Abstract

Meta-analytic studies have shown that mild traumatic brain injury (MTBI) has relatively negligible effects on cognitive functioning at 90 or more days post-injury. Few studies have prospectively examined the effects of MTBI in acute physical trauma populations. This prospective, cohort study compared the cognitive performance of persons who sustained a spinal cord injury (SCI) and a co-occurring MTBI (N = 53) to persons who sustained an SCI alone (N = 64) between 26 and 76 days (mean = 46) post-injury. The presence of MTBI was determined based on acute medical record review using a standardized algorithm. Primary outcome measures were seven neuropsychological tests that evaluated visual, verbal, and working memory, perceptual reasoning, and processing speed that controlled for potential upper extremity impairment. Persons who sustained SCI with or without MTBI had lower than expected performance across all neuropsychological tests, on average about 1 SD below the mean. Analysis of covariance indicated that persons with MTBI did not evidence greater impairment on any neuropsychological test. The aggregated effect size (Cohen's d) was −0.16. The strongest predictors of neuropsychological test scores were education, race, history of learning problems, and days from injury to rehabilitation admission. MTBI did not predict performance on any neuropsychological test. These findings are consistent with other controlled studies that indicate a single MTBI has negligible long-term impacts on cognition.

Introduction

Animal studies identified a pathophysiological correlate of mild traumatic brain injury (MTBI) over three decades ago (Ommaya & Gennarelli, 1974). Since then, MTBI has been examined in the emergency room (Alves, Macciocchi, & Barth, 1993), single- and multi-center longitudinal studies (Dikmen, Machamer, Winn, & Temkin, 1995; Levin et al., 1987), sports (Hinton-Bayre & Geffen, 2002; Macciocchi, Barth, Alves, Rimel, & Jane, 1996; McCrea et al., 2003), and spinal cord injury (SCI) rehabilitation (Davidoff, Morris, Roth, & Bleiberg, 1985).

Meta-analyses of controlled studies have found that a single MTBI has a moderate effect (Cohen's d = −0.41 to −0.54) on overall neuropsychological test performance 1–7 days post-injury (Belanger, Curtiss, Demery, Lebowitz, & Vanderploeg, 2005; Belanger & Vanderploeg, 2005; Schretlen & Shapiro, 2003), a small effect (d = −0.29) 7–30 days post-injury (Schretlen & Shapiro, 2003), negligible effects (d = −0.08) 30–89 days post-injury (Schretlen & Shapiro, 2003), and negligible effects (d = −0.12 to 0.04) 90 or more days post-injury (Binder, Rohling, & Larrabee, 1997; Schretlen & Shapiro, 2003).

Rohling and colleagues (2011) used a random effects meta-analytic technique to re-analyze existing outcome studies and establish effect sizes for multiple cognitive domains following MTBI. The mean effect size across all cognitive domains at 7 days post-injury was −0.39 and at 8–30 days post-injury was −0.32. At 31–92 days post-injury (mean = 58 days), the mean effect size across cognitive domains was −0.14, but moderate effects were observed for several individual domains including visual memory (−0.45), working memory (−0.34), and executive functioning (−0.32). The mean effect size at 93+ days (mean = 234 days) post-injury was −0.07 with only working memory (−0.19) showing an effect size above −0.10.

Meta-analytic techniques have provided important information on how MTBI affects cognitive functioning at various points in time post-injury, but all meta-analyses are constrained by methodological limitations in the studies selected for analysis. For example, MTBI investigations that exclude participants based on age, time from injury to assessment, co-occurring medical disorders, presence of pre-injury learning disorders, or that have high attrition can significantly reduce the generalizability of research findings to clinical practice (Corrigan et al., 2003; Luoto et al., 2013). Consequently, prospectively examining the effects of MTBI in a consecutive cohort of persons who have sustained co-occurring physical injuries and associated secondary conditions and receive standardized acute rehabilitation treatment may help address concerns about the representativeness and equivalence of MTBI and trauma control groups (Belanger et al., 2005; Dikmen & Levin, 1993; Rohling et al., 2011).

The traumatic SCI population has a high base rate of co-occurring MTBI (Davidoff et al., 1985, 1988; Macciocchi, Seel, Thompson, Byams, & Bowman, 2008), which provides an opportunity to study the acute effects of MTBI on cognitive functioning in a trauma cohort where the control group has equivalent physical injuries, secondary conditions, and treatment setting. Studying persons with traumatic SCI in the inpatient rehabilitation setting also allows for observation and analysis of medications and injury mechanisms that potentially contribute to lower than the expected neuropsychological test performance but are infrequently addressed in MTBI meta-analytic studies.

A number of factors other than MTBI that may explain neuropsychological test performance can also be examined in a prospective, cohort study. For instance Dikmen, Machamer, and Temkin (2001) report that socio-demographic factors are powerful predictors of low neuropsychological test performance. Moreover, pre-injury histories of low educational achievement or learning disabilities are not commonly examined in MTBI studies, despite evidence of their negative impact on test performance (Greiffenstein & Baker, 2003; Mapou, 2008).

In this prospective cohort study, we compared the neuropsychological test performance of persons who sustained traumatic SCI and documented co-occurring MTBI to a concurrent control group who sustained only a traumatic SCI. Demographically diverse participants were tested while receiving inpatient rehabilitation between 26 and 76 days post-injury, consistent with the 30–90-day post-injury epoch frequently reported in the meta-analysis literature (Rohling et al., 2011). We also examined socio-demographics, self-reported history of learning problems, markers of injury severity, and prescribed medications as potential covariates of neuropsychological test performance. Based on current research evidence from meta-analytic and standalone studies that used control groups, we hypothesized that there would be no differences in cognitive functioning between the SCI and MTBI and the SCI alone groups.

Methods

Participants

The current sample was a subset of a larger sample of persons participating in a National Institute of Rehabilitation Research (NIDRR) funded study on traumatic SCI and co-occurring TBI (Macciocchi et al., 2008). The study was conducted at an urban, NIDRR Model Spinal Cord Injury System center that is designated as a long-term acute care hospital. Persons with a traumatic SCI aged 16–59 admitted for rehabilitation over an 18-month period (2004–2005) were eligible for inclusion (n = 266). Due to the unavailability of bilingual examiners, non-English speaking persons were excluded from the study. Eighty percent of eligible persons consented to participate (n = 212) and 89% of enrolled participants completed all inpatient admission and outcome measurement (n = 189).

Participant data for the current study were included if persons had a medically documented traumatic SCI with either no documented brain injury or a medically documented, non-complicated MTBI as evidenced by negative imaging tests but either a period of post traumatic amnesia (PTA) < 24 h or a GCS total score of 11T, 13, or 14. The initial sample contained data on 140 persons with SCI alone or SCI and co-occurring MTBI. We excluded data on persons who previously sustained brain injury (n = 11) and persons tested before 26 days (n = 6) or after 92 days (n = 6). The final sample consisted of 117 participants.

Measures

Neuropsychological tests were selected to assess diverse cognitive skills while eliminating performance confounds due to upper extremity motor impairment associated with tetraplegia. The tests administered included Wechsler Adult Intelligence Scale-3 (WAIS-3) Digit Span (DS) and Letter-Number Sequencing (LNS) Tests (Wechsler, 1997); the Hopkins Verbal Learning Test 2nd Edition (HVLT-2; Brandt & Benedict, 2001); the Symbol Digit Modalities Test-Oral (SDMT-O; Smith, 1995); the Category Test-Short Form (SCAT; Wetzel & Boll, 2000); and the Continuous Visual Memory Test (CVMT; Trahan & Larrabee, 1988). All six tests are routinely used in clinical practice and research and have standardized administration and scoring procedures, good psychometric properties, and age corrected scores.

Procedure

The current study was approved by the host institution's Research Review Committee and informed consent was obtained from participants prior to enrollment. Participants were enrolled within 1 week of SCI acute rehabilitation admission. Trained research coordinators completed a structured clinical interview of the participant and when available a family member to document selected pre-injury medical issues and risk factors for cognitive impairment. A trained research coordinator administered all six neuropsychological tests during a single testing session. The research coordinator was blinded to brain injury diagnosis at the time of testing. MTBI diagnosis was established by two investigators (SNM and RTS) based on a standardized medical record review, which was completed 6 months prior to and independent of review or analysis of neuropsychological data (Macciocchi et al., 2008).

Data Analyses

Cramer's V was used to compare demographics and injury characteristics of the SCI alone and SCI and MTBI groups. Analysis of covariance (ANCOVA) was used to examine differences in neuropsychological test scores between the SCI alone and the SCI and MTBI groups. Eight covariates were selected based on their potential to impact neuropsychological test performance: highest education level achieved, self-reported history of learning problems (e.g., attention deficit, reading, or spelling problems), African-American race, ASIA C5-T1 motor subscale score, days from injury to rehabilitation admission (i.e., a surrogate marker of medical acuity/physical injury severity), injury due to motor vehicle crash (MVC), injury due to violence, and prescribed narcotic medications.

Eta squared (η2) was calculated for the ANCOVAs to identify the effect size of each of the eight covariates and the presence or the absence of MTBI controlling for all other model variables. Covariates with p ≥ .15 were removed one at a time from each model starting with the covariate with the highest p-value provided that removal did not change the direction or degree of the relationships between the presence of MTBI and neuropsychological test scores. Cohen's d was calculated using the standard deviation of the pooled sample to identify the effect size of MTBI versus no TBI on neuropsychological test scores. The percentage of persons with and without MTBI scoring 1.33 SDs below the mean or ninth percentile, 1.66 SDs below the mean or fifth percentile, and 2.0 SDs below the mean or second percentile were compared using the phi coefficient. Given the limited number of a priori between group tests for each research question, an alpha level of 0.05 was deemed appropriate.

Results

Table 1 provides participant characteristics data. Both the SCI and SCI + MTBI groups were predominately young (range 16–56) and men (82%) with equivalent mean education levels of 12 years (range 8–20 years). The sample was predominately White and African American. The incidence of self-reported, pre-injury learning problems was <15%.

Table 1.

Characteristics of SCI and MTBI and SCI alone samples

Characteristic Total (n = 117) SCI + MTBI (n = 53) SCI (n = 64) p-value 
Age (years) 27.85 ± 9.83 26.81 ± 9.87 28.70 ± 9.79 .302 
Gender (men, %) 82.1 88.7 76.6 .089 
Race    .420 
 White (%) 60.7 67.9 54.7  
 African American (%) 36.8 30.2 42.2  
 Other (%) 2.6 1.9 3.1  
Education (highest # years completed) 12.22 ± 2.00 11.89 ± 1.90 12.50 ± 2.06 .099 
Pre-injury learning problem (yes, %) 13.7 17.0 10.9 .344 
Injury Etiology    .000 
 Motor vehicle accident (%) 54.7 73.6 39.1  
 Violence (%) 21.4 3.8 35.9  
 Sporting injury (%) 14.5 7.5 20.3  
 Falls/flying object (%) 9.4 15.1 4.7  
Days from injury to rehabilitation admission 27.92 ± 14.11 30.19 ± 12.99 26.05 ± 14.81 .114 
SCI motor level and completeness    .696 
 C1-4, ASIA grade A–C (%) 6.0 5.7 6.3  
 C1-4, ASIA grade D (%) 5.1 5.7 4.7  
 C5-8, ASIA grade A–C (%) 30.8 34.0 28.1  
 C5-8, ASIA grade D (%) 5.1 7.5 3.1  
 T1-S3, ASIA grade A–C (%) 48.7 45.3 51.6  
 T1-S3, ASIA grade D (%) 4.3 1.9 6.3  
ASIA C5-T1 motor score at admission 34.59 ± 17.98 32.49 ± 18.52 36.33 ± 17.47 .252 
Days from injury to NP testing 45.92 ± 13.31 48.11 ± 13.05 44.11 ± 13.36 .106 
Narcotic medications (yes, %) 79.5 79.2 79.7 .953 
Characteristic Total (n = 117) SCI + MTBI (n = 53) SCI (n = 64) p-value 
Age (years) 27.85 ± 9.83 26.81 ± 9.87 28.70 ± 9.79 .302 
Gender (men, %) 82.1 88.7 76.6 .089 
Race    .420 
 White (%) 60.7 67.9 54.7  
 African American (%) 36.8 30.2 42.2  
 Other (%) 2.6 1.9 3.1  
Education (highest # years completed) 12.22 ± 2.00 11.89 ± 1.90 12.50 ± 2.06 .099 
Pre-injury learning problem (yes, %) 13.7 17.0 10.9 .344 
Injury Etiology    .000 
 Motor vehicle accident (%) 54.7 73.6 39.1  
 Violence (%) 21.4 3.8 35.9  
 Sporting injury (%) 14.5 7.5 20.3  
 Falls/flying object (%) 9.4 15.1 4.7  
Days from injury to rehabilitation admission 27.92 ± 14.11 30.19 ± 12.99 26.05 ± 14.81 .114 
SCI motor level and completeness    .696 
 C1-4, ASIA grade A–C (%) 6.0 5.7 6.3  
 C1-4, ASIA grade D (%) 5.1 5.7 4.7  
 C5-8, ASIA grade A–C (%) 30.8 34.0 28.1  
 C5-8, ASIA grade D (%) 5.1 7.5 3.1  
 T1-S3, ASIA grade A–C (%) 48.7 45.3 51.6  
 T1-S3, ASIA grade D (%) 4.3 1.9 6.3  
ASIA C5-T1 motor score at admission 34.59 ± 17.98 32.49 ± 18.52 36.33 ± 17.47 .252 
Days from injury to NP testing 45.92 ± 13.31 48.11 ± 13.05 44.11 ± 13.36 .106 
Narcotic medications (yes, %) 79.5 79.2 79.7 .953 

Notes: SCI = spinal cord injury; MTBI = mild traumatic brain injury; ASIA = American Spinal Injury Association Impairment Scale; C = cervical; T = thoracic; S = sacral; NP = neuropsychological.

All reported statistics are the mean ± SD or % of sample. Age, education, days from injury to rehabilitation admission and NP testing, and ASIA C5-T1 motor score p-values based on the one-way analysis of variance. Gender, race, pre-injury learning problem, injury etiology, and narcotic medication use p-values based on Cramer's V.

Participants in both the SCI alone and SCI and MTBI groups had equivalent SCI motor levels and completeness of injury. The SCI and MTBI group was more likely to be injured in a motor vehicle accident, whereas the SCI alone group was more likely to be injured as a result of violence or sporting injury. Of the participants who sustained MTBI, 93% had documented loss of consciousness lasting <30 min. PTA was present in 100% of participants with MTBI. The vast majority of participants were prescribed narcotic medications (80%). Neuropsychological tests were administered a mean of 46 days post-injury (range 26–76 days).

ANCOVA produced models of neuropsychological test performance that had effect sizes, that is, variance explained, as high as 0.350 for LNS and as low as 0.064 on the SCAT (Table 2). Completed education level (years) was a significant covariate on the five neuropsychological tests in which education corrected norms were not available. African-American race was a significant covariate in six of the seven neuropsychological test models and approached significance (p = .052) in the seventh model. Self-reported history of learning problems was significant on three tests of working memory and processing speed. No significant predictors of SCAT neuropsychological test performance were identified. Controlling for covariates, sustaining MTBI did not significantly contribute to neuropsychological test performance and had an effect size (η2) of 0.02 or less on all seven measures.

Table 2.

Parameter estimates and effect sizes for potential covariates and mild TBI on neuropsychological test scores

Tests model p, η2 Data reported Intercept Years education Race Learning problem Prescribed narcotics Days injury to rehabilitation Injury due to MVC Injury due to violence Sustained MTBI 
WAIS-3 LNS B 6.45 0.40 −2.13 −1.97 – – −1.06 −1.29 −0.65 
p < .001 95% CI 2.77, 9.94 0.14, 0.65 −3.27, −1.00 −3.33, −0.61   −2.23, 0.11 −2.95, 0.36 −1.73, 0.42 
η2 = 0.350 p-value .001 .003 <.001 .005   .076 .125 .232 
 η2  0.081 0.114 0.071   0.029 0.022 0.011* 
WAIS-3 DS B 7.79 0.36 −1.35 −1.40 −0.82 −0.03 – – −0.59 
p < .001 95% CI 4.37, 11.21 0.12, 0.59 −2.30, −0.39 −2.70, −0.09 −1.93, 0.28 −0.07, −0.00   −1.51, 0.34 
η2 = 0.252 p-value .000 .003 .006 .036 .142 .043   .210 
 η2  0.077 0.066 0.039 0.019 0.037   0.014 
SDMT-Oral B 49.47 – −4.52 −5.16 – −0.18 −3.97 −8.13 0.86 
p < .001 95% CI 44.34, 54.61  −8.82, −0.21 −10.35, 0.03  −0.31, −0.06 −8.37, 0.43 −14.25, −2.01 −3.10, 4.81 
η2 = 0.236 p-value .000  .040 .051  .005 .076 .010 .669 
 η2   0.038 0.034  0.069 0.028 0.059 0.002 
HVLT Delayed B 22.35 1.78 −8.03 – – – – – −3.20 
p < .001 95% CI 7.35, 37.35 0.63, 2.94 −12.43, −3.63      −7.43, 1.03 
η2 = 0.212 p-value .004 .003 .000      .137 
 η2  0.078 0.112      0.020 
HVLT Total B 25.90 1.49 −7.23 – – – – – −2.77 
p < .001 95% CI 12.05, 39.75 0.43, 2.55 −11.33, −3.14      −6.71, 1.16 
η2 = 0.194 p-value .000 .006 .001      .165 
 η2  0.065 0.099      0.017 
CVMT Total B 26.23 1.76 −11.90 – – – −5.40 – −1.14 
p < .001 95% CI 6.11, 46.35 0.24, 3.28 −18.28, −5.51    −11.84, 1.03  −7.49, 5.22 
η2 = 0.178 p-value .011 .024 .000    .099  .724 
 η2  0.045 0.109    0.024  0.001 
SCAT B 48.7 – −4.07 – −4.16 – – – −2.38 
p = .059 95% CI 43.72, 53.67  −8.17, 0.03  −8.99, 0.67    −6.34, 1.59 
η2 = 0.064 p-value .000  .052  .09    .238 
 η2   0.034  0.026    0.013 
Tests model p, η2 Data reported Intercept Years education Race Learning problem Prescribed narcotics Days injury to rehabilitation Injury due to MVC Injury due to violence Sustained MTBI 
WAIS-3 LNS B 6.45 0.40 −2.13 −1.97 – – −1.06 −1.29 −0.65 
p < .001 95% CI 2.77, 9.94 0.14, 0.65 −3.27, −1.00 −3.33, −0.61   −2.23, 0.11 −2.95, 0.36 −1.73, 0.42 
η2 = 0.350 p-value .001 .003 <.001 .005   .076 .125 .232 
 η2  0.081 0.114 0.071   0.029 0.022 0.011* 
WAIS-3 DS B 7.79 0.36 −1.35 −1.40 −0.82 −0.03 – – −0.59 
p < .001 95% CI 4.37, 11.21 0.12, 0.59 −2.30, −0.39 −2.70, −0.09 −1.93, 0.28 −0.07, −0.00   −1.51, 0.34 
η2 = 0.252 p-value .000 .003 .006 .036 .142 .043   .210 
 η2  0.077 0.066 0.039 0.019 0.037   0.014 
SDMT-Oral B 49.47 – −4.52 −5.16 – −0.18 −3.97 −8.13 0.86 
p < .001 95% CI 44.34, 54.61  −8.82, −0.21 −10.35, 0.03  −0.31, −0.06 −8.37, 0.43 −14.25, −2.01 −3.10, 4.81 
η2 = 0.236 p-value .000  .040 .051  .005 .076 .010 .669 
 η2   0.038 0.034  0.069 0.028 0.059 0.002 
HVLT Delayed B 22.35 1.78 −8.03 – – – – – −3.20 
p < .001 95% CI 7.35, 37.35 0.63, 2.94 −12.43, −3.63      −7.43, 1.03 
η2 = 0.212 p-value .004 .003 .000      .137 
 η2  0.078 0.112      0.020 
HVLT Total B 25.90 1.49 −7.23 – – – – – −2.77 
p < .001 95% CI 12.05, 39.75 0.43, 2.55 −11.33, −3.14      −6.71, 1.16 
η2 = 0.194 p-value .000 .006 .001      .165 
 η2  0.065 0.099      0.017 
CVMT Total B 26.23 1.76 −11.90 – – – −5.40 – −1.14 
p < .001 95% CI 6.11, 46.35 0.24, 3.28 −18.28, −5.51    −11.84, 1.03  −7.49, 5.22 
η2 = 0.178 p-value .011 .024 .000    .099  .724 
 η2  0.045 0.109    0.024  0.001 
SCAT B 48.7 – −4.07 – −4.16 – – – −2.38 
p = .059 95% CI 43.72, 53.67  −8.17, 0.03  −8.99, 0.67    −6.34, 1.59 
η2 = 0.064 p-value .000  .052  .09    .238 
 η2   0.034  0.026    0.013 

Notes: HVLT = Hopkins Verbal Learning Test; WAIS-3 DS = Wechsler Adult Intelligence Scale-3 Digit Span; SCAT = Short Category Test; WAIS-3 LNS = Wechsler Adult Intelligence Scale-3 Letter-Number Sequencing; CVMT = Continuous Visual Memory Test; SDMT = Symbol Digit Modality Test; η2 = eta-squared; B = beta coefficient; CI = confidence interval; – = covariate was not significant at p ≥ .15 and removed from the model without changing the nature of the relationships between other variables.

ASIA C5-T1 motor subscale score was not significant in any model and was not included in Table 2. SDMT and SCAT norms control for education level. Bold values indicate P < .05.

Univariate comparison of neuropsychological test performance between the SCI and MTBI and SCI alone groups revealed no differences on verbal memory (HVLT Total and Delayed), working memory (DS and LNS), perceptual reasoning (SCAT), visual memory (CVMT), or processing speed (SDMT-O; Table 3). Effect sizes (Cohen's d) for each neuropsychological test ranged from −0.26 to 0.08 with an aggregated effect size for the overall battery of −0.16 at a mean of 46 days post-injury.

Table 3.

Estimated marginal means (95% confidence intervals) and effect sizes (Cohen's d) for MTBI on neuropsychological test scores

Cognitive domains Tests SCI + MTBI
 
SCI
 
F-value p-value Pooled SD Cohen's d 
Mean 95% CI Mean 95% CI 
Verbal memory HVLT Delayed (T)* 37.73 34.65, 40.80 40.93 38.09, 43.77 2.25 .137 12.43 −0.26 
Verbal memory HVLT Total (T)* 38.56 35.69, 41.43 41.34 38.71, 43.96 1.95 .165 11.45 −0.24 
Working memory WAIS-3 DS (SS)* 9.30 8.63, 9.98 9.89 9.28, 10.50 1.59 .210 2.73 −0.22 
Perceptual reasoning SCAT (T)* 41.54 38.61, 44.47 43.92 41.26, 46.58 1.41 .238 10.84 −0.22 
Working memory WAIS-3 LNS (SS)* 8.64 7.89, 9.39 9.30 8.62, 9.97 1.45 .232 3.04 −0.21 
Visual memory CVMT Total (T)* 39.30 34.77, 43.83 40.44 36.35, 44.53 0.13 .724 17.18 −0.07 
Processing speed SDMT-Oral (T)* 38.91 36.12, 41.70 38.06 35.54, 40.57 0.18 .669 10.73 0.08 
Cognitive domains Tests SCI + MTBI
 
SCI
 
F-value p-value Pooled SD Cohen's d 
Mean 95% CI Mean 95% CI 
Verbal memory HVLT Delayed (T)* 37.73 34.65, 40.80 40.93 38.09, 43.77 2.25 .137 12.43 −0.26 
Verbal memory HVLT Total (T)* 38.56 35.69, 41.43 41.34 38.71, 43.96 1.95 .165 11.45 −0.24 
Working memory WAIS-3 DS (SS)* 9.30 8.63, 9.98 9.89 9.28, 10.50 1.59 .210 2.73 −0.22 
Perceptual reasoning SCAT (T)* 41.54 38.61, 44.47 43.92 41.26, 46.58 1.41 .238 10.84 −0.22 
Working memory WAIS-3 LNS (SS)* 8.64 7.89, 9.39 9.30 8.62, 9.97 1.45 .232 3.04 −0.21 
Visual memory CVMT Total (T)* 39.30 34.77, 43.83 40.44 36.35, 44.53 0.13 .724 17.18 −0.07 
Processing speed SDMT-Oral (T)* 38.91 36.12, 41.70 38.06 35.54, 40.57 0.18 .669 10.73 0.08 

Notes: HVLT = Hopkins Verbal Learning Test; WAIS-3 DS = Wechsler Adult Intelligence Scale-3 Digit Span; SCAT = Short Category Test; WAIS-3 LNS = Wechsler Adult Intelligence Scale-3 Letter-Number Sequencing; CVMT = Continuous Visual Memory Test; SDMT = Symbol Digit Modality Test; CI = confidence interval; SS = scaled score; T = T-score; SD = standard deviation.

Neuropsychological test battery mean effect size, Cohen's d = −0.16.

Comparison of binary classifications of lower than expected neuropsychological test performance, that is, 1.33 SDs below the mean or ninth percentile, 1.66 SDs below the mean or fifth percentile, and 2.0 SDs below the mean or second percentile showed a greater percentage of persons in the SCI and MTBI group who scored 1.33 SDs below the mean on the WAIS-3 DS (Table 4). No other differences between the SCI and MTBI and SCI alone groups were found.

Table 4.

Percent of participants with low neuropsychological test scores in SCI and MTBI versus SCI group

Cognitive domain Test SD≤ % SCI + MTBI % SCI p-value 
Verbal memory HVLT Total (T) −1.33 43.4 34.9 .351 
HVLT Total (T) −1.66 32.1 22.2 .232 
HVLT Total (T) −2.00 28.3 14.3 .063 
HVLT Delayed (T) −1.33 49.1 38.7 .265 
HVLT Delayed (T) −1.66 32.1 24.2 .347 
HVLT Delayed (T) −2.00 30.2 21.0 .256 
Working memory WAIS-3 DS (SS) −1.33 20.8 6.3 .019 
WAIS-3 DS (SS) −1.66 9.4 1.6 .055 
WAIS-3 DS (SS) −2.00 1.9 1.6 .893 
WAIS-3 LNS (SS) −1.33 19.2 19.0 .980 
WAIS-3 LNS (SS) −1.66 7.7 9.5 .729 
WAIS-3 LNS (SS) −2.00 3.8 4.8 .811 
Visual memory CVMT Total (T) −1.33 45.3 35.9 .305 
CVMT Total (T) −1.66 34.0 34.4 .963 
CVMT Total (T) −2.00 34.0 29.7 .621 
Perceptual reasoning SCAT (T) −1.33 40.4 34.9 .547 
SCAT (T) −1.66 21.2 15.9 .466 
SCAT (T) −2.00 15.4 9.5 .339 
Processing speed SDMT-Oral (T) −1.33 43.4 53.1 .295 
SDMT-Oral (T) −1.66 26.4 34.4 .353 
SDMT-Oral (T) −2.00 22.6 26.6 .625 
Cognitive domain Test SD≤ % SCI + MTBI % SCI p-value 
Verbal memory HVLT Total (T) −1.33 43.4 34.9 .351 
HVLT Total (T) −1.66 32.1 22.2 .232 
HVLT Total (T) −2.00 28.3 14.3 .063 
HVLT Delayed (T) −1.33 49.1 38.7 .265 
HVLT Delayed (T) −1.66 32.1 24.2 .347 
HVLT Delayed (T) −2.00 30.2 21.0 .256 
Working memory WAIS-3 DS (SS) −1.33 20.8 6.3 .019 
WAIS-3 DS (SS) −1.66 9.4 1.6 .055 
WAIS-3 DS (SS) −2.00 1.9 1.6 .893 
WAIS-3 LNS (SS) −1.33 19.2 19.0 .980 
WAIS-3 LNS (SS) −1.66 7.7 9.5 .729 
WAIS-3 LNS (SS) −2.00 3.8 4.8 .811 
Visual memory CVMT Total (T) −1.33 45.3 35.9 .305 
CVMT Total (T) −1.66 34.0 34.4 .963 
CVMT Total (T) −2.00 34.0 29.7 .621 
Perceptual reasoning SCAT (T) −1.33 40.4 34.9 .547 
SCAT (T) −1.66 21.2 15.9 .466 
SCAT (T) −2.00 15.4 9.5 .339 
Processing speed SDMT-Oral (T) −1.33 43.4 53.1 .295 
SDMT-Oral (T) −1.66 26.4 34.4 .353 
SDMT-Oral (T) −2.00 22.6 26.6 .625 

Notes: HVLT = Hopkins Verbal Learning Test; WAIS-3 DS = Wechsler Adult Intelligence Scale-3 Digit Span; SCAT = Short Category Test; WAIS-3 LNS = Wechsler Adult Intelligence Scale-3 Letter-Number Sequencing; CVMT = Continuous Visual Memory Test; SDMT = Symbol Digit Modality Test; CI = confidence interval; SS = scaled score; T = T-score; SD = standard deviation. Bold values indicate P < .05.

Discussion

Persons who sustained traumatic SCI and co-occurring MTBI did not evidence significantly greater impairment on neuropsychological tests when compared with persons who sustained a traumatic SCI alone. The aggregated MTBI effect size of −0.16 at 26–76 days post-injury in our sample is consistent with the MTBI aggregated effect size of −0.14 at 31–92 days post-injury reported by Rohling and colleagues (2011). Socio-demographic, pre-injury, and medical covariates explained more variance in neuropsychological test performance than MTBI. Consistent with previous research, education was positively associated with better neuropsychological test performance, whereas African-American race and self-reported pre-injury history of learning problems contributed to lower test performance (Dikmen et al., 2001; Greiffenstein & Baker, 2003; Mapou, 2008). Days from injury to rehabilitation admission and injury mechanism also evidenced stronger associations with neuropsychological test performance than a medically documented MTBI. Our findings reaffirm the critical importance of using control groups sampled from the same population when studying the effects of MTBI on cognitive functioning.

We found lower than expected neuropsychological test performance in our SCI sample with and without MTBI, with mean age-corrected test scores typically falling 1 SD below normative expectations. Several factors may explain overall low test performance in our sample. First, our sample was a traumatically injured cohort and markers of severity (i.e., days from injury to rehabilitation admission) and injury etiology (i.e., MVCs or violence) showed evidence of negatively affecting cognitive test performance. Furthermore, 36.8% of persons in our study identified themselves as African American, a rate that is almost three times the estimated percentage of African Americans in the U.S. census. Based on model beta coefficients, African-American participants scored approximately 0.4–1.2 SDs lower on all seven neuropsychological tests, which may be attributable to unequal access to educational opportunities, cultural differences, and lack of cultural equivalence in cognitive measures (Dotson, Kitner-Triolo, Evans, & Zonderman, 2008; Manly, Jacobs, Touradji, Small, & Stern, 2002; Manly, Schupf, Tang, & Stern, 2005). Finally, the high rates of low scores found in persons with SCI alone is consistent with other recent studies demonstrating relatively high base rates of low scores in normative samples (Binder, Iverson, & Brooks, 2009).

Our data can inform clinical practice in several ways. Our findings were consistent with the existing body of evidence from controlled studies that reports a single MTBI has limited effects on cognitive functioning at 30–89 days post-injury. Our analyses also demonstrate that pre-injury, demographic, and medical factors explain more variance in neuropsychological test scores than MTBI. Clinicians must consider pre-injury learning problems and co-occurring trauma as potential contributors to lower than expected neuropsychological test performance following MTBI. Considering race, ethnicity, culture, and education is also essential in post-MTBI test score interpretation. Lastly, using lower than expected neuropsychological test scores as the primary basis for diagnosing MTBI >30 days post-injury will likely result in high rates of false positive diagnoses.

Our study has several methodological strengths and limitations. Participants with and without MTBI were enrolled from a consecutive, demographically diverse sample of persons with SCI who had equivalent age, education, gender, motor level and completeness of injuries, days from injury to rehabilitation, medications, and rehabilitation treatment, which minimized the potential bias in post-injury environmental factors and non-equivalence of the comparison group. Acute medical records were rigorously retrieved and reviewed using a standardized diagnostic algorithm to determine the history of MTBI prior to examining covariates or neuropsychological test data. The percentage of persons diagnosed with TBI in our initial SCI study was the highest rate reported in an inpatient SCI rehabilitation sample (Macciocchi et al., 2008). The consent rate was high (80%) and the attrition rate was quite low (11%). Finally, the socio-demographic and injury-related covariates examined provide empirically plausible alternative explanations for lower than expected neuropsychological test performance in the SCI population.

With regard to study limitations, persons with SCI often have risk factors for reduced cognitive efficiency including depression, pain disorders, and insomnia (Sipski & Richards, 2006), which we were unable to systematically examine, but merit exploration in future MTBI studies. We did not conduct effort testing, which may have identified suboptimal effort as a covariate of lower than expected test performance in both the SCI and SCI + MTBI groups. Suboptimal effort can lower test performance by at least 1 SD (Green, Rohling, Lees-Haley, & Allen, 2001). Future studies of MTBI using trauma populations should consider using effort measures to exclude suboptimal effort as an explanation for lower than expected test scores. Our sample was predominately young, that is, <40 years old, and our results cannot be generalized to older persons with MTBI. Our sample also was not large but generated models with high power. Lastly, eta-squared and Cohen's d are effect size calculations for samples and are not corrected for populations.

In conclusion, few controlled studies have been conducted in populations with an MTBI and other significant trauma during the acute recovery period. We found no evidence that a single MTBI negatively impacted cognitive functioning following co-occurring SCI. These results support prospective, controlled studies conducted over the past two decades (Rohling et al., 2011). Further research on the impact of physical trauma and secondary conditions on neuropsychological test performance is warranted. The evidence from this study and the research literature indicate that clinicians who observe lower than expected neuropsychological test performance >30 days following a single MTBI must consider alternative explanations for test performance including socio-demographic factors, history of learning disorders, and co-occurring medical conditions.

Funding

Funding for this study was provided by National Institute of Disability and Rehabilitation Research: H113G03004.

Conflict of Interest

None declared.

References

Alves
W.
Macciocchi
S.
Barth
J.
Postconcussive symptoms after uncomplicated mild head injury
Journal of Head Trauma Rehabilitation
 , 
1993
, vol. 
8
 
3
(pg. 
48
-
59
)
Belanger
H. G.
Curtiss
G.
Demery
J. A.
Lebowitz
B. K.
Vanderploeg
R. D.
Factors moderating neuropsychological outcomes following mild traumatic brain injury: A meta-analysis
Journal of the International Neuropsychological Society
 , 
2005
, vol. 
11
 
3
(pg. 
215
-
227
)
Belanger
H. G.
Vanderploeg
R. D.
The neuropsychological impact of sports-related concussion: A meta-analysis
Journal of the International Neuropsychological Society
 , 
2005
, vol. 
11
 
4
(pg. 
345
-
357
)
Binder
L. M.
Iverson
G. L.
Brooks
B. L.
To err is human: “abnormal” neuropsychological scores and variability are common in healthy adults
Archives of Clinical Neuropsychology
 , 
2009
, vol. 
24
 
1
(pg. 
31
-
46
)
Binder
L. M.
Rohling
M. L.
Larrabee
G. J.
A review of mild head trauma. Part I: Meta-analytic review of neuropsychological studies
Journal of Clinical and Experimental Neuropsychology
 , 
1997
, vol. 
19
 
3
(pg. 
421
-
431
)
Brandt
J.
Benedict
R.
Hopkins Verbal Learning Test-Revised manual
 , 
2001
Lutz
Psychological Assessment Resources
Corrigan
J. D.
Harrison-Felix
C.
Bogner
J.
Dijkers
M.
Terrill
M. S.
Whiteneck
G.
Systematic bias in traumatic brain injury outcome studies because of loss to follow-up
Archives of Physical Medicine and Rehabilitation
 , 
2003
, vol. 
84
 
2
(pg. 
153
-
160
)
Davidoff
G.
Morris
J.
Roth
E.
Bleiberg
J.
Cognitive dysfunction and mild closed head injury in traumatic spinal cord injury
Archives of Physical Medicine and Rehabilitation
 , 
1985
, vol. 
66
 
8
(pg. 
489
-
491
)
Davidoff
G.
Thomas
P.
Johnson
M.
Berent
S.
Dijkers
M.
Doljanac
R.
Closed head injury in acute traumatic spinal cord injury: Incidence and risk factors
Archives of Physical Medicine and Rehabilitation
 , 
1988
, vol. 
69
 
10
(pg. 
869
-
872
)
Dikmen
S.
Levin
H.
Methodological issues in the study of mild head injury
Journal of Head Trauma Rehabilitation
 , 
1993
, vol. 
8
 
3
(pg. 
30
-
37
)
Dikmen
S.
Machamer
J.
Temkin
N.
Mild head injury: Facts and artifacts
Journal of Clinical and Experimental Neuropsychology
 , 
2001
, vol. 
23
 
6
(pg. 
729
-
738
)
Dikmen
S.
Machamer
W.
Winn
H. R.
Temkin
N.
Neuropsychological outcome at 1-year post head injury
Neuropsychology
 , 
1995
, vol. 
9
 
1
(pg. 
80
-
90
)
Dotson
V. M.
Kitner-Triolo
M.
Evans
M. K.
Zonderman
A. B.
Literacy-based normative data for low socioeconomic status African Americans
The Clinical Neuropsychologist
 , 
2008
, vol. 
22
 
6
(pg. 
989
-
1017
)
Green
P.
Rohling
M. L.
Lees-Haley
P. R.
Allen
L. M.
3rd.
Effort has a greater effect on test scores than severe brain injury in compensation claimants
Brain Injury
 , 
2001
, vol. 
15
 
12
(pg. 
1045
-
1060
)
Greiffenstein
M. F.
Baker
W. J.
Premorbid clues? Preinjury scholastic performance and present neuropsychological functioning in late postconcussion syndrome
The Clinical Neuropsychologist
 , 
2003
, vol. 
17
 
4
(pg. 
561
-
573
)
Hinton-Bayre
A. D.
Geffen
G.
Severity of sports-related concussion and neuropsychological test performance
Neurology
 , 
2002
, vol. 
59
 
7
(pg. 
1068
-
1070
)
Levin
H. S.
Mattis
S.
Ruff
R. M.
Eisenberg
H. M.
Marshall
L. F.
Tabaddor
K.
, et al.  . 
Neurobehavioral outcome following minor head injury: A three-center study
Journal of Neurosurgery
 , 
1987
, vol. 
66
 
2
(pg. 
234
-
243
)
Luoto
T. M.
Tenovuo
O.
Kataja
A.
Brander
A.
Ohman
J.
Iverson
G. L.
Who gets recruited in mild traumatic brain injury research?
Journal of Neurotrauma
 , 
2013
, vol. 
30
 
1
(pg. 
11
-
16
)
Macciocchi
S.
Barth
J. T.
Alves
W.
Rimel
R. W.
Jane
J. A.
Neuropsychological functioning and recovery after mild head injury in collegiate athletes
Neurosurgery
 , 
1996
, vol. 
39
 
3
(pg. 
510
-
514
)
Macciocchi
S.
Seel
R. T.
Thompson
N.
Byams
R.
Bowman
B.
Spinal cord injury and co-occurring traumatic brain injury: Assessment and incidence
Archives of Physical Medicine and Rehabilitation
 , 
2008
, vol. 
89
 
7
(pg. 
1350
-
1357
)
Manly
J. J.
Jacobs
D. M.
Touradji
P.
Small
S. A.
Stern
Y.
Reading level attenuates differences in neuropsychological test performance between African American and White elders
Journal of the International Neuropsychological Society
 , 
2002
, vol. 
8
 
3
(pg. 
341
-
348
)
Manly
J. J.
Schupf
N.
Tang
M. X.
Stern
Y.
Cognitive decline and literacy among ethnically diverse elders
Journal of Geriatric Psychiatry and Neurology
 , 
2005
, vol. 
18
 
4
(pg. 
213
-
217
)
Mapou
R.
Adult learning disabilities and ADHD. Research informed assessment
 , 
2008
New York
Oxford University Press
McCrea
M.
Guskiewicz
K. M.
Marshall
S. W.
Barr
W.
Randolph
C.
Cantu
R. C.
, et al.  . 
Acute effects and recovery time following concussion in collegiate football players: The NCAA Concussion Study
Journal of American Medical Association
 , 
2003
, vol. 
290
 
19
(pg. 
2556
-
2563
)
Ommaya
A. K.
Gennarelli
T. A.
Cerebral concussion and traumatic unconsciousness. Correlation of experimental and clinical observations of blunt head injuries
Brain
 , 
1974
, vol. 
97
 
4
(pg. 
633
-
654
)
Rohling
M. L.
Binder
L. M.
Demakis
G. J.
Larrabee
G. J.
Ploetz
D. M.
Langhinrichsen-Rohling
J.
A meta-analysis of neuropsychological outcome after mild traumatic brain injury: Re-analyses and reconsiderations of Binder et al. (1997), Frencham et al. (2005), and Pertab et al. (2009)
The Clinical Neuropsychologist
 , 
2011
, vol. 
25
 
4
(pg. 
608
-
623
)
Schretlen
D. J.
Shapiro
A. M.
A quantitative review of the effects of traumatic brain injury on cognitive functioning
International Review of Psychiatry
 , 
2003
, vol. 
15
 
4
(pg. 
341
-
349
)
Sipski
M. L.
Richards
J. S.
Spinal cord injury rehabilitation: State of the science
American Journal of Physical Medicine and Rehabilitation
 , 
2006
, vol. 
85
 
4
(pg. 
310
-
342
)
Smith
A.
Symbol Digit Modalities Test
 , 
1995
Los Angeles
Western Psychological Services
Trahan
D.
Larrabee
G.
Continuous Visual Memory Test professional manual
 , 
1988
Odessa
Psychological Assessment Resources
 
Wechsler
D.
Administration and scoring manual for the Wechsler Adult Intelligences Scale-Third Edition
 , 
1997
San Antonio
The Psychological Corporation
 
Wetzel
L.
Boll
T. J.
Short Category Test booklet
 , 
2000
Los Angeles
Western Psychological Services