Prevalence and predictors of renal dysfunction among people living with HIV on antiretroviral therapy in the Southern Highland of Tanzania: a hospital-based cross-sectional study

Introduction infection with Human Immunodeficiency Virus (HIV) and use of antiretroviral therapy (ART) poses a significant risk of developing renal dysfunction in people living with HIV (PLHIV). Renal dysfunction contributes to the morbidity and mortality of PLHIV. There is limited information on renal dysfunction among PLHIV in the Southern Highland, the highest HIV prevalent area in Tanzania. We conducted a study to estimate the magnitude and predictors of renal dysfunction among PLHIV on ART. Methods a cross-sectional study was conducted at Njombe Town Council Hospital from December 2019 to April 2020, recruiting 396 participants. Serum was obtained to measure creatinine level then calculated glomerular filtration rate (GFR) using CKD-EPI and the Bedside Schwartz equations. The participants' informations were collected using a structured questionnaire. Data analysis was performed using STATA version 15; a modified Poisson regression model was used to estimate prevalence ratios (PR). The level of significance was specified at 0.05. Results the overall prevalence of renal dysfunction defined as GFR less than 90 mL/min/1.73 m2 was 20.7%, which increased by 4% as the age increases. The prevalence of renal dysfunction was higher in PLHIV on ART for more than six months to 24 months compared to their counterparts. Likewise, obese individuals had a 2.5 times higher prevalence of renal dysfunction than normal individuals. Conclusion there is a relatively high prevalence of renal dysfunction among PLHIV on ART, predicted by age, duration on ART, and nutrition status.


Introduction
Renal dysfunction is a common health problem worldwide, with epidemiological variations from one population to another [1]. Renal dysfunction is the kidney's impairment or abnormal functioning that leads to a decline of glomerular filtration. In Africa, 2% to 41% of people in the general population and 11% to 90% of people living with HIV (PLHIV) had renal dysfunction manifesting as Chronic Kidney Disease (CKD) in the year 2017 [2]. In Tanzania, studies reported a high prevalence of renal dysfunction among PLHIV in tertiary hospitals as 85.6% at Bugando Medical Centre, Northern Tanzania [3], and 32.8% at Muhimbili National Hospital, Eastern Tanzania [4].
Renal dysfunction contributes to significant morbidity and mortality for PLHIV [5]. Older people are at higher risk of developing renal dysfunction than the younger ones [6,7]. HIV replication in renal cells that damage the kidney and exposure to some antiretroviral drugs promotes renal dysfunction [8,9]. Additionally, the use of ART increases life expectancy but also is associated with increased risk of other non-communicable conditions such as diabetes mellitus and hypertension that contribute to renal dysfunctions [10].
Experience from clinical settings in Tanzania suggests infrequent routine monitoring of kidney status to PLHIV. Similarly, it is not common to calculate the glomerular filtration rate (GFR) in care and treatment clinics (CTC), leading to underreporting of renal dysfunction cases. There is insufficient scientific information on the magnitude of renal dysfunction in Southern Highland, the area with the highest prevalence of HIV in Tanzania [11]. The current study gives out the magnitude and factors associated with renal dysfunction among PLHIV on ART attending primary CTC in Njombe Region. The findings may help relevant health authorities to consider routine screening and subsequent renal disease treatment among PLHIV. Sample size and sampling: a minimum study sample size was estimated using Leslie and Kish formula; the calculation considered the prevalence of renal dysfunction among adult PLHIV starting ART in Mwanza, Tanzania [3]. A study enrolled a total of 356 PLHIV on ART. Self-reported and documented pregnant women were excluded because of a reported dynamic nature of GFR during the gestational period [12]. PLHIV were stratified based on duration on antiretroviral therapy use, as follows stratum 1: on ART for zero months up to six months, stratum 2: on ART for more than six months to 24 months, and stratum 3 -on ART above 24 months. Proportional to size sampling was used to determine the number of subjects to be selected from each stratum [13].
Data collection: information was collected using a structured questionnaire administered by trained CTC staff using the Swahili language (Tanzanian, native language). We collected socio-demographic information and clinical information from participants, and some extracted from participants' files. Weight in kilogram was measured using a standard weighing machine, while height in centimeter was measured using a height scale then converted into meters.
Laboratory procedures: venous blood from the arm's antecubital area was collected in a tube without anticoagulant by a trained phlebotomist. Blood was centrifuged within one hour at 1500 x g (relative centrifugal force) for 10 minutes at ambient temperature to obtain serum. The serum was placed into cryotubes and stored at -20°C until testing. All specimens were transported to the National Health Laboratory, Quality Assurance, and Training Center (NHL-QATC) at Dar Es Salaam for testing. Serum creatinine values were obtained using an automated chemistry analyzer, Cobas Integra 400 plus (Roche Diagnostic Ltd, Rotkreuz, Switzerland). The study used the NHL-QATC standard operating procedure for processing clinical chemistry samples. The quality control (normal and pathological) were performed before processing study samples and after every thirty samples.
Estimation of glomerular filtration rate: glomerular filtration rate (GFR) for participants aged 15 years and above was calculated using Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equation: Where Scr is serum creatinine, κ is 0.7 for females and 0.9 for males, α is -0.329 for females and -0.411 for males, min indicates the minimum of Scr/κ or 1, and max indicates the maximum of Scr/κ or 1 [16]. CKD-EPI equation shows high performance and allows better renal function staging in HIV infected adults [17]. The Bedside Schwartz equation was used to calculate GFR for children less than 15 years old: GFR= 41.3 x (height/Scr), whereby Scr is serum creatinine in mg/dL and height in meters [16,18]. GFR was used to estimate each participant's renal function status; renal dysfunction was defined when GFR is less than 90 mL/min/1.73 m 2 .
Data management and analysis: the analysis was performed by using Stata version 15.1. The continuous variables were presented as mean (standard deviation) or median (interquartile ranges). Categorical variables were summarized in terms of frequencies and percentages. Chi-Square and Fisher exact tests were used to test the difference between proportions. The variables with a p-value of < 0.25 or considered necessary from the literature were retained in the regression model for controlling possible confounder. A modified Poisson regression model was employed to assess factors associated with renal dysfunction as an alternative to logistic regression because the prevalence was above 10% [19,20]. The strength of association was expressed using prevalence ratio (PR), defined as the prevalence of a disease among exposure over the prevalence of a disease among unexposed. Variables with p-value < 0.05 in multivariate analysis were considered statistically significant.

Results
Socio-demographic characteristics of study participants: a total of 396 PLHIV on ART were enrolled in this study. The majority of study participants (52.5%) were aged between 18 and 44 years, with a mean (± SD) of 41.5 (± 12.4) years. Most participants (59.6%) were female, and 71.5% were urban residents. More than half, 53.8%, were married or living with partners, 74.0% had primary education level, and 61.6% were farmers. At time of the study, 71.5% of study participants were on treatment for more than 24 months, and 98.5% were taking TDF containing regimen. The median duration on ART was 68 months. Overall, 96.0% had viral load suppression defined as HIV viral load less than 1000 copies/mL. More than half (54.8%) of participants had normal BMI. A total of 8 (2.0%) individuals had known diabetes mellitus, and 20 (5.0%) had known hypertension ( Table 1).
Prevalence of renal dysfunction among PLHIV on ART: the overall prevalence of renal dysfunction was (20.7%, 95%CI, 17.0 -25.0). The renal dysfunction was significantly higher among participants aged 45 years old and above (28.2%) than the age group 18 -44 years (14.9%), p=0.006. There were no significant differences in the proportion of renal dysfunction between participants based on sex, marital status, and place of residence (Table 2). A high prevalence of renal dysfunction was observed among participants on ART for more than six months to 24 months (38.7%) and obese (48.2%) than others in the same categories, p < 0.001. A high proportion of renal dysfunction was detected among participants with hypertension (35% vs 20%) and diabetes mellitus (25.0% vs 20.6%); however, the difference was not significant ( Table 2).

Predictors of renal dysfunction among PLHIV on
ART: in univariate analysis, age, marital status, duration on ART, and BMI associated with renal dysfunction. With an increase in age by one year, the prevalence of renal dysfunction increases by 3% (cPR, 1.03, 95% CI, 1.0-1.1, p < 0.001). Those on ART for more than six months to 24 months had 2 times the prevalence of renal dysfunction than those on ART for six months or less (cPR, 2.3, 95% CI, 1.1-5.0, p = 0.04). Obese participants have 2.8 times prevalence compared to those with normal weight (cPR, 2.8, 95% CI, 1.8-4.2, p < 0.001). Individuals who are not living with a life partner (divorced/widowed/separated) had a 1.8 times prevalence of renal dysfunction than those living single (Table 3).
In multivariable analysis, age, duration on ART, and BMI were the only variable independently associated with renal dysfunction. One year increase in age had a 4% increase in prevalence of renal dysfunction (aPR, 1.04, 95%CI, 1.0-1.1, p < 0.001). Participants on ART for more than six months to 24 months had 2 times the prevalence of renal dysfunction than those on ART for six months or less (aPR, 2.1, 95%CI, 1.0-4.5, p = 0.04). Obese participants had 2.5 times prevalence than those with normal BMI (aPR, 2.5, 95%CI, 1.7-3.8, p < 0.001) ( Table 3).

Discussion
We found a considerable high prevalence by defining renal dysfunction as having GFR less than 90 mL/min/1.73 m 2 (20.7%). Age, duration on ART, and BMI were independent predictors of renal dysfunction among PLHIV on ART. The current study's finding on the prevalence of renal dysfunction is comparable with a study conducted in Spain in which the overall prevalence was 25.0% [21] and in Ethiopia, where the prevalence was 25.4% [22]. However, some studies have reported a higher prevalence of renal dysfunction than our study's finding [8,23,24]. The studies conducted in tertiary health care settings reported 32.8% in Eastern Tanzania [4] and 85.6% in Northern Tanzania [3]. Differences in the study setting might explain the observed differences in the prevalence of renal dysfunction. Tertiary health care settings are likely to have individuals with significant renal dysfunction compared to primary health care settings where we conducted our study. We found an association between renal dysfunction and age; as age increases, the prevalence of renal dysfunction also increases. Similar findings have been reported in studies conducted in Tanzania [3,25] and from different parts of the world [6][7][8]23]. The study conducted in Mwanza, Northern Tanzania suggested that individuals who start ART at older age developed renal impairment at large as measured by the GFR [3]. The relationship of age and renal dysfunction calls for screening of renal function status to older clients before ART initiation and regular monitoring PLHIV on ART.
The duration of ART was associated with renal dysfunction. PLHIV on ART for more than six months to 2 years had more prevalence of renal dysfunctions than those on ART for up to six months. The same finding was reported in France [26] and agreed with different studies that documented the association between renal impairment and the use of ART [3,8,27]. Our finding is supported by Soto et al. clinical evidence, which testifies ART toxicity, resulting in a severe proximal tubular injury and proximal tubulopathy to HIV [28]. Additionally, nearly all study participants (98.5%) were on a TDF-containing regimen, which may have contributed to our findings on renal impairment, as described in other studies [10,26,29].
The other factor we found associated with renal dysfunction in the current study is being obese as defined by BMI of ≥ 30 kg/m 2 . Close to half of individuals with obesity were found to have renal dysfunction and had approximately 3 times the prevalence of renal dysfunction than those with normal BMI. Several studies reported a similar association of obesity with renal dysfunction results [3,21,30]. The biological explanation of the study finding is that most obese individuals have an increased risk of diabetes mellitus and hypertension, the traditional predictors of renal impairment [3,30].
The current study's major strength is the use of modified Poisson regression as an alternative method for classical logistic regression. The method provided the genuine depiction of estimated measures of effect while logistic regression tends to overestimate the effect mostly when the prevalence is more than ten [19,20]. This study's weakness is the failure to establish the causal and effect relationship due to the study design; however, it provides a snapshot of the burden of renal dysfunction.

Conclusion
This study found a relatively high prevalence of renal dysfunction among PLHIV on ART predicted by age, duration on ART, and BMI. The findings suggest a need for routine screening and monitoring renal function status at CTC service delivery for early detection of kidney impairment for proper treatment.
Funding: this study received financial support from Tanzania Field Epidemiology and Laboratory Training Programme for data collection and laboratory work. The funder had no role in the study's design as well as the collection, analysis, interpretation of data, and writing of the manuscript.

What is known about this topic
 Renal dysfunction is a devastating health problem in general population globally;  Renal dysfunction is predicted by different factors such as age, use of ART, nutrition status and viral status.

What this study adds
 Renal dysfunction is wide spread problem, can also found in PLHIV attending CTC in primary health care settings;  Supporting the idea that there is possibility of recovering from renal dysfunction to PLHIV on ART for more than two years.

Competing interests
The authors declare no competing interest.

Authors' contributions
MNM conceived the study, designed and performed the study, participated in the field, performed data analysis, contributed to interpretation of results, and drafted the manuscript; LJU contributed to overall study design, data analysis and interpretation of data; MVM contributed to overall study design, data analysis, interpretation of data and critical review of the manuscript. All the authors have read and agreed to the final manuscript. (MoHCDGEC) for their financial support. Special thanks to the National Health Laboratory Quality Assurance and Training Centre (NHLQATC) for testing all samples. Special thanks to the participants, Njombe Regional officials, Njombe Town Council Hospital staffs, particularly CTC and laboratory staff, all for permission, support, and cooperation during data collection. Table 1: clinical characteristic of study participant during ART initiation and at time of the study Table 2: distribution of renal dysfunction by sociodemographic and clinical characteristics of study participants Table 3: factors associated with renal dysfunction among PLHIV on ART Figure 1: status of renal function among study participants