Keywords:
cognition - coronary artery disease - echocardiography - left ventricular function
Palavras-chave:
cognição - doença da artéria coronariana - ecocardiografia - função ventricular esquerda
It has been estimated that the elderly population will rise from 7% in 2007 to 22%
in 2050[1]. Progressive cognitive decline, which may become dementia, is one of the most important
health problems in old age[1]. From the first neurocognitive sign to obvious dementia, there is usually a long
transitional period with a broad range of presentations. This transitional period
may have unremarkable signs of neurodegeneration without any clinically significant
symptoms (predementia state)[2]. Elucidating risk factors for possible impending dementia can help with the design
of preventive strategies and specific managements[3].
Vascular risk factors including hypertension, diabetes mellitus, and hyperlipidemia
are also prevalent in the elderly population, which can lead to higher incidences
of cerebrovascular accidents and myocardial infarction[4]. These factors may also carry some risk for neurodegeneration, even in situations
where cerebrovascular accidents and myocardial infarction are absent[5]. For example, coronary artery disease has been found to be an independent risk factor
for vascular dementia[4],[5],[6]. A review study reported that heart failure also correlated with decreased memory
registration, delayed free recall, working memory, executive function, and speed of
information processing. Among the various cognitive domains, language and visuospatial
performance were less affected by heart failure[7].
Although some studies have tried to evaluate the relationship between clinical presentations
of ischemic heart disease and cognitive functioning, few studies have used a definitive
index of coronary artery disease for this purpose[8]. Precise echocardiographic findings, including left ventricular systolic and diastolic
indices, left atrial morphologic parameters, cardiac output and aortic root diameter,
were used to show correlation between cognitive functioning and echocardiographic
indices[9]. Regarding methods of cognitive evaluation, there have been a few studies that specifically
administered neuropsychological tests to determine cognitive performance[10],[11].
In this study, we sought to find an association between cognitive functioning and
specific parameters of cardiac echocardiography and coronary artery angiography. In
addition, for a thorough evaluation of the cognitive profile, we used not only a neuropsychiatric
screening tool but also specific neuropsychological tests to obtain objective measurements
for each cognitive domain.
METHODS
Study design and participants
This study was approved by the Regional Bioethics Committee of Isfahan University
of Medical Sciences. We discussed this project with all of the participants and obtained
written informed consents.
Participants in this prospective study were the patients who underwent elective coronary
artery angiography at Sina and Chamran hospitals' cardiac catheterization facilities
in Isfahan, Iran in 2016. The inclusion criteria were: age 60 years or older, and
education of primary school or higher levels. The participants were referred during
the four weeks after elective coronary artery angiography for possible ischemic heart
disease. We screened for those fitting the inclusion criteria through telephone calls.
Based on their medical documents and hospital records, patients with a history of
coronary artery bypass graft or percutaneous coronary angioplasty, head trauma, serious
medical or neurological diseases, major psychiatric disorders, substance or medication-related
disorders, and dementia were not included. The included participants underwent a semi-structured
neuropsychiatry interview, assessments for cognitive functioning with specific neuropsychological
tests, and echocardiography. Patients who could not perform the neuropsychological
tests were excluded from the study. Baseline characteristics including age, gender,
educational level, weight, height, smoking and alcohol use, medical history of stroke,
diabetes mellitus, hypertension, hyperlipidemia, hypo/hyperthyroidism, and current
drugs were recorded. Specific laboratory tests were also investigated.
Cognitive assessments
The Neuropsychiatry Unit Cognitive Assessment Tool (NUCOG) was used to evaluate the
patients' cognitive function in the following domains: attention, visuospatial, memory,
executive function, and language. The maximum score for each domain was 20 (with a
total score of 100). In the Persian version of the NUCOG, the cutoff points for separating
mild cognitive impairment from normal individuals and patients with dementia were
86.5 and 75 respectively[8]. The Tower of London (TOL) test was selected to assess deficits in planning and
set shifting of executive functioning. The Color Trail Test (CTT) was used for evaluation
of processing speed/attention in CTT-part 1 and divided attention and executive function
in CTT-part 2[12]. Verbal memory was assessed, using the Auditory Verbal Learning Test (AVLT)[13]. Two sets of tasks were used to assess language function, including the Persian
Picture Naming Battery[14] and the Persian Diagnostic Aphasia Battery[15]. These tests were administered by a well-trained neuropsychologist.
Evaluation of cardiac function
Two-dimensional echocardiography, with a GE Vivid 3 echocardiography device (General
Electric Company, Milwaukee, WI, USA), was performed according to the American Society
of Echocardiography standards, by a cardiology resident who was blinded to the neuropsychiatric
tests. The selected parameters included left ventricular ejection fraction using the
volume method and visual estimation. Indices of diastolic function on the basis of
the peak mitral inflow velocity of the early rapid filling wave (E) and late filling
wave caused by atrial contraction (A) and myocardial peak early velocity (e') from
septal mitral annulus tissue doppler images were obtained, and the E/e' ratio was
calculated. To assess the left atrial size, the end-systolic anteroposterior diameter
in long axis view and the end-systolic left atrial surface area in 4-chamber view
were measured[16]. Coronary artery angiography records were reviewed. The Gensini score was calculated
by multiplying the severity of stenosis by the segment location and collateral adjustment
factor. Higher scores in the Gensini indicated more severe coronary artery disease[17].
Statistical analysis
All statistical analyses were carried out using the SPSS 18 (SPSS crop. Chicago, IL,
USA). Quantitative and qualitative variables were calculated as mean ± standard deviation
and proportion in percentage, respectively. Pearson's correlation was used to identify
the strength and direction of any correlation between cardiac and cognitive quantitative
variables with statistical significance level p ≤ 0.05. Based on the NUCOG cutoff
point of 86, patients were divided into two categories (normal and impaired cognition)
to compare demographic, clinical and neuropsychological variables. The association
between the cardiovascular parameters and cognitive variables were also evaluated,
using Multivariate Analysis of Variance, which was adjusted for potential confounders.
RESULTS
Of 153 referred patients, 60 did not meet the full inclusion criteria and eight individuals
were excluded due to their refusal to participate or inability to perform the tests.
A total of 85 patients were enrolled in the study. [Table 1] shows the baseline characteristics and statistical measures. The mean sample age
was 65.78 ± 5.13 and 85.9% were men. The NUCOG score average was 80.56 ± 8.30.
Table 1
Demographic characteristics and clinical and biochemical parameters of the study population.
Clinical variables
|
n (percentage) mean ± SD
|
Demographic features
|
|
Age
|
65.78 ± 5.13
|
|
Male gender (%)
|
73 (85.9)
|
|
Education:
|
|
|
5–8 years
|
30 (35.3)
|
|
9–12 years
|
27 (31.8)
|
|
> 12 years
|
28 (32.9)
|
|
History of smoking (%)
|
33 (38.3)
|
|
BMI
|
25.16 ± 3.31
|
Medical history
|
|
Diabetes mellitus (%)
|
25 (29.4)
|
|
Hypertension (%)
|
34 (40)
|
|
CVA (%)
|
6 (7.1)
|
|
Hypothyroidism (%)
|
5 (5.9)
|
|
History of smoking (%)
|
33 (38.3)
|
Laboratory data
|
|
Fasting plasma glucose (mg/dl)
|
109.49 ± 28.85
|
|
HbA1C (%)
|
6.27 ± 1.56
|
|
Total cholesterol (mg/dl)
|
151.42 ± 32.54
|
|
Triglycerides (mg/dl)
|
144.63 ± 69.18
|
|
HDL (mg/dl)
|
41.84 ± 8.97
|
|
LDL (mg/dl)
|
86.66 ± 64.58
|
|
Creatinine (mg/dl)
|
1.14 ± 0.39
|
Cognitive measures
|
|
NUCOG total score
|
80.56 ± 8.30
|
|
NUCOG A (attention)
|
13.32 ± 2.67
|
|
NUCOG B (visuospatial)
|
17.62 ± 2.10
|
|
NUCOG C (memory)
|
14.46 ± 2.68
|
|
NUCOG D (executive function)
|
15.64 ± 3.09
|
|
NUCOG E (language)
|
19.37 ± 1.10
|
|
CTT-time 1
|
99.23 ± 38.22
|
|
CTT-time 2
|
208.62 ± 88.07
|
|
TOL-score
|
34.81 ± 1.49
|
|
TOL-time
|
286.08 ± 74.96
|
|
P-DAB (AQ)
|
94.81 ± 3.11
|
|
P-PNB-score
|
47.24 ± 7.17
|
|
P-PNB-time
|
139.8 ± 67.6
|
|
AVLT (5th trial)
|
9.52 ± 2.4
|
|
AVLT (after 30 minutes)
|
6.56 ± 2.91
|
Echocardiography measures
|
|
LVEF (%)
|
54.94 ± 8.36
|
|
E (m/s)
|
0.67 ± 0.17
|
|
A (m/s)
|
0.83 ± 0.16
|
|
E/A
|
0.83 ± 0.24
|
|
e' (m/s)
|
0.06 ± 0.01
|
|
E/e'
|
11.04 ± 3.28
|
|
LA diameter (cm)
|
3.46 ± 0.47
|
|
LA area (cm2)
|
15.28 ± 3.71
|
|
Gensini Score
|
36.20 ± 23.86
|
Data are presented as mean ± SD and n (%). BMI: body mass index; CVA: cerebrovascular
accident; HDL: high density lipoprotein; LDL: low density lipoprotein; NUCOG: Neuropsychiatry
unit cognitive assessment tool; CTT: Color trail test; TOL: Tower of London test;
P-DAB: Persian diagnostic aphasia battery; P-PNB: Persian picture naming battery;
AVLT: Auditory verbal learning test; LVEF: left ventricular ejection fraction; E:
early mitral valve flow velocity; A: late mitral valve flow velocity; e': diastolic
lengthening velocity; LA: left atrial.
Considering all the patients in a group, the correlation of echocardiography and angiography
indices with NUCOG scores, CTT, and AVLT tests are shown in [Table 2]. The left ventricular ejection fraction was positively correlated with the NUCOG
score (P = 0.005, R = 0.30), CTT-parts 1 and 2 (p = 0.02, R = 0.024 and p = 0.01,
R = 0.27), AVLT 5th trial and after 30 minutes (p = 0.03, R = 0.23 and p = 0.01, R = 0.26). The left
atrial area and dimension were positively correlated with CTT tests as shown in [Table 2], and higher Gensini scores were also associated with higher CTT-times 1 and 2 (p
= 0.001, R = 0.35 and p = 0.01, R = 0.27).
Table 2
Correlation of cardiac parameters with cognitive tasks.
Variable
|
Gensini score
|
LVEF (%)
|
LAD (cm)
|
LA area (cm2)
|
E (m/s)
|
e' (m/s)
|
E/e'
|
R
|
p
|
R
|
p
|
R
|
p
|
R
|
p
|
R
|
p
|
R
|
p
|
R
|
P
|
NUCOG-total
|
-0.10
|
0.33
|
0.30
|
0.005
|
-0.11
|
0.27
|
-0.18
|
0.09
|
0.003
|
0.98
|
-0.007
|
0.95
|
-0.04
|
0.66
|
A (attention)
|
-0.12
|
0.25
|
0.18
|
0.9
|
-0.02
|
0.79
|
-0.2
|
0.06
|
-0.08
|
0.42
|
-0.07
|
0.52
|
-0.04
|
0.7
|
B (visuospatial)
|
-0.16
|
0.13
|
0.23
|
0.03
|
-0.03
|
0.76
|
-0.03
|
0.77
|
-0.004
|
0.97
|
0.05
|
0.6
|
-0.12
|
0.27
|
C (memory)
|
0.0
|
0.99
|
0.18
|
0.08
|
-0.18
|
0.08
|
-0.17
|
0.10
|
-0.05
|
0.59
|
0.03
|
0.72
|
-0.11
|
0.30
|
D (executive function)
|
-0.04
|
0.68
|
0.26
|
0.01
|
-0.11
|
0.28
|
-0.16
|
0.14
|
0.1
|
0.34
|
-0.04
|
0.66
|
0.10
|
0.34
|
E (language)
|
-0.04
|
0.70
|
0.16
|
0.14
|
-0.01
|
0.86
|
0.002
|
0.98
|
0.06
|
0.57
|
0.09
|
0.4
|
-0.11
|
0.29
|
CTT-time 1
|
0.35
|
0.001
|
-0.24
|
0.02
|
0.28
|
0.008
|
0.28
|
0.008
|
0.12
|
0.27
|
-0.01
|
0.89
|
0.11
|
0.31
|
CTT-time 2
|
0.22
|
0.03
|
-0.27
|
0.01
|
0.2
|
0.06
|
0.25
|
0.01
|
0.1
|
0.36
|
0.11
|
0.31
|
0.01
|
0.88
|
AVLT (5th trial)
|
-0.03
|
0.74
|
0.23
|
0.03
|
-0.09
|
0.4
|
-0.2
|
0.06
|
-0.03
|
0.74
|
-0.03
|
0.74
|
0.03
|
0.74
|
AVLT (30 minutes)
|
-0.1
|
0.32
|
0.26
|
0.01
|
-0.03
|
0.74
|
-0.06
|
0.56
|
-0.01
|
0.89
|
-0.04
|
0.71
|
0.07
|
0.50
|
NUCOG: Neuropsychiatry unit cognitive assessment tool; CTT: Color trail test; AVLT:
Auditory verbal learning test; LVEF: left ventricular ejection fraction; E: early
mitral valve flow velocity; e': diastolic lengthening velocity; LA: left atrial; LAD:
Left Atrial Diameter.
In addition, there was a positive correlation between the left atrial diameter and
area with the TOL-time (p = 0.003, R = 0.32 and P = 0.001, R = 0.34 respectively).
No association was found between other cardiac indices and cognitive measures (p >
0.5)
On the NUCOG cutoff, patients were divided into normal individuals and patients with
impaired cognition. Of the latter, 73% had an abnormal NUCOG score and 84% were men
in this group. Using Multivariate Analysis of Variance with age and sex adjustment,
there was a positive association between the Gensini score and CTT-time 1 (P = 0.01,
SE = 0.26 and 95% CI [0.05, 0.47]) in patients with impaired cognition. The e' as
an index of diastolic function was negatively correlated with CTT-time 2 (p = 0.05,
SE = 6.52 and 95%CI [-2.53, 0.00]) in the same group. There was a negative correlation
between the e' and AVLT immediate free recall of the 5th trial in the normal cognition group (p = 0.04, SE = 0.003, 95%CI [-0.001, 0.00]).
Other data are summarized in [Table 3].
Table 3
Age and sex adjusted associations between echocardiography and coronary artery angiography
measures and cognitive functioning using Multivariate Analysis of Variance (MANOVA).
Variable
|
Impaired cognition
|
Normal Cognition
|
Based on NUCOG < 86
|
Based on NUCOG ≥ 86
|
Number (%)
|
62 (72.94)
|
23 (27.05)
|
Age
|
66.06 ± 5.22
|
65.00 ± 4.90
|
Male (%)
|
5 (83.87)
|
19 (82.60)
|
Effect
|
CTT-time 1
|
CTT-time 2
|
AVLT (5th trial)
|
AVLT (30 minutes)
|
CTT-time 1
|
CTT-time 2
|
AVLT (5th trial)
|
AVLT (30 minutes)
|
p
|
t
|
p
|
t
|
p
|
t
|
p
|
t
|
p
|
t
|
p
|
t
|
p
|
t
|
p
|
t
|
Gensini score
|
0.01
|
2.53
|
0.50
|
0.66
|
0.11
|
1.59
|
0.47
|
-0.72
|
0.51
|
0.67
|
0.36
|
0.92
|
0.88
|
-0.14
|
0.48
|
0.70
|
LVEF (%)
|
0.16
|
-1.14
|
0.28
|
-1.07
|
0.55
|
-0.59
|
0.31
|
1.01
|
0.55
|
-0.60
|
0.26
|
1.17
|
0.16
|
-1.45
|
0.80
|
0.24
|
LA diameter (cm)
|
0.20
|
1.18
|
0.23
|
1.13
|
0.97
|
0.03
|
0.82
|
0.22
|
0.31
|
1.03
|
0.57
|
-0.56
|
0.76
|
-0.30
|
0.51
|
0.67
|
LA area (cm2)
|
0.24
|
1.18
|
0.17
|
1.37
|
0.45
|
-0.74
|
0.70
|
0.37
|
0.89
|
0.13
|
0.77
|
0.27
|
0.32
|
-1.02
|
0.17
|
1.43
|
E(m/s)
|
0.17
|
1.37
|
0.94
|
-0.07
|
0.94
|
-0.07
|
0.88
|
-0.14
|
0.90
|
0.11
|
0.66
|
-0.44
|
0.52
|
0.65
|
0.41
|
-0.84
|
e'(m/s)
|
0.46
|
-0.74
|
0.05
|
1.93
|
0.59
|
0.53
|
0.73
|
-0.33
|
0.97
|
0.03
|
0.80
|
0.24
|
0.04
|
-2.17
|
0.52
|
-0.64
|
NUCOG: Neuropsychiatry Unit COGNITIVE ASSESSMENT TOOL; CTT: Color TRAIL TEST; AVLT:
Auditory VERBAL LEARNING TEST; LVEF: left ventricular ejection fraction; E: early
mitral valve flow velocity; e': diastolic lengthening velocity; LA: left atrial.
DISCUSSION
In this study, we investigated the relationship between echocardiography and coronary
artery angiography parameters; and cognitive function, using a neuropsychiatric screening
tool and precise neuropsychological tests.
We found that the ejection fraction, as the index of cardiac systolic function, was
significantly correlated with global cognition, based on the total NUCOG scores (p
= 0.005). Considering distinct cognitive domains of the NUCOG, the ejection fraction
was correlated with visuospatial (NUCOG subscale B) (p = 0.03) and executive function
(NUCOG subscale D) (p = 0.01). It was also associated with processing speed/attention
(p = 0.02) in the CTT-part 1 and divided attention and executive function in the CTT-part
2 (p = 0.01).
In a prospective study on 44 elderly outpatients with documented heart failure, scores
less than 26 on the Montreal Cognitive Assessment were detected in more than 70% of
the patients[18]. In a systematic review, Vogels et al.[19] compared the risk of cognitive impairment in a pooled sample of 2,937 heart failure
patients with 14,848 controls. They found that the risk of cognitive impairment was
1.62 greater in the heart failure group. They also reported that diminished performance
of psychomotor speed/attention (assessed using the Trail Making Test-A) was correlated
with heart failure[19]. Athilingam et al. reported that the mean Montreal Cognitive Assessment score was
lower in patients with systolic heart failure than diastolic failure, especially in
the visuospatial ability, psychomotor speed, and executive function domains[20]. In contrast to these reports, Park et al. showed that there was no association
between “peak velocities during systole” and cognitive function; however, left ventricular
diastolic function had strong association with global cognition, which was assessed
with “a community screening instrument for dementia”[21]. Visual memory and processing speed had also a significant association with left
ventricular global function. They reported that left ventricular diastolic function
correlated with working memory and fluency of speech[21]. Although, we could not find the same results for diastolic dysfunction in the total
sample of our study, there was significant relationship between e' and immediate free
recall of verbal learning (p = 0.04) in patients with a total NUCOG score of 86 or
more.
Although the exact mechanisms explaining the link between systolic dysfunction and
cognitive decline remain unknown, improvement of cardiac output by cardiac transplantation
and resynchronization therapy has often resulted in better cognitive functioning[22],[23]. Cerebral hypoperfusion, silent cerebral infarction, impaired cerebrovascular autoregulation,
atrial fibrillation, and endothelial dysfunction have been suggested mechanisms for
decreasing cognitive performance in heart failure[11]. Interestingly, patients in our study had mean ejection fraction scores of 54.94
± 8.36, which could be considered as preserved ejection fraction (or mild reduction:
45 < ejection fraction < 55)[16]. It means that even slight reduction of the ejection fraction may lead to significant
cognitive changes. In a longitudinal nondemented population-based study with echocardiography
assessments at baseline and after five years follow up, van den Hurk et al. concluded
that cognitive decline can be observed in the early stages of left ventricular dysfunction
and heart failure[11]. It seems that from the above-mentioned mechanisms, impaired cerebrovascular autoregulation
and endothelial dysfunction are possibly more affected with even mild reduction of
the ejection fraction[24].
Our patients were selected from referrals to a cardiac catheterization unit. Logically,
it might be hypothesized that the main contributor of ejection fraction reduction
in most of them was ischemic events (as we found regional wall motion abnormality
during echocardiography). We concluded that the ischemia, not the reduced ejection
fraction per se, could negatively affect cognition. We found significant correlation between the
diastolic dysfunction index (e') with the CTT-part 2 in patients with impaired cognition
([Table 3]).
In our study, the Gensini score, as the index for severity of coronary artery disease,
was significantly correlated with processing speed/attention (P = 0.001) of the CTT-part
1 and divided attention and executive function of the CTT-part 2 (P = 0.03). When
participants were divided into two distinct groups of patients with normal cognition
and patients with impaired cognition (based on the NUCOG cutoff point), Gensini scores
were significantly associated with the CTT-part 1.
In a cross-sectional study of postmenopausal women with a history of myocardial infarction,
the prevalence of cognitive decline was double that of their counterparts without
myocardial infarction[25]. In a case-control study on patients who had recently been admitted to cardiac catheterization
facilities, Barekatain et al. showed that in patients with mild cognitive impairment,
a greater degree of coronary stenosis correlated with a greater loss of gray matter
in specific brain regions. However, they could not show this in the cognitively normal
patients[8].
Vidal and colleagues, in their cross-sectional study, showed that lower scores on
the speed of processing and executive function, but not memory, were strongly correlated
with atherosclerotic burden, indirectly estimated by coronary artery calcium load,
measured with CT scans[26].
In our study, the left atrial surface area, which is a significant predictor of cardiovascular
morbidity and mortality outcomes, was associated with processing speed/attention (p
= 0.008) in the CTT-part 1, divided attention/executive function of CTT-part 2 (p
= 0.01), and the TOL-time (p = 0.001). In one study on 108 healthy people without
cardiovascular disease, a larger left atrium was associated with cognitive impairment
based on the Mini-Mental State Examination[27]. In a study by van den Hurk et al., the left atrial volume index was specifically
correlated with a lower speed of information processing[11]. An enlarged left atrium may have a possible risk for thrombogenesis due to blood
stasis. This can result in a propensity for microembolization in cerebral arteries,
which in turn may have adverse effects on cognition[27]. In addition to possible consequences of left atrial enlargement, comorbidities
such as hypertension, left ventricular dysfunction, and coronary artery disease may
also be responsible for cognitive impairment[27].
On the other hand, left atrial enlargement may be an adaptive response to endothelial
dysfunction[28]. Thus, it may indirectly point to one of the possible mechanisms of the brain degenerative
process[28].
Assessment of memory in the patients in the present study showed that the AVLT-5th trial (indicator of immediate recall) (p = 0.03) and the AVLT-free delayed recall
(after 30 minutes) (p = 0.01) had a significant correlation with the ejection fraction.
In one study on 251 outpatients with stable heart failure, cognitive impairment was
found in 58% of the sample. Immediate and delayed verbal memory was the most impaired
cognitive domain[29]. The results of a study by Almeida et al. showed that adults with heart failure
had a worse immediate and long-term memory and psychomotor speed than healthy controls.
Another finding of their study of heart failure patients, was the loss of volume in
the posterior association cortex, which plays an important role in the retrieval of
memory[30].
Limitations
The most important limitation in this study was its small sample size, which does
not permit generalization of the results. The overall cardiovascular risk factors
in patients who were referred for catheterization were high. This might result in
an underestimation of the effects, under powering the extrapolation of the study's
results. This part of our study did not have a follow-up. Thus, the causal relationship
between ischemic heart disease or left ventricular dysfunction and cognitive impairment
could not be concluded. Although the duration of cardiac disease may also exaggerate
the deterioration of cognition and could have been included, we did not have accurate
and reliable data. Future studies based on a prospective cohort design should be performed
to determine whether the cognitive deficits are due to heart disease or not.
In summary, we showed that decreased processing speed/attention and executive function
may correlate with cardiac dysfunction and coronary artery disease. The prognostic
implication of coronary artery disease and diastolic dysfunction in patients with
possible cognitive impairment may be worse concomitantly, than in either alone. The
CTT may be recommended as a simple screening test for cognitive problems in elderly
patients with coronary artery disease or diastolic dysfunction.