One-time vitamin A supplementation of lactating sows enhances hepatic retinol in their offspring independent of dose size

¹ From the University of Wisconsin–Madison, Interdepartmental Graduate Program in Nutritional Sciences, Madison

² Presented in part orally at Experimental Biology 2004, Washington, DC.

³ Supported by NIH NIDDK grant 61973, USDA grant NRI 2003-35200-13754, and the University of Wisconsin Graduate School.

⁴ Address reprint requests to SA Tanumihardjo, Department of Nutritional Sciences, 1415 Linden Drive, Madison, WI 53706. E-mail: sherry{at}nutrisci.wisc.edu.

ABSTRACT  
Background:Single megadoses of vitamin A between 200 000 and 400 000 IU have been administered to lactating mothers to improve the vitamin A status of both mothers and breastfeeding infants. However, the most beneficial dosing regimen is not known.

Objective:The effect of megadoses of vitamin A supplements given to lactating sows on hepatic vitamin A concentrations in their nursing offspring was examined.

Design:Lactating sows were given a high (2.1 mmol), low (1.05 mmol), or control (0 mmol) dose of retinyl acetate in oil (n = 3 sows per treatment). Piglets nursed for 3 or 14 d, consumed a vitamin A–free diet for the next 4 d, and were then killed. Liver and serum samples were analyzed for vitamin A.

Results:After 3 d, piglets of the control, low-dose, and high-dose sows had different (P = 0.034) hepatic vitamin A concentrations, ie, 0.078 ± 0.004, 0.14 ± 0.053, and 0.13 ± 0.026 µmol/g, respectively. Liver vitamin A concentrations on day 18 were 0.069 ± 0.004, 0.14 ± 0.044, and 0.11 ± 0.026 µmol/g in the control, low-dose, and high-dose piglets, respectively (P = 0.017). Liver vitamin A concentrations in piglets of the low- and high-dose sows were not significantly different (day 3: P = 0.97; day 18: P = 0.59). Serum retinol concentrations were higher (P = 0.02) at early kill (0.95 ± 0.22 µmol/L) than at late kill (0.76 ± 0.24 µmol/L) but were not significantly different between groups.

Conclusions:Maternal vitamin A supplementation enhances liver vitamin A concentrations in offspring. Larger one-time doses are not more effective than are smaller doses. Additional research is needed to determine the most effective maternal dosing regimens for improving infant vitamin A status.

Key Words: Vitamin A supplementation • liver retinyl esters • pigs

INTRODUCTION  
Vitamin A deficiency (VAD) is known to be detrimental to health and is among the leading causes of preventable blindness in developing countries. The World Health Organization estimates that VAD affects between 100 and 250 million children worldwide (1, 2) and may, along with other nutrient deficiencies, be a major contributor to maternal death during childbirth (2). In communities at risk of VAD, supplementation of lactating women is recommended by the International Vitamin A Consultative Group (IVACG) for improving the vitamin A status of both mothers and their breastfeeding infants. In their most recent report, the IVACG advocates supplementing postpartum women with "400,000 IU as two doses of 200,000 IU at least1 d apart...as soon after delivery as possible and not more than 6 weeks later" (3). In contrast with the IVACG's split-dose supplementation recommendation, some interventions supply women one-time megadoses of 200 000–400 000 IU of vitamin A (4–6). The direct effect of a one-time maternal megadose of vitamin A on infant hepatic vitamin A reserves is unknown, because only surrogate measures of hepatic vitamin A are feasible in human studies.

Before this study was conducted we examined, in a lactating-sow model, the change in milk retinol concentrations in response to 3 doses of vitamin A. We measured milk retinol concentrations over a 48-h period and used the data to predict the benefit of vitamin A in infants. A theoretical average increase in human infant liver vitamin A was calculated to be 0.08 or 0.16 µmol/g for infants breastfeeding from mothers given 200 000 or 400 000 IU, respectively (7). Although the average increase in infant liver vitamin A theoretically doubles when mothers are given the higher dose, we were hesitant to state that the larger dose was more beneficial than the smaller dose. A wide variation in milk retinol concentrations among sows given the high dose resulted in no significant differences in milk retinol concentrations compared with sows given the low dose. Thus, we concluded that, because of the large variation in the maternal response to a higher dose, additional benefit to the infant from the high dose may not be realized when compared with the lower dose. The present study follows up on our previous work; we directly examined the effect of a single maternal dose of vitamin A on liver vitamin A stores in their offspring in a lactating-sow, nursing-piglet model. The swine was chosen as the model because of the similarity between the human and swine gastrointestinal tracts (8–12), use of swine as a model in previous vitamin A research (7, 13–15), the physiologic and anatomical similarity between human infants and piglets (8, 9), and the feasibility of collecting livers.

MATERIALS AND METHODS  
Sow diet and dosing regimen
Nine pregnant sows of the same crossbreed (Large White, Landrace) were housed in individual pens at the University of Wisconsin (UW)–Madison Swine Research and Teaching Center (SRTC) in Arlington, WI. Sows were fed a standard sow lactation diet; its composition was previously published (7). The UW Research Animal Resources Center approved all procedures.

Between 2 and 4 d after the sows gave birth, sows were randomly assigned to a treatment group and were given either a high (2.1 mmol), low (1.05 mmol), or control (0 mmol) dose of retinyl acetate (RAc) dissolved in corn oil (n = 3 sows per treatment). Because sow weights for this study were unavailable, the doses given were the same sizes as those used in our previous study (7). The sows of both studies were obtained from the SRTC, where they were similarly raised and were roughly the same age at the time of each study. In our earlier study, the dose was calculated to approximately replicate a 400 000-IU or 200 000-IU dose given to a 50-kg woman on a per kilogram basis, ie, 8.4 and 4.2 µmol RAc/kg, respectively.

To make the doses, crystalline RAc was dissolved, via sonication, into corn oil to make a stock solution. The concentration of the stock solution was determined by diluting 10 µL with 100 mL of hexanes and obtaining an absorbance value with the use of ultraviolet-visible spectroscopy. The stock solution was then diluted appropriately to create dosing solutions. The concentration of RAc/mL oil for each dosing solution was calculated by using the extinction coefficient E^1%_{1 cm} for retinol. E^1%_{1 cm} is the molar absorptivity of a 1% solution and therefore gives the solution's concentration in g/100 mL when used in Beer's Law. The high-dose solution contained 0.423 mmol retinol as RAc/mL oil and the low-dose solution contained 0.233 mmol retinol as RAc/mL oil. Thus, dose volumes containing 2.1 or 1.05 mmol RAc for the high or low doses were calculated as 5.0 and 4.5 mL solution, respectively. Sows in the control group were given 5 mL corn oil without added RAc. Doses were administered orally at the beginning of day 1 with the use of a 10-mL Terumo sterile syringe without a needle. The syringe was loaded with the dose and then inserted in the back corner of the sow's mouth and subsequently unloaded. The sows were watched to ascertain that they swallowed the solution. Sows that refused the dose were replaced.

The standard procedure at the SRTC is to breed a group of sows at the same time so that they can be cared for during pregnancy as a group; therefore, they farrow at approximately the same time. To guarantee that sows nursed similar numbers of piglets, the SRTC staff adjusted litter sizes by cross-fostering piglets (ie, removing piglets from a large litter and adding them to a smaller litter) as soon after birth as possible and before dosing began. Cross-fostering to equilibrate litter sizes makes competition among litter mates during nursing more uniform in terms of the number of piglets competing for food.

Ultimately, 3 male piglets per sow were required for this study. One sow in the group gave birth to only 2 male piglets. Therefore, the SRTC staff took a randomly chosen male piglet from a sow that farrowed 5 males and added it to the litter. This was done immediately after birth and before dosing began. After adjustment of litter sizes and before or on the same day as the sows were dosed, some piglets in several litters died. Sows nursed all of the surviving piglets in their adjusted litters throughout the study period. The mean number of live births and surviving piglets after litter adjustment per sow, as well as additional baseline information, are reported in Table 1. Baseline data were not significantly different between treatment groups.

View this table:
TABLE 1. Baseline characteristics of lactating sows given a control (0 mmol), low-dose (1.05 mmol), or high-dose (2.1 mmol) supplement of retinyl acetate¹

 
Piglet diet and tissue collection
After 3 d of nursing from dosed sows, 1 randomly selected male piglet per sow was electrically stunned and killed by exsanguination through jugular puncture; these piglets formed the early-kill group. Piglet weights were 2.43 ± 0.43 kg ( ± SD) and did not differ significantly between treatment groups (P = 0.44). Two samples of whole blood (8 mL each) were collected from jugular blood flow at the beginning of exsanguination. Immediately after the kill, livers were excised, blotted, weighed, and stored in brown plastic bottles on dry ice until they were brought to the laboratory and stored at –30 °C until analyzed. Blood samples were kept in the dark and allowed to clot at room temperature for 30 min before being transferred to our laboratory on the UW–Madison campus. Blood samples were centrifuged for 20 min (4000 rpm, 4 °C, 1250 On the morning of day 15 postdosing, 2 male piglets per sow were brought to the Livestock Laboratory on the UW–Madison campus, which formed the late-kill group. For sows that gave birth to >3 males, the staff at the STRC randomly selected the 2 males that were brought to campus. At the Livestock Laboratory, the piglets were randomly divided into 2 groups and housed in 2 pens to provide ample living space. On arrival, piglets were immediately weaned to a vitamin A–free oat and rice gruel, the composition of which was previously published (15). Piglets consumed the weaning diet for 4 d before being killed. We chose to put piglets on a vitamin A–free diet for a short period of time to determine whether the benefits of maternal supplementation would be sustained over a period of zero vitamin A intake, as might occur when young children of poor families in developing countries are weaned. This also allows for serum retinol concentrations to better reflect homeostatic concentrations when the dietary component is removed, a process that occurs between 2 and 4 d in rats (17).

Piglet weights at late kill, 6.35 ± 1.09 kg, were not significantly different between the maternal treatment groups. Livers and blood samples were collected and stored in the same manner as for the early-kill group. The timeline of sow dosing, piglet nursing, and killing is shown in Figure 1.

View larger version (14K):
FIGURE 1.. Experimental timeline. Sows (n = 3 per treatment group) were dosed with retinyl acetate (0, 1.05, or 2.1 mmol) at the beginning of day 1. Piglets in the early-kill group nursed from dosed sows for 3 d and then were killed. One male piglet per sow formed the early-kill group, resulting in 3 piglets per maternal treatment group. Piglets in the late-kill group (2 male piglets per sow or 6 piglets per maternal treatment group) nursed from dosed sows for 14 d and then were weaned to a vitamin A–free rice and oat gruel for 4 d before being killed. Liver and blood samples were collected from all piglets at the time of death.

 
Piglet serum analysis
Serum was analyzed for retinol with a method standardized in our laboratory with the use of 200 µL serum and equipment as described previously (15). In brief, after addition of RAc as an internal standard, serum proteins were denatured with 250 µL pure ethanol followed by 2 extractions with 300 µL hexanes, pooling organic layers. The organic layer was dried under inert gas and reconstituted in 40 µL of 75:25 methanol:dichloroethane, and 35 µL was injected into the HPLC system (injector: Rheodyne, Cotati, CA; detector: Shimadzu SPD-10A ultraviolet-visible radiation set at 350 nm, Kyoto, Japan; pump: Beckman, San Ramon, CA; data processor: Shimadzu C-R7A Chromatopac). A C₁₈, 5-µm, 15-cm reversed-phase column (Waters, Milford, MA) was used with a mobile phase of 89:11 methanol:water (0.73 g/L triethylamine as a modifier) at a flow rate of 1 mL/min. An HPLC purified retinol standard curve was run on the same system to quantify the retinol in the serum samples.

Piglet liver analysis
Livers were analyzed by using previously published methods (15). Briefly, a 1-g piece of liver randomly taken from the exterior surface was homogenized by mortar and pestle and dried by using 2–3 g anhydrous sodium sulfate. This mixture was extracted repeatedly by using dichloromethane and was brought to 50 mL volume. One milliliter of solution was dried under argon and reconstituted in 100 µL of 50:50 methanol:dichloroethane (by vol). Fifty microliters was injected into a dual-wavelength gradient HPLC system (detector: Waters 2487 Dual ; pump: Waters 600E Multisolvent Delivery System; data processor: Shimadzu CR7A plus). A Waters Resolve C₁₈ column (5 µm, 3.9 x 300 mm) served as the stationary phase, and absorbance was monitored at 325 nm. Two mobile phases were used: solvent A contained 93.5:7.5 acetonitrile:water (by vol) with 0.73 g triethylamine/L, and solvent B contained 85:10:5 acetonitrile:methanol:dichloroethane (by vol) with 0.73 g triethylamine/L. Each run started with 100% solvent A, with a switch to 100% solvent B by using a linear gradient from 0 to 3 min. From 3 to 35 min, 100% solvent B ran isocratically. Between 35 and 37 min, 100% solvent B switched to 100% solvent A by using a linear gradient. The flow rate was set at 2 mL/min. This flow rate and gradient scheme were used to optimize separation of retinol and its respective esters. The system was allowed to reequilibrate for 10 min between runs. Hepatic retinol and retinyl esters were quantified by using a retinyl butyrate standard curve and E^1%_{1 cm} for retinol. Retinol and retinyl ester values were summed (after they were individually quantified) and divided by the molecular weight of retinol to obtain the total vitamin A concentration (µmol/g liver) or total liver reserves (µmol/liver). Mean (±SD) piglet liver weights were not significantly different between treatment groups at either early or late kill and were 76 ± 15 and 161 ± 30 g, respectively.

Statistical analysis
Data from the early-kill piglets were ranked before analysis because of the small sample size. The ranked data were analyzed with PROC MIXED. Tukey's adjustments were used for comparison between maternal treatment groups. Sibling data from the late-kill piglets were averaged together before statistical analysis. The resulting values were not ranked. All late-kill data were analyzed by using PROC MIXED. Tukey's adjustments were applied to make comparisons between groups. The above statistical analyses were performed by using SAS (version 8.2; SAS Institute, Cary, NC). Paired t tests and regression analysis were applied where appropriate by using MINITAB (version 13; Minitab Incorporated, State College, PA). Data are presented as unranked means ± SDs. Results were considered significant at a P value 0.05.

RESULTS  
Early-kill piglets
Differences in liver vitamin A reserves (µmol/liver) of the ranked data among groups were nearly significant (P = 0.075). Liver reserves for early-kill piglets from the control-, low-, and high-dose groups were 5.7 ± 1.3, 12.2 ± 5.3, and 8.8 ± 2.3 µmol/g liver, respectively (Figure 2).

View larger version (16K):
FIGURE 2.. Mean (±SD) liver vitamin A reserves and concentrations in early-kill piglets. Because of the small sample size, data were ranked before ANOVA analysis, and Tukey's adjustments were used to make comparisons between groups. Liver vitamin A reserves were not significantly different between maternal treatment groups (P = 0.075). Liver vitamin A concentrations were significantly different between piglets from the 3 maternal treatment groups (P = 0.034). Piglets from the low-dose sows had higher liver concentrations than did the piglets from the control sows (P = 0.043). The difference in liver vitamin A concentrations between piglets from the high-dose sows and those from the control sows was nearly significant (P = 0.058). There was no significant difference in liver concentrations between piglets from the high-dose sows and those from the low-dose sows (P = 0.97). Means with different lowercase letters are significantly different (P < 0.05).

 
Hepatic concentrations of vitamin A (µmol/g) in early-kill piglets from the 3 maternal dose groups were different (P = 0.034). Mean hepatic concentrations in piglets from the control, low-dose, and high-dose sows were 0.078 ± 0.004, 0.14 ± 0.053, and 0.13 ± 0.026 µmol/g, respectively (Figure 2). Piglets from the low-dose sows had higher liver concentrations than did piglets from the control sows (P = 0.043), whereas the difference between piglets from the high-dose and control sows was nearly significant (P = 0.058). There was no significant difference in liver concentrations between piglets from the high- and low-dose sows (P = 0.97).

Serum retinol concentrations (µmol/L) were not significantly different between the 3 groups (P = 0.97; Table 2) and was not associated with hepatic vitamin A concentration (R² = 0.26).

View this table:
TABLE 2. Serum retinol concentrations in piglets killed 3 or 18 d after their mothers were given a control (0 mmol), low-dose (1.05 mmol), or high-dose (2.1 mmol) supplement of retinyl acetate¹

 
Late-kill piglets
Differences in hepatic vitamin A reserves (µmol/liver) in late-kill groups were significant (P = 0.026). Mean liver reserves for piglets from the control, low-dose, and high-dose sows were 11.3 ± 2.1, 24.9 ± 8.4, and 16.3 ± 1.9 µmol/liver, respectively (Figure 3). Liver reserves of piglets from high-dose sows were higher than those in the controls (P = 0.050). Liver reserves of piglets from low-dose sows tended to be greater than those of control sows (P = 0.078), whereas there was no significant difference between piglets from the low-dose and high-dose sows (P = 0.27).

View larger version (15K):
FIGURE 3.. Mean (±SD) liver vitamin A reserves and concentrations in late-kill piglets. Sibling data were averaged together, the resulting values were analyzed by using PROC MIXED, and Tukey's adjustments were used to make comparisons between groups. Liver vitamin A reserves were significantly different between maternal treatment groups (P = 0.026). Liver reserves of piglets from high-dose sows were higher than those of the control sows (P = 0.050), whereas liver reserves of piglets from low-dose sows tended to be greater than those of piglets from control sows (P = 0.078). Liver reserves were not significantly different (P = 0.27) between piglets from low-dose and high-dose sows. Means with different lowercase letters are significantly different (P < 0.05). Liver vitamin A concentrations were not significantly different between treatment groups (P = 0.017). Liver vitamin A concentrations in piglets from low-dose and high-dose sows tended to be greater than those in piglets from control sows (low-dose sows: P = 0.059; high-dose sows: P = 0.055). There was no significant difference between piglets from the low-dose and the high-dose sows (P = 0.59).  

 
Hepatic vitamin A concentrations (µmol/g) were also significantly different between groups at late kill (P = 0.017). Piglets from the control, low-dose, and high-dose sows had 0.069 ± 0.004, 0.14 ± 0.044, and 0.11 ± 0.026 µmol/g liver, respectively (Figure 3). Piglets from the low-dose sows tended toward higher liver vitamin A concentrations compared with piglets from the control sows (P = 0.059). The same trend was seen between piglets of high-dose sows compared with control sows (P = 0.055). Hepatic vitamin A concentrations of piglets from the high-dose sows were not significantly different from those of the low-dose piglets (P = 0.59).

Serum retinol concentrations (µmol/L) were not significantly different between groups (P = 0.33; Table 2) and had no relation to hepatic vitamin A concentration (R² = 0.31), a finding similar to that in early-kill piglets.

Serum retinol over time
Because there were no significant differences in serum retinol between maternal treatment groups at either time point, the early-kill data were combined as were the late-kill data to compare serum retinol over time. When the entire data set (late- and early-kill values) was ranked, the overall effect of time was nearly significant (P = 0.072), but the effect of treatment was not (P = 0.84 by two-factor ANOVA). Additionally, there was no time-by-treatment interaction (P = 0.76). A similar pattern was seen when the combined data set was not ranked. Serum retinol concentrations were 0.95 ± 0.22 and 0.76 ± 0.24 µmol/L at early and late kill, respectively, and these means were significantly different (P = 0.02; Table 2).

In summary, hepatic retinol concentrations in the offspring were enhanced by megadose supplementation with vitamin A to the mothers. This enhancement was maintained during lactation in a healthy animal model. However, there does not appear to be a dose-dependent accumulation. Vitamin A supplementation of the sows brought the piglets from a mean hepatic retinol concentration considered to be deficient (ie, 0.07 µmol/g liver) to one considered to be adequate.

DISCUSSION  
Attempts to improve vitamin A status through a single 200 000-IU oral dose administered to lactating mothers have been effective in both mothers (4, 18–20) and infants (4, 19, 20). Yet, the duration and type of improvements vary among studies. Maternal intervention with 200 000 IU vitamin A did not completely correct VAD in mothers and infants when the prevalence of VAD after supplementation was a measured outcome (4, 19). Thus, Rice et al (4) recommend the use of higher maternal doses, whereas Bahl et al (19) encourage additional supplementation strategies.

Although the IVACG recommendation is to supplement lactating women with 400 000 IU vitamin A as a split-dose (200 000 IU 1 d apart), some interventions have provided lactating women with 300 000–400 000 IU vitamin A as a single dose (5, 6, 21). A one-time dose is desirable in developing countries because it does not require follow-up and is less expensive. The trials using one-time doses >200 000 IU resulted in beneficial effects, such as increased vitamin A concentrations in mothers' breast milk, increased serum retinol concentrations in mothers, increased serum retinol concentrations in infants, and improved infant vitamin A status as assessed with the use of relative dose-response or modified relative dose-response tests. Again, similar to trials that evaluated doses of 200 000 IU, the type and duration of benefits were variable between studies. Moreover, effects of these dosing regimens on infant vitamin A status may be questionable because assessment was via surrogate measures of liver reserves, which have limitations as indicators of vitamin A status (16). Because there is overlap in outcomes from studies that used various one-time doses and no study has examined infant liver reserves directly, the following question arises: Is a one-time maternal dose of 400 000 IU more effective than a dose of 200 000 IU at improving infant vitamin A status?

To our knowledge, no previous study has addressed the question of the benefit to infants of larger compared with smaller doses of vitamin A to the mothers. Using a lactating-sow, nursing-piglet model, we examined how a high compared with a low maternal dose of vitamin A affected the vitamin A status of the offspring, assessed on the basis of liver vitamin A concentrations. Doses were chosen to parallel the 400 000 and 200 000 IU given to postpartum women on a body weight basis. Piglets of sows receiving vitamin A had significantly higher liver vitamin A concentrations (µmol/g) than did the control piglets at both 3 and 18 d after the intervention. Vitamin A reserves (µmol/g liver) of piglets from dosed sows were also greater than those of control piglets at 18 d. There were no significant differences in liver concentrations and reserves of vitamin A between piglets from the low- and high-dose sows at either time point. These data suggest that there are no differences in benefit to the infants between the 2 different maternal doses of vitamin A.

A potential explanation for the lack of difference between treatment groups is a physiologic limit to the amount of vitamin A that can be incorporated into milk. Our laboratory conducted previous research that provides evidence in support of this hypothesis. Sows given the same doses of vitamin A as used in this study showed no difference in milk retinol concentrations up to 48 h after dosing (7). Hepatic vitamin A concentration and the combined serum concentration of long-chain retinyl esters, up to 8 h postdose, were significantly higher in sows given the high dose than in sows given the low dose of vitamin A (22). Thus, we do not believe that intestinal absorption explains the lack of difference in milk concentrations between sows receiving a high compared with a low dose of vitamin A. Rather, we hypothesize that chylomicron vitamin A uptake by the mammary gland is limiting. As yet, the mechanisms by which vitamin A is incorporated into mammary tissue and breast milk are not well understood. Ross et al (23) recently showed that lipoprotein lipase may be a major player in vitamin A transfer from chylomicrons to mammary tissue. Saturated lipoprotein lipase activity after the smaller dose would explain why no additional vitamin A is incorporated into milk with the higher dose.

Although significantly more vitamin A was detected in the livers of piglets nursing from sows dosed with vitamin A, no differences in serum retinol concentrations were detected. Our study supports the conclusions of other studies that more rigorous assessment methods than serum retinol alone need to be considered when evaluating the results of supplementation trials (16, 24). Of note, serum retinol decreased over time. As the piglets' diet changed at weaning from a diet containing vitamin A to one containing no vitamin A, the decrease in serum retinol concentrations may have been due to a new homeostatic "set point." This may indicate a decrease in vitamin A utilization rate and a shift in vitamin A mobilization from the stellate cell to hepatocyte as the body adapts to the dietary change (25).

Regarding the safety of vitamin A megadoses, acute toxicity may be a concern for the mother. The tolerable upper intake level for lactating women in the United States is 2800 µg/d between the ages of 14 and 18 y and is 3000 µg/d between the ages of 19 and 50 y (26). This translates to the 400 000 IU dose being roughly 40 times the tolerable upper intake level, and it is unknown how undernourished, and possibly metabolically and immunologically compromised, women in developing countries metabolize such a bolus. We found evidence supporting the possibility that higher megadoses of retinyl ester induce greater vitamin A catabolism than do lower doses of vitamin A in vitamin A–sufficient sows (22). Again, the doses used were the same as those in this study. Higher circulating concentrations of retinyl ß-glucuronide, a metabolite of retinol, were found in the high-dose sows than in the low-dose and control sows (22). Additionally, higher concentrations of retinoyl ß-glucuronide, a metabolite of retinoic acid, were found in the high-dose sows than in the control sows (22). Serum retinol concentrations did not differ significantly between the high- and low-dose sows. This finding may indicate that the larger dose invokes a mechanism that protects against acute toxicity from the formation of retinoic acid, at least in a vitamin A–and protein-sufficient animal model. It is possible that women who are not regularly consuming adequate protein or vitamin A may not synthesize the enzymes required for this type of retinoid catabolism and, therefore, could be at higher risk of acute retinoic acid toxicity after supplementation with megadoses of vitamin A.

The effects of one-time maternal supplementation with vitamin A on infant vitamin A status in the present study have implications for the continued development of supplementation programs and policy concerning mothers and children with VAD. It appears that there are no added benefits to the infant when a larger one-time maternal dose of vitamin A (400 000 compared with 200 000 IU) is used. Thus, it is advisable—because of potential toxic side effects—to consider the welfare of the mother and use smaller doses of vitamin A in supplementation programs (100 000–200 000 IU), possibly multiple times, as per the IVACG recommendation. Convincing evidence in animals supports the ability of daily vitamin A intake during lactation to enhance milk vitamin A concentrations (27, 28). In humans, Tanumihardjo et al (29) showed that a daily intake of 8000 IU vitamin A for 35 d improved the status of lactating women in Indonesia, as assessed by both serum retinol concentrations and the modified-relative-dose-response test. Villard and Bates (30) showed that Gambian women given 2200 IU vitamin A/d during pregnancy and lactation had significantly greater (by 23%) breast milk vitamin A concentrations during lactation than did the group not receiving supplements.

Ideally, adequate daily intake of vitamin A through consumption of vitamin A–rich foods, foods containing provitamin A carotenoids, or daily low-dose supplements is preferable to one-time supplementation with a megadose of vitamin A. However, this is not always economically feasible in poor communities and may not correct VAD quickly enough to protect women from pregnancy-induced night blindness and children from VAD disorders (3). As changes to dietary habits are sought, supplementation remains the most useful method for reducing VAD. This study clearly showed no added benefit to the infants when 400 000 IU as opposed to 200 000 IU vitamin A was administered as a single bolus to the mother. Further investigation is required to determine whether multiple vitamin A doses over several days is most effective at prolonging benefits and reducing the occurrence of VAD in mothers and infants. Moreover, the timing of the doses administered to mothers postpartum needs to be evaluated for optimal effectiveness.

ACKNOWLEDGMENTS  
We thank Tom Crenshaw for animal care guidance and Kristina Penniston for analytic assistance.

ARV was responsible for orchestrating the study, the sample and statistical analysis, and the manuscript preparation. All work was conducted in the laboratory of SAT, who was responsible for the study design, for overseeing study operation, for preparing the doses, for input into the data analysis, and for the manuscript preparation and revision. None of the authors had any financial interest in the work or a conflict of interest with the sponsors of this study.

REFERENCES  

1. World Health Organization. Immunizations, vaccines, and biologicals. Vitamin A supplementation. May 2003. Internet: http://www.who.int/vaccines/en/vitamina.shtml (accessed 25 June 2004).
2. World Health Organization. Nutrition. Micronutrient deficiencies: combating vitamin A deficiency, the challenge. 3 September 2003. Internet: http://www.who.int/nut/vad.htm (accessed 25 June 2004).
3. Sommer A, Davidson FR. Assessment and control of vitamin A deficiency: the Annecy accords. J Nutr 2002;132:2845S-50.
4. Rice AL, Stoltzfus RJ, de Francisco A, Chakraborty J, Kjolhede CL, Wahed MA. Maternal vitamin A or ß-carotene supplementation in lactating Bangladeshi women benefits mothers and infants but does not prevent subclinical deficiency. J Nutr 1999;129:356-65.
5. Stoltzfus RJ, Hakimi M, Miller KW, et al. High dose vitamin A supplementation of breast-feeding Indonesian mothers: effects on the vitamin A status of mother and infant. J Nutr 1993;123:666-75.
6. Ayah R, Tedstone A, Marshall T, Friis H, Michaelsen KF, Mwaniki DK. High-dose maternal and infant vitamin A supplementation in rural Kenya. XXI IVACG Meeting. Washington, DC: IVACG Secretariat, 2003;45 (abstr T62).
7. Penniston KL, Valentine AR, Tanumihardjo SA. A theoretical increase in infants' hepatic vitamin A is realized using a supplemented lactating sow model. J Nutr 2003;133:1139-42.
8. Book SA, Bustad LK. The fetal and neonatal pig in biomedical research. J Anim Sci 1974;38:997-1002.
9. Mount LE. The climatic physiology of the pig. London: Edward Arnold (Publishers) Ltd, 1968.
10. Miller ER, Ullrey DE. The pig as a model for human nutrition. Annu Rev Nutr 1987;7:361-82.
11. Tumbleson M, ed. Swine in biomedical research. New York: Plenum Press, 1986.
12. Leddin DJ. Characterization of small intestinal water, sodium, and potassium transport and morphology in the pig. Can J Physiol Pharmacol 1992;70:113-5.
13. Arnhold T, Nau H, Meyer S, Rothkoetter HJ, Lampen AD. Porcine intestinal metabolism of excess vitamin A differs following vitamin A supplementation and liver consumption. J Nutr 2002;132:197-203.
14. Cooper DA, Olson JA. Properties of liver retinyl ester hydrolase in young pigs. Biochim Biophys Acta 1986;884:251-8.
15. Valentine AR, Tanumihardjo SA. Adjustments to the modified relative dose response (MRDR) test for assessment of vitamin A status minimize the blood volume used in piglets. J Nutr 2004;134:1186-92.
16. Tanumihardjo SA. Assessing vitamin A status: past, present and future. J Nutr 2004;134:290S-3.
17. Tanumihardjo SA. Vitamin A status assessment in rats using ¹³C₄-retinyl acetate and gas chromatography-combustion isotope ratio mass spectrometry (GCCIRMS). J Nutr 2000;130:2844-9.
18. Basu S, Sengupta B, Paladhi PK. Single megadose vitamin A supplementation of Indian mothers and morbidity in breastfed young infants. Postgrad Med J 2003;79:397-402.
19. Bahl R, Bhandari N, Wahed MA, Kumar GT, Bhan MK, the WHO/CHD Immunization-Linked Vitamin A Group. Vitamin A supplementation of women postpartum and of their infants at immunization alters breast milk retinol and infant vitamin A status. J Nutr 2002;132:3243-8.
20. Roy SK, Islam A, Molla A, Akramuzzaman SM, Jahan F, Fuchs G. Impact of a single megadose of vitamin A at delivery on breastmilk of mothers and morbidity of their infants. Eur J Clin Nutr 1997;51:302-7.
21. Humphrey JH, Quinn T, Fine D, et al. Short-term effects of large-dose vitamin A supplementation on viral load and immune response in HIV-infected women. J Acquir Immune Defic Syndr Hum Retrovirol 1999;20:44-51.
22. Penniston KL, Tanumihardjo SA. High doses of retinyl ester induce transient increases in 4-oxoretinol and ß-glucuronide metabolites in lactating sows. FASEB J 2004;18 (abstr 272.6).
23. Ross AC, Pasatiempo AMG, Green MH. Chylomicron margination, lipolysis, and vitamin A uptake in the lactating rat mammary gland: implications for milk retinoid content. Exp Biol Med 2004;229:46-55.
24. Tanumihardjo SA. Can lack of improvement in vitamin A status indicators be explained by little or no overall change in vitamin A status of humans? J Nutr 2001;131:3316-8.
25. Blomhoff R, Green MH, Green JB, Berg T, Norum KR. Vitamin A metabolism: new perspectives on absorption, transport, and storage. Physiol Rev 1991;71:951-90.
26. Institute of Medicine Food and Nutrition Board. Vitamin A. Dietary reference intakes: vitamin A, vitamin K, arsenic, boron, chromium, copper, iodine, iron, molybdenum, nickel, silicon, vanadium, and zinc. Washington, DC: National Academy Press, 2001:82-161.
27. Green MH, Green JB, Akohoue SA, Kelley SK. Vitamin A intake affects the contribution of chylomicrons vs. retinol-binding protein to milk vitamin A in lactating rats. J Nutr 2001;131:1279-82.
28. Haskell M, Brown K. Maternal vitamin A nutriture and the vitamin A content of human milk. J Mammary Gland Biol Neoplasia 1999;4:243-57.
29. Tanumihardjo SA, Muherdiyantiningsih, Permaesih D, et al. Daily supplements of vitamin A (8.4 µmol, 8000 IU) improve the vitamin A status of lactating Indonesian women. Am J Clin Nutr 1996;63:32-5.
30. Villard L, Bates C. Effect of vitamin A supplementation on plasma and breast milk vitamin A levels in poorly nourished Gambian women. Hum Nutr Clin Nutr 1987;41:47-58.