We conducted this cross sectional study in Dar es Salaam, where HIV prevalence was 8.9% in 2009. It accounts for about 300,000 people living with HIV/AIDS. The city has 44 HIV/AIDS care and treatment centers (CTCs) which provide ART programs to care about 4,000 children .
Participants and selection criteria
Participants of this study included pairs of 219 under-five ART-treated, HIV-positive children and their caregivers. These children were attending CTCs that provide RUTF treatment to severe undernourished children. In Dar es Salaam, a total of six out of 44 CTCs provide RUTF to such children. We excluded the CTC of the Muhimbili National Hospital (MNH). MNH is the referral center for complicated cases from the Northern zone and beyond. Being the highest tertiary hospital, its CTC also cares for more complicated and severe cases from the outside of the region. As a result, we selected five CTCs: Amana, Temeke, Mwananyamala, IDC, and Pasada CTCs. Children under five years of age in the selected CTCs totaled 1719. Amana CTC registered 350 children, Temeke CTC 396, IDC 250, Pasada 373, and Mwananyamala 350.
Of the total, about 58% of the children enrolled in CTCs were on ART. We excluded children with missing medical and ART-related information and those whose parents did not consent to participate in the study. We selected each alternative participant of this study from the list of the eligible participants (n = 670). We used Epi-Info Version 6 (CDC, Atlanta, USA) to calculate the minimum required sample size. We estimated that the ratio of children exposed to RUTF to those who were not exposed to RUTF in the intervention clinics to be 2:1. The proportion of underweight among ART-treated HIV-positive children was considered at 13.6% . We therefore estimated the minimum sample size at the power of 80 and with 95% CI to be 173 (115 RUTF treated and 53 RUTF naïve HIV-positive children under ART). In this study, we collected the data of 219 ART treated HIV-positive children (140 RUTF treated and 79 RUTF naïve HIV-positive children under ART).
We defined ART-treated HIV-positive children as children whose sero-status was confirmed to be positive using standard laboratory methods and were taking Highly Active Antiretroviral Therapy (HAART). According to the national guidelines for pediatric HIV/AIDS care in Tanzania, children are typically treated with a combination therapy consisting of three antiretroviral medicines, hence HAART .
RUTF Plumpynut® is given to children with severe wasting (Weight-for-height Z-score (WHZ) < −3SD) or underweight (Weight-for-age Z-score (WAZ) < −3SD) or both severe wasting and underweight . Such children are treated with the recommended dosage of 200 kilocalories per kilogram per day until they reach the target weight, which should be in 6 to 10 weeks . RUTF is a community-based intervention given to undernourished children who have no other clinical complications [3, 9, 19]. In this study, 140 HIV ART-treated, HIV-positive children who fulfilled the criteria received RUTF.
We measured children’s weight by two methods. For children who could not stand alone, we used a standardized hanging Salter scale® (UK) calibrated to 0.1 kg. For children who could stand alone we used a standardized Seka® digital scale (Brooklyn, USA). We measured height for the 24 months and older children  using a Seka® measuring rod calibrated to 0.5 cm. We used the same measure on a board for the children who were less than 24 months, and took their lengths in a recumbent position . We converted height and weight into height-for-age z-score (HAZ), weight-for-age z-score (WAZ), and weight-for-height z-score (WHZ)  by the Epi-Info ENA Ver. 3.5.1, 2008 (CDC, Atlanta, Georgia, USA) software, using WHO reference values [22, 23].
Low HAZ, WAZ, and WHZ are the measures of stunting, underweight, and wasting, respectively . Stunting reflects a chronic failure to receive adequate quality and quantity of nutrition over a long period of time; it may also signify a chronic recurrent illness . Stunting therefore may be reflected by failure to attain normal height for the age of a normal child [24, 26]. Wasting in most cases signifies an acute or recent and severe process of weight loss, which is associated with acute starvation or severe disease with nutritional deficit . Underweight is also a measure of acute undernutrition, but is not as robust as wasting in this regard, since the child may be underweight because of stunting, wasting or both .
According to the WHO Global Database on Child Growth and Malnutrition, an abnormal anthropometry is defined as a value of Z-score below −2 Standard Deviation (SD) or above +2SD. Z-scores less than -2SD are defined as moderate undernutrition, while less than -3SD are defined as severe undernutrition . These cut-off points define the central 95% of the reference distribution as the normality range .
We adopted the socio-demographic variables pertaining to children and their caregivers from the Tanzania Demographic and Health Survey (TDHS), women and household questionnaires . Information collected included education level, religion and marital status. We defined a caregiver as child’s caretaker, parent, or a guardian that is involved in child’s routine care and who accompanied the child to the clinic. Education was classified as either: low (up to primary school level) or high (higher than secondary school level). Religions were categorized into two major groups in Tanzania i.e. Christians and Muslims. Marital status was categorized into either currently married or not , caregivers who were divorced or widowed at the time of data collection were considered as not currently married.
We assessed food security by using the 6-item Household Food Security Scale (HFSS) . HFSS has been used in various settings including the US , Bolivia, Burkina Faso, Philippines , and the Caribbean . This scale is used to measure household food security by 12-month recall. Characterization is made based on the sum of affirmative responses; two or more affirmatives indicates ‘food insecurity’, while 5 or more affirmatives indicate ‘hunger’. In this study, Cronbach’s alpha for the HFSS was 0.72, with corrected item-total correlation ranging from 0.09 to 0.74.
We assessed economic status by using a Weighted Wealth Index incorporating household durable assets ownership such as paraffin lamp, television, radio, telephone, flat iron, refrigerator, bicycle, motor car, farm and electricity; housing and dwelling characteristics, such as main floor materials, house ownership, fuel for lighting and cooking, type of toilet, source of water, feeding characteristics, and household food satisfaction . Then, we constructed dichotomous variables and conducted principle component analysis (PCA) to reduce 42 items to 22 (loaded as factor 1). We also used factor loadings as item weights. We totaled item weights to yield the wealth index for each household [33–35].
We hired five nurse counselors and trained them for one day on interviewing technique and questionnaire contents. The first author and trained research assistants pretested the questionnaire. To ensure good interviewing environment, we recruited all research assistants from the clinics and assigned them to conduct data collection. We collected data by face-to-face interview with caregivers and anthropometric measurements of the children attending the clinic between September and October 2010.
We conducted descriptive analysis using Chi-square test and Independent sample T-tests to compare characteristics and nutrition status of children who were treated versus those who were not treated with RUTF. We conducted multivariate analysis to examine the effectiveness of RUTF on nutritional status of HIV positive children after adjusting other covariates and important confounders including ART treatment duration. Among RUTF treated children, we used bivariate analysis to evaluate the association of undernutrition and RUTF treatment duration; since the proportion of wasting and underweight was small for multivariate regression analysis. We set the statistical significance at P-value <0.05. We conducted analyses using PASW 18 (SPSS Inc., Chicago, Illinois, USA).
We obtained ethical clearance from the Ethical Committees of the University of Tokyo and Muhimbili University of Health and Allied Sciences to conduct this study. We also obtained permission to conduct research from health departments and facilities used for the study. Participation was voluntary, and confidentiality was ensured. Caregivers gave the written informed consent before starting the interview.