Weight Change and Its Predictors among Newly Initiated ART Clients in Dessie City, Northeast Ethiopia, July 2020. Longitudinal Data Analysis

The weight of HIV/AIDS patients is one of the classications WHO clinical staging of the diseases. A positive weight change in antiretroviral therapy patients is one of the expected clinical outcomes within a few months after the initiation of antiretroviral therapy in previously naïve patients. But the weight change varies across clients, and the reason for this variation and the effect of time-varying clinical proles on the weight of the clients is not well investigated. Method: A retrospective cohort study was conducted in Dessie City Health Facility in July 2020. The data were collected using a simple random sampling method in adult antiretroviral therapy clients who were enrolled to care between January to June 2019. Totally, 58 charts were reviewed within three months interval for 6 consecutive observations per chart. The data were entered into Epi-data, and analyzed using Stata 14. The effect of Panel and random effect model was assessed using Breusch and Pagan and Hausman's test, respectively. Finally, the Random Effect Generalize Least Square model was tted, and variables with a p-value less than 0.05 were considered as the predictors of weight change. A total 58 clients chart with 322 observations were assessed and the mean age (standard deviation) of participants were 37 (10) and 30 (51.7%) of them were female The absence of infection (β:1.85; WHO


Introduction
HIV infection and undernutrition are mainly affecting Sub-Sahara African Countries including Ethiopia (1). Weight of the patient is one of the most affected clinical parameters and an indicative measurement for treatment response.
Moreover, it is an important indicator for the classi cation of WHO clinical stage of the disease, which lies from unexplained weight loss in stage II into a severe wasting syndrome in stage IV (2).
A positive weight change among clients on antiretroviral therapy (ART) is one of the expected clinical outcomes within a few months after the initiation of ART in ART naïve patients (3). Most ART clients in Ethiopia are underweight (4), and a positive weight change shows a major indicator for clinical progress and one of the signs of gaining adequate nutrition (5). Weight of adult ART users attain the maximum weight within one year, and almost all adults' height keeps constant (6). Low baseline body weight is one of the main risks of death in naïve ART clients, and also early weight gain improves ART outcomes (1,7). Weight loss is one of an acute response for inadequate intake of nutrients, which induces a low immune response (5). The low immune response increases the diseases incidence severity and prolongs the duration of the diseases as well a poor weight gain is a potential sign of treatment failure (2). Low body mass index (BMI) decrease adherence status of patients to Antiretroviral (ARV) drugs (8).
So, taking a drug in previous naïve patients will expected a change in weight over time, as well as the interaction between weight and covariates, have a contribution to a change in weight in each patient (9). The weight change varies across clients some of them decrease, the other stagnant change and some others sharply increase, and the factor for this variation is not well investigated. ART providers measure the weight of their clients every month as a prognostic indicator and clinical compliance of ART drug, but there is no su cient evidence on the effect of timevariant clinical pro les and sequence of events for the contribution of weight change with its degree and directions (10). This study tries to assess the change of weight and determinant clinical pro les in Adult ART clients.

Methods And Materials
Study Area, Period and Study Design A retrospective cohort study was conducted in Dessie City Health Facility, which is 401 km away from Addis Ababa a capital city of Ethiopia. Overall, 514 clients started ART from January to June 2019. Based on the national modality of ART health service delivery, a newly ART initiated clients should visit the facility every month for one year and then may adopt to three or six months follow up a schedule following the client's health status and willingness (11). During their visit, clients were measured their clinical condition and recorded to ART follow up form then data stored into Smart care database (12).

Population
All adult ART users, who were taking ART drugs for the rst time from January 2019 to June 2019, were considered as source populations. Clients, who were attending at least three or more follow-ups, were included in the study.

Exclusion Criteria
Pregnant women, who were initiated ART between January to June 2019 and being pregnant from initiation to July 2020, were excluded from the study. Food by prescription, Three months follow up change on: regimen change: CPT status: CPT adherence, ARV adherence, Viral load measurement, OI, Clinical (T-stage), food by prescription.

Sample Size Determination
The sample size of this study was determined by standard longitudinal continuous variable outcomes, through the assumption that the number of subjects in each group is equal with a 95% level of Con dence and 80% of power (13).
Where N = the total sample size the distribution of z table andZ( 1-&beta;) is the normal distribution of the power δ 2 is the variance of outcome variable and T is the number of observations = is correlation coe cient in outcome variable and V 2 =is the mean deference in Advanced WHO stage and WHO early stage of the diseases at baseline.
Since there is a limited longitudinal study for the weight of Adult ART clients, a preliminary survey was conducted in the study area among 20 Adult ART clients for six observations. The sample size was calculated using the computed mean weight of early and advanced WHO Stage diseases, the standard deviation of weights, and the correlation coe cient of the weight of patients. Finally, the calculated sample size was 58 adult ART clients were eligible for the study with a maximum of six and a minimum of three observations.

Sampling Procedure
The total number of newly enrolled clients between January and June 2019 was 514. Among the newly enrolled clients in the study period, 421 and 93 were having three or more and less than three follow up, respectively.
Regarding the availability of necessary data 338 and 81 charts were complete and incomplete, respectively. During preparing the chart for data collection, 25 Charts were lost and 313 of the completed charts were availed for data collection. All the 313 clients' medical record numbers were listed and using simple random sampling technique 58 of them were select randomly through the lottery method.

Data collection Procedure
The data were extracted from the patient follow up chart and ART register in July, 2020. The checklist was developed from ART Guideline, ART register, and ART follow up forms. The baseline weight of the patients was taken at a follow-up chart, and other sociodemographic and clinical pro les were also taken at the same time.

Data Quality Control
The data were entered into epi-data version 3.1 and exported to Stata version 14.0. The data were entered by double data entry and check the consistency between the two data entry through double-checking and the exactness of missing data and only a few data were found inconsistent and managed through the veri cation of data from the extraction sheet.

Data Analysis Procedures
Data were described with mean and standard deviation and frequency present with

Ethical Issue
The ethical issue was secured from Wollo University College of Medicine and Health Science Ethical Review Committee and permission letters also were obtained from Dessie City Zonal Health Department and respective health facilities and all methods were performed in accordance with Helsinki Declaration. The information provided in each chart of adult ART clients was kept strictly con dential.

Sociodemographic Characteristics
A total of 58 adult HIV patient medical charts were assessed with their repeated follow up for about 322 observations. The mean age (SD) was 37 (10) years with a minimum of 19 years and a maximum of 61 years old.
Around 30 (51.7%) of them were female and, 24(41.4%) of them were also married. Regarding educational status and residence 25 (43%) and 39 (67.2%) were primary school level and urban dwellers, respectively. More than half 33(56.9%) of them does not disclose their HIV status to their families and relatives (Table 1).   The weights of ART clients were varying across sex, which is more varies at baseline among females, and also the variation continues over time. Unlike female, the variation in male at baseline lesser and also the variation in male slightly decrease over time (Fig. 1). Figure 1 Individual Pro le of Newly ART initiated ART Clients in Dessie City, July 2020.
The mean weight change shows a quadratic nature (Fig. 2), and there is a difference between males and females.
The mean weight among females is lower than male (Fig. 3). Figure 2 The mean Weight Chane in Dessie City, July 2020. Figure 3 The mean weight in Male and Female in Dessie City, July 2020 The Weight Changes across Months on ART.
The Repeated measure of ANOVA with a Wilk's Lamda statistics shows that there is a signi cant difference in the weight change of adult ART clients with a P-Value < 0.001. The variations were detected by wild statistics of Bonferroni methods and the result shows a signi cant difference every six months apart and more rather than within three months interval.
Adult ART clients gain weight 4.92 kg (C95%: 3.03-6.82) within one year after the initiation of ART clients. The mean weight changes after initiation of ART over 15 months of follow up is decreased to 3.28 kg. This study address change of weights after initiation of adult ART clients and its covariate predictors. The analysis shows that no opportunistic infection, the interaction between months on ART and absence of opportunistic infection, and WHO stage were statistically signi cant (Table 4). The interaction term of months on ART with the absence of opportunistic infection increases 0.09 kg of weight over time as a month on ART increases in adult ART clients. Adult ART clients with absence of opportunistic infection over a period of time in a month on ART increase their weight by 1.85 kg within the time span of their follow up.
Adult ART clients, who were experienced advanced WHO stage disease over a period of their follow up time a month on ART, were decrease their weight by 3.5 kg in their follow-up time (Table 4).

Discussion
The weight of adult on ART clients a positive progressive change on their follow-ups. A positively changed weight shows an improvement in their condition compared with their baseline. Most of the ART clients come in the advanced stage of HIV AIDS diseases (14), so taking ART drug that slows the replication of the virus and boosting their immunity over period of times (3). In a study conducted in Arba Minch, Ethiopia a short duration of taking ART shows a lower body mass index (15). Another study also shows a signi cant increment of weight within one year (6) it shows that 2.7 kg increases on average within one year (6). Monitoring weight of the patient in every follow up visit should consider the progress of patients towards the compliances of Antiretroviral Therapy and clinical conditions.
The interaction term of months on ART with absence of opportunistic infection was found statistically signi cant over time has a contribution of positive weight gain through times in adult ART clients. Adult ART clients with the experience of opportunistic infection in their follow-up period show a decrease in their weight over a period of time (16). Another study suggests that a longer stayed on ART increase the risk of acquiring an opportunistic infection (17). So, considering the nding of this interaction effect to keep clients free from any opportunistic infection for the good progress of weights for clients under ART.
Adult ART clients with absence of opportunistic infection over a period of time and had a positive contribution to weight changes though times. Adult ART clients not experiencing any opportunistic infection over a period of time show an increment in their weight during their follow-up period. Weight loss is the most common symptom during diagnosis (18) and after initiation of ART (19). Clients facing weight loss from previous follow up is the sign of developing new opportunistic infection (20). Opportunistic infections remain a challenge in patients receiving ART in resource-limited settings (21), so keeping clients free from any opportunistic infection is good progress toward positive weight gain and improves their clinical stage of diseases.
Adult ART clients, who were experienced advanced WHO stage disease over a period of their follow up time, were negatively associated with weight gain in their follow-up time. Adult ART clients who experience an advanced stage of the disease is a decline in their weight (16). The WHO stage is categorized based on different criteria among these criteria, severe weight loss and wasting syndrome are the criteria found in stage 3 and stage four, respectively(12) (22). This evidence is supported in different studies conducted in Arba Minch and Jimma, Ethiopia in which adult ART clients with advanced WHO stage signi cantly associated with undernutrition, which was measured using body mass index (15,23). A decline in the weight of Adult ART clients need crucial assessment on the stage of their diseases and therapy per the national guideline

Limitation of the Study
This study is not without limitation; 1) This study includes only patients' medical records with complete data, so this might have made a selection bias; 2) All the measurements documented in the chart were not sure about the calibration and consistency of measurements and also shares limitations of retrospective studies. Figure 1 Individual Pro le of Newly ART initiated ART Clients in Dessie City, July 2020.

Figure 1
Individual Pro le of Newly ART initiated ART Clients in Dessie City, July 2020.

Figure 2
The mean Weight Change in Dessie City, July 2020.

Figure 2
The mean Weight Change in Dessie City, July 2020.

Figure 3
The mean weight in Male and Female in Dessie City, July 2020