NOTE: all abstract citations are from the XI International Conference on AIDS, held in Vancouver, BC, July 7-12, 1996.
One of the most prominent aspects of the discussion of women and HIV/AIDS at the XI International Conference on AIDS was mother-to-child HIV transmission, alternately referred to as vertical transmission or perinatal HIV transmission (PHT). Too often, discussions about PHT regard the HIV positive woman simply as a vector of viral transmission. Conference organizers, presenters and attendees alike agreed that the best way to prevent PHT is to prevent HIV infection in women. This perspective is of the utmost importance for women, children and families worldwide.
An estimated 40% of the approximately 30 million HIV-infected persons worldwide are women of childbearing age. Today, the leading mode of HIV acquisition for children, by far, is PHT. The need for an effective and accessible method of preventing PHT is clear and urgent, and preventing PHT is an important part of a global strategy to reduce HIV/AIDS. A critical footnote to this discussion pertains to resources and access: 9 of 10 new infections in women occur in developing countries, where access even to oral AZT (Retrovir) -- let alone to the technology required to carry out the AZT prophylaxis regimen currently considered the gold standard for preventing PHT -- is extremely limited.
The gold standard for reducing mother-to-child HIV transmission in developing nations is the same regimen first studied in the U.S. in AIDS Clinical Trials Group (ACTG) protocol 076 (for previous BETA reports on 076 and PHT, see the December 1995 issue, pp. 43-44, 47; June 1995, pp. 41-43; and September 1994, pp. 68-74). The regimen gives oral AZT to HIV positive women during pregnancy, starting at 14-34 weeks of gestation, as well as intravenous infusions of the drug during delivery, then continues to give AZT for 6 weeks after delivery to infants. In the study which led to approval of this regimen, perinatal transmission rates were reduced by two-thirds: PHT occurred in approximately 25% of untreated HIV positive women and in 8% of women who followed the regimen. In August 1994 U.S. public health authorities officially recommended that all pregnant HIV-infected women be offered the regimen. Similar national public health recommendations followed shortly thereafter in Europe and Australia.
At the Vancouver AIDS conference, the first prospective studies on transmission rates since the beginning of the "era of 076" were presented. Reports also were made on the latest innovations and research efforts to develop strategies to reduce vertical transmission in both developed and developing nations. Interesting data are also emerging from Africa, where researchers are studying regimens that may be more practicable in developing nations, in terms of resources as well as cultural acceptability.
In her plenary address to the Vancouver audience, Yvonne Bryson, MD, of the University of California at Los Angeles (UCLA) School of Medicine, summarized findings from vertical transmission studies to date. Bryson stated that she envisioned finding a way to reduce perinatal transmission rates to less than 2%, and stressed the necessity of making strategies for reducing PHT available to all women worldwide. She said that researchers must continue to pursue answers to key questions, including why, without intervention, 25% of infants will be infected by their mothers and 75% will not. (Rates of infection in the absence of intervention tend to be higher in developing countries -- according to some estimates, nearly twice as high -- but the question remains the same.) Another important question for study is why multiple children born to the same HIV positive mother are not all HIV-infected. Researchers also are evaluating the transient HIV infection with seroreversion (loss of antibodies to HIV) now documented to have occurred in 25 infants worldwide.
Prospective Studies Conducted in the Era of 076
Various researchers presented data on perinatal transmission rates in North America and Europe since the beginning of the era of 076.
By the time of the Vancouver conference, researchers who had been tracking women and their infants in various settings, collecting and analyzing data, were able to deliver the first reports on the regimen's real-world efficacy and impact since its (theoretical) widespread availability. While most studies presented in Vancouver concluded that use of perinatal AZT successfully reduces transmission rates, findings also elucidated limitations of the regimen, identified new risk factors for perinatal transmission and suggested new avenues for research. Ultimately, AZT monotherapy as a means for preventing perinatal transmission may be replaced by the use of other antiviral agents such as nevirapine (Viramune) or combination regimens such as AZT/3TC (Epivir), and may be supplemented with techniques such as pre-delivery cleansing of the birth canal.
Several presentations focused on prevention of perinatal transmission (Abstract #TuC440-445). Researchers from the Centers for Disease Control and Prevention (CDC) delivered a report entitled "Declining mother-to-child HIV transmission following perinatal zidovudine [AZT] recommendations, United States," based on a study that evaluated risk factors for PHT since August 1994. The women and children in this study were also participants in a prospective CDC study known as the Perinatal AIDS Collaborative Transmission Study (PACTS). Since 1986, PACTS has enrolled and followed 1,183 mother/child pairs. Of the children born during the study, 1,019 were born before September 1, 1994 (i.e., before official approval of the 076 regimen), and 164 were born afterward.
Researchers gathered data on all the mother/child pairs and compared results for 2 groups that were divided according to the date of the child's birth. Final results were thus stratifed by delivery date: before and during the era of 076 (i.e., deliveries before and after September 1994). Investigators evaluated prenatal and newborn AZT use, maternal CD4 cell count and clinical status, date of delivery, duration of ruptured placental membranes and infant's HIV infection status. The PHT rate for children born before September 1994 was found to be 21%, compared to 11% for those born after September 1994. Perinatal AZT use increased from 17% before September 1994 to 80% after that date -- a large difference between the 2 groups. Another noteworthy, although incompletely understood, difference in risk factors between the 2 groups was that the proportion of deliveries that involved rupture of membranes (ROM) longer than 4 hours decreased from 55% in the earlier group to 37% in the later group.
Investigators found that the risk for PHT was lower when AZT was used and higher when ROM preceded delivery by more than 4 hours. Risk for PHT was also higher for mothers with fewer than 500 CD4 cells/mm3 and for mothers with class C HIV disease (i.e., with AIDS-defining opportunistic infections). In conclusion, investigators stated that transmission risk "has declined among PACTS enrollees following publication of the ZDV [AZT] regimen," and also concede that "lower transmission risk may be due not only to increasing ZDV [AZT] use but also to decreasing prevalence of other risk factors, such as long duration of ROM" (TuC440).
Another PACTS substudy that evaluated risk factors for PHT and the impact of AZT use involved only children born after September 1994, "who all had the opportunity to benefit from AZT use." Data was collected through interviews and chart reviews, and PHT rates were compared between 3 groups of mothers and children (singletons and first-born twins only). Organized on the basis of type of AZT use, the groups consisted of full, partial and non-users of AZT. Full use meant that both mother and infant received AZT in the manner prescribed by 076. Partial use meant that 1 or 2 of the 3 AZT components (prenatal, intrapartum, postnatal) were utilized, but not all 3. No use meant that neither the mother nor the newborn received AZT.
Of 269 mother/child pairs evaluated, 101 (38%) were full AZT users. Of 49 partial users, 27 (10%) received AZT only during pregnancy and 22 (8%) received only infant AZT. Finally, 119 (44%) received no AZT. Transmission rates for those 4 groups, respectively, were similar: 13.9%, 14.8%, 9.1% and 14.3%. After adjusting for AZT use, significant risk factors for PHT were maternal CD4 count less than 200 cells/mm3 and ROM greater than 4 hours before birth. Another statistically significant risk factor was prematurity, defined as gestational age less than 37 weeks.
Since the differences in PHT between users and nonusers of AZT were not statistically significant, and transmission in nonusers was unexpectedly low, researchers evaluated PHT risk factors by multivariate analysis. Relative risk for PHT in relation to ROM and prematurity was the same for users and nonusers of AZT; however, AZT users were more likely than nonusers to have lower CD4 cell counts.
Overall, PHT increased as the number of risk factors increased. Those with all 3 risk factors examined in this study -- CD4 cell count less than 200 cells/mm3, ROM greater than 4 hours and prematurity -- had the greatest likelihood of PHT. AZT seemed to have little impact on PHT. Yet the early findings of similar PHT rates between AZT users and nonusers, upon analysis, "does not mean that AZT is bad; it just means that nonusers had other anti-risk [protective] factors," said Richard Steketee of the CDC. Although not statistically significant, PHT was greater for women with class C disease and for women who had used AZT for more than 18 weeks. Prolonged ROM and prematurity, in this as in other studies, were found to be risk factors for both AZT users and nonusers, and need further study.
These findings underscore the importance of prenatal care and caution against continued use of medically induced ROM, a practice that involves purposely rupturing the membranes in an effort to speed labor. This study provides further evidence to identify maternal characteristics and delivery conditions as important risk factors that still require more study. Also in need of further study is the best time to begin prophylactic AZT (or other antiretrovirals) and the best manner in which to use it (TuC441).
A team from Switzerland presented PHT findings from a study that employed prophylactic AZT and elective cesarean sections (ECS). In the summer of 1994, Swiss authorities recommended that all pregnant HIV-infected women follow the 076 protocol. Like the CDC team, the Swiss team also sought to determine PHT rates since the national recommendation to use AZT during pregnancy was instituted. Women in this study were recruited from an ongoing cohort that began in 1986. Some of the centers involved in the study routinely used AZT; others did not.
Of 443 children enrolled by December 1, 1995, infection status was determined for 368: 302 were HIV negative and 62 were HIV positive, for an overall PHT rate of 16.8%. Risk factors associated with positive PHT were analyzed.
Since approval of the AZT regimen, 45 pregnant HIV positive women were enrolled, approximately 75% of whom used AZT. Only 2 of 23 children whose mothers used AZT were HIV positive, for a PHT rate of 8.7%. The 2 failures were attributed to (1) an infant receiving an AZT-resistant strain from its mother, and (2) seroconversion that occurred during pregnancy in a mother who did not begin prophylaxis until shortly before delivery.
Another factor associated with PHT risk reduction was ECS. All ECS were performed before onset of labor and with intact membranes. Conversely, vaginal birth was deemed a risk factor for PHT.
Still other factors associated with increased risk for PHT were maternal HIV symptoms, one or more previous abortions and pregnancy that lasted longer than 40 weeks. In conclusion, ECS was a significant factor associated with reduced risk of PHT, and a history of abortion and prolonged pregnancy were identified as new, possible risk factors. This interesting study is far from conclusive. Before recommending ECS for all pregnant, HIV positive women, a randomized, controlled clinical trial is essential. ECS itself involves considerable risk, especially to the HIV positive mother, which must be assessed before any policy could be reasonably implemented (TuC442).
In a French study of obstetric factors involved in PHT, Laurent Mandelbrot and others, on behalf of the French Pediatric HIV Study Group, studied records of pregnancy outcomes and serostatus for 1,033 infants followed to 18 months of age. Participants were enrolled between September 1985 and December 1992, before the approval of the 076 regimen. Antiretroviral use as well as duration of ROM and other factors were evaluated. The team found that the 20% transmission rate in this group was not related to mode of delivery nor to any procedure during labor and delivery. They concluded that "cesarean section...did not decrease mother-to-child transmission rate." Factors that were associated with PHT were procedures during pregnancy including amniocentesis and amnioscopy, blood in the amniotic fluid and premature delivery (before 34 weeks). Therefore, they recommend that "severe preterm labor be prevented and invasive procedures during pregnancy avoided" (TuC2603).
Preventing PHT in Developing Countries
On behalf of an African/European collaborative study conducted in Abidjan, Cote d'Ivoire, Stefan Wiktor presented results of a small study that gave a shorter, simpler AZT regimen (relative to 076) late in pregnancy. This pilot study measured viral load in plasma and cervicovaginal secretions to determine if the short, simple AZT regimen would successfully reduce viral levels and, potentially, reduce PHT rates, before launching a larger, randomized clinical trial of the regimen. The 2-pronged rationale for this smaller study comes from knowing that 1) AZT reduces PHT by reducing plasma and cervicovaginal viral load, and 2) most PHT occurs late in pregnancy.
Women in the pilot study took the same regimen that will be used in the larger trial. Beginning in October 1995, consenting clients at a public prenatal clinic in Abidjan were given 250 mg AZT twice daily beginning at 36 weeks of gestation. At the onset of labor, the women received a single dose of 500 mg AZT orally, and then 250 mg orally every 2 hours until delivery. No postpartum AZT was given to either mothers or infants.
Results were available for 4 of 15 women who have completed the regimen so far. AZT was taken for a median of 28 days. Only one woman had a detectable cervicovaginal viral load (CVL) at enrollment; within 2 weeks, CVL became undetectable. Plasma viral load (PVL) was reduced by 50% in 3 women after taking AZT for 2 weeks. A maximum decline in viral load was observed at 4 weeks after initiation of therapy. A statistically significant viral load rebound was seen after delivery. In one woman, postnatal PVL was more than twice the enrollment level. To evaluate whether the observed changes in viral load are due to AZT, immunologic alterations in pregnancy or alterations in plasma volume, a control group would be needed. There may be a postnatal immunologic "catch-up" period for women. These women and their children will be followed for 2 years. This short, simple regimen cannot address all potential routes of transmission or exposure in this population. For example, many women in West Africa continue to breast-feed well into the baby's second year of life, so that transmission could still occur long after the perinatal AZT prophylaxis regimen has ended (TuC443).
Another study conducted with the same group of women in Cote d'Ivoire, also preparatory to the planned, larger study of short-course oral AZT given late in pregnancy, had promising results. This small study was designed to determine compliance, safety and tolerance.
Women and babies were seen at regular intervals for evaluation of clinical health and bloodwork. Results from 15 women indicated good compliance, based on pill counts (a measure of compliance of limited value), which showed that 94% of pills were taken during the prenatal period. Women had no adverse reactions to the regimen at any time during pregnancy or afterward; no adverse events occurred among infants, nor were there any congenital abnormalities. Since the short course AZT regimen was well-tolerated and since compliance was high, the large-scale AZT efficacy trial is also expected to be well-tolerated. Since some women did not remember to take the AZT pills during labor, when they were overwhelmed by pain, the researchers plan to educate midwives to be more helpful during the intrapartum phase of the regimen (ThC412).
In West Africa, preliminary results of a Phase II study evaluated another short AZT prophylaxis regimen. Neither the efficacy of an AZT prophylaxis regimen when combined with breast-feeding nor the tolerability of the regimen in this locale are known. Investigators recruited women at prenatal clinics in Abidjan, Cote d'Ivoire, and Bobo Dioulasso, Burkina Faso. Consenting women were tested for HIV and, if infected, randomized to receive AZT or placebo from 37 weeks of gestation. Depending on site, 50-90% of women agreed to be tested, and 45-80% returned to receive their test results. Among 2,800 pregnant women tested, HIV seroprevalence was 12.1%. The regimen used consisted of AZT 250 mg twice daily until delivery, 500 mg when labor began and 250 mg twice daily for the first week after delivery (mothers only). All women received iron/folate supplementation and malaria prophylaxis.
As of January 1996, 37 women had enrolled in the AZT trial. Compliance has been good, and there have been no significant side effects. The relatively modest number of women who consented to testing, returned for test results and enrolled, suggests that interventions that require HIV testing may be relatively hard to implement in this region. Still, the simplified AZT regimen studied may be safe and acceptable to African women, and may help reduce PHT. Since conditions vary by setting, and since only very basic, simple technology may be available, alternative interventions for reducing PHT should be considered for use in Africa. The next step is to perform a Phase III efficacy trial as a further test of this regimen (TuC444).
Finally, Mark Johnson of Glaxo Wellcome presented pharmacokinetic data from a study of prophylactic combination 3TC/AZT in South African pregnant women and their infants. Data from studies of combination 3TC/AZT in HIV-infected adults showing that the combined effect of these 2 oral drugs is more powerful than either used alone provided the inspiration for this small, preliminary, dose-finding study. Pregnant women began taking the combination at week 38 of their pregnancy. Results of this study suggest that the combination regimen is as safe and as well tolerated by pregnant women as by other adults. However, the pediatric dose of 3TC should be reduced by one-half for infants (to 4 mg/kg/day), because its clearance in infants is slower than in older children. (The lower clearance rate in infants is more similar to the rate in adults than to the rate in older children.) (Tu.C.445).
A panel of African researchers presented results of studies that represent attempts to find effective and feasible interventions for reducing perinatal transmission in regions where resources and access to healthcare are limited.
A study conducted in Blantyre, Malawi, evaluated the effect of cleansing both the birth canal and the newborn on PHT and other postpartum outcomes. Malawi, a country in southeast Africa, has a population of 9 million people and an HIV seroprevalence rate in pregnant women of approximately 30%.
The agent used for cleansing was 0.25% chlorhexidine (CL), a broad-spectrum, water-based antiseptic that is widely used in hospitals. The regimen is easy for nurses and midwives to administer and inexpensive; the cost of the intervention per woman is $0.10 U.S. The simple procedure involves manually wiping the birth canal with CL, from the cervix outward through the vagina and over the external genitalia, prior to delivery. Immediately after delivery, the newborn also is cleansed by wiping with CL. In other studies, CL has been shown to reduce group B streptococcus, a significant cause of infant morbidity (sickness). In addition, it may reduce viral, bacterial and chlamydial infections.
More than 3,000 HIV positive and HIV negative women were enrolled from June to November 1994. They received either the CL cleansing intervention or no intervention. The PHT rate for women whose birth canals were cleansed prior to delivery and whose infants were cleansed immediately after birth was 26.9%, compared to 27.9% for those not receiving the intervention. The overall difference in terms of PHT between cleansing and not cleansing was very small, and not statistically significant. However, among women whose membranes were ruptured for more than 4 hours before delivery, there was a statistically significant relative reduction in PHT, which was 24.4% in the intervention arm and 38.8% in the no-intervention arm.
Several other favorable outcomes were associated with cleansing, including reduced infant morbidity and mortality, particularly related to bacterial sepsis (blood infection resulting in life-threatening low blood pressure). Babies who received the intervention were also hospitalized less often. The intervention benefitted mothers, too. Women who received the intervention were hospitalized less often and for shorter periods of time than mothers who did not receive the intervention. Intervention women also had significantly fewer infections due to sepsis, as well as fewer overall postpartum infections. No adverse reactions related to CL cleansing occurred among either mothers or babies.
In conclusion, although the intervention had no significant impact on PHT -- except when membranes were ruptured for more than 4 hours -- it did significantly reduce maternal and infant morbidity and mortality, especially related to sepsis. According to the researchers, "the safety, simplicity and inexpensiveness of this intervention encourage its adoption as a public health measure." They also asked whether the policy of rupturing membranes should be changed and whether women admitted with ruptured membranes should immediately receive the CL intervention. Currently, physicians in Malawi continue to debate these questions (ThC410).
A study conducted in Burkina Faso gauged the acceptability and feasibility of routinely offering HIV testing and counseling (HIV T&C;) to pregnant women. This study was nested within a larger study of short-course AZT prophylaxis during pregnancy. It was conducted because "Africa is presently the continent where these interventions are most urgently needed, but it is currently unknown whether African pregnant women are ready to even accept HIV T&C;." To assess this, researchers recruited women attending 2 prenatal clinics in Bobo Dioulasso between January and June 1995. They offered HIV T&C; to the women and documented their responses. Of 916 pregnant women, 90% accepted testing, 73% returned to the clinic to receive their test results -- 53% spontaneously and 20% only after a home visit by a social worker. Twelve percent were HIV positive. Women who were tested reported appreciation for both the support of social workers and the fact that testing was free of charge. Among the 10% who refused testing, the majority (57%) stated that they did so because they wanted their husband's approval or advice first. A further 34% cited fear of AIDS, 5% stated that they did not perceive themselves to be at risk and 4% said they wanted to consider it further.
The main conclusions are that both acceptance and benefits of HIV T&C; appear to be high in pregnant women in Burkina Faso. In addition, women's partners were found to play a key role in the decisions women made about HIV testing. Finally, without free testing and social support, "HIV T&C; may remain a limit to implementing some interventions to reduce mother-to-child transmission of HIV in developing countries." This, researchers add, "should stimulate research on interventions which do not require HIV T&C;, such as vaginal disinfection" (ThC411).
Glenda Gray, MD, from Soweto, South Africa, presented results of a study of formula-feeding vs breast-feeding, which examined key issues in breast-milk transmission of HIV. Factors known to influence breast-milk transmission include maternal factors (infectivity), infant factors (susceptibility) and viral factors (virulence of the strain, concentration in milk). Still other maternal factors include stage of HIV disease, nutritional status (vitamin A), antiretroviral use and physical factors, primarily the shape of the breast or nipples. Previously, vitamin A depletion in women with fewer than 400 CD4 cells/mm3 has been associated with a 20-fold increased risk of viral shedding in breast-milk.
In all, breast-feeding plays a clear role in PHT; an estimated one-third to one-half of HIV positive children worldwide reportedly may have been infected by this route. Experimental intervention strategies to reduce or prevent breast-milk HIV transmission include exclusive formula-feeding; withholding or pasteurizing the colostrum; early weaning, at 3 or 6 months; antiretroviral treatment (although probably prohibitively expensive in most developing countries); and vitamin A supplementation for mother, child or both.
Eighty-five percent of the world's 3 million PHT-infected children live in sub-Saharan Africa. Here, where breast-feeding is nearly universal and often continues into the child's second year of life, the new PHT reduction strategies such as perinatal AZT, even if available, would be greatly reduced in efficacy. The objective of this study was to determine the influence of breast-feeding on PHT and to compare early and late postnatal transmission in such a setting. Researchers gleaned valuable information about the timing of transmission and the comparative risks of formula-feeding and breast-feeding. These data are crucial for better informed public health policies on infant weaning and breast-feeding, especially in the developing world, where recommendations continue to advocate breast-feeding.
Investigators asked: (1) what is the rate of breast-milk transmission? (2) when does breast-milk transmission occur? (3) what factors influence breast-milk transmission? The study was prospective and observational. Pregnant HIV positive women recruited from a prenatal clinic chose whether they wanted to formula-feed or breast-feed, forming the study s 2 arms. There were no significant differences between participants in these 2 arms in terms of sociodemographics, mean CD4 cell count or mode of delivery. Mothers and children were followed for 18 months, with visits every 3 months. Follow-up visits included full clinical assessments, PCR viral load testing and CD4 monitoring.
The formula-feeding mothers gave birth to 49 (29%) infants, and breast-feeding mothers bore 114 (71%) infants. The PHT rate for formula-fed babies was 18%: 9 of the 49 babies were HIV positive. The PHT rate for breast-fed babies was 46%Ñ53 of the 114 babies acquired HIV from their mothers. The mean duration of breast-feeding was 3 months, with a range of 1 week to 18 months. Among breast-fed infants, 15% (8 of 53) of transmissions occurred after 6 months of age.
In conclusion, there were significant reductions in PHT among women who chose to formula-feed exclusively. Infants who were exclusively formula-fed were just as healthy as breast-fed infants. Formula-feeding appears to be safe in a poor but "urbanizing" community like Soweto. Early weaning, before 6 months of age, may be another way to significantly reduce risk of PHT (ThC415).
In the U.S., investigators with the Pediatric Spectrum of Disease Project (PSD) studied approximately 1,200 medical records of children born between 1988 and 1993 to HIV positive mothers in order to discern patterns of breast-feeding. About 80 babies were breast-fed at some point by their mothers. Information about prenatal care was available for 48 of 80 mothers who chose to breast-feed: 65% had prenatal care but learned of their HIV positive serostatus only after delivery, 21% had prenatal care and knew of their status before delivery, 8% had no prenatal care and 6% had prenatal care but an unclear HIV diagnosis date.
Other data contained in the PSD database suggest that prenatal care was not associated with refraining from breast-feeding. In U.S. women, what did seem associated with breast-feeding was when women learned their HIV test results. HIV positive women were much more likely to breast-feed if they were unaware of their serostatus until after the baby was born.
Thus, despite PHS recommendations to HIV positive mothers in the U.S. to refrain from breast-feeding, some women still do. This observation serves as yet another reason to offer prenatal care, including HIV counseling and testing, to all pregnant women. Prenatal care for HIV positive women should include education about the risks of breast-feeding (WeC 3583).
Strategies to reduce or prevent perinatal transmission interest politicians as well as public health officials and HIV-affected persons. While a political discussion would constitute a different article, the policy and legislative debates related to PHT warrant vigilant public attention.
In early July, as the Vancouver conference began, the American Medical Association narrowly voted in favor of mandatory HIV testing of pregnant women, against the recommendation of the CDC, which advocate mandatory counselling. There are many ramifications associated with this and other decisions.
The rights of pregnant women must not be forgotten in this discussion.
Vitamin A and Perinatal HIV Transmission in the U.S.
A few years ago, Richard Semba, MD, published the startling but promising results of a study of vitamin A in HIV positive women in Africa, which showed that vitamin A deficiency was associated with perinatal HIV transmission (PHT). Now, Semba and colleagues have released results of a similar study conducted in a cohort of women in the U.S.
In this cohort of 133 women, 44 transmitted HIV to their infants born between May 1986 and May 1994, i.e., before the approval and widespread use of the 076 AZT regimen; 89 women did not transmit HIV.
Perinatal transmission was associated with severe vitamin A deficiency, rupture of membranes (ROM) lasting longer than 4 hours and gestational age less than 37 weeks. After controlling for CD4 cell count and for duration of ROM, vitamin A deficiency and prematurity continued to be independent risk factors. CD4 count, type of delivery, age, body mass index and race were evaluated, but found not to be associated with PHT.
Among transmitting mothers, 16% had severe vitamin A deficiency (defined as less than 0.7 micromoles/liter) and 23% had marginal vitamin A deficiency (0.7-1.04 micromoles/liter). By comparison, 5% of nontransmitting mothers had severe and 24% had marginal vitamin A deficiency.
Women in this study, as in the African study, were significantly more likely to transmit HIV to their babies as vitamin A deficiency worsened. Clinical trials are underway to determine if giving supplemental vitamin A will help reduce PHT (TuC 2592). See also Perinatal HIV Transmission Prevention Strategies: an Economic Perspective
Leslie Hanna is Associate Editor of BETA.