Prevalence, Incidence and Risk Signature of Chronic Kidney Disease in Sub-Saharan Africa

Overview

Chronic kidney disease (CKD) is associated with increased cardiovascular morbidity and mortality. The prevalence of CKD is increasing worldwide and is assumed to also dramatically increase in Sub-Saharan Africa (SSA). Key shortcomings of available data on CKD in SSA are as follows: (i) Available data are based on single measurements and, therefore, cannot distinguish between harmless transient deterioration in kidney function and chronic kidney damage; (ii) Accurate information regarding renal protein loss, an important and early marker of kidney disease, is lacking; (iii) Cardiovascular risk factors for CKD, such as obesity, hypertension and diabetes, are often not searched for. Likewise non-classic potential risk factors, such as endemic infectious diseases, socioeconomic status and lifestyle have not been consistently recorded; (iv) Information to interrogate linked interaction over time between risk factors and development of CKD is unavailable. With this project, situated in a region representative of semi-rural SSA, we aim to fill this knowledge gap and (i) establish guideline conform prevalence data of CKD and its major cardiovascular risk factors, as well as (ii) prospectively define the incidence of cardiovascular- and non-classic risk factors of CKD. The data from (i) and (ii) is used to develop predictive models. A prospective cohort of 1200 individuals in a primary care facility will serve as study population. The population is representing a society in transition from rural to more urban lifestyle. In the pilot study, participants will be followed for one years and undergo the clinical and biomedical testing required to capture CKD and its classic and non-classic risk factors over time.

RenalTWO is a prospective cohort study collecting clinical, biological, lifestyle, and socioeconomic data. The study is systematically enrolling walk-in patients in a primary care setting from the Bagamoyo district hospital (BDH) catchment area attending the outpatient clinic (OPC) or the Fukayosi or Yombo dispensary. Patients are included if the patient is ≥ 18 years age, of African descent, living in the BDH catchment area, not presenting with acute trauma, and written consent or fingerprint in case of illiteracy is given. Three visits are planned: baseline, ≥ 90 days confirmation visit, and 1-year follow-up. Inclusion is initiated over a period of approx. 6 months, starting mid June 2023. Medical history of each participant will be taken, including cardiovascular diseases (cerebrovascular stroke, myocardial infarction, peripheral arterial occlusive disease, diabetes, AHT), history of prostate hyperplasia/obstructive disorders of the urinary tract, cancer, sickle cell disease, asthma, cold, endemic infection diseases (tuberculosis, malaria, schistosomiasis, filariasis) as well as common infections (hepatitis, respiratory tract infection, tonsillitis, urinary tract infection, vaginal discharge syndrome, skin infection). Further, the reason for attending the primary care facility as well as the main symptoms are recorded. To assess socio-economic backgrounds and information on lifestyle and nutrition, adapted questionnaires on the basis of WHO Stepwise approach are used. To gain insights into physical activity patterns, patients are invited use a step counter for one week. Each patient will undergo an assessment of vital signs, including height, weight, hip and waist circumferences, heart rate, and a measurement of office BP, according to 2020 International Society of Hypertension Global Hypertension Practice Guidelines. The following biological data are collected: Random or fasting blood glucose, glycated HbA1c, CRP, lipids, serum creatinine, cystatin C, full blood cell count, HIV urinary, albumin-to-creatinine ratio (ACR), urine dipstick test measures, alpha-microglobulin, and urine sediments in patients with CKD. To assess kidney morphology and size, POC kidney sonography is carried out. The data collection process is electronically, based on standardized questionnaires programmed using REDCap software. Surveys assessing socio-economic backgrounds, lifestyle, nutrition and medical history are translated to Kiswahili, the predominant language, and back translated for quality check. To minimise a selection-bias, the study will recruit approx.10% of the study population across two separate representative rural communities from the BDH catchment area. The estimated glomerular filtration rate (eGFR) is calculated using the CKD-EPI 2021 formula. CKD is defined and staged according to KDIGO guidelines, including albuminuria stages A1-A3 and GFR classification G1-G5. To define CKD according to KDIGO guidelines, stage A2 albuminuria (ACR \> 30mg/g) and/or eGFR \< 60ml/min/1.73m2 (G3b) are the respective cut-off values. CKD is considered being confirmed if stage A2 and/or G3b is measured twice with a minimum interval of 90 days. For statistical analyses, R software will be used. For descriptive statistics, discrete variables are expressed as counts (percentage) and comparison between groups is performed using the Pearson's Chi-square test or Fisher's exact test. Continuous variables are expressed as mean ± standard deviation (SD) if normally distributed, or as median and inter-quartile-range (IQR), if not normally distributed. The T-test or Mann-Whitney test are used for comparisons between groups. Power considerations: To estimate prevalence data with a 95% CI and using multivariate models, approximately 1000 patients are required assuming a CKD prevalence of 7-14%. However, the confirmation requirement criterium imposes a more conservative CKD prevalence estimate, and the loss of follow-up dictates an extra 10-20% of patients to be enrolled at baseline. Thus, the study aims to recruit 1200 patients. For multinomial outcomes, such as KDIGO-defined CKD risk groups, the Sison-Glaz method is used to calculate prevalence and corresponding 95% CI estimates. To develop predictive models for CKD outcome (albuminuria stage A2 and/or a GFR classification G3b), clinical, biological and lifestyle data will be tested as predictors. In an explorative approach, logistic regression and penalised regression is applied. For penalised regression, sparse principal component analysis (PCA), and further LASSO and ridge regression will be applied. Models will be internally validated 100 times, using 10-fold cross validation. Based on the optimal validation of these models, probability of CKD will be estimated dependent on the regression parameters, and graphically presented as nomograms. This will be providing the basis to further program digital applications for interactive tools visualizing risk signatures at a population and patient level. With longitudinal data, Markov multi-state transition modelling will be applied to predict disease trajectories and factors influencing change of disease state. Primary Objective: i) Prevalence of chronic kidney disease (CKD) and cardiovascular and non-classic risk factors of CKD Biological and clinical assessed: Albuminuria (quantitative albumin-creatinine-ration, dipsticks), kidney function, (creatinine, cystatin C, eGFR), CKD stages (KDIGO), arterial hypertension (standardised BP measurements), diabetes mellitus, pre-diabetes (random-, fasting blood glucose, HbA1c), dyslipidemia (lipid panel), BMI (weight in kg/height in meters squared), anaemia (haemoglobin), inflammation (CRP, leucocytosis), HIV, physical activity (step count). Questionnaire assessed: Malaria, tuberculosis, schistosomiasis, filariasis, hepatitis, respiratory tract infection, tonsillitis, urinary tract infection, pelvic discharge syndrome, prostatic obstructive syndrome, skin infections, cerebrovascular stroke, myocardial infarction, peripheral arterial disease, asthma, cold, heart burn, sickle cell disease, cancers, socio-economic background, lifestyle. ii) Incidence of chronic kidney disease (CKD) and cardiovascular- and non-classic risk factors of CKD: as described above iii) Integrating data from (i) and (ii) to interrogate the linked interaction over time between risk factors and development of CKD, thus developing an SSA-specific risk model: as described above Secondary objectives (non-hierarchical): i) Longitudinal assessed concordance and usefulness of glycated hemoglobin A1c (HbA1c) in patients with anaemia ii) Validation of urinary dipsticks in diagnosis of albuminuria in a setting with repeated measurement iii) Difference of eGFR calculations and their impact on CKD classification in a pre- dominant population of black Africans iv) Evaluation of kidney size as a predictor for CKD v) Evaluation of single point versus repeated biological data collection in accurate diagnosis of CKD vi) Evaluation of the quality of the single point versus repeated data collection on lifestyle behaviour and profiles as well as socioeconomic backgrounds.