Clustering of chronic kidney disease and cardiovascular risk factors in South-West Nigeria

Background: There exists a synergy between chronic kidney disease (CKD) and cardiovascular risk factors (CVRFs) with increased morbidity and poor outcomes. Objectives: Data relating to this clustering in black homogenous populations is scanty. We aim to investigate this relationship in Nigerian communities. Patients and Methods: It was a cross-sectional observation study from semi-urban communities in South-West Nigeria. We used modified World Health Organization (WHO) questionnaire on chronic diseases (WHO STEPS) to gather information on socio-demographic data, biophysical and clinical characteristics. Biochemical analysis of plasma samples was done. Results: We analyzed data of 1084 with mean age of 56.3 ± 19.9 years (33.4% female). Prevalence of stage 3 CKD was 14.2% (3a and 3b were 10.3% and 3% respectively). Prevalence of hypertension (systolic and diastolic blood pressure) and low high-density lipoprotein cholesterol (HDL-C) increased as clustering of cardiovascular (CV) risk factors (CVFRs) increased both in CKD and proteinuria (P < 0.05). CKD prevalence increases with number of risk factors. There was an inverse relationship between increasing risk factors and mean estimated glomerular filtration rate (eGFR) (P < 0.05). Clustering at least 2 CVRFs in the population with CKD compared to those without CKD was significantly higher (76.6% vs. 65.1%, OR: 1.8, 95% CI: 1.2-2.6, P = 0.005). Similarly, in a univariate analysis, albuminuria had an increased odds of clustering (69.7% vs. 59.6%, OR: 1.9, 95% CI 0.6-6.2, P = 0.409). Using multivariate logistic analysis, there is significantly increased odds of clustering when eGFR is <45 mL/min/1.73 m2 (OR: 2.66, 95% CI: 1.12-6.32) and microalbuminuria 1.74 (95% CI: 1.10-2.75). Conclusions: Reduced kidney function and proteinuria significantly clustered with CVRFs. This data suggests that individuals with CV clusters should be screened for CKD or vice versa and they should be considered for prompt management of their CVRFs.


Background
Chronic kidney disease (CKD) is an important public health concern with worse outcomes especially among blacks (1,2). There exists a synergy between CKD and cardiovascular (CV) risk factors resulting in increased co-morbidities, hospitalizations and mortality (3,4). Reduced estimated glomerular filtration rate (eGFR) is an independent risk factor for increased incidence of CV diseases (5,6) and is thought to be related to increased concentrations of inflammatory markers, systemic metabolic disorders, loss of night-time blood pressure dipping (7,8) and increasing left ventricular hypertrophy (9). Also, albuminuria is strongly independently associated with CV diseases resulting in increased intima-media thickness, left ventricular remodelling and myocardial ischaemia (10). It is a reflection of endothelial dysfunction and presents as early sign of nephropathy contributing additional prognostic dilemma to reduced eGFR, CV burden and reduced life expectancy. The severity of albuminuria and glomerular function has incremental effects on the poor outcomes of patients with CKD. CV diseases resulting from the interplay of these pathophysiological changes activate early aging of the CV system in a vicious cycle and account for most CKD-related deaths (11). The increased incidence of CKD and CV disease in individuals of black ethnicity has been established but the interplay between CKD and CV disease has not been well studied. There is paucity of information from Nigeria and other low-income countries in spite of the burden of CKD, risk factors and associated challenges. Resource limited countries, such as Nigeria, would therefore benefit from establishing good preventive and interventional policies if 'clustering' of CV disease and its risk factors is associated with CKD.

Objectives
The aim of this study was therefore to determine the burden of CV risk factors and their clustering with CKD.

Population selection
The data presented is from a cross-sectional study in 10 communities in Ekiti and Osun States both in South-western Nigeria. Ten communities from Ekiti North and Central senatorial districts were chosen based on their proximity to our hospital.

Instruments
The Modified World Health Organization questionnaire on chronic diseases (WHO STEPS) was used to gather information on socio-demographic data, biophysical and clinical characteristics. The instrument was administered by trained health professionals in the language (principally Yoruba) best understood by participants. The study was conducted over a 6-month period (January 2013 and May 2013).

Sample analysis and measurements
Urine samples were checked for proteinuria with COMBI 10 strips (Medi-Test Combi 10 SGL) and those that tested negative were checked for microalbuminuria using semi-quantitative methods of estimation with a Clinitek R analyser (Bayer HealthCare LLC) and Microalbumin 2-1 Combo strip of Teco Diagnostic USA. The latter was used when there was difficulty getting more of the Clinitek reagent strips. Blood samples were taken for serum creatinine estimation using the Jaffe reaction to analyse samples. CKD-MDRD 4-variable formula was used to estimate GFR using serum creatinine determined from blood samples collected. Samples were also analysed for lipids: total cholesterol (TC), high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C) and triglycerides (TG). Data regarding CV risk factors such as old age (≥60 years), hypertension, diabetes, elevated TC, HDL-C and LDL-C cigarette smoking, waist circumference (WC) and body mass index (BMI) were obtained. The number of risk factors present was counted and 'clusters' grouped if <2, 2, 3 or ≥4 risk factors were present. We further used some of the variables (systolic blood pressure [SBP], HDL-C, TC, smoking, age and gender) to calculate the group averages of pooled cohort equations (PCE) for cohorts with CKD as measured by eGFR <60 ml/min/1.73 m 2 , albuminuria. PCE measured the global CV risk and predict 10-year risk of atherosclerotic CVD (fatal and non-fatal stroke, non-fatal myocardial infarction or coronary death) (12). Elevated risk is reported if PCE was ≥7.5%. Hypertension was defined as SBP ≥140 mmHg and/or diastolic blood pressure (DBP) ≥90 mm Hg or use of blood pressure medications. Diabetes was defined as fasting blood sugar 126 mg/dL and non-fasting blood sugar ≥200 mg/dL or on medications for diabetes.
Methods of measurement of hypertension, blood sugar and indices of anthropometry were previously reported (13) were considered as cluster. Chronic Kidney Disease-Epidemiology Collaboration (CKD-EPI) formula was used to estimate kidney function.

Ethical issues
The research followed the tenets of the Declaration of Helsinki. We sought consent in each community through meeting with the community heads majority of who were Royal Fathers before the commencement of screening exercises in each community. Each participant signed consent form before they were admitted into the study. Ethics approval was also received from Ladoke Akintola University of Technology Teaching Hospital's research and ethics committee.

Data analysis
Data was analysed using SPSS version 20 (SPSS Inc., Chicago IL, USA). Standard descriptive statistics were used to examine baseline demographic characteristics, clinical and laboratory values. Descriptive statistics was used for continuous variables and summarized as mean ± standard deviation (SD). Categorical variables were reported as frequencies and proportions. Chi-square test was used to analyse associations in discrete variables. We examined the association between eGFR and albuminuria as features of CKD and CVRFs. Using univariate analysis, we also examined the relationship between these features of CKD and clusters of CV factors as 2, 3 and ≥4. We then determined the association and interaction effect of CKD as defined by eGFR and/or albuminuria, age and gender from univariate analysis as independent variables and clustering of at least two risk factors as dependent variable using multivariate logistic statistical analysis to obtain odds ratio. Level of significance was P < 0.05.

Discussion
This cross-sectional study from Nigeria has shown an association of CV risk factor clusters with CKD. Age, abdominal obesity by WC but not BMI, hypertension, and low HDL were significantly associated and clustered with CKD as measured by eGFR and albuminuria. Lemieux et al (14) described that 'hypertriglyceridaemic waist' has poor cardiometabolic profile and renal-related diseases. In Framingham study, low HDL-C co-existing with CKD was found to have a significant predictability of CVD (15). . Interestingly, the prevalence of low HDL-C has been reported to be high among Nigerians (16,17) and there may be need for further exploration of this     (18,19). A review by Stanifer et al also put prevalence of CKD at 12.4% in the urban and 16.9% in rural settlements of Sub-Saharan Africa (20). The high prevalence of proteinuria is ascribed to high prevalence of glomerular diseases and the various reports that blacks are at increased risk of proteinuria than Caucasians with CKD (21,22). There are reports of increased variants of APOL1 among blacks; Peralta et al (23) observed an increased excretion of albumin among this cohort with two alleles of these genes. Whether this could be generalized remains to be answered. CV profile of proteinuric cohort in our study is different from those with reduced eGFR. Proteinuria compared to reduced eGFR is clustered with majority of the traditional CV risk factors. Proteinuria with reduced eGFR portends increased CV risk factors and poor CV outcomes (24). Prevalence of clustering of CVRFs is higher among the CKD and proteinuric cohorts than general population studied. Reduced kidney function has increased odds of 1.8 to cluster with other risk factors. Increasing number of risk factors has an inverse relationship with eGFR. In ARIC study, CKD is an independent risk factor for CV diseases (25). Framingham heart study (15) and Muntner et al (26) reported a strong bond between CKD and other CV risk factors but not with CVD. Weiner et al (27), in a meta-analysis of pooled community studies reported CKD as an independent risk factor for CVD and outcomes. They found a significant association between CKD and CVD among black cohort. We showed that the presence of CKD using GFR <60 mL/min/1.73 m 2 only as well as proteinuria alone increased the risk of clustering in a univariate analysis. We further demonstrated that CKD 3b has an independent increased odd of 2.7 with clustering of CVRFs. Likewise, microalbuminuria had increased odds of clustering of 1.7 and macroalbuminuria showing a positive trend. This independence of reduced GFR and albuminuria was also reported by van der Velde et al (28). Fox et al in a meta-analysis of over a million populations reported high risk of CVD in reduced kidney function and proteinuria regardless of the presence of importantly established risk factors (5,6). Risk of clustering of CV factors with moderate to severe CKD (3a and 3b) according to KDIGO was three-fold and two-fold respectively higher than cohort with normal GFR. Parikh et al (15) in Framingham study, gave a similar report of an association between CKD 3b and CV diseases, but Foster et al (29) failed to establish the consistent association of CKD and CV risk factors. Plausible reasons are the estimating eGFR equation (CKD-EPI in our study) used which has been found to predict CV events more than MDRD equation (30), racial differences in the CV profiles and markers of CKD as shown in our study.

Conclusions
Reduced kidney function and proteinuria significantly clustered with CVRFs. This data suggests that individuals with CV clusters should be screened for CKD or vice versa and they should be considered for prompt management of their CVRFs. The possible genetic composition cum high prevalence of other CV risk factors such as hypertension in Nigeria and African population in general, may explain this clustering. And if this is the case, it is a call for urgent attention. The ongoing prospective study by the Human Heredity and Health (H3) in Africa Kidney Disease Research Network is expected to unravel some of these genetic conundrum associated with risk of CKD and CV risk factors in black population (31).

Limitations of the study
In spite of these significant findings, our study is limited by its cross-sectional nature as the causal relationship could not be established. We therefore suggest longitudinal studies in a larger population. Other limitations were the inability to repeat the urine dipstick tests which could have helped to confirm the proteinuria in keeping with NKFKDQI definition of CKD. However, the strength of this study lies in the homogeneity of the black populace in the communities. This would bridge the gap as regards clustering of CVD risk factors in CKD.

Conflicts of interest
We declare no conflict of interest in getting this work done and published.