Temporal relation between body mass index and renal function in individuals with hypertension and excess body weight

Results Of the 218 participants who attended the first examination, 150 were available for paired final analyses. At the end of follow-up, GFR decreased by 1.024 mL/min for each 1-kg/m 2 increment in BMI ( P < 0.03). When BMI was analyzed in quartiles, a positive graded relation with GFR changes was observed in quartiles 1 and 2 (individuals who maintained or lost weight), and a negative relation in quartiles 3 and 4 (individuals who gained weight, P = 0.05). A significant difference was observed between the smallest and highest BMI quartiles ( P = 0.01). At the end of follow-up, the 76 participants (51%) who gained weight (+4.6 ± 0.4 kg) showed a reduction in GFR (−2.99 ± 1.99 mL/min) of borderline significance ( P = 0.06) and a significant increase in fasting plasma glucose and triacylglycerol levels. Conversely, the 74 participants who maintained or lost weight showed no significant change in GFR and in fasting plasma glucose and triacylglycerol levels, although their blood pressure decreased significantly. Conclusions Our study showed a significant temporal association between changes in BMI and GFR in overweight and obese hypertensive patients. Keywords Obesity Body mass index Glomerular filtration rate Chronic kidney disease Hypertension Introduction Chronic kidney disease (CKD) is a worldwide public health problem [1] . In the United States, the number of patients with end-stage renal disease (ESRD) by the year 2010 is expected to reach 651 330 and the total Medicare ESRD program cost will exceed US $28 billion [2] . Data from the Brazilian guidelines for CKD showed 59 153 individuals with ESRD in 2004, with a progressively increasing incidence of 8% per year [3] . With the increasing prevalence of overweight and obesity and their impact on metabolic and cardiovascular diseases, more attention is now being focused on the relation between obesity and renal function. Through its close association with type 2 diabetes and hypertension, the two most important precursors of ESRD [4] , excess body weight has been identified as a major risk factor for the development of CKD and ESRD [5] . There is also epidemiologic evidence that excess body weight by itself, independently of blood pressure levels and presence of type 2 diabetes, significantly increases the risk of progression to CKD, including the risk for ESRD [6,7] . Additional risk factors that have been variably associated with CKD include cholesterol level abnormalities, smoking, and alcohol consumption [8–10] . However, because many studies in obese individuals have not been longitudinally designed, the effects of concomitant diseases, such as diabetes and hypertension, cannot be completely separated. In addition, few studies have evaluated body mass index (BMI) as a potential risk factor for the development of kidney disease, and available results are still controversial. In addition, most of the literature relating excess body weight to renal disease is limited to primary outcomes, such as CKD or ESRD, and the temporal relation between BMI and the emergence of early renal dysfunction remains unknown. Thus, the present study, conducted in hypertensive individuals with excess body weight, assessed whether variations in BMI would reflect on eventual changes in the estimated glomerular filtration rate (GFR) not only at baseline but also prospectively over time. Materials and methods This was a cross-sectional, longitudinal study carried out in overweight and obese hypertensive patients originated from the Hypertension Clinic—CLINEX—of the Rio de Janeiro State University from January to December 2000. The Hypertension Clinic is a Brazilian reference center in the State of Rio de Janeiro for multidisciplinary care of hypertensive patients with cardiometabolic risk factors. The exclusion criteria were a GFR <60 mL/min and/or previous nephrectomy, obstructive uropathy, acute renal insufficiency, renovascular hypertension, polycystic renal disease, glomerulopathy of several etiologies, or tubulointerstitial nephritis. Smokers and those with a history of cardiovascular disease or stroke were maintained in the study, unless an acute episode of any of those disorders had occurred within the 3 mo preceding the selection. Two hundred eighteen individuals, aged 25 to 70 y, without gender distinction were enrolled. Participants were evaluated every 6 mo by physicians and nutritionists and were strongly recommended to follow an energy-restricted diet and to exercise regularly. The study protocol was approved by the committee on ethics of Pedro Ernesto University Hospital. Patients were considered hypertensive if their systolic and/or diastolic blood pressure levels were ≥140 mm Hg and ≥90 mm Hg, respectively [11] , or if they were on antihypertensive therapy. Diabetes mellitus was diagnosed when fasting glucose levels were ≥126 mg/dL or when patients were using insulin or an oral antidiabetic medication for at least 8 wk [12] . The diagnosis of dyslipidemia followed the Adult Treatment Panel III (ATP III) criteria [13] : total cholesterol ≥200 mg/dL, low-density lipoprotein cholesterol (LDL-C) ≥130 mg/dL, and triacylglycerols ≥150 mg/dL, or use of cholesterol-lowering drugs. Blood pressure was measured by the auscultatory method in the dominant arm after a resting period of at least 10 min in the sitting position. A standard mercury sphygmomanometer was used, with the appropriate cuff size for each patient. The first value was discarded and the mean of at least three readings was used in the analysis. Height and weight were measured according to standardized procedures, and BMI was calculated. Anthropometric measurements were taken twice, and the mean values were used for all analyses. The cutoff points used to define overweight and obesity followed the World Health Organization classification [14] . All blood samples were collected after a 12-h fasting period. Fasting plasma glucose was determined by use of the glucose-oxidase method. Serum lipid profile, including total cholesterol, triacylglycerols, and high-density lipoprotein cholesterol, was estimated by enzymatic methods. LDL-C was calculated by the Friedewald formula, when values of triacylglycerols were lower than 400 mg/dL [15] . Uric acid was measured by an automated technique. Serum urea nitrogen and plasma creatinine were assessed by a kinetic method. GFR was estimated by using the Modification of Diet in Renal Disease (MDRD) equation, in which GFR (mL/min per 1.73 m 2 ) = 186 × (serum creatine) −1.154 × (age) −0.203 × (0.742 if female) × (1.210 if African American) (conventional units) [16] . This equation does not require weight because the results are reported normalized to 1.73-m2 body surface area, which is an accepted average adult surface area. The MDRD equation is superior to other methods for estimating GFR. Direct comparisons of the MDRD equation with other equations, such as Cockcroft-Gault and even 24-h urine collection, have proven its superiority [17] . We analyzed BMI as a continuous variable and as quartiles. Continuous and quartile-based analyses of BMI values consisted of the following models. 1. Cross-sectional relation of BMI values and GFR at baseline examination using multivariable regression models, adjusting for sex, smoking, blood pressure, presence of diabetes and dyslipidemia, clinical history of myocardial infarction and stroke, and antihypertensive therapy. 2. Relations of baseline BMI values to changes in GFR at the end of follow-up, adjusting for sex, smoking, blood pressure, presence of diabetes and dyslipidemia, clinical history of myocardial infarction and stroke, and antihypertensive therapy, using a multivariable regression analysis. 3. Association between changes in BMI values and changes in GFR at the end of follow-up, adjusting for sex, smoking, blood pressure, presence of diabetes and dyslipidemia, clinical history of myocardial infarction and stroke, and antihypertensive therapy, using a multivariable regression analysis. Secondary analyses were performed to assess the possible differences between cardiometabolic risk factors and GFR in subjects who gained weight during follow-up (group 1 [G1]) and those who maintained or lost weight during the same period (group 2 [G2]). For all analyses, two-tailed P < 0.05 was considered statistically significant. Continuous variables were expressed as mean and standard error of the mean. Comparisons between two continuous variables were performed by using Student's t test. Changes in the values of the variables at the end of follow-up were assessed by subtracting arithmetically the values obtained in 2006 from those registered in 2000. STATA 8.2 (STATA Corp., College Station, TX, USA) was used for statistical analysis. Results Of the 218 participants who attended the first examination, complete data from 150 were available for the paired final analyses. The average interval between the examinations at baseline and at the end of follow-up was 5.8 y. A significant decrease was observed in the measurements when comparing those taken at baseline and at the end of follow-up in the following variables: blood pressure levels, heart rate, and values of total cholesterol and LDL-C. Conversely, a significant increase was observed in the measurements when comparing those taken at baseline and at the end of follow-up in the following variables: levels of triacylglycerols and fasting plasma glucose ( Table 1 ). At baseline, all participants were hypertensive, 71% had dyslipidemia, 32% were diabetic, 11% were smokers, and 5% and 3% had a history of acute myocardial infarction and stroke, respectively. In regard to medication, 4% of patients were on insulin therapy, 37% were taking statins, 77% used angiotensin-converting-enzyme inhibitors, 44% used calcium antagonists, 51% used β-blockers, 85% used diuretics, and 25% were taking oral antidiabetic agents. Cross-sectional relations between baseline BMI and GFR At baseline, in the analyses modeling BMI as a continuous variable, no significant relation with GFR values was observed. In quartile-based analyses, no significant relation between BMI and GFR ( P = 0.63) was observed, not even when the lowest and highest BMI quartiles were compared ( P = 0.19). Baseline BMI values and changes in GFR There was no relation between BMI values, modeled as a continuous variable, and GFR changes over time ( P = 0.12). In the quartile-based analyses, baseline BMI did not relate to GFR changes at the end of the 5.8-y follow-up ( P = 0.39). Changes in BMI values and GFR during follow-up During follow-up, the changes in BMI ranged from −6.9 to +9.6 kg/m 2 . The GFR decreased by 1.024 mL/min for every 1-kg/m 2 increment in BMI, with statistical significance ( P = 0.03), after adjustments for sex, smoking, blood pressure levels, presence of diabetes and dyslipidemia, clinical history of myocardial infarction and stroke, and antihypertensive therapy. In the quartile-based analyses, a positive graded relation between BMI changes and GFR changes was observed in quartiles 1 and 2, corresponding to those subjects who maintained or lost weight (G2), and a negative relation was observed in quartiles 3 and 4, corresponding to those who gained weight during follow-up (G1; Fig. 1 ). A significant difference between the smallest and highest BMI quartiles was also observed ( P = 0.01). In the 76 participants (51%) who gained weight during follow-up (mean weight change 4.6 ± 0.4 kg, P < 0.0001), a reduction of borderline statistical significance in GFR was observed (mean change −2.99 ± 1.99 mL/min, P = 0.06). Those subjects also had a significant increase in fasting plasma glucose and serum triacylglycerol levels, despite significant decreases in total cholesterol and LDL-C. Conversely, the 74 participants (49%) who maintained or lost weight during follow-up (mean decrease 4.1 ± 0.3 kg) showed neither a significant temporal modification in GFR nor significant changes over time in fasting plasma glucose and serum triacylglycerols. Nevertheless, like those participants who gained weight, they had a statistically significant decrease in total cholesterol and LDL-C levels. However, only participants in G2 exhibited a significant decrease in systolic and diastolic blood pressures ( Table 2 ). In addition, changes in GFR, glucose, and triacylglycerol levels were significantly different in G1 and G2 ( Table 2 ). Discussion Major findings In the present study, based on a sample of hypertensive individuals with excess body weight, temporal changes in BMI were associated with a GFR variation during the 5.8-y follow-up. Models analyzing BMI values as a continuous variable yielded results consistent with quartile-based models. When the participants were categorized according to their weight variation at follow-up, a significant decrease in GFR and a significant increase of fasting plasma glucose and triacylglycerol levels was observed only in subjects who gained weight compared with those in individuals who lost or maintained their weight. In addition, subjects loosing or keeping their weight stable exhibited a consistent and significant control of their blood pressure over time. Comparison with previous studies Previous large cohort studies have shown a strong association between obesity and the risk for CKD and ESRD [6–9,18,19] . Whether hypertension or diabetes is a prerequisite for developing CKD in overweight and obese patients has recently been assessed in several studies. In a Swedish population-based, case–control study, overweight and obesity were evaluated as risk factors for advanced CKD. After adjustment for confounding variables, men and women with a BMI of at least 25 kg/m 2 showed a greater chance for developing advanced CKD, independently of concomitant diabetes or hypertension [19] . Another large, historic cohort study extended these findings to subjects with ESRD receiving maintenance dialysis or a renal transplant. Patients were categorized according to BMI as normal weight to class III obesity. After adjusting for confounding variables, the association between ESRD and excess body weight remained significant and continued to increase in a stepwise fashion across categories of BMI, starting at a threshold of 25 kg/m 2 . BMI was also a strong predictor of ESRD development, even in individuals with hypertension and diabetes mellitus [6] . In the present study, we evaluated prospectively the consequences of body fat mass variations on renal function at a point of the natural history of renal disease antedating the development of overt chronic kidney dysfunction. In addition, our results showed that a significant decrease in GFR (−0.51 mL/min per year) occurred temporally only in those individuals who gained weight, but not in those who lost or whose weight remained stable, although this decrease would not be considered a major loss according to the rate of decline in renal function with age [20] . Potential mechanisms There is scarce information in the literature evaluating the early repercussions of excess body weight on renal function. A study on the effects of obesity on renal function and urinary albumin excretion in normotensive subjects and never-treated hypertensive patients demonstrated that excess weight was associated with renal hyperfiltration and hyperperfusion, irrespective of the presence of hypertension, and that obesity magnified the effect of hypertension on albuminuria [21] . In another analysis on renal hemodynamics in lean and obese hypertensive subjects, renal vascular resistance was significantly decreased in obese hypertensive subjects compared with that in lean hypertensive subjects, thus explaining the glomerular hyperfiltration and renal vasodilatation that resulted in loss of renal functional reserve, an early mark of renal dysfunction in subjects with excess body weight [22] . A central clinical feature of the present study was that all participants had overweight- or obesity-associated hypertension. This association can be explained by increased renal sodium reabsorption that leads to impairment of the renal-pressure natriuresis and requires higher levels of blood pressure to maintain the balance between intake and urinary output [23] . According to mechanistic evidence, increased renal tubular reabsorption is accompanied by a compensatory increase in GFR and increased renal blood flow due to afferent arteriolar dilatation and efferent arteriolar constriction, by tubular salt reabsorption, and an increase in angiotensin II levels [24] . In addition, BMI is an important determinant of renal hemodynamic adaptation to a high sodium intake in young obese men without kidney disease [25] . Thus, hyperfiltration is the hallmark of obesity-associated kidney dysfunction because it exacerbates the physical stress on the glomerular wall and causes the appearance of microalbuminuria or proteinuria, even before major structural changes occur [26] . Interestingly, and according to the hyperfiltration hypothesis, we cannot rule out the possibility that the 4.22-mL/min difference of estimated GFR values at baseline in individuals in G1, although not statistically significant when compared with those in G2, could have affected the final values of GFR at follow-up. What other additional mechanisms would be involved in the renal dysfunction of our patients with excess body weight? In this regard, hyperglycemia has been considered a possible additional factor. It is worth noting that the significant elevation in fasting plasma glucose (change in plasma glucose +10.44 ± 3.82 mg/dL during follow-up) observed exclusively in our patients who gained weight could lead to glomerular structural changes and nephron loss, when combined with other metabolic and profibrotic effects of hyperinsulinemia, resulting in a decrease in GFR [27] . In addition, increased glucose filtration is a supplementary mechanism enhancing sodium reabsorption in the proximal tubule, by the sodium–glucose cotransporter [28] . There is clear-cut evidence that hypertriglyceridemia reflects the over-accumulation of free fatty acids in several non-adipose tissues with subsequent toxicity secondary to fatty acid non–β-oxidation [29] . Serum triacylglycerols increased more than 40% in the group with weight gain, reaching values well above the recommended thresholds of the ATP III guidelines. Although there is no clinical evidence showing that increased accumulation of lipids in the kidney tissue has deleterious effects on renal function and structure, obesity is experimentally associated with marked accumulation of fatty tissue, especially in the renal sinuses [30] . The clinical response of blood pressure to antihypertensive agents deserves to be commented on. Although both groups had been using a similar protocol to decrease blood pressure, including the prescription of angiotensin-converting enzyme inhibitors that exert well-known nephro-protective effects [31] , only the patients who lost weight or maintained their weight stable exhibited a significant decrease in blood pressure. This finding confirms the central role of hypertension in the development of obesity-associated renal disease [32] and emphasizes that weight decrease and adequate blood pressure control may be as effective as plasma glucose control in obese patients for preventing the development of progressive renal insufficiency [33] . Limitations The present study has some limitations. First, the follow-up was relatively short. Considering that the natural history of CKD usually lasts much longer than 5.8 y, our findings are medium-term outcomes. Second, the interpretation of our findings may be also limited by potential misclassification of kidney function by using estimated GFR, obtained through the use of the original, not calibrated, MDRD study equation. Moreover, because muscle mass affects creatinine generation, and the equation used to estimate GFR includes creatinine level, the association of higher BMI with the risk for CKD may be partly related to the greater muscle mass of our obese patients. However, the MDRD study equation has been recommended by the National Kidney Foundation for GFR estimation and was the only equation used to estimate GFR in all participants at baseline and the end of follow-up. Third, despite adjustments for a large number of potential confounders of the association between BMI and renal dysfunction, residual confounding remains a possible alternative explanation for our findings given the study's observational design. Fourth, we were unable to account for other covariates known to be associated with body weight and renal dysfunction, including microalbuminuria and adipocytokines. In conclusion, our study shows a significant temporal association between changes in BMI and GFR variation in overweight and obese hypertensive patients. References [1] A.S. Levey J. Coresh E. Balk A.T. Kausz A. Levin M.W. Steffes National Kidney Foundation practice guidelines for chronic kidney disease: evaluation, classification, and stratification Ann Intern Med 139 2003 137 147 [2] J.L. Xue J.Z. Ma T.A. Louis A.J. Collins Forecast of the number of patients with end-stage renal disease in the United States to the year 2010 J Am Soc Nephrol 12 2001 2753 2758 [3] J.E.R. Junior The Brazilian guidelines for chronic kidney diseases J Bras Nefrol 24 Suppl 1 2004 S1 S3 [4] USRDS the United States Renal Data System Am J Kidney Dis 42 Suppl 5 2003 1 230 [5] A. Sachse G. Wolf New aspects of the relationship among hypertension, obesity, and the kidneys Curr Hypertens Rep 10 2008 138 142 [6] C.Y. Hsu C.E. McCulloch C. Iribarren J. Darbinian A.S. Go Body mass index and risk for end-stage renal disease Ann Intern Med 144 2006 21 28 [7] R.P. Gelber T. Kurth A.T. Kausz J.E. Manson J.E. Buring A.S. Levey Association between body mass index and CKD in apparently healthy men Am J Kidney Dis 46 2005 871 880 [8] B. Stengel M.E. Tarver-Carr N.R. Powe M.S. Eberhardt F.L. Brancati Lifestyle factors, obesity and the risk of chronic kidney disease Epidemiology 14 2003 479 487 [9] C.S. Fox M.G. Larson E.P. Leip B. Culleton P.W.F. Wilson D. Levy Predictors of new-onset kidney disease in a community-based population JAMA 291 2004 844 850 [10] E.S. Schaeffner T. Kurth P.E. de Jong R.J. Glynn J.E. Buring J.M. Gaziano Alcohol consumption and the risk of renal dysfunction in apparently healthy men Arch Intern Med 165 2005 1048 1053 [11] A.V. Chobanian G.L. Bakris H.R. Black W.C. Cushman L.A. Green J.L. Izzo Seventh report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure Hypertension 42 2003 1206 1252 [12] American Diabetes Association diagnosis and classification of diabetes mellitus Diabetes Care 31 Suppl 2008 S55 S60 [13] Executive summary of the Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III) JAMA 285 2001 2486 2497 [14] World Health Organization Physical status: the use and interpretation of anthropometry Technical report series 854 1995 World Health Organization Geneva Available at: http://www.who.int Accessed July 9, 2008 [15] W.T. Friedewald R.I. Levy D.S. Frederickson Estimation of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge Clin Chem 18 1972 499 502 [16] A. Levey J.P. Bosch J.B. Lewis T. Greene N. Rogers D. Roth A more accurate method to estimate glomerular filtration rate from serum creatinine: a new prediction equation Ann Intern Med 130 1999 461 470 [17] National Kidney Foundation Education Program (internet homepage). National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health, US Department of Health and Human Services; last reviewed December 28, 2005. Available at: http://www.nkdep.nih.gov . Accessed July15, 2008. [18] K. Iseki Y. Ikemiya K. Kinjo T. Inoue C. Iseki S. Takishita Body mass index and the risk of development of end-stage renal disease in a screened cohort Kidney Int 65 2004 1870 1876 [19] E. Ejerblad C.M. Fored P. Lindblad J. Fryzek J.K. McLaughlin O. Nyrén Obesity and risk for chronic renal failure J Am Soc Nephrol 17 2006 1695 1702 [20] R.D. Lindeman J. Tobin N.W. Shock Longitudinal studies on the rate of decline in renal function with age J Am Geriatr Soc 33 1985 278 285 [21] J. Ribstein G. du Cailar A. Mimran Combined renal effects of overweight and hypertension Hypertension 26 1995 610 615 [22] I.M.D. Pecly V. Genelhu E.A. Francischetti Renal functional reserve in obesity hypertension Int J Clin Pract 60 2006 1198 1203 [23] J.E. Hall The kidney, hypertension, and obesity Hypertension 41 2003 625 633 [24] I.M. Wahba R.H. Mak Obesity and obesity-initiated metabolic syndrome: mechanistic links to chronic kidney disease Clin J Am Soc Nephrol 2 2007 550 562 [25] J.A. Krikken A.T. Lely S.J.L. Bakker G. Navis The effect of a shift in sodium intake on renal hemodynamics is determined by body mass index in healthy young men Kidney Int 71 2007 260 265 [26] P. Metcalf J. Baker A. Scott C. Wild R. Scragg E. Dryson Albuminuria in people at least 40 years old: effect of obesity, hypertension, and hyperlipidemia Clin Chem 38 1992 1802 1808 [27] J.R. Henegar S.A. Bigler L.K. Henegar S.C. Tyagi J.E. Hall Functional and structural changes in the kidney in the early stages of obesity J Am Soc Nephrol 12 2001 1211 1217 [28] A. Geneidy R. Solomon Obesity and renal disease C.S. Mantzoros Obesity and diabetes 2006 Humana Press Totowa, NJ 319 331 [29] J.M. Weinberg Lipotoxicity. Kidney Int 70 2006 1560 1566 [30] T.M. Dwyer H.L. Mizelle K. Cockrell P. Buhner Renal sinus lipomatosis and body composition in hypertensive, obese rabbits Int J Obes Relat Metab Disord 19 1995 869 874 [31] P. Ruggenenti A. Perna G. Gherardi G. Garini C. Zoccali M. Salvadori Renoprotective properties of ACE-inhibition in non-diabetic nephropathies with non-nephrotic proteinuria Lancet 354 1999 359 364 [32] A.K. Bidani K.A. Griffin Pathophysiology of hypertensive renal damage Hypertension 44 2004 595 601 [33] J.R. Sowers S. Haffner Treatment of cardiovascular and renal risk factors in the diabetic hypertensive Hypertension 40 2002 781 788