Description of Post-Stroke Cognitive Disorders at Abidjan University Hospitals, Côte d’Ivoire ()
1. Introduction
Stroke is a major public health problem worldwide. According to [1], stroke is the most common neurological disorder in developed countries, where it is the leading cause of acquired non-traumatic physical disability, the second leading cause of dementia after Alzheimer’s disease, and the third leading cause of death after myocardial infarction and cancer. A major cause of depression among patients and their families is the cost of care, estimated at 2.4% of total healthcare expenditure worldwide. Post-stroke cognitive dysfunction is the second most common cause of severe neurocognitive impairment after Alzheimer’s disease (AD) [2]. Despite the significant consequences, such as poorer prognosis in terms of vitality and function, cognitive deficits after stroke are poorly understood. The consequences of these disorders lead to a significant risk of disability and subsequent institutionalization in most developed countries [3] [4]. Researchers have shown that cognitive symptoms after stroke are frequent in the acute phase (around 80% of surviving patients), and persist in 38% to 73% of cases [5] [6]. However, this symptomatology is often neglected compared to other possible deficits (motor and/or sensory) [7]. In the least developed countries, the impact of cognitive disorders is still poorly assessed because they are not a priority in national health programs [8]. In Côte d’Ivoire, as in less developed countries, data on post-stroke cognitive disorders are still insufficient. Our study is designed to contribute to a better understanding of cognitive disorders after stroke. In the present study, we determine the frequency and the risk factors related to the cognitive status of post-stroke patients treated in the neurology department of the University Hospitals of Cocody and Treichville (Côte d’Ivoire).
2. Method
2.1. Setting of the Study
This study describes the cognitive status of post-stroke patients seen in the neurology departments of the university hospital centers of Cocody and Treichville (Côte d’Ivoire). It is a cross-sectional and analytical study that lasted 12 months (from September 2020 to September 2021), excluding the month of June.
2.2. Method Description
We screened different cognitive functions (global cognitive functioning, executive functions, language and memory) among stroke patients. We determined the frequencies of patients with impaired functions and non-impaired functions. We first of all determined the frequencies of patients with global cognitive functioning impairment and patients without global cognitive functioning impairment by demographic characteristics (age group, gender, educational level, employment status) and by clinical characteristics (vascular diseases, and indexed stroke). For age group, we categorize patients into 2 groups: young (inferior to 65 years) and old (superior to 65 years). About gender, we categorize patients into 2 groups: men and women. Regarding educational level, patients were divided into 3 groups: low level (less than 7 years of education); medium level (less than 14 years of education); high level (at least 14 years of education). For employment status, patients were divided into 3 groups (public, private and liberal employment). Clinical characteristics were concerned with hypertension, diabetes, and cardiopathy for vascular diseases. The injured hemisphere and affected area were taken into account for indexed stroke. We then determined related stroke risk factors according to demographic and clinical characteristics of stroke patients by significant p-value.
2.3. Inclusion Criteria
Subjects included in the study were:
Patients who gave written informed consent;
Patients in the subacute phase of stroke (period of 5 to 15 days after stroke) are documented by computed tomography (CT) or brain magnetic resonance imaging (MRI);
Patients without pre-stroke cognitive impairment (IQ code ≤ 3.4);
Patients with sufficient visual, auditory, and verbal or written expression skills to perform neuropsychological tests adequately.
2.4. Technical Equipment
The cognitive status of the patients was assessed using the following psychometric tests:
Informant Questionnaire on Cognitive Decline in the Elderly (IQ-Code) [9];
Montreal Cognitive Assessment (MoCA), French version 7.1 [10];
Frontal Assessment Battery (FAB) [11];
French version of the Language Screening Test (LAST) [12];
Digit Span Test (forward/backward) [13].
2.5. Statistical Analysis
Statistical tests were performed using R. Associations between variables were analyzed using the Chi square test. Binary logistic regression was used to identify possible risk factors for cognitive impairment after stroke. Graphs were presented using Excel 2019.
3. Results
During the period of the study, 724 patients were hospitalized. Among them, 309 patients (42.68%) were admitted for diverse neurological pathologies and 415 stroke patients (57.32%) were enrolled in the study. The latter’s were selected based on selection criteria. Eventually, 145 patients (34.94%) aged 30 to 87 years were evaluated as shown by the flowchart (Figure 1).
3.1. Frequencies of Different Cognitive Functions
The frequency of patients with global cognitive functioning impairments was significantly higher (86.21%) than that of patients without impairment (13.79%). Global cognitive functioning is highly impaired after stroke. The frequency of
Figure 1. Flowchart of the studied population.
patients with executive function problems was 81.38% against 18.62% without the problem. Executive functions are highly impaired after stroke. The frequency of patients with impaired language was 54.48%. However, 45.52% of the stroke patients were not impaired. Language is moderately affected after a stroke. The frequency of short-term memory disorders was 34.48% and that of the working memory disorders was 24.83%. On the contrary, the frequency of patients without short-term memory disorders was 65.52% and the patients without working memory disorders were 75.17%. Our figures suggest that short-term memory disorders and working memory disorders are not frequent after a stroke.
3.2. Overall Cognitive Functioning Status of Patients According to Socio-Demographic Characteristics
3.2.1. Patients’ Overall Cognitive Functioning by Age Group
As shown in Figure 2, the frequency of elderly patients with global cognitive functioning impairment was 95%, which was higher than the frequency of young patients, which was 82.86%.
Figure 2. Distribution of patients’ overall cognitive functioning by age group. The frequency of patients with global cognitive functioning impairment was higher in older patients (95%) than in younger patients (82.86%).
3.2.2. Overall Cognitive Functioning of Patients by Gender
Figure 3 shows that the frequency of male patients with impaired global cognitive functioning, 90.48%, is higher than the frequency of female patients, 82.93%.
3.2.3. Overall Cognitive Functioning of Patients by Level of Education
Figure 4 shows that after stroke, 93.65% of patients with a low level of education had problems with global cognitive functioning, compared with 85.48% of patients with a medium level of education and 65% of patients with a high level of education. The frequency of patients with a high level of education without global cognitive functioning impairment was higher (35%) than that of patients with a medium or low level of education (14.52% and 6.35% respectively).
3.2.4. Patients’ Overall Cognitive Functioning According to Working Status
Figure 5 shows that the frequency of patients in free practice with global cognitive functioning impairment after stroke (89%) was significantly higher than that of patients in private and public practice (81.48% and 77.78%) respectively. The frequency of patients in public practice without global cognitive functioning impairment was significantly higher (22.22%) than that of patients in private and free practice (18.52% and 11%) respectively.
3.3. Overall Cognitive Functioning Status of Patients According to Clinical Characteristics
The frequency of patients with hypertension and GCFI was higher (88.63%) than those without hypertension (82.45%). The frequency of patients with diabetes and GCFI was higher (90%) than those without diabetes (85.6%). The frequency of patients with cardiopathy and GCFI was higher (90%) than those without cardiopathy (85.92%).
Figure 3. Distribution of patients’ global cognitive functioning by gender. The frequency of patients with global cognitive functioning impairment was higher in men (90.48%) than in women (82.93%).
Figure 4. Distribution of patients’ overall cognitive functioning by level of education. The frequency of patients with problems in overall cognitive functioning was higher in patients with a low level of education than in patients with a medium level of education and a high level of education.
3.4. Related Risk Factors
3.4.1. Sociodemographic Risk Factors
The p-values for age group, gender and employment status were not significant. Age, sex and employment status have no effect on the onset of global cognitive impairment. The p-value is significant for educational level. The level of education therefore had an effect on the onset of global cognitive impairment (Table 1).
3.4.2. Vascular Diseases Risk Factors
The p-values are not significant. Hypertension, diabetes and cardiopathy had no effect on the occurrence of cognitive functioning impairment (Table 2).
Figure 5. Distribution of global cognitive functioning according to working status. The frequency of patients with impaired global cognitive functioning was higher in patients with liberal status than in patients with private and public status.
Table 1. Sociodemographic risk factors.
Variables |
GCFI |
OR |
Chi 2 |
P value |
Yes (n = 125) |
No (n = 20) |
Age group |
Old |
38 |
2 |
3.93 [0.86 - 17.79] |
3.59186 |
0.058066 |
Young |
87 |
18 |
Gender |
Women |
68 |
14 |
1.95 [0.71 - 5.42] |
1.70767 |
0.191292 |
Men |
57 |
6 |
Educational level |
Low |
59 |
4 |
- |
10.5276 |
0.005177** |
Medium |
53 |
9 |
- |
High |
13 |
7 |
- |
Employment status |
Public |
14 |
4 |
- |
2.23869 |
0.326497 |
Private |
22 |
5 |
- |
Liberal |
89 |
11 |
- |
(5% significance level). *significant, **very significant. The p-values for age group, gender and employment status were not significant. Age, gender and employment status have no effect on the onset of global cognitive impairment. The p-value is significant for educational level. The level of education therefore has an effect on the onset of global cognitive functioning impairment. GCFI: global cognitive functioning impairment.
3.4.3. Risk Factors of Indexed Stroke
1) Hemorrhagic stroke
The odds ratio shows that patients injured on the left hemisphere are likely to present GCFI 4 times in comparison to patients injured on the right hemisphere. But the p-value is not significant. Either left or right hemisphere injury had no effect on the occurrence of global cognitive functioning impairment (Table 3).
The odds ratio shows that patients injured on the left hemisphere are 4.16 times
Table 2. Vascular diseases risk factors.
Variables |
GCFI |
OR |
Chi 2 |
P value |
Yes (n = 125) |
No (n = 20) |
Hypertension |
Without hypertension |
47 |
10 |
- |
- |
- |
With hypertension |
78 |
10 |
1.92 [0.70 - 5.30] |
|
0.20039 |
Diabetes |
Without diabetes |
107 |
18 |
- |
- |
- |
With diabetes |
18 |
2 |
1.52 [0.35 - 10.64] |
|
0.61108 |
Cardiopathy |
Without cardiopathy |
116 |
19 |
- |
- |
- |
With cardiopathy |
9 |
1 |
1.61 [0.26 - 31.20] |
|
0.66021 |
(5% significance level). The p-values are not significant. Hypertension, diabetes and heart disease had no effect on the occurrence of global cognitive functioning impairment.
Table 3. Risk factors of indexed stroke (hemorrhagic stroke).
Variables |
GCFI |
OR |
Chi2 |
P value |
Yes (n = 55) |
No (n = 10) |
Injured Hemisphere |
Right |
34 |
4 |
- |
- |
- |
Left |
21 |
6 |
0.24 [0.04 - 1.14] |
|
0.088937 |
Affected area |
Lobar |
10 |
6 |
- |
|
- |
Deep |
45 |
4 |
9.61 [2.07 - 56.45] |
|
0.005731** |
(5% significance level). *significant, **very significant. The p-value is very significant for the deep area. The deep area lesion is associated with the occurrence of global cognitive functioning impairment.
less likely to have global cognitive functioning impairment than patients injured on the right hemisphere. Moreover, the p-value is very significant (0.005731). The deep area lesion is associated with the occurrence of post-stroke cognitive functioning impairment. Patients with a deep lesion are 9.61 times more likely to have global cognitive functioning impairment than those with a lobar lesion (Table 3).
2) Ischemic stroke
The odds ratio shows that patients injured on the left hemisphere are 1.34 times more likely to have global cognitive functioning impairment than patients injured on the right hemisphere. However, the p-value is not significant. Either left or right hemisphere injury had no effect on the occurrence of global cognitive functioning impairment (Table 4).
The odds ratio value shows that patients with a deep lesion are 1.87 times more likely to have impaired global cognitive functioning than patients with a lesion in the MCA area. Patients with a lesion in the PCA area are 57.1 times less likely to have impaired global cognitive functioning than patients with a lesion in the MCA area. In addition, the p-value is significant (0.0480493). The lesion located in the posterior cerebral artery area is associated to the occurrence of global cognitive functioning impairment (Table 4).
Table 4. Risk factors of indexed stroke (ischemic stroke).
Variables |
GCFI |
OR |
Chi2 |
P value |
Yes (n = 70) |
No (n = 10) |
Injured Hemisphere |
Right |
37 |
6 |
- |
- |
- |
Left |
33 |
4 |
1.34 [0.32 - 6] |
|
0.6850554 |
Affected area |
MCA |
46 |
6 |
- |
- |
- |
ACA |
6 |
0 |
5.51 × 106 [1.46 × 10−51] |
|
0.9923178 |
PCA |
4 |
3 |
1.75 × 10−1 [3 × 10−1 - 1.06] |
|
0.0480493* |
Deep |
14 |
1 |
1.87 [2.8 × 10−1 - 37] |
|
0.5776079 |
(5% significance level). *significant, **very significant. MCA: middle cerebral artery, ACA: anterior cerebral artery, PCA: posterior cerebral artery. The p-value is significant for the PCA. The lesion located in posterior cerebral artery area is associated to the occurrence of global cognitive functioning impairment.
4. Discussion
In our study, 81.38% of patients had problems with executive function and 54.48% had problems with language. Vidović et al. found that language deficits after stroke are common, affecting up to 80% of patients, and may regress with appropriate rehabilitation [14]. Our results showed a frequency of 34.48% and 24.83% for disorders of short-term memory and working memory, respectively. According to [15], about 30% develop memory problems in the year after stroke. Our study found 86.21% cognitive impairment. Other studies have also found around 80% impairment after stroke [6]. Burton and al., estimated that 80% would be impaired in the acute phase [16]. In foreign developed countries, a cross-sectional Chinese study conducted by [17] in Changsha (China) had a frequency of 41, 8%. A cross-sectional study conducted in India by [18] revealed a frequency of 66, 66%. Another Chinese study on post-stroke cognitive impairment (PSCI), revealed an overall prevalence of 80.97% [19]. A study conducted in Indonesia by Pinzon et al revealed a frequency of 68.2% [20]. According to Jokinen and collaborators, 83% of patients showed impairment in at least one cognitive domain [21]. In comparison, the frequencies of PSCI in reported studies in various countries move away from each other’s. The frequency observed in our study remains higher, but it is not so far from those of studies conducted in African developing countries. Studies similar to ours have been conducted in Ghana by [22], in Nigeria by [23], Uganda by [24]. These authors found frequencies of 72.8%, 67.4%, and 66.4%. The higher frequency in our study may be due to when the patients were assessed. In our study, patients were assessed in the subacute stage (at least 5 days after stroke), while in the studies mentioned, patients were assessed in the post-acute phase (at least 3 months after stroke). Timing of cognitive assessment is crucial for accurate diagnosis, as retested scores often show improvements [25] [26]. Some studies reported that cognitive problems continue to stabilise and even improve beyond the first 3 months after stroke [27]. Moreover, we performed assessment with the MoCA test, while authors of studies named above used the MMSE. Incidentally, Zhao and colleagues have shown that the MoCA is more sensitive to mild cognitive impairment than the MMSE [28].
Although our study did not find a significant effect of age on the risk of cognitive impairment after stroke (Table 1), other studies show that age influences cognitive decline after stroke. Age is a risk factor for both stroke and cognitive decline. There is evidence that cognitive decline after stroke increases with age after 65 [29]. Our study did not find that gender had a significant effect on the onset of cognitive impairment (Table 1). Others researchers showed that gender was not identified as a factor associated with the onset of post-stroke cognitive impairment [30] [31]. In some studies, gender is linked to post-stroke cognitive issues [32]. Our research reveals education impacts cognitive function. High education levels protect against cognitive decline after a stroke [22] [33]. Indeed, a higher level of education increases the brain’s cognitive reserve, which may lead to better compensation for ageing and brain damage [34]. Some authors found education level doesn’t impact cognitive decline. Singh-Manoux, Zahodne & colleagues studied education’s effect on cognition in 7454 Whitehall II & 1023 Victoria Longitudinal Study participants [35] [36]. Both studies suggested that there was no significant difference of the rate of decline in cognitive function between each group of the education level. The education level is a conflictive risk factor. It could influence the expression of the cognitive impairment in patients. The cohort study conducted by Elbaz and collaborators, on 4010 participants suggested the higher education was associated with better cognitive performances [37]. Our study did not show that employment status was related to cognitive impairment (Table 1). Yet, there is evidence suggesting that the occupation have effects on the prevalence of the cognitive impairment. Singh-Manoux and collaborators also suggested that the individuals with high occupations that were defined as the administrative positions had a more obvious cognitive decline than other occupations [35]. Another study conducted by Douiri and collaborators, proposed a higher prevalence of cognitive impairment after ischemic stroke in the manual workers [38]. Our study did not show that vascular diseases risk factors i.e. hypertension, diabetes, cardiopathy were related to post-stroke cognitive functioning impairment (Table 2). On the contrary, a study has shown the effect of those factors on post-stroke cognitive impairment [39]. Table 3 showed that deep affected area is related to post-stroke cognitive impairment in hemorrhagic stroke (p = 0.005). Table 4 showed that affected area of posterior cerebral artery is related to post-stroke cognitive functioning impairment in ischemic stroke (p = 0.048). About it, Munsch and collaborators suggested that the onset of post-stroke cognitive impairment could be related to the location of cerebral infarction [40].
5. Limitations of the Study
Cross-sectional studies cannot establish causality. Timing of patient assessment (subacute phase) may overestimate cognitive problems (86.21% of survivors). Restricting the study to public health facilities may introduce bias due to the social status of the population. The study may be biased by the exclusion of a significant number of participants from the study consortium, particularly for illiteracy (n = 82, 19.76%) and aphasia (n = 35, 8.43%). In addition, the use of the MoCA test, which has not been validated in sub-Saharan African populations, could lead to misinterpretation of cognitive impairment in participants from these countries, potentially overestimating the impact. Despite these limitations, the study provides valuable data on cognitive impairment after stroke in sub-Saharan Africa. This can inform public awareness, policy recommendations and future research into effective interventions.
6. Conclusion
This study provides data on frequencies and related risk factors of cognitive status among patients in subacute stage of stroke in Abidjan. Global cognitive functioning and executive functions were highly impaired, language was moderately impaired while memory was less impaired. The level of education and the affected brain area were risk factors associated with the occurrence of cognitive functioning disorders. The interest of this study is the importance of early screening of cognitive disorders after a stroke for many cognitive functions have been found to be impaired. The cognitive care of stroke patients should benefit from these findings. We plan to integrate further aspects such as the patient’s lifestyle in the research of post-stroke cognitive disorders in order to find more associated factors.
Acknowledgements
We especially and attentively thank the family members of the Biology and Health Laboratory, Felix Houphouët-Boigny University and the medical staff of the University Hospital who contributed significantly to the completion of this project.
Consent to Participate
Informed consent was obtained from the patient.
Authors’ Contribution
NNA and ON contributed to writing this manuscript; ON and KKL contributed to revising the manuscript. TNA and KKL contributed to the study design. NNA performed all the data analyses.