Early View
Original Research Article
Open Access

Cognitive impairment in patients with heart failure: an international study

Ercole Vellone

Department of Biomedicine and Prevention, University of Rome Tor Vergata, Rome, Italy

Search for more papers by this author
Oronzo Chialà

Department of Biomedicine and Prevention, University of Rome Tor Vergata, Rome, Italy

Search for more papers by this author
Josiane Boyne

Department of Cardiology, Maastricht University Medical Centre, Maastricht, The Netherlands

Search for more papers by this author
Leonie Klompstra

Department of Social and Welfare Studies, Faculty of Health Science, Linköping University, Linköping, Sweden

Search for more papers by this author
Lorraine S. Evangelista

Sue & Bill Gross School of Nursing, University of California Irvine, Irvine, CA, USA

Search for more papers by this author
Maria Back

Division of Physiotherapy, Department of Medical and Health Sciences, Linköping University, Linköping, Sweden

Search for more papers by this author
Tuvia Ben Gal

Department of Cardiology, Rabin Medical Center (Beilinson Campus), Petah Tikva, Israel

Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel

Search for more papers by this author
Jan Mårtensson

Department of Nursing, School of Health and Welfare, Jönköping University, Jönköping, Sweden

Search for more papers by this author
Anna Strömberg

Division of Physiotherapy, Department of Medical and Health Sciences, Linköping University, Linköping, Sweden

Search for more papers by this author
Tiny Jaarsma

Corresponding Author

E-mail address: tiny.jaarsma@liu.se

Department of Social and Welfare Studies, Faculty of Health Science, Linköping University, Linköping, Sweden

Correspondence to: Tiny Jaarsma, Department of Medical and Health Sciences, Linkoping University, Linkoping, Sweden. Email: tiny.jaarsma@liu.se

Search for more papers by this author
First published: 18 December 2019

Abstract

Aims

Cognitive impairment (CI) in heart failure (HF) patients has mostly been studied in single countries in specific health care settings. Sociodemographic and clinical predictors of the global CI and CI dimensions are still unclear. We described CI in a diverse HF population recruited in several countries and in different health care settings and investigated sociodemographic and clinical factors associated with the global and specific CI dimensions in HF patients.

Methods and results

A secondary analysis from the baseline data of the Wii‐HF trial. Patients (n = 605) were enrolled in Sweden, Italy, Israel, The Netherlands, Germany, and the United States. We used the Montreal Cognitive Assessment to evaluate CI and the 6 minute walk test (6MWT) to measure exercise capacity. Patients were on average 67 years old (SD, 12), and 86% were in New York Heart Association Class II and III. The mean Montreal Cognitive Assessment score was 24 (SD, 4), and 67% of patients had at least a mild CI. The item evaluating short‐term memory had a considerable proportion of low scoring patients (28.1%). Worse CI was associated with patients' older age, lower education, and lower 6MWT scores (R2 = 0.27). CI dimension scores were differently associated with specific clinical and demographic variables, but the 6MWT scores were associated with five out of seven CI dimension scores.

Conclusions

CI is an important problem in HF patients, with specific challenges in regard to memory. Exercise capacity is a modifiable factor that could be improved in HF patients with the potential to improve cognition and other outcomes in this population.

1 Introduction

Heart failure (HF) is a prominent public health problem in the western world as a result of its rising prevalence and incidence.1 Because of the ageing population and the improved treatment of cardiac diseases, there are about 15 million people affected by HF in Europe1 and more than 6 million in the United States.2 HF is associated with several symptoms (e.g. shortness of breath and fatigue) that impair quality of life3 and is also associated with high mortality rates, recurrent hospitalizations, and high direct and indirect costs.4-6

Cognitive impairment (CI) is a common co‐morbid condition in patients with HF.7 Several studies have shown that CI in HF ranges between 20% and 80%,7, 8 and that CI is associated with poor HF outcomes, such as death, hospital readmission, and poor self‐care.9 Despite that, the pathophysiology of CI in patients with HF remains unclear,10 although several studies have succeeded in identifying its sociodemographic and clinical predictors. Sociodemographic predictors of CI in HF include older age,11 female gender, and lower education;12 and clinical predictors include decreased physical activity,13 co‐morbidities14 and HF severity [i.e. higher New York Heart Association (NYHA) Class and lower ejection fraction].15

Although several studies have addressed CI in patients with HF,7, 8 the existing studies have mostly been conducted in small samples, in single countries, and specific health care settings8 (e.g. only rehabilitation or outpatient settings). Also, the existing studies report inconsistencies among sociodemographic and clinical predictors of CI (i.e. gender, age, education, time after diagnosis, NYHA class, ejection fraction, comorbidities, and exercise capacity) and have not identified the sociodemographic and clinical predictors of specific cognitive dimensions in patients with HF (e.g. memory). Since patients with HF receive care in a large diversity of health care settings and their cognitive dimensions might be affected by different sociodemographic and clinical variables, we conducted this study to (i) describe CI in a diverse HF population recruited in several countries and in different health care settings and (ii) investigate sociodemographic and clinical factors (i.e. gender, age, education, time after diagnosis, NYHA class, ejection fraction, comorbidities, and exercise capacity) associated with global and specific CI in patients with HF.

2 Methods

2.1 Design

This study is a secondary analysis from the baseline data of the Wii‐HF study,16 a randomized controlled trial (ClinicalTrial.gov identifier: NCT01785121) aimed at evaluating the effect of exergaming on improving exercise capacity in patients with HF enrolled in six countries.

2.2 Sample and settings

In the Wii‐HF study, patients with HF were enrolled in different settings in Sweden, Italy, Israel, The Netherlands, Germany, and the United States. In Sweden, patients were enrolled in outpatient clinics of five hospitals, two university hospitals, two county hospitals, and one community hospital. In Italy, all patients were enrolled in a rehabilitation hospital; in Israel, patients were recruited from an advanced HF treatment centre; in The Netherlands and the United States, patients were enrolled in an outpatient clinic of a university hospital; and in Germany, from a private cardiology practice. Inclusion criteria in each country were (i) diagnosis of HF with reduced or preserved ejection fraction according to the European Society of Cardiology guidelines1 in NYHA functional Class I‐IV; (ii) age older than 18 years; and (iii) able to speak and understand the language of the enrolled country. The exclusion criteria were (i) unable to use the Wii computer because of visual, hearing, and motor impairment; (ii) unable to complete the research instruments; and, (iii) life expectancy shorter than 6 months. CI was not an exclusion criterion; however, if patients were not able to participate because of severe CI (e.g. dementia) as estimated by the study team, the patient was not eligible for the study. Two patients were excluded from the study because of their CI.

2.3 Data collection

Several instruments were used in the parent study,16 but for this analysis we used the instruments described next. The Montreal cognitive assessment (MoCA),17 is a 30‐item valid and reliable instrument that is used worldwide (also in patients with HF) to measure cognition. It evaluates seven cognitive domains as follows: visuospatial/executive, naming, attention, language, abstraction, delayed recall, and orientation. From the MoCA it is possible to obtain a total score (ranging between 0 and 30) and a dimension score with a different range across the dimensions: a lower score means worse CI both for the total and dimension scores. As defined in prior studies,18 the following categories of CI can be identified according to the MoCA score: a score between 27 and 30 means ‘no to light CI’; a score between 18 and 26 means ‘mild CI’, a score between 10 and 17 means ‘moderate CI’, and a score lower than 10 means ‘severe CI’.

The Charlson comorbidity index (CCI)19 is a widely used instrument that evaluates the presence of 19 co‐morbid conditions according to their gravity. Each condition could have a possible score from 1 to 6, with a higher score meaning worse co‐morbidity.

The 6 minute walk test (6MWT)20 is a valid, reliable, and objective test of submaximal exercise capacity that measures the distance an individual is able to walk over a total of 6 min on a hard, flat surface. The 6MWT score is obtained from the metres the patient walks in 6 min, with a higher score meaning better exercise capacity. A 6MWT score >300 is associated with lower event‐free survival at 36 months.21

Sociodemographic variables were collected for this study through self‐report, including gender, age, education, and marital status. Clinical variables, including time since diagnosis (in months), NYHA functional class and ejection fraction were abstracted from patients' medical records.

2.4 Ethical considerations

The protocol of the study was approved by the competent Institutional Review Board in each centre of the six countries in which the patients were enrolled (Sweden DNR 2012/247‐31, The Netherlands NL48647.068.14/METC141085; Italy 0052838/272/U.V.F/1, 2014; Israel 0022‐13‐RMC; Germany S22(a)/2015; United States UCI IRB HS# 2016‐2955). Before data collection, patients were fully informed about the study protocol. Data collection began only after patients signed the informed consent form.

2.5 Statistical analysis

Descriptive statistics, including means, standard deviations, and frequencies were used to analyse sociodemographic and clinical variables as well as MoCA, CCI, and 6MWT scores. Comparison across MoCA score categories (no to light CI, mild CI, moderate CI, and severe CI), and patient sociodemographic and clinical variables were performed with χ2 and ANOVA as appropriate. Spearman's correlations were used to identify the significant correlations among the variables of interest. Variables significantly and independently associated with the total MoCA and dimension scores were identified with a series of regression analyses (enter procedure). In each regression, the independent variables were gender, age, education, time after diagnosis in months, NYHA class, ejection fraction, CCI, and 6MWT scores. Data were analysed with the statistical software IBM SPSS 21.

3 Results

The total sample included 605 patients with HF (Table 1). The sample mean age was 67 (SD, 12), and most were male (71%). In 68% of cases, patients were educated to primary or high school level, and 72% were married. The average CCI score was 2.51 (SD, 1.71), meaning that on average patients suffered from one to two diseases other than HF. The most prevalent comorbidity was diabetes (27%). Eighty‐six percent of participants were in NYHA Class II and III, and the mean ejection fraction was 38.7% (11.9). The aetiology of HF was ischemic in 42% of patients. The 84% of patients were treated with angiotensin‐converting enzyme inhibitor/angiotensin II receptor blockers/angiotensin receptor neprilysin inhibitors, 87% with B‐Blocker, 48% with mineralocorticoid receptor antagonist, 10% with digoxin, and 65% with digoxin.

Table 1. Sociodemographic and clinical characteristics of enrolled patients (n = 605)
Characteristic N (%)
Total sample (n = 605) Sweden (n = 333) Italy (n = 96) Israel (n = 60) The Netherlands (n = 84) Germany (n = 24) United States (n = 8)
Age (mean, SD) 66.9 (11.7) 67.2 (11.8) 71.4 (9.7) 60.5 (12.9) 65.4 (10.6) 69.1 (9.1) 58.1 (8.8)
Gender (male) 430 (71.1) 229 (68.8) 70 (72.9) 44 (73.3) 64 (76.2) 17 (70.8) 6 (75.0)
Education
Primary School 145 (24.2) 85 (25.9) 36 (37.5) 6 (10.0) 8 (9.5) 10 (21.7) 0 (0)
High School 266 (44.4) 140 (42.7) 36 (37.5) 32 (53.3) 50 (59.5) 6 (25.0) 2 (28.6)
University 177 (29.3) 100 (30.5) 23 (24.0) 21 (35.0) 25 (29.8) 3 (12.5) 5 (71.4)
Other 11 (1.8) 3 (0.9) 1 (1.0) 1 (1.7) 1 (1.2) 5 (20.8) 0 (0)
Marital status
Married/relationship 430 (71.7) 234 (71.1) 56 (58.3) 50 (83.3) 64 (76.2) 21 (87.5) 5 (71.4)
Single/divorced/widowed 164 (27.3) 93 (28.3) 37 (38.5) 10 (16.7) 20 (23.8) 2 (8.3) 2 (28.6)
Other 6 (1.0) 2 (0.6) 3 (3.1) 0 (0) 0 (0) 1 (4.2) 0 (0)
Charlson Comorbidity Index Score (mean, SD) 2.51 (1.71) 2.22 (1.40) 2.91 (1.85) 3.61 (2.57) 2.24 (1.18) 3.08 (2.50) 2.50 (1.77)
Diabetes 159 (26.6) 71 (21.8) 34 (35.4) 23 (38.3) 19 (22.6) 10 (41.7) 2 (25.0)
COPD 108 (18.2) 24 (7.4) 43 (44.8) 21 (35.0) 17 (20.5) 1 (4.2) 2 (25.0)
Stroke 58 (9.7) 32 (9.9) 13 (13.5) 8 (13.3) 5 (6.0) 0 (0) 0 (0)
NYHA class
I 54 (8.9) 28 (8.8) 4 (4.2) 4 (6.7) 15 (21.1) 1 (4.2) 2 (25.0)
II 350 (57.9) 202 (63.1) 48 (50.0) 33 55.0) 45 (63.4) 17 (70.8) 5 (62.5)
III 171 (28.3) 90 (28.1) 41 (42.7) 23 (38.3) 10 (14.1) 6 (25.0) 1 (12.5)
IV 4 (0.7) 0 (0) 3 (3.1) 0 (0) 1 (1.4) 0 (0) 0 (0)
EF (mean, SD) 38.7 (11.9) 63.4 (10.9) 43.53 (9.8) 32.52 (10.5) 41.68 (13.0) 45.54 (14.9) 39.21 (15.9)
HF aetiology
Ischemic 249 (42.1) 127 (39.6) 19 (19.8) 32 (53.3) 61 (73.5) 9 (37.5) 1 (14.3)
Non‐ischemic 342 (57.9) 194 (60.4) 77 (80.2) 28 (46.7) 22 (26.5) 15 (62.5) 6 (85.7)
Medications
ACE inhibitor/ARB/ARNI 506 (83.6) 317 (95.5) 40 (41.7) 53 (88.3) 72 (85.7) 20 (83.3) 4 (50.0)
B‐Blocker 522 (86.9) 312 (94.3) 52 (54.2) 54 (91.5) 77 (91.7) 21 (87.5) 6 (85.7)
MRA 290 (47.9) 164 (49.5) 26 (27.1) 43 (74.1) 47 (56.0) 8 (33.3) 2 (28.6)
Digoxin 61 (10.1) 26 (7.8) 5 (5.2) 15 (25.9) 10 (11.9) 5 (20.8) 0 (0)
Diuretics 392 (64.8) 195 (55.7) 81 (84.4) 38 (63.3) 58 (69.0) 19 (82.6) 4 (50)
6MWT score (mean, SD) 402.9 (141.7) 449.6 (110.3) 222.4 (114.3) 442.2 (114.6) 432.9 (125.9) 280.5 (116.4) 423.2 (57.5)
  • ACE, angiotensin‐converting enzyme; ARB, angiotensin II receptor blockers; ARNI, angiotensin receptor neprilysin inhibitors; COPD, chronic obstructive pulmonary disease; EF, ejection fraction; HF, heart failure; NYHA, New York Heart Association; MRA, mineralocorticoid receptor antagonist; 6MWT, 6 minute walking test.

The mean total MoCA score was 24.53; 33% of the sample had no or a light CI; 62 % had mild CI; 5% had moderate CI; and 0.3% had severe CI (Table 2). Table 3 reports the number (and percentage) of patients per each MoCA item score. Item 9, evaluating delayed recall, had a considerable proportion of patients (28.1%) who had a low score. The supporting information table reports the mean, SD, minimum, and maximum score per each MoCA item.

Table 2. Cognitive impairment in the sample (n = 605)
Mean (SD) or N (%)
Moca Total sample (n = 605) Sweden (n = 333) Italy (n = 96) Israel (n = 60) The Netherland (n = 84) Germany (n = 24) USA (n = 8)
MoCA total score 24.35 (3.84) 25.02 (2.95) 20.53 (4.98) 24.45 (3.18) 25.73 (3.05) 25.61 (2.87) 24.13 (6.64)
MoCA categories
No to Light CI (27–30) 195 (32.7) 119 (36.1) 7 (7.3) 16 (27.6) 38 (46.3) 11 (47.8) 4 (50.0)
Mild CI (18–26) 371 (62.1) 204 (61.8) 68 (70.8) 42 (72.4) 42 (51.2) 12 (52.2) 3 (37.5)
Moderate CI (10–17) 29 (4.9) 7 (2.1) 19 (19.8) 0 (0) 2 (2.4) 0 (0) 1 (12.5)
Severe CI (0–9) 2 (0.3) 0 (0) 2 (2.1) 0 (0) 0 (0) 0 (0) 0 (0)
MoCA items
Item n.1 Visuospatial/executive (potential score: 0–5) 3.75 (1.23) 3.83 (1.12) 2.75 (1.38) 3.92 (1.02) 4.33 (0.93) 4.38 (1.10) 3.75 (1.75)
Item n.2 Naming (potential score: 0–3) 2.91 (0.36) 2.97 (0.19) 2.75 (0.65) 2.86 (0.35) 2.87 (0.43) 3.00 (0.00) 2.87 (0.35)
Item n.3 Attention (potential score: 0–2) 1.75 (0.53) 1.78 (0.48) 1.48 (0.74) 1.85 (0.36) 1.82 (0.44) 1.79 (0.41) 1.88 (0.35)
Item n.4 Attention (potential score: 0–1) 0.94 (0.23) 0.97 (0.17) 0.84 (0.36) 0.92 (0.28) 0.99 (0.11) 0.96 (0.20) 0.88 (0.35)
Item n.5 Attention (potential score: 0–3) 2.69 (0.69) 2.74 (0.63) 2.35 (0.95) 2.66 (0.76) 2.88 (0.33) 2.87 (0.34) 2.25 (1.16)
Item n.6 Language (potential score: 0–2) 1.76 (0.52) 1.85 (0.44) 1.53 (0.60) 1.54 (0.68) 1.81 (0.45) 1.71 (0.46) 1.75 (0.71)
Item n.7 Language (potential score: 0–1) 0.58 (0.49) 0.61 (0.49) 0.50 (0.50) 0.53 (0.50) 0.61 (0.49) 0.30 (0.47) 0.88 (0.35)
Item n.8 Abstraction (potential score: 0–2) 1.67 (0.58) 1.68 (0.57) 1.56 (0.63) 1.56 (0.62) 1.74 (0.54) 2.00 (0.00) 1.75 (0.71)
Item n.9 Delayed recall (potential score: 0–5) 2.49 (1.58) 2.67 (1.46) 1.40 (1.47) 2.68 (1.69) 2.87 (1.60) 2.65 (1.70) 2.62 (1.60)
Item n.10 Orientation (potential score: 0–6) 5.82 (0.55) 5.92 (0.31) 5.36 (1.05) 5.95 (0.22) 5.86 (0.35) 5.96 (0.21) 5.50 (0.93)
  • MoCA, Montreal Cognitive Assessment.
Table 3. Number (and percentage) of patients per each MoCA item score
MoCA item N (%)
Score 0 Score 1 Score 2 Score 3 Score 4 Score 5 Score 6
Item n.1 Visuospatial/executive (potential score: 0–5) 3 (0.5) 35 (5.8) 65 (10.7) 112 (18.7) 174 (29) 211 (35.2)
Item n.2 Naming (potential score: 0–3) 4 (0.7) 3 (0.5) 36 (6) 558 (92.8)
Item n.3 Attention (potential score: 0–2) 26 (4.3) 101 (16.8) 474 (78.9)
Item n.4 Attention (potential score: 0–1) 33 (5.5) 567 (93.7)
Item n.5 Attention (potential score: 0–3) 13 (2.2) 41 (6.8) 65 (10.8) 482 (80.2)
Item n.6 Language (potential score: 0–2) 25 (4.2) 97 (16.1) 479 (79.7)
Item n.7 Language (potential score: 0–1) 253 (42.2) 346 (57.8)
Item n.8 Abstraction (potential score: 0–2) 34 (5.7) 131 (21.8) 436 (72.5)
Item n.9 Delayed recall (potential score: 0–5) 104 (17.3) 60 (10.8) 109 (18.2) 157 (26.2) 103 (17.2) 67 (11.2)
Item n.10 Orientation (potential score: 0–6) 1 (0.2) 1 (0.2) 0 (0) 3 (0.5) 13 (2.2) 60 (10.8) 522 (87)
  • MoCA, Montreal cognitive assessment.

Table 4 reports the differences among the variables in relation to the MoCA score categories as follows: no to light CI, mild CI, moderate CI, and severe CI. Patients in the moderate CI and severe CI were significantly older, with less education, with higher CCI scores, in a more advanced NYHA classification and with a lower score at the 6MWT.

Table 4. Differences among the variables in relation to the MoCA score categories
Characteristic MoCA score category
No to light CI (Scores 27–30) M (SD) or N (%) Mild CI (Score 18–26) M (SD) or N (%) Moderate CI (Score 10–17) M (SD) or N (%) Severe CI (Score 0–9) M (SD) or N (%) P
Age 62.1 (12.6) 68.6 (10.3) 77.1 (9.4) 83 (4.2) 0.000
Gender n.s.
Male 132 (31.1) 273 (64.4) 18 (4.2) 1 (0.2)
Female 63 (36.4) 98 (56.6) 11 (6.4) 1 (0.6)
Education 0.000
Primary School 30 (20.8) 98 (68.1) 16 (11.1) 0 (0.0)
High School 83 (31.7) 171 (65.3) 7 (2.7) 1 (0.4)
University 78 (44.8) 92 (52.9) 4 (2.3) 0 (0.0)
Marital status n.s.
Married/relationship 145 (34.3) 262 (61.9) 14 (3.3) 2 (0.5)
Single/divorced/widowed 48 (29.4) 102 (62.6) 13 (8.0) 0 (0.0)
CCI Score 0.025
1 78 (40.6) 107 (55.7) 7 (3.6) 0 (0.0)
2 51 (30.5) 108 (64.7) 7 (4.2) 1 (0.6)
3 27 (29.3) 59 (64.1) 6 (6.5) 0 (0.0)
>4 24 (19.5) 89 (72.4) 9 (7.3) 1 (0.8)
NYHA class 0.000
I 24 (44.4) 28 (51.9) 2 (3.7) 0 (0.0)
II 127 (37.0) 206 (60.1) 10 (2.9) 0 (0.0)
III 31 (18.2) 120 (70.6) 17 (10.0) 2 (1.2)
IV 0 (0.0) 4 (100) 0 (0.0) 0 (0.0)
EF 38.4 (12.2) 38.7 (11.8) 40.6 (12.6) 44.0 (5.6) n.s.
6MWT score 463.3 (117.4) 385.9 (139.5) 229.3 (116.4) 163.5 (4.9) 0.000
  • Comparisons were performed with χ2 or ANOVA as appropriate.
  • CCI, Charlson comorbidity index; EF, ejection fraction; NYHA, New York Heart Association; MoCA, Montreal cognitive assessment; 6MWT, 6 minute walking test.

3.1 Correlation and regression analysis

In the correlation analysis, lower MoCA total scores (meaning worse CI) were significantly associated with older age, lower education, higher NYHA class, lower CCI scores, and lower 6MWT scores (Table 5). When all the above variables were entered in a regression model (Table 6), lower total MoCA scores (worse CI) were independently associated with patients' older age, lower education, and lower 6MWT scores; the same variables were significantly and independently associated with lower visuospatial/executive dimension scores. Lower naming dimension scores were associated with longer time from HF diagnosis and worse 6MWT scores; lower attention dimension scores were associated with lower education and worse 6MWT scores; lower language dimension scores were associated with lower education and 6MWT scores; lower abstraction dimension scores were associated with lower ejection fraction but with a P level of 0.052; lower delayed recall dimension scores were associated only with older age; finally, lower orientation dimension scores were associated with lower 6MWT scores.

Table 5. Correlations among the studied variables
Variable 1 2 3 4 5 6 7 8
1. Total MoCA score 1
2. Gender (Male = 1; Female = 2) 0.015 1
3. Age −0.322** **P < 0.01. 0.026 1
4. Education 0.244** **P < 0.01. 0.000 −.118** **P < 0.01. 1
5. Time after HF diagnosis (in months) 0.010 −0.041 0.015 −0.020 1
6. NYHA class −0.261** **P < 0.01. 0.107* *P < 0.05 and 0.243** **P < 0.01. −0.123** **P < 0.01. −0.021 1
7. Ejection fraction −0.036 0.105* *P < 0.05 and 0.171** **P < 0.01. −0.053 −0.095 0.028 1
8. Charlson comorbidity index score −0.236** **P < 0.01. −0.162** **P < 0.01. 0.221** **P < 0.01. −0.066 0.141** **P < 0.01. 0.264** **P < 0.01. −0.001 1
9. Six minute walk test score 0.396** **P < 0.01. −0.125** **P < 0.01. −0.397** **P < 0.01. 0.183** **P < 0.01. 0.052 −0.458** **P < 0.01. −0.230** **P < 0.01. 0−.287** **P < 0.01.
  • Lower MoCA scores mean worse cognitive impairment; each number of the first row corresponds to the variable with the
  • same number in the first column; and numbers in the cells are correlation coefficients.
  • HF, heart failure; NYHA, New York Heart Association; MoCA, Montreal cognitive assessment.
  • * P < 0.05 and
  • ** P < 0.01.
Table 6. Regression analysis predicting MoCA total score and MoCA dimension scores
Variable Total MoCA score Visuospatial/executive Naming Attention Language Abstraction Delayed recall Orientation
Gender (Male = 1; Female = 2) 0.053 −0.006 −0.093 −0.003 0.091 −0.095 0.095 0.061
Age ‐0.210*** ***P < 0.001. −0.166** **P < 0.01, −0.053 −0.099 0.032 −0.076 −0.186** **P < 0.01, −0.108#
Education 0.152** **P < 0.01, 0.168** **P < 0.01, 0.033 0.133* *P < 0.05, 0.213*** ***P < 0.001. 0.091 −0.070 −0.012
Time after HF diagnosis (in months) 0.023 0.042 −0.133* *P < 0.05, 0.011 −0.026 −0.079 0.033 0.081
NYHA class 0.010 −0.048 0.059 −0.063 0.018 −0.020 −0.029 0.022
Ejection fraction 0.070 0.082 0.017 0.000 0.093 0.109¥ 0.043 0.050
Charlson comorbidity index score ‐0.082 −0.026 0.010 −0.045 −0.072 −0.065 −0.007 −0.026
Six minute walk test score 0.311*** ***P < 0.001. 0.217*** ***P < 0.001. 0.135* *P < 0.05, 0.130* *P < 0.05, 0.249*** ***P < 0.001. 0.013 0.106 0.258*** ***P < 0.001.
R2 0.266 0.183 0.050 0.098 0.143 0.049 0.059 0.097
  • Lower MoCA scores mean worse cognitive impairment. Numbers in columns are standardized betas. ¥P 0.052; #P 0.055
  • HF, heart failure; NYHA, New York Heart Association; MoCA, Montreal cognitive assessment.
  • * P < 0.05,
  • ** P < 0.01,
  • *** P < 0.001.

4 Discussion

To our knowledge, this is the first international study to describe CI in a large diverse HF population recruited in six countries and in different health care settings and the first study to identify sociodemographic and clinical variables (i.e. gender, age, education, time after diagnosis, NYHA class, ejection fraction, comorbidities, and exercise capacity) independently associated with specific cognitive dimensions. These two aspects are important because prior studies have described CI in patients with HF but mostly in single countries (e.g. United States)7 or in single settings (e.g. outpatient settings).22 HF is a common health problem across the globe, and patients with HF are cared for in several health care settings (e.g. outpatient settings, rehabilitation, and private cardiology practices). Consequently, our findings provide a better and comprehensive representation of CI in patients with HF. This might have important clinical and research implications.

Sixty‐two percent (n = 371) of our sample had mild CI. This percentage is similar to that identified by Harkness et al.23 in a sample of 44 patients with HF in an outpatient setting (70%) but higher than that identified by Gallagher et al.,24 where 55% of patients with HF (n = 124) had a score <26 on the MoCA. The percentage of patients with CI falls within the range of prior studies that have been included in systematic reviews.10 Our finding corroborates further that CI is an important problem in patients with HF.

Looking at the MoCA scores across the six countries in which we enrolled the patients (Table 2), we saw significant differences across these subsamples. For example, Italian patients had the lowest score on the MoCA when compared with patients of other countries. However, Italian patients were enrolled in a rehabilitation hospital and could have been more cognitively compromised than other patients enrolled, for example, in outpatient settings, as in Sweden. Further studies with more comparable patients should be conducted to see if CI in patients with HF might be influenced by different approaches in HF treatment and cultural factors across countries. Interestingly, we found that 28.1 % scored 0 or 1 for delayed recall (short‐term memory), meaning that our population had important memory problems, which confirms other studies that found that memory and executive functions were cognitive areas particularly damaged in patients with HF.25 This finding is important to consider when delivering patient education and when asking patients to perform self‐care tasks, such as monitoring deterioration or taking medications.

In this study, we analysed sociodemographic and clinical variables independently associated with the global CI and with impairment of specific cognitive dimensions. Interestingly, we saw that older age, education, and exercise capacity were predictors of the global CI, a finding that was not observed across the specific domains. For example, time after diagnosis and exercise capacity were predictors of the naming dimension, and age was a predictor of delayed recall. To our knowledge, no prior studies have previously identified sociodemographic and clinical predictors of specific cognitive domains, and these new findings may be helpful to guide future investigations and interventions in patients with HF. For example, while patient's age predicted the visuospatial/executive domain, it was not associated with the naming domain. Interestingly, exercise capacity was associated with global CI and five out of the seven cognitive domains. This finding is important because exercise capacity could be improved in patients with HF. Prior studies have already found that physical activity was associated with better cognition in patients with HF.26

This study has strengths and limitations. A strength was the large multinational sample of patients with HF from multiple settings. We have given a wide description of CI in patients with HF. At the same time, our study strengths could be considered a limitation because the sample we enrolled was heterogeneous and selected with convenience procedures. Another limitation is that our study was a cross‐sectional and secondary analysis with patients originally selected to evaluate the effectiveness of exergaming to improve exercise capacity in patients with HF. Even though CI was not an exclusion criterion for enrolling patients and only two patients were excluded from the study because their CI was an impediment to play with the Wii computer, we might have had a selection bias. However, the distribution of CI in our population was not different from other studies. Another limitation of our study was that we used a screening tool to evaluate CI that could be better evaluated with other tools with better sensitivity and specificity.

Our study has important practical implications because patients with HF who are more cognitively compromised have worse outcomes, such as lower self‐care,22 higher mortality,27 and more frequent hospitalizations.28 Our findings showed that exercise capacity, a modifiable factor, was independently associated with CI. This means that improving physical activity has the potential to improve direct (e.g. cognition) and indirect (e.g. lower mortality rates) outcomes in patients with HF. Improving physical activity in HF is advocated in the international guidelines.1 We have also discovered from our analysis that specific cognitive domains have different predictors that may be useful to know in clinical practice. For example, a patient's older age was associated with worse delayed recall (memory) but not with abstraction. Practically, this means that older patients with HF may find it difficult to memorise the medications they have to take but not the fact that diuretics help to eliminate water from the body (Abstraction).

This study has several research implications. Since we observed different predictors for specific cognitive domains, it would be important to conduct further studies to confirm our findings. Another important implication is the use of longitudinal designs and randomized controlled trials to examine if increased physical activity is associated with improved cognition. A recent pilot study29 found that a programme of combined aerobic exercise and cognitive training improved memory in patients with HF, but further studies are needed to clarify this association.

Acknowledgement

The following persons played a vital role in data collection. Norrköping: L Nestor, C Norrman, M Viklander, A Waldemar, RM Petterson; M Wärfman; Jönköping: E Lundberg, H Sköldbäck, M Sahlin; Linköping: A Gylling, M Huss, M Jonsson, P Wodlin, L Hjelmfors; Stockholm: E Hägglund, U Lennmark; Nyköping: E Säfström; Italy: R Corsi, G Alberto Ortali; The Netherlands: HP Brunner‐La Rocca, M Spanjers, A van de Voorde, G Cleuren; Israel: S Donanhirsh, Y Navon, V Yaari B. Avraham; Germany: A Hagenow, A Kuntzsch; and United States: J Ardo, J Nguyen, M Cacciata.

    Conflict of interest

    None to declare

    Funding

    This work was supported through the Swedish National Science Council (K2013‐69X‐22302‐01‐3, 2016‐01390), Swedish National Science Council/Swedish Research Council for Health, Working Life and Welfare (VR‐FORTE) 2014‐4100, The Swedish Heart and Lung Association E085/12, The Swedish Heart and Lung Foundation (20130340 and 20160439), the Vårdal Foundation (2014–0018), the Medical Research Council of Southeast Sweden (FORSS 474681).

      The full text of this article hosted at iucr.org is unavailable due to technical difficulties.