![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
|
|
|
|
|
||
Program in International Nutrition, Department of Nutrition, University of California, Davis, CA 95616
2To whom correspondence should be addressed. E-mail: LHAllen{at}UCDavis.edu.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONSOURCES of VITAMIN AEVALUATING RISK OF EXCESSIVE...BIOLOGICAL INDICATORS OF...EVALUATING THE RISK OF...LITERATURE CITED |
|---|
KEY WORDS: • vitamin A toxicity • vitamin A kinetics • women • infants • lactation
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONSOURCES of VITAMIN AEVALUATING RISK OF EXCESSIVE...BIOLOGICAL INDICATORS OF...EVALUATING THE RISK OF...LITERATURE CITED |
|---|
|
SOURCES of VITAMIN A |
|---|
|
TOPABSTRACTINTRODUCTIONSOURCES of VITAMIN AEVALUATING RISK OF EXCESSIVE...BIOLOGICAL INDICATORS OF...EVALUATING THE RISK OF...LITERATURE CITED |
|---|
Humans normally obtain preformed vitamin A from animal products in the diet and provitamin A carotenoids (mainly ß-carotene) from fruits, vegetables and oils. Consuming natural sources of vitamin A rarely results in toxicity. The exception is toxicity resulting from excessively high intakes of carnivore liver on a continued basis. Liver contains 3000–5000 µg of retinol per 100 g. Although this level of intake is rare, a case study reported vitamin A toxicity symptoms in a 7-mo-old infant who was given 
12,100 µg of retinol/d, mostly in chicken liver (1
). The vitamin A content of this food is high (5000 µg/g).
Supplements
In industrialized countries, daily vitamin and mineral supplements are commonly consumed by women of reproductive age and children. An informal local (California) survey of such supplements showed that almost all provide 1500 µg of retinol activity equivalents (RAE)2 [5000 international units (IU)] in retinyl palmitate per day for adults, and 750 µg of RAE (2500 IU) for children age 2–5 y. Such supplements are less common in developing regions, but international organizations such as the United Nations Children’s Fund (UNICEF) are considering replacing the current iron-folate supplements with multiple micronutrient supplements, which are likely to contain approximately the recommended daily intake of preformed retinol.
In many developing countries, it is more common for infants and children under 5 y of age to be provided with a single high dose retinyl palmitate supplement at regular intervals, with various coverage rates. The current dosing schedule in the Expanded Program on Immunization (EPI) is as follows: 200,000 IU (60 mg) to mothers in the first 4–6 wk postpartum; 25,000 IU (7.5 mg) to their infants at 6, 10 and 14 wk of age; and 100,000 IU (30 mg) at 9 mo (2
). However, a multicenter trial found that this level and frequency of supplementation had little effect on infant vitamin A status and did not reduce rates of morbidity (2
). A recent World Health Organization (WHO) consultation on this topic therefore recommended that high dose vitamin A supplementation of women in vitamin A–deficient areas be increased to 200,000 IU (60 mg) at delivery plus another 200,000 IU before 8 wk postpartum or a single dose not exceeding 10,000 IU (3 mg) per day or 25,000 IU (7.5 mg) once a week at any time postpartum (3
). The new recommendation for infants aged 0–5 mo is three doses of 50,000 IU (15 mg) at least 1 mo apart, preferably at DPT (diphtheria, pertussis, tetanus) contacts (6, 10 and 14 wk). At age 6–11 mo, the recommendation is a single dose of 100,000 IU (30 mg) (e.g., at measles vaccination at 9 mo) followed by 200,000 IU every 6 mo between 12 and 59 mo of age, which can be given at any health or immunization contact.
Fortified foods
Fortified foods provide additional vitamin A in some settings. In industrialized countries important sources of preformed fortificant vitamin A include fortified milk (about 150 µg per cup), fortified breakfast cereals (10–100% of the recommended intake per serving) and margarine (approximately 50 µg per pat) (4
). In some developing countries, approaches to increasing vitamin A intake include fortification of sugar, oil, margarine, milk, wheat flour, corn flour, instant noodles and rice. Infant formulas and prepared infant cereals are usually fortified with the vitamin.
Based on the potential for individuals to consume large amounts of vitamin A from one or more of the above sources, it is important to ensure that the total intake of the vitamin from all available sources is unlikely to be excessive, particularly given the new recommendations to increase the amount given to infants and postpartum women as high dose supplements, the push toward vitamin A fortification of more foods and the possibility that the current iron-folate supplements available through UNICEF will be reformulated to include the daily recommended intake of vitamin A.
|
EVALUATING RISK OF EXCESSIVE INTAKES OF VITAMIN A |
|---|
|
TOPABSTRACTINTRODUCTIONSOURCES of VITAMIN AEVALUATING RISK OF EXCESSIVE...BIOLOGICAL INDICATORS OF...EVALUATING THE RISK OF...LITERATURE CITED |
|---|
Only preformed retinol and other retinoids can cause acute toxicity. Absorption of retinol remains efficient (approximately 90%) even when dietary levels are high (5
), although absorption from high dose supplements is closer to 50%. ß-Carotene and other provitamin A carotenoids from foods are not toxic, because the efficiency of absorption falls as carotenoid intake increases (6
). There are no reports of high ß-carotene intakes from foods ever having caused vitamin A toxicity. The effect of high or low intakes of preformed retinol on the absorption and conversion of ß-carotene or other provitamin A carotenoids to retinol is not known.
Until recently, it was assumed that about one-third of a dose of dietary carotenoids was absorbed and half that amount was converted to retinol, resulting in a bioconversion factor of 6:1 for ß-carotene to retinol (7
). This bioconversion factor has been used in most food composition tables to convert carotenoids to retinol equivalents. However, during the past decade it has become apparent that absorption of carotene from plant sources, especially from vegetables, is substantially less than one-third that absorbed from a dose given in oil. Several recent estimates of ß-carotene absorption from a diet consisting mainly of vegetables show that absorption is about one-half what was previously assumed (8
,9
). Based on such studies, the Institute of Medicine estimated recently that 1 RAE is equal to 12 µg of ß-carotene (10
), instead of the 6 µg of ß-carotene estimate used previously (4
,11
).
When a dose of preformed vitamin A is administered as retinyl esters, it is deesterified in the intestinal lumen, incorporated into lipid micelles and transported into intestinal cells. Within the intestinal cells, most of the free retinol is reesterified and incorporated into the lipid phase of chylomicrons, which are released into the lymph and later into the systemic circulation. A small percentage of free retinol is oxidized and absorbed as retinoic acid through the portal vein. Most of the retinyl esters in chylomicrons and chylomicron remnants are taken up by the liver. However, retinol can also be taken up directly from lipoproteins by the mammary gland and possibly by other extrahepatic tissues (12
,13
). In the liver, retinyl esters are deesterified and converted to retinol. This retinol can be 1) bound to retinol-binding protein and released into the circulation, 2) reesterified and stored in the liver or 3) oxidized and excreted in bile. At high vitamin A intakes, retinyl esters also appear in the circulation, as do increasing amounts of retinoic acid. The latter and especially all-trans-retinoic acid—is of particular concern because of its potent effect on cell differentiation and gene expression.
ß-Carotene and other provitamin A carotenoids can be cleaved into retinal and reduced to retinol in intestinal cells, esterified, packaged into lipoproteins as a retinyl ester, taken up by the liver and either stored as retinyl esters or circulated bound to retinol-binding protein, as described for preformed vitamin A. These carotenoids can also be converted to retinoic acid, ß-apo-carotenals and ß-apo-carotenoic acids within the intestinal cells (14
).
Acute toxicity from high doses
Acute toxicity can occur when an individual consumes a large dose of preformed vitamin A. By far the majority of information about tolerance to acute high dose supplements of vitamin A has been obtained in public health programs, such as EPI, the Integrated Management of Childhood Illness (IMCI) program, and national vitamin A days. Symptoms of acute toxicity are usually mild and transient.
Expanded Programme of Immunization (EPI) recommends a 1-mo interval between doses for infants. To the best of our knowledge, a dose of 50,000 IU (15 mg) of vitamin A is safe for infants before 6 mo, but doses larger than that should not be given, especially to children under the age of 4 mo (3
).
In young children, 100,000 IU (30 mg) of vitamin A at 6–11 mo and 200,000 IU (60 mg) every 3 to 6 mo for children aged 12–60 mo, gives few side effects and is effective for reducing mortality (15
). In 1980 a 300,000-IU dose—a higher level than is now recommended—produced a 6% incidence of vomiting and a 16% incidence of transient diarrhea in Indonesian preschoolers (16
). Doses of 100,000 and 200,000 IU caused nausea, vomiting, headache, diarrhea and fever among some children in the Philippines (17
). The incidence of these symptoms was higher with the larger of the two doses and among the smallest children during the first 2 y of life. Thus, transient nausea and vomiting are expected in 5–10% of the youngest children receiving these doses (3
). It is theoretically possible for an infant to receive multiple high doses through EPI as well as on vitamin A days and as part of the IMCI program at closer intervals than 1 mo. The effect of closely spaced high doses of vitamin A on liver stores of the vitamin is estimated in the last section of this article.
In adults symptoms of acute vitamin A toxicity include nausea and vomiting, increased cerebrospinal fluid pressure, headaches, blurred vision and lack of muscular coordination (5
). The symptoms have been reported to occur with single or short term doses of about 15 mg (50,000 IU) in adults (18
). However, the current recommendation for postpartum women is 400,000 IU given as two doses of 200,000 IU 1 d apart within 6 wk of delivery (2
). To evaluate whether postpartum women can tolerate this level of supplementation, Iliff et al. (19
) conducted a randomized double-blind trial in which 390 women who received 400,000 IU of vitamin A as a single dose were compared with 380 women who received a placebo. Overall, <5% of women in both treatment groups reported headache, nausea, vomiting, blurred vision or drowsiness; there was no difference in the rates of these symptoms among treated and control women (19
). These data suggest that dosing at this level is well tolerated by postpartum women.
In infants an excessive dose can cause transient bulging of the fontanelle. A dose of 25,000 IU (7.5 mg) or 50,000 IU (15 mg) increases the incidence of this symptom (20
). Other symptoms of more severe chronic toxicity after repeated long term consumption of very high doses include vomiting, weight loss, fever, headache, bone abnormalities and pain, an enlarged liver and increased intracranial pressure (21
). The median lethal dose for a single dose in young monkeys is very high: 168 mg (560,000 IU) per kg of body weight (168–302-mg total dose). These monkeys weighed 1–1.8 kg, and all monkeys that received >300 mg/kg died. None died at a dose of 100 mg/kg (22
). Interpretation of these data are complicated by the fact that the monkeys also received high doses of vitamins D and E. In a case study, an infant who weighed 2.5 kg died when given 3.3 mg/kg of preformed vitamin A daily for 11 d after birth, a total of 119 mg (21
). Other factors may have contributed to the death of this underweight infant.
Chronic toxicity from persistent consumption of moderate or high doses
Chronic toxicity could occur in an individual who commonly consumes supplements or frequently eats foods with high vitamin A content, such as liver (3
) and can lead to severe liver damage. It is thought to require ingesting at least 30 mg of preformed retinol per day [100,000 IU or about 40 times the recommended dietary allowance (RDA)] for months or years (23
–26
). However, there are case studies of liver damage in a 63-y-old woman who took 14 mg/d (20 times the RDA) for 10 y (27
) and a 36-y-old man who took 15 mg/d (17 times the RDA) for 12 y (28
). Several other case studies report liver toxicity at intakes above about 8 mg/d or 10 times the RDA (25
,29
). Healthy consumers of supplements providing about twice the RDA for vitamin A in the United States had elevated serum retinyl ester concentrations (24
).
Tolerable upper levels (UL) for vitamin A intake
The Institute of Medicine has recently recommended UL for vitamin A (10
). These values and the criteria used to define them are listed in Table 1
for children and women of childbearing age. UL is defined by the Institute of Medicine as "the highest level of daily nutrient intake that is likely to pose no risks of adverse health effects for almost all individuals in the general population. As intake increases above the UL, the risk of adverse effects increases." When possible, the UL is based on a no-observed-adverse-effect level (NOAEL), which is the highest intake at which no adverse effects have been reported. If the data are inadequate to determine a NOAEL, the lowest-observed-adverse-effect level (LOAEL) is used—i.e., the lowest intake at which an adverse effect has been observed. The UL is severalfold lower than the NOAEL or LOAEL; the latter are divided by an uncertainty factor, the size of which depends on factors such as the severity of the adverse effect and level of uncertainty about the data, which provides a margin of safety. There are more data on which to base the NOAEL and LOAEL for infants than there are for children over 12 mo of age. The LOAEL for children was extrapolated from that of adults (men and women beyond childbearing age) "given the dearth of information and the need for conservatism" (see Ref. 10
; p. 107). In general, the UL values for vitamin A were deliberately set conservatively.
|
). This is probably a rather conservative estimate. One very controversial analysis has been published in which teratogenicity was thought to be associated with an intake of as little as 3000 µg of preformed vitamin A per day or 1500 µg/d from supplements in addition to food sources (31
). In contrast, Mills et al. (32
) reported no association between birth defects and periconceptional vitamin A exposure at doses >8000 IU (2400 µg) or >10,000 IU (3000 µg) per day in a case-control study. Numerous studies show no risk from intakes <3 mg/d, but there are too few studies to be able to define a dose-response relationship in the range of 3000 to 7000 µg/d.
It is likely that increased blood levels of retinoic acid compounds are responsible for the teratogenic effects of vitamin A. Retinoid-induced teratogenicity has been reported in women who were exposed to high doses of retinoic acid or its metabolites within the first 6 wk of pregnancy (33
), but there is limited evidence that exposure to high doses of retinol or retinyl esters in early pregnancy has teratogenic effects (34
). A recent report indicates that blood levels of retinoic acid and 13-cis-retinoic acid in healthy nonpregnant women consuming 10,000 IU/d (3 mg/d) were similar to the physiologic levels of these retinoids in pregnant women. Even at dosages of 30,000 IU of retinol per day (9 mg/d), blood levels of retinoic acid compounds in nonpregnant women were similar to the physiologic levels of these retinoids in pregnant women. This suggests that doses higher than 30,000 IU/d may be required to increase retinoic acid concentrations to levels that are likely to produce teratogenic effects (35
). At excessive vitamin A intakes consumed over 20 d (250 µg kg-1 d-1 or a total of about 1 million IU), blood levels of retinoic acid compounds have been shown to increase two- to sevenfold (36
).
|
BIOLOGICAL INDICATORS OF TOXICITY |
|---|
|
TOPABSTRACTINTRODUCTIONSOURCES of VITAMIN AEVALUATING RISK OF EXCESSIVE...BIOLOGICAL INDICATORS OF...EVALUATING THE RISK OF...LITERATURE CITED |
|---|
Chronic vitamin A toxicity produces liver abnormalities including perisinusoidal fibrosis, hypertrophy of Ito cells and a spontaneous green fluorescence of sinusoidal cells. Alcohol abuse, hepatitis, some medications and liver disease hasten this liver damage, which can be fatal. The available data were evaluated by the Institute of Medicine and used to derive the UL for men and women who are not of reproductive age (10
). Because hepatotoxicity is severe and irreversible and is associated with very high liver vitamin A concentrations, it is not a very useful endpoint for estimating the safety of vitamin A supplementation or fortification programs, where much lower levels are encountered.
High vitamin A concentrations in liver
Normal vitamin A concentrations in liver of adults in the United States are reported to be about 100 µg/g with a range of 14 to 160 µg/g (37
). Plasma vitamin A concentrations increase when concentrations in the liver reach about 300 µg of retinol per gram, a concentration at which circulating retinyl esters also increase (5
). Olson suggested that 300 µg per gram of liver indicates vitamin A toxicity (5
).
A review of the liver vitamin A concentrations of full-term stillborn infants in the United States showed a range from 10 to >100 µg/g with most values at the lower end of this range (median, 11 µg/g) (38
). Although there are few data, concentrations in liver increase with age, reaching levels of 30–126 µg/g in U.S. children aged 13–24 mo (39
). Although rare, cases of vitamin A toxicity have been reported in young children of well-intentioned parents who provide liver to their children on a daily basis. One case study describes hypervitaminosis A in two boys who consumed an average of 690 µg of vitamin A per day (2530 IU/d) in chicken liver and 135–759 µg/d (450–2500 µg/d) from supplements. One child developed toxicity symptoms at age 2 y, and the other developed symptoms at 6 y of age (40
).
High serum retinol concentrations
Serum retinol concentrations tend to plateau at the vitamin A intake levels usually consumed in industrialized countries, falling below the cutoff value for marginal deficiency of 0.70 µmol/L (20 µg/dL) when intakes are deficient and rising to values of >3.5 µmol/L (100 µg/dL) only when intakes are very high. Therefore, in the United States, serum retinol concentrations are not correlated with usual levels of vitamin A intake (41
) or with dietary supplement use (24
,42
,43
). The median serum retinol concentration in the National Health and Nutrition Examination Survey (NHANES III) ranges from 1.2 to 2.2 µmol/L (34 to 63 µg/dL) across age and gender groups. Olson suggested that a serum retinol concentration of 100 µg/dL is associated with vitamin A toxicity in infants (44
). In California, we frequently find concentrations of 50–80 µg/dL in healthy young adults. There is no generally accepted serum retinol concentration that signifies hypervitaminosis A.
The presence of esterified retinol in fasting plasma is, however, an early reflection of high vitamin A intakes (26
). Increases in retinyl esters are seen with intakes 1–2 times the RDA for vitamin A in the form of supplements consumed in addition to food (24
). In most cases, the supplements were consumed regularly for several years. The median retinyl ester concentration in NHANES III was 0.15–0.19 µmol/L. There is little information about the threshold vitamin A intake above which retinyl ester concentrations begin to increase, and there is little evidence that the retinyl esters are toxic. Instead, they are indicators of a high vitamin A intake.
Production of toxic metabolites
Retinoic acid can be produced from retinol or provitamin A carotenoids. There are several forms of retinoic acid including all-trans-, 13-cis-, and 13-cis-oxo-retinoic acid. Concentrations in plasma are much lower than that of retinol, but they increase two- to fourfold after daily doses of retinyl palmitate well above dietary levels (36
) and after consuming 2 mg of turkey liver per kilogram of body weight (45
). Retinoic acid is the active form of retinol and is required by almost every cell for normal function. However, there is concern about the potential toxicity of higher levels of retinoic acid. 13-cis-Retinoic acid is a known teratogen. Elevated blood levels of retinoic acid and other retinoid metabolites may explain many of the symptoms of vitamin A toxicity. Retinoids may destabilize membranes and cause them to rupture and also alter gene expression (46
). Much remains to be learned about the role of retinoic acid in vitamin A toxicity and the association between vitamin A intake and retinoic acid production. Currently, there is insufficient information to define a vitamin A intake value at which side effects and toxicity from retinoic acid become a problem. However, serum levels increase with vitamin A intakes >3000 µg/d (>10,000 IU/d) (47
). To put high dose supplements into perspective, about 150 g of liver from domestic animals contains 25–75 mg of retinol (80,000 to 250,000 IU). Infants and children with some forms of cancer are treated long term with daily doses exceeding 150,000 IU without apparent side effects. Future studies on high dose supplements should test their effect on plasma and breast milk retinoic acid concentrations.
Breast milk retinol concentrations
Increasing the vitamin A content of breast milk, by strategies such as maternal high dose supplementation in early lactation and food fortification, can make an important and sustained contribution to infant stores of the vitamin. High concentrations of retinol in breast milk potentially could serve as a marker of high vitamin A intake. In Guatemala, breast milk vitamin A levels of 100–400 µg/dL have recently been reported in lactating women in periurban neighborhoods of Guatemala City (unpublished data). It is probable that these high vitamin A concentrations are related to consumption of vitamin A–fortified sugar. There is currently no cutoff for elevated milk retinol. Assuming that average breast milk production is about 780 mL/d and the UL of retinol intake for infants is 600 µg/d (10
), the UL would be equivalent to about 77 µg/dL. In several studies in which high doses of vitamin A were given to lactating women, milk concentrations reached an average of 70 µg/L and caused no clinical signs of toxicity in the infants (48
). More work is needed to explore the potential benefits and side effects of high dose vitamin A, fortified foods and other vitamin A supplements on breast milk retinol concentrations.
Proposed cutoffs for monitoring toxicity in pubic health programs
Based on the information in the preceding sections, we propose tentative cutoffs that can be useful for monitoring whether vitamin A intakes are sufficiently excessive to lead to toxicity (Table 2
). Unlike the UL, these should not be interpreted as being unsafe levels of intake. Instead, they are levels in the diet or in body tissues that indicate there may be some risk of toxicity and are more suited than the UL for evaluating programs in which high dose supplements are provided.
|
|
EVALUATING THE RISK OF TOXICITY FROM VITAMIN A IN FOODS AND SUPPLEMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONSOURCES of VITAMIN AEVALUATING RISK OF EXCESSIVE...BIOLOGICAL INDICATORS OF...EVALUATING THE RISK OF...LITERATURE CITED |
|---|
are used to assess the potential risk of toxicity in specific scenarios meant to improve the vitamin A status of deficient populations. These include a sensitivity analysis of the potential for toxicity, if, under field conditions, children inadvertently received large doses in addition to the recommended schedule. Question 1. What is the evidence of toxicity of vitamin A intake in a developed country, based on existing dietary intake (including fortification) plus supplementation?
Vitamin A intake data are collected in two major dietary surveys in the United States. NHANES consists of a 1-d survey of individuals aged 2 mo and older. NHANES III, the most recent of these surveys, was completed in 1994. Almost 34,000 individuals were surveyed but data from 18,000 people were used for the dietary analysis. The second survey is the Continuing Survey of Food Intakes by Individuals (CSFII) conducted by the U.S. Department of Agriculture. It captures information on 60,000 individuals with 1 or 2 d of dietary information (generally 2 d). It includes questions such as "Do you take a vitamin supplement? What kind? How often?" It does not provide quantitative information about vitamin A intake from supplements.
We used NHANES III data to estimate the following: 1) percentiles of dietary retinol intake by children 1–3 y of age and by nonpregnant nonlactating women (NPNL) aged 19–30 y; 2) percentiles of total dietary vitamin A intake (retinol and provitamin A carotenoids, micrograms of RAE using a 12:1 factor for converting ß-carotene to retinol) for both groups; and 3) percentiles of total vitamin A intake (micrograms of RAE) from supplements for both groups.
These survey data are useful for assessing percentiles of vitamin A intake by age and gender in the United States. However, there are few data on the intakes of pregnant or lactating women in these data sets, so estimates of their intake are not included. Also, information on the use of supplements is limited and considered to be potentially unreliable statistically because of the small sample size. Furthermore, the data on supplement use are grouped together for children 1–8 y of age and for NPNL 19 y of age and older, so it is not possible to separate supplement use by narrower age ranges.
Data on percentiles of vitamin A intake are presented in Table 3
and Table 4
for women and children, respectively. The data on dietary retinol intake from CSFII, shown in column 2 of Tables 2
and 3
, is very similar to that from the NHANES III.
|
|
3655 µg of RAE/d, which exceeds the UL of 3000 µg/d (10
). However, a daily intake of 4500 µg of RAE/d (the NOAEL) among women of reproductive age is not associated with any adverse effects.
For children aged 1–3 y, the median vitamin A intake from food was 484 µg of RAE/d, and the 95th percentile was 1259 µg of RAE/d. The median intake from supplements was 721 µg of RAE/d, and the 95th percentile was 1482 µg of RAE/d. About 25% of children exceed the UL of 600 µg/d (10
) from diet alone, and, of those given supplements, 75% exceed the UL from supplements alone. However, there is no evidence that these levels of intake cause toxicity in the United States, and they are below the lowest level of intake at which adverse effects have been reported for children in this age range. (The LOAEL is 6000 µg/d for children 1–3 y of age, based on daily consumption of this amount for several months.)
Question 2. What is the evidence of toxicity of vitamin A intake in a developing country based on dietary intake (including fortification) plus supplementation?
The intake of retinol from foods is very low for most individuals in developing countries. Estimates for usual vitamin A and ß-carotene intake were obtained from the Mexico Nutrition Collaborative Research Support Program in rural Mexico (NCRSP). Although these data obviously apply only to this population sample, the main purpose of the following analyses is to provide an example that can be modified for other population groups (4
,11
). Individuals in Mexico each provided multiple days (typically 24) of dietary intake data over the course of a year, permitting reasonable estimates of vitamin A intake in a community with relatively low intakes of preformed vitamin A. For the exercise here, data were used only when there were at least 12 d of dietary data per individual. None of the foods contained vitamin A as a fortificant. Intakes were expressed as percentiles for comparison with the NHANES values as shown in Table 5
and Table 6
for women and children, respectively.
|
|
80% of that consumed by U.S. women (224 µg of RAE/d)]. Had each woman consumed a vitamin supplement containing 1500 µg of RAE, the 95th percentile of retinol intake would have been about 2000 µg of RAE/d. This is below the UL of 3000 µg of RAE/d for women of reproductive age. The median retinol intake of children (53 µg/d) was similar to that of their mothers because preschoolers consume more dairy products. Their median provitamin A carotenoid intake (31 µg of RAE/d) is about 40% that of U.S. children (80 µg of RAE/d). The highest 5% of consumers ingested 386 µg of retinol + 83 µg of RAE carotene for a total of 469 µg of RAE/d. Had each child consumed a supplement (750 µg of RAE/d), then 5% of them would be consuming >1176 µg of RAE/d, which is above the UL of 600 µg of RAE/d. However, it is below the LOAEL of 6000 µg of RAE/d for children 0–3 y of age and substantially below the 15-mg (50,000 IU) dose associated with a small risk of bulging fontanelle and other symptoms of toxicity.
There is relatively little information about the contribution of fortified foods to vitamin A intake in developing countries. In urban Guatemala, three fortified foods (sugar, Incaparina and margarine) provided 55% of the total dietary intake of vitamin A for preschoolers as shown in Table 7
(49
). The median intakes from these sources for children who consumed them were 81, 94 and 61 µg/d, respectively, and the 75th percentile values were 143, 217 and 122 µg/d, respectively. Adding retinol intake from the usual diet (469 µg of RAE/d) to the 500 µg/d at the 75th percentile of consumption from fortified foods suggests that there is little risk of toxicity at intakes of 1 mg of RAE/d. If preschoolers ate usual amounts of all four of the main vitamin A–fortified foods (sugar, Incaparina, margarine and Cerelac), they would have consumed close to 600 µg of retinol in addition to their usual diet, which would amount to a total vitamin A intake of 1.1 mg of RAE/d. If they took a supplement (750 µg) in addition to all these fortified foods and consumed dietary retinol at the level of the 95th percentile, their total retinol intake would have been about 1819 µg/d. Again, this level of intake exceeds the UL of 600 µg of RAE/d for this age group, but it is well below the LOAEL of 6000 µg of RAE/d, which is the lowest level of intake at which adverse effects have been reported. Between 1996 and 2000, rural Indonesian children aged <5 y had access to vitamin A–fortified foods, including margarine, dry milk, condensed milk, some instant noodles, complementary foods and infant formula. However, their median intake of vitamin A from fortified foods was zero, and the 95th percentile of intake from these sources was only about 500 µg/d (de Pee, S., unpublished data). The median intake from retinol-rich foods was 120–200 µg/d across the 4 y, and the 95th percentile was 700-1100 µg/d. Thus, the intake of vitamin A in either vitamin A–rich or fortified foods was still low in rural Indonesia.
|
,51
) and that 2.2% of the body pool is catabolized daily by children (52
,53
). The value for the fractional catabolic rate of 2.2%/d was estimated from our data on plasma retinol kinetics of rural Peruvian children (52
). Some of these children had mild upper respiratory tract infections but no other clinically evident illness. Of 80 children who had sufficient serum for C-reactive protein analyses to detect subclinical infection, 51% were normal (<10 mg/dL), 39% were slightly elevated (10–19 mg/dL) and only 10% were very elevated.
Briefly, a population-level plasma retinol kinetic curve was constructed among Peruvian children 12–24 mo of age to estimate the fractional catabolic rate of retinol in this age group. Because a large number of blood samples (24 over a 75-d period) is required to construct a plasma kinetic curve for an individual child, we chose to construct a population-level plasma kinetic curve for the full population of subjects instead of doing individual curves for each child. For development of the population kinetic curve, an oral dose of [2H4]retinyl acetate (14 µmol) was administered to 107 children. Two blood samples were drawn from each child to measure the plasma isotopic ratio [2H4]retinol:retinol. Each child was scheduled to have a blood sample (5 mL) drawn 3 d after administration of the isotope and another sample at 1 of 23 time points over a 75-d study period. (The day 3 plasma isotopic ratios were used to examine the relationship between the early postdose plasma isotopic ratio and estimates of vitamin A pool size in this population.) Children were assigned so that there were about five children at each of the 23 time points. Plasma isotopic ratios were measured at 24 time points total as follows: 2, 3, 4, 5, 6, 8, 24 and 32 h daily on days 2–8 and days 10, 12, 15, 18, 23, 29, 35, 42 and 75. The plasma kinetic data were plotted as [2H4]retinol:retinol/([2H4]retinol:retinol + 1) versus time. The fractional catabolic rate was estimated by fitting an exponential model of the form ratio/(ratio + 1) = a0 · exp(-b0 · day) + a1 · exp(-b1 · day) + a2 · exp(-b2 · day) to the observed plasma kinetic data. The fractional catabolic rate for retinol was estimated as the second coefficient (b2) of the final exponential term, which was -0.022 for this population. Because the fractional catabolic rate was estimated by using the population-level kinetic curve, we do not have estimates of the fractional catabolic rate for each individual child. Thus, it was not possible to calculate a standard deviation for the rate. However, the 95% confidence interval (-0.014; -0.030) was calculated for the coefficient that was used to estimate the fractional catabolic rate as an estimate of the variability of the rate in the population studied.
Although these children may be representative of children in developing countries, their vitamin A turnover rate may be higher than that of younger infants—especially compared with infants who are exclusively breast-fed—because of the high incidence of infections that typically occurs between ages 12 and 24 mo. For this reason, the analyses that follow were also done with the estimated catabolic rate of 0.5%/d for adults (44
,50
) and an intermediate value of 1.5%/d.
Using this approach, we can estimate how much vitamin A is retained and how long liver vitamin A stores are sustained at an acceptable level after administration of high dose vitamin A supplements. As discussed earlier, >300 µg of retinol per gram of liver has been proposed as a threshold for potential risk of vitamin A toxicity, because at this concentration there is an increase in plasma retinyl esters and retinol. This is likely to be a conservative cutoff because no association with liver damage or clinical symptoms has been reported at this liver retinol concentration. Previously, we used this kinetic approach to estimate the effect of the WHO multicenter trial on the vitamin A stores of infants who received 25,000 IU at 6, 10 and 14 wk and 9 mo of age; a control group received 100,000 IU at 9 mo of age (52
). As shown in Figure 1
, our estimates of the amounts of vitamin A retained at the various time points were consistent with the results of the trial; in infants who received vitamin A at each of the EPI contacts, little of the dose remained at 6 mo of age, and liver stores (assessed as the modified retinol dose response) were depleted at 9 and 12 mo. In the following section, we use the same approach to estimate the impact of high dose supplementation on infant vitamin A stores, according to the newly proposed, higher dosing schedules (50,000 IU at 6, 10 and 14 wk of age; 100,000 IU at 6–11 mo of age; and 200,000 IU at 
12 mo of age) and to evaluate the risk of toxicity if children received extra high dose supplements by accident.
|
Using the kinetic approach described above, we estimated the impact of this proposed dosing schedule on infant liver vitamin A concentrations. Liver concentrations of the vitamin were estimated at each interval, assuming that 50% of the dose is retained; the fractional catabolic rate is 0.5%/d, 1.5%/d or 2.2%/d, and liver weight is 4.5% of body weight. The calculations ignore any effects of dietary vitamin A (other than breast milk); it is assumed that complementary foods supply much lower amounts than are consumed in breast milk or as high dose supplements.
In Figure 2
, estimates are based on the infant receiving 50,000 IU of vitamin A at 6, 10 and 14 wk (EPI doses); 100,000 IU of vitamin A at 9 mo (measles vaccine dose); 200,000 IU of vitamin A at 15 mo (on a vitamin A day); and 200,000 IU of vitamin A at 16 mo (a dose during treatment of measles or diarrhea). In this example, the mother receives 200,000 IU in the first 8 wk postpartum, and the vitamin A concentration of breast milk is assumed to be 50 µg/dL. For infants, these are the doses now recommended for EPI (3
).
|
105 and 125 µg/g at the catabolic rates of 2.2%/d and 1.5%/d, respectively. The maximum value occurs with administration of the high dose supplement (60 mg) at 16 mo, which is only 30 d after the previous high dose supplement (60 mg). With the slower adult rate of catabolism (0.5%), which is very unlikely to be the catabolic rate in infants, concentrations would peak at 180 µg/g, which is still lower than the proposed cutoff of 300 µg per gram of liver for vitamin A toxicity. The concentrations at other time points are much lower in relation to the peak value at 16 mo for all three catabolic rates, particularly at 2 y of age.
When the proposed new dosing schedule for the mother (3
) is also considered, postpartum she will receive two 200,000-IU (60-mg) supplements within 2 mo of giving birth, for a total of 400,000 IU. In this instance, we assumed that these women will have subsequently higher breast milk vitamin A concentrations, 70 µg/dL, based on a supplementation trial in Indonesia where mothers were given 300,000 IU 1–3 wk postpartum (48
). We assume this higher concentration of retinol in breast milk made almost no difference to the estimated peak liver concentration of the vitamin, because the amount consumed in breast milk is so much less than the amount given directly to the child in high dose supplements (Fig. 3
).
|
. Question 4. Is there any risk for a child older than 1 y who unintentionally receives one 200,000-IU dose every month?
The parameters used to generate the liver vitamin A values shown in Figure 4
are that the infant receives 50,000 IU of vitamin A at 6, 10 and 14 wk (EPI doses); 100,000 IU of vitamin A at 9 mo (a dose given at the same time as measles vaccine); and 200,000 IU of vitamin A monthly from 12 to 24 mo. This unlikely scenario is presented to illustrate the effect of frequent high doses on liver retinol concentrations. It is assumed that the mother receives 200,000 IU in the first 8 wk postpartum and that breast milk contains 50 µg/dL.
|
130 µg/g for the catabolic rate of 2.2%/d and 180 µg/g for the rate of 1.5%/d, which is below the suggested cutoff value of 300 µg/g (Fig. 4)
. However, at the adult catabolic rate of 0.5%/d, the liver concentration reaches 400 µg/g, which exceeds the cutoff value for potential toxicity. It is very unlikely that children in this age group would have a catabolic rate similar to that of adults. The adult catabolic rate is 4 times lower than the observed catabolic rate for Peruvian children 12–24 mo of age. At 12 mo of age, liver stores of infants who have never been supplemented are estimated to be about half (4.8 versus 9.8 mg) those of infants who are supplemented with 50,000 IU at 6, 10 and 14 wk (EPI doses) and with 100,000 IU at 9 mo (with measles vaccine). Nevertheless, if children were to be supplemented with 200,000 IU monthly between 12 and 24 mo of age, their liver concentrations during this time would be almost exactly the same regardless of whether they received high dose supplements during the 1st y, and the peak liver concentration would occur at the same time (age 17–18 mo). The reason is that the amount of vitamin A provided by the monthly high dose supplements (60-mg dose) given between 12 and 24 mo of age is much larger than the 5–10 mg from earlier supplements remaining in the liver at age 12 mo.
Based on these estimates, providing the new recommended doses for the EPI schedule during the 1st y of life, and 200,000 IU at health contacts during the 2nd y of life, is unlikely to cause excessive liver vitamin A accumulation. Indeed, given the plateau in the two most likely catabolic rate scenarios (2.0 and 1.5%), even monthly dosing with 200,000 IU throughout the 2nd y of life should be relatively safe. However, because the fractional catabolic rate of older children (age >24 mo) is not known, it is not yet possible to assess whether monthly dosing of older children with 200,000 IU is safe.
Question 5. What is the likelihood of toxicity when the mother is supplemented on schedule, and the child is supplemented on schedule and through campaigns and then the child receives an extra dose 1 wk after a campaign?
For these estimations, it is assumed that the infant receives 50,000 IU of vitamin A at 6, 10 and 14 wk (EPI doses); 100,000 IU of vitamin A at 9 mo (measles vaccine dose); 200,000 IU of vitamin A at 15 mo (on a vitamin A day); and an additional 200,000 IU of vitamin A at 15 mo, 1 wk (a random dose on a visit to a health clinic). The mother receives 200,000 IU in the first 8 wk postpartum, and breast milk contains 50 µg/dL.
The highest liver concentration occurs when the 200,000-IU dose at 15 mo is followed by a similar dose 1 wk later (Fig. 5
). The peak concentration is 130 µg/g for the catabolic rate of 2.2%/d, 140 µg/g for the rate of 1.5%/d, and nearly 200 µg/g for the adult rate of 0.5%/d. All these concentrations are below the cutoff for toxicity. As indicated in Figure 5
, the high liver concentration is temporary, falling to 50–75 µg/g for the catabolic rates of 2.2%/d and 1.5%/d, respectively, and to 150 µg/g for the adult rate of 0.5%/d by 18 mo. Assuming a high concentration of vitamin A in breast milk of 70 µg/dL makes little difference to these estimates (only 5 µg per gram of liver), because the contribution of breast milk is so much less than that of the high dose supplements.
|
There are few data on which to establish thresholds for excessive vitamin A intake or vitamin A concentrations in tissues. For preschoolers, NOAEL has been set at 6000 µg/d, and the UL of intake is only one-tenth that amount, 600 µg/d. In the United States, the 95th percentile intake from foods is 1259 µg of RAE/d (including fortified foods), which exceeds the UL but is below the NOAEL. The 95th percentile of intake from foods and supplements, 2741 µg of RAE/d, is still below the NOAEL for this age group. For women of reproductive age, NOAEL has been set at 4500 µg/d, and UL has been set at 3000 µg/d. In the United States, the 95th percentile of vitamin A intake from foods (including fortified foods) for NPNL aged 19–30 y is 1112 µg of RAE/d, which is below UL and NOAEL. When vitamin A from supplements is included in the estimate of total daily vitamin A intake, the 95th percentile of intake (3655 µg of RAE/d) exceeds the UL but is below the NOAEL for women of reproductive age.
In developing countries, among low-income populations, the intake of vitamin A from foods is far less than that observed in industrialized countries. There are few data on the contribution of fortified foods to total dietary vitamin A intake in developing countries. Data from Guatemala suggest that a maximum of 400–600 µg/d may have come from fortified foods in the 1990s. Preschoolers who consume fortified foods in addition to a diet that contains vitamin A–rich foods may exceed the UL of 600 µg/d, but it is very unlikely that their total dietary vitamin A intake will reach or exceed the NOAEL of 6000 µg/d. For women of reproductive age, the Nutrition Collaborative Research Support Program data from Mexico indicate that the 95th percentile of vitamin A intake from unfortified foods was 650 µg/d in the 1980s, which is far below the UL of 3000 µg/d and NOAEL of 4500 µg/d. It is unlikely that intake from fortified foods or from supplements providing the RDA will result in intakes that exceed the NOAEL for women in this age group. Future surveys should monitor the intake of vitamin A from foods, fortified foods and supplements over time to better estimate trends in intake from these sources in developing countries.
Biological indicators of vitamin A toxicity
Few data are available to set thresholds for tissue concentrations associated with vitamin A toxicity. Olson (5
) suggested a cutoff for liver vitamin A concentration of 300 µg/g and a cutoff for plasma of 100 µg/dL. More information is needed on blood and breast milk levels of retinoic acid that are associated with vitamin A toxicity. This information is especially important for fully assessing the risk of toxicity associated with administering high dose vitamin A supplements in public health programs.
Our estimates of vitamin A retention from high dose supplements given to infants indicate that the amount of vitamin A retained from a single high dose supplement of 50,000 or 100,000 IU can greatly exceed the amount of vitamin A retained from breast milk but only for a short time (1 mo after administration of the dose). Because the vitamin A retained from supplements is rapidly depleted, a constant daily supply of vitamin A from breast milk and/or complementary foods is very important to sustain liver vitamin A concentrations at acceptable levels (>20 µg/g). Thus, administration of high dose supplements according to the proposed new dosing schedule for the EPI will provide a temporary boost in stores after administration of each of the doses, but these doses still would not be sufficient to maintain adequate liver vitamin A concentrations throughout the 1st y of life.
According to our estimates, administration of 200,000 IU at health contacts between 12 and 24 mo of age, in addition to the recommended supplements during the 1st y of life, should be perfectly safe. Even administration of 200,000 IU (60 mg) monthly between 12 and 24 mo of age is likely to be safe. Only in the highly unlikely situation that a child’s catabolic rate is close to that of an adult and the child receives 200,000 IU monthly between 12 and 24 mo of age will liver vitamin A concentrations exceed the proposed cutoff of 300 µg/g. This is unlikely because the adult rate is 4 times lower than the observed rate for children in this age range.
Editor’s note
The method used to estimate liver vitamin A stores needs additional refinement as more data are obtained on plasma retinol kinetics in infants and young children. However, the authors believe the estimates presented here are the best possible, given the limitations of the current data and that they fit well with observed changes in vitamin A status in high dose supplementation trials.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
3 Abbreviations used: CHD, Child Health Division; CSFII, Continuing Survey of Food Intakes by Individuals; EPI, Expanded Programme on Immunization; IMCI, Integrated Management of Childhood Illness; LOAEL, lowest-observed-adverse-effect level; MRDR, modified relative dose-response; NCRSP, Nutrition Collaborative Research Support Program; NHANES, National Health and Nutrition Examination Survey; NOAEL, no-observed-adverse-effect level; NPNL, nonpregnant nonlactating women; RAE, retinol activity equivalent; RDA, recommended dietary allowance; RE, retinol equivalents; UL, tolerable upper level; UNICEF, United Nations Children’s Fund; WHO, World Health Organization.
|
LITERATURE CITED |
|---|
|
TOPABSTRACTINTRODUCTIONSOURCES of VITAMIN AEVALUATING RISK OF EXCESSIVE...BIOLOGICAL INDICATORS OF...EVALUATING THE RISK OF...LITERATURE CITED |
|---|
1. Mahoney, C., Margolis, M., Knauss, T. & Labbe, R. (1980) Chronic vitamin A intoxication in infants fed chicken liver. Pediatrics 65:893-897.
2. WHO/CHD (1998) Randomised trial to assess benefits and safety of vitamin A supplementation linked to immunisation in early infancy. Lancet 352:1257-1263.[Medline]
3. Ross, D. (2002) Recommendations for vitamin A supplementation. J. Nutr. Suppl. 132:2902S-2906S.
4. West, C. E., Eilander, A. & van Lieshout, M. (2002) Consequences of revised estimates of carotenoid bioefficacy for dietary control of vitamin A deficiency in developing countries. J. Nutr. Suppl. 132:2920S-2926S.
5. Olson, J. (1993) Vitamin A. Present Knowledge in Nutrition 1993:176-191 The Nutrition Foundation Washington, DC. .
6. Brubacher, G. & Weiser, H. (1985) The vitamin A activity of beta-carotene. Int. J. Vitam. Nutr. Res. 55:5-15.[Medline]
7. National Research Council (1989) Recommended Dietary Allowances 10th ed. 1989 National Academy Press Washington, DC. .
8. Van het Hof, K., Brouwer, I., West, C., Hademan, E., Steegers-Theunissen, R., van Dusseldorp, M. & Westrate, J. (1999) Bioavailability of lutein from vegetables is five times higher than that of beta-carotene. Am. J. Clin. Nutr. 70:261-268.
9. Micozzi, M. S., Brown, E. D., Edwards, B. K., Bieri, J. G., Taylor, P. R., Khachik, F. & Beecher, G. R. (1992) Plasma carotenoid response to chronic intake of selected foods and ß-carotene supplements in men. Am. J. Clin. Nutr. 55:1120-1125.
10. Institute of Medicine (2001) Dietary Reference Intakes: Vitamin A, Vitamin K, Arsenic, Boron, Chromium, Copper, Iodine, Iron, Manganese, Molybdenum, Nickel, Silicon, Vanadium and Zinc 2001 National Academy Press Washington, DC. .
11. Miller, M., Humphrey, J. H., Johnson, E., Marinda, E., Brookmeyer, R. & Katz, J. (2002) Why do children become vitamin A deficient?. J. Nutr. Suppl. 132:2867S-2880S.
12. Green, M. H., Green, J. B., Akohoue, S. A. & Kelley, S. K. (2001) Vitamin A intake affects the contribution of chylomicrons vs. retinol binding protein to milk vitamin A in lactating rats. J. Nutr. 131:1279-1282.
13. Davila, M., Norris, L., Cleary, M. & Ross, A. (1985) Vitamin A during lactation: relationship of maternal diet to milk vitamin A content and to the vitamin A status of lactating rats and their pups. J. Nutr. 115:1033-1041.
14. Erdman, J. W., Bierer, T. L. & Gugger, E. T. (1993) Absorption and transport of carotenoids. Ann. NY Acad. Sci. 691:76-85.[Medline]
15. Beaton, G. H., Martorell, R., L’Abbe, K. A., Edmonston, B., McCabe, G., Ross, A. C. & Harvey, B. (1993) Effectiveness of vitamin A supplementation in the control of young child morbidity and mortality in developing countries. Nutrition Policy Discussion Paper 13 1993 World Health Organization Geneva, Switzerland. .
16. Kusin, J., Sinaga, H. & Smit, E. (1980) Vitamin A status of preschool children in Suka Village, North Sumatra after an oral massive dose.Meeting on Vitamin A Deficiency and Xerophthalmia. WHO Nutrition Series 80 1980 World Health Organization Geneva, Switzerland. .
17. Florentino, R. F., Tanchoco, C. C., Ramos, A. C., Mendoza, T. S., Natividad, E. P., Tangco, J. B. & Sommer, A. (1990) Tolerance of preschoolers to two dosage strengths of vitamin A preparation. Am. J. Clin. Nutr. 52:694-700.
18. Bendich, A. & Langseth, L. (1989) Safety of vitamin A. Am. J. Clin. Nutr. 49:358-371.
19. Iliff, P., Humphrey, J. & Mahomva, A. (1999) Tolerance of large doses of vitamin A given to mothers and their babies shortly after delivery. Nutr. Res. 129:1437-1446.
20. Humphrey, J. H. & Ichord, R. N. (1999) Safety of Vitamin A for post partum women, young children, and infants (with special reference to) clinical significance of vitamin A-induced bulging fontanelle 1999 Prepared for: World Health Organization, Department of Nutrition for Health and Development Geneva, Switzerland. .
21. Bush, M. & Dahms, B. (1984) Fatal hypervitaminosis A in a neonate. Arch. Pathol. Lab. Med. 108:838-842.[Medline]
22. Macapinlac, M. & Olson, J. (1981) A lethal hypervitaminosis A syndrome in young monkeys following a single intramuscular dose of a water-miscible preparation containing vitamins A, D2 and E. Int. J. Vitam. Nutr. Res. 51:331-341.[Medline]
23. Smith, F. & Goodman, D. (1976) Vitamin A transport in human vitamin A toxicity. N. Engl. J. Med. 294:805-808.[Abstract]
24. Krasinski, S., Russell, R., Otradovec, C., Sadowsi, J., Hartz, S., Jacob, R. & McGandy, R. (1989) Relationship of vitamin A and vitamin E intake to fasting plasma retinol, retinol-binding protein, retinyl esters, carotene, 
-tocopherol, and cholesterol among elderly people and young adults: increased plasma retinyl esters among vitamin A supplement users. Am. J. Clin. Nutr. 49:112-120.
