Association of statin use and development of renal dysfunction in type 2 diabetes—The Hong Kong Diabetes Registry

Methods A consecutive cohort of 5264 diabetic patient recruited between 1996 and 2005 underwent detailed assessments. Renal dysfunction was defined as first estimated glomerular filtration rate <60 ml/min/1.73 m 2 , or, the first hospitalisation with a diagnosis of renal disease as coded by the International Classification of Disease, Ninth Revision. Drug use was quantified using the proportion of exposure time from baseline to event/death/censored time, as appropriate. Results In this cohort (male: 47.3%, median age: 55 years, median duration of diabetes: 6.0 years), none had renal dysfunction at baseline. During a median follow-up period of 4.9 (quartiles: 2.77, 7.04) years, 703 patients (13.4%) developed renal dysfunction, 1275 patients (22.2%) were exposed to statins. After controlling for baseline risk factors, multivariable adjusted hazard ratio of statin use for development of renal dysfunction was 0.32 (95% CI 0.21–0.50, p < 0.0001). Conclusion Use of statins was associated with reduced risk of developing renal dysfunction in type 2 diabetes and this association was independent of baseline risk factors. Keywords Type 2 diabetes Renal dysfunction Statin 1 Introduction Amongst all diabetes-associated complications, diabetic nephropathy has probably the most detrimental consequences resulting in dialysis and heightened cardiovascular risks [1,2] . Asians, including Chinese patients with type 2 diabetes, have a greater propensity to develop renal complications, affecting 60% of them as compared to 30–40% in Caucasians given similar disease duration [3–5] . Chronic kidney disease is associated with dyslipidaemia, typified by low levels of high-density lipoprotein cholesterol (HDL-C) and high triglyceride [6,7] . Moreover, large scale epidemiologic and clinical studies suggested that dyslipidaemia per se, in addition to established risk factors of hypertension and hyperglycaemia, may play a part in the development of renal dysfunction. In the Physician's Health Study, high total cholesterol to HDL-C ratio was associated with significantly increased risk of developing renal disease in initially healthy men observed for over a decade [8] . Recently, our group has reported the association between hypertriglyceridaemia and incident chronic kidney disease amongst Chinese patients with type 2 diabetes [9] . The precise mechanisms by which dyslipidaemia generates renal injury are yet to be elucidated. Apart from renal microcirculatory dysfunction brought on by atherosclerosis [10] , lipid accumulation in renal parenchyma can result in upregulation of various inflammatory cytokines, which in turn stimulate macrophage infiltration and foam cell formation [11] . Other mechanisms include direct interaction of lipoproteins with receptors expressed on mesangial cells causing mesangial cell proliferation and matrix deposition. Exposure of podocytes to oxidized lipids can also lead to structural aberration and diminution of nephrin, the latter being a protein of the slit diaphragm that has anti-apoptotic property [12] . Results from several post-hoc and meta-analyses of clinical studies have supported potential renoprotective effects of lipid-lowering drugs [13–15] . In the current study, using a prospective cohort of Chinese patients with type 2 diabetes with detailed documentation of phenotypes at baseline, drug usage during follow-up and clinical outcomes, we investigate whether the use of statins is associated with reduced risk of renal dysfunction. 2 Methods 2.1 Patients The Prince of Wales Hospital is the teaching hospital of the Chinese University of Hong Kong and serves a population of over 1.2 million. The Hong Kong Diabetes Registry was established in 1995 as part of a quality improvement program. The referral sources included general practitioners, community clinics, other specialty clinics and patients discharged from the Prince of Wales Hospital or other regional hospitals. Enrolled patients with hospital admissions within 6–8 weeks prior to assessment accounted for less than 10% of all referrals. The 4-h assessment of complications and risk factors was performed on an outpatient basis, modified from the European DiabCare protocol [16] . Once a patient has undergone comprehensive assessment, he/she was considered to have entered this study cohort and would be followed up till time of death. Patients were excluded if they had type 1 diabetes as defined by acute presentation with diabetic ketoacidosis, heavy ketonuria (>3+) or continuous requirement of insulin within 1 year of diagnosis, if they had unknown diabetes type, or if they were non-Chinese or had unknown nationality. In addition, patients with baseline estimated glomerular filtration rate (eGFR) <60 ml/min/1.73 m 2 , who were on dialysis or had missing baseline eGFR were excluded. Finally, we excluded patients who used statin before entering analysis. Ethical approval was obtained from the Chinese University of Hong Kong Clinical Research Ethics Committee. The Declaration of Helsinki was adhered and informed consent was obtained from all patients for data analysis and research purpose at the time of assessment. 2.2 Endpoints For this analysis, the clinical endpoints, including discharge diagnoses of hospital admissions and mortality, were censored on 30th July 2005. Details of all medical admissions of the cohort by that date were retrieved from the Hong Kong Hospital Authority (HA) Central Computer System. This system records admissions to all public hospitals which provide 95% of total hospital bed-days in Hong Kong. Hong Kong has a heavily subsidized health care system and the majority of patients with chronic diseases are followed up at hospital clinics where medications are dispensed on site. Drug use data were extracted from the HA Central Computer System that records all drug dispensary data in all public hospitals including start and end dates of the use of each drug of interest. These databases are matched by a unique identification number, the Hong Kong Identity Card number, which is compulsory for all residents in Hong Kong and used by all government departments and major organizations. Hospital discharge diagnoses coded by the International Classification of Diseases, Ninth Revision (ICD-9), were used to identify the first hospitalisation with renal disease event, defined as (1) hospitalisation with diagnosis of diabetes with renal manifestations (Code 250.4), chronic kidney disease (Code 585) or unspecified renal failure (Code 586), or (2) dialysis (ICD-9 procedure code: 39.95) or peritoneal dialysis (ICD-9 procedure code: 54.98). During the observation period, 22,276 measurements of serum creatinine in the patients used in the analysis (2317 patients did not have a serum creatinine on follow-up) were retrieved from the HA Central Computer System for this prospective cohort. Glomerular filtrate rate was calculated in each data point to identify the first event of eGFR <60 ml/min/1.73 m 2 . Thus, the endpoint of the study was renal dysfunction as defined by the first eGFR <60 ml/min/1.73 m 2 or the first hospitalisation with renal disease. 2.3 Clinical measurements Details of clinical assessments and laboratory assays have been described [17] . Fasting blood samples were taken for measurement of plasma glucose, glycohemoglobin (HbA 1C ), lipid profile (total cholesterol, HDL-C, triglycerides and calculated low-density lipoprotein cholesterol [LDL-C]), and renal function. A sterile, random spot urine sample was used to measure albumin to creatinine ratio (ACR). All laboratory assays were performed at the Department of Chemical Pathology, the Prince of Wales Hospital, which is accredited by the Royal College of Pathologists of Australasia. This study used the abbreviated modification of diet in renal disease (MDRD) formula recalibrated for Chinese to calculate GFR expressed in ml/min/1.73 m 2 : eGFR = 186 × [SCR × 0.011] −1.154 × [age] −0.203 × [0.742 if female] × 1.233, where SCR is serum creatinine expressed as μmol/l (original mg/dl converted to μmol/l) and 1.233 is the adjusting coefficient for Chinese [18] . 2.4 Statistical analyses The Statistical Analysis System (Release 9.10) was used to perform the statistical analysis (SAS Institute Inc., Cary, USA). Follow-up time was calculated as the period in year from the first enrolment since 1 December 1996 to the date that end point was reached, death date or censored date. Cox proportional regression was used to derive hazard ratios for development of renal dysfunction due to the use of statins while controlling for baseline covariates including age, sex, current/ex-smoking status, duration of diabetes, systolic/diastolic blood pressure (SBP/DBP), body mass index (BMI), HbA 1C , total cholesterol, HDL-C, triglyceride, LDL-C, urine ACR, eGFR, and use of angiotensin converting enzyme (ACE) — inhibitors, angiotensin receptor blockers (ARB), insulin and aspirin during follow-up, and, use of oral anti-diabetic drugs and anti-hypertensive drugs other than ACE-inhibitors/ARB at baseline. To adjust for covariates that displayed non-linearity, restricted cubic splines with 4 knots were applied in the Cox proportional hazard regression model [19,20] . There is no consensus on the optimal method in quantifying drug use in cohort studies. We have described a method, so-called “adjusted time proportion of drug use” to measure drug exposure [21,22] . In brief, drug use was quantified as the proportion of drug exposure time during the entire follow-up period from baseline to event/death/censored time, whichever was earlier. Then, the time proportion of drug use was further divided by the sum of 1 and the number of years from termination of the drug to the event/death/censored time. Thus, the adjusted time proportion of the use a lipid-lowering drug was coded “1” if the drug was used for the whole period of time until censored. If the drug was stopped early, a number between 0 and 1 was coded. Subjects who comply to the use of a particular drug over a long duration may represent a distinctively health-conscious group. Better clinical outcomes seen with longer duration of drug use may not be due to drug effect, but rather, the product of general health promoting behaviour in patients who adhere to treatment. This type of selection bias, termed “healthy volunteer bias”, is particularly problematic in observational studies that assess drug effects [23] . Any drug quantification schemes that incorporate absolute duration of drug use may suffer from this bias. For example, use of drugs at the endpoint and use of the drug in the last 20% of the follow-up time are quantification methods that can introduce healthy volunteer bias. As an alternative, we used “adjusted time proportion” which is a ratio of duration of drug use over total follow-up time. This quantification of drug exposure is “free” from absolute measurements of drug use duration and theoretically will reduce healthy volunteer bias. Using “adjusted time proportion”, a patient who is followed for 5 years and has used statins for the entire follow-up period and another patient who is followed for 2 years and has also used statins for the 2 years would both be coded to 1. For this study, we compared the model fit (−2 log likelihood) of the “adjusted time proportion” model with the traditional yes/no coding scheme and the “time proportion” model. The “adjusted time proportion” method had the best model fit, i.e., it had the least −2 log likelihood value (see Appendix A ). Therefore, the “adjusted time proportion” was used in this analysis to quantify all drug exposure during follow-up. Proportionality and functional form were checked using the Supremum test, which is implemented using the ASSESS statement in the SAS procedure PROC PHREG. [24] . A two-sided p value <0.05 was considered to be significant unless indicated otherwise. 3 Results From 1995 to 2005, 7920 patients with type 2 diabetes were enrolled in the Registry. As detailed drug use information in the HA Central Computer System were only available from 1 December 1996, we limited the analysis to 7387 patients who underwent comprehensive assessment on or after 1 December 1996. After excluding patients with type 1 diabetes or with unknown diabetes type ( n = 328), who were non-Chinese ( n = 45), with renal dysfunction at baseline ( n = 1010) or were using statin at baseline ( n = 740), a total of 5264 patients were eligible for analysis. The cohort had a median age of 55 (quartiles 45, 66) years and median duration of diabetes of 6 (quartiles 1, 10) years. During a median of 4.90 (quartiles 2.77, 7.04) years of follow-up, 13.4% ( n = 703) of the patients developed renal dysfunction (612 patients with eGFR <60 ml/min/1.73 m 2 and 275 patients with hospitalisation with renal diseases amongst whom 184 patients overlapped with eGFR <60 ml/min/1.73 m 2 ). Table 1 compares the baseline demographic, clinical and biochemical characteristics of patients with and without renal dysfunction. Patients in the former group were older, had diabetes for longer, worse metabolic profile and higher blood pressure. They were also more likely to have microvascular (retinopathy and neuropathy) and macrovascular (coronary artery disease, cerebrovascular disease and peripheral vascular disease) complications than those who did not develop renal dysfunction. Overall, 22.9% ( n = 161) of patients who progressed to renal dysfunction, and, 24.4% ( n = 1114) of patients who maintained normal renal function, used statins during the follow-up period. The frequency of statin use ( p = 0.3804) and its adjusted time proportions (0.45 versus 0.50, p = 0.2630) were similar between patients who developed renal dysfunction and those who did not during follow-up. The relationship between the use of lipid-lowering agents and development of renal dysfunction was further assessed by Cox proportional regression analysis. After adjustment for baseline covariates, the hazard ratio of use of statins for renal dysfunction during follow-up was 0.64 (95% CI 0.46–0.91, p = 0.0121) per adjusted time proportion. Further adjustment for adjusted time proportions of using ACE-inhibitors, ARB, insulin and aspirin, as well as uses of oral anti-diabetic drugs and anti-hypertensive drugs at baseline, lowered the hazard ratio to 0.32 (95% CI 0.21–0.50, p < 0.0001). However, further consideration of non-linear associations with renal dysfunction did not greatly change the magnitude of the association (hazard ratio: 0.34, 95% CI 0.21–0.54, p < 0.0001) ( Table 2 ). 4 Discussion In this prospective analysis of patients with type 2 diabetes with detailed documentation of risk factors, complications, treatments and clinical outcomes, 24.2% of patients were commenced statins. Furthermore, use of statins, analysed as adjusted time proportions of drug exposure, was associated with about two third risk reduction for the development of renal dysfunction. The latter was independent of all baseline risk factors and disease-modifying drugs. Of note, the association between statin use and renal dysfunction was independent of baseline lipid levels. In recent years, the potential renoprotective effect of lipid-lowering therapy has been speculated upon. In the post-hoc analysis pooling 18,569 patients from 3 trials (Cholesterol and Recurrent Events trial, Long-term Intervention with Pravastatin in Ischemic Disease, and West of Scotland Coronary Prevention Study), treatment with pravastatin retarded the rate of renal function loss by 8% amongst patients with at least stage 3 chronic kidney disease [13] . In the Heart Protection Study comprising 20,536 patients one quarter of whom had diabetes, simvastatin was associated with a smaller decline in GFR compared to placebo (5.9 versus 6.7 ml/min/1.73 m 2 ), irrespective of baseline renal function [14] . More recently, Sukhija et al. analysed data of 197,551 patients retrieved from the Department of Veterans Affairs database in the United States [25] . Renal dysfunction, as defined by doubling of serum creatinine or increase in serum creatinine by 0.5 mg/dl, developed in 2.9% of statin users compared with 4.7% of non-users over 3 years of observation. This translated to a reduction of 13% in the odds of developing renal dysfunction. By comparison, the reduction in risk of progressing to renal complication in our cohort is significantly greater in magnitude. Our cohort consists of high risk patients with type 2 diabetes, a significant proportion of whom also had albuminuria. Any apparent renal benefits of statins may be amplified in patients who are inherently at higher risk for renal complications. Not all studies support a beneficial effect of statin on renal function. In the recently published post-hoc analysis of the Anti-hypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial, treatment with pravastatin in 10,060 patients with hypertension and dyslipidaemia showed no significant difference in the frequency of end-stage renal disease and the rate of change in GFR compared to placebo over a 6-year period [26] . Experimental evidences suggest that the renoprotective effect of statins is in part mediated by lipid-independent mechanisms. In a prospective study of 86 subjects with diabetic nephropathy, treatment with simvastatin but not cholestyramine resulted in significant reduction in albuminuria and stabilization of GFR despite similar reduction in plasma cholesterol in both groups [27] . In another trial involving 63 proteinuric but normocholesterolaemic subjects, pravastatin therapy reduced proteinuria with no association between degrees of lipid lowering and reduction in proteinuria [28] . Statins decreased pro-inflammatory cytokines, suppressed extracellular matrix production and exhibited inhibitory effects on the Rho/Rho kinase pathway [29,30] . Rho is a G-protein that mediates intracellular signaling with diverse effects on vasculature and renal parenchyma. Dysregulated activity of Rho/Rho kinase results in increase in vascular smooth muscle tone, endothelial dysfunction and oxidative stress, thereby promotes atherosclerosis [31] . In the diabetic kidneys, Rho/Rho kinase upregulates several pathogenic cytokines including tumour growth factor β and connective tissue growth factor as well as generates reactive oxygen species. In this regard, statin blocks isoprenylation which is a pivotal step in the activation of Rho. Aside from its anti-inflammatory and anti-fibrotic properties, statins influences intra-renal haemodynamics by reducing angiotensin and aldosterone levels [32] . In the current analysis, the association between statin use and development of renal dysfunction was independent of baseline lipid levels. As repeated measurements of lipid levels were not available, our results cannot confirm pleiotropic property, if any, of statin in renoprotection. Our study has several limitations. First, due to the observational nature of this study, there might be other uncontrolled factors inherent in subjects taking statins that could affect renal function. Second, while we have quantified the duration of drug exposure, other important parameters such as types of statins and dosages were not examined. Third, we did not examine the relationships between changes in lipid levels and magnitude of risk reduction since not all patients had repeat lipid measurements after start of treatment. Last, serum creatinine on follow-up was not periodically measured in all patients in the cohort but based on available data from a Hong Kong Hospital Authority database, which included all the measurements in all public hospitals. Patients who did not have serum creatinine on follow-up were younger, had shorter duration of diabetes, lower HbA 1c , total cholesterol, systolic blood pressure and smaller BMI. This group of patients were also less likely to have microvascular and macrovascular complications. Therefore, this group of patients was a low risk group and were less likely to have renal complications during follow-up than those who had serum creatinine data on follow-up. Nevertheless, if their serum creatinine levels were periodically monitored, renal complications might have been detected in some of these patients. Despite these limitations, an important strength of our study is the large sample size with comprehensive baseline and longitudinal assessments as well as the relatively long observation period. Although we did not examine the total dosage of statins used, local pharmaco-epidemiological studies suggest that the average daily dosage used in our population was comparable to that recommended by the manufacturer [33] . We have used adjusted time proportion of drug usage to quantify drug exposure to minimise healthy volunteer bias. We applied rigorous statistical methods including use of spline function, a novel method in the Cox model [19] , to correct for confounders, which takes non-linearity into account. In this large prospective cohort of Chinese patients with type 2 diabetes, we showed that the use of statins was associated with 68% risk reduction of renal dysfuncion. Despite aggressive control of risk factors such as blood pressure and blood glucose as well as the use of ACE-inhibitors or ARB, patients with type 2 diabetes have large residual risk for progressive deterioration of renal function [34] and end-stage renal disease [35] . In the absence of large scale clinical trials with hypothesis defined a priori, our findings concord with previous analysis and provides clinically relevant information to guide clinical practice, especially in ethnic groups at high risk for diabetic renal disease. Conflict of interest There are no conflicts of interest. Acknowledgements We thank all medical and nursing staff of the Prince of Wales Hospital Diabetes Centre in recruiting and managing these patients. Special thanks are extended to HA Head Office and Chief Pharmacy Office. Appendix A Risk association of renal dysfunction with use of statins Use of statins during follow-up Hazard ratio 95% confidence interval p Value −2 log Likelihood Use of statins versus non-use of statins a 0.53 0.42–0.66 <0.0001 9235.99 Time proportion of using statins during follow-up b 0.42 0.29–0.62 <0.0001 9253.58 Adjusted time proportion of using statins during follow-up c 0.34 0.21–0.54 <0.0001 6045.60 All models were adjusted for age, sex, BMI, LDL-C, HDL-C, total cholesterol, triglyceride, HbA 1c , SBP, DBP, log 10 (ACR+1), eGFR, duration of diabetes at enrolment, use of oral anti-diabetic drugs and anti-hypertensive drugs other than ACE-inhibitors and ARB at baseline. Spline functions of all the continuous baseline variables were used for better adjustment for possible non-linear associations. a Adjusted for ACE-inhibitors, ARB, insulin and aspirin during follow-up, using “yes/no coding”. b Adjusted for ACE-inhibitors, ARB, insulin and aspirin during follow-up, using “time proportions”. c Adjusted for ACE-inhibitors, ARB, insulin and aspirin during follow-up, using “adjusted time proportions”. References [1] C.T. Valmadrid R. Klein S.E. Moss B.E. Klein The risk of cardiovascular disease mortality associated with microalbuminuria and gross proteinuria in persons with older-onset diabetes mellitus Arch. Intern. Med. 160 2000 1093 1100 [2] H.N. Gerstein J.F. Mann Q. Yi B. Zinman S.F. Dinneen Albuminuria and risk of cardiovascular events, death and heart failure in diabetic and nondiabetic individuals J. Am. Med. Assoc. 286 2001 421 426 [3] J.A. Critchley H.L. Zhao B. Tomlinson W. Leung G.N. Thomas J.C. Chan Management of nephropathy in patients with type 2 diabetes Chin. Med. J. 115 2002 129 135 [4] A.Y. Wu N.C. Kong F.A. de Leon C.Y. Pan T.Y. Tai V.T. Yeung An alarmingly high prevalence of diabetic nephropathy in Asian type 2 diabetic patients: the microalbuminuria prevalence (MAP) study Diabetologia 48 2005 1674 1675 [5] A. Karter A. Ferrara J. Liu H. Moffet L. Ackerson J. Selby Ethnic disparities in diabetic complications in an insured population J. Am. Med. Assoc. 287 2002 2519 2527 [6] P. Muntner J. Coresh J.C. Smith J. Eckfeldt M.J. Klag Plasma lipids and risk of developing renal dysfunction: the atherosclerosis risk in communities study Kidney Int. 58 2000 293 301 [7] C.S. Fox M.G. Larson E.P. Leip B. Culleton Predictors of new onset kidney disease in a community-based population J. Am. Med. Assoc. 291 2004 844 850 [8] E.S. Schaeffner T. Kurth G.C. Curhan R.J. Glynn K.M. Rexrode C. Baigent Cholesterol and the risk of renal dysfunction in apparently healthy men J. Am. Soc. Nephrol. 14 2003 2084 2091 [9] A.O. Luk W.Y. So R.C. Ma A.P. Kong R. Ozaki V.S. Ng Metabolic syndrome predicts new onset of chronic kidney disease in 5829 patients with type 2 diabetes: a 5-year prospective analysis of the Hong Kong Diabetes Registry Diabetes Care 31 2008 2357 2361 [10] V.M. Campese J. Park HMG-CoA reductase inhibitors and the kidney Kidney Int. 71 2007 1215 1222 [11] C.K. Abrass Cellular lipid metabolism and the role of lipids in progressive renal disease Am. J. Nephrol. 24 2004 46 53 [12] B. Bussolati M.C. Deregibus V. Fonsato S. Doublier T. Spatola S. Procida Statins prevent oxidized LDL-induced injury of glomerular podocytes by activating the phosphatidylinositol 3-kinase/AKT-signaling pathway J. Am. Soc. Nephrol. 16 2005 1936 1947 [13] M. Tonelli L. Moye F.M. Sacks T. Cole G.C. Curhan Effect of pravastatin on loss of renal function in people with moderate chronic renal insufficiency and cardiovascular disease J. Am. Soc. Nephrol. 14 2003 1605 1613 [14] R. Collins J. Armitage S. Parish P. Sleigh R. Peto Heart Protection Study Collaborative Group. MRC/BHF heart protection study of cholesterol lowering with simvastatin in 5,963 people with diabetes: a randomized placebo-controlled trial Lancet 361 2003 2005 2016 [15] S. Sandhu N. Wiebe L.F. Fried M. Tonelli Statins for improving renal outcomes: a meta-analysis J. Am. Soc. Nephrol. 17 2006 2006 2016 [16] K. Piwernetz P.D. Home O. Snorgaard M. Antsiferov K. Staehr-Johansen M. Krans Monitoring the targets of the St. Vincent declaration and the implementation of quality management in diabetes care: the DiabCare initiative Diabetic Med 10 1993 371 377 [17] P.C. Tong K.F. Lee W.Y. So M.H. Ng W.B. Chan M.K. Lo White blood cell count is associated with macro- and microvascular complications in Chinese patients with type 2 diabetes Diabetes Care 27 2004 216 222 [18] Y.C. Ma L. Zuo J.H. Chen Q. Luo X.Q. Yu Y. Li Modified glomerular filtration rate estimating equation for Chinese patients with chronic kidney disease J. Am. Soc. Nephrol. 17 2006 2937 2944 [19] F. Harrell Regression Modelling Strategies with applications to Linear Models, Logistic Regression and Survival Analysis 2001 Spinger-Varlag New York, Inc. New York [20] W.Y. So X. Yang R.C. Ma A.P. Kong C.W. Lam C.S. Ho Risk factors in V-shaped risk associations with all-cause mortality in type 2 diabetes — the Hong Kong Diabetes Registry Diabetes Metab. Res. Rev. 24 2008 238 246 [21] X. Yang W. So G.T. Ko R.C. Ma A.P. Kong C.C. Chow Independent associations between low-density lipoprotein cholesterol and cancer among patients with type 2 diabetes mellitus CMAJ 179 2008 427 437 [22] X. Yang W.Y. So R. Ma G. Ko A. Kong C. Lam Effects of albuminuria and renal dysfunction on development of dyslipidaemia in type 2 diabetes—the Hong Kong Diabetes Registry Nephrol. Dial. Transplant. 23 2008 2834 2840 [23] R.I. Horwitz S.M. Horwitz Adherence to treatment and health outcomes Arch. Intern. Med. 153 1993 1863 1868 [24] D.Y. Lin L.J. Wei Z. Ying Checking the Cox model with cumulative sums of martingale-based residuals Biometrika 80 1993 557 572 [25] R. Sukhija Z. Bursac P. Kakar L. Fink C. Fort S. Satwani Effect of statins on the development of renal dysfunction Am. J. Cardiol. 101 2008 975 979 [26] M. Rahman C. Baimbridge B.R. Davis J. Barzilay J.N. Basile M.A. Henriquez Progression of kidney disease in moderately hypercholesterolemic, hypertensive patients randomized to pravastatin versus usual care: a report from the Anti-hypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial (ALLHAT) Am. J. Kidney Dis. 52 2008 412 424 [27] G. Tonolo M. Velnssi E. Brocco C. Abaterusso A. Carraro G. Morgia Simvastatin maintains steady patterns of GFR and improves AER and expression of slit diaphragm proteins in type II diabetes Kidney Int. 70 2006 177 186 [28] T.M. Lee S.F. Su C.H. Tsai Effects of pravastatin on proteinuria in patients with well-controlled hypertension Hypertension 40 2000 67 73 [29] A. Gojo K. Utsunomiya K. Taniguchi T. Yokota S. Ishizawa Y. Kanazawa The Rho-kinase inhibitor, fasudil, attenuates diabetic nephropathy in streptozotocin-induced diabetic rats Eur. J. Pharm. 568 2007 242 247 [30] S. Patel R.M. Mason J. Suzuki A. Imaizumi T. Kamimura Z. Zhang Inhibitory effect of statins on renal epithelial-to-mesenchymal transition Am. J. Nephrol. 26 2006 381 387 [31] K. Budzyn P.D. Marley C.G. Sobey Targeting Rho and Rho-kinase in the treatment of cardiovascular disease Trends Pharm. Sci. 27 2006 97 104 [32] M.A. Bayorh A.A. Ganafa D. Eatman M. Walton G.Z. Feuerstein Simvastatin and losartan enhance nitric oxide and reduce oxidative stress in salt-induced hypertension Am. J. Hypertens. 18 2005 1496 1502 [33] A. Wierzbicki D. Mikhailidis R. Wray M. Schacter R. Cramb W. Simpson Statin-fibrate combination: therapy for hyperlipidemia: a review Curr. Med. Res. Opin. 19 2003 155 168 [34] H.H. Parving H. Lehnert J. Brochner-Mortensen R. Gomis S. Anderesen P. Arner The effect of irbesartan on the development of diabetic nephropathy in patients with type 2 diabetes N. Engl. J. Med. 345 2001 870 878 [35] B.M. Brenner M.E. Cooper D. de Zeeuw W.F. Keane W.E. Mitch H.H. Parving Effects of losartan on renal and cardiovascular outcomes in patients with type 2 diabetes and nephropathy N. Engl. J. Med. 345 2001 861 869