The Cognitive Profile of Elderly Patients With Mild... : The Journal of Head Trauma Rehabilitation (2024)

TRAUMATIC BRAIN INJURY (TBI) is a worldwide problem and has an estimated annual incidence of hospitalized patients ranging from 100 to 300 per 100 000,¹ with the majority (80%-90%) comprising a mild traumatic brain injury (mTBI). Given the rapidly aging population, the number of the elderly sustaining mTBI is steadily increasing.² mTBI can initially cause cognitive impairments, which can be objectified by neuropsychological tests in the first few weeks to months after the injury,^3,4 and which usually recover over time.^5,6 Cognitive impairments can be distinguished from cognitive complaints. Cognitive complaints refer to the subjectively experienced symptoms mentioned by patients. These are frequent in the early stage after injury⁷ and can persist in the chronic phase post-injury in a subgroup of patients (15%-20%).⁸ Despite the increase in incidence of mTBI among elderly patients, only a few studies to date have investigated cognition specifically in this patient group alone. Differences in injury mechanism, with falls as most prevalent in the elderly leading to more focal damage, and vulnerability to secondary damage due to age-related changes in the brain, emphasize the need to investigate the elderly category separately from the overall population of patients with mTBI.⁹

As every life can be complex, intact cognitive functions are essential for mTBI patients to reintegrate in the community and participate in social and productive activities. In the overall population of patients with mTBI, subtle cognitive deficits in information processing speed, attention, and memory are common in the acute phase but the majority of patients are expected to recover at 3 months post-injury.^3–5,10 On a group level, no differences in cognitive functions are found between patients with mTBI and healthy controls (HCs) 1 year after injury.^5,6 Yet, on an individual level there may be variability in cognitive performance. Studies on cognitive functions in elderly patients with mTBI are scarce but show a consistent profile of subtle cognitive deficits at 3 months post-injury,^11,12 possibly affecting daily life activities and therefore community reintegration. These studies focused only on prospective memory. To our knowledge, no studies so far investigated a more extensive cognitive profile and its association with community reintegration of the elderly who sustained an mTBI.

It is hypothesized that cognitive reserve may be a modifying factor moderating the relationship between the brain damage caused by the TBI and cognitive test performance.^13,14 The concept of cognitive reserve can be defined as a compilation of intellectual activities that shape network efficiency, processing capacity, and flexibility of the brain.^14,15 When cognitive reserve is high, individuals are thought to be better able to overcome the consequences of brain-related diseases. Several studies have investigated the impact of cognitive reserve on cognitive functioning following acute brain injury, measured by different proxy variables such as premorbid intelligence quotient (IQ), education level, occupation, or a questionnaire.¹⁶ Although research suggests that lower cognitive reserve is a risk factor for the development of cognitive impairment after TBI,^10,17 few studies have used a standardized, multidimensional measure to examine the effects of cognitive reserve.

Besides subtle cognitive impairments, patients with mTBI can also report cognitive complaints such as forgetfulness, slowness, and lack of concentration. These complaints persist in a subgroup of patients, but are not always interfering with daily life functioning, as the majority of patients with complaints resume their daily activities to preinjury levels.¹⁸ Prior studies have suggested that, in the overall group of mTBI patients, these complaints are not associated with cognitive impairment, but that they can be associated with mental distress factors such as anxiety and depression.^7,19 However, it has not been investigated yet to which extent this is the case in the elderly mTBI population.

The primary aim of this study was to assess cognitive impairment in elderly patients with mTBI in the subacute phase after trauma, considering the possible moderating role of cognitive reserve. In addition, we aimed to investigate the presence of cognitive complaints in this patient group and to determine whether these complaints are associated with the both the presence of cognitive impairments and the indices of mental distress (anxiety and depression). Insight into the cognitive sequelae of mTBI in older adults might allow early detection of specific targets for treatment.

METHODS

Participants and procedures

Patients were assessed in the subacute phase, 2 weeks to 6 months after trauma. They were part of a larger prospective observational cohort study (the RECONNECT study²⁰). A consecutive cohort was included from 2018 until 2021 at the University Medical Center Groningen (a level 1 trauma center). Eligibility criteria were 60 years or older, having sustained a mTBI with diagnosis based on Glasgow Coma Scale (GCS) scores of 13 to 15 with loss of consciousness and/or posttraumatic amnesia. Exclusion criteria were insufficient comprehension of the Dutch language, a history of brain injury or psychiatric disorder (for which the patient was admitted in the past), substance use disorders, inability to comply with follow-up, and suspected neurodegenerative disease, to allow investigation of the specific effects of mTBI. Eligible patients were examined with a short neuropsychological assessment. Comorbidities (such as diabetes and cardiovascular disease) and presence of preinjury mental health issues were documented. Cognitive test results were reviewed by a licensed clinical neuropsychologist to detect patients with suspected dementia, which were excluded. HCs were recruited via advertisem*nts. Exclusion criteria comprised: a previous admission for a TBI, a psychiatric, neurological or substance use disorder, or suspected neurodegenerative disease. Patients with mTBI and HCs were group-matched for sex and education level. The study was approved by the Medical Ethical Committee of the University Medical Center Groningen. All participants gave written informed consent, prior to participation.

Measurement instruments

Neuropsychological assessment

A short neuropsychological test battery of 1 hour was used to assess cognitive functions.

Memory

The Digit Span, subtest of the Wechsler Adult Intelligence Scale (WAIS-III), is a measure of working memory.²¹ Patients have to repeat several digit sequences both forward and backward. Scores are determined by the total amount of correctly repeated series (maximum 30). Percentile scores are corrected for the influence of age. The 15-word test (15-WT) is the Dutch version of the Rey Auditory Verbal Learning Test measuring verbal episodic memory.²² Fifteen spoken words are presented and patients are asked to repeat as many words as possible immediately afterward, in 5 trials. For this immediate recall (IR) the maximum score is 75. After 15 to 20 minutes, the patient is asked again to reproduce the words to measure retention—delayed recall (DR) (maximum score of 15). Percentile scores of the 15-WT IR are corrected for the influence of sex, age, and education level. Percentile scores of the 15-WT DR are corrected for the total score in IR. The word naming test (WNT), is a subtest of the Groninger Intelligence Test²³ measuring semantic memory fluency. Patients produce as many words as possible during 1 minute, belonging to a well-known category (1 = animals; 2 = professions). The total score is the sum correct words produced in each of the 2 categories. Percentile scores are corrected for the influence of sex, age, and education level.

Attention, processing speed, and executive control

The Trail Making Test part A (TMT-A) measures psychom*otor speed.²⁴ The Trail Making Test part B (TMT-B) measures switching attention.²⁴ The score is the amount of time it takes (in seconds) to finish the trail. The Trail Making Test index A/B (TMT-A/B) measures divided attention and cognitive flexibility, corrected for speed. Percentile scores are corrected for the influence of sex, age, and education level. The Controlled Oral Word Association Test, or Letter Fluency Test, measures executive control.²⁵ Patients have to produce words in 1 minute, which start with a certain letter. After 3 trials (initial letters D-A-T), the sum of correctly produced words is the total score. Percentile scores are corrected for the influence of sex, age, and education level.

Questionnaires

The Head Injury Symptom Checklist²⁶ questionnaire contains the most commonly described posttraumatic complaints. For this study, we only used 3 items that focused specifically on cognitive complaints: forgetfulness, concentration and slowness. Postinjury levels of complaints were established with values ranging from 0 = never, 1 = seldom, to 2 = often. Postinjury levels were also dichotomized to report absence (0) or presence (1) of complaint(s).

The Hospital Anxiety and Depression Scale (HADS)²⁷ measures anxiety (HADS-A) and depressive (HADS-D) symptoms using 7 items each resulting in a subscale score with a maximum of 21. A score of 8 or more on a subscale indicates the presence of anxiety or depression.

The Cognitive Reserve Index questionnaire (CRIq)²⁸ is a questionnaire providing a standardized measure of the cognitive reserve. The CRI scores were categorized using the cut-off scores as recommended by the authors: low (≤70), low-moderate (70-84), moderate (85-114), moderate-high (115-130), and high (≥130).

The Community Integration Questionnaire (CIQ)²⁹ measures community integration using 15 items assessing the following daily life activities: participation at home (scores ranging from 0 to 10), social participation (scores ranging from 0 to 12), and productive activities (scores ranging from 0 to 7), the sum resulting in a total score (range 0-29), with higher scores indicating greater overall community reintegration.

The Glasgow Outcome Scale Extended (GOSE)³⁰ measures functional outcome after TBI on a scale ranging from 1 (death) to 8 (complete recovery).

Statistical analysis

To summarize participant characteristics, descriptive statistics were calculated. Differences between groups were analyzed using t tests, χ² tests, Mann-Whitney U tests, and Fisher's exact test. The Shapiro-Wilk test showed no normal distribution of cognitive complaints and scores on the HADS. Norm scores (percentile scores) of the neuropsychological tests (15WT, WNT, letter fluency, trial making task, and Digit Span task test) were used for analysis to control for factors such as sex, age, and/or education level. Neuropsychological test scores of patients were compared to a mean of 50 (which is the expected mean in a healthy population). A percentile score of less than 10 or a t-score of less than 37 on a neuropsychological test was labeled as impaired. An average percentile score across tests was calculated to provide a global estimate of cognitive performance. Spearman's ρ correlations were performed to compare the CRIq, CIQ, and GOSE scores with the neuropsychological test scores and education level. Separate moderator regression analysis was performed to test whether the relationship between injury severity (presence or absence of intracranial injury and the GCS, <15 or 15) and cognitive outcome was moderated by our measure of cognitive reserve. Cognitive complaints of patients and HCs were compared using χ² tests. Cognitive complaints were compared with neuropsychological test scores using Spearman's ρ correlations. Missing values were present but did not exceed 5%, except for the 15WT (IR: 11.5% and DR: 14% missing values), CRIq, and GOSE (8% missing values). An α level of .05 was used. All statistical analyses were performed using the Statistical Package for the Social Sciences (SPPS), version 28.0 (IBM Corp, Armonk, New York), and a regression PROCESS macro system (version 4.3)³¹

RESULTS

Participant characteristics

A total of 52 patients with mTBI aged from 61 to 88 years at the time of injury were assessed. Also, 42 HCs aged from 60 to 80 years participated in the study and filled in the questionnaires, to be able to compare their psychological profile with the patient group. Demographic characteristics are described in Table1. The interval from injury to assessment was 2 to 6 months with a mean of 3.5 months. Patients and controls were well-matched for sex and education level, but not completely for age. Levels of community integration and functional outcome were determined in patients.

Neuropsychological test performances and cognitive reserve

Mean scores of patients on neuropsychological measures and percentage of patients scoring impaired are shown in Table2. After Bonferroni-Holm correction, patients with mTBI had significantly lower scores on 15 WT-IR, WNT-animals, WNT-professions and letter fluency, compared with the norm scores, with moderate to large effect sizes.

Scores of the CRIq, CIQ, and GOSE are shown in Table2. Only between the CRIq and the Digit Span task a moderate positive correlation was found (r_s = 0.49, p < .001). A significant, very strong positive correlation was found between the CRIq and education level (r_s = 0.71, P < .001). Cognitive reserve did not moderate the relationship between injury severity as defined by either intracranial injury (b = −0.17, P = .422) or GCS (b = −0.05, P = .744) and cognitive outcome. After Bonferroni-Holm correction, no associations were found between cognitive functions and CIQ scores and cognitive functions and scores on the GOSE.

Cognitive complaints

Cognitive complaints were present in almost half of the patients with mTBI: 57% experienced forgetfulness, 43% lack of concentration, and 45% reported slowness. Overall, patients experienced significantly more often cognitive complaints than HCs (see Table3). Cognitive complaints were compared to neuropsychological test scores. After Bonferroni-Holm correction, there was a significant, moderate negative correlation between complaints of concentration and TMT-B (r_s = −0.42, P = .003), but no other significant correlations were found for specific tests. Complaints of concentration also had a moderate negative correlation with the mean percentile score across tests (r_s = −0.34, P = .017).

Positive moderate correlations were found between all subjective cognitive complaints and anxiety and depression, respectively, except for forgetfulness with anxiety (see Table4). Mann-Whitney U tests showed no differences between elderly patients with mTBI and HCs in the level of anxiety and depression. However, when applying the cut-off score of 8, 10% of elderly patients with mTBI could be defined as depressed, and 10% as anxious. In the HC group no one met the cut-off point for depression, and only 2% could be defined as anxious. Fisher's exact test showed no differences between patients with mTBI and HC when using the cut-off scores for depression and anxiety. Premental health status was not associated with either cognitive complaints or the HADS scores.

DISCUSSION

This study shows that cognitive impairments in memory, processing speed, attention, and executive function are present in elderly patients with mTBI in the subacute phase (2 weeks to 6 months post-injury). Cognitive reserve was not associated with neuropsychological test performance, with the exception of the Digit Span task. Patients with mTBI reported cognitive complaints of forgetfulness, mental slowness, and lack of concentration more frequently than HCs. Interestingly, these complaints, except for complaints of concentration, were not associated with cognitive impairments, indicating a partial discrepancy between complaints and cognitive functions. Mental distress symptoms were significantly associated with these cognitive complaints.

Elderly patients with mTBI performed lower than a healthy norm group on memory, attention, processing speed, and executive function tests in the subacute phase, contrary to the typical findings in the overall mTBI population.^4,10 A possible explanation for this finding is increased vulnerability to secondary damage after mTBI due to normal age-related neural and vascular changes in the brain.⁹ This might cause more severe impairment in cognitive functions after a similar trauma than in younger patients.³² Also, it is conceivable that recovery rate in the elderly is slower than in younger adults, suggesting that a follow-up at a later timepoint post-injury might show less impairment in the elderly.³² Future studies should focus on assessing elderly patients over a longer period post-injury.

To date, this is the first study to investigate the relationship between cognitive reserve and cognitive impairment in elderly patients with mTBI. Contrary to expectations, no significant relationship was found, but a strong correlation existed between our measure of cognitive reserve and education level (0.71). This was not unexpected, given that one-third of the cognitive reserve questionnaire is about educational attainment. Nevertheless, this suggests a substantial overlap between these 2 constructs. Given that we used for each test, except one, norm scores, which were corrected for age, sex, and education level, it is not surprising that there was significant association with our measure of cognitive reserve. The only test for which norm scores did not correct for level of education was the Digit Span task. This was also the only neuropsychological task that correlated with our measure of cognitive reserve. Furthermore, the relationship between injury severity and cognitive outcome was not moderated by cognitive reserve. These findings contradict previous research on younger adults with TBI, which suggests that lower cognitive reserve is a risk factor for poorer cognitive functioning after mTBI.¹⁰ In this study, Stenberg etal¹⁰ used a cognitive composite score that was derived from norm scores, which were corrected for the influence of age and sex, but half of these norm scores disregarded education level. Hence, it was more likely to find an association between cognitive reserve and this cognitive composite score. Another explanation could be that advancing age might diminish the effect of cognitive reserve as a protective factor for the decline of cognitive functions, specifically after acute brain injuries. Due to normal aging, there are already functional and structural alterations in the brain: for example, volume loss, increased synaptic pruning, and reduced plasticity,³³ which may affect cognitive capacities. Hence, it is conceivable that older patients may exhibit a slower adaptation rate due to, for example, structural alterations associated with aging especially when encountering acute brain injury. In line with this, Moretti etal³³ described that normal age-related cognitive decline might be accelerated due to a TBI, as a result of the complex interaction between aging and TBI.

Complaints of forgetfulness, mental slowness, and concentration were reported in approximately half of the elderly patients with mTBI, which was significantly higher than in the HC group. Complaints of concentration were linked to decreased cognitive functioning across various domains, suggesting that complaints of concentration may in fact reflect impairments in objective cognitive functioning. This is an interesting finding, as no prior studies to our knowledge investigated this in the elderly mTBI population and it seems to deviate from what is typically known in the overall mTBI population. However, more research into the relationship between cognitive complaints and cognitive functioning in elderly patients with mTBI is necessary. Additionally, we found an association between worse performance on a test for executive functioning, namely the TMT-B, and complaints of concentration, which is consistent with prior findings in the overall mTBI population.³⁴ This finding could indicate that the TMT-B is a sensitive test to detect the possible nature of the complaints of concentration. Underlying this relationship might be that patients with cognitive impairments as a result of mTBI have to put in more cognitive effort, which may be interpreted as complaints in concentration. In conclusion, especially complaints of concentration may be an indicator for cognitive impairment in the elderly population.

For complaints of forgetfulness and mental slowness, we found no link with objective cognitive functioning. This could indicate a more generic effect, where other factors, such as mental distress, play a role in the presence of those complaints. In line with this hypothesis, anxiety and depressive symptoms were significantly associated with forgetfulness, concentration, and mental slowness. In the overall population with mTBI, these cognitive complaints do not necessarily limit daily functioning.¹⁸ Therefore, future studies should focus particularly on whether cognitive complaints influence the daily activities in the elderly mTBI population.

This study has some limitations. First, it cannot be ruled out that some patients experienced cognitive deficits for reasons unrelated to the mTBI, even though they were screened by a licensed neuropsychologist on dementia. We found lowered neuropsychological test performances in the cognitive domains associated with mTBI and not in domains associated with neurodegenerative disease, such as memory retention, which suggests that patients with mild cognitive impairment or dementia were successfully excluded. A methodological issue to consider is the lack of younger mTBI patients and the lack of more detailed information available on healthy elderly peers participating in this study. For example, information about comorbidities or cognitive functions of the healthy elderly could further enhance our knowledge on factors playing a role in outcomes after mTBI. Comparing patients with healthy peers would have contributed to our understanding of the relationship between our cognitive reserve measure and neuropsychological test performance. Although we were unable to conduct these analyses, the validity of the CRIq as a measure of cognitive reserve was examined in a validation study by Nucci etal.²⁸ The assumption that elderly mTBI patients score lower on neuropsychological tests than younger patients or healthy peers should be tested in a study including these groups. Furthermore, as cognitive reserve is a latent construct, measured in a variety of ways, it would be interesting to further investigate the findings of the current study, using different measures of cognitive reserve. For example, given that cognitive reserve might allow for more flexible strategy use, it is conceivable that this could be captured by executive function tasks.³⁵ Another issue to consider is that HCs were not fully matched for age. Elderly patients with mTBI were slightly older than HCs, which might have contributed to more frequently reported cognitive complaints in this group when compared with HCs. The results on cognitive complaints should therefore be interpretated with some caution.

CONCLUSIONS

This study shows cognitive impairments in memory and attention, processing speed, and executive functions in elderly patients in the subacute phase after mTBI. These impairments were not significantly associated with our measure of cognitive reserve. Furthermore, cognitive reserve did not moderate the relationship between injury severity and cognitive outcome. Difficulties in concentration could be an indicator for decreased overall objective cognitive functioning. Complaints of forgetfulness and mental slowness possibly represented more generic complaints, where other factors such as mental distress may play a role. For translation of our findings into clinical practice, we recommend careful evaluation of these patients in the subacute phase, to provide clinicians with guidelines for specific targets for treatment, focusing either on cognitive impairment, decreased mental well-being or both.

REFERENCES