A High Body Roundness Index Is More Closely Associated With Urinary Albumin-creatinine Ratio Than Traditional Adiposity Indices in Chinese Population: a Cross-sectional Report From the Reaction Study

Objective: Body roundness index (BRI) has been proposed to be a promising index of body fat distribution. The association between BRI and increased albuminuria is unclear. Therefore, the aim of this study is to examine the discriminative ability of BRI for increased albuminuria and compare the association of BRI and traditional adiposity indices (body mass index: BMI, waist circumstance: WC, waist-to-hip ratio: WHR) with albuminuria. Methods: This cross-sectional study was nested in an ongoing REACTION study. A total of 43591 participants aged over 40 years were recruited across seven different regional provinces. Increased albuminuria was dened as urinary albumin to creatinine ratio (UACR) ≥ 30 mg/g. Multiple logistic regression analyses were performed to detect the association between BRI and UACR and compare the discriminative ability of traditional (BMI, WC, WHR) and new adiposity indices (waist-to-height ratio: WHtR, BRI) with UACR. Results: Participants with increased UACR exhibited increased age, blood pressure, blood glucose, poor control of lipid level, decreased estimated glomerular ltration rate (eGFR)(cid:0)and higher prevalence of diabetes, hypertension and cardiovascular events. Multiple logistic models showed that compared with traditional adiposity indices (BMI, WC, WHR), the BRI index remained signicantly associated with UACR especially in women, indicating the strong discriminative power for increased albuminuria. Stratied analysis revealed that the strong positive association of the BRI index with increased albuminuria also occurred in people who were young, women, poor control of hypertension and blood glucose, and eGFR ≥ 90 ml/min/1.73². The BRI were closely with albuminuria in the Chinese


Introduction
Elevated urinary albumin to creatinine ratio (UACR) has been recommended as an indicator of chronic kidney disease (CKD), 1 and a key risk factor for many chronic diseases. Romero-Aroca et al. indicated that compared with the estimated glomerular ltration rate (eGFR), UACR shows a stronger association with diabetic retinopathy. 2 Moreover, UACR can not only be predictive of cardiovascular disease (CVD), but also has important value in evaluating the left ventricular dysfunction. 3 A recent study has demonstrated that UACR is closely associated with preeclampsia and could be used as an early predictor of preeclampsia in the pre-gestational diabetic population. 4 Furthermore, UACR has been con rmed associated with vascular damage. Not only in diabetic population, but also in patients with acute lacunar infarction, increased UACR is a signi cant indicator for cerebral small vessel diseases. 5 A growing body of evidence have reported that obesity is a key driver of CKD, 6,7 and increased UACR as an early indicator of CKD could effectively re ect the injury of glomerular and tubular cell function.
Recently, body mass index (BMI), an easily applicable measure of obesity, has been reported to be closely associated with CKD progression in hypertensive patients. 8 Framingham study cohort has suggested that high BMI could be independently predicting of the renal dysfunction, 9 and it is valuable in predicting the onset of albuminuria. 10 However, several studies has demonstrated that it is the accumulation of visceral fat rather than subcutaneous fat that promotes the development of CKD. [11][12][13] Although BMI, waist circumstance (WC) and waist-to-hip ratio (WHR) are commonly used adiposity indices with the advantages of convenience, they provided limited information on the body fat distribution and body shape. [14][15][16] Thus, it is important to nd a simple and effective screening tool for albuminuria to reduce the incidence of CKD.
In recent years, a new anthropometric index, body roundness index (BRI), have been proposed as an indicator of body fat distribution and body shape, because it has a strong and stable relationship with the abdominal adipose accumulation. BRI, a predictor of body fat and visceral adipose tissue volume, has been con rmed to have the potential to identify the status of diabetes. [17][18][19] It is well known that diabetes is an additional cause of CKD and obesity is the main pathogenetic pathway linking diabetes and CKD. However, the relationship between fat distribution and albuminuria is still unknown.
To the best of our knowledge, there are limited studies involving the relationship between BRI and albuminuria and it is still unclear that whether traditional adiposity indices or newly developed adiposity indices are more valuable in estimating the renal dysfunction. Therefore, the aim of this study is to examine the association between BRI with UACR and compare the discriminative ability of the traditional and new anthropometric indices for UACR.

Study population and design
This cross-sectional study was nested in a longitudinal REACTION (Risk Evaluation of Cancers in Chinese diabetic Individuals) study, which was designed to investigate the association between T2DM and the risk of cancer among the Chinese population. Details of the REACTION study have been previously reported. 20,21 The present study used baseline investigation data from seven centers across China. Initially, a total of 45130 participants aged over 40 years were recruited from May 2011 to December 2011. Participants diagnosed with kidney or other related diseases, those using ACEI/ARB medicines and those with missing data were excluded as shown in Figure 1. Finally, 43591 participants were enrolled in the present study.
Before carrying out the investigation, the staff received extensive training, including the standardized questionnaire and data collection. The protocol of present study was approved by the Committee on Human Research at Rui-Jin Hospital a liated with the School of Medicine, Shanghai Jiao Tong University. This study adhered to the principles of the Declaration of Helsinki. Written informed consents were obtained from all participants before data collection.

Data collection
Data collection were performed by the same trained staff, according to standardized operational procedures. All the participants received the same comprehensive examinations including a standard questionnaire, anthropometric measurements, venous blood collection, and 75-g oral glucose tolerance test (OGTT) or bread meal test. The questionnaire included the history of diabetes, hypertension, acute/chronic nephritis, nephritic syndrome, kidney stones, CVD, diabetes, alcohol habits, and smoking habits. Regular smokers were de ned as those who smoked at least one cigarette per day. Occasional smokers were participants who smoked less than one cigarette per day or less than 7 cigarettes per week.
Regular drinkers were de ned as participants who consumed alcohol at least once a week for over six months. Occasional drinkers were de ned as participants who drank less than once a week.
Height, weight, WC and hip circumstance (HC) were measured by the same well-trained staff after participants were required to wear light clothing and take off the shoes. Height and weight were clinically measured in light clothing, using the same device with the precision of 0.01 m and 0.1 kg, respectively. WC was measured by the same staff placing a tape horizontally between the inferior costal margin and the superior border of iliac crest at the end of expiration and that was measured to the nearest 0.01m. HC was recorded at the maximum circumference over the buttocks, to the nearest 0.01m.
After resting for ve minutes, participants' blood pressure and pulse were measured three times with in one-minute intervals by the same staff. The average of blood pressure was calculated and used in the statistical analysis. The pulse rate was measured while the blood pressure was recorded. After 8-10 of fasting overnight, the rst fasting blood samples of all the participants were obtained. Patients without a history of T2DM underwent a 75 g OGTT; they were required to drink 300 mL of a glucose solution containing 75 g of glucose within 5 minutes. After 2 hours, the second venous blood sampling was obtained by the same well-trained staff. Fasting blood glucose (FBG), 2 h post-load blood glucose (PBG), serum triglycerides (TG), total cholesterol (TC), high-density lipoprotein cholesterol (HDL), low-density lipoprotein cholesterol (LDL), and serum creatinine (Cr), Haemoglobin A1c (HbA1c), alanine transferase (ALT), aspartate transferase (AST), and gamma-glutamyl transferase (GGT) were respectively measured in every center.
The variables were de ned as follows: hypertension (any self-reported history of hypertension or systolic blood pressure SBP ≥ 140 mmHg or diastolic blood pressure DBP ≥ 90 mmHg), T2DM (FBG ≥ 7.0 mmol/L or PBG ≥ 11.1 mmol/L simultaneously or any self-reported history of diabetes), cardiovascular events (any self-reported history of coronary heart disease, stroke, and myocardial infarction). According to the WHO criteria, prediabetes was de ned as follows: 6.1 ≤ FBG < 7.0 mmol/L or 2h PBG < 11.1 mmol/L. Based on WHO criteria, prediabetes was further divided into 3 groups as follows: impaired fasting glucose (IFG): 6.1 ≤ FBG < 7.0 mmol/L and PBG < 7.8 mmol/L; impaired glucose tolerance (IGT): FBG < 6.1 mmol/L and 7.8 ≤PBG < 11.1 mmol/L; and IFG+IGT: 6 1 ≤ FBG < 7.0 mmol/L and 7.8 ≤ PBG <11.1 mmol/L. eGFR was calculated based on the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI). 22 De nition of UACR group and anthropometric indices The concentration of urine albumin and creatinine were measured by collecting the rst urine specimens in the morning. The de nition of UACR was calculated using the following formula: urinary albumin (mg)/urinary creatinine (g). All seven centers used the same normal value range and unit of measurement as the following, normo-albuminuria:<30 mg/g; increased albuminuria ≥30 mg/g. BMI was calculated using the formula: BMI = weight kg /height m 2 . WHtR was calculated as the WC divided by the height. ABSI and BRI were calculated using the following formula: BRI = 364.2-365.5×(1 - The BRI score was converted to a z-score using the following formula: BRI minus BRI mean divided by BRI SD . BMI, WC, WHR, WHtR were also transformed to z-score by the same equation. Given difference of the unit change in the logistic regression analysis, the z-score were conducted to made the comparison meaningful. 23
Data were presented as median (25th percentile-75th percentile) for continuous variables of non-normal distribution. Category variables were presented as percentage (%). Differences in the continuous variables among the two subgroups of UACR were analyzed using the Kruskal-Wallis test. The category variables were tested using the chi-square test. Multiple logistic regression analysis was conducted to detect the association of traditional and new adiposity indices with UACR. The conventional risk variables related to renal dysfunction were adjusted, 24,25 and confounding factors that when added to the model, changed the matched odds ratio (OR) by at least 10% were also selected for adjustment. Model I was adjusted for age and sex. To further correct the effects of the confounding factors, model II was adjusted for centers, age, sex, BMI, SBP, DBP, FBG, PBG, HbA1c, HDL, LDL, TG, GGT, eGFR, pulse, centers, smoking status, drinking status, and medication history. In order to thoroughly investigate the association and compare the difference in different levels of age, sex, eGFR, the control of blood pressure and glucose, strati ed analyses were performed based on the subgroups of age (Age < 55 years, 55-64 years, ≥ 65 years), sex (men, women), the subgroup of eGFR (eGFR≥ 90 mL/min per 1.73 m 2 , eGFR < 90 mL/min per 1.73 m2), blood glucose (normal: FBG <7.0 mmol/L and/or PBG < 11.1 mmol/L, prediabetes : IFG: 6.1 ≤ FBG < 7.0 mmol/L and PBG < 7.8 mmol/L, IGT: FBG < 6.1 mmol/L and 7.8 ≤ PBG < 11.1 mmol/L, IFG + IGT: 6.1 ≤ FBG < 7.0 mmol/L and 7.8 ≤ PBG < 11.1 mmol/L,diabetes: FBG ≥ 7.0 mmol/L and PBG ≥ 11.1 mmol/L or any self-reported history of diabetes), blood pressure (normal blood pressure SBP <140 mmHg and diastolic blood pressure DBP < 90 mmHg, hypertension SBP ≥ 140 mmHg or DBP ≥ 90 mmHg or any self-reported history of hypertension).
Associations between adiposity indices and the risk of increased UACR in total population Table 2 shows the signi cant association between all adiposity indices and UACR in the total population in Model I. After further adjusting for confounding factors in model II, the association between newly developed adiposity indices (WHtR and BRI) remained signi cant, indicating the stability of the association between WHtR, BRI and increased UACR (WHtR: OR: 1.08, 95% CI 1.02-1.13 P=0.0075; BRI: OR: 1.22, 95% CI 1.11-1.33 P=0.0001). However, the signi cant association between BMI, WHR, WC and increased UACR vanished, suggesting the inferior discriminative ability for albuminuria.

Associations between adiposity indices and increased UACR in strati ed analyses
To verify the discriminative power of all adiposity indices for albuminuria, strati ed analyses were performed in subgroups as shown in increased albuminuria. To further explore the association of the effect of blood glucose on UACR, the participants with prediabetes were divided into IFG, IGT, and IFG+IGT. Table 4 shows that only BRI had a signi cant association with increased UACR in the IGT and IGT+IFG groups (IGT: FBG < 6.1 mmol/L and 7.8 ≤PBG < 11.1 mmol/L: OR: 1.20, 95% CI 1.08-1.33, P=0.0005; IGT+IFG: 6 1 ≤ FBG < 7.0 mmol/L and i7.8 ≤ PBG <11.1 mmol/L: OR:1.34, 95% CI 1.03-1.74, P=0.0320), suggesting the additional risk of poor control of 2h post-load blood glucose on increased UACR. Moreover, to thoroughly verify the association of BRI with UACR, strati ed analysis was also conducted based on different renal functions, and we found similar results at different level of eGFR. When participants had a normal eGFR≥90 mL/min per 1.73 m2, the odds of having an increased UACR positively in participants with increased BRI (BRI: OR:1.26, 95% CI 1.13-1.41, P<0.0001). However, no signi cant association was found in participant with eGFR<90 mL/min per 1.73 m2. The above results show that compared with traditional adiposity indices (BMI, WC, WHR), the new adiposity index BRI presented a superior discriminative ability for albuminuria, suggesting the advantage of BRI as an easy and effective screening tool for early kidney failure.

Main ndings
In the present study, we found that BRI was positively associated with UACR and showed a superior discriminative ability for UACR than traditional adiposity indices (BMI, WHR and WC), suggesting its advantage of being an accurate discriminator for the risk of albuminuria to early predict renal lesion. Strati ed analysis revealed that participants with higher BRI were more likely to have albuminuria than those with lower BRI, especially in participants who were young (age <60 years), those with abnormal blood glucose (FBG ≥ 7.0 mmol/L and PBG ≥ 11.1 mmol/L), and those with normal blood pressure (SBP ≥ 140 mmHg or diastolic blood pressure DBP ≥ 90 mmHg) and eGFR (eGFR≥90 mL/min per 1.73 m2).
To the best of our knowledge, the is the rst multicenter, large sample investigation to explore the association between BRI and UACR in the Chinese elderly population. Therefore, effective prevention and intervention are important for UACR, and modi cation of the body fat distribution and body shape may contribute to effectively reduce the incidence of adverse events in participants with poor control of blood glucose and blood pressure.
It is well known that obesity is a leading cause of chronic diseases, including CVD, T2DM, hypertension and CKD. Considering the use of computed tomography (CT) and magnetic resonance imaging (MRI) to detect body composition is accurate but expensive and inconvenient, previous studies proposed to use anthropometric indices to investigate the association between obesity and the risk of chronic diseases. Traditionally, BMI is a commonly recommended anthropometric index of the body fat. BMI has been reported to be valuable in predicting increased UACR.9 As well, it is closely associated with CKD among hypertensive patients as a predictor of the onset of albuminuria. Conversely,8,10 other studies opposed that BMI only indicates general obesity with less accuracy but not abdominal body fat, which plays a more signi cant role in the progression of chronic diseases.26 Emerging evidence supported that it is abdominal obesity rather than general obesity that has superior predictive value for CVD risk. 27 Although BMI has been the most commonly used adiposity index, it has limited power to distinguish peripheral from central body fat, fat and lean mass.28,29 Therefore, WC and WHR have been proposed as surrogate indices for abdominal obesity. Previous studies found that both WC and WHR are better predictors of CVD risk than BMI.30-32 A recent study found WC to be more strongly associated with cardiometabolic risks than BMI, while another meta-analysis of 32 studies suggested WC, WHR and BMI have similar associations with the incidence of T2DM. 33,34 In this present study, no signi cant difference was observed among WC, WHR, and BMI, which is consistent with previous studies. The results might be partially explained by the following reasons. First, WC and WHR have the insu cient capability of distinguishing visceral fat from subcutaneous fat, resulting in overestimating the visceral fat tissue in subjects. Second, WC and WHR may underestimate or overestimate the visceral fat without the consideration of height, which is an indicator of nutritional status during childhood.
In recent years, WHtR was also found to be a better screening tool for cardiometabolic risk factors than BMI and WC in a meta-analysis.32 A cross-sectional study revealed that WHtR was closely associated with cardiometabolic risk factors.35 Moreover, another case-control study based on Chinese population found that WHtR showed the superior value in the assessment of CKD than BMI and WC, which is in line with our ndings.34 BRI, a novel and promising obesity-related index combines height and WC and has been used to quantify the body shape and predict visceral fat tissue volume.20 BRI has been proven to be more valuable in indicating body fat and visceral fat tissue than BMI and WC.20 Some studies investigated that BRI is a better predictor of hyperuricemia than BMI in women. It is a superior index to investigate the association between obesity and hyperuricemia than WC and WHtR. Similarly, another cross-sectional study demonstrated that BRI presents a high predictive value to detect dyslipidemia in women.36,37 However, there is limited evidence on the comparison of the discriminative ability of BRI with other adiposity indices in identifying the risk of UACR. This present study was the rst to access the superior ability of BRI to identify the risk of increased UACR compared with the traditional adiposity indices, indicating the advantage of BRI as an accurate and stable indicator of albuminuria.
Though the mechanism underlying the relationship between obesity and albuminuria is complex and unclear, it may be partially explained by the following reasons. It is well established that obesity can result in hemodynamic, hormonal and metabolic changes. First, obesity can induce increased renal plasma ow, glomerular pressure and ltration fraction, resulting in an increase in albuminuria.38,39 Second, it is reported that obesity activates the sympathetic nervous system and renin-angiotensin system (RAS) and physical compression of the kidneys, resulting in the increase of renal tubular sodium reabsorption and volume expansion, especially in the presence of visceral obesity. In addition, obesity can cause renal vasodilation and glomerular hyper ltration. The results of these changes further lead to glomerular injury and urinary protein excretion.40 Third, the relationship between visceral obesity and albuminuria may be linked to insulin resistance (IR). Accumulating evidence suggested that a characteristic feature of visceral obesity is increased free fatty acid (FFA) levels, which may contribute to the peripheral IR and dyslipidemia. 41 The podocyte insulin signaling plays an important role of the maintenance of the integrity of the glomerular ltration barrier. IR and lipo-toxicity induce the podocyte dysfunction, disrupting the integrity of the glomerular ltration barrier. This may be the initial step of diabetic nephropathy and the presence of albuminuria.42-44 The above-mentioned reasons might be possibly explained why the discriminative ability of BRI for the risk of UACR is superior to other adiposity indices.
In the clinical practice, it is convenient and highly cost-effective to use obesity-related indices to evaluate the visceral obesity and UACR risk. BRI showed the superior potential to be an accurate and stable indicator in the assessment of UACR. Hence, more attention should be paid to people with a high BRI to improve the distribution and deposition of visceral adiposity instead of only losing weight, in order to reduce CVD risk in clinical practice. We believe that taking BRI as part of management strategy is bene cial for reducing the diabetic nephropathy.

Limitations
Though the current study conducted in an aggregation of multi-community and the sample of our study was large, not only involved in 7 regions, but representing the general population in China. Several limitations, however, should be noted. Firstly, due to the design of this cross-sectional study, we can just investigate the association of different anthropometric indices with UACR and compared this relationship between new and traditional adiposity not cause. Second, though we excluded participants with related renal dysfunction and ACEI/ ARB medications, we did not detect other medications, possibly affecting the relationship. Furthermore, considering the convenience and price of measurements, we did not detect the distribution of body fat and abdominal fat percentage. Therefore, the conclusions should be draw with caution and more perfective studies should be needed to clarify the mechanisms underlying the relationship between anthropometric indices, especially BRI, and albuminuria.

Conclusions
In conclusion, this current study found that newly developed visceral adiposity index BRI was more closely associated with UACR compared with traditional adiposity indices (BMI, WHR, WC) and showed a superior discriminative ability for UACR in the general Chinese population. Considering that BRI is an accurate indicator of visceral body fat and body shape, people with borderline blood glucose, normal blood pressure and eGFR should pay more attention to control body shape and visceral body fat to reduce the risk of albuminuria and slow CKD progression rather than just weight loss for a low BMI. It is signi cant to identify people at high risk of albuminuria to take effective measures to decrease the incidence of CKD in the early status.

Declarations
Availability of data and materials The datasets used to support this study are not freely available due to participants' privacy protection.   Model II adjusted for age, sex, BMI, center, SBP, DBP, pulse, LDL, HDL, TG, FBG, PBG, HbA1c, ALT, AST, GGT, eGFR, BMI Z score, WHR Z score, ABSI Z score, BRI Z score, smoking status, drinking status, history of hypertension, cardiovascular history, diabetes history, medication history. Figure 1