Cognitive Profile in Persons With Systolic and Diastolic Heart Failure


Ponrathi Athilingam, PhD, College of Nursing, University of South Florida, 12901 Bruce B. Downs Boulevard, MDC 22, Tampa, FL 33612


©2012 Wiley Periodicals, Inc.

Persons with heart failure (HF) have four times the risk of having cognitive impairment compared with the general population and display different patterns of cognitive impairment. This secondary analysis of a published cross-sectional study of 90 community-dwelling adults examined the Montreal Cognitive Assessment (MoCA) scores and HF differentiated as systolic and diastolic HF. Mean MoCA score was 22.9 (standard deviation±2.31) in persons with systolic HF (n=69) and 24.8 (standard deviation±2.76) in persons with diastolic HF (n=21) with statistically significant mean difference between groups (t=−2.025, P=.030). Independent t test on the eight MoCA domain scores and systolic and diastolic HF indicated significance on visuo-spatial/executive function (P=.026), attention (P=.049), abstraction (P=.014), and delayed recall (P=.048). Findings from this study support the need for including persons with systolic and diastolic HF in future researches on identifying varying cognitive profiles to plan tailored cognitive intervention.

Heart failure (HF) is a growing public health concern with an estimated 5 million Americans with this condition.1 There are many different ways to categorize HF, including the side of the heart involved and whether the abnormality is due to insufficient contraction resulting in systolic HF with low ejection fraction (EF) (≤40%) or insufficient relaxation causing diastolic HF with preserved or normal EF (>40%) or both.2 Systolic and diastolic HF appear to be separate syndromes with distinctive morphologic and functional changes, although signs, symptoms, and prognosis are very similar.3 Both systolic and diastolic HF are associated with cognitive impairment and may at times manifest primarily as delirium in hospitalized patients or as mild cognitive impairment in otherwise stable persons with HF.1 The different physiological characteristics of systolic and diastolic HF may have different effects on cognitive function among these individuals.4 Hence, the aim of the secondary analysis of a published correlational study was to examine cognitive profiles among persons with systolic and diastolic HF.

Although the underlying mechanism for cognitive impairment in HF remains unclear, several pathophysiological hypotheses have been proposed including: (1) low EF and cardiac output,4–6 (2) embolic episodes of the failing heart resulting in cerebral ischemia and cerebral hypoperfusion,7,8 (3) impaired cerebral reactivity,9 and (4) autoneurotoxicity or neural cell death from the episodic decompensation of HF resulting in hypoxia and causing loss of brain plasticity.10,11 Most of the brain’s cognitive functions are based on the coordinated interactions of large numbers of neurons that are distributed within and across different specialized brain areas.11,12 Areas in the brain responsible for short-term memory function, including the hippocampus and its output fibers of the fornix, projecting to the septum and to the anterior thalamus via the mammillary bodies, were injured or damaged in 13 persons with systolic HF (EF <40%) compared with 49 healthy controls,12 which was supported in (n=17) persons with systolic HF compared with age-matched healthy participants (n=50)13 and also in an animal model.14 Significantly more periventricular and total white matter hyperintensities, lacunar and cortical infarcts, and global and medial temporal lobe atrophy on magnetic resonance imaging studies were identified among persons with systolic HF (n=58) compared with medical participants with no HF (n=48).15

The reported prevalence of cognitive impairment in persons with HF ranges from 25% to 75%, with most prior studies including persons with low EF or systolic HF.16 A population-based study of 14,089 participants reported that HF patients are 1.51 times more likely to have cognitive impairment (95% confidence interval [CI], 1.15–1.96) than those without HF.17 In a systematic review of 22 controlled studies with a pooled sample of 2937 persons with HF and 14,848 control patients, the odds ratio for cognitive impairment among patients with HF was 1.62 (95% CI, 1.48–1.79; P=.0001).18 Additionally, a case-control study reported that persons with HF have four times the risk of having cognitive impairment compared with the general population.19 Recently published studies that compared adults with systolic HF, medical participants, and/or healthy controls have reported that persons with systolic HF scored poorly on Cambridge Cognitive Examination of the Elderly compared with controls,20,21 were impaired in ≥3 cognitive domains,22 and 25% of persons with systolic HF demonstrated cognitive impairment compared with 15% of the cardiac controls and 4% of the healthy controls.16 Results of these reviews on systolic HF and cognitive function are presented in Table I.

Table I.   Summary of Literature Review on Recently Published Case Control Studies That Compared Cognitive Impairment Among Persons With HF, Cardiac Control, and/or Healthy Controls
AuthorsDesignSampleCognitive MeasuresResults
  1. Abbreviations: CAD, coronary artery disease; COMCOG, Cambridge Cognitive Examination of the Elderly; HF, heart failure; SHF, systolic heart failure.

Almeida et al19Longitudinal studySHF 77, medical/cardiac controls 73, healthy controls 81COMCOGSHF patients declined 0.9 points on COMCOG over 2 years (P=.022) compared with controls but was not significant when compared with persons with coronary artery disease (P=.133)
Beer et al20Case control studySHF 31, controls without HF 24CAMCOGSHF patients had lower CAMCOG scores than controls (93.5±6.1 vs 99.9±2.4, P<.001) and had significantly lower scores on visuo-spatial, executive function, visual memory, and verbal learning
Pressler et al16Case control studySHF patients (n=249), healthy controls (n=63), and medical participants (n=102)Wechsler Test of Adult Reading, Boston Naming Test, digit span, Hopkins Verbal Learning Test, figure copy and memory recall, digit symbol, Trial making A and B, and controlled oral word associationSHF patients had poorer memory, slower psychomotor speed, and worse scores on executive function. In addition, 24% of the HF patients were impaired in ≥3 cognitive domains compared with 14% of the healthy and 12% of the medical participants
Sauve et al18Case control studySHF 55 and healthy controls 50Neurobehavioral cognitive status examinationCognitive impairment was independently associated with HF (odds ratio, 4.47; confidence interval, 1.75–11.43; P<.002)
Vogles et al22Case control studySHF 62, cardiac controls 53, healthy controls 42Neuropsychological assessments of mental speed, executive function, memory, language, and visuo-spatial function25% (P=.04) of SHF patients were cognitively impaired in executive function, memory, language, mental speed, and attention, compared with 15% of the cardiac controls and 4% of the healthy controls

Although evidence indicates that persons with systolic HF perform worse on cognitive function, a recent study reported an association of early stages of diastolic dysfunction with poorer cognitive scores on attention domain (r=−0.16; 95% CI, 0.26–0.05) and executive functioning (r=−0.17; 95% CI, 0.28–0.05) with no significant association to systolic dysfunction.23 After controlling for age, sex, and New York Heart Association (NYHA) functional class, persons with systolic HF had a mean score of −1.04 for immediate memory compared with −0.38 for the group with diastolic HF (P=.01).24 Persons with systolic HF continued to perform significantly worse on immediate memory (P=.03) and delayed memory (P<.001); these were significantly lower compared to persons with diastolic HF (P<.00).24 To our knowledge, these were the only data we have that compared cognitive scores differentiated by systolic and diastolic HF.24 Although, there have been considerable advances in the management of systolic HF, progress on the management of diastolic HF and research in this area is overdue.2,3

The aim of this secondary analysis of a published cross-sectional study of 90 persons with HF was to examine the existing data on cognitive scores measured by Montreal Cognitive Assessment (MoCA), a simple cognitive screening tool that has 8 cognitive domains that are most commonly affected in persons with HF.16,19,26 Therefore, MoCA total scores and domain scores were used in this secondary analysis to ascertain differences in cognitive profile among persons with systolic and diastolic HF. HF was differentiated as systolic HF (EF ≤40%) and diastolic HF (EF >40%) from the available data.25

Research Methods

Method and Sample

Existing data from a cross-sectional correlational study that enrolled 90 community-dwelling adults 50 years and older with a clinical diagnosis of HF as defined by the International Statistical Classification of Diseases and Related Health Problems–Ninth Revision (ICD-9-CM); NYHA functional classification class I to III; and the American Heart Association HF stages B, C, and D were used. Persons were excluded if they were on continuous oxygen, listed for heart transplant as a United Network for Organ Sharing status 1A or 1B, supported by ventricular assist device or home inotropic therapy, enrolled in a palliative or hospice care program, had a clinical diagnosis of dementia or Alzheimer’s disease, or had a clinical history of stroke.

Dependent Variable

Participants were screened for cognitive impairment using the MoCA.26 The MoCA total score and 8 cognitive domain scores were used in this secondary analysis. MoCA has a total scoring range of 0 to 30. A score of 26 to 30 is considered normal, 22 to 25 is considered mild cognitive impairment, 17 to 21 is considered moderate cognitive impairment, and a score below 17 is considered dementia. Cronbach’s alpha for the total MoCA score was 0.83.26 It has been validated for use in persons with mild cognitive impairment and showed high sensitivity as a screening tool for persons with HF,25,27 stroke,28 and cardiovascular disease.29 The MoCA incorporates the clock drawing test, which has been reported to be 50% more sensitive among participants with HF when compared with Mini-Mental Status Examination and Boston Aphasia Examination.30 Recently, the psychometric properties of MoCA were tested using the classical test theory among clinical and nonclinical community samples and reported a Cronbach’s alpha of 0.64, 0.66 among community samples, and 0.77 among clinical sample indicating that the MoCA is best suited to quantify cognitive impairment in clinical samples such as persons with HF.31 The MoCA is a one-page scale that involves a paper and pencil test in a question and answer format conducted by an interviewer and takes about 10 to 12 minutes to administer. To minimize ceiling effect on education, one point was added to the MoCA total score for participants with an education less than high school.26 As is standard for the MoCA, scores from memory domain were not included in the total MoCA score. A separate domain score was determined from the memory test that tested the patient’s ability to remember and repeat 5 unrelated things in two attempts. Participants were given 1 point if they remembered all 5 items in 2 attempts and scored 0 if were unable to remember all 5 items in 2 attempts.

Independent Variable

The independent variable used in this secondary analysis was EF by echocardiogram obtained from medical record within 6 months and was coded as systolic (≤40%) and diastolic HF (>40%).

Other Variables

Demographic and clinical variables obtained from demographic and clinical questionnaire, clinical variables including HF stages, etiology of HF, knowledge of medications, and medication lists were used. NYHA functional classification was determined based on the patient’s symptoms on the day of the study by a cardiology nurse practitioner. Other variables that may have potential association with cognitive impairment were collected and included depression measured using the Geriatric Depression Scale-15 (GDS-15) with a clinical cut-off score of >6 indicating depression,32,33 comorbidity data using Modified Cumulative Illness Rating Scale,34 Physical Self-Maintenance Scale for understanding activities of daily living,35 and 6-Minute Walk Test to assess functional ability.36

Procedure and Data Collection

The human subject review board of the University of Rochester approved the original study. Participants were recruited from the University of Rochester–affiliated cardiology clinics in 2007. The principal investigator collected all of the data.

Data Analyses

The secondary analysis used total MoCA score and MoCA domain scores as dependent variables and systolic and diastolic HF as an independent variable. Means and standard deviations (SDs) were calculated on continuous variables and frequencies for categorical variables differentiated by systolic and diastolic HF to describe the sample. Pearson correlation coefficients on continuous variables and Spearman’s Rho nonparametric correlation coefficients on nominal and categorical variables were calculated to examine the relationships between them and the cognitive score. A step-wise multiple regression model was performed first for the covariates identified to be related to the MoCA score and then for the systolic and diastolic HF. Analyses were performed separately on systolic and diastolic HF on the total MoCA score and the 8 MoCA domain scores. Independent t test was calculated to understand mean differences between systolic and diastolic HF on the MoCA score and MoCA domain scores.


A total of 67% of participants in the diastolic HF group were older than 65, compared with 35% among persons with systolic HF. More persons with systolic HF (54%) had less than a high school education, compared with 33% of persons with diastolic HF. An overwhelming 78% of participants in the systolic HF group had a family history of HF, compared with 27% among persons with diastolic HF. A total of 57% of participants had NYHA class II HF and none had class IV HF. More details on demographic and clinical data among the groups are shown in Table II.

Table II.   Demographic and Clinical Characteristics of the Study Sample
 DescriptionSystolic HF (n=69)Diastolic HF (n=21)
  1. Abbreviations: ACE, angiotensin-converting enzyme inhibitor; ARB, angiotensin receptor blocker; BB, β-blocker; BP, blood pressure; CRT, cardiac resynchronization therapy; GDS-15, Geriatric Depression Scale-15; HF, heart failure; MICR, Modified Cumulative Illness Rating Scale; NYHA, New York Heart Association; PSMS, Physical Self-Maintenance scale; 6MWT, Six-Minute Walk Test; SOS, significant others.

Age, y>6562.1±8.62434.863.1±10.71466.7
Education in yearsHigh school or less13.7±2.93753.614.2±2.4733.3
Marital statusMarried 4971 1466.7
RaceCaucasian 5376.8 885.7
Living arrangementWith spouse or SOS 5681.2 1676.2
Family historyYes 5478.3 626.8
Etiology of HFIschemic 3652.2 1571.4
NYHA classificationClass I 6 8.7 2 9.5
Class II 3956.5 1257.2
Class III 2434.8 733.3
HF stageStage B 710.1 628.6
Stage C 5884.1 1466.6
Stage D 4 5.8 1 4.8
Systolic BP 123.6±23  132.6±22.5  
Diastolic BP 68.7±11.5  68.0±10.9  
Mean arterial pressure 87.2±14.7  89.6±13.4  
ACE inhibitorYes 80±17.95275.474.7±24.41781.0
ARB 70.0±25.81115.942.3±7.4314.3
CRT  2434.8 2 9.5
GDS-15Score >63.43±2.78811.64.71±4.05523.8
MICR 21.7±4.4  23.14±5.1  
PSMS 58.8±3.2  58.5±3.5  
6MWTDistance in meters649.7±134.0  645.7±169.4  

Bivariate analysis indicated no significant association between MoCA score in both systolic and diastolic HF groups with sex, NYHA class, HF stages, physical ability scores, and functional ability measured by 6-Minute Walk Test. Although more persons with diastolic HF (23.8%) scored high (score >6) on the depression score (GRD-15) compared with 11.6% of persons with systolic HF, there was no association between depression score and MoCA score.

Age, education, knowledge of medications, medication use (only angiotensin-converting enzyme inhibitor), and mean arterial blood pressure were associated with MoCA score and were treated as covariates. In a step-wise regression model, covariates accounted for 37% of the variance (P<.001) in the model. When systolic and diastolic HF data were added, there was only a 2% increase in variance and the model was not significant (P=.882). In addition, coefficient statistics was not significant for age (t=−1.09; P=.28) and mean arterial blood pressure (t=0.46; P=.64).

Cognitive Profile in Systolic and Diastolic HF

Participants with systolic HF (n=69) had a mean MoCA score of 22.9 (SD±2.31) compared with participants with diastolic HF (n=21) with a mean of 24.8 (SD±2.76), indicating a lower mean score for participants with systolic HF with statistically significant mean difference among groups by t test (t=−2.025, P=.030). A total of 61% (42 of 69) of participants with systolic HF scored below 26 on the MoCA (a value suggestive of mild cognitive impairment) compared with 52% (11 of 21) of participants with diastolic HF.

Independent t test on systolic and diastolic HF on the 8 MoCA domain scores was performed. Levene’s Test of Equality of variance was significant (P>.05) on visuo-spatial/executive function, attention, abstraction, and delayed recall; therefore, on these variables, Levene’s Test of Equality of variance was not assumed. Table III displays the t test results comparing the mean MoCA domains scores on systolic and diastolic HF indicating association with visuo-spatial/executive function (P=.026), attention (P=.049), abstraction (P=.014), and delayed recall (P=.048).

Table III.   Cognitive Impairment Differentiated by Systolic and Diastolic HF on MoCA Total Score and 8 MoCA Domains Score
 Systolic HF (n=69)Diastolic HF (n=21) t Test for Equality of Means
Mean/SDMean/SD t df2-Tailed
  1. Abbreviations: HF, heart failure; MoCA, Montreal Cognitive Assessment; SD, standard deviation. aEquality of variance not assumed if Levene’s Test of Equality of variance was significant (P<.05).

MoCA total score22.9±2.3124.81±2.99 −2.02536.45.030a
Visuo-spatial/executive function3.74±1.02 3.95±1.07 −1.78749.74.026a
Naming Animals 2.8±0.26 2.95±0.22 0.56888.571
Memory1.25±0.69 1.24±0.700.32388.748
Attention5.19±0.97 5.43±0.871.53246.98.049a
Language2.01±0.88 2.09±0.990.72488.471
Abstraction1.64±.066 1.52±0.51 −1.83328.66.014a
Delayed recall3.19±1.39 3.05±1.16 −1.54624.32.048a
Orientation5.98±0.12 5.95±0.22 −0.74988.456

Although the mean MoCA scores were significantly different among groups, 74% of participants with systolic HF scored poorly on the MoCA domain scores on visuo-spatial/executive function compared with 57% of participants with diastolic HF (P=.026), language domain (68% vs 57%; P=.47), and attention domain (51% vs 38%; P=.049). On the contrary, 95% of participants with diastolic HF scored poorly on delayed recall domain compared with 84% of participants with systolic HF (P=.048) and abstraction domain (48% vs 26%; P=.01). Memory score was similar among both groups, with comparable percentage of participants unable to remember 5 items in 2 attempts and was not statistically significant (P=.75). When visuo-spatial and executive functions were analyzed as individual items, similar number of participants were impaired on visuo-spatial function tested by copying a cube (60% vs 40%; P=.23) and with the clock-drawing test (47% vs 50%; P=.31) and executive function tested with Trial Making (30% vs 0%; P=.08).


Results from this study added additional data indicating that mean cognitive scores are different among participants with systolic and diastolic HF and are statistically significant. Although age was considered a covariate, it was not a predictor in the regression model. However, 67% of participants in the diastolic HF group were older than 65 years compared with 35% among persons with systolic HF, indicating that age may be a predictor in a larger sample. The mean MoCA score was lower among persons with systolic HF compared with persons with diastolic HF and the domains of impairment were different, a notion supported by Bauer.24 The data from our study also support the results of different cognitive domains of impairment among persons with systolic and diastolic HF.24 This difference was also supported by Bauer24 among community-dwelling persons (n=249) with systolic HF (mean EF 25.7%±8.1%) who performed poorly on visuospatial ability, psychomotor speed, and executive function,16 which was true among our participants with systolic HF, and was also supported in a case control study of persons with systolic HF.4 Additionally, a recent study reported early stages of diastolic dysfunction was associated with lower cognitive scores on attention domain and executive functioning, whereas our data indicated that these domains are impaired among persons with systolic HF, which was supported by other researchers.4,16 Our data also indicated that memory domain that required remembering 5 items during two trials was similarly impaired among both groups of participants with systolic and diastolic HF, a domain commonly impaired in persons with HF.37

Also, the current literature is consistent in that patients with severity of HF defined by NYHA class III to IV symptoms have a higher prevalence of cognitive impairment than persons with class I and II.17,22,27 Our study included stable community-dwelling adults with an aim to identify early, subtle cognitive impairment; therefore, 67% of our participants had class I or II HF and no participants had class IV HF. This stability and chronicity of HF may explain nonassociation of cognitive scores and NYHA classification among our participants. Seventy-seven percent of participants with diastolic HF in our study were older than 65 years; however, age was not associated with total MoCA score in this study. On the other hand, age was a stronger predictor for cognitive impairment among persons with HF in several published studies,5,18,22 yet there is limited evidence to suggest an association of diastolic HF and cognitive impairment among this population. Coexisting depression was the strongest predictor of nonadherence among persons with HF.38 Although more persons with diastolic HF were screened (23.8%), indicating depression based on GDS-15 scores, depression score showed no association with MoCA score or MoCA domain scores. This may reflect a restriction in range of depression scores and lack of data on medication use for depression.

Multiple studies have documented impairment of one or more cognitive domains in individuals with systolic HF with scant data on cognitive impairment among persons with diastolic HF.16,23,24 Additionally, there is no current standard for assessing or screening for cognitive impairment in persons with HF. Although the science surrounding cognitive impairment in HF has progressed, future research needs to focus on cognitive impairment with a simple, domain-specific cognitive screening instrument rather than multiple cognitive batteries that are time-consuming, because fatigue, a common HF symptom, may potentially affect completion of multiple test batteries.39 In addition, the neuropsychological measures used in published research studies are inconsistent, thus making comparison difficult for use in HF population.4,5,18,22 In a recently published study, a MoCA score below 26 was a significant predictor of self-care in older hospitalized patients with HF.40 Thus the use of a domain-specific simple cognitive screening tool such as the MoCA may allow researchers and clinicians to create a more accurate cognitive profile of individuals with systolic and diastolic HF and make referrals as needed for detailed neuropsychological testing.27

In addition, because of multiple organ involvement in the HF syndrome and altered tissue perfusion, cytokines and inflammatory biomarkers may be produced both by the myocardium and extramyocardial sources in HF, indicating a need to focus on the exact mechanisms of how these cytokines participate in the molecular and cellular processes of memory formation and cognitive function.41 Future research should focus on validating MoCA as a screening instrument for HF, especially test-retest stability, since this has not been tested in our study or the few HF studies that used MoCA.25,27


This was a cross-sectional study with no control patients. We used the MoCA scale, which is a simple yet detailed cognitive screening instrument, and used the 8 domain scores that were not compared against detailed neuropsychological battery. Systolic and diastolic HF were determined from EF values that were extracted from chart review from within 6-month data. This inconsistency in EF measurement is not an accurate assessment for comparing with cognitive profile.

Conclusions and Clinical Implication

Patient adherence to medication regimens and self-care management of symptoms is always a concern for healthcare professionals working to minimize readmission rates for HF. Given the higher prevalence of cognitive impairment among persons with systolic and diastolic HF, these individuals need to be carefully screened and evaluated for cognitive impairment before these deficits alter their ability to perform activities of daily living and compliance with HF medication and self-care management. Persons with diastolic HF need to be included in future research to design tailored interventions.