ensures protection from the adverse consequences of
either too much or too little vitamin A for the mother, fetus and newborn. Adequate status is
maintained by intakes at the level of the RDA. Where habitual diets are up to three times the
RDA, there is no demonstrated benefit in taking a supplement during pregnancy or lactation.
Where habitual vitamin A intakes are low, a pregnant women and her developing fetus are
expected to benefit without risk from daily vitamin A intakes of 10 000 IU from diet and/or
supplement or 25 000 IU weekly. Lactating women can safely be given a high-dose vitamin A
supplement during the infertile postpartum period of 200 000-300 000 IU, which will maintain
breast milk concentrations for at least 6 months.
In areas with a high prevalence of iron-deficiency anaemia and VAD, vitamin A supplements
will improve the maternal response to iron supplementation but are not a substitute for iron
prophylaxis. The benefit to mother, fetus and newborn of supplements to HIV-l-infected
mothers who are vitamin A deficient remains to be demonstrated by studies that are in
progress.
Limited current evidence suggests that concentrated food sources of vitamin A (e.g. animal
liver) are utilized differently than preformed vitamin A supplements of similar retinol
equivalence. This finding requires confirmation. Meanwhile, where median intakes of pregnant
women exceed the RDA three-fold, countries should be encouraged to monitor and control
levels of vitamin A in animal livers. Occasional consumption of cooked liver (e.g. weekly or less
frequently in about a 50-70 g serving size) by pregnant women poses no significant teratogenic
risk while providing a valuable source of other nutrients such as iron and folic acid.
Teratogenicity of vitamin A in humans
Animal studies have demonstrated that both too little and too much vitamin A during critical
periods in embryonic development can have teratogenic consequences. There are limited
human data that directly link teratogenicity in women exposed early in pregnancy to high doses
of preformed vitamin A, i.e. retinol and retinyl esters. However, teratogenic effects from
naturally occurring metabolites of vitamin A, e.g. trans-retinoic acid, 13-cis retinoic acid, and
their 0x0-derivatives, are well documented from case-studies of women directly exposed to
high doses of preformed retinoic acid derivatives within the first six weeks of pregnancy.
Extensive epiderniologic studies have produced no evidence for teratogenicity in humans after
6 weeks of pregnancy. Other, as yet unpublished, reports of toxic non-teratogenic
developmental effects have not been sufficiently documented to date. A review was presented
summarizingwhat is known about the developmental toxicity of inadequate and excess vitamin
A during human pregnancies. A condensed version of the review appears below.
Vitamin A and human developmental toxicity

(Dr Edward Lammer)
It is generally accepted that physiological serum concentrations of retinol and its metabolites
are non-teratogenic and that retinoic acid plays an essential role in controlling many aspects of
normal embryogenesis. In the embryo, retinoic acid (like several isomers) acts as a ligand that
binds to a nuclear hormone receptor. This ligand-receptor complex then acts to coordinate
expression of many other genes by binding to a regulatory element (DNA sequence) of a target
gene, including genes coding for other transcription factors. Experimental evidence indicates
that the embryonic concentration of retinoic acid determines at least some of the differential
specificity of genetic regulatory control attributable to retinoic acid. This suggests that tight
control over retinoic acid levels in embryonic tissues is of critical importance, and that the
teratogenic effects probably result from receptor-mediated processes. Thus, the potential for
vitamin A teratogenicity stems from intakes that increase maternal blood levels of retinoic acid
isomers, not retinol or retinyl esters. How much of an increase is "too much" is unknown.
Given the global magnitude of VAD, it is surprising that there are so few observations of
malformed babies delivered to pregnant women who have symptoms of VAD. Countries in
which VAD is endemic do not appear to have a higher prevalence of birth defects than other
countries. Only a handful of published case reports of teratogenicity can be found and they
describe eye malformations and central nervous system problems. However, no systematic
studies of VAD during pregnancy have been reported.
Experimental studies strongly suggest that retinoic acid (all-trans-RA, 13-cis-RA, 4-0x0-all-
trans-RA, 4-0x0-13-cis-RA), and not retinol, is the proximate teratogen. Thus, most concern
about excessive maternal intake relates to preformed vitamin A supplements that produce
elevations in maternal serum retinoic acid levels, rather than to increases in maternal serum
retinol or retinyl esters. The potential for human teratogenicity from excess maternal
consumption of preformed vitamin A is based on the unequivocal demonstration of human
teratogenicity of 13-cis-retinoic acid (isotretinoin) when it is taken during early pregnancy.
Daily therapeutic doses result in peak serum concentrations of isotretinoin of >200 ng/rnl,
compared to endogenous levels of 1-4 ng/ml. Isotretinoin causes a characteristic pattern of
spontaneous abortion, premature delivery and malformations that involve central nervous
system, cranio-facial and cardiac development. The magnitude of the teratogenic risk is
unusually high when isotretinoin is used beyond the 15th day after conception but is not
increased if it is stopped within the first 15 days after conception. Additional adverse outcomes
of pregnancy, although less severe, have been reported when isotretinoin was used exclusively
during the fetal period of development (beginning more than 60 days after conception). Thus,
there is no "safefftime during pregnancy when a high dose of vitamin A might be given without
putting a fetus at some risk, except during the first 15 days after conception.
A review of evidence of human developmental toxicity from excessive preformed vitamin A

intake included case reports and controlled studies. At least six relatively complete case reports
of adverse pregnancy outcome associated with daily intake of 25 000 IU, or more, preformed
vitamin A have been published. No pattern of anomalies is obvious among these cases, except
for an unexpected number of renal/urinary tract anomalies. Other later case reports, published
after identification of the retinoic acid (isotretinoin) embryopathy, more closely resemble that
phenotype, probably because of reporting bias.
Four casecontrol studies of vitamin A supplement intake during pregnancy were reviewed. For
most, the measure of exposure was ingestion of 10 000 IU or more per day of retinol or retinyl
palmitate. Among the studies, the frequency of such "exposure" was 0.14%, 0.3%, 1.3% and
0.5%. In three of these studies, cases were selected for the presence of a malformation that
involved a structure that was embryologically derived, at least in part, from cranial neural crest
cells (e.g. orofacial clefts, conotruncal cardiac defects, ventricular septal defects). These
malformations were targeted because experimental and human studies of retinoic acid
teratogenesis indicated that activities of cranial neural crest cells were adversely affected. The
results of these studies do not demonstrate human teratogenicity when exposure is quantified
as >l0 000 IU preformed vitamin A per day.
Four cohort studies of vitamin A use during pregnancy were summarized. Two were small and
not controlled. In the third there was no excess of birth defects among the offspring of women
who took a multivitamin that had 6000 IU of preformed vitamin A. The population of the fourth
study was a large, prospectively followed cohort of women referred to an (academic) urban
medical centre for prenatal diagnostic services. In many studies of supplement users, high
vitamin A exposure was defined as more than 10 000 IU of retinol per day. The overall
incidence of birth defects was low (1.5%), indicating substantial under-ascertainment of
abnormalities. There were 10 malformed outcomes among the higher-dose group. Using least
squares regression, the authors showed that the prevalence of cranial neural crest outcomes
(n=7)among the maternal intake category of >l0 000 IU/day was greater than the prevalence
in the lowest exposure category by 1.7% of births, or 1baby in 57. The authors concluded that
"1 of every 57 babies is born with a birth defect attributable to the high vitamin A intake of the
mother" among women who take more than 10 000 IU of preformed vitamin A from
supplements. The strengths and limitations of this research are discussed in detail.
Because retinoic acid in therapeutic doses is a human teratogen, and because many of the
embryological roles of vitamin A are carried out by conversion to retinoic acid, it makes sense
that preformed supplemental vitamin A must be teratogenic at some unknown dose that is "too
h i g h . The cohort study described above attempted to determine this threshold dose by
defining the high dose as >l0 000 IU of preformed vitamin A per day. Such a cutoff, if correct,
is of public health importance because this dose is so close to the upper range of normal dietary
intakes of pro-vitamin and preformed vitamin A, and because women of reproductive age have
been advised to take a daily multivitamin supplement to prevent malformations. Recent studies
strongly suggest that periconceptional supplements of vitamin A that are close to, but less than
10 000 IU/day, and that are given as a component of a multivitamin, are much more likely to
be associated with reduced, rather than increased, risk of malformations.
Although many of the human teratology studies of vitamin A use have methodological
limitations, the preponderance of the data does not allow an inference to be made that 10 000
IU/day is a threshold. Human teratogenic effects from supplemental vitamin A intake
presumably result from increased maternal blood levels of retinoic acid. However, preliminary
results presented at this consultation (U. Wiegand, personal communication) of human
pharmacokinetic studies using various supplemental doses of preformed vitamin A from 10 000
to 30 000 IU/day do not show significant increases of maternal serum levels of retinol or
retinoic acid. This lack of evidence of any significant change in serum levels after
supplementation with commonly available unit doses of preformed vitamin A makes it highly
unlikely. that such supplements are teratogenic in humans. Further pharmacokinetic research
will be valuable in providing information on how data on vitamin A doses can be more
biologically modelled to reflect changes in maternal blood levels that can potentially be induced
with supplements. This will be useful to epidemiological research.
Mechanism of ieraiogenic action
Discussion of the background paper highhghted several pertinent issues. There was agreement
that the mechanism by which vitamin A exerts teratogenic effects is now understood to be
mediated through the influence of high concentrationsof some retinoic acid metabolites on gene
function at critical periods of organogenesis and embryonic development. This can occur upon
exposure to the spike in circulatingchylomicra-solubilized retinyl-esters that follows a mother's
ingestion of a large dose of preformed vitamin A during the first 6 weeks following conception.
Potential teratogenicity of metabolites is related to the area under the blood-level curve and the
half-life of the metabolite. Primate and human pharmacokinetic studies show that risk of
teratogenicity from high blood levels of vitamin A metabolite, all-trans and 13-cis-retinoic acid,
after single-dose supplementation decreases after 5 days. Within this period, circulating
chylomicra-solubilized retinyl ester levels are reduced by depositing retinol in reserve tissues,
primarily the liver, and metabolite levels return to a physiological range. Retinol in liver
reserves-following deesterification-is bound to its carrier protein and subsequently
mobilized under tight homeostatic control, further complexed with transthyretin, to maintain
circulating levels of the trimolecular complex between 300-800 pg/l over a wide range of liver
reserves and usual dietary intakes. As a consequence, materno-fetal transfer of RBP-R is well
regulated and concentrations of endogenously generated, potentially teratogenic metabolites
remain physiological. As noted earlier, epidemiologic evidence for teratogenicity comes
primarily from studies in which preformed teratogenic retinoic acid metabolites were directly
ingested.
Threshold levels for ieraiogenicify
There are no available data showing to what extent circulating levels of potentially teratogenic
retinoic acid metabolites become elevated, and for how long they remain elevated, when single
doses of 100 000- 200 000 IU (30 000-60 000 pg RE) retinyl ester are given, i.e. when there could
be exposure to spikes of potential teratogens. These dosage levels are commonly used to
supplement young children and non-pregnant lactating mothers. It is not clear whether the
potential teratogenicity of metabolites is correlated better with the area under the blood-level
curve or with peak blood levels. Research is needed, therefore, to determine the
pharmacokinetics of metabolites of vitamin A following high doses of retinyl esters given to
women who enter pregnancy with elevated, as well as low, vitamin A stores. It is possible to
conduct such research in women of reproductive age while they are protected from becoming
pregnant.
For obvious ethical reasons, prospective studies of maternal dosage levels associated with
teratogenicity cannot be undertaken in humans. The closest approximation is non-human
primate studies. Participants were apprised of ongoing studies in which retinyl palmitate
clearly had teratogenic effects -60% abnormal fetuses -at an exposure level of 80 000 IU/kg
(24 000 pg RE/kg). The threshold in monkeys replete with vitamin A-8% of fetuses showed
minor defects- was close to 20 000 IU/kg (6000 pg RE/kg). The teratogenic dose in primates,
therefore, is 20-30 times the human RDA per day during pregnancy (800 pg RE), or the
equivalent of 50 000-80 000 IU/day (15 000-24 000 pg RE/day) for pregnant women.
Nevertheless, participants cautioned that exposure to a daily dose might be less toxic than a
single high dose because of metabolic adaptation. Neither teratogenicity nor toxicity is observed
in species exposed to high-dose supplements of beta-carotene.
Unpublished and published research was presented regarding systemic concentrations of

vitamin A and its metabolites in female volunteers who received daily oral doses of vitamin A
for three weeks. Systemic concentrations of vitamin A metabolites (retinoic acid and 13-cis
retinoic acid) following doses of 10 000 and 30 000 IU were compared with the concentrations
in women not receiving vitamin A supplements during the first trimester of pregnancy.
Systemic concentrations after 10 000 IU were not signhcantly different from those before intake
and were well within the physiological concentrations in pregnant women who were not
supplemented during the first trimester and who gave birth to healthy babies. Even after doses
of 30 000 IU, concentrationsin serum were only slightly elevated, and just beyond the range of
physiological concentrations (2).
The recently reported study by Rothman et al. (3) was extensively discussed because it has
raised concern internationally. Results suggested that there is an increased risk of birth defects
(1 in 57, but with a wide range down to 1in 300 births) among women consuming more than
15 000 IU (4500 pg RE) vitamin A daily in their total diet, or above 10 000 IU (3000 kg RE) from
a supplement alone. The period of ingestion associated with birth defects was obtained
retrospectively and was described as between 2 weeks before and 7 weeks after conception. No
birth defects were seen in mothers with similar levels of ingestion after 7 weeks of pregnancy.
These doses are low enough for there to be concern about the potential teratogenicity of doses
within currently recommended ranges. The legitimate criticisms of the Rothrnan et al. study
raised in published letters to the editors of New England Journal of Medicine and The
Lancet -about poor quantitative measurement of ingested vitamin A, inappropriate assignment
of birth defects associated with vitamin A, and bias in the sampled population -do not explain
away the results of the study. Indeed, some of these criticisms would tend to mask results
thereby implying that in reality the relationship is even greater than reported. The Rothman
et al. study, therefore, needs to be examined within a broader context.
In view of the dose-response metabolite data, reviewed earlier, the consultation pointed out that
the production of retinoic acid metabolites necessary to produce the level of teratogenicity
reported does not occur at vitamin A dosage levels of 10 000-15 000 IU (3000-4500 pg RE), but
only at levels above 30 000 IU (9000 kg RE). Hence, a three-fold higher intake than that
suggested by the Rothman et al. study did not raise systemic concentrations of vitamin A
metabolites sufficiently to cause a teratogenic effect. This raises the question of whether the
doses that caused the teratogenicity in the Rothman et al. study were in fact much higher (e.g.
extrapolation from the study's actual findings to a 100 000 IU [30 000 pg RE] threshold leaves
large opportunity for error). Participants in the consultation therefore questioned the reliability
of theevidence from the Rothman et al. study regarding increased teratogenic risk from regular
intake of a 10 000 IU (3000 pg RE) supplement during pregnancy since even a two-to-three-fold
higher intake did not raise potentially teratogenic metabolite levels in the blood.
Participants also pointed out that, in vitamin A replete women, added dosing results in more
retinoic acid production. In contrast to much of the developing world where inadequate
vitamin A intake is prevalent, the women in the study by Rothman et al. were from a
population where VAD rarely occurs and, therefore, they were likely to be replete in vitamin
A stores. Newly ingested vitamin A is unlikely to endure in the circulation when the liver stores
are depleted, but circulating elevated esters may persist when the liver is replete. No added
benefit is expected from vitamin A supplementationin replete situations, so there is no rationale
for such supplementation whether or not there is a risk.
Evidence from regional registries of birth defects
An additional point that was noted was the lack of consistency of the conclusions of Rothman
et al. with findings from other regional registries of birth defects in Europe and the USA. In
Europe, a registry was started in 1996 by the European Network of Teratology Information
Services (ENTIS) in Rome. In the USA, registries are maintained at the Centers for Disease
Control and Prevention in Atlanta, Georgia, and in California. These registries have recorded
no increase in birth defects that could be attributed to excess intakes by pregnant women of
preformed vitamin A, i.e. regular intakes of retinol or retinyl palmitate above 10 000 IU (3000
pg RE) from multivitamin and/or single vitamin A supplements.
In response to the controversy stimulated by the Rothman et al. article, ENTIS conducted an
analysis of pregnancy outcomes observed prospectively among a cohort of women counseled
in recent years by the 11participating teratology information services that used vitamin A
supplements in excess of 10 000 IU (3000 pg RE) or more before 70 days of gestation. The study
cohort was compared to two convenience control cohorts, one made up of women exposed to
vitamin A supplements beyond 70 days gestation and the other made up of women with non-
teratogenic drug exposures. Detailed exposure information was collected directly during the
risk assessment consultation. A structured survey questionnaire on pregnancy outcomes,
addressed to attending doctors or women with infants who had cranio-facial defects, was
administered by mail or by telephone 2-8 weeks after expected delivery dates.
Among women exposed to 10 000 IU or above before 70 days of gestation, 10 out of 302 women
(3.3%) delivered babies with birth defects compared to 6 out of 113 women (5.3%)who were
exposed to the same amount after 70 days. However, only one case was attributable to cranial
nerve crest origin and this occurred in the group where exposure was at a level of 100 000 IU
(30 000 pg RE) after 70 days gestation. There was no statistical difference in the occurrence of
birth defects between exposed and control groups. In addition, there was no evidence of a
relationship between increased vitamin A exposure and spontaneous and induced abortion. It
is noteworthy that there was a lack of any indication from the European registry of congenital
anomalies attributable to exposure to vitamin A in excess of 10 000 IU (3000 pg RE). However,
the sample size in this study is insufficient to validate or discard confidently the results from
the study by Rothman et al.
Little information is available on safe levels of vitamin A in humans exposed at weekly or
monthly intervals at different periods of gestation. Participants saw no reason to question the
safety of the current recommended exposure for women of reproductive age of 10 000 IU (3000
pg RE) daily throughout pregnancy. Participants also agreed that, after the first trimester,
exposure to levels in excess of 10 000 IU of preformed vitamin A or retinoic acid is less risky.
However, caution is advisable because some preliminary data in humans suggest mild
abnormalitiesin CNSrelated performance (e.g. speech and language, and verbal IQ tests) from
exposure to 13-cis retinoic acid during the fetal period of development. These observations,
which need verifying, would extend the period for possible toxic exposures into the second
trimester of pregnancy.
The effects of supplement exposure could differ between well nourished and poorly nourished
women, and according to whether exposure is from a concentrated supplement or a food
source. Participants noted recent data that indicate less risk of elevated circulating metabolites
from concentrated food sources of vitamin A, such as animal liver, than from supplements (4).
These findings-which need to be confirmed -reinforce the current recommendation for safely
improving the vitamin A status of pregnant women from dietary food sources, including
animal liver, where this is programmatically feasible.
On the basis of available data, participants concluded that there is no teratogenic risk from
preformed vitamin A supplement of 10 000 IU (3000 pg RE) given to pregnant women who
habitually consume less than the RDA. However, there is no justification for daily supplements
at a level above 8000 IU (2400 pg RE) for pregnant women who habitually consume vitamin A
at the level of the RDA (800 pg RE) or above. A weekly supplement of 25 000 IU (7500 pg RE)
given to women who regularly under-consume vitamin A is unlikely to produce peak levels of
blood metabolites above their physiological ranges or to exceed tissue storage potential. On the
contrary, the potential benefits derived from correcting maternal deficits outweigh the risks.
Determinants of the reiurn of fertility

From a programmatic perspective, the benefits that vitamin-A-deficient women and their
nursing infants can obtain from periodic exposure to high-dose preformed vitamin A
supplement must occur within a "safe period during which there is contact with the health
system. Such contact varies with the health service infrastructure. It is important, therefore, to
identdy the determinants of postpartum infertility and the practical indicators available in the
field for identdying the risk of the return of fertility. The relevant issues are summarized below.
Physiological determinants of ihe reiurn of fertility postpartum

and the duration of lactational infertility in developing countries
(Dr Kaihy I. Kennedy and Dr Alan S. McNeilly)
Lactational infertility depends on the neuroendocrine stimulus of suckling. Suckling affects
ovarian activity by suppressing the normal pattern of release of the gonadotropin-releasing
~ A - - d . d & - ~ -
. . Pppu-."
hormone (GnRH) in the hypothalamus and by increasing the sensitivity of the hypothalamus
and pituitary to the negative effects of estradiol. As suckling declines, there is a progressive
escape from this inhibition, a gradual return to pulsatile secretion of GnRH/luteinizing
hormone (LH), and eventually a return to the positive effects of estradiol that are necessary to
generate the pre-ovulatory LH surge. The pattern of suckling plays a crucial role in maintaining
the disruption of the release of GnRH and the consequent suppression of fertility.
The onset of the first postparhun menses is a clear indication that ovarian activity has occurred,
that fertility is returning, and that a family planning method should be used if pregnancy is to
be avoided. During lactational amenorrhoea there is, by definition, no menstruation. Bleeding
of any sort is due to a withdrawal of steroid support to the uterine endometrium after a period
of exposure to increased steroid. Thus any bleeding during lactation is indicative of ovarian
activity or, more accurately, the end of a period of steroid secretion by the ovary. In the absence
of ovulation there is no corpus luteum, no progesterone secretion, and hence no progesterone
withdrawal to precipitate the normal menstrual bleeding, although bleeding of shorter than
normal duration may occur after withdrawal of estradiol. If the first ovulation precedes the first
menstruation during breast-feeding, it is usually associated with abnormal function of the
resulting corpus luteum, which secretes progesterone in levels that are too low to support a
pregnancy.
The Bellagio Consensus was achieved in 1988 on the basis of data about the return of fertility
in 13 prospective studies in eight countries. The consensus states that "...the maximum birth
spacing effect of breastfeeding is achieved when a mother 'fully' or nearly fully breastfeeds and
remains amenorrhoeic ...When these two conditions are fulfilled, breastfeeding provides more
than 98 % protection from pregnancy in the first six months." Prospective clinical trials have
been carried out to test the indicators identified in reaching the Bellagio Consensus, and the
consensus has been operationalized as the lactational amenorrhea method (LAM) of family
planning. These trials involved women who were actively using LAM to avoid pregnancy. The
studies found that the rate of pregnancy at the end of six months postpartum in amenorrhoeic,
fully breast-feeding women was indeed less than 2%.
A prospective study of over 4000 women in seven countries was conducted by WHO to study
infant feeding practices and the return of menses. It was not a study of women using LAM, but
it found that the rate of pregnancy was less than 1%at the end of six months among breast-
feeding women who perceived that they were still amenorrhoeic, regardless of how much or
how little food supplement they were giving to their infants.
Thus, it was determined at a follow-up meeting in 1995, that "the Bellagio Consensus clearly
has been upheld by subsequent prospective research.
A large array of factors secondary to the breast-feeding stimulus have been assumed to affect
the duration of lactational infertility. Among the many potential factors are characteristics of
the infant, such as general health (including gestational age at delivery, size/weight and
nutritional status) and the intensity of suckling. Potential maternal factors include age, parity,
ethnicity, employment and psychological, environmental and socioeconomic factors. Maternal
nutritional status is associated with lactational infertility, but probably acts by mediating the
stimulus of suckling.
Data from the DHS in 47 countries are available on the percentage of women still breast-feeding
at increasing times postpartum. Such data help in making informed decisions about the risk of
pregnancy in developing countries. Summary data by region presented in Table 2 illustrate the
wide ranges encountered among countries even within the same region.
Table 2. Ranges of percentage of women still breast-feeding, amenorrhoeic, abstaining and

insusceptible to pregnancy, by region and time since delivery
Breast-feeding Amenorrhoeic Abstaining Insusceptible
2-3 months postpartum
Sub-Saharan Africa 87-100 73-93 12-93 84-99
Asia/Near East/
North Africa
Latin America 61-96 46-90 22-55 53-93
4-5 months uostuartum
Asia/Near East/
North Africa
Latin America 45-95 19-82 8-35 32-85
6-7 months postpartum
Asia/Near
69-99
East/North Africa
Latin America 35-95 8-67 35726 15-73
The data summarized in Table 3 show the positive relationship between the breast-feeding
stimulus and the duration of lactational amenorrhoea that has been observed prospectively in
clinical studies. The longest durations of both breast-feeding and lactational amenorrhoea are
generally observed in sub-Saharan Africa. In no sub-Saharan African country studied was the
median duration of breast-feeding less than 17 months, or the median duration of lactational
amenorrhoea less than 8 months. Only in Latin America were median durations of breast-
feeding observed to be less than one year. However, there are likely to be subgroups of women
with short durations of both lactation and infertility, as well as subgroups with long durations,
in all countries.
The level of breast-feeding stimulation required to induce lactational infertility appears to vary
widely. A practical measure of a woman's natural capacity to respond to the breast-feeding
stimulus is the duration of her previous lactational amenorrhoea, which obviously excludes
women breast-feeding their first child.
Clinical studies of lactational infertility, as well as the DHS data, suggest that strategies for
vitamin A supplementation should vary from region to region and country to country, just as
the duration of lactational arnenorrhoea or postparturn insusceptibility varies. Within a regional
or country strategy, advice to individual women to use contraception before or when their
LAM protection expires may be an appropriate adjunct to vitamin A supplementation.
Table 3. Ranges of median durations of breast-feeding, amenorrhoea,

abstinence and insusceptibility to pregnancy by region
Breast-feeding Amenorrhoeic Abstaining Insusceptible
Sub-Saharan
17-28 months 8-17 months 1-19 months 13-22 months
Africa
Asia/Near East/ 12-36+ months 4-10 months 1-4 months 4-11 months
North Africa
Latin America 4-18 months 3-12 months 2-5 months 3-13 months
HRP-sponsored muliicentre study
Participants in the consultation reviewed the unpublished results of a seven-country

multicentre study carried out in Australia, China, Chile, Guatemala, India, Nigeria and Sweden
using a standardized protocol to examine the duration of lactational amenorrhoea in relation
to breast-feeding practices. Highly significant centre effects were seen in the duration of
lactational amenorrhoea. In addition to the centre differences, 10 factors were significantly
related to the duration of amenorrhoea: seven of these were infant feeding characteristics, and
the remaining three were parity, body mass index and frequency of infant illness. A strong
correlation was found in all centres between the length of amenorrhoea when breast-feeding
the previous child and the duration of lactational amenorrhoea in the study.
Among the cohort of women with previous breast-feeding experience, median durations of
arnenorrhoea ranged from 122 days in India to 282 days in China. Cumulative pregnancy rates
during breast-feeding and lactational amenorrhoea (woman's perception) from all centres were
0.1% (95%C1 = 0.0-0.3) at day 112 (3.5 months) postpartum, and 0.8% (95%C1 = 0.2-1.4) at day
182 (6 months).
DHS data on risk of concepiion
Data from DHS surveys in 27 countries were evaluated from a different perspective to address
the question of risk of subsequent conception following a birth under different feeding regimes.
Three approaches to data analysis were used, namely:
the duration of "closed birth intervals" assuming a pregnancy duration of 9 months;
construction of tables for the probability of conception according to breast-feeding
status;
susceptibility to conception.
Each approach has limitations, but the inaccuracies were thought to give a conservative bias.
Consideration of the data from these viewpoints indicated that the "best case" scenario of
exclusive and full breast-feeding showed an approximately 2% conception susceptibility as early
as two months postpartum. Participants emphasized that susceptibility was not equivalent to
probability of conception, i.e. many physiological and social factors can decrease the likelihood
of conception following the first postpartum ovulation. For example, up to 80% of first
postpartum ovulations do not result in a viable pregnancy even if conception occurs, and
cultural factors sometimes dictate extended periods of postpartum sexual abstinence.
Country experience
Bangladesh provided an example for situational analysis and evaluation of the length of the
postpartum non-conception period under real community conditions. In Bangladesh, with a
population of 120 million, infant mortality is about 78 per 1000 live births, low birth weight
occurs at a rate of 40%,contraceptive use rates are 46%among married women (median age at
marriage is 18.3 years), mean duration of breast-feeding is 28.6 months and postpartum
amenorrhoea averages nearly 12 months. A review of experience from longitudinally collected
data on fertility return from the Matlab study area confirmed a reduced risk of the return of
fertility among fully breast-feeding women. For breast-feeding women, the overall risk of
conception was 4 per 10 000 at 6 months postpartum. However, among women who abruptly
stopped breast-feeding, most because their infants died, the cumulative probability of
conception was 30 per 10 000. Hence, the return to fertility was much faster when breast-
feeding was abruptly interrupted in the early postpartum lactational period. This confirms in
a practical setting the experimental findings reviewed in the background paper.
Participants agreed that the available data indicate that suckling is the stimulus that drives the
mechanism that postpones the recovery of fertility. Breast-milk substitutes or supplements
affect the return of fertility to the extent that they decrease the suckling stimulus. Among
women in affluent societies, "supplementation" is often a substitute for breast-feeding and the
overall amount of suckling is decreased, leading to an earlier return of fertility. Less affluent
women may eventually supplement breast-feeding however, they may or may not concurrently
reduce suckling.
The importance of exclusive breast-feeding (i.e. nothing else but breast-feeding) and full breast-
feeding (i.e. minor amounts of low-caloric additions) in extending the time to fertility return is
confirmed on a population basis across a wide variety of communities. Although the three
indicators associated by the Bellagio Consensus with less than 2% risk of pregnancy-
amenorrhoea, the first 6 months postpartum, and full or nearly full breast-feeding -are reliable
guidelines for populations, there is no field-applicable certain indicator of infertility on an
individual basis. However, individual risk will be low if the breast-feeding woman is in
compliance with the Bellagio guidelines. These guidelines provide, therefore, simple indicators
that could be used by field-workersin areas of endemic VAD to screen women of reproductive
age who are eligible to receive a high-dose vitamin A supplement but who are in contact with
the health services only beyond 2-3 months postpartum.
For women who are not breast-feeding there are limited data on which to base an estimate of
the infertile postpartum period. Although it often takes 5-10 weeks for normal fertility to
, .p--
return, the first ovulation can occur within 27 days of delivery. The cycle of such an early
ovulation will occasionally have normal characteristics.
Policy and programme implications
Participants identified four scenarios for giving vitamin A supplements, through public health
programmes, primarily to improve the vitamin A status of the infant before six months of age
in order to realize the benefit of protection against morbidity and mortality beyond six months.
The scenarios, each of which have implications for safe dosage and frequency of administration,
were:
1. Maternal supplementation during pregnancy for mothers whose habitual intakes
are above the RDA or below the RDA, from 1-60 days following conception, and
after 60 days following conception.
2. Supplementation'for mothers in the first six months postpartum.
3. Direct supplementation of infants before six months of age.
4. Supplementation both for mothers during the "safe" infertile postpartum period
and for infants under six months of age.
Working groups were formed to discuss each of the scenarios and to reach consensus on safe
doses and timing-so that policy and programmes could be drawn up without waiting for
additional research-and to identify research required before current guidelines should be
altered. The consensus views on recommendations and research needs for each of the scenarios
were as follows.
Recommendations on doses and timing
1. Maternal supplementation during pregnancy

(Either during thefirst 60 hysfbllowing conception when there is a teratogenic risk or after thefirst 60
daysfollowing conception,for women whose habitual intakes are above the RDA or below the RDA)
For fertile women, independent of their vitamin A status, 10 000 IU (3000 pg RE) is the
maximum daily supplement to be recommended at any time during pregnancy.
Where VAD is endemic among children under school age and maternal diets are low in vitamin
A, health benefits are expected for the mother and her developing fetus, with little risk of
detriment to either, from:
either a daily supplement not exceeding 10 000 IU vitamin A (3000 pg RE) at any
time during pregnancy;
or a weekly supplement not exceeding 25 000 IU vitamin A (8500 pg RE). In this
regard:
a single dose > 25 000 IU is not advisable, particularly between day 15 and day 60
following conception (day 0);
beyond 60 days after conception, the advisability of providing a single dose of
> 25 000 IU is uncertain; any risk for non-teratogenic developmental toxicity is likely
to diminish as pregnancy advances. In the case of a pregnant woman who may be
reached only once during pregnancy, health workers should balance possible
benefits from an improved vitamin A status against potential risks of adverse
consequences from receiving a supplement.
Where habitual vitamin A intakes exceed at least three times the RDA (about 8000 IU or 2400
yg RE), there is no demonstrated benefit from taking a supplement. On the contrary, the
potential risk of adverse effectsincreases with higher intakes -above 10 000 IU -if supplements
are routinely ingested.
2. Supplementaiion for mothers In the first six months postpartum
(Single high-dose supplement above 25 000 IU, and usually at a level of 200 000 IU, during the safe
period ofpostpartum infertilityfor mothers in vitamin-A-d&cient areas)
At the population level

Mothers who are not breast-feeding will benefit from a high-dose supplement given safely
during the first 28 days (4 weeks or 1 month) postpartum. Although the risk of conception
beyond this point is poorly documented, normal fertility does not usually return for 5-10
weeks. Beyond 6 weeks, therefore, non-lactating mothers should be given no more than 10 000
IU daily. Direct supplementation of the non-breast-fed infant 6 months of age, who is not
given a fortified breast-milk substitute, with as much as 50 000 IU (15 000 pg RE) is the
recommended safe intervention to meet the infant's need for vitamin A.
Mothers who are breast-feeding will benefit from a high-dose supplement given up to 60 days
(8 weeks or 2 months) postpartum, as will their nursing infants - through higher levels of
vitamin A in breast milk. The risk of pregnancy is related to menstrual status of the breast-
feeding mother. If she has resumed menstruation she is regarded as fertile. If she is
arnenorrhoeic, the risk of pregnancy rises after 60 days, in some circumstances reaching 1%-2%
by 6 months postpartum. In some very high-risk areas (e.g. where there is a high prevalence of
clinical symptoms of VAD in mothers) where a high percentage of contact with the health
system occurs within 8-12 weeks, the risk of extending the period for giving a high-dose
supplement from week 8 to week 12 (estimated 2.8% conceptions) might be offset by important
benefits to the mother (relief of symptoms) and the nursing infant (increased breast-milk
vitamin A consumption and consequent increased likelihood of decreased risk of mortality).
For individuals
Mothers who are not breast-feeding, provide a supplement within 28 days of delivery;
otherwise give a supplement directly to the infant.
Mothers who are breast-feeding:

if using reliable contraception, give supplement anytime;
if practising postpartum abstinence, give supplement anytime;
if amenorrhoeic, give supplement up to 6 months postpartum;
- Safe vitamin A dosage during pregnancy and lactation
-
if not amenorrhoeic, give supplement to mother at time of next menstruation (an
indication that conception has not occurred) or give supplement to child.
3. Direct supplementation of infants before six months of age

(In areas of endemic vitamin A deficiency)
Firm evidence of benefits to breast-feeding infants of direct supplementation before six months
of age is insufficient. Studies are in progress to clarify the benefits/risks of single
supplementation at 50 000 IU (15 000 pg RE) at birth or thereafter, or multiple supplementation
at 25 000 IU (7500 pg RE).
Infants who are not breast-fed and who are not given fortified breast-milk substitutes should
receive a 50 000 IU supplement, preferably by about 2 months of age-otherwise at any time
within the first 6 months of life. As an alternative, two doses of 25 000 IU can be given with an
interval of a month or more in-between.
4. Supplementation both for mothers during the "safe" infertile

postpartum period and for infants under six months of age
There is currently insufficient information to make firm programmatic recommendations

regarding risks and benefits to infants on the basis of this supplementation strategy. A WHO-
sponsored trial is being conducted in three countries to clarify this issue and results should be
available within the next year.
Research needs
1. Direct supplementation of infants

Comparative study of effects of a single 50 000 IU dose given before 6 weeks versus multiple
doses (25 000 or 50 000 IU) given in accord with vaccination schedules at approximately 6,10
and 14 weeks.
Outcomes to be evaluated:
mortality/morbidity reduction and vitamin A status.
2. Direct supplementation of mother up to 8 weeks postpartum (possibly longer where
infertility is assured)
Effect of high-dose maternal supplementation (200 000-300 000 IU) on reduction in infant
mortality/morbidity up to three years of age, and on maternal vitamin A status.
acute effects in mother, i.e. serum and milk retinol and retinol metabolites;
long-term effects on mortality/morbidity and vitamin A status in infancy up to
three years of age, and in mothers;
effects of partial weaning/cessation of breast-feeding on mortality, morbidity
and vitamin A status and return of fertility.
3. Direct supplementation of both mother and infant (the consultation noted that WHO-
sponsored studies are in progress in three countries using the following protocol)
Effect of 200 000 IU vitamin A supplement to mother before 8 weeks postpartum and 25 000
IU to the infant in accordance with vaccination schedules at approximately 6,10 and 14
weeks.
mortality/morbidity of infants up to three years of age, and vitamin A status
of both infant/young children and mothers.
4. Risk/benefit analysis:
benefit of 200 000 IU vitamin A supplement given to mother at birth on reduction in
mortality at birth, at 2 months and at 4 months (followed until three years of age
where possible) versus supplementation of infants only after 6 months;
benefit to the mother in terms of reduction of incidence of diarrhoea and night
blindness.
5. Pharmacokinetic studies using preformed vitamin A and retinoic acid metabolites given to
non-pregnant women residing both where there is a history of VAD and where intakes are
known to be adequate, using:
daily doses of 10 000 IU;
weekly doses of 25 000 IU.
6. Developmental psychological studies of children who were exposed in utero to doses of
preformed vitamin A 225 000 IU, and those exposed to weekly doses of vitamin A (25 000
IU) via breast milk. W l e acknowledging that such research is methodologically
challenging, the results are needed before routine daily dosages >l0 000 IU or weekly doses
>25 000 IU can be recommended at any time during pregnancy.) Ongoing human studies
of the behavioural teratogenesis of 13-cis-retinoic acid will provide guidance in selecting
an efficient and specific battery of testing tools for use in such research.
7. In countries with adequate vitamin A status, development of systems for monitoring
vitamin A content of foods with high concentrations (animal livers) and associated risks of
human hypervitaminosis and of teratogenicity in pregnant women.
8. Exploratory research to develop a new type of supplement that has advantageous
absorption characteristics such as those reported for foods like liver, i.e. delayed absorption
with reduced levels of teratogenic retinoid metabolites.
9. Studies on maternal factors that optimize transfer of vitamin A to breast milk (e.g. effect
of other nutrient deficits or disease and mammary gland concentrations of vitamin A
relative to circulating levels).
10. Level and duration of elevated retinoic acid metabolites in breast milk following high-
dose maternal supplementation.
Coniad points with the health services
Country-specific information on the contacts of mothers and infants with the health services,
and the coverage of these services, is important for selecting appropriate supplementation
programmes. Data available from WHO'S Maternal and Newborn Health/Safe Motherhood
Rograrnrne have been used as the basis of the summary global estimates contained in Table 4
for the percentages of antenatal care contacts, place of birth, and level of training of birth
attendants.
Table 4. Global estimate of the percentage of births receiving antenatal care,

the place of birth, and the level of training
- of home birth attendants
Births in Antenatal
Place of birth
millions care (%)
Institution
Home (%)
(%)
Trained birth Untrained
attendant birth attendant
World 141 64 44 16 40
More
developed
Less developed 126 59 37 18 45
Asia 83 57 33 23 44
Africa 31 59 34 21 45
Latin America 12 72 66 10 24
Additional information on delivery by a medical professional (physician or nurse), contacts with

infants under 2 months of age, and contacts with mother/infant within 8 weeks of delivery was
provided by DHS surveys from 27 countries. The data are summarized by region in Table 5.
Table 5. Estimates from 27 countries, by region, of contacts
with the health services from birth and up to 8 weeks
Age 2 Contact 5
Region Health professional (%l months at 8 weeks
vaccination after birth
Physician Nurse Total (X) (% )
Africa 6.2 34.3 33.0 59.1
(14 countries) (0.3-14.1) (10.0-56.4) 40'5 (19.2-59.9) (23.0-79.8)
Latin America 42.0 17.2 25.7 75.8
(6 countries) (10.5-70.9) (3.2-31.9) 59'2 (13.9-40.0) (52.3-94.7)
South-East Asia
5.1 30.1 35.2 7.9 39.7
(Indonesia)
Eastern
25.5 16.3 17.8 55.5
Mediterranean 41.8
(5.9-50.3) (7.2-37.0) (2.8-43.2) (31.9-88.1)
(4 countries)
Western pacific
26.0 26.8 52.8 17.2 59.0
(Philippines)
Generic scenarios for populations where supplementation

programmes are needed
Although limited, contact data are useful in developing generic scenarios to assist in
programme planning. The scenarios in Table 6 are suggested only as a guide. Country-specific
situation analysis is required before country-specific programme decisions can be made.
Table 6. Suggested situation analysis guideline for selecting appropriate
countrv-specific vitamin A su~~lementation Droarammes
Case scenario number*
1 2 3 4 5 6
non-breast-
Breast-fed infants
fed infants
Adolescent weekly/ - - - - -
girls monthly
1st antenatal daily/ - - daily/ - -

care visit weekly weekly
(mother)
Delivery kit 200 000 IU - 200 000 IU - - -

(mother) at delivery at delivery
1st postpartum - - 200 000 IU - 50 000 IU

contact at C60 days to infant
post-
partum to
lactating
mother
BCG contact - 200 000 IU - - 25 000 IU -

within 2 to mother to infant
months
Child DlT 1& 25 000 IU 25 000 IU 25 000 IU 25 000 IU 25 000 IU 25 000 IU

3rd contact (alternative
to single
larger
dose)
Routine 100 000 IU 100 000 IU 100 000 IU l00 000 IU l00 000 IU start at 12
supplement to at 6-12 at 6-12 at 6-12 at 6-12 at 6-12 months at
child 2 6 months, months, months, months, months, 200 000 IU
months at 4-6 200 000 IU 200 000 IU 200 000 IU 200 000 IU 200 000 IU
monthly afterwards afterwards afterwards afterwards afterwards
intervals
-
Measles 25 000 IU 25 000 IU 25 000 IU 25 000 IU 100 000 IU l00 000 IU

immunization
contact (if no
routine
supplement)
*Descriptionof case scenarios for areas with endemic vitamin A deficiency:
1. "Ideal" situation, i.e. the highest recommended supplementation schedule (starts early
with a woman of reproductive age, continues through pregnancy, and extends through
infancy and the child's vulnerable years).
2. Breast-feeding and supplementation linked to vaccination schedule.

3. Preferred over case 2 if safe delivery kits are available because of probable wider
coverage.
4. Good alternative because it covers the mother during early pregnancy as well as the
child's early years.
5. Targeted exclusively at immunization and infant. Although this misses the mother, this is
the situation in many developing countries.
6. A difficult situation in which the infant should be supplemented at the earliest contact
and should receive doses linked with vaccination schedule as well.
References
1. Buss et al. Human & Experimental Toxicology, 1994,13:33-43.

2. Wiegand W, Hartmann S, Wyss R. Endogenous concentrations of retinoids in pregnantfemales
and their relevance for the risk assessnzent of vitamin A. Presentation at the Symposium on
Pharmacokinetics/Pharmacodynamics in the Developing System and Impact on Risk
Assessment, Little Rock, AR, April, 1996.
3. Rothrnan et al. New England journal of medicine 1995; 333:1369-1373.
4. Buss et al., op. cit.
Annex 1
List of participants
Dr Christopher J. Bates Professor Jean-PierreHabicht

Head of Micronutrient Support Facility Division of Nutritional Sciences
MRC Dunn Nutrition Unit Savage Hall
Milton Road Cornell University
GB-Cambridge CB4 1XJ Ithaca, NY 14853
Tel: 44-1223426-356 USA
Fax: 44-1223-426 617 Tel: 1-607-255-4419
e-mai1:chris.ba [email protected] Fax: 1-607-255-2608/607 255 1033
e-mail: [email protected]
Dr VCronique Braesco
ChargCe de Recherche Dr Najia Hajji
Responsable Pquipe "vitarnine" Head of Family Planning
Institut National de la Recherche Agronomique Ministry of Health
(INRA) Rabat, Morocco
CRNM - BP 321 Fax: 212-769.10.82
58 rue Montalembert Or c/o WR - Morocco
63009 Clermont-Ferrand Tel: 690 510 Rabat
France
Tel: 33 73 60 82 70 Dr Kathy I. Kennedy
Fax: 33 73 60 82 72 Maternal and Child Health Research
2201 South Fillmore St.
Dr Anna Coutsoudis Denver, CO 80210
Senior Scientist USA
University of Natal Medical School Tel: 1-303-758-5494
Department of Paediatrics and Child Health Fax: 1-303-758-5660
P.O. Box 17039 e-mail: [email protected]
Congella 4013
South Africa Dr Edward Lammer
Tel: 27 31 260 4489 Department of Medical Genetics
Fax: 27 31 260 4388 Children's Hospital
e-mail: [email protected] Oakland, CA 94609-1809
USA
Dr Luis Andres de Francisco Serpa Tel: 510-428-3550
Project Director, MCH-FP Project Fax: 510-450-4678
International Centre for Diarrhoea1 Disease e-mail: [email protected]
Research Bangladesh
G.P.O. Box No. 128 Dr Bo Lonnerdal
Dhaka, 1000 Professor of Nutrition and Internal Medicine
Bangladesh Department of Nutrition
Tel: 871-751-60 University of California, Davis
Fax: 880-2-883 116 or 880-2-886 050 Davis, CA 95616 USA
e-mail: andres%[email protected]. th Tel: 916 752-8347
Fax 916 752-3564
e-mail: bllonnerda [email protected]
Professor Pierpaolo Mastroiacovo Dr Kathleen M. Rasmussen
Professor of Preventive and Social Paediatrics Associate Professor
Chief of Birth Defects Unit Division of Nutritional Sciences
Paediatric Unit 111Savage Hall
Catholic University Cornell University
Rome, Italy Ithaca, New York 14853-6301
Tel: 33 81 344 - 370-1905 - 39-6 Tel: 1-607-255-2290
Fax: 33 81 211 - 370-1904 Fax: 1-607-255-1033 or 1-607-255-2290
e-mail: [email protected] e-mail: [email protected]
Dr JosephMulinare Dr Shea Rutstein

Chief, Prevention Section Deputy Director
Division of Birth Defects and Developmental Macro International, Inc., Suite 300
Disabilities 11785 Beltsville Drive
Centers for Disease Conh-ol and Prevention Calverton, Maryland 20705
4770 Buford Highway USA
Mail Stop F 45 Tel: 301-572-0950
Atlanta, CA 30307, USA Fax: 301-572-0999
Tel: 1-770-488-7190 e-mail: [email protected]
Fax: 1-770-488-7197
e-mail: [email protected]
Dr Prema Ramachandran
Adviser (Health)
Planning Commission
Government of India
Yojana Bhavan
New Delhi - 110001 India
Tel: 91 11371 4058
Fax: 91 11371 7681
WHO Secretariat
Dr F.S. Antezana, Assistant Director-General

Dr G. Clugston, Director, Programme of Nutrition
Dr B. Underwood, Programme of Nutrition
Mrs R. Saadeh, Programme of utrition
Mrs E. Ahrnan, Maternal and Newborn Health/Safe Motherhood
Dr R.J. Guidotti, Maternal and Newborn Health/Safe Motherhood
Dr J. Zupan, Maternal and Newborn Health/Safe Motherhood
Dr H. von Hertzen, Technology Development and Assessment,
Special Programme of Research, Development and Research
Training in Human Reproduction
Dr S. Khanum, Regional Adviser/Nutrition, South-East Asia Regional Office
Vitamin A is essential for normal maintenance and functioning r f
body tissues, and for growth and development. This is also the case
during pregnancy, when the fetus makes demands on the mother's
vitamin A stores, and during the postpartum period when the newborn
is growing rapidly and, in most cultures, depends on breast milk to
obtain adequate amounts mf the vitamin. Although the increased
requirement during pregnancy is relatively small, in many countries
where vitamin A deficiency (VAD) is endemic, women often
experience deficiency symptoms such as night blindness that
continue during the early period of lactation.
Beyrnd 4 to 6 morlths pestrartum breast milk from deficient mrthers
is likely to contain insufficient vitamin A to build
-or even maintain-vitamin A stores in nursing infants.
Providing a diet adequate in vitamin A-neither too little nor too much-
is the safest solution to meeting needs during pregnancy and lactation.
However, this is not easily accrmplished in situations r f poverty and
where food with appropriate vitamin A crntent is in short supply and/or
expensive. in such situations the recommended approach is to provide a
vitamin A supplement during pregnancy at a dosage and frequency
that will safely meet the needs of growing maternal and fetal tissue and will
potentially build maternal Lady stores in anticipation of lactation.
However, using hith-dmse vitamin A supplements ta build maternal stores
during pregnancy creates a dilemma because of the vitamin's potential
teratogenicity during the early stages of pregnancy.
WHO has received requests for programmatic guidance m the safe

use of vitamin A supplements by women of reproductive age.
Mew data have become available concerning the return of menses relative
to breast-feeding practices and country-specific contacts with the health system.
To review these data and other information, WHO convened a consultation
to consider both the safe dosage of vitamin A during pregnancy and
the first six months postpartum, and the relevant policy and programme
implications. This document, which will be of particular interest for
managers of national VAD prevention and control programmes, presents
the recommendations of participating experts in nutrition, teratology,
reproductive physiology and population-based surveys, who have experience
in both basic research and its public health applications.
Progran~~neof Nutrition
Fam~lyand Reproductive Health

A and Lactation: Safe Vitamin Dosage During Pregnancy

Uploaded by

Copyright:

Available Formats

A and Lactation: Safe Vitamin Dosage During Pregnancy

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

A and Lactation: Safe Vitamin Dosage During Pregnancy

Uploaded by

Copyright:

Available Formats

WHO/NUT/98.

Safe vitamin A dosage

World Health Organization

Safe vitamin A dosage

World Health Organization

This document is not a formal publication of the World Health Organization,

The views expressed in documents by named authors are solely

Report of the consultation: Safe vitamin A dosage during

Annex l List of participants ............................................. 33

CNS central nervous system

1. Maternal supplementation during pregnancy.

1. Maternal supplementafion during pregnancy

At the population level

Mothers who are breast-feeding:

3. Direct supplementation of infants before six months of age

supplementation at 50 000 IU (15 000 pg RE) at birth or thereafter, or multiple supplementation

4. Supplementation both for mothers during the "safe" inferfile

There is currently insufficient information to make firm programmatic recommendations

Obieciives of the consultatton

1. To review evidence of VAD during pregnancy and/or lactation from countries

Vitamin A status during pregnancy and lactatton

1RE = l p g all-trans retinol = 0.0035 pm01 = 3.33 IU

The relation of maternal vitamin A status to other conditions

Breast milk composition

Maternal diet influences the level of several vitamins

In Bangladesh, there is currently no national vitamin A supplementation programme that

Immunization-linked vitamin A supplementation programmes during infancy need to be

Countries where vitamin A intakes habitually

Teratogenicity of vitamin A in humans

Vitamin A and human developmental toxicity

A review of evidence of human developmental toxicity from excessive preformed vitamin A

Mechanism of ieraiogenic action

Threshold levels for ieraiogenicify

Unpublished and published research was presented regarding systemic concentrations of

Evidence from regional registries of birth defects

Determinants of the reiurn of fertility

Physiological determinants of ihe reiurn of fertility postpartum

Table 2. Ranges of percentage of women still breast-feeding, amenorrhoeic, abstaining and

Table 3. Ranges of median durations of breast-feeding, amenorrhoea,

HRP-sponsored muliicentre study

Participants in the consultation reviewed the unpublished results of a seven-country

DHS data on risk of concepiion

Policy and programme implications

Recommendations on doses and timing

1. Maternal supplementation during pregnancy

2. Supplementaiion for mothers In the first six months postpartum

At the population level

Mothers who are breast-feeding:

3. Direct supplementation of infants before six months of age

4. Supplementation both for mothers during the "safe" infertile

There is currently insufficient information to make firm programmatic recommendations

1. Direct supplementation of infants

Coniad points with the health services

Table 4. Global estimate of the percentage of births receiving antenatal care,

Additional information on delivery by a medical professional (physician or nurse), contacts with

Generic scenarios for populations where supplementation

1st antenatal daily/ - - daily/ - -

Delivery kit 200 000 IU - 200 000 IU - - -