The majority of pediatric HIV infections are acquired through maternal-infant transmission, with reported rates varying from 15 to 50% depending on the geographic location [1–8]. Rates in the United States have been reported at about 15–30%. Factors associated with increased risk of vertical transmission include low maternal CD4+ cell count or percentage [3,8–10], preterm birth , duration of membrane rupture (DMR) , mode of delivery , maternal p24 antigenemia [11,12], syncytium-inducing or slow growing viral strain , high maternal viral load [14,15], hard drug use (that is, use of drugs including heroin, cocaine or methadone, or injecting drug use) during pregnancy [9,16] and unprotected heterosexual intercourse during pregnancy . A recent study found that maternal vitamin A deficiency among pregnant women in Malawi was associated with a fourfold increased risk of maternal-infant transmission . A study from Kenya reported a 20-fold increased risk of detectable HIV-1 DNA in the breastmilk of severely vitamin A-deficient pregnant women with low CD4+ cell counts (<400 × 106/l) , raising the possibility that some of the excess risk of perinatal transmission among the vitamin A-deficient women in the Malawi study was associated with the breastfeeding. However, it is unknown whether vitamin A deficiency is associated with maternal-infant transmission of HIV in developed countries. We conducted a study of vitamin A and maternal-infant transmission in a group of prospectively followed HIV-infected pregnant, non-breastfeeding women in two metropolitan areas in the US.
The study population was drawn from HIV-seropositive women who were followed at the University of Maryland (Baltimore, Maryland, USA) and Montefiore Medical Center/Albert Einstein College of Medicine (Bronx, New York, USA) as part of a multicenter, prospective study of maternal-infant transmission, which has been described in detail elsewhere . Women were enrolled during pregnancy or up to 2 weeks postpartum. Women were screened in the antenatal clinic for HIV antibody using enzyme immunoassay (EIA) with confirmatory Western blot (Abbott Laboratories, North Chicago, Illinois, USA, up to Feb. 1988, and thereafter Dupont, Wilmington, Delaware, USA). After informed consent, HIV-infected women were enrolled in the study. None of these women breastfed their infants. A serum sample was obtained by venipuncture and stored at −70°C. Maternal prenatal care and primary pediatric care were provided by study personnel. Mothers were seen during each trimester, at delivery, and periodically up to 6 months postpartum. Demographic information, HIV risk factors, medical and drug use histories were collected and CD4+ cell count and percentage were determined at each visit as part of a standard protocol. Information on zidovudine use during pregnancy was obtained from maternal self-report and medical record abstraction. Intrapartum obstetric information was collected by medical chart review. The timing of the last menstrual period, or the Ballard score, was abstracted from the medical chart and used to determine the gestational age. Infants were scheduled for evaluation by medical history, physical examination and phlebotomy at birth to 1 week of life, and every 2–3 months during the first 18 months of life. CD4 lymphocyte subset analysis was performed at the Centers for Disease Control and Prevention (CDC), Atlanta, Georgia, USA, using flow cytometry. HIV testing was performed at the CDC using the EIA and confirmatory Western blot for women once they were enrolled in the study and for infants. Specimens were considered positive by Western blot when two of three core envelope proteins were present (p24, gp41 and gp120/160). Viral culture and polymerase chain reaction (PCR) were performed as previously described . All women who delivered from May 1986 through May 1994 were included in the current analysis if there was an available serum sample during the third trimester and the infection status of their child was known.
Serum vitamin A was measured by high-performance liquid chromatography on third trimester samples in a masked fashion . Quality control was monitored with vitamin A reference standards from the National Institute of Standards and Technology (Gaithersburg, Maryland, USA).
A child was considered infected if one of the following applied: (i) two separate blood specimens were positive for HIV by PCR; (ii) presence of HIV antibodies persisting after 15 months of age [21,22]; or (iii) the 1987 CDC criteria  were met for class P2 with either an AIDS-defining illness or HIV-related death and one positive PCR.
A child was considered uninfected if no AIDS-defining illness had been diagnosed and HIV serology was negative after 15 months of age; or no AIDS-defining illness and at least two samples were negative by PCR and no samples tested positive.
Vitamin A deficiency in adults was defined as serum vitamin of 0.70–1.04 μmol/l and severe deficiency as <0.70 μmol/l . Serum vitamin A levels were divided into three groups as per convention: severe deficiency, <0.70 μmol/l; mild to moderate deficiency, 0.70–1.04 μmol/l; normal, ≥ 1.05 μmol/l. Gestational age was categorized as <37 (preterm, prematurity) and ≥ 37 (term) weeks . DMR was categorized as <4 and ≥ 4 h . Maternal body mass index (BMI) at the time the specimen was obtained was calculated as weight/height squared  as an indicator of protein-energy status. Median BMI was presented because there are no standard reference charts available on pregnant women for converting BMI to percentiles. CD4+ cell counts were categorized as ≥ 500, 200–499 and <200 × 106/l. Percentage of CD4+ lymphocytes was categorized as <14, 14–28 and >28%.
Univariate comparisons for categorized variables were tested with Fisher's exact test or χ2 test. Mantel-Haenszel χ2 test for trend was used to compare vitamin A categories and CD4 categories by transmission group. Medians were compared using Wilcoxon rank-sum test or Kruskal-Wallis test as appropriate. Multivariate logistic regression analysis was used to determine the independent association of vitamin A deficiency with transmission while controlling for effect modifiers. Interaction terms and confounding were evaluated; interaction terms were not significant. The full model included factors found to be significant on univariate analysis in this sample or risk factors reported in the literature (maternal age, mode of delivery, DMR, infant gestational age and CD4+ percentage, injecting drug use, and race/ethnicity). Given the high correlation (r = 0.97) between gestational age and maternal BMI, only gestational age was included in the multivariate model. Modelling was accomplished by sequential removal of each variable. Final selection was based on significant likelihood ratio statistics and examination of the change in the regression coefficients . The adjusted odds ratio (AOR) and the 95% confidence intervals (CI) are presented. All tests are two-tailed with α= 0.05 as significant. All analyses were performed using SAS statistical software (SAS Institute, Cary, North Carolina, USA).
Forty-four HIV-infected women who transmitted HIV to their infants (TR) and 89 women who did not transmit HIV to their infants (NTR) were evaluated for third trimester serum vitamin A levels. The sample consisted of 70% of the HIV-seropositive women enrolled in the two sites who had infants with known infection status; the other 30% did not have an available third trimester serum sample either because they were enrolled postpartum or because there was inadequate sample stored. At the time of the study, the maternal-infant transmission rate in the Baltimore cohort was 25% and in the Montefiore cohort was 24.5% (total number of infected children divided by the total number of children born to HIV-seropositive mothers with known infection status). The study sample was 52% (70 out of 133) black, 34% (45 out of 133) Hispanic and 14% (18 out of 133) white with 88% (117 out of 133) aged ≤ 35 years and 22% (29 out of 133) reporting injecting drug use during pregnancy. Forty-three per cent (56 out of 129) had CD4+ counts ≥ 500 × 106/l, 42% (54 out of 129) had 200–499 × 106/l, and 15% (19 out of 129) had <200 × 106/l; median CD4+ count was 471 for TR mothers and 455 for NTR mothers. Fifty-five per cent (71 out of 130) had CD4+ percentage of <14%, 37% (48 out of 130) had CD4+ percentage of 14–28%, and 8% (11 out of 130) had CD4+ percentage of ≥ 29%. Twelve per cent (16 out of 133) had delivery by Cesarean section, 51% (66 out of 129) had DMR of ≥ 4h, and 22% (28 out of 133) had preterm births. Twenty-three per cent (31 out of 133) of the women had serum vitamin A levels consistent with mild-to-moderate deficiency (0.70–1.04 μmol/l) and 9% (nine out of 133) had levels consistent with severe deficiency (<0.70 μmol/l). Except for the distribution of DMR, there was no difference in demographic and clinical characteristics of this study sample compared with the total study population of women enrolled in the prospective study in Baltimore and the Bronx (data not shown). This study sample had significantly fewer women with DMR ≥ 4 h than the total population of women enrolled in the prospective study: 51% (66 out of 129) versus 65% (245 out of 1611; P = 0.02).
The distribution of demographic and clinical characteristics by maternal transmission status is presented in Table 1. The distribution of vitamin A levels by transmission group tended to be different (P = 0.10). However, a significantly greater proportion of TR mothers than NTR mothers had serum vitamin A levels consistent with severe deficiency (16 versus 6%; P = 0.05).
A greater proportion of the TR mothers than NTR mothers had premature infants (33 versus 15% <37 weeks gestation; P = 0.02), longer DMR (63 versus 45% ≥ 4 h; P = 0.06), and delivery by Cesarean section (20 versus 8%; P = 0.04). There were no significant differences between TR and NTR mothers in age distribution, race/ethnicity, CD4+ cell count, or injecting drug use during pregnancy (Table 1). A significantly greater proportion of TR mothers had CD4+ percentage <14% than NTR mothers: 16% (seven out of 44) compared with 5% (four out of 86; P = 0.03); 64% of women with CD4+ percentage <14% transmitted HIV to their infants compared with 31% with CD4+ percentage 14–28%, and 30% with ≥ 29% (P = 0.09). Except for CD4+ cell count, there were no significant differences for demographic and clinical characteristics between women who took zidovudine and those who did not take zidovudine during pregnancy. This study was carried out prior to the use of zidovudine during pregnancy to prevent vertical transmission; zidovudine use was associated with clinical indications of AIDS. As expected, those women who took zidovudine some time during the pregnancy had a significantly lower median CD4+ cell count (332 versus 475 × 106/l; P = 0.03); 38% of those who took zidovudine had CD4+ cell counts of lower than 200 × 106/l compared with 13% of those women who did not take zidovudine (P = 0.09).
The frequency distribution (range) of vitamin A levels by race was similar, although mean vitamin A levels were significantly lower in blacks (mean vitamin A, 1.15 μmol/l) than in Hispanics (mean vitamin A, 1.37 μmol/l) and whites/other (mean vitamin A, 1.33 μmol/l; analysis of variance, P = 0.03). Seventeen per cent of blacks had Cesarean sections compared to 10% of white/other and 7% of Hispanics (P = 0.26). A greater proportion of blacks had preterm births (30% for blacks, 12% for Hispanics and 11% for white/other; P = 0.03) and longer DMR (58% blacks had DMR ≥ 4 h compared with 28% for whites and 50% for Hispanics; P = 0.08). A greater proportion of women with ≥ 4 h DMR were vitamin A-deficient compared to women with <4 h DMR (57 versus 47%; P = 0.30). There was no difference in the distribution of vitamin A by delivery type or gestational age.
With increasing levels of prematurity, a greater proportion of women had prolonged DMR: 46% of women with term infants had prolonged DMR compared to 62% of women with premature infants (35–36 weeks gestation) and 73% of women with severely premature infants (<35 weeks gestation; P = 0.03). A greater proportion of women who had prolonged DMR had premature infants (29 versus 14%; P = 0.04).
All but one of the Cesarean section deliveries were non-scheduled and were performed secondary to maternal/fetal complications. The one scheduled Cesarean section was in a woman with a previous Cesarean section delivery. Women who had Cesarean section deliveries had significantly longer DMR (28 versus 10 h; P = 0.03). Although there was no significant difference in the proportion of preterm infants (<37 weeks gestation) by delivery type (20 versus 31%; P = 0.33), a significantly greater proportion of the women who had Cesarean section deliveries had severely premature infants (<35 weeks gestation), 31% (five out of 16) versus 9% (10 out of 115; P = 0.02). The gestational age for all the preterm deliveries among the women who had Cesarean section deliveries was <35 weeks.
Table 2 shows the results of the final multivariate logistic regression model. Severe vitamin A deficiency was independently associated with a 5.05 (95% CI, 1.20–21.24) increase in the risk of maternal-infant HIV transmission after controlling for the effects of CD4+ percentage, gestational age, DMR and mode of delivery. Cesarean section and prematurity were also independently associated with an increased risk of HIV transmission (Cesarean section: AOR, 3.75; 95% CI, 1.10–12.87; prematurity: AOR, 2.25; 95% CI, 1.22–4.13). CD4+ percentage was not independently associated with increased risk of transmission. There was no evidence of an interaction between Cesarean section and gestational age or gestational age and DMR. Injecting drug use during pregnancy was evaluated in the multivariate model and did not contribute significantly to the model, nor did it change the AOR of any of the covariates in the final model or main effect variable in the final model. Although race was not significantly associated with transmission or with vitamin A distribution, the final model was also run controlling for possible confounding effects of race. The resulting AOR remained relatively unchanged compared to the AOR when race was not included (data not shown).
Given the increased risk of transmission associated with Cesarean section and the fact that all women who had Cesarean section delivery had severely premature infants, a third model was run with gestational age as premature (35–36 weeks gestation) and severely premature (<35 weeks gestation). The results are shown in Table 3. Severe vitamin A deficiency was associated with a four-to-fivefold increased risk (AOR, 4.57; 95% CI, 1.08–19.44) and severe prematurity with a six-to-sevenfold increased risk (AOR, 6.51; 95% CI, 1.52–27.86) of maternal-infant HIV transmission in the final model. Cesarean section was no longer a significant factor.
Although vitamin A deficiency is generally regarded as infrequent in the United States, this study suggests that vitamin A deficiency is relatively common (30%) among inner-city HIV-infected pregnant women. Furthermore, HIV-infected women who were severely vitamin A-deficient during pregnancy had four-to-five-fold increased risk of transmitting HIV infection to their infants. This study corroborates findings from Malawi  and Kenya  where maternal vitamin A deficiency may increase the risk of maternal-infant transmission of HIV and extends them in that this association is seen in a population of non-breastfeeding women. These findings suggest that the increased rate of maternal-infant transmission seen among vitamin A deficient breastfeeding African women does not wholly result from post-natal transmission associated with high viral load in breastmilk.
Pregnancy, HIV infection, low socioeconomic status and race may all be associated with vitamin A deficiency. Pregnant women are at a higher risk of vitamin A deficiency because of increased demands for vitamin A by the developing fetus . Night-blindness, the earliest clinical manifestation of vitamin A deficiency, is common enough among pregnant women in some cultures to be considered a symptom of pregnancy along with morning sickness . In the early part of this century in Europe, epidemics of night-blindness were described among pregnant women  prior to improvements in diet. HIV infection may predispose women to vitamin A deficiency as a result of many possible factors including low intake and malabsorption of vitamin A-rich foods, liver disease, and abnormal urinary losses of vitamin A during infection .
Recently, it has been demonstrated that vitamin A deficiency is common among pregnant women of low socioeconomic status  and among minority groups . Our study population in Baltimore and New York City consisted primarily of black and Hispanic women of low socioeconomic status, many former or current injecting drug users. Although this study was limited in that we did not perform a dietary assessment, a low intake of vitamin A-rich foods and lack of use of antenatal vitamin supplements were among the likely factors contributing to vitamin A deficiency in these women in addition to HIV infection. Several factors in this population, including injecting drug use and HIV infection, may contribute to vitamin A deficiency.
Other risk factors for increased maternal-infant transmission in this study included Cesarean section and preterm delivery. Women with Cesarean section delivery had a four-to-fivefold increased risk of transmitting HIV to their infants. Other studies that have examined the relationship between mode of delivery and maternal-infant HIV transmission reported a protective factor or no association with maternal-infant transmission [9,32,33]. It should be noted that all but one of the Cesarean sections in the present study were performed secondary to maternal or fetal complications in women with other risk factors for vertical transmission. Thus, in this sample of pregnant women Cesarean section delivery was not performed for the purpose of reducing the likelihood of maternal-infant transmission and might explain why Cesarean section delivery was associated with increased risk of maternal-infant transmission, results that contrast with other reports where Cesarean section delivery was found to be protective . Further analysis of the women who had Cesarean section delivery revealed that all had severely premature infants. When severe prematurity was included in a final multivariate analysis Cesarean section delivery was no longer associated with increased risk of maternal-infant transmission. Our results suggest that any evaluation of Cesarean section and risk of maternal-infant transmission should consider the reason for Cesarean section delivery.
The twofold increased risk associated with prematurity is in accordance with previous findings of increased risk of maternal-infant transmission with premature birth [1,34]. Others have reported an increased risk with low birth weight [9,14,35]. In our study, low birth weight and infant gestational age were extremely highly correlated. Increased fetal susceptibility to infection associated with immature fetal immune development in these infants may be a possible mechanism for the increased risk of HIV acquisition in these infants. Other maternal factors, such as infection, may lead to in utero growth retardation and increased HIV transmission .
We were unable to corroborate recent reports of an independent association of prolonged DMR and increased maternal-infant HIV transmission with prolonged DMR . In our sample a significantly greater proportion of women had prolonged DMR with increasing levels of premature delivery. The six-to-sevenfold increased transmission risk associated with severe prematurity in our sample of women may have modified the association of DMR and maternal-infant transmission.
Recent studies report increased odds of 1.9–3.1 of maternal-infant HIV transmission associated with maternal hard drug use (including heroin, cocaine or methadone, or injecting drug use) during pregnancy [9,14]. In our study, injecting drug use was not associated with maternal-infant transmission, which is consistent with previous results from the New York City Collaborative Perinatal Transmission Study Group, which found no relationship between hard drug use and maternal-infant HIV transmission . Recent data from this group indicate that hard drug use was associated with prolonged DMR and gestational age, factors associated with increased maternal-infant transmission (P. B. Matheson, personal communication, 1996). Hard drug use was not independently associated with transmission when DMR and gestational age were considered, suggesting that hard drug use is a marker for other risk factors of maternal-infant transmission rather than a risk factor itself. In our study, only injecting drug use was considered; crack and cocaine use by routes other than injection were not included, which may also contribute to the lack of association found in the present study.
The relationship of low CD4+ cell count or CD4 percentage as a measure of increased severity of disease and increased maternal-infant HIV transmission has been reported in several studies [3,8–10,36], whereas others have not found this association [1,34]. In our study, there was a significant association of low CD4+ lymphocyte percentage and increased risk of maternal-infant transmission on univariate analysis that was not maintained with multivariate analysis. There was no association with CD4+ cell count, possibly due to the nature of the distribution of CD4+counts in our sample and the imprecise nature of the measurement. It is not entirely clear why our results differ from dominant findings reported from other studies of low CD4+ cell count and increased risk of maternal-infant transmission.
The fourfold increased transmission risk associated with severe vitamin A deficiency in our study is consistent with the fourfold increased transmission risk among African women  with severe vitamin A deficiency and may be related to the role vitamin A as an immune enhancer. Vitamin A is essential for maintaining the integrity of the mucosal surfaces, for modulating healthy antibody responses, and for the function and growth of T and B cells . The increased risk of maternal-infant transmission associated with maternal vitamin A deficiency may result from altered mucosal immunity, altered pathology of the reproductive tract, or abnormalities of the placenta. Vitamin A deficiency is also associated with atrophy of the thymus, spleen and lymphoid tissue . Thus, another possible mechanism for increased risk of maternal-infant HIV transmission associated vitamin A deficiency may be related to aberrant fetal development of the immune system and increased fetal susceptibility to infection. Recent data suggest that vitamin A deficiency is associated with increased viral load in breastmilk, another potential mechanism by which vitamin A deficiency could have an impact on maternal-infant transmission . Vitamin A deficiency is associated with chorioamnionitis in animal models ; in humans, chorioamnionitis is associated with increased maternal-infant transmission of HIV .
Significant micronutrient deficiencies may occur during HIV infection. This study is limited in that other micronutrients were not assessed. Micronutrient deficiencies which may be common during pregnancy and play a role in immunity and birth outcomes include iron, folate and zinc. Further studies are needed to determine whether these and other micronutrient deficiencies are associated with maternal-infant transmission of HIV.
It is not known whether vitamin A deficiency in HIV-infected pregnant women augments maternal-infant transmission or is a marker for advanced HIV disease. With the generalized inflammation that may occur with the opportunistic infections associated with HIV disease, mobilization of vitamin A from the liver can be impaired despite adequate vitamin A reserves, and abnormal urinary losses of vitamin A can occur . There are currently four clinical trials in progress in sub-Saharan Africa where both vitamin A deficiency and HIV infection are common . These trials, which involve antenatal micronutrient supplementation during the second and third trimester in over 3000 HIV-infected pregnant women, may determine whether vitamin A supplementation could reduce maternal-infant transmission of HIV.
The authors greatly appreciate the programming assistance of D. Buono.
1. Goedert JJ, Mendez H, Drummond JE, et al.
: Mother-to-infant transmission of human immunodeficiency virus type 1: association with prematurity or low anti-p120
2. Ryder RW, Nsa W, Hassig SE, et al.
: Perinatal transmission of the human immunodeficiency virus type 1 to infants of seropositive women in Zaïre
. N Engl J Med
3. Blanche S, Rouzioux C, Guihard Moscato ML, et al.
: A prospective study of infants born to women seropositive for human immunodeficiency virus type 1
. N Engl J Med
4. European Collaborative Study: Risk factors for mother to child transmission of HIV-1
5. Dabis F, Msellati P, Dunn D, et al.
: Estimating the rate of mother-to-child transmission of HIV. Report of a workgroup in methodological issues Ghent (Belgium), 17–20 February 1992
6. Oxtoby MJ: Perinatally acquired human immunodeficiency virus infection
. Pediatr Infect Dis J
7. St Louis ME, Kamenga M, Brown C, et al.
: Risk of perinatal HIV-1 transmission according to maternal immunologic, virologic and placental factors
8. Thomas PA, Weedon J, Krasinski K, et al.
: Maternal predictors of perinatal human immunodeficiency virus transmission
. Pediatr Infect Dis J
9. Landesman SH, Kalish LA, Burns DN, et al.
: Obstetrical factors and the transmission of human immunodeficiency virus type 1 from mother to child
. N Engl J Med
10. Semba RD, Miotti PG, Chiphangwi JD, et al.
: Maternal vitamin A deficiency and mother-to-child transmission of HIV-1
11. Jackson JB, Kataaha P, Hom DL, et al.
: β2 -microglobulin, HIV-1 p24 antibody and acid-dissociated HIV-1 p24 antigen levels: predictive markers for vertical transmission of HIV-1 pregnant Ugandan women
12. Boyer PJ, Dillom M, Navaie M, et al.
: Factors predictive of maternal-fetal transmission of HIV-1. Preliminary analysis of zidovudine given during pregnancy and/or delivery
13. Scarlatti G, Hodara V, Rossi P, et al.
: Transmission of human immunodeficiency virus type (HIV-1) from mother to child correlates with viral phenotype
14. Dickover RE, Garratty EM, Herman SA, et al.
: Identification of levels of maternal HIV-1 RNA associated with risk of perinatal transmission. Effect of maternal zidovudine treatment on viral load
15. Thea DT, Steketee R, Bornschlegel K, Pliner V, Brown T and the New York City Perinatal HIV Transmission Collaborative Study Group: The effect of maternal viral load on the risk of perinatal transmission of HIV-1
. XI International Conference on AIDS
. Vancouver, July 1996 [abstract 5492].
16. Rodriquez EM, Mofenson LM, Chang B-H, et al.
: Association of maternal drug use during pregnancy with maternal HIV culture positivity and perinatal HIV transmission
17. Matheson PB, Thomas PA, Pliner V et al.
: Heterosexual behavior during pregnancy and the perinatal transmission of human immunodeficiency virus type 1 (HIV-1)
18. Nduati RW, John GC, Richardson BA, et al.
: Human immunode-ficiency virus type 1-infected cells in breast milk: association with immunosuppression and vitamin A deficiency
. J Infect Dis
19. Rogers MF, Ou CY, Rayfield M, et al.
: Use of polymerase chain reaction for the early detection of the proviral sequence of human immunodeficiency virus in infants born to seropositive mothers
. N Engl J Med
20. Craft NE, Wise SA, Soares JH Jr: Optimization of an isocratic high performance liquid chromatographic separation of carotenoids
. J Chromat
21. Centers for Disease Control: Classification system for human immunodeficiency virus (HIV) infection in children under 13 years of age
22. Working Group on Antiretroviral Therapy: National pediatric HIV Resource Center. Antiretroviral and medical management of human immunodeficiency virus-infected child
. Pediatr Infect Dis J
23. Wallingford JC, Underwood BA: Vitamin A deficiency in pregnancy
. In Vitamin A Deficiency and Its Control
. Edited by Bauernfeind JC. Orlando: Academic Press; 1986:101–152.
24. Behrm RE, Vaughan VC Sr, Nelson WE (Eds): Nelson Textbook of Pediatrics
. Philadelphia: WB Saunders; 1987.
25. Bray GA: Definition, measurement and classification of the syndromes of obesity
. Int J Obesity
26. Hosmer DW, Lemeshow S (Eds): Applied Logistic Regression
. New York: John Wiley and Sons; 1989.
27. Dixit DT: Night-blindness in the third trimester of pregnancy
. Ind J Med Res
28. Birnbacher T: Ueder den Frühlingsgipfel der epidemischen Mangelhemeralopie und die pathogenetische Bedeutung des Frühjahrs
. Wiener Klin Wochenschr
29. Stephensen CB, Alvarez JO, Kohatsu J, Hardmeier R, Kennedy JI Jr, Grammon RB Jr: Vitamin A is excreted in the urine during acute infection
. Am J Clin Nutr
30. Duitsman PK, Cook LR, Tanumihardjo SA, Olson JA: Vitamin A inadequacy in socioeconomically disadvantaged pregnant Iowan women as assessed by the modified relative dose response (MRDR) test
. Nutr Res
31. Pilch SM: Analysis of vitamin A data from the health and nutrition examination surveys
. J Nutr
32. European Collaborative Study: Caesarean section and risk of vertical transmission of HIV-1 infection
33. Abrams EJ, Matheson PB, Thomas PA, et al.
: Neonatal predictors of infection status and early death among 392 infants at risk of HIV-1 infection monitored prospectively from birth
34. Halsey NA, Boulos R, Holt E, et al.
: Transmission of HIV-1 infections from mother to infants in Haiti. Impact on childhood mortality and malnutrition
35. Lepage P, Msellati P, Van de Perre P, Hitimana D-G, Dabis F: Characteristics of newborns and HIV-1 infection in Rwanda
36. Lepage P, Van de Perre P, Msellati P, et al.
: Mother-to child transmission of human immunodeficiency virus type 1 (HIV-1) and its determinants: a cohort study in Kigali, Rwanda
. Am J Epidemiol
37. Semba RD: Vitamin A, immunity, and infection
. Clin Infect Dis
38. Blackfan KD, Wolbach SB: Vitamin A deficiency in infants: a clinical and pathological study
. J Pediatr
39. Noback CR, Takahashi YI: Micromorphology of the placenta in the rats reared on marginal vitamin-A-deficient diet
. Acta Anat
40. Semba RD: An overview of the potential role of vitamin A in mother-to-child transmission of HIV-1
. Acta Pediatrica
1997 (in press).