Introduction
Postoperative cognitive dysfunction is a major concern of medical community. High risk of postoperative cognitive dysfunction (POCD) in cardiac patients is confirmed. Incidence of this event can reach 70% in these patients [1-3]. POCD impairs compliance with treatment, reduces effectiveness of recovery and rehabilitation and can complicate restoration of working capacity. All these aspects increase costs in health care system [4, 5].
Modern guidelines emphasize the need for a comprehensive interdisciplinary study of postoperative cognitive impairment including neuropsychological testing and neuroimaging [6]. Prevention of postoperative cognitive dysfunction also contains various unresolved issues. Obviously, prevention of cognitive disorders and restoration of postoperative cognitive functions in cardiac patients is very important objective of modern medicine. This issue requires new methodological approaches.
Previous studies established and substantiated the positive effect of physical training on general well-being, cognitive and motor functions in advanced age patients [7, 8]. Dual-task exercise (simultaneous motor and cognitive tasks) is perspective specialized rehabilitation approach for various brain pathologies [7, 9]. This paradigm is perspective, since most everyday actions include both these components. There are various dual-task exercises, for example, combination of two motor tasks (simultaneous training of strength and balance), motor-cognitive dual tasks (cognitive tasks involving memory, various types of attention and executive functions combined with physical exercises, simple movements or maintaining a stable posture) [8, 10]. Cognitive training with dual-task exercises ensures more stable improvement in cognitive or motor function in advanced age patients compared to other types of exercises [7, 11]. Some authors reported a wider activation of brain regions following dual-task exercises, i.e. this approach may be perspective for postoperative cognitive recovery [9, 12]. However, He et al. [13] emphasized unclear clinical effectiveness of dual-task exercises in patients with ischemic brain damage and need for further research.
Paradigms of dual-task exercises significantly vary in different studies, that can explain contradictory data [9, 14, 15]. Therefore, the components of dual-task exercises should consider characteristics and severity of clinical status and form of brain lesion. Dual-task exercises for early postoperative cognitive recovery in cardiac patients must have special requirements. In this case, dual task should involve certain cortical regions, first of all, frontal and parietal ones. These zones are the most vulnerable to perfusion impairment. The components of dual-task exercise should be interesting for the patient during multiple repetitions and have acceptable subjective difficulty. In case of computerized version, the patient’s computer skills should be considered.
POCD in cardiac patients may be considered as a form of vascular cognitive disorder characterized primarily by impairment of executive functions [16]. Therefore, the components of dual-task exercises for cardiac patients should include the tasks for executive control and motor components.
An important objective in analysis of dual-task effectiveness for cognitive rehabilitation is assessment of brain compensatory resources and specific rearrangements of brain activity during simultaneous motor and cognitive exercises. Efficiency of one or both tasks will be reduced if brain resources required for cognitive exercise exceeds the individual’s cognitive reserve. This effect is called as dual-task cost [11]. Changes in brain activity associated with dual-task exercises can be studied using electroencephalography (EEG). G. Borghini et al. [17] have previously proposed the power of EEG potentials in theta range to characterize cognitive processes associated with multitasking.
Considering the above-mentioned data, the purpose of this study was to analyze postoperative dynamics of neurophysiological status in cardiac patients using various dual-task exercises for cognitive rehabilitation.
Material and methods
Patients
A cohort prospective study included 47 patients aged 40 — 75 years. All patients admitted for cardiac surgery to the Research Institute for Complex Issues of Cardiovascular Diseases. Most patients (n = 45) underwent isolated CABG, 5 patients — CABG with carotid endarterectomy (CEE). Study design was developed in accordance with the WMA Declaration of Helsinki (1964) and approved by the Local Ethics Committee. All patients signed an informed consent. Exclusion criteria: life-threatening rhythm disturbances, severe heart failure (NYHA class IV), severe comorbid diseases (chronic obstructive pulmonary disease, malignancies, drug addiction, brain diseases and/or injuries, dementia and depression). In addition to standard preoperative examination, all patients underwent neuropsychological testing and electroencephalography 3—5 days prior to surgery and in 7—10 days after intervention. Immediately before cardiac surgery, all patients were pseudo-randomized into 3 groups: control group, group of postoperative cognitive training I and II. Postoperative examination was impossible in some patients. Thus, the control group included 21 patients after randomization, postoperative cognitive training group I — 11 patients, postoperative cognitive training group II — 15 patients. Clinical and anamnestic characteristics of patients are shown in Table 1.
Surgical interventions
Cardiac surgeries included isolated coronary artery bypass grafting (CABG) and combined procedures (CABG + CEE). The number of patients undergoing isolated CABG and simultaneous CABG + CEE was matched in all study groups. All operations were performed under normothermic cardiopulmonary bypass and combined anesthesia (propofol and sevoflurane). Mean time of cardiopulmonary bypass and aortic cross-clamping was similar in all groups (Table 2). Intraoperative invasive control of hemodynamics and cerebral oxygenation was performed in all cases. There were no disturbances.
Neuropsychological examination
The Montreal Cognitive Assessment Scale (MoCA) was used for initial estimation of cognitive status. Depressive symptoms were assessed using the BDI-II questionnaire, situational and personal anxiety — using the Spielberger-Khanin questionnaire.
Software psychophysiological complex “Status PF” was used for an extended assessment of cognitive status [3]. We analyzed the following cognitive domains: neurodynamics (motor and executive functions), attention (target attention and attention volume) and short-term memory (mechanical symbolic and verbal memory, figurative memory).
EEG
Monopolar EEG (62 standard leads of international system 10–20; bandwidth 0.1–50.0 Hz; sampling rate 1000 Hz) was recorded under calm wakefulness with closed eyes. Neuvo SynAmps2 system (Compumedics, Charlotte, NC, USA) with a modified 64-channel cap consisting Ag/AgCl electrodes (QuikCap; Neurosoft, El Paso, TX, USA) was applied. We have previously described EEG technique [3]. EEG biopotential power within the bandwidth 1–50.0 Hz was calculated automatically in the Scan 4.5 software package (Compumedics, Charlotte, NC, USA). The values were averaged for theta1 (4-6 Hz), theta2 (6—8 Hz), alpha1 (8-10 Hz), alpha2 (10-13 Hz), beta1 (13-20 Hz) and beta2 (20-30 Hz) ranges. We obtained the total values of biopotential power in the left and right hemispheres. The indices of EEG activity were also calculated: mean power ratio for theta and alpha rhythms, theta and beta rhythms in accordance with the equation “theta1 + theta2) / (alpha1 + alpha2)” and “(theta1 + theta2) / (beta1 + beta2)”, respectively. Cognitive training
Recovery of cognitive functions in early postoperative period after cardiac surgery was achieved using two approaches. The first one was postural training using hardware-software stabilographic complex “Stabilan-01-2” (Taganrog, Russia) and simultaneous solving an open-type problem J.P. Guildford “Unusual using an ordinary object”. The second approach was a simple visual-motor task using “Status PF” software psychophysiological complex and solving an open-type problem “Unusual using an ordinary object”. Everyday trainings were initiated after the 4th postoperative day in patients randomized to the cognitive rehabilitation group. Exercises were conducted in the morning at the specialized room. Duration of one exercise ranged from 7 to 20 min, mean number of trainings in both groups was 5.
Statistical analysis
Statistical analysis was performed using STATISTICA 10 software (StatSoft, Inc.). Clinical, anamnestic data and cognitive parameters were characterized by abnormal distribution. Therefore, non-parametric methods with Kruskal-Wallis test were used. EEG data were normalized via logarithm and assessed using ANOVA variance analysis.
Results
In all patients, cardiac surgery was not complicated by major cerebrovascular events, including stroke, myocardial infarction, repeated emergency revascularization, etc.
All patients had similar preoperative neuropsychological parameters (MoCA scale) (Table 1). Preoperative extended neuropsychological testing using Status PF battery revealed no significant differences in cognitive status. Similarly, baseline EEG parameters were comparable.
Table 1. Baseline clinical and anamnestic characteristics of patients in all groups
|
Variable |
Patients without cognitive training (n=21) |
Patients with cognitive training I (n=11) |
Patients with cognitive training II (n=15) |
p-value |
|
Age, years |
64 [52,5; 69] |
61 [59; 68] |
62 [57; 69] |
n/s |
|
Male/female, n |
15/6 |
9/2 |
10/5 |
n/s |
|
Education, n (%) |
n/s |
|||
|
secondary and secondary specialized |
16 (76) |
8 (73) |
10 (67) |
|
|
higher |
5 (24) |
3 (27) |
5 (33) |
|
|
LV EF, % |
59 [52; 63] |
60 [51; 63] |
64 [60; 68] |
n/s |
|
Diabetes mellitus type 2, n (%) |
5 (24) |
2 (18) |
4 (27) |
n/s |
|
Carotid artery stenosis, n (%) |
n/s |
|||
|
no |
12 (57) |
7 (64) |
9 (60) |
|
|
< 50% |
6 (29) |
3 (27) |
5 (33) |
|
|
> 50% |
3 (14) |
1 (9) |
1 (7) |
|
|
Class of angina, n (%) |
n/s |
|||
|
0-I |
3 (14) |
2 (18) |
3 (20 |
|
|
II-III |
18 (86) |
9 (82) |
12 (80) |
|
|
Heart class NYHA class, n (%) |
n/s |
|||
|
I-II |
16 (76) |
8 (73) |
10 (67) |
|
|
III |
5 (24) |
3 (27) |
5 (33) |
|
|
MoCA score |
25 [23; 26] |
26 [24; 26] |
27 [24; 27] |
n/s |
|
BDI-II score |
2 [2; 5] |
2 [1; 6] |
2,5 [2; 5] |
n/s |
Note. Quantitative data are presented as median, 25 and 75 quartiles; n/s — differences are statistically insignificant (p<0.05).
Both variants of the approved cognitive training programs showed an acceptable subjective difficulty of exercises for patients after cardiac surgery. Most of the patients were interested in exercises until the final stage of training, although decrease in interest was more common in the 2nd group.
Next, we analyzed EEG and cognitive changes in early postoperative period after cardiac surgery compared to preoperative values depending on the type of cognitive training.
Analysis of cognitive changes is presented in Table 3. An increase or decrease of cognitive parameter by 10% or more during repeated testing was significant. According to Table 2, any type of cognitive training was followed by positive changes in all cognitive domains: neurodynamics (higher rate of motor reactions, less number of missed signals), attention (increase of attention volume, higher number of processed characters in Bourdon test on the 1st (workability) and the 4th minute (less exhaustion)). Memory for numbers and syllables was improved.
Table 2. Intraoperative characteristics of patients in all groups
|
Variable |
Patients without cognitive training (n=21) |
Patients with cognitive training I (n=11) |
Patients with cognitive training II (n=15) |
p-value |
|
Surgery, n (%): |
n/s |
|||
|
Isolated CABG |
19 (90) |
10 (90) |
13 (87) |
|
|
Simultaneous CABG and CEE |
2 (10) |
1 (10) |
2 (13) |
|
|
CPB time, min |
83.5±32.2 |
85.7±32.3 |
86.7±30.3 |
n/s |
|
Aortic cross-clamping time, min |
57.8±15.1 |
58.3±15.6 |
53.7±19.9 |
n/s |
|
ICA occlusion time, min |
22.1±3.7 |
22.4±3.5 |
22.6±3.7 |
n/s |
|
CPB temperature, °C |
35.5±0.2 |
35.7±0.3 |
35.7±0.7 |
n/s |
|
Revascularization index |
2.5±0.7 |
2.5±0.4 |
2.6±0.5 |
n/s |
Table 3. Postoperative changes in cognitive parameters in all groups
|
Variable |
Patients without cognitive training (n=21) |
Patients with cognitive training I (n=11) |
Patients with cognitive training II (n=15) |
|
Visual-motor reaction time, sec |
–12% |
–17% |
–21% |
|
Functional mobility of nerve processes: the number of errors, n |
+11% |
ns |
+12% |
|
Functional mobility of nerve processes: the number of missed signals, n |
ns |
+24% |
–14% |
|
Brain performance: missed signals, n |
ns |
–16% |
–22% |
|
Attention volume, score |
+17% |
ns |
+12% |
|
Bourdon test: the number of processed symbols, n |
|||
|
within the 1st minute |
+19% |
+10% |
+27% |
|
within the 4th minute |
ns |
+11% |
–11% |
|
"Memorizing 10 numbers" test, n |
ns |
+21% |
ns |
|
"Memorizing 10 syllables" test, n |
ns |
+16% |
+13% |
|
"Memorizing 10 words" test, n |
ns |
ns |
ns |
Note. ns — no significant dynamics of the index (difference <10%).
We can emphasize improvement in more indicators related to executive functions and attention (6 parameters) in the training group II compared to the training group I (5 parameters). Improvement of short-term memory and attention was observed in the training group I.
At the next stage, we analyzed the power of EEG activity in each frequency range. The factors GROUP (control group, postoperative cognitive training I and II) x TIME OF SURVEY (before, after surgery) x LATERALITY (left, right hemisphere) were incorporated in variance analysis (ANOVA). We found higher power of theta1 rhythm in postoperative period after cardiac surgery is observed in all patients. Significant factor is TIME OF SURVEY (F1.45 = 6.33; p = 0.016). This factor was also significant in beta1 range (F1.45 = 5.57; p = 0.023). In all patients, higher power of biopotentials was observed in postoperative period compared to preoperative data (Fig. 1).
Fig. 1. EEG changes in all patients undergoing cardiac surgery with and without dual-task cognitive training.
Gray columns — preoperative parameters, black columns — postoperative parameters; * — significant differences between postoperative and preoperative data.
We also analyzed EEG data (theta/alpha and theta/beta) using ANOVA with introduction of the factors GROUP (control group, groups of postoperative cognitive training I and II) x TIME OF SURVEY (before, after surgery). In all patients, we found higher postoperative theta/alpha activity index compared to preoperative values (factor TIME OF SURVEY, F1.45 = 5.43; p = 0.02). At the same time, theta/beta activity index decreased in postoperative period compared to preoperative value (Fig. 2).
Fig. 2. EEG changes in all patients undergoing cardiac surgery with and without dual-task cognitive training.
◊, * — significant differences between postoperative and preoperative data.
Discussion
Our pilot study revealed positive changes in cognitive domains of memory and attention after dual-task exercise type I while exercise type II ensured improvement of all domains (neurodynamics, attention and memory). We can also note improvement of more parameters related to executive functions and attention (n=6) after dual-task exercise type II compared to exercise type I (n=5). In the control group, we observed only a slight improvement of postoperative attention.
Some authors have previously reported combination of aerobic training and tasks for attention and executive functions as a successful variant of dual-task exercise [18, 19]. The authors emphasize a positive effect of these elements on cognitive abilities and recommend careful increase of difficulty of the tasks depending on individual abilities of the patient. However, this is true for cognitively and clinically intact persons. In people with cognitive impairment, dual-task cognitive training should be longer to achieve positive effect compared to healthy people [18]. Patients after cardiac surgery are limited in their physical capabilities due to certain factors accompanying early postoperative period (pain syndrome, asthenia and surgical complications). In this regard, the choice of motor task is not so wide. We opted for postural training and simple visual-motor task to maximize motor load tolerance in this cohort of patients. Dual tasks consisting postural training have been previously used to restore cognitive function after traumatic brain injury [20]. However, we found no significant between-group differences in effectiveness of cognitive functions among patients with and without cognitive training in early postoperative period. Therefore, insufficient intensity of motor load may be assumed.
Insignificant clinical effect of dual-task exercises may be also explained by significant interactions between cognitive and motor components. Deterioration of cognitive component was demonstrated in older people undergoing dual-task exercise combining complex cognitive tasks with any motor tasks. Moreover, older patients worse perform motor tasks combined with complex cognitive tasks [12]. In our study, open-type cognitive exercise can be attributed to complex cognitive tasks. Therefore, patients could not effectively perform the proposed training option in early postoperative period due to limited cognitive resources.
Nevertheless, dual-task cognitive training can be effective for improving cognitive function in cardiac patients. Indeed, this approach ensures better transfer effect (improved performance in other cognitive tasks in post-workout period compared to baseline level) relative to isolated cognitive training or motor exercise [21-23]. Some authors previously found transfer effect change depending on complexity and modality of the task [22]. According to some data, effectiveness of dual tasks in cognitive recovery is ensured via more efficient coordination of cognitive processes.
An important fact of our study is comparable severity of perioperative ischemic brain damage according to EEG data in all groups (increase of the power of the theta1 and beta1 rhythm, negative changes of the theta/alpha and theta/beta indexes). Nevertheless, improvement of only short-term memory and attention was observed in the training group I, while in the training group II — improvement of executive functions and attention. Of course, it is a success of training.
Conclusion
In early postoperative period (8-11 days), all patients had signs of perioperative ischemic brain damage according to EEG data (increase of the power of the theta1 and beta1 rhythm, negative changes of the theta/alpha and theta/beta indexes). Dual-task cognitive training improves the indicators of all cognitive domains (neurodynamics, attention and short-term memory). Further improvement of the approaches to postoperative dual-task cognitive training with load intensification and individual support for patients undergoing cardiac surgery is required.
Study limitations
Small sample size, possible low intensity of dual tasks.
Financing
The study was supported by the Russian Foundation for Basic Research and the Kemerovo Region, project No. 20-415-420005.
The authors declare no conflicts of interest.