Previous studies of cognitive decline in patients with neurocognitive disorder due to traumatic brain injury (NCD-TBI) have often failed to control for baseline factors such as premorbid intelligence. The purpose of the current study was to estimate and compare cognitive function among three groups (controls, complicated mild/moderate TBI, and severe TBI) after controlling for premorbid intelligence.
Severity of TBI was classified as complicated mild/moderate or severe based on duration of loss of consciousness and brain neuroimaging results. Premorbid intelligence quotients (IQs) were estimated with the Oklahoma Premorbid Intelligence Estimate. There were no differences in premorbid intelligence between the groups, which were also matched for age and education. Current cognitive function was evaluated with the Wechsler Adult Intelligence Scale-Fourth Edition.
Comparison of current cognitive function among the three groups indicated significant group differences for all indexes and subtest scores. Processing speed showed the highest effect size. However, only working memory differed significantly between the two NCD-TBI groups.
The present findings suggest that mental memory manipulation processes seem to be more sensitive to TBI severity than are perceptual-motor processes. Specifically, both auditory rehearsal/discrimination and mental alertness/manipulation will be most strongly influenced by TBI severity.
Neurocognitive disorder due to traumatic brain injury (NCD-TBI) is often accompanied by cognitive decline [
Here, current cognitive function is assessed with the Wechsler Adult Intelligence Scale-Fourth Edition (WAIS-IV) [
Although previous studies have consistently highlighted differences in WAIS between patients with differing severities of NCD-TBI, as well as between patients with NCD-TBI and controls, no consideration has been given to premorbid intelligence [
As this suggests, neuropsychological assessment to determine cognitive functioning before and after NCD-TBI onset is essential to fully understand the effects of brain damage [
The current study was based on the hypothesis that there are no significant differences in premorbid intelligence among three groups: controls, patients with complicated mild/moderate TBI, and patients with severe TBI. We used the recently developed Oklahoma Premorbid Intelligence Estimate (OPIE) to estimate premorbid intelligence [
Initially, 103 Korean patients with NCD-TBI were enrolled in the study. They all consisted of inpatients and outpatients who were receiving care in the department of psychiatry. TBI groups were classified based on severity of trauma, which was determined by duration of loss of consciousness and abnormal neuroimaging results [
To control for the effects of age and education level on premorbid intelligence, participants in the control and severe TBI groups were matched for these variables with participants in the complicated mild/moderate TBI group. Data for healthy controls comprised partial data from the standardization study of the Korean WAIS-IV [
All the study procedures were reviewed and approved by the Local Institutional Ethics Review Board (WKUH 2017-04-023). All the participants were given informed consent before study.
The Korean Wechsler adult intelligence scale-fourth edition (WAIS-IV) was used to assess current cognitive function of patients with NCD-TBI [
To account for premorbid intelligence, premorbid IQ of TBI patients was estimated. The Oklahoma premorbid intelligence estimate (OPIE) method was recently developed and has been validated in clinical settings. The OPIE includes regression formulas developed by combining demographic variables and WAIS subtest scores [
In the present study, premorbid intelligence was estimated with the Korean version of the OPIE method for the WAIS-IV (OPIE-IV) [
The full formulas used in the current study were as follows: [67.173+1.722×(information subtest raw score)+0.161×(age)+0.995×(education level)]. Education level was scored as 1 for <8 years, 2 for 9–11 years, 3 for 12 years, 4 for 13–15 years, and 5 for ≥16 years.
One-way ANOVA was conducted to examine differences in demographic variables among the groups. Furthermore, frequency analysis of the descriptive classification for both premorbid and current IQ was also performed. Subsequently, a paired t-test was used to identify any difference between estimated premorbid and current intelligence within groups. Effect sizes (Cohen’s d) were calculated for statistically significant differences [
However, based on current IQ, 32% of patients in the complicated mild/moderate TBI group had “Average” and “Low average” ratings, while only 17% of the severe TBI group had “Average” and “Low average” ratings. In contrast, current full-scale IQ for controls suggested that more than 84% had “Average” and “Low average” ratings. Finally, “Extremely low” ratings accounted for 0%, 40.6%, and 67.7% of the control, complicated mild/moderate TBI, and severe TBI groups, respectively.
Based on scaled subtest scores, there were significant group differences with “large” effect sizes for all subtests except matrix reasoning. Moreover, symbol search and coding subtests accounted for over 50% of the variance (59.6% and 52.9%, respectively).
In this study, patients with NCD-TBI were first classified based on the severity of their injury. Subsequently, differences between premorbid and current cognitive function were investigated and domains in which patients differed from healthy adults were identified.
The current research was conducted to account for estimated premorbid IQ among healthy controls, patients with complicated mild/moderate TBI, and patients with severe TBI. Consequently, there was no significant difference in estimated premorbid IQ among the three groups (
A frequency analysis of the Wechsler’s descriptive classification was also conducted (
There were significant within-group differences in estimated premorbid and current IQ for patients with both complicated mild/moderate TBI and severe TBI (
These findings are inconsistent with previous studies. In a study by Donders and Strong [
In the current study, there was a significant difference between NCD-TBI groups in digit span and arithmetic, two subgroups subsumed within working memory. Therefore, both digit span and arithmetic are more susceptible to the effects of TBI severity than are other subtests [
We also investigated differences in general ability and cognitive proficiency, for both of which there were significant differences among the groups. In particularly, the magnitude of between-group differences for cognitive proficiency were associated with a very high effect size (η2=0.438). This is presumably because cognitive proficiency includes subtests that belong to both working memory and processing speed domains. However, both general ability and cognitive proficiency did not demonstrate sensitivity to TBI severity.
The present study contributes to the understanding of the impact of NCD-TBI on cognitive function with the following main findings: 1) While previous studies have compared current IQ for of patients with NCD-TBI to those of healthy adults, the preliminary aim of this study was to compare current IQ between these groups after controlling for estimated premorbid IQ. Ensuring the same baseline IQ can aid the interpretation and comparison of current cognitive function between groups; 2) The OPIE-IV method adopted in this study is the most recent premorbid intelligence estimation protocol to be developed. In particular, the Korean OPIE-IV method used in this study is less sensitive to variations in brain injury patients and should not overestimate or underestimate premorbid intelligence; 3) Although previous studies have reported processing speed as an indicator to compare different NCD-TBI groups, our results indicate that working memory is more sensitive to TBI severity. Moreover, processing speed most accurately distinguishes TBI from healthy adults. However, among cognitive domains tested, working memory accounts for the most variance in the discrimination of NCD-TBI groups according to severity; 4) The current study also presented a comparison of general ability and cognitive proficiency among the groups. As with IQ, there was a significant difference among the groups, with a greater proportion of the variance accounted for by cognitive proficiency than by general ability.
Despite the strengths of the current study, it is subject to some limitations that could be addressed by further research. The sample size of the study is adequate for the statistical analyses performed. However, studies with larger samples should be conducted to increase the statistical power. Additionally, unlike previous studies, working memory, but not processing speed, was observed to significantly discriminate TBI according to severity. This novel finding requires replication. Lastly, this study did not employ strict criteria for the classification of NCD-TBI participants. The present study employed both loss of consciousness and neuroimaging results to stratify TBI patients. The application of all the requisite criteria was difficult because of insufficient or unreliable information provided due to the nature of emergency medicine and potentially subjective reporting by the observer. However, future studies should be undertaken with the application of at least 3–4 criteria for stratification of NCD-TBI patients based on injury severity.
This study was supported by Wonkwang University in 2019.
The authors have no potential conflicts of interest to disclose.
Conceptualization: Sang-Yeol Lee. Data curation: Sang-Yeol Lee. Formal analysis: Min-Jung Soh, Hye-Jin Lee. Funding acquisition: Sang-Yeol Lee. Investigation: Kyu-Sic Hwang, Min-Jung Soh, Hye-Jin Lee. Methodology: Seung-Ho Jang, Kyu-Sic Hwang. Project administration: Seung-Ho Jang. Resources: Kyu-Sic Hwang, Min-Jung Soh, Hye-Jin Lee. Software: MinJung Soh, Hye-Jin Lee. Supervision: Sang-Yeol Lee. Validation: Kyu-Sic Hwang, Min-Jung Soh, Hye-Jin Lee. Visualization: Kyu-Sic Hwang, MinJung Soh, Hye-Jin Lee. Writing—original draft: Sang-Yeol Lee, Kyu-Sic Hwang. Writing—review & editing: Sang-Yeol Lee, Seung-Ho Jang.
Participant baseline characteristics
Variable | Normal control group (N=32) | Complicated mild/moderate TBI group (N=32) | Severe TBI group (N=31) | F | p |
---|---|---|---|---|---|
Mean±SD | Mean±SD | Mean±SD | |||
Age | 49.15±13.28 | 52.43±12.84 | 49.13±13.36 | 0.665 | 0.517 |
Years of education | 9.66±3.06 | 9.88±2.93 | 10.16±3.55 | 0.199 | 0.820 |
Estimated premorbid IQ | 89.52±8.56 | 89.14±8.35 | 88.35±7.15 | 0.171 | 0.843 |
IQ: intelligence quotient, TBI: traumatic brain injury
Frequency analysis of premorbid and current full-scale IQ, by group
Descriptive classification |
Normal control group (N=32) |
Complicated mild/moderate TBI group (N=32) |
Severe TBI group (N=31) |
|||
---|---|---|---|---|---|---|
Estimated premorbid IQ, N (%) | Current IQ, N (%) | Estimated premorbid IQ, N (%) | Current IQ, N (%) | Estimated premorbid IQ, N (%) | Current IQ, N (%) | |
Very superior | 0 (0) | 0 (0) | 0 (0) | 0 (0) | 0 (0) | 0 (0) |
Superior | 0 (0) | 0 (0) | 0 (0) | 0 (0) | 0 (0) | 0 (0) |
High average | 1 (3.1) | 3 (9.4) | 1 (3.1) | 0 (0) | 0 (0) | 0 (0) |
Average | 14 (43.8) | 21 (65.6) | 13 (40.6) | 4 (12.5) | 14 (45.2) | 2 (6.5) |
Low average | 15 (46.9) | 3 (9.4) | 14 (43.8) | 6 (18.8) | 13 (41.9) | 3 (9.7) |
Borderline | 2 (6.3) | 5 (15.6) | 4 (12.5) | 9 (28.1) | 4 (12.9) | 5 (16.1) |
Extremely low | 0 (0) | 0 (0) | 0 (0) | 13 (40.6) | 0 (0) | 21 (67.7) |
the classification is followed by descriptive classification of intelligence by Wechsler, which refers “Very superior” for 130 and above, “Superior” for 120 to 129, “High average” for 110 to 119, “Average” for 90 to 109, “Low average” for 80 to 89, “Borderline” for 70 to 79, and “Extremely low” for 69 and below.
IQ: intelligence quotient, TBI: traumatic brain injury
Within group differences between premorbid and current full-scale IQ
Group | Estimated premorbid IQ | Current IQ | Difference value | t | p | d |
---|---|---|---|---|---|---|
Mean±SD | Mean±SD | |||||
Normal control group (N=32) | 89.52±8.56 | 94.72±11.89 | -5.20 | -1.85 | 0.074 | NS |
Complicated mild/moderate TBI group (N=32) | 89.14±8.35 | 70.53±17.48 | 18.61 | 8.61 |
0.000 | 1.44 |
Severe TBI group (N=31) | 88.35±7.15 | 62.65±15.87 | 25.70 | 12.67 |
0.000 | 2.27 |
p<0.001.
IQ: intelligence quotient, TBI: traumatic brain injury, NA: not significant
Differences in current cognition, by group
Current score |
Group |
Statistic |
|||||
---|---|---|---|---|---|---|---|
Index | Subtest | Normal control (N=32) | Complicated mild/moderate TBI (N=32) | Severe TBI group (N=31) | F | p | η2 |
Mean±SD | Mean±SD | Mean±SD | |||||
VCI | 96.50±12.86 | 78.53±17.18 | 72.74±15.66 | 20.67 |
0.000 | 0.310 | |
SI | 8.78±3.06 | 5.81±3.50 | 4.16±2.91 | 17.27 |
0.000 | 0.273 | |
VC | 9.78±3.00 | 5.97±3.34 | 5.19±2.97 | 19.98 |
0.000 | 0.303 | |
IN | 9.31±2.43 | 6.66±3.18 | 6.13±3.18 | 10.63 |
0.000 | 0.188 | |
PRI | 94.34±12.03 | 78.50±17.18 | 73.65±18.18 | 14.12 |
0.000 | 0.235 | |
BD | 9.50±3.10 | 6.59±3.78 | 5.00±3.36 | 14.02 |
0.000 | 0.234 | |
MR | 8.12±2.56 | 6.06±3.63 | 5.74±3.59 | 5.18 |
0.007 | 0.101 | |
VP | 9.34±2.73 | 6.63±2.73 | 5.81±3.25 | 12.82 |
0.000 | 0.218 | |
WMI |
94.94±12.03 | 79.63±19.71 | 68.61±15.23 | 21.60 |
0.000 | 0.320 | |
DS |
8.94±2.30 | 6.03±4.20 | 3.87±3.12 | 18.75 |
0.000 | 0.290 | |
AR |
8.94±2.54 | 6.53±3.02 | 5.06±2.52 | 16.54 |
0.000 | 0.264 | |
PSI | 101.72±14.58 | 68.81±15.80 | 62.77±13.44 | 64.97 |
0.000 | 0.585 | |
SS | 10.41±3.12 | 4.10±2.78 | 3.03±2.20 | 67.72 |
0.000 | 0.596 | |
CD | 9.66±3.10 | 4.16±2.73 | 3.13±2.36 | 51.76 |
0.000 | 0.529 | |
Full-Scale IQ | 94.72±11.89 | 70.53±17.48 | 62.65±15.87 | 38.02 |
0.000 | 0.453 | |
GAI | 94.38±12.89 | 75.09±17.96 | 68.68±17.66 | 21.27 |
0.000 | 0.316 | |
CPI | 97.22±11.47 | 68.41±17.76 | 58.97±15.03 | 56.09 |
0.000 | 0.549 |
means that significant univariate differences between complicated mild/moderate and severe TBI group,
p<0.01,
p<0.001.
K-WAIS-IV: Korean Wechsler Adult Intelligence Scales-Fourth Edition, TBI: traumatic brain injury, VCI: Verbal Comprehension Index, PRI: Perceptual Reasoning Index, WMI: Working Memory Index, PSI: Processing Speed Index, BD: block design, SI: similarities, DS: digit span, MR: matrix reasoning, VC: vocabulary, AR: arithmetic, SS: symbol search, VP: visual puzzles, IN: information, CD: coding, LN: letter number sequencing, GAI: General Ability Index, CPI: Cognitive Proficiency Index