Obesity is common in chronic kidney disease and associates with greater antihypertensive usage and proteinuria: evidence from a cross‐sectional study in a tertiary nephrology centre

Summary Obesity is a treatable risk factor for chronic kidney disease progression. We audited the reporting of body‐mass index in nephrology outpatient clinics to establish the characteristics of individuals with obesity in nephrology practice. Body‐mass index, clinical information and biochemical measures were recorded for patients attending clinics between 3rd August, 2018 and 18th January, 2019. Inferential statistics and Pearson correlations were used to investigate relationships between body‐mass index, type 2 diabetes, hypertension and proteinuria. Mean ± SD BMI was 28.6 ± 5.8 kg/m2 (n = 374). Overweight and obesity class 1 were more common in males (P = .02). Amongst n = 123 individuals with obesity and chronic kidney disease, mean ± SD age, n (%) female and median[IQR] eGFR were 64.1 ± 14.2 years, 52 (42.3%) and 29.0[20.5] mL/min/BSA, respectively. A positive correlation between increasing body‐mass index and proteinuria was observed in such patients (r = 0.21, P = .03), which was stronger in males and those with CKD stages 4 and 5. Mean body‐mass index was 2.3 kg/m2 higher in those treated with 4‐5 versus 0‐1 antihypertensives (P = .03). Amongst n = 59 patients with obesity, chronic kidney disease and type 2 diabetes, 2 (3.5%) and 0 (0%) were prescribed a GLP‐1 receptor analogue and SGLT2‐inhibitor, respectively. Our data provides a strong rationale not only for measuring body‐mass index but also for acting on the information in nephrology practice, although prospective studies are required to guide treatment decisions in people with obesity and chronic kidney disease.


| BACKGROUND
Obesity, hypertension and type 2 diabetes mellitus (T2DM) constitute inter-related pandemics that have increased the prevalence of chronic kidney disease (CKD). 1 Diabetic kidney disease (DKD) and obesityrelated glomerulopathy (ORG) are the two main drivers of CKD in people with obesity. 2 DKD develops in approximately 40% of people with T2DM, with higher prevalence amongst non-Caucasian ethnicities. 3 DKD is associated with significant increases in cardiovascular and all-cause mortality, with the majority of excess cardiovascular and all-cause mortality attributable to diabetes occurring in those with kidney disease. 4 ORG is a distinct cause of CKD characterized by subnephrotic range proteinuria, glomerulomegaly and progressive renal functional loss. 2 In the absence of routine histopathological confirmation of CKD aetiology, the true prevalence of ORG is unknown, although 4% to 10% of people with obesity have significant proteinuria (≥1+ by urine dipstick or uACR ≥30 mg/mmol). 2 Reducing the severity of obesity with metabolic surgery decreases the incidence of albuminuria and end-stage kidney disease (ESKD) over long-term follow-up. [5][6][7] In a single-centre study of 105 individuals with type 2 diabetes and albuminuria who underwent gastric bypass surgery, median reductions in albuminuria of 80.7% were achieved over mean 13-month follow-up. 8 Postoperative reductions in proteinuria occur independently of improvements in blood pressure and metabolic control, suggesting that weight-and glycaemia-independent mediators may contribute to the renoprotective effects of the procedure. 9,10 GLP-1 receptor analogues (GLP1RAs) and sodium-glucose cotransporter-2 inhibitors (SGLT2is), which modify the course of DKD independently of their antihyperglycaemic properties, may improve CKD outcomes for people with obesity and/or T2DM. 11,12 In addition, integrated diabetology and nephrology care slows the progression of CKD in people with diabetes. 13 Given the increasing recognition of obesity as an important driver of the onset and progression of CKD in people with and without T2DM, 1,2 and recent advances in medical and surgical treatment approaches to obesity, 14,15 there exists a rationale to investigate the impact of multi-disciplinary clinics, which emphasize intentional weight loss on CKD progression and cardiovascular mortality in people with obesity. Prior to establishing such a clinic, we aimed to understand associations between obesity, CKD and other obesity complications in individuals with obesity attending general nephrology clinics at our tertiary referral centre. What is already known about this subject • Obesity is recognized to exacerbate proteinuria amongst people with chronic kidney disease. The magnitude of this relationship, and how it is influenced by gender and chronic kidney disease stage, is inadequately described.
In addition, whether or not a continuous relationship exists between increasing BMI and proteinuria amongst people with obesity and chronic kidney disease is unknown.
• Data on the incorporation of GLP-1 receptor analogues and SGLT2-inhibitors, which have metabolic, weight and renal benefits in people with type 2 diabetes, into real-world clinical practice for the management of high-risk patients with chronic kidney disease, are lacking.
• Outside of NHANES surveys in the United States, reports on the prevalence of obesity and associated comorbidity burden amongst people with chronic kidney disease are sparse. In addition, insights from such data are limited by the lack of granular information on physician-assigned diagnoses of chronic kidney disease aetiology.

What this study adds
• Obesity is over-represented in people with chronic kidney disease compared with background rates in the general population. Over 60% of the chronic kidney disease burden amongst people with obesity is attributable to diabetic kidney disease and hypertensive nephropathy; conversely, alternative chronic kidney disease aetiologies, including primary glomerular diseases, obstructive nephropathy and interstitial renal diseases, remain common in people with obesity and account for over 35% of chronic kidney disease cases in this setting.
• Amongst people with obesity and chronic kidney disease, increasing BMI associates with greater proteinuria and antihypertensive usage. The relationship between increasing BMI and proteinuria is stronger in males and those with advanced chronic kidney disease (stages 4 and 5). As a means of reducing the severity of hypertension and proteinuria, and consequently accelerated chronic kidney disease progression, intentional weight loss strategies should be explored in these particularly high-risk subgroups of people with obesity and chronic kidney disease.
• The majority of people with obesity, chronic kidney disease and type 2 diabetes in our cohort were treated with weight-promoting diabetes therapy (insulin and sulphonylureas), rather than newer therapies with weight-lowering effects as well as established end-organ handled as per the General Data Protection Regulation guidelines (EU), 2016/679. All procedures performed were in accordance with the ethical standards of the institutional audit committee and with the 1964 Helsinki declaration and its later amendments. Amongst people with obesity and CKD, cross-sectional relationships between BMI, Chronic Kidney Disease-Epidemiology Collaboration (CKD-EPI) estimated glomerular filtration rate (eGFR), urine protein-to-creatinine ratio (uPCR), and usage of antihypertensives and glucose-lowering medications are also reported. People with ESKD on haemodialysis or with a prior kidney transplant were excluded. Pregnant women, women in the post-partum period for less than 3 months and individuals who self-reported anabolic steroid use or with missing clinical and laboratory data were excluded.

| Clinical information
Body height and weight were routinely measured in nephrology outpatient clinics between 3 rd August, 2018 and 18 th January, 2019 using a Seca 701 electronic scale and Seca 220 stadiometer, respectively.
Body-mass index (BMI) (kg/m 2 ) was calculated: body weight (kg)/ (height × height [m 2 ]). 16 Only the first BMI measurement was considered for individuals that attended nephrology clinics on multiple occasions during the data collection period. BMI was categorized according to the World Health Organization criteria. 16 Clinical information (demographics, diabetes type, diabetes complications, office blood pressure, cardiovascular comorbidities and medication usage) was recorded from hospital outpatient and discharge records at or before the study entry date. CKD was defined according to 2012 KDIGO consensus criteria, including impaired glomerular filtration (eGFR <60 mL/min/BSA) or proteinuria sustained for ≥3 months, or abnormal renal histopathology or abnormal renal imaging. 17  and obstructive nephropathy were assigned only with radiological confirmation. Chronic pyelonephritis and NSAID-and lithium-related CKD were collectively categorized as interstitial renal disease. In the absence of supportive renal histopathological evidence, ORG was not assigned as a distinct CKD aetiology. Haemoglobin was measured on the Sysmex XE-5000 differential analyser (Sysmex Europe GmbH, Norderstedt, Germany). Urine protein was measured using turbidimetry with benzethonium chloride, traceable to a NIST standard. CKD-EPI eGFR, calculated using standard formulae, was recorded and expressed as mL/min/body surface area (BSA). 18

| Statistical analyses
RStudio version 3.6.1 was used for analysis. BMI distribution and prevalence of BMI categories at study enrolment were summarized by descriptive statistics. Categorical variables are presented as frequencies and percentages and were compared between groups using χ 2 tests. Fisher's exact tests were used to compare categorical variables between groups when χ 2 test assumptions were violated (frequency of values <5 in ≥1 cell of contingency table). Continuous variables with normal and skewed distributions are presented as mean ± SD and median [interquartile range], respectively. One-way between-group ANOVAs and Kruskal-Wallis tests were used to assess for differences across the BMI categories of obesity (three groups) in continuous variables with normal and skewed distributions, respectively. Independent sample t tests and Wilcoxon ranksum tests were used to assess for gender differences (two groups) in continuous variables with normal and skewed distributions, respectively. Univariate relationships between BMI, uPCR and HbA 1c were investigated by Pearson correlations. The functions "ggscatter" and "ggboxplot" from the R package "ggpubr" were used to generate scatterplots and boxplots, respectively. 19 P <.05 was considered statistically significant. benefits in the kidney (GLP-1 receptor analogues and SGLT2-inhibitors). In our study cohort reflective of realworld contemporary nephrology practice, obesity was infrequently addressed as a modifiable risk factor for chronic kidney disease progression.

| RESULTS
3.1 | Overweight and obesity are common amongst people attending nephrology clinics BMI was measured in 384 individuals over the 6-month study period; n = 10 people were excluded for: active haemodialysis treatment (n = 3), missing clinical and laboratory data (n = 3), pregnancy (n = 2), <3 months post-partum (n = 1), and anabolic steroid use (n = 1). Table 1 summarizes the demographic and anthropometric characteristics of the remaining 374 individuals included in downstream analysis.

| Clinical and laboratory characteristics of individuals with obesity attending nephrology clinics
In total, 132 people with obesity (BMI ≥ 30 kg/m 2 ) attended nephrology clinics during the study period. Nine patients did not have confirmed CKD. Of these, four were attending nephrology clinics for blood pressure management, two for evaluation of high serum creatinine, one for evaluation of proteinuria, one for nephrolithiasis and one for recurrent urinary tract infections. Table 2 summarizes the clinical characteristics of the remaining n = 123 individuals with obesity and CKD, stratified by obesity class. As observed in the cohort as a whole, males were more likely to have obesity class 1 while females were more likely to have obesity classes 2 and 3 (P = .04). The majority of the cohort had hypertension (87%) and dyslipidaemia (63%), while almost 25% had established coronary artery disease. Over 60% and 50% of people were treated with renin-angiotensin-aldosterone system (RAAS) blockade and statins, respectively. Overall, 50 (40.7%) T A B L E 1 Baseline characteristics of the study cohort (n = 374) 3.3 | Increasing BMI associates with higher uPCR and greater antihypertensive medication usage in individuals with obesity and CKD

| DISCUSSION
This study provides insight into the BMI distribution of males and females attending general nephrology clinics. Obesity (35.3%) was common in outpatient nephrology practice. Obesity class 1 was more common in men, while normal weight and obesity classes 2 and 3 were more frequently observed in females. Amongst individuals with obesity and CKD, a positive correlation was observed between increasing BMI and uPCR, a relationship, which was stronger in males and those with more advanced CKD stages (eGFR <30 mL/min/BSA). BMI was higher in individuals treated with 4-5 compared with 0-1 antihypertensives. Individuals with obesity, CKD and T2DM were frequently prescribed RAAS blockade, statins, insulin and sulphonylureas, but infrequently prescribed diabetes therapy with weight-lowering effects, including GLP1RAs and SGLT2is.
Obesity was more common in people with CKD than expected based on trends in the general Irish population. In the 2015 Healthy Ireland survey of 6,142 people aged ≥15 years, the prevalence of obesity was 23%. 20 An additional 12% of people attending nephrology clinics in Ireland are affected by obesity compared with the general population. The 2015 Healthy Ireland survey also identified that compared with men, women are more likely to be of normal weight (men: 31%, women: 44%) and less likely to be affected by overweight (men: 43%, women: 31%) or obesity (men: 25%, women 22%). 20 Findings on gender differences in BMI distribution in our study thus mirror background patterns in the Irish population. 20 Globally, obesity is more common in women than men and the majority of patients recruited to studies of metabolic surgery as a treatment for obesity have been female (approximately 70%). [22][23][24] The prevalence of obesity amongst people with CKD is even  One-way between-groups ANOVA was used to assess for variation in normally distributed continuous variables across obesity classes. b χ 2 analysis or Fisher's exact test was used to analyse for differences in categorical variables across obesity classes.
c Kruskal-Wallis test was used to assess for variation across obesity classes in continuous variables that were not normally distributed.
and present in over 85%, 60% and 50% of patients, respectively. No major differences in cardiovascular comorbidities or CKD stage were observed across incremental categorical classes of obesity. This may be partly related to the relatively low number of patients (n = 18) with obesity class 3 (BMI ≥ 40 kg/m 2 ) and CKD. Obesity-related glomerular hyperfiltration, which is an independent predictor of adverse cardiovascular outcomes, and the obesity paradox may have contributed to the high burden of cardiovascular disease even amongst individuals with less severe obesity (WHO classes I and II) in our cohort. 27,28 However, a linear relationship between BMI and risk of ORG does not exist in people with obesity. Almost half the total number of biopsyproven cases of ORG occurred in those with BMI <40 kg/m 2 in a single-centre series from the United States. 29 Thus, as the complication burden amongst people with obesity and CKD is similar across obesity classes, and obesity classes 1 and 2 are much more prevalent than obesity class 3, it is plausible that the maximum benefit of intentional weight loss strategies in nephrology practice may be achieved in those with obesity classes 1 and 2. Indeed, the Microvascular Outcomes after Metabolic Surgery study, which randomized individuals with T2DM and microalbuminuria to medical therapy or medical therapy plus metabolic surgery, selectively recruited individuals with class 1 obesity. 30 Further prospective studies of intentional weight loss strategies across the spectrum of obesity severity in people with CKD are required.
Despite the lack of observed difference in CKD severity across obesity classes 1 to 3 in our cohort, when BMI was treated as a continuous variable, a significant positive correlation between increasing BMI and increasing uPCR was observed amongst people with obesity and CKD. Importantly, this association was stronger in males and those with CKD stages 4 and 5. Mean BMI was also 2.3 kg/m 2 higher Only two patients were treated with a GLP1RA and none were treated with an SGLT2i. The current study was conducted prior to the publication of the CREDENCE trial, 12 which will likely increase the proportion of such patients treated with SGLT2is in our practice.
Nevertheless, median eGFR in our study cohort was 29 mL/min/ BSA; SGLT2is are not yet licensed in this setting due to concerns that they may lack glycaemic efficacy in those with eGFR <45 mL/min/ BSA. 41  Misclassification of individuals as being affected by obesity using BMI as the sole measure of body composition is well described. 48 This phenomenon may be more common in people with CKD as the accompanying changes in extracellular fluid volume alter body weight independently of adiposity. 49 We excluded certain individuals from our cohort, including women during pregnancy and the post-partum period, to minimize the impact of such misclassification on our study findings. Differences in the burden of cardiovascular comorbidities and CKD severity across obesity stages may have been more readily apparent if patients were stratified on a measure of abdominal adiposity. Waist circumference and/or waist-to-height ratio should be measured in future studies to assess the impact of obesity in nephrology practice. Given the aforementioned changes in fluid status in CKD, accurate delineation of lean and fat mass using DEXA scanning or bioimpedance should also be considered. 50

| CONCLUSIONS
In conclusion, obesity was common amongst individuals attending nephrology clinics. Class 1 obesity was more common in males; normal weight and obesity classes 2 and 3 were more frequently observed in females. The severity of obesity did not significantly T A B L E 3 Medical therapy of individuals with obesity (BMI ≥ 30 kg/m 2 ), type 2 diabetes mellitus and chronic kidney disease attending nephrology clinics, stratified by obesity class (n = 59) Abbreviations: ACE, angiotensin-converting enzyme; ARB, angiotensin-II receptor blocker; DPP4i, dipeptidyl peptidase-4 inhibitor; GLP1RA, glucagon-like peptide-1 receptor analogue; IQR, interquartile range; N/A, not applicable; SD, standard deviation; SGLT2i, sodium-glucose co-transporter-2 inhibitor. a One-way between-groups ANOVA was used to assess for variation in normally distributed continuous variables across obesity classes. b χ 2 analysis or Fisher's exact test was used to analyse for differences in categorical variables across obesity classes.
c Kruskal-Wallis test was used to assess for variation across obesity classes in continuous variables that were not normally distributed.
influence the prevalence of cardiovascular complications or CKD stage amongst people with obesity and CKD. Increasing BMI associated with increasing proteinuria amongst people with obesity and CKD, a relationship which was stronger in males and those with CKD stages 4 and 5. Increasing BMI also associated with greater antihypertensive usage in patients with obesity and CKD. The majority of people with obesity, T2DM and CKD were treated with medications that promote weight gain, including sulphonylureas and insulin; GLP1RAs and SGLT2is were infrequently prescribed in this setting. Prospective studies of medical and surgical intentional weight loss strategies in people with obesity and CKD evaluating renal, cardiovascular, and mortality outcomes are warranted.