Effects of Atorvastatin and Rosuvastatin on Renal Function in Patients With Type 2 Diabetes Mellitus

Accepted Manuscript Effects of Atorvastatin and Rosuvastatin on Renal Function in Patients with Type 2 Diabetes Mellitus Chao-Lun Lai, MD, PhD, Hsu-Wen Chou, PhD, Kinwei Arnold Chan, MD, ScD, MeiShu Lai, MD, PhD PII:

S0002-9149(14)02272-3

DOI:

10.1016/j.amjcard.2014.12.009

Reference:

AJC 20843

To appear in:

The American Journal of Cardiology

Received Date: 28 September 2014 Revised Date:

3 December 2014

Accepted Date: 5 December 2014

Please cite this article as: Lai C-L, Chou H-W, Chan KA, Lai M-S, Effects of Atorvastatin and Rosuvastatin on Renal Function in Patients with Type 2 Diabetes Mellitus, The American Journal of Cardiology (2015), doi: 10.1016/j.amjcard.2014.12.009. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

Effects of Atorvastatin and Rosuvastatin on Renal Function in Patients with Type 2 Diabetes Mellitus Chao-Lun Lai, MD, PhDa,b,c, Hsu-Wen Chou, PhDc,

a

RI PT

Kinwei Arnold Chan, MD, ScDd,e, Mei-Shu Lai, MD, PhDc,f,*

Department of Internal Medicine and Center for Critical Care Medicine,

Department of Internal Medicine, College of Medicine, National Taiwan University,

M AN U

b

SC

National Taiwan University Hospital Hsin-Chu Branch, Hsin-Chu, Taiwan;

Taipei, Taiwan; cInstitute of Epidemiology and Preventive Medicine, College of Public Health, National Taiwan University, Taipei, Taiwan; dDepartment of Medical Research, National Taiwan University Hospital, Taipei, Taiwan; eGraduate Institute of

TE D

Oncology, College of Medicine, National Taiwan University, Taiwan; fCenter of Comparative Effectiveness Research, National Center of Excellence for Clinical Trial

EP

and Research, National Taiwan University Hospital, Taipei, Taiwan

AC C

Running Title: Atorvastatin, rosuvastatin and renal function *Corresponding author: Dr. Mei-Shu Lai, Institute of Epidemiology and Preventive Medicine, College of Public Health, National Taiwan University, Taipei, Taiwan; Room 518, No. 17, Hsu Chow Road, Taipei, Taiwan, 100; Tel: +886-2-3366-8018; Fax: +886-2351-1955 E-mail address: [email protected] (M-S Lai)

1

ACCEPTED MANUSCRIPT Abstract We performed this population-based study to investigate the effect of atorvastatin and

RI PT

rosuvastatin on renal function in patients with type 2 diabetes. From the Taiwan National Health Insurance (NHI) Pay-for-Performance (P4P) program for diabetes

mellitus database, 2006-2009, type 2 diabetic patients aged 40-100 years with the first

SC

prescription of atorvastatin or rosuvastatin were identified. All the data were linked to

M AN U

the NHI claims database, 2000-2010, to construct a longitudinal health care data. The Modification of Diet in Renal Disease (MDRD) equation was used to calculate the estimated glomerular filtration rate (eGFR) and the eGFR between baseline and the end of follow-up (maximum 2 years) was compared. Totally, 3601 new users of

TE D

atorvastatin and 1968 new users of rosuvastatin were included. The median follow-up was 238 days in atorvastatin users and 210 days in rosuvastatin users. The eGFR at

EP

baseline was 72.3 ± 25.9 mL/min/1.73 m2 in atorvastatin users and 73.7 ± 27.3

AC C

mL/min/1.73 m2 in rosuvastatin users. In both statin groups, we found no significant change in eGFR (+0.1 mL/min/1.73 m2, 95% confidence interval [CI] -0.4 to 0.7, p=0.62 in atorvastatin users; -0.1 mL/min/1.73 m2, 95% CI -0.8 to 0.6, p=0.77 in rosuvastatin users). In conclusion, neither treatment with atorvastatin nor rosuvastatin was associated with a significant change of renal function in type 2 diabetic patients. Key Words: statin; atorvastatin; rosuvastatin; kidney; glomerular filtration rate.

2

ACCEPTED MANUSCRIPT Several studies have refuted the adverse effect of statin therapy on renal function.1-9 However, a few large-scale observational studies reported an association

RI PT

between statins, especially high potency statins, and acute kidney injury.10-12 The aim of this study was to evaluate the effects of two high potency statins, atorvastatin and rosuvastatin, on renal function in adult type 2 diabetic patients in an ethnic Chinese

SC

population. A retrospective cohort study using the longitudinal National Health

M AN U

Insurance (NHI) claims data of Taiwan was conducted. Methods

Taiwan launched a single-payer NHI Program since 1995. As of 2007, > 98% of the total Taiwanese population is covered by the program. Patient identification

TE D

numbers, gender, birthdays, dates of hospital admission and discharge, diagnoses and drugs dispensed are available in the NHI claims database. The diagnoses are coded

EP

according to the International Classification of Diseases Ninth Revision Clinical

AC C

Modification (ICD-9-CM) system. Under the NHI system, most healthcare services are reimbursed on a fee-for-service basis. Since 2001, the NHI program has implemented a Pay-for-Performance (P4P) program for diabetes mellitus.13 Hospitals and community clinics with qualified physicians can voluntarily apply to participate in the NHI P4P program. The participating physicians then can enroll patients in the NHI P4P program. The NHI P4P program reimburses participating clinicians with

3

ACCEPTED MANUSCRIPT additional physician fees and case management fees in addition to regular reimbursement for healthcare services to increase comprehensive follow-up visits

RI PT

including annual diabetes-specific examinations such as laboratory tests. Clinical characteristics of the enrolled patients, including habits of smoking and alcohol

drinking, body height, body weight, body mass index, and key follow-up laboratory

SC

data concerning diabetic care, were reported by the hospitals themselves periodically

M AN U

and were entered into the P4P-specific database automatically.13

Our data came from two databases. One database contained information collected from the NHI P4P program for the period from January 2006 to December 2010 and was used to obtain major clinical characteristics of patients and physicians.

TE D

The patients’ NHI P4P records were then linked to the NHI claims database through 2000 to 2010 by patient identification number to identify comorbidities, drug

EP

exposures and medical utilizations during the baseline period. To comply with privacy

AC C

regulations, personal identifiers were encrypted and all data were analyzed anonymously. The study protocol was approved by the Institutional Review Board of the National Taiwan University Hospital. All type 2 diabetic patients aged between 40 and 100 years with the first prescription of atorvastatin or rosuvastatin were identified from the NHI P4P database through 2006 to 2009 and their longitudinal claims data were extracted from the NHI

4

ACCEPTED MANUSCRIPT claims database. The date on which the first statin was prescribed was operationally set as the index date. Background characteristics and comorbidities of the enrolled

RI PT

subjects were assessed within the baseline 6-month period prior to the index date. Patients who had received any type of statin or any type of renal replacement therapy during the baseline 6-month period were excluded. Besides, we also excluded subjects

SC

receiving > 1 statin at the index date, subjects without baseline serum creatinine level,

M AN U

and subjects with abnormally high baseline serum alanine aminotransferase (ALT) level (>1000 IU/L). Only subjects with at least one report of serum creatinine level during the follow-up period were retained in the study cohort.

All subjects were followed from their index dates until they had a prescription of

TE D

another type of statin, no refill for statins after 1.5 times the duration of their last prescription of statins, died, withdrawal from health insurance coverage, initiation of

EP

any renal replacement therapy, maximum 2 years of follow-up, or reached the end of

AC C

the study at 31st December 2010, whichever came first. The abbreviated Modification of Diet in Renal Disease (MDRD) equation14 was

used to calculate the estimated glomerular filtration rate (eGFR). The baseline eGFR was calculated based on the last record of serum creatinine level prior to the index date and the follow-up eGFR was calculated based on the last record of serum creatinine level prior to the end of follow-up (last observation carried forward [LOCF]

5

ACCEPTED MANUSCRIPT approach). The change of eGFR between end of follow-up and baseline was calculated for each subject accordingly.

RI PT

In addition to baseline age at prescription of specific study drugs and gender, we assessed the potential confounders such as smoking, alcohol drinking, body mass index, key laboratory data at baseline, comorbidities, history of exposure to

SC

nephrotoxic drugs, specialty of prescribing physicians and medical utilizations within

M AN U

the baseline 6-month period. The laboratory data (also using LOCF approach) included fasting blood glucose, glycated hemoglobin (HbA1c), ALT, uric acid, low-density lipoprotein cholesterol (LDL-C), and baseline eGFR. The stage of chronic kidney disease (CKD) was defined according to baseline eGFR.15 Most of the

TE D

comorbidities were extracted based on the revised ICD-9-CM coding algorithms for Elixhauser Index16 except for myocardial infarction (410.x, 412.x) and

EP

cerebrovascular disease (362.34, 430.x-438.x). Only those comorbidities with

AC C

prevalence of >1% were retained in the analysis (Table 1). The list of potential nephrotoxic drugs were acetaminophen, aspirin, non-steroidal anti-inflammatory drugs, allopurinol, penicillins, cephalosporins, sulfonamides, aminoglycosides, quinolones, thiazide diuretics, loop diuretics, angiotensin-converting-enzyme inhibitors (ACEIs), angiotensin receptor blockers (ARBs), H2-receptor antagonists, and proton pump inhibitors.17,18 The specialty of prescribing physicians included

6

ACCEPTED MANUSCRIPT family medicine, internal medicine, cardiology, nephrology and endocrinology. Medical utilizations were defined as number of outpatient visits and number of

RI PT

hospitalizations during baseline 6-month period. Categorical data are presented in contingency tables and continuous variables are presented as mean ± standard deviation. The χ2 test and two-sample t-test were used

SC

to test differences between atorvastatin users and rosuvastatin users. The eGFR

M AN U

between end of follow-up and baseline period within each statin group was compared by paired t-test. Multiple linear regression model was applied to compare the change of eGFR between two statin groups. Baseline eGFR and all the potential confounders listed in Table 1 except serum creatinine level were included as regressors. The

TE D

primary analysis was repeated for different subgroups such as genders, age groups, HbA1c levels, CKD stages, and exposure to ACEIs/ARBs. We also performed a

EP

stratified analysis according to the maximum daily dosages of statins as patients ever

AC C

receiving atorvastatin 80 mg/daily or rosuvastatin 40 mg/daily in any prescription during the follow-up period were classified as high dose users and others were classified as low dose users. All analyses were performed with SAS software, version 9.2 (SAS Institute, Inc., Cary, North Carolina). All p values reported are two-sided and the significant level was set at <0.05.

7

ACCEPTED MANUSCRIPT Results

Totally, 3601 incident users of atorvastatin and 1968 incident users of

RI PT

rosuvastatin were included in this study (Figure 1). Compared with rosuvastatin users, atorvastatin users were older and were more likely to be women, have habits of

SC

cigarette smoking and alcohol consumption, and have lower levels of fasting glucose, HbA1c, and LDL-C. Atorvastatin users were more likely to have congestive heart

M AN U

failure, paralysis, chronic pulmonary disease, collagen vascular disease, depression, exposure to aspirin and ACEIs within baseline 6-month period than rosuvastatin users. Atorvastatin was more likely to be prescribed by family physicians while rosuvastatin

TE D

was more likely to be prescribed by endocrinologists (Table 1).

At baseline, the mean eGFR was 72.3 ± 25.9 mL/min/1.73m2 in atorvastatin

EP

users and 73.7 ± 27.3 mL/min/1.73m2 in rosuvastatin users. The mean daily dosage

AC C

was 13.3 ± 10.4 mg for atorvastatin and 8.9 ± 4.7 mg for rosuvastatin. After a median follow-up of 238 days in atorvastatin users (interquartile range [IQR]: 124 days, 461 days) and 210 days in rosuvastatin users (IQR: 116 days, 402 days), we observed no change in eGFR within either statin group (change of eGFR: +0.1 mL/min/1.73m2, 95% confidence interval [CI]: -0.4 ~ 0.7 mL/min/1.73m2 in atorvastatin users; and -0.1 mL/min/1.73m2, 95% CI: -0.8 ~ 0.6 mL/min/1.73m2 in

8

ACCEPTED MANUSCRIPT rosuvastatin users). After accounting for baseline differences, no significant difference existed in the change of eGFR between atorvastatin users and rosuvastatin users

RI PT

(adjusted p=0.27) (Table 2 and Figure 2).

The lack of difference in change of eGFR between atorvastatin users and

SC

rosuvastatin users was consistent in most of the pre-specified subgroups (Figure 3), including patients with different genders, different age groups, different levels of

M AN U

HbA1c at baseline, ever/never exposure to ACEIs/ARBs at baseline, high dose users (mean daily dosage 52.2 ± 26.3 mg in atorvastatin users and 22.2 ± 13.3 mg in rosuvastatin users) and low dose users (mean daily dosage 13.1 ± 9.7 mg in

TE D

atorvastatin users and 8.8 ± 4.5 mg in rosuvastatin users). Both atorvastatin users and rosuvastatin users showed a similar phenomenon of deterioration in eGFR in patients

EP

with underlying CKD in stage 1-2 and improvement in eGFR in patients with

AC C

underlying CKD in stage 3 and stage 4-5.

Discussion

In this ethnic Chinese type 2 diabetic population, the effect of two high potency

statins, atorvastatin and rosuvastatin, on renal function was investigated using insurance claims database including key laboratory tests related to diabetic care. We found that both atorvastatin and rosuvastatin were not associated with a significant

9

ACCEPTED MANUSCRIPT change of eGFR. Although there was deterioration in eGFR in patients with underlying CKD in stage 1-2, significant improvement in eGFR in patients with

RI PT

underlying CKD in stage 3 and stage 4-5 were noted in both high potency statin groups.

Much evidence concerning renal effect of statins comes from clinical trials. Both

SC

the Heart Protection Study,8 and the GREACE study7 found improvement in

M AN U

creatinine clearance after statin treatment. Several meta-analysis with pooling of results from clinical trials also found renal protective effect in high potency statins including atorvastatin and rosuvastatin.2,4-6,9 Even though our study did not show a significant improvement in renal function in both atorvastatin and rosuvastatin groups,

TE D

our findings were in line with the results from clinical trials that atorvastatin and rosuvastatin did not influence renal function adversely.

EP

In contrast to clinical trials, several large-scale observational studies10,11

AC C

including one study from our colleagues12 using administrative data reported association between acute kidney injury and treatment with statins, especially high potency ones. Although these studies possessed huge case number, they relied on only diagnoses in claims database without details of laboratory tests. Our study used a specific claims database with results of key laboratory tests including serum creatinine level for calculation of eGFR for each subject. The different conclusions

10

ACCEPTED MANUSCRIPT between our study and previous observational studies10-12 underscore the limitation of observational studies depending on only diagnoses in administrative databases.

RI PT

The influence of statins on renal function in patients with CKD is of interest. In our study, both atorvastatin and rosuvastatin were associated with a decline of eGFR

in patients with stage 1-2 CKD. Conversely, a significant improvement in eGFR was

SC

found in patients with stage 3 and stage 4-5 CKD in both statin groups. Nevertheless,

M AN U

two Cochrane reviews stated that statins did not impact on the decline in renal function in patients with CKD.3,19 Whether our findings reveal a different impact of statin therapy in patients with different stages of CKD or nothing but the phenomenon of “statistic regression towards the mean”20,21 deserves further investigation.

TE D

Some limitations of our study have to be acknowledged. Firstly, most clinical trials involving statin therapy have long follow-up durations of up to 4-5 years.7,8 In

EP

contrast, our study had a relatively short median follow-up duration of about 200 days

AC C

because changes in prescriptions of statins are common in Taiwan22 and we defined the follow-up being censored if the patients had been prescribed a different type of statin. However, previous observational studies reported that high potency statins were associated with acute kidney injury which could develop within 120 days11 to 180 days10 after starting statin treatment. Thus, the median follow-up duration of >200 days in our study was comparable to those of the two observational studies10,11

11

ACCEPTED MANUSCRIPT mentioned previously and our study seemed sufficient to provide clinical evidence. Secondly, because the physicians and patients voluntarily apply for participation in

RI PT

the NHI P4P program for diabetes mellitus in Taiwan, patients enrolled in the P4P program may have better compliance and better control of diabetes than general

population. However, the same situation exists in all clinical trials. Thirdly, most of

SC

the subjects included in our study were in stage 1-3 CKD. The sparse case number in

M AN U

stage 4-5 CKD led to very wide 95% CIs regarding the estimates of change of eGFR in subgroup analysis. Acknowledgement

The authors would like to express their appreciation for the support provided by

TE D

Shu-Ting Chen for editing and helping with the SAS programming. Disclosures

EP

The authors have no conflicts of interest to disclose.

AC C

Sources of Funding

This study was supported by a grant from the National Health Insurance

Administration, Ministry of Health and Welfare, Executive Yuan, Taiwan (DOH101-NH-9014). The funding agency did not have any input in study design, data analysis, and interpretation of findings, or in the decision to submit the manuscript for publication.

12

ACCEPTED MANUSCRIPT 1.

Fassett RG, Robertson IK, Ball MJ, Geraghty DP, Coombes JS. Effect of

atorvastatin on kidney function in chronic kidney disease: A randomised double-blind

2.

RI PT

placebo-controlled trial. Atherosclerosis 2010;213:218-224. Bangalore S, Fayyad R, Hovingh GK, Laskey R, Vogt L, DeMicco DA, Waters

DD, Treating to New Targets Steering Committee and Investigators. Statin and the

SC

Risk of Renal-Related Serious Adverse Events: Analysis from the IDEAL, TNT,

M AN U

CARDS, ASPEN, SPARCL, and Other Placebo-Controlled Trials. Am J Cardiol 2014;113:2018-2020. 3.

Palmer SC, Navaneethan SD, Craig J, Johnson DW, Perkovic V, Hegbrant J,

Strippoli GF. HMG CoA reductase inhibitors (statins) for people with chronic kidney

4.

TE D

disease not requiring dialysis. Cochrane Database Syst Rev 2014;5:CD007784. Savarese G, Musella F, Volpe M, Paneni F, Perrone-Filardi P. Effects of

EP

atorvastatin and rosuvastatin on renal function: A meta-analysis. Int J Cardiol

5.

AC C

2013;167:2482-2489.

Takagi H, Umemoto T. A meta-analysis of randomized trials for effects of

atorvastatin on renal function in chronic kidney disease. Int J Cardiol 2011;152:242-244. 6.

Wu Y, Wang Y, An C, Dong Z, Liu H, Zhang Y, Zhang M, An F. Effects of

rosuvastatin and atorvastatin on renal function: meta-analysis. Circ J

13

ACCEPTED MANUSCRIPT 2012;76:1259-1266. 7.

Athyros VG, Mikhailidis DP, Papageorgiou AA, Symeonidis AN, Pehlivanidis

RI PT

AN, Bouloukos VI, Elisaf M. The effect of statins versus untreated dyslipidaemia on renal function in patients with coronary heart disease. A subgroup analysis of the

Greek atorvastatin and coronary heart disease evaluation (GREACE) study. J Clin

Collins R, Armitage J, Parish S, Sleigh P, Peto R, Heart Protection Study

M AN U

8.

SC

Pathol 2004;57:728-734.

Collaborative Group. MRC/BHF Heart Protection Study of cholesterol-lowering with simvastatin in 5963 people with diabetes: a randomised placebo-controlled trial. Lancet 2003;361:2005-2016.

Sandhu S, Wiebe N, Fried LF, Tonelli M. Statins for Improving Renal Outcomes:

TE D

9.

A Meta-Analysis. J Am Soc Nephrol 2006;17:2006-2016.

EP

10. Corrao G, Soranna D, Casula M, Merlino L, Porcellini MG, Catapano AL.

AC C

High-potency statins increase the risk of acute kidney injury: evidence from a large population-based study. Atherosclerosis 2014;234:224-229. 11. Dormuth CR, Hemmelgarn BR, Paterson JM, James MT, Teare GF, Raymond CB, Lafrance JP, Levy A, Garg AX, Ernst P, Canadian Network for Observational Drug Effect Studies (CNODES). Use of high potency statins and rates of admission for acute kidney injury: multicenter, retrospective observational analysis of

14

ACCEPTED MANUSCRIPT administrative databases. BMJ 2013;346:f880. 12. Chung YH, Lee YC, Chang CH, Lin MS, Lin JW, Lai MS. Statins of high versus

Pharmacoepidemiol Drug Saf 2013;22:583-592.

RI PT

low cholesterol-lowering efficacy and the development of severe renal failure.

13. Lee TT, Cheng SH, Chen CC, Lai MS. A pay-for-performance program for

SC

diabetes care in Taiwan: a preliminary assessment. Am J Manag Care 2010;16:65-69.

M AN U

14. Levey AS, Coresh J, Balk E, Kausz AT, Levin A, Steffes MW, Hogg RJ, Perrone RD, Lau J, Eknoyan G, National Kidney Foundation. National Kidney Foundation Practice Guidelines for Chronic Kidney Disease: Evaluation, Classification, and Stratification. Ann Intern Med 2003;139:137-147.

TE D

15. National Kidney Foundation. K/DOQI clinical practice guidelines for chronic kidney disease: evaluation, classification, and stratification. Am J Kidney Dis

EP

2002;39:S46-75.

AC C

16. Quan H, Sundararajan V, Halfon P, Fong A, Burnand B, Luthi JC, Saunders LD, Beck CA, Feasby TE, Ghali WA. Coding algorithms for defining comorbidities in ICD-9-CM and ICD-10 administrative data. Med Care 2005;43:1130-1139. 17. Naughton CA. Drug-induced nephrotoxicity. Am Fam Physician 2008;78:743-750. 18. Pannu N, Nadim MK. An overview of drug-induced acute kidney injury. Crit

15

ACCEPTED MANUSCRIPT Care Med 2008;36:S216-223. 19. Navaneethan SD, Pansini F, Perkovic V, Manno C, Pellegrini F, Johnson DW,

RI PT

Craig JC, Strippoli GF. HMG CoA reductase inhibitors (statins) for people with chronic kidney disease not requiring dialysis. Cochrane Database Syst Rev 2009;2:CD007784.

SC

20. Bland JM, Altman DG. Regression towards the mean. BMJ 1994;308:1499.

M AN U

21. Bland JM, Altman DG. Some examples of regression towards the mean. BMJ 1994;309:780.

22. Lai CL, Shau WY, Chang CH, Chen MF, Lai MS. Statin Use and Cataract Surgery: A Nationwide Retrospective Cohort Study in Elderly Ethnic Chinese Patients.

AC C

EP

TE D

Drug Saf 2013;36:1017-1024.

16

ACCEPTED MANUSCRIPT Figure Legends Figure 1. Patient flow chart. ALT = alanine aminotransferase; RRT = renal

RI PT

replacement therapy.

Figure 2. Mean with 95% confidence interval of the change of eGFR in atorvastatin

SC

and rosuvastatin groups. eGFR = estimated glomerular filtration rate (mL/min/1.73

M AN U

m2).

Figure 3. Effects of atorvastatin and rosuvastatin on change of estimated glomerular filtration rate in subgroups. ACEI = angiotensin-converting-enzyme inhibitor; ARB =

TE D

angiotensin receptor blocker; CI = confidence interval; CKD = chronic kidney disease; ΔeGFR = change of estimated glomerular filtration rate; HbA1c = glycated

EP

hemoglobin. *comparison between atorvastatin group and rosuvastatin group by

AC C

multiple linear regression model with adjustment for potential confounders.

17

ACCEPTED MANUSCRIPT Table 1 Baseline characteristics of enrolled subjects. Variable

Atorvastatin (n=3601) 62.8 ± 10.5

Rosuvastatin (n=1968) 61.7 ± 10.2

P value

<0.0001 1683 ( 47% ) 980 ( 50% ) 0.029 1672 ( 46% ) 835 ( 42% ) 0.004 1679 ( 47% ) 813 ( 41% ) <0.001 26.4 ± 3.7 26.5 ± 3.9 0.27

Fasting glucose (mg/dL) Glycated hemoglobin (%) Alanine aminotransferase (IU/L) Uric acid (md/dL) LDL-C (mg/dL) Creatinine (mg/dL) eGFR (mL/min/1.73m2) Chronic kidney disease stage 1 2 3 4 5 Comorbidities Myocardial infarction Cerebrovascular disease Congestive heart failure Cardiac arrhythmia Valvular heart disease Peripheral vascular disorder Hypertension Paralysis Other neurological disorder Chronic pulmonary disease Hypothyroidism Liver disease Peptic ulcer disease, excluding bleeding

151.8 ± 50.0 7.9 ± 1.6 29.1 ± 22.5

TE D

EP

AC C

155.1 ± 57.4 8.0 ± 1.7 28.2 ± 19.6 6.2 ± 1.6

SC 1.8 32.2 0.5 25.9

M AN U

6.3 ± 136.4 ± 1.1 ± 72.3 ±

RI PT

Age (years) Men Smoker Alcohol use Body mass index (kg/m2) Baseline

139.7 ± 34.5 1.1 ± 1.1 73.7 ± 27.3

0.037 0.019 0.25 0.36 <0.001 0.31 0.06 0.37

856 1608 964 152 21

( 24% ) 476 ( 24% ) ( 45% ) 876 ( 45% ) ( 27% ) 541 ( 27% ) ( 4% ) 62 ( 3% ) ( 1% ) 13 ( 1% )

22 236 145 117 51 83 1981 25 42 277 38 296 265

( 1% ) 14 ( 1% ) 0.65 ( 7% ) 128 ( 7% ) 0.94 ( 4% ) 40 ( 2% ) <0.0001 ( 3% ) 74 ( 4% ) 0.32 ( 1% ) 23 ( 1% ) 0.44 ( 2% ) 41 ( 2% ) 0.59 ( 55% ) 1046 ( 53% ) 0.18 ( 1% ) 5 ( 0% ) 0.035 ( 1% ) 18 ( 1% ) 0.38 ( 8% ) 111 ( 6% ) 0.004 ( 1% ) 31 ( 2% ) 0.09 ( 8% ) 173 ( 9% ) 0.46 ( 7% ) 125 ( 6% ) 0.16 (To be continued in the next page)

ACCEPTED MANUSCRIPT Table 1 Baseline characteristics of enrolled subjects. (Continued from the previous page) Variable

Atorvastatin (n=3601)

Cancer Collagen vascular disease Psychosis Depression Exposure to nephrotoxic drugs Acetaminophen Aspirin NSAIDs Allopurinol Penicillins Cephalosporins

131 109 26 116

( ( ( (

1559 131 1693 104 561 648

( 43% ) 804 ( 4% ) 44 ( 47% ) 936 ( 3% ) 48 ( 16% ) 294

AC C

EP

( ( ( (

3% 2% 1% 2%

) ) ) )

0.25 0.015 0.18 0.035

( 41% ) ( 2% ) ( 48% ) ( 2% ) ( 15% )

0.08 0.004 0.70 0.33 0.53

RI PT

60 38 21 44

P value

SC

) ) ) )

M AN U

TE D

Sulfonamides Aminoglycosides Quinolones Thiazide diuretics Loop diuretics ACEIs Angiotensin receptor blockers H2-receptor antagonists Proton pump inhibitors Specialty of prescribing physicians Family medicine Internal medicine Cardiology Nephrology Endocrinology Medical utilizations Outpatient visits Hospitalizations

4% 3% 1% 3%

Rosuvastatin (n=1968)

281 121 117 206 281 666 1207 501 136

( ( ( ( ( ( ( ( ( (

18% 8% 3% 3% 6% 8% 18% 34% 14% 4%

456 580 223 79 1930

( 13% ) 71 ( 16% ) 290 ( 6% ) 124 ( 2% ) 52 ( 54% ) 1391

14.9 ± 11.9 0.1 ± 0.4

) ) ) ) ) ) ) ) ) )

355 159 65 67 107 159 316 639 253 87

( ( ( ( ( ( ( ( ( (

18% 8% 3% 3% 5% 8% 16% 32% 13% 4%

) ) ) ) ) ) ) ) ) )

0.97 0.72 0.91 0.76 0.66 0.72 0.023 0.43 0.27 0.24 <0.0001

( 4% ) ( 15% ) ( 6% ) ( 3% ) ( 71% )

14.4 ± 11.5 0.1 ± 0.4

0.16 0.55

Values are expressed as mean ± standard deviation or number (percentage). ACEI = angiotensin-converting-enzyme inhibitor; eGFR = estimated glomerular filtration rate; LDL-C = low-density lipoprotein cholesterol; NSAID = non-steroidal anti-inflammatory drug.

ACCEPTED MANUSCRIPT Table 2 Comparison of estimated glomerular filtration rate between study groups. eGFR

Atorvastatin

Rosuvastatin

(mL/min/1.73m2)

mean (95% CI)

mean (95% CI)

72.3 ( 71.4 - 73.1 ) 72.4 ( 71.5 - 73.3 )

73.7 ( 72.5 - 74.9 ) 0.06 73.6 ( 72.4 - 74.8 ) 0.13

0.1 ( -0.4 -

-0.1 ( -0.8 -

0.7

0.62

)

0.77

0.6

)

RI PT

Baseline End of follow-up Change Crude p‡

Crude Adjusted p* p†

0.27

AC C

EP

TE D

M AN U

SC

CI = confidence interval; eGFR = estimated glomerular filtration rate. *comparison between atorvastatin group and rosuvastatin group by two-sample t-test. †comparison between atorvastatin group and rosuvastatin group by multiple linear regression model with adjustment for potential confounders. ‡comparison between end of follow-up and baseline by paired t-test.

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT