This study was a randomized, community-based trial testing whether dosing newborn infants in the first few days after delivery with a large dose of vitamin A could reduce early infant morbidity, mortality and improve growth. Pregnant women in 2 districts of Tamil Nadu State in southern India were identified and recruited for participation. When the child was born, study staff were notified and traveled to the house to collect information and provide the vitamin A or placebo dose. Children were visited every two weeks until they were 6 months of age to collect information about vital status and morbidity. All children were discharged from the study at 6 months after growth was assessed and they received a 100,000 IU dose of vitamin A.

I. Background

A variety of studies over the past 15 years have demonstrated that clinical vitamin A deficiency is associated with increased mortality in pre-school aged children (1,2) and that supplementation of populations with endemic sub-clinical vitamin A deficiency can significantly reduce total mortality (3-8). A recent meta-analysis of 8 clinical trials showed that vitamin A supplementation of children 6 months through 5 years of age reduced mortality by an average of 23% (9). While the average impact was 23%, there was significant variation in the results of these studies. The impact was as high as 54% among children in an underprivileged area of Tamil Nadu supplemented with small weekly doses of vitamin A (4).

While the story of the impact of vitamin A supplementation in children 6 months and older is clear, such is not the case for younger infants. Traditionally, it has been thought that the full-time breast feeding practices associated with early infancy protected the infant from becoming vitamin A deficient. It was not until the onset of weaning or the misuse of bottle-feeding that clinical signs of xerophthalmia appeared (1). Recent evidence from a variety of locations suggests that these traditional assumptions about the adequacy of breast -feeding in assuring adequate vitamin A nutriture are suspect and likely unfounded. It is well known that infants are born with very low body stores of vitamin A and they depend on regular breast feeding or other sources to build body stores (13, 14). Therefore, if the mother's breast milk has a lower than normal concentration of vitamin A, the infant will not be able to meet daily requirements and increase body reserves. It has become clear that many women in developing countries have poor vitamin A status themselves and that their breast milk concentrations may be only half that found among women with adequate vitamin A status (15). In addition, it has been shown in Indonesia that supplementation of women post-partum can improve the vitamin A concentration in their blood and breast milk (16).

While it is becoming clearer that young infants in developing countries with endemic vitamin A deficiency are at significant risk of being vitamin A deficient in spite of adequate breast feeding behaviour, there is little data concerning the impact of supplementing this group with vitamin A. Two major studies have addressed this issue, both done by the same group at Johns Hopkins University, with opposite findings. Humphrey et al. have reported, in a randomized clinical trial of 2067 infants born in a hospital in Bandung, Indonesia, that direct supplementation of newborn infants within 24 hours of birth with 50,000 IU of vitamin A was safe (10) and associated with a 64% reduction in infant mortality (11). In contrast, West et al. studied over 11,000 infants under six months of age in their large intervention trial in southern Nepal (17). They found no difference in mortality between those supplemented with vitamin A versus those given placebo (relative risk =1.11, 95% CI 0.86 - 1.42). However, there were significant and potentially important differences between these two studies. First, the time of dosing was different. The Indonesian study dosed infants within 24 hours of birth whereas the Nepal study dosed infants when they were first identified to the study during their distribution rounds every four months. On average then, the age at first dosing was 2 months after the highest risk period and exposure to dietary and other environmental factors. It may be that the time of dosing is critically important for the development of early immune function or other protective mechanisms. Support for the Indonesian findings has come from a recently discussed study in Bangladesh where supplementation of mothers' post-partum was associated with a reduction in infant mortality (12).

Given the apparent controversy in the scientific literature on early infant dosing with vitamin A and the fact that this remains the highest risk period for mortality during childhood, more data are urgently needed to resolve this confusion. If the Indonesian findings can be replicated, the policy implications for developing countries are clear. An expansion of the "safe delivery" concept to include supplementation of the infant and mother with vitamin A would enhance nutrition and safe motherhood program impact in an extremely cost-efficient fashion. As practices in India move increasingly in the direction of attended childbirth, the opportunity to intervene becomes easier with significant potential for having a dramatic impact on infant mortality.

Goals and Specific Aims

The overall goal of this project is to improve the health and survival of young children in India through improvement in vitamin A status. More specifically this project will:

1. Evaluate the efficacy of vitamin A supplementation of newborn infants (50,000 IU) on mortality within the first 6 months of life using a community-based distribution and outcome assessment strategy.
2. Evaluate the efficacy of vitamin A supplementation of newborn infants (50,000) on the incidence and severity of morbidity within the first 6 months of life. Specific morbidities of interest include diarrhea and acute respiratory infections.
3. Evaluate the efficacy of vitamin A supplementation of newborn infants (50,000 IU) on growth of infants over the first 6 months of life.

These specific aims will be accomplished by conducting a community-based, randomized trial of vitamin A supplementation to newborn infants within 48 hours of birth in underprivileged rural communities of Tamil Nadu.

II. Study Protocol

Overview: The overall design of this study will be that of a randomized, community-based clinical trial. The intervention to be tested for impact on infant mortality, morbidity and growth is supplementation of 50,000 IU of vitamin A to the newborn within 48 hours after delivery. Outcomes of interest are mortality within the first 6 months of life, the incidence and severity of diarrhea and acute respiratory infections, and growth.

This study will be population-based such that all live births within a defined population will be enrolled. Women residing in study villages will be identified by village-based field staff when they become pregnant. Informed consent will be obtained at that time as well as baseline data regarding family and household characteristics. In addition, field workers for the study will ask the woman where she plans to deliver the child, at home, in a maternity home, or in the hospital. The woman will be assigned a study number and randomized at the time of enrollment to either the vitamin A or placebo groups. Actual supplementation will occur at the time of delivery.

Study staff will attend all deliveries within 24 hours of the birth. At that time, the child will be dosed with either vitamin A or placebo according to their treatment assignment. Birth weight will be measured, and details of the delivery will be collected. Following delivery, a short vital status and morbidity assessment will be conducted every two weeks for the first 6 months of life. If a child death is reported, supervisors will be dispatched to the household to verify the death and to conduct a verbal autopsy. All children will receive 200,000 IU oral vitamin A at the time of the 6 month visit and discharged from the study (200,000 IU is the dose currently provided by the government of Tamil Nadu to children 6 months of age and older).

The analysis will focus on estimation of mortality rates for the vitamin A and placebo groups. These will be compared using a person-time approach as well as survival analysis techniques. The two groups will be compared on baseline characteristics and any imbalance will be controlled using standard adjustment techniques including logistic regression and proportional hazards models. Results will be reported in the scientific literature, at national and international meetings.

Study Population:

The study population chosen for this project will include all villages in two rural blocks in Tamil Nadu. The two blocks chosen for inclusion in the study are Kariaypatti and Natham. ASSEFA (Association for Sarva Seva Farm) is a non-governmental organisation involved in community development, education, and health programs for underprivileged rural communities. Within the ASSEFA area of a block, the health activities consist of a maternal and child health program, which runs antenatal clinics and produces and delivers weaning foods to young children and pregnant women. The engine for this program is the Health Committee in each village. In most ASSEFA program areas, registration of pregnant women by ASSEFA staff or volunteers is virtually complete. Non-ASSEFA villages are slightly better off and served primarily through a variety of government programmes.

Eligibility: In order to maximise the generalisability of the results of this study, the criteria for eligibility will be broad and inclusive. All pregnancies that occur within the geographic boundaries of the participating villages, which result in a live born infant(s), will be eligible for recruitment. This will include multiple births, those with congenital anomalies, and those born early or with low birth weight. Provision of informed consent will be the only eligibility criterion in addition to geographic eligibility.

Recruitment and enrollment: Pregnant women will be identified from a variety of sources, the ASSEFA Health Committee Members, the antenatal care clinics run on a monthly basis by ASSEFA or the government in the study areas, and other women's development workers in the area. Once a pregnant woman is identified, the cluster worker will explain the study and attempt to enroll the subject. If she agrees, then the cluster worker will proceed with the assignment of a study ID number and completion of baseline data collection form, including a determination of where the women plans to deliver her child.

For those women who are registered, but plan to deliver outside the village, special actions will be taken. There are three types of these women:

i) Those who go outside the village to deliver their child and are not reachable by our study teams (more than 20 kms away from the village). These women will be excluded from participation in the trial, but these babies will be followed after they return to the village, only for 6-month mortality.

ii) Those who go outside the village to deliver in another study village. These women will be transferred to the appropriate Cluster Worker so the treatment doses can be given as per protocol and the infants will be following in this location until they return to their home village. At this time, the file will be transferred back to the appropriate Cluster Worker in the home village.

iii) Those who go outside the village, not to study villages, but who are reachable by special efforts of our study supervisory staff. Reachable will be defined as being within 20 km of the Block Office. These women will be treated and infants dosed by the supervisor of the cluster worker. They will not be followed for morbidity, but will be followed for mortality based on home village family reports until they return to the village. After returning to the home village, they will continue to be followed for mortality and growth only (not morbidity).

A fourth special situation are those women who move into one of our study villages temporarily to deliver from outside the study area. These women will be included and enrolled as soon as they arrive in our study villages and they will be followed for morbidity and mortality after birth until they return home. After they leave the village, mortality only will be followed by reports from the family in the study village.

Randomisation: Randomisation will be conducted at the individual level, stratified by geographic area and randomized in blocks of size four to insure equal numbers of babies in each treatment group. Each pregnant woman will be assigned a unique study number at the time of enrollment. Each set of capsules (two 25,000 IU capsules for the newborn) will be packaged separately and labelled with the woman's name, address, and ID number to avoid mix-ups at the time of delivery. Treatment assignment codes will be kept in a locked filing cabinet in a sealed envelope at a location separate from the study headquarters.

Definition and Measurement of Outcomes: There are one primary and two secondary outcomes that will be evaluated in this study. The primary outcome is vital status within the first 6 months of life. Deaths will be ascertained during the vital status and morbidity assessments conducted every two weeks during a home visit. Any child found to have died within the preceding two weeks will initiate the process of death verification and cause of death determination. Cause of death determination will identify both immediate and underlying causes by two independent reviews of the verbal autopsy information by paediatricians. If there is disagreement regarding the cause of death, a third paediatrician will review the data and conduct a consensus meeting to arrive at an agreed upon cause of death.

Secondary outcomes include the incidence and severity of diarrhea and acute respiratory infections over the first six months of life, clinic utilisation for illnesses during the first six months, and growth over the first six months. Episodes of diarrhea, dysentery, acute respiratory illness (ARI), and clinic utilisation will be based on a maternal interview conducted every two weeks. Diarrhea will be defined as 6 or more loose, watery stools in a 24 hour period. Dysentery will be defined as diarrhea with the additional history of blood or mucus in the stool. The incidence of episodes as well as their duration will be assessed during the data collection process. Growth will be assessed only at the six-month visit where weight, length, and mid-upper arm circumference will be assessed using standard protocols.

Sample Size Requirements: Previous research on the impact of vitamin A supplementation among young infants in Tamil Nadu and Indonesia suggests that we may expect a mortality reduction of between 30% and 50%. Because even the lower end of this range of impact would be of significant importance, sample size calculations have been done using a minimal detectable reduction of 30%.

The infant mortality rate (IMR) in Tamil Nadu as a whole is approximately 60/1000 live births. The population selected for inclusion in this study is poorer economically and has worse health status than the average for Tamil Nadu. It is likely, therefore, that the infant mortality rate in this area is higher than the state as a whole. We have estimated it at 70/1000 live births. The study design for this project calls for vital status assessment to conclude at 6 months of age. Therefore the expected mortality rate in the first six months will be lower than the total IMR. Mortality rates decline dramatically with age, even in the first year of life. Reasonable estimates are that 75% of total infant mortality occurs in the first six months of life. Applying this figure to the estimated IMR of 70/1000 live births gives a 6-month infant mortality rate (IMR-6) of 52.5/1000 live births.

Assuming a type I error of 5% (2-sided), 80% power, an IMR-6 of 52.5/1000 in the placebo group, and a minimum detectable reduction of 30% in the vitamin A group, the sample size required in each group is approximately 2700 live births. If we assume a 10% loss to follow-up this brings the number required in each group to approximately 3000, or a total of 6000 live births.

The crude birth rate in Tamil Nadu is approximately 20/1000 population. In order to enroll 6000 live births in one year, we must ascertain all births in a total population of approximately 300,000 or a population of 150,000 over a two-year enrollment period. We have chosen to enroll deliveries for a period of 24 months.

References

1. Sommer A. Nutritional Blindness. New York, Oxford University Press, 1983.
2. Sommer A, Tarwotjo I, Hussaini G, et al. Increased mortality in children with mild vitamin A deficiency. Lancet 1983; 2:585-588.
3. Sommer A, Tarwotjo I, Djunaedi E, et al. Impact of vitamin A supplementation on childhood mortality. A randomized controlled community trial. Lancet 1986; 1:1169-1173.
4. Rahmathullah L, Underwood BA, Thulasiraj RD, et al. Reduced mortality among children in southern India receiving a small weekly dose of vitamin A. N Eng J Med 1990; 33:929-935.
5. Daulaire NMP, Starbuck ES, Houston RM et al. Childhood mortality after a high dose of vitamin A in a high risk population. Brit Med J 1992; 304:207-210.
6. West KP Jr, Pokhrel RP, Katz J, et al. Efficacy of vitamin A in reducing pre-school child mortality in Nepal. Lancet 1991; 338:67-71.
7. Muhilal, Permeisih D, Idjradinata YR, et al. Vitamin A fortified monosodium glutamate and health, growth, and survival of children: a controlled field trial. Am J Clin Nutr 1988; 48:1271-1276.
8. Ghana VAST Study Team. Vitamin A supplementation in northern Ghana: effects on clinic attendances, hospital admissions, and child mortality. Lancet 1993;342:7-12.
9. Beaton GH, Martorell R, Aronson KJ, et al. Effectiveness of vitamin A supplementation in the control of young child morbidity and mortality in developing countries. ACC/SCN State of the Art Series, Nutrition Policy Discussion Paper No. 13, December 1993, Geneva.
10. Agoestina T, Humphrey JH, Taylor GA, et al. Safety of one 52 umol (50,000 IU) oral dose of vitamin A administered to neonates. Bull WHO 1994;72(6):859-868.
11. Humphrey JH, Agoestina T, Wu L, et al. Impact of neonatal vitamin A supplementation on infant morbidity and mortality. J Pediatrics 1996;128:489-496.
12. de Francisco A, Yasui Y, Chakraborty J. Vitamin A supplementation given to mothers after delivery reduces infant mortality and increases symptoms of morbidity. Presented at the XVI IVACG Meeting, October 24-28, 1994, Chiang Rai, Thailand.
13. Olson JA. Liver vitamin A reserves of neonates, pre-school children and adults dying of various causes in Salvador, Brazil. Arch Latinoam Nutr 1979; 29:521-545.
14. Olson JA, Gunning DB, Tilton RA. Liver concentrations of vitamin A and carotenoids as a function of age and other parameters of American children who died of various causes. Am J Clin Nutr 1984; 39:903-910.
15. Wallingford JC, Underwood BA. Vitamin A deficiency in pregnancy, lactation and the nursing child. In Bauernfeind JC (ed). Vitamin A Deficiency and Its Control. Orlando, Florida, Academic Press, 1986: 101-152.
16. 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; 123:666-675.
17. West KP Jr, Katz J, Shrestha SR, et al. Mortality of infants under six months of age supplemented with vitamin A: a randomized, double-masked trial in Nepal. Am J Clin Nutr 1995; 62:143-148.

Inclusion Criteria:

• All live born children

Exclusion Criteria:

• Children who die prior to study staff visiting the home to enroll the child and deliver the assigned intervention.