This study was conducted at the Day clinic of the Bamenda Regional Hospital located in Bamenda, the capital of the North West Region of Cameroon. Bamenda regional hospital is the leading tertiary care hospital in the region, providing care and treatment to over 5000 PLWHIV as in 2016. The North West Region was ranked fourth (5.1%) in the CAMPHIA 2017 HIV prevalence ranking in Cameroon [3].

This was a comparative hospital-based cross-sectional study involving both HIV positive and negative adults within the age group of 21 to 60 years. Study participants were recruited at the Day clinic of the Bamenda Regional Hospital. Recruitment of participants took a span of 4 months from March to June 2016.

Study population: Participants were recruited from HIV-infected patients who came to the centre for drug refill or routine check-up during the study period or healthy individuals who came to the clinic for voluntary HIV counseling and testing (VCT). The study population was recruited under 3 study categories: participants who were HIV negative, HIV infected participants not yet on HAART (HAART-naïve group) and HIV infected participants on HAART (HAART experienced group).

Adults, males and females aged between 21 to 60 years who voluntarily accepted to take part in this study and signed the written consent form, were recruited into the study. This is because according to the Cameroon code for the classification of adults, adulthood begins at the age of 21 years. There are reports of a higher probability of having reduced renal function among adults older than 60 years [22]. Therefore in an effort to reduce renal effect due to age, only participants younger than 60years were considered in this study. Participants who had a blood pressure >140/90mmHg and/or fasting blood sugar >120mg/dl were excluded. Participants who were obese, suffering from urinary track infection or any other acute disease, had history of kidney transplant or dialysis, or those with known kidney, liver or heart diseases as well as pregnant women were also excluded from the study. Participants on long-term medications other than HAART were not included.

Considering prevalence of renal dysfunction among HIV patients on HAART of 14.3% [16], at a 5% error margin, the following statistical formula was used to determine the sample size for this study.

Where n =sample size, Z= constant= 1.96, p = prevalence= 14.3%, e= error margin.

Serum and Urine Creatinine levels were analyzed spectrophotometrically with an ERMA BIOCHEMICAL ANALYZER (ERMA INC. Tokyo Japan, Model AE-600N) following the manufacture’s instructions. Urea and urine proteins were measured using a 3000 EVOLUTION spectrophotometer. Serum and urine creatinine concentrations were then used to calculate Creatinine clearance (Crcl) while serum urea and creatinine were used to calculate blood urea nitrogen (BUN) and BUN-creatinine ratio.

Renal impairment was classified according to the National Kidney Foundation clinical practice guideline based on the GFR determined by the Crcl method. Accordingly, estimated creatinine clearance values ≥ 90ml/min/1.73m², 60-89 ml/min/1.73 m², 30-59 ml/min/1.73 m², 15-29 ml/min/1.73 m² and < 15ml/min/1.73 m² were interpreted as normal, mild, moderate, severe impairment and the most severe as kidney failure [23]. Renal dysfunction was defined as Crcl <60ml/min/1.73m². Reference ranges for renal function tests were set as follows; Crcl >90ml/min/1.73m2, BUN 10-20mg/dl, BUN-Creatinine ratio 10-20, and urine protein <150mg/24hrs.

Codes were assigned to each participant so as to observe strict confidentiality. The results were entered in a secured log book. The questionnaires and logbooks were treated with strict confidentiality. Data was then entered into Microsoft excel version 10, cross-checked for errors and then transferred into the statistical package for social sciences (SPSS) version 21 for analysis. The differences between group means were compared using the analysis of variance (ANOVA) or student t-test. Fisher’s Least Significant difference multiple comparison was used for post hoc analysis for significant ANOVA comparison. Categorical data were compared using chi-square test. Pearson’s correlation was used to correlate variables. Statistical significance was set at p < 0.05.

A total of 206 adults visiting the Day Clinic of the Bamenda Regional Hospital were enrolled into the study. Three participants were excluded due to a blood pressure of >140/90mmHg and after laboratory analysis, two more patients were excluded for positive Nitrite and Leucocyte in their urine samples leaving behind a total number of 201 participants for the data analysis. The study population was sub-divided into three study groups; 65(32.3%) participants were HIV negative, 68(33.8%) were HIV positive patients not yet on HAART (HAART-naïve group) and 68(33.8%) were HIV patients on HAART (HAART experienced group). There were 57(28.4%) males and 144 (71.6%) females. The participants had ages ranging from 22 to 60 years with a mean of 37.4±9.6 years. The majority of the participants, 64(31.8%) were within the age group of 41 to 50 years. The HAART experienced group had the highest mean age, BMI, SBP/DBP and FBS among the study categories.

Table 1 summarises the sociodemographic characteristics of the study population.

Fig. (1) describes the distribution of participants on the various HAART regimens. A majority of the participants on HAART (82.4%) were on tenofovir-based regimen and had a mean duration on treatment of 6.95±3.74years.

Fig. (1). Distribution of the HAART group according to the HAART regimen.

The study population had a mean 24 hours urine proteins of 206.65±201.17mg/24hours. The cut-off value for a normal 24 hours urine protein was set at <150mg/24hours. Of the total study population, 101(50.2%) participants had normal urine protein concentrations and 100(49.8%) had abnormal urine protein concentrations. The HAART-naïve group had a proteinuria prevalence of 58.8%, HAART experience group had 54.4% and 35.4% for the HIV negative group. Fig. (2) summarizes proteinuria in the study population.

Table 2 below displays the prevalence of renal dysfunction within the three study groups. Of the total study population, 33 (16.4%) had Crcl <60ml/min/1.73m². The prevalence of renal dysfunction was significantly different among the different study groups (p= 0.022). The HAART experienced group had a higher prevalence of renal dysfunction (26.5%) than the HAART-naive group (11.8%).

This study demonstrated a significant difference in the mean renal function markers (BUN, BUN-Creatinine ratio, Crcl and Urine protein) among the different study groups (p-values 0.001, 0.003, 0.026 and 0.017 respectively) as shown in Table 3.

Table 4 displays the multiple comparison of the renal function between the study groups using Fisher’s Least Significance difference (LSD). The mean BUN concentration was significantly higher in the HAART group as compared to the HIV negative groups (p= 0.017) and lower in the HAART Naïve group when compared to the HAART groups (p=0.000). The mean BUN-Creatinine was significantly lower in the HAART Naïve than the HAART-experienced group (p= 0.001), mean Crcl significantly decreased in the HAART group when compared to the HIV Negative groups (p= 0.005) and the mean urine protein was significantly higher in the HAART Naïve than in the HIV Negative group (p= 0.007). However, though the HAART-naive group had a higher urine protein and the HAART experience group had decreased Crcl, these differences were not statistically significant (P= 0.150 and 0.053 respectively).

Fig. (2). Proteinuria prevalence amongst the study groups.

To ascertain whether age could affect the renal function, the renal markers were compared among age categories of the study participants as shown in Table 5. It was demonstrated that the mean BUN and mean Crcl were significantly different among the age groups (p-values were 0.046 and <0.001 respectively).

Renal functions in patients with different CD4+ counts were also analyzed as shown in Table 6. Comparing the means of the renal markers between the different CD4 count categories, only the mean BUN level was statistically different between the categories (p= 0.048).

Though there were differences in the mean BUN, BUN-Creatinine ratio, Crcl and Urine proteins between the participants with different durations on HAART, these differences were not statistically significant.

This study demonstrated that the renal function markers were not associated with the patients’ adherence to treatment. Pearson’s correlation showed a non-significant correlation between duration on HAART and renal function. CD4 count showed a significant negative correlation with urine protein and positive correlation with BUN (r= 0.200, p= 0.020) and BUN-Creatinine ratio (r= 0.246, p=0.004).

This study consisting of HIV outpatients and healthy HIV negative participants attending the Day clinic of the Bamenda Regional Hospital, demonstrated a prevalence of renal dysfunction among HAART-naive patients of 11.8% and 26.5% among HAART experienced patients. The demonstrated prevalence of renal dysfunction in HAART-naïve patients (11.8%) was similar to 14.4% reported by Odongo et al. of Uganda [20]. However, Kahsu et al. of Ethiopia [24] reported a higher prevalence among HAART-naive patients of 30.1%. This difference could be because, instead of the conventional 24 hours urine creatinine clearance as used in this study, Kahsu et al. used the Cockcroft-Gault equation to estimate the creatinine clearance. Hypertensive and diabetic patients included in the Kahsu et al. study could be a reason for this disparity. The prevalence of renal dysfunction among HIV patients on HAART of 26.5% in this study was similar to the 29% reported in a Tanzanian cohort by Mpondo et al. [17]. Nevertheless, this prevalence was higher than the 14.3% reported by Nsagha et al. [14]. This discrepancy could be due to the difference in the method used to estimate the GFR. Nsagha et al. used the modification of diet in renal disease (MDRD) equation to estimate the GFR. The high prevalence of renal dysfunction recorded in this study may also be due to the fact that most of the patients were on Tenofovir-based regimens. Also, the participants on HAART in this study had a high mean blood pressure, age and FBS which could have also contributed to the higher renal dysfunction reported.

In this study, the mean Crcl significantly decreased in HIV patients on HAART when compared to the HIV negative controls, indicating that HAART could have a devastating effect on the kidney. However when renal function was compared between the HAART Naïve and HAART experienced group, there was no significant difference in Urine protein and Crcl. Thus the decreased renal function might only be the effect of HIV. Our result was in line with the observation of Obirikorang et al. who reported renal function to not be associated with the use of HAART [23]. However, a different observation was made from that of Nsagha et al. [14] who reported HAART to be nephrotoxic. This could be due to the fact that Nsagha et al. did not consider HIV infection as a determinant of renal function thereby attributing the cause of renal dysfunction entirely to the effect of HAART. Patients on HAART had a higher mean BUN and BUN-Creatinine ratio when compared to the HAART-naive patients. This result was different from the report of Kahsu et al. [24] who demonstrated that HAART-naive participants have a higher mean BUN. Our study demonstrated high proteinuria among HAART-naïve participants of 58% when compared to HIV negative participants, which could mean that HIV increases the risk of developing proteinuria. It was demonstrated that though the prevalence of proteinuria was higher in the HAART-naïve participants than the HAART experienced participants, this difference in mean urine protein was not significant between these two categories. This could further signify that HAART does not reverse the damages caused by HIV. Our prevalence of 58% was not in line with Kaze et al. [25] of Cameroon who reported a prevalence of 36%. This disparity could have been because, instead of the gold standard 24 hours urine protein as used in this study, the dipstick urinalysis was used in the study by Kaze et al. which is less sensitive.

This study demonstrated older age to be associated with decreased Crcl. This was in agreement with the report of Mpondo et al. [17] where age also reported to be associated with a decrease in renal function. This study also demonstrated no association between decreased renal function and the type of HAART regimen, the duration on HAART as well as the CD4 counts. This was in agreement with the report of Kamga et al. [26] who reported that the effect of HAART on the renal function was irrespective of the HAART regimen used, the duration of treatment or the CD4 counts. Our result was not in agreement with that reported by Nsagha et al. [14] who reported the HAART regimen tenofovir, lamivudine and efavirenz to be the most nephrotoxic. This study also reported no association between decreased renal function and patient’s adherence to treatment.

This study can be considered limited as microalbumin was not measured, thus a total lower urine protein was not included. However, a quantitative measurement of urine proteins gives a good picture of the renal function. Another limitation in this study was that the evaluation of proteinuria after 3 months was not repeated to differentiate acute from chronic kidney disease; nonetheless, the BUN-Creatinine ratio was calculated which could serve this purpose.