Early screening of chronic kidney disease patients among the asymptomatic adult population in Bangladesh

1Department of Public Health and Informatics, Jahangirnagar University, Savar, Dhaka-1342, Bangladesh 2Centre for Health Equity Training Research & Evaluation (CHETRE), UNSW, Sydney, Australia 3Department of Genetic Engineering and Biotechnology, University of Dhaka, , Dhaka, Bangladesh 4Department of Biochemistry and Cell Biology, Bangladesh University of Health Sciences, Dhaka, Bangladesh 5Pothikrit Centre for Health Studies (PCHS), Sangckriti Bikash Bhaban 1/E/1 Paribag, Dhaka, Bangladesh


Introduction
Chronic kidney disease (CKD) is a serious public health concern that is the major cause of morbidity and mortality worldwide. CKD is defined by the kidney damage that is usually detected through certain markers including albuminuria or estimated glomerular filtration rate (eGFR <60 mL/ min/1.73 m 2 ) (1). The prevalence of CKD among the general population worldwide was estimated at around 11%-13% (2), whereas the prevalence was reported to be higher in developing countries compared to developed countries (3). According to an earlier report (4), in 2015 around 109.9 million people from developed countries had CKD while these numbers were around 387.5 million for developing and lower income countries. In South-East Asia, the prevalence of CKD was ranging from 10.6% in Nepal to 23.3% in Pakistan (3). Bangladesh is also experiencing huge burdens in the total number of CKD patients. In Bangladesh, the overall prevalence of CKD was estimated at around 14% (5), while these numbers varied substantially depending on age and other

Key point
In a screening program among the 400 subjects in the Thakurgaon district of Bangladesh to identify people with the risk of CKD, 18.2% of subjects with CKD were f asymptomatic in our study. Since a substantial number of adults were found to be unaware of the existence of CKD hence large-scale prevention programs should be undertaken to reduce the classical risk factors of these disorders. factors (5). According to a study conducted in Dhaka, the prevalence of CKD was 26% for people aged 30 years or above (6), and 13% for people aged 15 years or above (7). CKD is influenced by several factors including age, gender, socioeconomic conditions, and health risk factors like cardiovascular diseases (CVD), diabetes, hypertension, obesity, etc (8)(9)(10). According to previous reports, the prevalence of CKD among the diabetic patients was 34% (11), around 26.8% to 47.2% among the CVD patients (12,13), around 30.1% among the young obese people (14). Along with the effect on health, CKD also impacts social life (15), whereas financial burden is identified as the most common form of social burden. Most of the CKD patients are at higher risk of developing end-stage renal diseases (ESRDs) that often require kidney transplantation and dialysis (16).
Substantial number of people in Bangladesh are at risk of CKD without being diagnosed (8). Many patients in their early stage of CKD are asymptomatic and they do not take any measures to treat which posed them serious health risks (17). Therefore, the actual number of people with CKD cannot be determined as the first and second stages of CKD are not easily detectable. Thus, the screening of CKD patients in their early stages could offer valuable opportunities for effective interventions that could reduce the risk of possible CKD outcomes. However, no such study has been conducted so far showing early detection of CKD among the Bangladeshi population. Only the overall prevalence of CKD among the Bangladeshi population and its associated risk factors were reported by few studies so far (6,10,18).

Objectives
Here, this study attempted to screen CKD patients in their early asymptomatic stage among the adult Bangladeshi population. This study will also consider the association between CKD with other risk factors including demographic and baseline factors, lifestyle and clinical factors among the Bangladeshi population. We also aimed to perform logistic analysis to find out the adjusted associations between CKD with other risk factors.

Study participants
This study was conducted to generate epidemiological data of CKD among the asymptomatic adult population from Thakurgaon district in Bangladesh. This screening program was conducted between January 2013 to December 2013 among 400 subjects who were willing to participate in the study. All the participants were chosen randomly. However, participants with previous history of kidney diseases, or taking drugs that could affect kidney functions were excluded from the study. Then we interviewed all the participants using a close-ended structured questionnaire. The questionnaire included several questions regarding demographic information, lifestyle factors like smoking habit and alcohol consumption habit.

Sample size
Sample size was calculated using the formula: n= z 2 p (1-p)/ d 2 ; where z= Standard normal deviation at 95% confidence level, p= assumed target proportion to have a particular characteristic, d= margin of error and n= sample size.
Considering the proportion of patient with particular characteristics (p) = 50% =0.5 (since the actual number is unknown), z = 1.96, margin of error (d) = 5% = 0.05; then at 95% confidence interval, sample size (n) becomes 384. However, due to 5% non-response, we added 16 samples for conducting this study. So, total sample size was 400.

Biochemical measurement
The participants' blood pressure, waist and hip circumference, weight and height were measured at the time of hospital visit and then body mass index (BMI) and waist-hip ratio (WHR) were calculated accordingly. Blood samples (5 mL) were collected for the measurement of serum fasting blood glucose, serum albumin and serum creatinine. Blood specimens were centrifuged to separate serum and were preserved immediately at -20°C for the future analysis.
Serum glucose (fasting and 2 h after breakfast) was measured by glucose oxidase (GOD-PAP) method (19) (Randox Laboratories Ltd., UK). Serum creatinine was estimated by colorimetric method using an auto-analyzer (20). Proteinuria was assessed by Dipstick method in most outpatient settings to measure the urine protein concentration (21). Urine albumin was measured by BNII ProSpec nephelometer, and urine creatinine was measured by Jaffe method using the Modular-P chemistry analyzer (Roche/Hitachi, Switzerland) (22). For determination of estimated glomerular filtration rate (eGFR), modification of diet in renal disease (MDRD) formula version 4 was used (23,24). According to K/DOQI and Kidney Disease (25), we should have at least three months history or three positive test results. Since we conducted a screening program and tested only once, we used the term "likely CKD" instead of CKD.

Ethical issues
The study was performed according to the ethical standards of human experimentation in accordance with the Helsinki Declaration. Here, written consent was taken from each respondent and the study was approved by the appropriate Ethics committee. The participants were notified about the aim and purpose of this study. Anonymity, confidentiality and voluntary participation were ensured. The participants' safety was assured by not doing any kind of harm to them. Sensitive and polite wording were used.

Statistical analysis
Data analysis was done using SPSS program version 24 software (SPSS Inc., Chicago, USA). Quantitative variables were presented as mean ± S.D. Categorical variables were shown in number and percentage. Since, the data followed normal distributions, for analysing categorical variables, Pearson's χ2 (chi-square) test was performed and for continuous variables, student's t-test was performed. The level of significance of each result was set at P < 0.05. All statistical tests were 2-sided. Binary logistic regression analysis was also performed to determine the adjusted associations between CKD and other risk factors. Table 1 showed the demographic characteristics of the study subjects. Here, the likely CKD subjects were termed as CKD group and non-CKD subjects as control group. Among the total 400 subjects, 73 (18.2%) subjects were diagnosed with likely CKD, and the rest 327 (81.8%) were control. The mean age of the subjects was found 48.94 ± 11.00 for control, and 53.09 ± 11.11 for the CKD group, which differed significantly (P = 0.01). Within the age categories, most participants in the control group were belonged to 30-40 years (29.1%), followed by 41-50 years (28.4%), 51-60 years (26.6%) and >60 years (15.9%). Whereas, most of the participants in the CKD group were belonged to >60 years (30.1%), followed by 51-60 years (27.4%), 41-50 years (26%), and 31-40 years (16.4%). Almost half of the subjects in both groups (67.3% for control and 63% for CKD group) were found male. The mean BMI for the control group was 21.40 ± 5.25, and that for the CKD group was 24.75 ± 5.14, which was significantly higher (P = 0.00). Among the BMI categories, most of the subjects belonged to normal weight (59% in the control group, and 41.1% in the CKD group). Regarding educational qualification, over half of the subjects (52.9% in control, and 52.1% in CKD) in both groups were found with primary level education. Most of the subjects in both groups were domestic helper (33% in control, and 37% in CKD group) and day labourer (42.5% in control, and 41.1% in CKD group). Furthermore, most of the study subjects were found married (94.5% in control, and 95.9% in CKD group). Finally, the WHR was within normal category for 54.8% of CKD subjects and within low and high-risk categories for 9.6% and 35.6% of CKD subjects respectively, whereas in case of control group, the WHR categories belong to normal, low and high-risk were 58.1%, 14.7% and 27.2%, respectively.

Prevalence of CKD among the asymptomatic adult subjects
Here, among the total 400 subjects, 82% (327) were in the control group that means their eGFR was found >90 mL/ min/1.73m 2 with no proteinuria and kidney damage. On the other hand, 15% (60) of total subjects were found with stage 1 (eGFR of > 90 mL/min/1.73 m 2 with evidence of Volume 5, Issue 1, 2020 kidney damage). Only 3% of total subjects were found with stage-2 (mild renal insufficiency with eGFR of ≥ 60-89 mL/min.1.73 m 2 ). Since we conducted our study only among the undiagnosed populations, we did not find any subjects with stage 3, 4 and 5 ( Figure 1).

Proteinuria and albumin-creatinine ratio significantly affected CKD
All subjects were tested for proteinuria, albuminuria and urine creatinine. Table 2 showed overall proteinuria status and albumin-creatinine ratio (ACR) among the CKD and control subjects. Among the 73 CKD subjects, over half (58.9%) exhibited +1 proteinuria, 24.6% exhibited +2 proteinuria, and 6.5% exhibited +3 proteinuria, while these proportions represented only 7.6%, 1.5% and 0.4%, respectively in case of control subjects. The difference in proteinuria status between the CKD and control groups was statistically significant (P = 0.00). On the other hand, ACR was found less than 30 mg/g for 92% of control subjects, but this proportion belonged to only 8.2% in case of CKD subjects. About 78% of subjects with CKD had ACR within 31-300 mg/g, and 13.8% subjects had ACR >300 mg/g. However, in case of control subjects, only 08% subjects had ACR within 31-300 mg/g range, and none had ACR >300 mg/g. The differences in ACR between the CKD group and the control group were significant as well (P < 0.01).

Association between CKD with other risk factors
We also investigated the association between CKD with other risk factors (Table 3). Here, we observed that smoking, diabetes, obesity, and hypertension were significantly (P < 0.05) associated with CKD. According to our study, the risks of developing CKD increased 1.8 times due to smoking, 10.68 times due to having diabetes, 7.43 times due to obesity, and 2.08 times due to hypertension. However, our study didn't find any significant correlation between CKD and alcohol consumption.

Logistic regression analysis
Finally, we performed a logistic regression analysis to find out the adjusted association between CKD and other risk factors. Here in our study, the presence or absence of CKD was considered as dependent variable, and age, gender, BMI, diabetes, smoking habit, hypertension, and WHR were considered as independent variables. Independent variables were further categorized considering age (30-40 years), gender (male), BMI (normal weight), smoking habit (non-smoker), alcohol consumption (nonconsumer), WHR (normal), hypertension (no), diabetes (no) as reference group.

Discussion
CKD is a worldwide public health threat, but the dimension of this problem in Bangladesh is probably not fully appreciated. Since patients in the early stages of CKD are largely asymptomatic (8), they often do not take any measures to treat which eventually progressed over time to ESRD. This prevented finding out the actual prevalence of CKD in the poor country. This study, for the first time, aimed to screen of CKD patients in their early stages, to avoid adverse outcomes of CKD through early detection and intervention. Here, we found that the overall prevalence of likely CKD was 18.2% among the undiagnosed asymptomatic Bangladeshi subjects, where most of them were in stage 1 (82% of CKD subjects). Among the associated risk factors, we had identified diabetes, hypertension, obesity, and smoking that significantly affected CKD. Stages 1 and stage 2 of CKD appear to be very common as reported in the National Kidney Foundation (NKF) K/ DOQI guidelines (25). Our study showed 82% of likely CKD subjects were in stage 1, and 18% were in stage 2. These findings supported earlier study that reported prevalence of CKD in stage 2 was around 12% (8). Our study showed an overall prevalence of CKD around 18.2% which also correlated with previous reports showing overall prevalence of CKD among Bangladeshi population (around 14% to 19.5%) (5,8). Since the symptomatic participants avoided takeing part in screening, our study did not find any subjects in CKD stage 3, 4 and 5. This is more alarming since large proportion of asymptomatic individuals in our study had found with CKD. Thus, this study surely showed the effectiveness of identifying a large number of individuals in the early stages of CKD, which is alarming for a developing country like Bangladesh. In our study, the association between proteinuria with CKD were investigated. Prevalence of proteinuria was significantly (P = 0.0001) higher among previously documented CKD patients. Overall, 10.75% of the participants in our study showed +1 (30-100 mg/dL) positive dipstick proteinuria. This value was lower than the 15% found in Congo (28) and 11% found in Bolivia (29). It had been reported that proteinuria accounts for 4-1% of the population in Japan (30), 7%-9% in Taiwan (31) and 0%-5% in China (32). This discrepancy between studies may be partly due to various definitions (microalbuminuria versus dipstick proteinuria) and the criteria of selection (random versus without random sample and/ or highrisk population versus whole population) applied in each survey. As it is a key prognostic finding and it has a significant beneficial role for early diagnosis of CKD, screening of urine for proteinuria using the dipstick test is being used successfully (33).
According to previous studies, CKD is strongly associated with age since ageing is associated with gradual loss of total glomeruli (34). Our study also found that most of the subjects with likely CKD were in the age group of above 60 years (30.1%). Nevertheless, our study showed that younger age group (30-40 years) accounted for 16% of total likely CKD, which supported earlier study that reported a significant proportion of middle-aged people developed CKD in Bangladesh (10). When we adjusted our data, we found that age above 60 years increased the chance of developing CKD by 3.39 folds (ORs: 3.39) than the younger age group (30-40 years).
Iseki et al reported that the prevalence of CKD was lower among people who were married (35). In this study, we saw that among the CKD patients, 95.9% were married. This might be because the mean age for marriage among the rural Bangladeshi people was 18.3 for women and 25.7 for men according to Bangladesh Bureau of Statistics-2018 (36), and most of the subjects in our study after 30 got married which prevented us to get enough unmarried subjects. Furthermore, the positive association between education and health is well established (37). In the present study, we observed that 60% of the study population had completed primary level education, so health related awareness was very poor. Furthermore, individuals with higher occupational status are less likely to develop a poor health outcome than those with a lower occupational status (38). In addition, job related stress has a negative impact on the health outcome of an individual although its impact appears to vary across the occupational classes (39). These findings are also reflected in our study also, in our study, we saw that most of our study subjects were domestic helpers (53%) and day labour (24.4%).
Diabetes, hypertension, obesity, and smoking are previously considered as risk factors for developing CKD (4,8,10). Our study, like previous studies, also found a strong association between CKD and diabetic group (P <0.01), between CKD and obese group (P = 0.00), CKD and hypertension group (P = 0.009), and CKD and smoking habit group (P = 0.03). Although unlike other studies (5), the association between hypertension and CKD after adjustments were not found significant, our adjusted data revealed that diabetes, smoking, obesity and hypertension increased the chance of developing CKD by 7.46, 2.36, 3.98 and 1.16 folds, respectively. Among the other confounding risk factors, our study did not find any significant association between CKD and alcohol consumption, between CKD and gender, and CKD and WHR.

Conclusion
In conclusion, we suggested that early screening for CKD among the general population was utmost important since around 18.2% subjects with CKD were found asymptomatic in our study. This early screening could be used to identify the people with risk of CKD and developing the disease. It will ultimately save more expenses and reduce burden of disease. Early detection of CKD should lead to treatment that prevents or delays the occurrence of significant morbidity and mortality. A country like Bangladesh with meagre number of nephrologists and limited resources, early prevention through screening can be the most valuable and effective strategy.

Limitations of the study
This investigation was conducted in a limited area, therefore large studies on this subject are suggested.