Statin Therapy in Non-ischemic Cardiomyopathy: evidence from a Population Based Cohort and Mendelian Randomization Analysis

doi:10.21203/rs.3.rs-4270769/v1

Aims Statin therapy is widely used for atherosclerotic cardiovascular disease. However, evidence of benefits in patients with non-ischemic cardiomyopathy (NICM) remains controversial. Thus, this study aimed to examine the association between statin use and mortality risk in NICM patients and investigate the causal role of statin therapy on NICM.

Methods and Results We included NICM patients from a heart failure (HF) surveillance cohort abstracted from ARIC study. Kaplan-Meier estimates and logistic regression analysis were used to investigate the associated risk. In mendelian randomization (MR) analysis, summary-level GWAS data for statins SNPs were obtained through UK Biobank while data for NICM were obtained from FinnGen GWAS. Five MR methods were used to investigate the potential causality between statin therapy and NICM. 372 NICM patients were included in the population-based cohort, with 44% (164/372) reported as statin users. Overall, the 28-day and 1-year mortality rate were 5.11% (19/372) and 27.96% (104/372), respectively. In logistic regression analysis, statin use was associated with lower risk of 28-day (Odds Ratio [OR] 0.14, P=0.009) and 1-year mortality (OR 0.81, P=0.37) in the non-adjusted models. In MR analysis, estimates were concordant and similar in size in all MR methods, supporting a protective effect of statin use against NICM. The Cochran’s Q test suggested no evidence of heterogeneity in the instrumental variables while MR-Egger test showed no sign of directional pleiotropy.

ConclusionThis study provides evidence of a beneficial effect of statin use on NICM. This could have significant implications for current HF management in NICM patients.

Non-ischemic cardiomyopathy

heart failure

statin

Mendelian Randomization

Non-ischemic cardiomyopathies (NICM) include a heterogeneous group of myocardium diseases free of ischemic causes and usually manifest as ventricular dilatation and systolic dysfunction (1, 2). The major cause of death in NICM patients were progressive heart failure (HF) and sudden cardiac death (SCD) secondary to ventricular arrhythmias. Thus, except for targeting the specific causes, optimal managements for HF and arrhythmias are fundermental to mortality risk reduction in this group of patients.

Angiotensin converting enzyme inhibitors (ACEI), beta-blockers and mineralocorticoid receptor antagonist (MRA) have been recommended as evidence-based therapies for HF (3). Although statin therapy has been reported to have survival benefits, current HF guidelines do not recommend its routine use unless comorbiding with other indications, such as established coronary artery disease or hypercholesterolemia (4).

The benefit of statin therapy in HF remained disputed. A 2014 meta-analysis involving 45,110 HF patients showed that statin use was associated with reductions in all-cause mortality and rehospitalization for HF (5). However, data from two large randomized trials (CORONA and GISSI-HF) suggested no benefit of statin use in HF patients (6, 7). Some studies even reported detrimental role for statin use when presenting with low cholesterol levels (8).

It is unclear whether the inconsistent findings were attributed to different etiologies of HF. For ischemic HF, the improvement in cholesterol profiles and regression of atherosclerosis are well-accepted mechanisms for improved outcomes (4). However, except for cholesterol lowering effect, statins would reduce the synthesis of coenzyme Q10 and selenoprotein, having a chance for cardiomyopathy deterioration (9, 10). Thus, the question of whether patients with NICM benefit from statin therapy remains to be answered.

To better delineate the effect of stain therapy in NICM, we restricted our study to NICM patients and evaluated the mortality risk associated with statin use. In addition, as Mendelian Randomization (MR) analysis uses genetic variants as instrumental tools and is less likely to be biased by confounding or reverse causation (11), we further performed MR analysis to help investigate the causal role of statin therapy on NICM.

Population based cohort

The Atherosclerosis Risk in Communities Study (ARIC) study is an observational, population-based cohort recruiting adults from 4 US communities: Forsyth County, North Carolina; Washington County, Maryland; Jackson, Mississippi; and Minneapolis, Minnesota. Study participants have been prospectively followed since the baseline visit (1987 to 1989)(12). From 2005 through 2014, medical records for cohort residents ≥55 years of age hospitalized with heart failure were abstracted and formed a HF community surveillance study (13, 14).

Demographics (age, sex, smoking status, BMI, systolic and diastolic blood pressure, heart rate), medical histories (history of hypertension, diabetes), and laboratory values (BNP, hemoglobin) were collected from the hospital record by trained abstractors. For the purposes of this analysis, the last laboratory values were analyzed as suggested in previous study (13). Left ventricular ejection fraction (EF) was prioritized first for inpatient transthoracic echocardiography reports and replaced with EF obtained from other imaging methods if even the within 2-year transthoracic echocardiography reports was not available. Patients were defined as statin users if they answered “yes” for the question of “Lipid Lowering Statins At hospital discharge” and non-users if they answered “no”. 28-day and 1-year mortality were considered as the outcome of the population study. All cause deaths within 28 days and within 1 year were ascertained by linking patients records within the national Death Index. NICM herein was referred to idiopathic or dilated cardiomyopathy.

Initially, 496 patients were identified as having idiopathic or dilated cardiomyopathy. Patients were excluded if comorbiding with ischemic cardiomyopathy (n=66), with missing information on statins use (n=17) or without outcome data for 28-day or 1-year mortality (n=41).

All research activities are approved by local institutional review boards from the 4 ARIC communities and all participants were received the written informed consent from the original study. Informed consent was not required as all data used in this study were anonymized.

Statistical analysis

Baseline characteristics were compared between statin users and non-users. Data were presented as mean± SD (standard deviation) or median ± IQR (interquartile range) for continuous variables and number (percentage) for categorical variables. The Kruskal-Wallis or student t-test were used for continuous variables and Chi-square test for categorical variables.

Kaplan-Meier estimates were used to evaluate the 28-day and 1-year mortality risk between statin users and non-users. In addition, logistic regression analysis was applied to calculate the odds ratios (ORs) and 95% confidence intervals (CI) of the mortality risk with statin use. We constructed two models. One was not adjusted (model 1) and the other was adjusted for age, sex, history of hypertension, diabetes, body mass index, systolic blood pressure, diastolic blood pressure and smoking status (model 2).

Mendelian Randomization Analysis

MR uses publicly available summary-level GWAS data. Data for statins SNPs were obtained through UK Biobank, with 462933 individuals’ information of statin use (atorvastatin, rosuvastatin and simvastatin, separately). Data for NICM were obtained from FinnGen GWAS, including 11400 cases and 175752 controls of predominantly European ancestry. All data sources used (exposure from UK Biobank; outcome from FinnGen) have existing approvals from their respective Institutional Review Boards.

We conducted five MR methods, including inverse variance weighted (IVW), MR-Egger, weighted median, simple mode weighted mode to investigate whether potential causality between statin therapy and NICM exist. Effect size estimates (beta) and standard errors (SEs) for SNP-exposure and SNP-outcome associations from independent samples are used to estimate the causal exposure-outcome association. All MR studies rely on 3 basic assumptions as described in previous studies (11). Each instrumental variable (IV) was constructed from SNPs showing GWAS significant association (P<5×10^–8) with the respective trait. All included SNPs were independent with the linkage disequilibrium (r2=0.001, kb=10,000). Palindromic SNPs with intermediate allele frequencies were removed during analyses. Leave-one-out analysis was applied to estimate the effect of one single SNP. Heterogeneity between IVs was tested using the Cochran’s Q statistic. MR-Egger intercept test was used to examine any possible horizontal pleiotropic effects.

All statistical analyses were performed using R version 4.1.2 or Stata 15.1 (StataCorp/SE, College Station, TX). “TwoSampleMR” R package was used for MR analyses (15).

Population-based cohort

Finally, a total of 372 patients were included in the population-based cohort, with 44%(164/372) reported as statin users and the remaining 56% (208/372) as non-users. As summarized in Table 1, statin users were less likely to be male (47.60% vs 32.93%, P=0.004), had higher proportion of hypertension (83.57% vs 91.46%, P=0.03) and diabetes (35.27% vs 53.05%, P=0.001), higher BMI (27.62 vs 30.11 kg/m2, p=0.004) but lower BNP (958 vs 1432 pg/ml, P=0.01) and a trend of higher baseline ejection fraction (30.6% vs 27.7%, P=0.15).

Overall, the 28-day and 1-year mortality rate of the studied patients were 5.11% (19/372) and 27.96% (104/372), respectively. Compared to non-users, statin users had lower risk of death (1.22% vs 8.17% for 28-day; 25.61% vs 29.81% for 1-year). The Kaplan-Mirer survival curves also suggested that statin users had better 28-day (Log-rank test P=0.002) and 1-year survival than non-users (Log-rank test P=0.26) (Figure 1).

Likewise, in logistic regression analysis (Table 2), statin use was associated with lower risk of 28-day mortality (OR 0.14, P=0.009). The results remained consistent when adjusted for potential confounders (OR 0.10, P=0.03). Similar findings were observed for the 1-year mortality outcome. Compared to non-users, statins users had 19% and 17% lower risk of 1-year mortality in the non-adjusted (OR 0.81, P=0.37) and adjusted models (OR 0.83%, P=0.50).

Mendelian Randomization Analysis

Associations for individual SNPs for the exposure traits (atorvastatin, rusovastatin and simvastatin) are presented in supplemental Table S1/S2/S3. After removing one palindromic SNP (rs10738606), 20 were remaining for use in the Atorvastatin-NICM MR analyses. 6 SNPs were used for rusovastatin-NICM and 34 for simvastatin-NICM MR analyses. Table 3 shows causal-effect estimates of statin use on NICM from MR analyses. Estimates were concordant and similar in size in all MR methods, supporting a protective effect of statin use against NICM.

Figure 2 shows scatter plots of the SNP-outcome associations against the SNP-statins associations, allowing visualization of the causal effect estimate for each individual SNP on NICM. The leave-one-out analysis suggested that the observed result was not the effect of a single SNP (supplemental Figure S1). The Cochran’s Q test in the IVW and MR-Egger method (all P values >0.05) suggested no evidence of heterogeneity in the IVs. MR-Egger test suggested no sign of directional pleiotropy in our analysis (P values of MR-Egger intercept: P=0.12 for atorvastatin, 0.92 for rusovastin, 0.05 for simvastatin).

Our findings suggest that statin therapy was associated with lower mortality risk in HF patinets with NICM. MR analyses further supported that statin use causally confered a protective effect against NICM.

NICM is a heterogeneous group of myocardium diseases which generally lead to clinical manifestations of HF (2). The prevalence of NICM was estimated to be 2-15%, according to different diagnostic criteria (2, 16). As specific etiologies are not easily identified in many cases, treatment strategy mainly foused on HF management and arrhythmic prevention. However, except the three evidence-based HF drugs, there are limited evidence of survival benefit for other pharmaceutical treatment.

Previous studies have suggested that statin therapies were associated with reduced risks of arrhythmia and mortatliy in NICM patients (5, 17-20). However, a 2006 randomized placebo-controlled crossover trial suggested a neutral effect (21) and the short-term benefit of statin therapy remains unclear. In our study, we included a larger group of NICM patients and reported the 28-day and 1-year mortality risk as the endpoint. The results suggested that statin therapy did confer survival benefit in NICM patients, particularly for the early 28 days, which provided additional information to current knowledge.

Increasing evidence suggested that statins had pleotropic properties that help improve clinical outcomes (4), independently of its lipid-lowering capacity. The anti-inflammatory actions, the abilities to improve endothelial function, reverse remodeling or normalize sympatho-excitation are several proposed mechanisms why statin therapy could decrease mortality risk in non-ischemic HF patients (4, 19, 22). However, the causal relationship between statin use and NICM has not been well addressed yet.

MR analysis has been widely used as a tool to estimate the causal relationships between exposures and outcomes. The consistency of our results across different MR methods greatly strengthens the causal protective role of statin use in NICM. In addition, our findings suggested that the protective effects were observed in different types of statins (including atorvastatin, rusovastatin and simvastatin), suggestive of a robust effect of statin thrapy. The data presented here not only reassure clinicians considering statin as a treatment strategy in NICM patients, but also provides causal evidence of beneficial effect of statin use on NICM.

Limitations

Several limitations should be acknowledged. First, given the nature of an observational study, our analyses were limited by the abstracted information from the HF surveillance cohort. Data on lipid profiles, the prescription types, therapeutic dosage and duration of statin therapy were not available. Second, in MR analysis, horizontal pleiotropy is an important concern as it may introduce bias if the SNPs were associated with confounders through pathways not involving statin use. However, results from MR-Egger, median weighted and mode analyses, which are less susceptible to horizontal pleiotropy, were similar to the IVW estimates. In addition, as specific etiologies of NICM were not available, we were not able to perform subgroup analysis in current study. Finally, the generalizability to other population should be cautious as the analyzed participants were mostly of European or American ancestry.

This study provides evidence of a beneficial effect of statin use on NICM. Statin therapy not only conferred long-term benefits, but also associated with reduced short-term mortality risk in NICM patients. This could have significant implications for current HF management in NICM patients.

Acknowledgement The authors thank the staff and participants of the ARIC study and BioLINCC for their important contributions.

Conflict of Interest Disclosures: None reported.

Sources of Funding This work was supported by the Guangdong Peak Project [DFJH201802] to Z.C.J. and the National Natural Science Foundation of China [82371963], Guangdong Basic and Applied Basic Research Foundation [2021A1515012232, 2023A1515011366] and the NSFC lauching fund of the Guangdong Provincial People’s Hospital [8197070830, 8207070477, 8227070211] to F.H.W.

Declaration of Helsinki Our study complies with the Declaration of Helsinki, and our locally appointed ethics committee has approved the research protocol. Informed consent has been obtained from the subjects in each of the original cohorts.

Data Availability Statement The datasets are publically available upon reasonable request.

Bhaskaran A, De Silva K, Rao K, et al. Ventricular Tachycardia Ablation in Non-ischemic Cardiomyopathy. Korean Circ J. 2020;50(3):203-19. doi:10.4070/kcj.2019.0292
Wu AH. Management of patients with non-ischaemic cardiomyopathy. Heart. 2007;93(3):403-8. doi:10.1136/hrt.2005.085761
Heidenreich PA, Bozkurt B, Aguilar D, et al. 2022 AHA/ACC/HFSA Guideline for the Management of Heart Failure: A Report of the American College of Cardiology/American Heart Association Joint Committee on Clinical Practice Guidelines. Circulation. 2022;145(18):e895-e1032. doi:10.1161/cir.0000000000001063
Verma N, Figueredo VM. Statins in non-ischaemic cardiomyopathy: an update on our current clinical and pathophysiological understanding. Int J Clin Pract. 2011;65(11):1156-64. doi:10.1111/j.1742-1241.2011.02753.x
Wang JQ, Wu GR, Wang Z, Dai XP, Li XR. Long-term clinical outcomes of statin use for chronic heart failure: a meta-analysis of 15 prospective studies. Heart Lung Circ. 2014;23(2):105-13. doi:10.1016/j.hlc.2013.07.012
Kjekshus J, Apetrei E, Barrios V, et al. Rosuvastatin in older patients with systolic heart failure. N Engl J Med. 2007;357(22):2248-61. doi:10.1056/NEJMoa0706201
Tavazzi L, Maggioni AP, Marchioli R, et al. Effect of rosuvastatin in patients with chronic heart failure (the GISSI-HF trial): a randomised, double-blind, placebo-controlled trial. Lancet. 2008;372(9645):1231-9. doi:10.1016/s0140-6736(08)61240-4
Horwich TB, Hamilton MA, Maclellan WR, Fonarow GC. Low serum total cholesterol is associated with marked increase in mortality in advanced heart failure. J Card Fail. 2002;8(4):216-24. doi:10.1054/jcaf.2002.0804216
Marcoff L, Thompson PD. The role of coenzyme Q10 in statin-associated myopathy: a systematic review. J Am Coll Cardiol. 2007;49(23):2231-7. doi:10.1016/j.jacc.2007.02.049
Moosmann B, Behl C. Selenoprotein synthesis and side-effects of statins. Lancet. 2004;363(9412):892-4. doi:10.1016/s0140-6736(04)15739-5
Sanderson E, Glymour MM, Holmes MV, et al. Mendelian randomization. Nat Rev Methods Primers. 2022;2. doi:10.1038/s43586-021-00092-5
The Atherosclerosis Risk in Communities (ARIC) Study: design and objectives. The ARIC investigators. Am J Epidemiol. 1989;129(4):687-702.
Caughey MC, Sueta CA, Stearns SC, Shah AM, Rosamond WD, Chang PP. Recurrent Acute Decompensated Heart Failure Admissions for Patients With Reduced Versus Preserved Ejection Fraction (from the Atherosclerosis Risk in Communities Study). Am J Cardiol. 2018;122(1):108-14. doi:10.1016/j.amjcard.2018.03.011
Chang PP, Chambless LE, Shahar E, et al. Incidence and survival of hospitalized acute decompensated heart failure in four US communities (from the Atherosclerosis Risk in Communities Study). Am J Cardiol. 2014;113(3):504-10. doi:10.1016/j.amjcard.2013.10.032
Hemani G, Zheng J, Elsworth B, et al. The MR-Base platform supports systematic causal inference across the human phenome. Elife. 2018;7. doi:10.7554/eLife.34408
Follath F. Nonischemic heart failure: epidemiology, pathophysiology, and progression of disease. J Cardiovasc Pharmacol. 1999;33 Suppl 3:S31-5. doi:10.1097/00005344-199906003-00004
Goldberger JJ, Subacius H, Schaechter A, et al. Effects of statin therapy on arrhythmic events and survival in patients with nonischemic dilated cardiomyopathy. J Am Coll Cardiol. 2006;48(6):1228-33. doi:10.1016/j.jacc.2006.05.053
Laufs U, Wassmann S, Schackmann S, Heeschen C, Böhm M, Nickenig G. Beneficial effects of statins in patients with non-ischemic heart failure. Z Kardiol. 2004;93(2):103-8. doi:10.1007/s00392-004-1005-0
Horwich TB, MacLellan WR, Fonarow GC. Statin therapy is associated with improved survival in ischemic and non-ischemic heart failure. J Am Coll Cardiol. 2004;43(4):642-8. doi:10.1016/j.jacc.2003.07.049
Li X, Liu XP, Liu XH, et al. [Effect of statin therapy on mortality in patients with non-ischemic dilated cardiomyopathy]. Zhonghua Yi Xue Za Zhi. 2010;90(28):1974-7.
Bleske BE, Nicklas JM, Bard RL, et al. Neutral effect on markers of heart failure, inflammation, endothelial activation and function, and vagal tone after high-dose HMG-CoA reductase inhibition in non-diabetic patients with non-ischemic cardiomyopathy and average low-density lipoprotein level. J Am Coll Cardiol. 2006;47(2):338-41. doi:10.1016/j.jacc.2005.06.087
Buber J, Goldenberg I, Moss AJ, et al. Reduction in life-threatening ventricular tachyarrhythmias in statin-treated patients with nonischemic cardiomyopathy enrolled in the MADIT-CRT (Multicenter Automatic Defibrillator Implantation Trial with Cardiac Resynchronization Therapy). J Am Coll Cardiol. 2012;60(8):749-55. doi:10.1016/j.jacc.2012.03.041

Table1 Baseline characteristics of the included patients.

	Total(n=372)	Non-users (n=208)	Users (n=164)	P value
Age, years	68.96(8.75)	68.90(9.09)	69.04(8.33)	0.88
Sex, male(%)	153(41.13%)	99(47.60%)	54(32.93%)	0.004
Hypertension	323(87.06%)	173(83.57%)	150(91.46%)	0.03
Diabetes	160(43.13%)	73(35.27%)	87(53.05%)	0.001
Current smoker	64(17.20%)	35(16.83%)	29(17.68%)	0.83
BMI, kg/m2	28.78(7.82)	27.62(7.43)	30.11(8.07)	0.004
SBP, mmHg	137.70(32.59)	134.91(34.03)	141.13(30.47)	0.07
DBP, mmHg	79.92(20.03)	79.31(19.74)	80.67(20.42)	0.52
HR, beats/min	91.51(25.87)	93.13(27.61)	89.46(23.40)	0.18
BNP, pg/ml	1224(1039-1409)	1432(1151-1712)	958(744-1172)	0.01
HGB, g/L	11.69(1.91)	11.66(1.89)	11.73(1.95)	0.74
LVEF,%	29.02(14.45)	27.76(14.64)	30.60(14.11)	0.15

BMI, body mass index; SBP, systolic blood pressure; DBP, diastolic blood pressure; HR, heart rate; HGB, hemoglobin; LVEF, left ventricular ejection fraction.

Table2 Logistic regression of statin use with the study outcomes.

	Odds Ratio	95%CI	P value
28-day mortality
Model 1	0.14	0.03-0.61	0.009
Model 2	0.10	0.01-0.84	0.03
1-year mortality
Model 1	0.81	0.51-1.28	0.37
Model 2	0.83	0.48-1.43	0.50

CI, confidence interval;

Model 1: non-adjusted model;

Model 2: adjusted for age, sex, history of hypertension, diabetes, body mass index, systolic blood pressure/diastolic blood pressure and smoking status.

Table 3 Estimates of causal association between different types of statins use and NICM using two-sample Mendelian randomization (MR) analysis.

Exposure	Outcome	Method	SNPs*	Beta	SE	P value	Q	Q_pval	Egger_intercept	Egger_se	Egger_pval
Atorvastatin	NICM	MR Egger	20	-8.90	2.76	<0.001	20.71	0.29	0.017	0.011	0.12
		Inverse variance weighted	20	-4.79	1.20	<0.001	23.79	0.20
		Weighted median	20	-6.96	1.63	<0.001
		Simple mode	20	-8.02	2.79	0.01
		Weighted mode	20	-7.20	2.03	<0.001
Rosuvastatin	NICM	MR Egger	6	-12.07	29.46	0.70	4.90	0.30	-0.005	0.043	0.92
		Inverse variance weighted	6	-15.23	5.04	<0.001	4.91	0.43
		Weighted median	6	-19.69	6.64	<0.001
		Simple mode	6	-19.15	10.83	0.14
		Weighted mode	6	-20.34	9.61	0.09
Simvastatin	NICM	MR Egger	34	-3.25	0.85	<0.001	33.62	0.39	0.013	0.006	0.05
		Inverse variance weighted	34	-1.78	0.47	<0.001	38.03	0.25
		Weighted median	34	-1.75	0.64	0.01
		Simple mode	34	-2.01	1.15	0.09
		Weighted mode	34	-2.25	0.74	<0.001

NICM, non-ischemic cardiomyopathy.

*Removing the following SNPs (rs10738606)for being palindromic with intermediate allele frequencies in atorvastatin-SNP association.

No competing interests reported.

supplements.docx

Statin Therapy in Non-ischemic Cardiomyopathy: evidence from a Population Based Cohort and Mendelian Randomization Analysis

Status:

Version 1

Abstract

Figures

Introduction

Methods

Results

Discussion

Conclusion

Declarations

References

Tables

Additional Declarations

Supplementary Files

Status:

Version 1