Original Article

Association Between Use of Statins and Outcomes in Heart

Failure With Reduced Ejection Fraction
Prospective Propensity Score Matched Cohort Study of 21 864 Patients in
the Swedish Heart Failure Registry
Urban Alehagen, MD, PhD; Lina Benson, MSc; Magnus Edner, MD, PhD;
Ulf Dahlström, MD, PhD; Lars H. Lund, MD, PhD

Background—In heart failure (HF) with reduced ejection fraction, randomized trials of statins did not demonstrate improved
outcomes. However, randomized trials may not always be generalizable. The aim was to determine whether statins are
associated with improved outcomes in an unselected nationwide population of patients with HF with reduced ejection
fraction overall and in relation to ischemic heart disease (IHD).
Methods and Results—In the Swedish Heart Failure Registry, 21 864 patients with HF with reduced ejection fraction (age
± SD, 72±12 years; 29% women), of whom 10 345 (47%) were treated with statins, were studied. Propensity scores for
statin use were derived from 42 baseline variables. The associations between statin use and outcomes were assessed
with Cox regressions in a population matched 1:1 based on propensity score and age and in the overall population
with adjustment for propensity score and age. The primary outcome was all-cause mortality; secondary outcomes were
cardiovascular mortality; HF hospitalization; and combined all-cause mortality or cardiovascular hospitalization. Survival
at 1 year in the matched population was 83% for statin-treated versus 79% for untreated patients (hazard ratio, 0.81; 95%
confidence interval, 0.76–0.86; P<0.001). In the unmatched population, 1-year survival was 85% for statin-treated versus
79% for untreated patients, hazard ratio after adjustment for propensity score and age was 0.84 (95% confidence interval,
0.80–0.89; P<0.001). No examined baseline variables interacted with statin use except for IHD (P=0.001), with a hazard
ratio of 0.76 (95% confidence interval, 0.70–0.82, P<0.001) with IHD and 0.95 (95% confidence interval, 0.85–1.07;
P=0.430 without IHD. Statin use was also associated with reduced risk for all 3 secondary outcomes.
Downloaded from http://ahajournals.org by on August 6, 2022

Conclusions—In an unselected nationwide population of patients with HF with reduced ejection fraction, statins were
associated with improved outcomes, specifically in the presence of IHD. This contrasts with previous randomized
controlled trials. Additional randomized controlled trials with more generalized inclusion or focused on IHD may be
warranted.  (Circ Heart Fail. 2015;8:252-260. DOI: 10.1161/CIRCHEARTFAILURE.114.001730.)
Key Words: heart failure ■ outcome assessment ■ prognosis

I schemic heart disease (IHD) is a major contributor to hos-

pitalization and mortality in the Western hemisphere.1,2
Treatment with 3-hydroxy-3-metylglutaryl coenzyme A reduc-
effects in HF with reduced ejection fraction (HFREF) are more
controversial,17–19 and the randomized Controlled Rosuvatatin
Multinational Trial in Heart Failure (CORONA) and Gruppo
tase inhibitors (statins) reduce total cholesterol and low-den- Italiano per lon Studio della Sopravvivenza Nell Insufficienza
sity lipoprotein, and also increase high-density lipoprotein.3,4 Cardia (GISSI) did not demonstrate benefits of statins.20,21
Although benefits of statin treatment have been demonstrated
in many patient groups, the strongest documentation and the
Clinical Perspective on p 260
greatest benefits are in IHD. Effects include not only reduction Although randomized controlled trials (RCTs) are the gold
of all-cause and cardiovascular mortality but also of cardiovas- standard for assessing treatment effects, they may have lim-
cular events5–8 and reduced new onset heart failure (HF) after ited generalizability. Rigorous observational studies may add
myocardial infarction.9–11 Pleiotropic effects may include antifi- information about treatment patterns and associations with
brotic and antihypertrophic effects, upregulation of endothelial outcomes in unselected patients and real-world situations.
NO production, vasodilation and decreased platelet aggrega- Furthermore, the extensive evidence and recommendations
tion,12 inhibition of inflammatory cytokines and neurohormonal for statins in IHD may provide a conundrum for clinicians in
activation, and reversal of myocardial remodeling.13–16 However, interpreting the absence of benefit in HFREF RCTs.
Received August 21, 2014; accepted January 6, 2015.
From the Departments of Cardiology and Medical and Health Sciences, Linköping University, Linköping, Sweden (U.A., U.D.); Departments of Clinical
Science and Education (L.B.) and Medicine (M.E., L.L.), Karolinska Institutet, Stockholm, Sweden; and Department of Cardiology, Karolinska University
Hospital, Stockholm, Sweden (M.E., L.L.).
Correspondence to Urban Alehagen, MD, PhD, Department of Cardiology, Linköping University Hospital, SE-581 85, Linköping, Sweden. E-mail
urban.alehagen@liu.se
© 2015 American Heart Association, Inc.
Circ Heart Fail is available at http://circheartfailure.ahajournals.org DOI: 10.1161/CIRCHEARTFAILURE.114.001730

252
 Alehagen et al   Statin Use in Heart Failure   253

Therefore, we tested the hypothesis that statins are asso-

ciated with improved outcomes in an unselected nationwide
population of patients with HFREF and assessed a potential
interaction with IHD.

Methods
The Swedish Heart Failure Registry (RiksSvikt) has been pre-
viously described.22 Inclusion criteria are clinician-judged HF.
Echocardiography is overwhelmingly performed with standardized
and accredited protocols requiring Simpson and M-mode determina-
tion of EF, but in this nationwide registry, EF is not independently
adjudicated and we cannot rule out that some EF values are by visual
estimation. In the registry, ≈80 variables are recorded at discharge
from hospital or after outpatient clinic visits on a case record form
and entered into a central web-based database. The database is run
against the Swedish death registry, monthly. The protocol, registra-
tion form, and annual report are available at www.rikssvikt.se. The
establishment of the registry and this study were approved by a mul-
tisite Ethics Committee and conform to the Declaration of Helsinki.
Individual patient consent was not required, but the patients were in-
formed of entry into national registries and allowed to opt out.
Mortality was determined from the Population Registry, causes of
death from the Cause of Death Registry, and hospitalization from the
Patient Registry. For cardiovascular mortality and hospitalization,
we included ICD-10 diagnoses I00–I99; and for HF hospitalization,
ICD-10 diagnoses I50, I42–I43, I25.5, I11.0, I13.0, and I13.2 were
included. For ischemic hospitalization, ICD-10 diagnoses I20–I25.2
were included. The causes of death are based on death certificates,
with potentially limited reliability. A recent study suggests a 55% ac-
curacy of death certificates in cardiovascular disease in a Swedish
community.23 However, causes of hospitalization are more reliable.
The positive predictive value for most hospitalization diagnoses in the
Patient Registry is 85% to 95%,24 and a HF diagnosis was verified in
Figure 1. Flow chart. EF indicates ejection fraction.
Downloaded from http://ahajournals.org by on August 6, 2022

86% to 91% of HF hospitalizations.25

Between May 11, 2000 and April 23, 2012, 73 392 registrations
were obtained in the registry from 66 of 77 hospitals and 97 of 1011
primary healthcare centers in Sweden. Registrations with statin use
missing, EF ≥40% or EF missing, inclusions before January 1, 2001
or after December 31, 2011, and repeat registrations for the same
patients were excluded, yielding 21 864 patients with HFREF (flow
chart, Figure 1).
matched population (Figure 3). Cox regression models were used to
estimate the hazard ratio (HR) and corresponding 95% confidence
interval (CI) in the overall population, crude and adjusted for pro-
pensity score as a continuous covariate, as well as in the matched
population.

Statistical Analysis
Descriptive data are presented as numbers (n) and percentages or means
with SD or median with interquartile range (IQR) as appropriate and
compared with Student unpaired t tests or χ2 tests as appropriate.

Propensity Scores
Propensity scores for treatment with statins were estimated for each
patient by logistic regression with 42 clinically relevant baseline vari-
ables as independent variables and statin treatment as the dependent
variable. All continuous variables were modeled using restricted
cubic splines (3 degrees of freedom). The propensity score is the
propensity from 0 to 1 to receive treatment, given a set of known vari-
ables, and is used to adjust for potential selection bias, confounding
and differences between treatment group in observational studies.26,27
Missing values were handled by estimating 1 logistic regres-
sion model for each pattern of missing values. Each individual
then received the propensity score that incorporated all variables
with nonmissing values for that individual. An age- and propensity
score–matched population was constructed28 with matching 1:1
without replacement, based on age difference ≤5 years and pro-
pensity score difference ≤0.1. This yielded 5381 patients in each
group (Figure 2). Figure 2. Box plot presentation of propensity scores for statin
use in the unmatched and matched populations. Note: Boxes
Outcomes represent median and interquartile range; whiskers represent
The primary outcome, all-cause mortality, by statin treatment was minimum and maximum (if not outliers). Outliers are displayed
assessed in the overall population with Kaplan–Meier analyses with circles and are defined as observations >1.5 times the inter-
and plotted in the same figure as Kaplan–Meier survival for the quartile range from the first or third quartile, respectively.
 254  Circ Heart Fail  March 2015

Figure 3. Kaplan–Meier plot illustrating all-cause mortality

according to statin use in the overall and matched populations.

The proportional hazards assumption was tested by scaled

Schoenfeld residuals and the presence of extreme outliers were as-
sessed by dfbetas. No violations to the proportional hazards assump-
tion or possibly influential outliers were found.
Interactions between statin therapy and clinically relevant vari-
ables were modeled with Cox regression and presented in a Forest
plot for all-cause mortality for the matched population (Figure 4).
Downloaded from http://ahajournals.org by on August 6, 2022

Continuous variables were dichotomized for visual interpretation.

Interaction analyses render similar results as subgroup analyses with
the additional benefit of being able to statistically test for differences
in associations between statin use and outcomes between subgroups.
The secondary outcomes namely cardiovascular mortality; HF hos-
pitalization; and combined all-cause mortality or cardiovascular hos-
pitalization were also assessed with Cox regression in the matched
population. Patients not experiencing an event were censored at
December 31, 2011. In addition, for the end point cardiovascular
mortality and HF hospitalization, censoring also occurred if the pa-
tient died from any other cause.
For all analyses, the level of significance was set to 5%, and all Figure 4. Forest plot illustrating hazard ratios for all-cause
reported P values are 2-sided. Statistics were performed in R v 2.15.3 mortality associated with statin use in predefined subgroups
(R foundation for Statistical Computing, Vienna, Austria). in the matched population after adjustment for the interac-
tion between statin use and the variable on the y axis. Note:
Squares represent the hazard ratio and the lines represent the
Results 95% confidence interval (CI). Continuous variables were ana-
lyzed as such but dichotomized at clinically relevant cutoffs for
Study Population display in this figure. BMI indicates body mass index; LVEF,
Baseline characteristics of the overall and matched study popu- left ventricular ejection fraction; MAP, mean arterial pressure;
lations are presented in Table 1. A total of 21 864 patients with NYHA, New York Heart Association functional class; and RAS,
renin–angiotensin system.
HFREF, age 72±12 years, 29% female, were included. Overall,
47% were treated with statins. There were several differences
between the groups. Treated versus untreated patients were of remained slightly more common in the treated group and EF
similar age but were more commonly men and with diabetes was slightly lower in the untreated group (Table 1), and propen-
mellitus and with higher EF. However, the greatest difference sity scores were now nearly identically distributed (Figure 2).
was presence of IHD, in 74% of treated versus 35% of untreated
patients. Also, statin-treated patients were more commonly on Outcomes
angiotensin-converting enzyme inhibitors/angiotensin receptor
blockers and β-blockers (Table 1). Thus, in Table 1, there are Primary Outcome: All-Cause Mortality
considerable standardized differences in baseline variables and in In the overall population, survival was at 1, 3, and 5 years,
Figure 2, an imbalance in propensity scores between the groups. 85%, 68%, and 53% in treated versus 79%, 61%, and 48% in
After propensity score and age matching, the standardized untreated patients (Figure 3), crude HR 0.80 (95% CI, 0.76–
differences were small, although diabetes mellitus and IHD 0.83; P<0.001; Tables 2 and 3).
 Alehagen et al   Statin Use in Heart Failure   255

Table 1.  Baseline Characteristics of the Unmatched and Matched Populations of the Study Population Containing Participants With
Heart Failure and Ejection Fraction <40%
Unmatched Population Matched Population
Missing No Statins Standardized No Statins Standardized
Variables No. Information, % (n=11 519) Statins (n=10 345) Difference, % (n=5381) Statins (n=5381) Difference, %
Characteristics
 Follow-up time, median 0 703 (4005) 757 (4015) 678 (3685) 746 (4005)
d (range)
 Age, mean y (SD) 1 0 72 (14) 72 (10) 6* 73 (11) 73 (11) 0
Sex 2 0
 Women, n (%) 3737 (32) 2586 (25) 16* 1579 (29) 1555 (29) 1
 Men, n (%) 7782 (68) 7759 (75) 16* 3802 (71) 3826 (71) 1
Civil status 3 5
 Married, n (%) 6106 (56) 6414 (65) 18* 3075 (61) 3089 (60) 1
 Single, n (%) 4781 (44) 3474 (35) 18* 2004 (39) 2039 (40) 1
Living arrangements 4 26
 Independent, n (%) 8128 (94) 7218 (97) 14* 3862 (95) 3758 (95) 1
 Institution, n (%) 551 (6) 251 (3) 14* 206 (5) 187 (5) 1
Clinic 5 3
 Cardiology, n (%) 6213 (56) 5662 (57) 2 2893 (56) 2872 (55) 1
 Medicine/geriatrics, n (%) 4961 (44) 4340 (43) 2 2311 (44) 2322 (45) 1
Location 6 0
 Inpatient, n (%) 6764 (59) 5615 (54) 9* 3113 (58) 3067 (57) 2
 Outpatient, n (%) 4749 (41) 4726 (46) 9* 2265 (42) 2310 (43) 2
Follow-up referral speciality 7 8
 Primary care, n (%) 3278 (31) 2261 (23) 18* 1498 (31) 1379 (28) 7*
Downloaded from http://ahajournals.org by on August 6, 2022

 Cardiology or internal 6698 (64) 6995 (72) 18* 3173 (65) 3397 (68) 7*
medicine, n (%)
 Other, n (%) 472 (5) 417 (4) 1 220 (4) 210 (4) 1
Follow-up referral 8 8 4369 (42) 4360 (45) 7* 2115 (43) 2095 (42) 2
to outpatient HF
nurse clinic, n (%)
Medical history
 Duration of HF, mo 9 1
≥6, n (%)
   5648 (49) 5494 (53) 8* 2920 (55) 2775 (52) 5*
  <6, n (%) 5806 (51) 4795 (47) 8* 2428 (45) 2574 (48) 5*
 
Year of registration 10 0
  2001–2005, n (%) 1924 (17) 1234 (12) 14* 907 (17) 745 (14) 8*
  2006–2011, n (%) 9595 (83) 9111 (88) 14* 4474 (83) 4636 (86) 8*
 
NYHA class 11 23
  I, n (%) 794 (9) 693 (9) 0 362 (9) 334 (8) 2
  II, n (%) 3891 (43) 3513 (45) 2* 1792 (43) 1787 (44) 0
  III, n (%) 3706 (41) 3321 (42) 2* 1734 (42) 1747 (43) 1
  IV, n (%) 577 (6) 366 (5) 8* 238 (6) 236 (6) 0
 LVEF, % 12 0
  <30%, n (%) 6210 (54) 4923 (48) 13* 2931 (54) 2636 (49) 11*
  30% to 39%, n (%) 5309 (46) 5422 (52) 13* 2450 (46) 2745 (51) 11*
 
Smoking 13 21
  Current, n (%) 1382 (16) 1233 (15) 3* 575 (14) 650 (15) 4
  Former, n (%) 3621 (41) 4181 (49) 17* 1927 (45) 1910 (45) 2
  Never, n (%) 3873 (44) 3062 (36) 15* 1706 (41) 1711 (40) 1
 Diabetes mellitus, n (%) 14 1 2038 (18) 3433 (33) 36* 1450 (27) 1352 (25) 4

(Continued )
 256  Circ Heart Fail  March 2015

Table 1.  Continued

Unmatched Population Matched Population
Missing No Statins Standardized No Statins Standardized
Variables No. Information, % (n=11 519) Statins (n=10 345) Difference, % (n=5381) Statins (n=5381) Difference, %

 Hypertension, n (%) 15 3 4363 (39) 4852 (48) 18* 2353 (46) 2381 (46) 0
 IHD, n (%) 16 4 3807 (35) 7435 (74) 78* 2786 (56) 2776 (54) 5*
 DCM, n (%) 17 3 2466 (22) 1444 (14) 20* 910 (18) 910 (18) 2
 HCM, n (%) 18 27 186 (2) 127 (2) 3* 107 (3) 62 (2) 8*
 Valvular disease, n (%) 19 3 2606 (23) 1870 (19) 12* 1167 (22) 1088 (21) 4*
 Atrial fibrillation/flutter, 20 1 5636 (49) 3898 (38) 23* 2421 (45) 2415 (45) 0
n (%)
 Pulmonary disease, n (%) 21 3 1871 (17) 1644 (16) 1 945 (18) 904 (17) 2
 Valvular surgery, n (%) 22 1 632 (6) 542 (5) 1 285 (5) 310 (6) 2
 Revascularisation, n (%) 23 2 1380 (12) 4857 (48) 78* 1240 (24) 1383 (26) 6*
Laboratory examination
 Potassium, mean 24 43 4.2 (0.5) 4.2 (0.4) 1 4.2 (0.4) 4.2 (0.4) 3
mmol/L (SD)
 Hb, mean g/L (SD) 25 0 134 (18) 133 (17) 5* 134 (17) 134 (17) 2
 Creatinine clearance, 26 6 66 (37) 68 (34) 6* 64 (33) 65 (34) 4
mean mL/min (SD)
 NT-proBNP, mean 27 72 6338 (7955) 5501 (7259) 11* 6271 (8035) 5808 (7554) 6
ng/L (SD)
Medical treatment
 ACEI/ARB, n (%) 28 0 9885 (86) 9501 (92) 19* 4717 (88) 4839 (90) 7*
β-Blocker, n (%)
  29 0 9937 (87) 9572 (93) 20* 4832 (90) 4868 (91) 2
 Diuretic, n (%) 30 0 9357 (81) 8170 (79) 6* 4446 (83) 4312 (80) 6*
Downloaded from http://ahajournals.org by on August 6, 2022

 Aldosterone antagonist, 31 0 3706 (32) 3452 (34) 3 1748 (33) 1771 (33) 1
n (%)
 Digoxin, n (%) 32 0 2451 (21) 1470 (14) 18* 997 (19) 970 (18) 1
 Nitrates, n (%) 33 0 1415 (12) 2234 (22) 25* 1009 (19) 912 (17) 5*
 Aspirin, n (%) 34 0 4658 (41) 6912 (67) 53* 2895 (54) 2904 (54) 0
 Oral anticoagulants, 35 0 4566 (40) 3790 (37) 6* 2136 (40) 2148 (40) 0
n (%)
Examinations
 Mean arterial pressure, 36 1 90 (14) 90 (13) 5* 90 (14) 90 (13) 1
mean mm Hg (SD)
 Pulse pressure, 37 1 50 (17) 52 (16) 10* 51 (17) 51 (16) 2
mean mm Hg (SD)
 Pulse frequency, 38 7 76 (17) 72 (15) 24* 74 (16) 74 (15) 2
mean bpm (SD)
 BMI, mean (SD) 39 54 26 (5) 27 (5) 21* 26 (5) 26 (5) 0
 Chest radiograph, 40 32 4403 (54) 3342 (49) 12* 1960 (52) 1888 (51) 2
cardiomegaly, n (%)
 Chest radiograph, 41 32 3875 (48) 2955 (43) 10* 1662 (44) 1714 (47) 5*
congestion, n (%)
 ECG
  QRS duration, mean 42 31 116 (29) 117 (28) 3* 118 (29) 117 (28) 2
ms (SD)
ACEI indicates angiotensin-converting enzyme inhibitors; ARB, angiotensin II receptor blocker; BMI, body mass index; DCM, dilated cardiomyopathy; HCM, hypertrophic
cardiomyopathy; HF, heart failure; IHD, ischemic heart disease; LVEF, left ventricular ejection fraction; NT-proBNP, N-terminal fragment of pro-brain natriuretic peptide;
NYHA class, New York Heart Association functional class; Standardized difference, the difference between the means for the 2 groups divided by the mutual SD.
Mean arterial pressure: derived as systolic blood pressure × one-third plus diastolic blood pressure × two thirds.
Column no. indicates the variables included in the derivation of the propensity score.
* P<0.05.
 Alehagen et al   Statin Use in Heart Failure   257

Table 2.  Survival at 1, 3, and 5 Years, Respectively, in the be interpreted with caution and by no means should dictate care
Unmatched Population in any way. However, our study raises important questions and
No Statins Statins sheds light on (1) statin use and associations with outcomes in an
unselected community population in contrast to RCT populations
Time, y No. at Risk Survival 95% CI No. at Risk Survival 95% CI
and (2) lends strength to the notion that statins are associated with
1 8410 0.79 0.78–0.80 7892 0.85 0.84–0.85 benefit in IHD, possibly also in the presence of HFREF.
3 4326 0.60 0.60–0.61 4022 0.68 0.67–0.69
5 1668 0.48 0.46–0.49 1382 0.53 0.52–0.55 How to Interpret Our Results in Relation to the
CI indicates confidence interval. Literature
The discrepancy in our study versus RCTs may have numerous
In the matched population, the corresponding figures at 1, 3, explanations. Rosuvastatin, studied in CORONA and GISSI-
and 5 years were 83%, 65%, and 53% versus 79%, 58%, and HF, is hydrophilic and relies on active transport into hepato-
45%, matched HR 0.81 (95% CI, 0.76–0.86; P<0.001; Tables 3 cytes, whereas, for example, simvastatin and atorvastatin are
and 4). In the overall population, HR after adjustment for propen- lipophilic and may be expected to reach higher levels so that it
sity score and age was 0.84 (95% CI, 0.80–0.89; P<0.001; Table could be expected that the levels of those statins are higher in
3). The hypothetical number needed to treat for 1 year to prevent the myocardial tissue.29,30 This may be particularly important
1 death from any cause was 25 (95% CI, 24.4–25.6) patients. because a potential mechanism of benefit may be regulation of
cardiac inflammation rather than systemic lipid levels. Indeed,
Interaction Analysis for All-Cause Mortality in a recent meta-analysis of 13 RCTs evaluating lipophilic
In Figure 4, a Forest plot illustrates the association between statins in HFREF, Liu et al31 demonstrated decreased mortal-
statin use and all-cause mortality in subgroups after adjust- ity, and hospitalization in the statin treatment groups. We do
ment for interactions between prespecified clinically impor- not have information on specific statins used in our patients.
tant variables and statin treatment. There were no interactions However, the Swedish National Board of Health and Welfare
except for a highly significant interaction with IHD (P=0.001); maintains annual prescription data. Overall in Sweden, filled
with HR of 0.76 (95% CI, 0.70–0.82; P<0.001) with IHD and statin prescriptions per 1000 inhabitants were distributed as
0.95 (95% CI, 0.85–1.07, P=0.430) without IHD. In contrast, follows: in 2006 hydrophilic statins: rosuvastatin 4; pravas-
there was no interaction with diabetes mellitus. tatin 11; fluvastatin 2; and lipophilic statins: simvastatin 206;
atorvastatin 40. The corresponding figures for 2013 were as
follows: rosuvastatin 13; pravastatin 6; fluvastatin 0.09; sim-
Secondary Outcomes
Downloaded from http://ahajournals.org by on August 6, 2022

vastatin 285; atorvastatin 66. Lovastatin is not available in

For the propensity score and age-matched population, HRs for
Sweden. Thus, lipophilic statins were indeed overwhelmingly
the secondary outcomes are listed in Table 3. Statin use was
used in Sweden overall and most likely in our patients.
associated with reduced risk for all 3 secondary outcomes, ie,
Conceivably, RCTs may have failed to demonstrate the ben-
cardiovascular mortality, HF hospitalization, and composite
efit of statins because of inadequate power or less relevant end
all-cause mortality or cardiovascular hospitalization.
points. Indeed, HF hospitalization was not an end point, but in
a post hoc analysis of CORONA, the cumulative number of
Discussion hospitalizations (which is increasingly being used as an end
In this large prospective registry of patients with HREF, use of
statins was associated with reduced all-cause mortality, cardio-
Table 4.  Association Between Statin Use and Outcomes
vascular mortality, HF hospitalization, and the combined out-
come of all-cause mortality and cardiovascular hospitalization. Outcome Population n Events/n HR 95% CI P Value
Our findings were consistent, both in a matched population based All-cause Overall/crude 8053/21 864 0.80 0.76–0.83 <0.001
on propensity score and age and the overall population with mortality
adjustment for propensity score and age. Importantly, there was All-cause Overall/ 8053/21 864 0.84 0.80–0.89 <0.001
a strong interaction with IHD, where statin treatment was asso- mortality adjusted for
ciated with reduced all-cause mortality in patients with but not propensity
without IHD. There was no interaction with diabetes mellitus. score as
covariate
Two major RCTs in HFREF20,21 have failed to show a ben-
efit from statins. Therefore, our observational findings should All-cause Match I 4214/10 762 0.81 0.76–0.86 <0.001
mortality

Table 3.  Survival at 1, 3, and 5 Years, Respectively, in the Cardiovascular Match I 3001/10 762 0.80 0.75–0.87 <0.001
mortality
Matched Population
Heart failure Match I 4083/10 762 0.92 0.86–0.97 0.005
No Statins Statins hospitalization
Time, y No. at Risk Survival 95% CI No. at Risk Survival 95% CI Composite all- Match I 7240/10 762 0.90 0.86–0.95 <0.001
1 3582 0.79 0.78–0.80 3719 0.83 0.82–0.84 cause mortality
+ cardiovascular
3 1739 0.58 0.57–0.60 1828 0.65 0.64–0.66 hospitalization
5 653 0.45 0.43–0.47 650 0.51 0.49–0.52 CI indicates confidence interval; and HR, hazard ratio.
CI indicates confidence interval. Matching group I: 0.1 of the propensity score and age and ≤ 5 yrs of follow-up.
 258  Circ Heart Fail  March 2015

point to enrich HF trials) were significantly reduced by 15%
to 20%.32 Interestingly, in a post hoc analysis of the Treating to
New Target (TNT) study (high- versus low-dose atorvastatin in
IHD), the higher dose resulted in lower risk of HF hospitaliza-
tion overall, and in particular in patients with pre-existing HF,
suggesting a potential direct effect on HF11. This could support
the argument that statins might affect the development of HF.
RCTs provide better evidence for a treatment effect in com-
parison with observational studies but may have other limitations.
Older patients28,33,34 and women33,35 are consistently excluded
both in cardiovascular and cancer trials, potentially compromis-
ing external validity of RCTs.33 Patients with HF included in
RCTs may have been more ill and cared for by HF specialists
active in HF trials, whereas patients with milder HF and predomi-
nantly IHD may have been cared for by general cardiologists or
those practicing predominantly in the ischemia field. Indeed, in
CORONA patients were more severely ill than in our study (eg,
60% versus 40% in NYHA class III), at a later stage of HF where
the ability of the statins to modify the disease may be limited.
There are additional indications that the populations differed
(Table 5), with patients in our study more frequently women,
with more impaired left ventricular function, atrial fibrillation,
and beta blocker treatment and less dilated cardiomyopathy,
hypertension, and digitalis treatment than in RCTs.
Additional observational studies in HFREF suggest risk reduc-
tion for all-cause mortality associated with statin use.36 In a pro-
pensity-scored matched analysis, including HFREF, Ouzounian
et al34 reported reduced all-cause mortality associated with statin
use, however only, like in our study, in those with IHD. The same
was reported about hospitalization for HF, acute coronary events,
or ischemic stroke. Gastelurrutia et al35 also observed reduced
mortality in a HFREF population in an observational study.
In comparison with the above observational studies, the pres-
ent study may be more unselective and generalizable because
patients were included from 86% of all Swedish hospitals.
Thus, we confirm previous smaller and less well-adjusted
observational studies because our results are in contrast to
major recent RCTs.
The major RCTs20,21 have not shown a benefit from, and hence
HF guidelines do not recommend statin use in HFREF (class III).
However, both the AHA/ACC and the ESC guidelines on treat-
ment of IHD state that statin treatment is recommended and they
do not make exceptions for or qualify recommendations based on
HFREF.37–39 The well-established benefits of statins in IHD and
diabetes mellitus may have made clinicians reluctant to include
such patients in CORONA and GISSI and risk assignment to
placebo, which may have compromised generalizability of those
RCTs. Although the RCTs did include a substantial proportion of
patients with IHD, specifically patients with predominantly IHD
and perhaps milder HF may nonetheless not have been cared for
by clinicians participating in the HF RCTs and may have been
excluded based on clinician recommendation or own preference,
given the extensive knowledge, including in the general public, of
the benefits of statins. Indeed, in GISSI-HF,21 only 57% of those

Table 5.  Comparison of Some Important Clinical Variables Between the 2 RCTs GISSI-HF, CORONA, and the Swedish Heart Failure
Downloaded from http://ahajournals.org by on August 6, 2022

Registry Study
GISSI-HF CORONA Swedish Heart Failure Registry
Variable Statins No statins Statins No statins Statins No statins
Age, mean y 68 68 73 73 73 73
Women, % 24 21 24 24 29 29
NYHA class II, % 61 64 37 37 44 43
III, % 36 34 61 62 43 42
IV, % 3 2 1 2 6 6
LVEF, mean % 33 33 31 31 … …
LVEF <30%, mean, % … … … … 49 54
LVEF 30–39%, mean % … … … … 51 46
LVEF >40%, mean % 10 10 … … … …
IHD, % 40 40 60 60 54 56
DCM, % 35 34 … … 18 18
Hypertension, % 55 54 … … 46 46
Diabetes mellitus, % 27 25 30 29 25 27
Atrial fibrillation, % 19 20 24 23 45 45
NT-proBNP, ng/L … … 1522 (median) 1404 (median) 3160 (median) 3146 (median
ACEI/ARB inhibitors, % 94 93 91 92 90 88
Beta blockers, % 63 62 75 75 91 90
Digitalis, % 40 40 34 32 18 19
Diuretics, % 90 90 89 88 80 83
ACEI/ARB indicates angiotensin-converting enzyme inhibitors or angiotensin receptor blockers; DCM, dilated cardiomyopathy; IHD, ischemic heart disease; LVEF, left
ventricular ejection fraction; NT-proBNP, N-terminal fragment of pro-brain natriuretic peptide; NYHA, New York Heart Association functional class; and RCT, randomized
controlled trial.
Ellipsis are based on missing information from the original publication.
No ANOVA tests have been performed because the original data from the 2 RCTs were not at hand.

screened were randomized, and the corresponding figure is not
given in CORONA.20 The 2 RCTs included considerable propor-
tions with IHD and there was no interaction between statin treat-
ment and IHD, but as alluded to, many patients with HFREF and
IHD may have been excluded from these RCTs. In contrast, there
was a strong interaction with statin use in our study (P=0.001),
where the HR was 0.76 (P<0.001) with IHD and 0.95 (P=0.430)
without IHD. Taken together, these observations and recom-
mendation in guidelines, other than those for HF, suggest that
the question of statin use in patients with concomitant IHD and
HFREF may not be resolved.
Limitations
An observational registry-based study is subject to bias and
confounding. Our study entails propensity score matching and
adjustment for most variables that may affect the choice to pre-
scribe a statin (bias) and affect outcome independent of statins
(confounding). A standardized difference between groups of
≤10% is generally considered inconsequential.40 Of all the 42
variables, only 1 had a standardized difference of >10% (EF,
somewhat lower in the untreated patients, standardized differ-
ence 11%). Importantly, we cannot rule out residual confound-
ing from unknown or unmeasured variables. The present study
should merely be viewed as additional information on general-
izability of RCTs and the potential role of statins in HFREF and
concomitant IHD and in no way dictate treatment decisions.
Conclusions
In this large nationwide observational study of patients with
Downloaded from http://ahajournals.org by on August 6, 2022
HFREF, statins were associated with reduced all-cause mortal-
ity, cardiovascular mortality, cardiovascular hospitalization, and
combined all-cause mortality and HF hospitalization. This con-
trasts with previous neutral RCTs. There was an interaction with
IHD, suggesting potentially reduced mortality in patients with
but not without IHD. Our findings in no way support the use of
statins but do suggest that RCTs with more generalized inclusion
and focused on presence and severity of IHD may be warranted.
Sources of Funding
The study was supported by grants from The County Council of
Östergötland, The Swedish Heart and Lung Foundation, and the
University of Linköping to Dr Alehagen and The County Council
of Stockholm, The Swedish Heart and Lung Foundation, and the
Swedish Research Council to Dr Lund, and grants from the Swedish
Heart Failure Research Foundation to Drs Alehagen, Dahlstrom, and
Lund. The funding organizations had no role in the design, manage-
ment, analysis, and interpretation of the data, preparation, review, or
approval of the article.
Disclosures
None.
References
1. Bolognese L. Changing patterns of ST elevation myocardial infarction
epidemiology. Am Heart J. 2010;160:S1–S3.
2. Rizzello V, Lucci D, Maggioni AP, Giampaoli S, Greco C, Di Pasquale G,
Pallotti MG, Mureddu GF, Di Chiara A, Boccanelli A; IN-ACS Outcome
Investigators. Clinical epidemiology, management and outcome of acute
coronary syndromes in the Italian network on acute coronary syndromes
(IN-ACS Outcome study). Acute Card Care. 2012;14:71–80. doi:
10.3109/17482941.2012.655296.
Alehagen et al   Statin Use in Heart Failure   259
3. Hunninghake DB, Mellies MJ, Goldberg AC, Kuo PT, Kostis JB, Schrott
HG, Insull W Jr, Pan HY. Efficacy and safety of pravastatin in patients
with primary hypercholesterolemia. II. Once-daily versus twice-daily
dosing. Atherosclerosis. 1990;85:219–227.
4. Rubenfire M, Maciejko JJ, Blevins RD, Orringer C, Kobylak L, Rosman
H. The effect of pravastatin on plasma lipoprotein and apolipoprotein
levels in primary hypercholesterolemia. The Southeastern Michigan
Collaborative Group. Arch Intern Med. 1991;151:2234–2240.
5. Randomised trial of cholesterol lowering in 4444 patients with coronary
heart disease: the Scandinavian Simvastatin Survival Study (4S). Lancet.
1994;344:1383–1389.
6. Andrews TC, Raby K, Barry J, Naimi CL, Allred E, Ganz P, Selwyn AP.
Effect of cholesterol reduction on myocardial ischemia in patients with
coronary disease. Circulation. 1997;95:324–328.
7. Sacks FM, Pfeffer MA, Moye LA, Rouleau JL, Rutherford JD, Cole TG,
Brown L, Warnica JW, Arnold JM, Wun CC, Davis BR, Braunwald E.
The effect of pravastatin on coronary events after myocardial infarction
in patients with average cholesterol levels. Cholesterol and Recurrent
Events Trial Investigators. N Engl J Med. 1996;335:1001–1009. doi:
10.1056/NEJM199610033351401.
8. Aronson D, Mutlak D, Lessick J, Kapeliovich M, Dabbah S, Markiewicz
W, Beyar R, Hammerman H, Reisner S, Agmon Y. Relation of statin
therapy to risk of heart failure after acute myocardial infarction. Am J
Cardiol. 2008;102:1706–1710. doi: 10.1016/j.amjcard.2008.07.057.
9. Strandberg TE, Holme I, Faergeman O, Kastelein JJ, Lindahl C, Larsen
ML, Olsson AG, Pedersen TR, Tikkanen MJ; IDEAL Study Group.
Comparative effect of atorvastatin (80 mg) versus simvastatin (20 to 40
mg) in preventing hospitalizations for heart failure in patients with pre-
vious myocardial infarction. Am J Cardiol. 2009;103:1381–1385. doi:
10.1016/j.amjcard.2009.01.377.
10. Kjekshus J, Pedersen TR, Olsson AG, Faergeman O, Pyörälä K. The
effects of simvastatin on the incidence of heart failure in patients with
coronary heart disease. J Card Fail. 1997;3:249–254.
11. Khush KK, Waters DD, Bittner V, Deedwania PC, Kastelein JJ,
Lewis SJ, Wenger NK. Effect of high-dose atorvastatin on hospital-
izations for heart failure: subgroup analysis of the Treating to New
Targets (TNT) study. Circulation. 2007;115:576–583. doi: 10.1161/
CIRCULATIONAHA.106.625574.
12. Laufs U, Liao JK. Post-transcriptional regulation of endothelial ni-
tric oxide synthase mRNA stability by Rho GTPase. J Biol Chem.
1998;273:24266–24271.
13. Luo JD, Zhang WW, Zhang GP, Guan JX, Chen X. Simvastatin inhibits
cardiac hypertrophy and angiotensin-converting enzyme activity in rats
with aortic stenosis. Clin Exp Pharmacol Physiol. 1999;26:903–908.
14. Brili S, Tousoulis D, Antonopoulos AS, Antoniades C, Hatzis G, Bakogiannis
C, Papageorgiou N, Stefanadis C. Effects of atorvastatin on endothelial func-
tion and the expression of proinflammatory cytokines and adhesion mol-
ecules in young subjects with successfully repaired coarctation of aorta.
Heart. 2012;98:325–329. doi: 10.1136/heartjnl-2011-300287.
15. Jain MK, Ridker PM. Anti-inflammatory effects of statins: clinical evi-
dence and basic mechanisms. Nat Rev Drug Discov. 2005;4:977–987.
doi: 10.1038/nrd1901.
16. Zheng C, Azcutia V, Aikawa E, Figueiredo JL, Croce K, Sonoki H, Sacks
FM, Luscinskas FW, Aikawa M. Statins suppress apolipoprotein CIII-
induced vascular endothelial cell activation and monocyte adhesion. Eur
Heart J. 2013;34:615–624. doi: 10.1093/eurheartj/ehs271.
17. Hayashidani S, Tsutsui H, Shiomi T, Suematsu N, Kinugawa S, Ide
T, Wen J, Takeshita A. Fluvastatin, a 3-hydroxy-3-methylglutaryl
coenzyme a reductase inhibitor, attenuates left ventricular remodel-
ing and failure after experimental myocardial infarction. Circulation.
2002;105:868–873.
18. Bleske BE, Nicklas JM, Bard RL, Brook RD, Gurbel PA, Bliden KP,
Rajagopalan S, Pitt B. Neutral effect on markers of heart failure, in-
flammation, 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;47:338–341. doi: 10.1016/j.jacc.2005.06.087.
19. Bauersachs J, Galuppo P, Fraccarollo D, Christ M, Ertl G. Improvement
of left ventricular remodeling and function by hydroxymethylglutaryl
coenzyme a reductase inhibition with cerivastatin in rats with heart fail-
ure after myocardial infarction. Circulation. 2001;104:982–985.
20. Kjekshus J, Apetrei E, Barrios V, Böhm M, Cleland JG, Cornel JH,
Dunselman P, Fonseca C, Goudev A, Grande P, Gullestad L, Hjalmarson
A, Hradec J, Jánosi A, Kamenský G, Komajda M, Korewicki J, Kuusi
T, Mach F, Mareev V, McMurray JJ, Ranjith N, Schaufelberger M,
 260  Circ Heart Fail  March 2015

Vanhaecke J, van Veldhuisen DJ, Waagstein F, Wedel H, Wikstrand J;
CORONA Group. Rosuvastatin in older patients with systolic heart fail-
ure. N Engl J Med. 2007;357:2248–2261. doi: 10.1056/NEJMoa0706201.
21. Gissi-HF Investigators, Tavazzi L, Maggioni AP, Marchioli R, Barlera S,
Franzosi MG, Latini R, Lucci D, Nicolosi GL, Porcu M, Tognoni G. Effect of
rosuvastatin in patients with chronic heart failure (the GISSI-HF trial): a ran-
domised, double-blind, placebo-controlled trial. Lancet. 2008;372:1231–9.
22. Jonsson A, Edner M, Alehagen U, Dahlström U. Heart failure registry: a
valuable tool for improving the management of patients with heart fail-
ure. Eur J Heart Fail. 2010;12:25–31. doi: 10.1093/eurjhf/hfp175.
23. Eriksson A, Stenlund H, Ahlm K, Boman K, Bygren LO, Johansson LA,
Olofsson BO, Wall S, Weinehall L. Accuracy of death certificates of
cardiovascular disease in a community intervention in Sweden. Scand J
Public Health. 2013;41:883–889. doi: 10.1177/1403494813499653.
24. Ludvigsson JF, Andersson E, Ekbom A, Feychting M, Kim JL,
Reuterwall C, Heurgren M, Olausson PO. External review and valida-
tion of the Swedish national inpatient register. BMC Public Health.
2011;11:450. doi: 10.1186/1471-2458-11-450.
25. Ingelsson E, Arnlöv J, Sundström J, Lind L. The validity of a diagno-
sis of heart failure in a hospital discharge register. Eur J Heart Fail.
2005;7:787–791. doi: 10.1016/j.ejheart.2004.12.007.
26. D’Agostino RB Jr. Propensity score methods for bias reduction in the
comparison of a treatment to a non-randomized control group. Stat Med.
1998;17:2265–2281.
27. Rosenbaum PR, Rubin DB. Reducing bias in observational studies using sub-
classification on the propensity score. J Am Statist Assoc. 1984;79:516–524.
28. Sekhon J. Multivariate and propensity score matching software with au-
tomated balance optimization: the matching package for R. J Stat Softw.
2011;42:1–52.
29. Ikeda U, Shimpo M, Ikeda M, Minota S, Shimada K. Lipophilic statins
augment inducible nitric oxide synthase expression in cytokine-stimulat-
ed cardiac myocytes. J Cardiovasc Pharmacol. 2001;38:69–77.
30. Sakamoto T, Kojima S, Ogawa H, Shimomura H, Kimura K, Ogata Y,
Sakaino N, Kitagawa A; MUSASHI-AMI Investigators. Usefulness of
hydrophilic vs lipophilic statins after acute myocardial infarction: sub-
analysis of MUSASHI-AMI. Circ J. 2007;71:1348–1353.
31. Liu G, Zheng XX, Xu YL, Lu J, Hui RT, Huang XH. Effects of lipophilic
Downloaded from http://ahajournals.org by on August 6, 2022
35. Gastelurrutia P, Lupón J, de Antonio M, Urrutia A, Díez C, Coll R,
Altimir S, Bayes-Genis A. Statins in heart failure: the paradox be-
tween large randomized clinical trials and real life. Mayo Clin Proc.
2012;87:555–560. doi: 10.1016/j.mayocp.2012.02.018.
36. Krum H, Latini R, Maggioni AP, Anand I, Masson S, Carretta E, Ingrillì
F, Pettinati G, Glazer R, Tognoni G, Cohn J. Statins and symptomatic
chronic systolic heart failure: a post-hoc analysis of 5010 patients en-
rolled in Val-HeFT. Int J Cardiol. 2007;119:48–53. doi: 10.1016/j.
ijcard.2006.07.106.
37. Stone NJ, Robinson JG, Lichtenstein AH, Bairey Merz CN, Blum CB,
Eckel RH, Goldberg AC, Gordon D, Levy D, Lloyd-Jones DM, McBride P,
Schwartz JS, Shero ST, Smith SC Jr, Watson K, Wilson PW, Eddleman KM,
Jarrett NM, LaBresh K, Nevo L, Wnek J, Anderson JL, Halperin JL, Albert
NM, Bozkurt B, Brindis RG, Curtis LH, DeMets D, Hochman JS, Kovacs
RJ, Ohman EM, Pressler SJ, Sellke FW, Shen WK, Smith SC Jr, Tomaselli
GF; American College of Cardiology/American Heart Association Task
Force on Practice Guidelines. 2013 ACC/AHA guideline on the treatment
of blood cholesterol to reduce atherosclerotic cardiovascular risk in adults: a
report of the American College of Cardiology/American Heart Association
Task Force on Practice Guidelines. Circulation. 2014;129:S1–S45.
38. Anderson JL, Adams CD, Antman EM, Bridges CR, Califf RM, Casey
DE Jr, Chavey WE 2nd, Fesmire FM, Hochman JS, Levin TN, Lincoff
AM, Peterson ED, Theroux P, Wenger NK, Wright RS, Smith SC Jr; 2011
Writing Group Members; ACCF/AHA Task Force Members. 2011 ACCF/
AHA focused update incorporated into the ACC/AHA 2007 guidelines for
the management of patients with unstable angina/non-ST-elevation myocar-
dial infarction: a report of the American College of Cardiology Foundation/
American Heart Association Task Force on Practice Guidelines. Circulation.
2011;123:e426–e579. doi: 10.1161/CIR.0b013e318212bb8b.
39. Task Force Members, Montalescot G, Sechtem U, Achenbach S,
Andreotti F, Arden C, Budaj A, Bugiardini R, Crea F, Cuisset T, Di
Mario C, Ferreira JR, Gersh BJ, Gitt AK, Hulot JS, Marx N, Opie LH,
Pfisterer M, Prescott E, Ruschitzka F, Sabaté M, Senior R, Taggart DP,
van der Wall EE, Vrints CJ; ESC Committee for Practice Guidelines,
Zamorano JL, Achenbach S, Baumgartner H, Bax JJ, Bueno H, Dean V,
Deaton C, Erol C, Fagard R, Ferrari R, Hasdai D, Hoes AW, Kirchhof P,
Knuuti J, Kolh P, Lancellotti P, Linhart A, Nihoyannopoulos P, Piepoli

statins for heart failure: a meta-analysis of 13 randomised controlled tri-
als. Heart Lung Circ. 2014;23:970–977. doi: 10.1016/j.hlc.2014.05.005.
32. Rogers JK, Jhund PS, Perez AC, Böhm M, Cleland JG, Gullestad L, Kjekshus
J, van Veldhuisen DJ, Wikstrand J, Wedel H, McMurray JJ, Pocock SJ.
Effect of rosuvastatin on repeat heart failure hospitalizations: the CORONA
Trial (Controlled Rosuvastatin Multinational Trial in Heart Failure). JACC
Heart Fail. 2014;2:289–297. doi: 10.1016/j.jchf.2013.12.007.
33. Cherubini A, Oristrell J, Pla X, Ruggiero C, Ferretti R, Diestre G, Clarfield
AM, Crome P, Hertogh C, Lesauskaite V, Prada GI, Szczerbinska K,
Topinkova E, Sinclair-Cohen J, Edbrooke D, Mills GH. The persistent ex-
clusion of older patients from ongoing clinical trials regarding heart failure.
Arch Intern Med. 2011;171:550–556. doi: 10.1001/archinternmed.2011.31.
34. Ouzounian M, Tu JV, Austin PC, Chong A, Liu PP, Lee DS. Statin thera-
py and clinical outcomes in heart failure: a propensity-matched analysis.
J Card Fail. 2009;15:241–248. doi: 10.1016/j.cardfail.2008.10.026.
MF, Ponikowski P, Sirnes PA, Tamargo JL, Tendera M, Torbicki A, Wijns
W, Windecker S; Document Reviewers, Knuuti J, Valgimigli M, Bueno
H, Claeys MJ, Donner-Banzhoff N, Erol C, Frank H, Funck-Brentano C,
Gaemperli O, Gonzalez-Juanatey JR, Hamilos M, Hasdai D, Husted S,
James SK, Kervinen K, Kolh P, Kristensen SD, Lancellotti P, Maggioni
AP, Piepoli MF, Pries AR, Romeo F, Rydén L, Simoons ML, Sirnes PA,
Steg PG, Timmis A, Wijns W, Windecker S, Yildirir A, Zamorano JL.
2013 ESC guidelines on the management of stable coronary artery dis-
ease: the Task Force on the management of stable coronary artery disease
of the European Society of Cardiology. Eur Heart J. 2013;34:2949–3003.
40. Patel K, Fonarow GC, Kitzman DW, Aban IB, Love TE, Allman RM,
Gheorghiade M, Ahmed A. Angiotensin receptor blockers and outcomes
in real-world older patients with heart failure and preserved ejection frac-
tion: a propensity-matched inception cohort clinical effectiveness study.
Eur J Heart Fail. 2012;14:1179–1188. doi: 10.1093/eurjhf/hfs101.

CLINICAL PERSPECTIVE
Statins improve outcomes in patients with ischemic heart disease (IHD) and are recommended in guidelines. Observational
studies have suggested that they may be beneficial in heart failure not only in the presence but also in the absence of IHD.
This hypothesis could not be confirmed in 2 randomized controlled trials,20,21 where statins in heart failure with reduced
ejection fraction did not improve outcomes, regardless of presence of IHD, and heart failure guidelines do not recommend
statins now. Questions remain about the generalizability of the trials and how to confront the conflicting evidence and rec-
ommendations in IHD. We performed a large observational study in the Swedish Heart Failure Registry, where associations
between statin use and outcomes were assessed in 21 864 patients with heart failure with reduced ejection fraction, with
matching by and adjustment for propensity scores for statin use. We observed that statin use was associated with reduced all-
cause mortality, cardiovascular mortality, heart failure hospitalization, and combined all-cause mortality or cardiovascular
hospitalization. Statins were associated with reduced mortality in patients with but not without IHD. Given previous trials,
our findings do not support general use of statins in heart failure with reduced ejection fraction, but whether statins may be
beneficial in broader generalized populations, in particular with IHD, require further study.