Under-nutrition and Associated Factors Among Children on ART in Southern Ethiopia: a Multicenter Facility Based Cross-sectional Study


 Background: Malnutrition is very common in HIV infected individual due to decreased food intake, altering digestion, absorption and altering metabolism and by increasing energy need. Even though data from different settings are necessary to tackle it, evidences are limited especially in case of nutritional status of HIV-infected children. Hence, this study aims to assess the nutritional status and associated factors among children on antiretroviral therapy. Methods: An institutional-based cross-sectional study was conducted among 383 HIV-positive children in Southern Ethiopia. Data were collected using interviewer administered questionnaire and anthropometry measurement. Data were coded and entered into Epi-Data Version 3.1, and analyzed using SPSS Version 25. Bivariable and multivariable binary logistic regression models were used to identify factors associated with nutritional status and variables with p-values ˂0.05 in multivariable logistic regression were considered as statistically significant factors.Results: The prevalence of wasting among HIV positive children in Southern Ethiopia selected Hospitals was 36.3% (95% CI: 31.6, 41.0), while stunting on the same study population was 5.5% (95% CI: 3.4, 7.8). Rural residence (AOR = 4.1, CI: 2.0, 8.4), lack of maternal education (AOR =9.3, CI: 5.0, 17.3), low CD4 counts (<500) (AOR =4.9, CI: 2.3, 10.4), using unprotected water source (AOR = 3.2, CI: 1.8, 5.8), having non-biological mother (AOR =4.2, CI: 1.9, 9.2) and recurrent oral lesion (AOR =2.2, CI:1.2, 4.2) were significantly associated with wasting. Furthermore, history of hospital admission (AOR =4.9, CI: 1.6, 15.0), recurrent oral lesion (AOR =3.9, CI: 1.1, 14.1), low CD4 counts (< 500) (AOR =3.5, CI: 1.0-12.0), advanced WHO clinical stage III (AOR =4.0, CI: 1.1, 14.2) were statically associated with stunting. Conclusion: This study found that the prevalence of under-nutrition among HIV-positive children in Ethiopia was significantly high. Rural residence, lack of maternal education, low CD4 count, recurrent oral lesion, having none-biological caretakers and unprotected source of water were significantly associated with childhood wasting. On the other hand, history of hospital admission, recurrent oral lesion, advanced WHO clinical stage and low CD4 counts were significantly associated with stunting of HIV positive children. Therefore, timely identification and monitoring of nutritional problems should be necessary to enhance the effectiveness of ART treatment and to prevent further related complications.


Background
The pandemic of human immunode ciency virus (HIV) is one of the major public health problems and associated with a range of long and short-term consequences (1). At the end of 2019, approximately 38.0 million people were living with HIV globally, of which 1.8 million were children (age 0-14 years) (2). Ethiopia is one of the Sub-Saharan Africa (SSA) countries which suffer from the global burden of HIV-infection. By the end of 2018, an estimated 56,514 children under age of 15 were living with HIV. Of which, around 2,994 were newly infected with HIV (3).
Malnutrition is one of the major causes of death for HIV positive children (4). Human Immune de ciency Virus (HIV) infection and malnutrition often coexist, which increases the risk of morbidity and mortality (5). Malnourished children have lowered resistance to infection and are more likely to die from common childhood illness. Children living with human immunode ciency virus (CLHIV) are physically stunted and underweight compared to non-infected children (6).
Maintaining good nutritional status remains very challenging issue for HIV-positive children. The problem is related with inadequate dietary intake, the effect of anti-retro Virus therapy (ART), and the HIV-infection itself (5). People with HIV/AIDS often do not eat enough as the illness and the drugs taken for it alter the food taste, decrease appetite, and inhibit the body rate of food absorption.
In Ethiopia, few studies have been conducted to assess nutritional status of CLHIV (5)(6)(7)(8)(9). However, to the best of our knowledge, there was no study conducted to explore the nutritional status and associated factors among HIV positive children (< 15 years of age) in the study area. Current and up-to-date evidence regarding nutritional status in HIV-positive children is essential for policy makers and clinicians to take appropriate actions. Therefore, the ndings of this study will highlight the magnitude and associated factors of malnutrition among HIV-positive children with implications to improve health workers' interventions, to ensure treatment effectiveness, and to accelerate the reduction of HIV related morbidity and mortality of children.

Study area, design, and period
An institutional-based cross-sectional study was conducted from February to March 2021 among HIV-infected children on ART in Southern Ethiopia. The study was carried out in three selected governmental hospitals (i.e.,Otona Teaching and Referral Hospital, Halaba District Hospital, and Duramie General Hospital). These hospitals provide service for more than six million people in the Region. All the three hospitals provide chronic HIV care and follow up services for HIV infected clients. Now a days, there are approximately 579 children (<15 years of age) receiving ART follow up service in these hospitals.
Study participants, sample size, and sampling technique All con rmed HIV-positive children (aged <15years) taking ART in Southern Regional State governmental hospitals were the target population. All HIV-infected children who had ART follow up at the selected hospitals were the study population. However, children with incomplete baseline medical information were excluded. Furthermore, a child who does not have care taker or parents to undertaken the consent, care takers diagnosed to have mental problem or children who have physical malformation and seriously ill were excluded for the study.
The minimum required sample size was determined using a single population proportion formula. To calculate our sample size, the following statistical assumptions were considered: 60.2% proportion (p) of malnutrition from a study done in East and West Gojjam Zones, Amhara, Northwest, Ethiopia (8); 5% margin of error; 10% nonresponse rate; and 95% con dence intervals (CI).
The calculated sample size was 369. After considering a 10 % non-response rate, the nal sample size of our study was 406.
This study was conducted in three randomly selected governmental hospitals. From the beginning, a sampling frame was prepared using the patient's medical registration number from each hospital's ART registration logbook. Then the total sample sizes were proportionally allocated for each hospital. Finally, study participants were taken from each of the three selected hospitals using computer generated simple random sampling technique.
Data collection tool and procedure The data abstraction checklist was developed from the current Ethiopian Federal Ministry of Health ART clinic intake and follow-up forms. Data were collected by trained health professionals through anthropometric measurement, face to face interview, and clinical record review. Training about the objectives of the study, contents of the tool, and data collection procedures was given for data collectors and supervisors for one day. Pretest was carried out at Sodo health center. During data collection time, care givers who had malnourished child were linked to therapeutic feeding center. Besides, weight and height were measured for each study participant, and nutritional advice was given to all caregivers. The assigned supervisors and principal investigator closely monitored and supervised the whole data collection process.

Operational de nitions
Under-nutrition: -was de ned when the children having either W/H or H/A or W/A z-score <-2 SD of the median value of WHO standard (10,11).

Data management and statistical analysis
The consistency and completeness of the collected data were examined during data management and analysis.
Data were entered into Epi Data Version 3.1 and analysis was done using Statistics Package for Social Science (SPSS) Version 25. The anthropometric measurements was converted into Z-scores using WHO Anthro Plus software version 3.2.2. Frequencies and cross tabulations were used to check for missed values of variables and to describe the study population in relation to relevant variables. Moreover, percentages, proportions, and summary statistics (mean, median) were used to summarize the study population characteristics. Binary logistic regression analysis was implemented to assess the association of factors against the outcome variable.
Variables with p-values < 0.25 in the bivariable analysis were entered into the nal model to control the effects of confounders and identify signi cant factors. Adequacy of the model to t the outcome variable with the predictors was checked using Hosemer-Lemeshow test for goodness of t. In the multivariable analysis, variables with p-values less than 0.05 at 95% CIs were considered as statistically signi cant factors. Finally, the strength and the direction of association were assessed using odds ratios with their correspondence 95% CIs.

Socio demographic characteristics of study participants
Out of 406 study participants, 383 were included in this study with response rate of 94.3%. Nearly half of the study participants' (50.4%) were boys and 157 (41%) were from rural areas. Children age less than 60 months were 124 (32.4%), while 52 (13.6%) of the study participants was between 60 and 120 months. The majority (73.6%) of caretakers were unmarried, and most (54.6%) of the caretakers were unable to read and write. Among the caretakers 152 (39.7%) were daily laborer and more than half (61.9%) of them have greater than four family in the house they live (Table 1).   Prevalence of under-nutrition The overall prevalence of wasting was 36.3% (95% CI: 31.6, 41.0) (Fig. 1). Moreover, the prevalence of stunting on the same study population was 5.5% (95% CI: 3.4, 7.8) (Fig. 2).

Factors associated with wasting
Bivariable and multivariable logistic regression analysis were conducted to determine factors associated with under-nutrition. In the multivariable logistic regression analysis, residency, maternal education, current CD4 count, having recurrent oral lesion, marital status of mother, caretaker's relation with child and source of water were signi cantly associated factors with wasting of HIV positive children. Subsequently, the likely hood for the presence of wasting was 4.1 times more likely among HIV positive children who lives in rural area (AOR 4.08; 95% CI: 1.98, 8.40) compared to who lives in urban. Regarding maternal education, wasting was 9.3 times (AOR = 9.33; 95% CI: 5.02,17.35) more likely to occur in those HIV positive children who have mother who was unable to read and write as compared to those who was able to read and write. Moreover, HIV positive children who had current CD4 count less than 500 cell/mm 3 were 4.9 times more likely to have wasting (AOR = 4.91; 95% CI: 2.33,10.37) as compared to their counterparts. The odds of wasting was 3.2 times more likely among the HIV positive children who uses unprotected water source (AOR = 3.23; 95% CI: 1.79, 5.79) compared to who uses protected water source. Additionally, HIV positive children those who have non-biological mother were 4.2 times more likely (AOR = 4.17; 95% CI: 1.89, 9.19) to have wasting than those whose caretakers was biological mothers. Lastly, HIV positive children who have recurrent oral lesion were 2.2 times more likely (AOR = 2.22; 95% CI: 1.17, 4.23) to have wasting than those who did not complain oral lesion (Table 4).

Discussion
In the current study, the prevalence of Wasting among children living with HIV/AIDS was 36.3 % (95% CI: 31.6, 41). This is in line with a study done in Nigeria which was 33.5% (12).
In this study, we also explored factors associated with under-nutrition among HIV positive children on ART in Southern Ethiopia. The results indicated that different factors were signi cantly associated with wasting and/or stunting. Accordingly, children who had recurrent oral lesion were more likely to develop wasting. This is in line with previous studies conducted in Ethiopia (Gojam, North Wollo) and Cameron (8, 13,16). This is because children with oral lesion have di culty of swallowing which reduces the amount of food intake that leads to nutritional imbalance less than body requirement.
This study revealed that children in the advanced WHO clinical stages were more likely to be wasted. This is in line with previous studies conducted in Ethiopia (5,16). This can be explained by the fact that HIV positive children who have advanced WHO clinical stage of disease are more vulnerable to opportunistic infections, making them susceptible to under nutrition by decreasing intake, altering digestion, absorption and metabolism as well as by increasing energy need (17). Additionally, children who lived in rural area were four times more likely to be wasted. This is in line with a study conducted in eastern Ethiopia (5). This is due to the fact that there is difference between rural and urban residents in access to health facilities for early management of malnutrition, in the level of awareness towards balanced diet, and infrastructure to access the health facilities.
This study revealed that wasting is associated with educational status of mother. This can be explained, because mothers who are unable to read and write are more prone to have knowledge de cit secondary to unable to read literatures and magazines, which talks about nutritional requirements for children living with HIV/AIDS. These mothers may also have lack of awareness on early management of malnutrition.
Finding from this study also revealed that, HIV-positive children who uses unprotected water source was more prone to wasting. The possible explanation for this, those HIV positive children who use unprotected water sources are more vulnerable to develop water borne disease like worms because of their immunocompromised status, which predisposes them to have malnutrition (18,19). Moreover, the occurrence of wasting was more likely in those HIV positive children who had none biological care takers. This might be due to the fact that, children whose caretakers are none-biological mothers are less likely to get breast feeding and they are also more prone to have poor drug adherence and good nutritional supplementation.
This study also found that children with history of hospital admission were associated with stunting. The reason for hospital admission among HIV positive children is most of the time due to opportunistic infection, which decreases food intake. Besides, HIV positive children classi ed as WHO clinical stage III are more likely to have stunting. This is supported by the study conducted in eastern Ethiopia (5). This can be explained by the fact that HIV positive children classi ed as advanced clinical stage of disease are more vulnerable to opportunistic infections, making them susceptible to under nutrition by decreasing food intake, absorption, altering digestion, and metabolism as well as by increasing energy need (20).
Additionally, CD4 count from 350 to 499 cell/mm3 was signi cantly associated with stunting. This could be explained by those whose CD4 count between 350 and 499 cell/mm3 may have low attention of care by health workers compared to those who have CD4 count less than 350 cell/mm3 (20). Furthermore, children who had recurrent oral lesion were more likely to be wasting. This is because children with oral lesion have di culty of swallowing which leads nutritional imbalance less than body requirement.

Limitations
Before interpreting the ndings, this study has its own limitation that must be considered. Since the study was done based on cross-sectional study design, it did not establish the possible cause and effect relationship between independent and dependent variables. There might be potential recall bias among respondents answering questions relating to events happening in the previous time. As the survey was conducted during a dry season, it was di cult to entertain the seasonal variations.

Conclusion
The nding of this study demonstrated that, the prevalence of wasting and stunting among HIV positive children was relatively high. Living in rural area, unable to read and write of caretakers, low CD4 count, recurrent oral lesion, none-biological caretakers and unprotected source of water were found to be signi cantly associated with wasting. On the other hand, history of hospital admission, recurrent oral lesion, advanced WHO clinical stage and low CD4 count were factors signi cantly associated with the occurrence stunting among HIV positive children.
Declarations of the article. All coauthors participated in literature search, analysis and interpretation of data, drafting the article and revising it for important intellectual content, and approve the nal version of the manuscript.

Availability of data and materials
The data sets used and/or analyzed during the current study are available from the Corresponding author upon reasonable request.

Consent for publication
Not applicable Ethics approval and consent to participate The study was conducted after obtaining ethical clearance from Wolaita Sodo University, College of health sciences, and medicine through ethical letter with protocol number 0437/2020, written on November 23, 2020.
The study was also done as per the declaration of Helsinki. An informed written consent was obtained from the care taker or parents of study participants after the purpose and procedures of study was fully informed. To maintain con dentiality participants name and unique ART number were not included in the data collection tool. Moreover, con dentiality of data was kept at all levels of the study and not used for any other purposes than the stated study objectives.