Dichotic listening: expanded norms and clinical application

Transcription

1 Archives of Clinical Neuropsychology 17 (2002) Dichotic listening: expanded norms and clinical application John E. Meyers a, *, Richard J. Roberts b, John D. Bayless c, Kurt Volkert a, Paul E. Evitts d a Mercy Rehabilitation Clinic, 500 Jackson Street, Ste 340, Sioux City, IA 51101, USA b VA Medical Center, Iowa City, IA, USA c Department of Psychiatry, University of Iowa, Iowa City, IA, USA d University of Northern Iowa, Cedar Falls, IA, USA Accepted 29 September 2000 Abstract The object of this study was to provide an expanded normative base for the Dichotic Word Listening Test (DWLT), with particular emphasis on the performance of older individuals. The normative study consisted of 336 community living volunteers. These new norms were used to compare several groups of neurologically impaired patient groups. DWLT was found to be sensitive to the presence of brain injury, and also to the degree of acute injury as measured by loss of consciousness. The results of the short form version of the DWLT test showed 100% specificity and 60% sensitivity for mildly brain-injured patients to 80% sensitivity for more severely brain-injured patients. The respective sensitivities for Left CVA and Right CVA were 55% and 88%. The present findings suggest that the DWLT is a valid and easy to use clinical tool. D 2001 National Academy of Neuropsychology. Published by Elsevier Science Ltd. Keywords: Dichotic listening; Neuropsychology; Adult; Brain injury Numerous studies have demonstrated that experimental dichotic listening procedures are sensitive to cerebral dysfunction due to various types of neurologic disease processes and different forms of brain injury. Briefly, dichotic word listening is assessed by presenting a single word to a subject s ear through stereo headphones, while simultaneously presenting a different word (usually matched for syllable length) to the other ear; the subject is asked to try to repeat both words. In their group of patients with lateralized lesions, Sparks, Goodglass, and Nickel (1970) found the lesion effect in which there was a loss of relative listening * Corresponding author. Tel.: /01/$ see front matter D 2001 National Academy of Neuropsychology. PII: S (00)

2 80 J.E. Meyers et al. / Archives of Clinical Neuropsychology 17 (2002) effectiveness in the contralateral ear. That is, their group of patients with right hemisphere lesions produced left ear extinctions on double simultaneous presentation of auditory stimuli. However, left hemisphere lesions produced ipsilateral (that is, paradoxical ) as well as contralateral ear extinctions. This ipsilateral effect was explained by a model in which information presented to the right ear is more directly routed to left hemisphere language areas, but information presented to the left ear is routed first to the right temporal lobe, then to the left hemisphere via anterior commissural fibers. Left-sided lesions affecting these fibers could thus produce a relative left ear extinction. Anatomical localization was elucidated by Rubens, Johnson, and Speaks (1978) who found that such left hemisphere lesions must extend deeply into left central or parietal white matter (p. 396). Further anatomical refinement was provided by Damasio and Damasio (1979), who indicated that the pathway leaves the auditory cortex traveling backward and upward to arch around the lateral ventricles and joins the callosum in its posterior region (p. 644). They noted that damage to any portion of the interhemispheric auditory pathway produces left ear extinction if the subject has left hemisphere speech dominance (p. 644). Rubens, Froehling, Slater, and Anderson (1985) frequently observed left ear suppressions in their group of patients with multiple sclerosis. This deficit was presumably due to degradation of auditory pathways through demyelinization, although such demyelinization did not have to be severe for the dichotic suppression to occur. They noted that their consonant vowel combinations made the task difficult, and were concerned that tasks using high contrast, real word pairs might decrease score variability in normals at the cost of reduced sensitivity to cerebral dysfunction. More recently, dichotic listening effects have been noted in aphasia (Bouma & Ansink, 1988), demyelinating disorders (Rao et al., 1989), primary degenerative dementia (Mohr, Cox, Williams, Chase, & Fedio, 1990), seizures (Roberts, Varney, Paulsen, & Richardson, 1990), and closed head injury (Levin et al., 1989). Roberts and colleagues developed norms for an abbreviated form of the Damasio stimulus tape, and recommended its use for standard clinical assessment. For example Roberts et al. (1990), Springer, Garvey, Varney, & Roberts (1991), and Verduyn, Hilt, Roberts, and Roberts (1992) illustrated the use of dichotic listening as a marker for subtle electrophysiological dysfunction in persons with brain injury. Richardson, Springer, Varney, Struchen, and Roberts (1994) suggested its routine use in evaluation for closed head trauma. Subsequently, Roberts et al. (1994) standardized a dichotic listening procedure suitable for everyday neuropsychological practice, the Dichotic Word Listening Test (DWLT; Auditec of St. Louis, 1991). It was their purpose to detect more robust deficits in bilateral and unilateral central auditory processing secondary to cerebral dysfunction, rather than subtle ear advantages resulting from lateralization of language processing. The task requires no special equipment apart from a portable stereo cassette player (CD format is also available) with headphones. Twenty unilaterally presented practice items allow the subject to adjust volume for optimal levels, ensure proper headphone placement, and provide a screening for potentially confounding hearing loss/impairment in speech discrimination. All items are common, fourth grade level English words. The 60 dichotic word pairs comprise the long form, with the first 30 pairs used as the short form version; short and long forms were found to be comparable.

3 J.E. Meyers et al. / Archives of Clinical Neuropsychology 17 (2002) The DWLT yields three index scores; the Left Index and Right Index scores are the overall number of words correctly repeated in respective single ears. The Both Ear Index is the total number of items in which both words of the dichotic pair are correctly repeated. It is important to differentiate defective performance in the Both Ear Index from bilateral suppression, which occurs when both the Left and Right ear Indices are in the defective range. Norms for university students were provided, as well as for urban/rural adults and children. Canadian norms were also collected to provide cross-cultural information. DWLT performances of a brain-injured sample were presented as well. In their group of control subjects, age and education were not found to be significant factors, but the groups did not contain a large number of persons in late adulthood. In a study of patients with multiple seizure like symptoms, Springer et al. (1991) found no statistical difference in test scores 6 weeks apart (test retest). DWLT was also found to have good sensitivity, with 80% of their subjects showing impairment in DWLTL performance. Roberts et al. (1994) also found that the short form DWLT retained the same psychometric properties as the longer version. Although there are other types of dichotic listening tasks (Spreen & Strauss, 1998), the DWLT used for this study was selected because preliminary norms had already been collected and could be added to, and the format of the single word presentation makes the test capable of being administered to a wide variety of clinical populations. The purpose of the present investigation was two-fold: (a) to provide more extensive and current normative data for the 30-item short form of the DWLT for use with older adults; and (b) to examine the clinical utility of the DWLT for detecting impaired neurocognitive processing in adults with cerebrovascular accidents and adults with head trauma of varying severity. 1. Experiment Method Participants The sample included 136 of the participants originally gathered by Roberts et al. (1994) from the university and rural samples. Of these 136 participants, 53 were male and 83 were female; 118 were right handed and 18 were left handed; all were Caucasian. The mean age of this pool of participants was years (S.D. = 12.3). This group had 13.8 years of education (S.D. = 1.7). For the present study, 200 community dwelling volunteers were added to the normative pool. Of these 200 participants, 76 were male and 124 were female; 192 were right handed and 8 were left handed; 192 were Caucasian, 1 African American, and 1 of mixed heritage. The mean age was 44.6 years (S.D. = 19.2) and the sample had a mean of 13.8 years of education (S.D. = 2.62). All participants (in both pools of participants) denied any history of neurological disease or hearing loss. If a participant failed two or more of the unilateral practice items or were

4 82 J.E. Meyers et al. / Archives of Clinical Neuropsychology 17 (2002) wearing hearing aids, they were excluded from the study sample. All participants in both data sets were living in the Midwest Procedure All participants were administered the DWLT task in the standard format, using the standard instructions. Prior to beginning the test, the earphones were checked to determine that they were working properly, and had equal loudness in each ear when listened to by the tester. Participants were given the first 10 practice items, which serve as a hearing screening test, and to familiarize the participant with the procedure. Following the practice items, the first 30 items of the DWLT were administered. This comprised the short form DWLT as identified by Roberts et al. (1994). Several methods were used to assess the comparability of the data collected by Roberts et al. (1994) with the community volunteers. There was no significant difference between these two groups in education, gender, or ethnic background ( P >.05), but the Roberts et al. (1994) group was comprised of younger participants [t(1,334) = 7.15, P <.000] and had more left-handed participants. A c 2 showed a significant difference in the groups for handedness ( P =.002). There was no significant correlation within either group between handedness and the DWLT scores. Therefore, it was felt that data from the two groups could be combined. The difference in age between the two groups was expected as one of the purposes of this study was to extend the age of the normative groups. Data from the year-old participants in both groups were compared using a t test of independent samples. This age group was selected as it represented the largest overlap in participants in an age group for each sample. The Roberts et al. (1994) sample was generally younger (though not exclusively), and the current sample was older (though not exclusively). It was necessary to find a general point of overlap to assess the similarity of the two samples. No significant differences were found in mean Left, Right, and Both Index performances (all P s >.05). Subjects from the two groups were therefore appropriate for pooling, resulting in a total sample of 336 participants. Further analysis of the effects of handedness on the data pool overall were also performed and will be discussed later. Subjects were divided into age groups for ages 16 19, and each succeeding decade of life. The total sample consisted of 129 males and 207 females, 310 were right handed and 26 were left handed. In the combined sample 1 participant was African-American, 334 were Caucasian, and 1 was of mixed ethnic background Results Demographic data and Left, Right, and Both ear Index performances are summarized in Table 1. A c 2 comparing gender and handedness with the DWLT Indices (Left, Right, Both) was performed, all comparisons showed a P >.05. Race was excluded from the analysis because of the limited number of non-caucasian participants. A MANCOVA was performed with dependent variables defined as the DWLT ear scores, independent

5 J.E. Meyers et al. / Archives of Clinical Neuropsychology 17 (2002) Table 1 Summary of normative data by age with 5th percentile Group n Age Education Left Right Both (0.7) 12.1 (0.8) 27.1 (1.7) 27.5 (1.3) 26.3 (1.7) 5th percentile cutoff (2.8) (1.5) 26.8 (1.6) 27.1 (1.5) 25.2 (2.0) 5th percentile cutoff (3.0) 14.3 (2.4) 26.9 (1.9) 27.2 (1.9) 25.3 (2.6) 5th percentile cutoff (2.8) 14.2 (2.6) 26.6 (2.2) 26.0 (2.5) 23.7 (3.5) 5th percentile cutoff (2.7) 13.5 (2.8) 26.1 (2.1) 25.6 (3.0) 23.1 (3.4) 5th percentile cutoff (3.1) 13.9 (2.5) 25.2 (3.0) 25.6 (2.4) 22.5 (3.1) 5th percentile cutoff (5.1) 12.3 (2.0) 21.2 (4.7) 22.3 (5.1) 16.9 (6.1) 5th percentile cutoff Total (17.8) 13.8 (2.3) 26.2 (2.7) 26.3 (2.7) 24.0 (3.8) Left = DWLT Left ear correct score. Right = DWLT Right ear correct score. Both = DWLT Both correct score. variables were defined as age and education, and covariates were handedness, race, and gender. None of the covariates, main effects, or interactions were significant ( P >.05). Consistent with the previous 1994 DWLT study, these factors were not significant in their effect on the DWLT variables. Education was only slightly correlated with the Left ear performance (r =.11, P <.05), although there was a significant correlation, the effect was insubstantial and not clinically relevant. Education was not significantly correlated with Right or Both ear indices ( P s >. 05). The 70+ age group was different from all other groups as indicated by ANOVA (see next paragraph for discussion) and so a separate correlation with education was made. No significant correlation with education was found for this age group ( P >.05). Mean index score performances for the different age groups were then assessed using a single factor ANOVA with Tukey HSD post hoc comparisons. Tukey was selected as it presented a better data fit compared with visual inspection of the data contained in Table 1. The age breakdown as identified in Table 1 was felt by the authors to present the data in a clinically useful method. There were significant differences among the seven age groups in all three Index Scores; Left [ F(6,329) = 19.87, P <.001], Right [ F(6,329) = 13.79, P <.001], and Both [ F(6,329) = 25.53, P <.001]. Table 2 summarizes the significant differences among the seven age groups. The post hoc comparisons revealed that Group 7 (70 79 year olds), performed significantly more poorly than all other age groups. Within the subsample between the ages of 16 and 69, performance on all three indices (Left, Right, and Both) was significantly negatively correlated with age: Left r =.23, P <.001, Right r =.31, P <.001, Both r =.38, P <.001. A comparison using a paired samples t test showed that for the participants age 16 to 39 there was a right ear advantage [t(1,190) = 2.54, P =.012]; however, this advantage was not observed for the age 40 and above participants [t(1,129) = 1.04, P =.30].

6 84 J.E. Meyers et al. / Archives of Clinical Neuropsychology 17 (2002) Table 2 Results of Tukey post hoc test of difference between the six study groups, ages years Groups 1 (16 19) 2 (20 29) 3 (30 39) 4 (40 49) 5 (50 59) 6 (60 69) Left ,2,3,4 Right 4,5,6 5,6 5,6 1,2,3 1,2,3 Both 4,5,6 4,5,6 4,5,6 1,2,3 1,2,3 1,2,3 All P s <.015. Group 7 was different from all other groups. Based on these findings, cutoff scores for defective performance (5th percentile) were developed for each age group and are indicated in Table 1. The 5th percentile was selected as this is a common standard used in the Benton Neuropsychology Laboratory (Benton, Hamsher, Varney, & Spreen, 1983) and has previously yielded coherent results when used by Grote, Pierre-Louis, Smith, Roberts, and Varney (1995) Discussion The findings of the first experiment were similar to the findings of Roberts et al. (1994) in that gender and handedness were not significant variables for the DWLT data. Unlike their study, it was found that age was a significant variable in DWLT performance, presumably because of the inclusion of older participants in the present study. An expanded normative sample is particularly important if the DWLT is to be used in studying the neurocognitive effects of disorders common among older persons. From visual inspection of the data, there appears to be little question that DWLT performance declines with advancing age, justifying the use of age-stratified norms. Beginning around age 40, there was a slight decline in mean task performance relative to that of younger participants, but thereafter, the mean performance in the year old group declines more sharply, and the 60-year-old and older group showed even lower mean performance. By the seventh decade, cutoffs for defective performance require that greater than 50% of items be failed for a patient to be labeled impaired. The division of the current age groups appears to be justified, as there are few marked differences in mean performance among adjacent age groups. Specifically, it can be observed from Tables 1 and 2 that there are only small differences in performance for ages 16 to 39, 40 59, and 60 and above groups. For younger participants in the standardization sample, there was a modest but consistent right channel preference, but this mean right ear advantage for processing of verbal material virtually disappeared after age 40. These findings are significant in that Roberts et al. (1994) also found no age effect in their younger sample of subjects. The diminution and ultimate disappearance of the right ear advantage for dichotic processing of verbal stimuli were consistent with previous findings by Koenig (1957). Koenig (1957) found that the older adults ability to detect small changes in pitch, a skill important for understanding speech, may begin to diminish as early as the fourth decade.

7 2. Experiment Method J.E. Meyers et al. / Archives of Clinical Neuropsychology 17 (2002) Participants A large database of clinical patients who had received DWLT as part of a neuropsychological battery was examined. Participants who had an identified traumatic head injury (TBI) that resulted in the need for medical care were selected. Only individuals who were not involved in litigation or disability proceedings at the time of the assessment were included. These individuals were referred due to continuing cognitive complaints and were seen for assessment to identify treatment needs. Groups were then separated based on the duration of unconsciousness (LOC). LOC was defined as the time from initial injury until the capacity to follow a simple command (i.e. squeeze hand, open eyes, and stick out tongue). The schema for operationally defining mild, moderate, and severe traumatic brain injury has been previously used by Dikman, Machamer, Winn, and Temkin (1995) and by Volbrecht, Meyers, and Kaster-Bundgaard (2000). The Mild TBI (MiTBI) group consisted of 40 participants with less than an hour of LOC, as documented by witnesses or medical personnel. The mean age for this group was 27.6 years (S.D. = 10.2), with 12.9 years of education (S.D. = 2.3). Participants were evaluated 22.8 (mean) months postinjury (S.D. = 47.4). Demographically, 14 were female and 26 were male; all were right handed and all were Caucasian. The Moderate TBI (ModTBI) group consisted of 28 individuals with 1 7 days of LOC documented in their medical records. The mean age for this group was 26.4 years (S.D. = 10.2), with 13.0 years of education (S.D. = 2.0). They were evaluated 22.9 months (mean) postinjury (S.D. = 74.1) and had a mean LOC of 3.8 days (S.D. = 1.8). There were 9 females and 19 males; 24 were right handed and 4 were left handed; 25 were Caucasian, 1 was African American, and 2 were of mixed ethnic background. The Severe TBI (SevTBI) group consisted of 28 individuals with 8 days or more LOC documented in their medical records. The mean age for this group was 29.2 years (S.D. = 13.4) with 11.5 years of education (S.D. = 1.8). They were evaluated 67.9 months (mean) postinjury (S.D. = 86.6) and had a mean of 30.0 days of LOC (S.D. = 26.1). There were 14 females and 14 males; 18 were right handed and 10 were left handed; 27 were Caucasian and 1 was Native American. In addition, the performance of two stroke groups, left hemisphere stroke (LCVA) and right hemisphere stroke (RCVA), were also examined. The LCVA group consisted of 27 individuals with identified (based on CT/MRI) strokes involving only the left hemisphere of the brain. The demographic makeup of this group was: 14 female and 13 male; 26 were right handed and 1 was left handed; all were Caucasian. The mean age was 55.8 years (S.D. = 15.9) with a mean of 12.9 years of education (S.D. = 2.5). They were evaluated 14.9 (mean) months postinjury (S.D. = 22.7). The RCVA group consisted of 17 individuals with identified (based on CT/MRI) strokes involving only the right hemisphere of the brain. The demographic makeup of this group was: 6 female and 11 male; 14 were right handed and 3 were left handed; all were Caucasian. The mean age was 59.5 years (S.D. = 15.5) with a mean of 13.1

8 86 J.E. Meyers et al. / Archives of Clinical Neuropsychology 17 (2002) years of education (S.D. = 3.9). They were evaluated 9.2 months (mean) postinjury (S.D. = 14.0). An additional group of Control participants were also collected as part of the larger data collection described in Experiment 1. This group consisted of 32 normal community dwelling individuals who had no history of neurological disease, depression, or special educational intervention for learning difficulties. The demographic makeup of this group was: 15 female and 17 male all were right handed; 31 were Caucasian, and 1 was of mixed racial background. The mean age for this group was 36.9 years (S.D. = 19.7) with 13.7 (mean) years of education (S.D. = 3.4). The control group subjects had also been included in the normative sample identified in Experiment 1, but these normal community volunteers had also undergone an extensive neuropsychological battery, similar to that administered to the clinical patients in Experiment Procedure All participants were administered the same general 3 h battery of neuropsychological tests (Volbrecht et al., 2000) by a licensed psychologist, graduate student, or masters level technician Results The means and standard deviations for various demographic factors in the TBI clinical groupings based on severity of injury (Controls, MiTBI, ModTBI, and SevTBI) are presented in Table 3. It can be observed that the more severe the TBI (i.e. longer LOC) the more impaired is the DWLT performance. An analysis of variance and multiple Tukey follow-up comparisons revealed a significant effect of age for the three TBI groups and the Control group. They differed in age [ F(3,124) = 3.68, P =.0l] and education [ F(3,124,) = 3.93, P =.01], and in DWLT performance for Left [ F(3,124) = 17.83, P <.001], Right [ F(3,124) = 9.72, P <.001], and Both [ F(3,124) = 27.62, P <.001] indices. The three injured groups all had a significantly younger mean age and less education relative to that of the normal comparison control group. However, as indicated in Experiment 1, education was not a significant influencing factor in DWLT performance ( P >.05). Table 3 Comparison groups for clinical sample groups with means and (standard deviations) Group FSIQ #Left #Right #Both TT MilTBI 96.3 (11.1) 23.1 (4.5) 26.7 (2.5) 20.8 (5.0) ModTBI 91.7 (13.2) 21.6 (6.2) 25.1 (4.2) 19.1 (6.4) SevTBI 80.8 (14.8) 16.5 (9.3) 22.0 (7.8) 12.2 (8.4) LCVA 87.8 (14.5) 19.6 (7.4) 19.8 (9.0) 12.8 (9.4) (20.7) RCVA 89.7 (13.0) 9.4 (7.8) 23.9 (5.6) 6.2 (6.5) (9.8) Controls (13.7) 27.5 (1.8) 27.8 (1.7) 25.8 (2.6) MilTBI = mild traumatic brain injury; ModTBI = moderate traumatic brain injury; SevTBI = severe traumatic brain injury; LCVA = left cerebral vascular accident; RCVA = right cerebral vascular accident; FSIQ = WAIS-R, III Full Scale IQ Score; TT = token test (Spreen and Strauss 1998); Controls = normal control participants.

9 J.E. Meyers et al. / Archives of Clinical Neuropsychology 17 (2002) Comparison of the three TBI groups revealed that they did not differ on age [ F(2,93) =.43, P =.64], but did differ on education [ F(2,93) = 4.63, P =.01]; the SevTBI group having less education and the ModTBI having the highest educational level. However, the Moderate TBI group did not have better performance on DWLT than did the MiTBI group. The time postinjury at the time of assessment was also different among the groups [ F(2,93] = 4.27, P =.01). However, a longer recovery time did not produce improved scores. The left and right ear scores on DWLT were not significantly correlated with time postinjury ( P >.05). DWLT Both scores did correlate (r =.21, P =.03) with time postinjury; however, when the data were corrected for age, the correlation between the Both score and time postinjury was not significant ( P >.05). Therefore, any effect of time postinjury was not a significant factor in performance on the DWLT. Performance on the DWLT variables also was significantly different among the groups: Left [ F(3,93) = 8.37, P <.001], Right [ F(3,93) = 7.15, P =.001], and Both indices [ F(3,93) = 14.90, P <.001]. A consistent finding regarding the TBI groups was that longer durations of LOC were associated with poorer performance on the DWLT variables. The DWLT score for the Left ear (r =.47), Right ear (r =.47), and Both ears (r =.58) were significantly correlated ( P >.05) with LOC. Using the 5th percentile cutoffs for each age group, no control subject failed any DWLT variable. In the MiTBI sample 20/40 (50%) patients failed the Left Index, 5/40 (13%) failed the Right Index, and 19/40 (48%) failed the Both Index score. The ModTBI group failed the Left 15/28 (54%), Right 11/28 (39%), and Both 16/28 (57%) indices. The SevTBI group failed Left 19/28 (68%), Right 11/28 (39%), and Both 23/28 (82%) indices. Another way to examine the data was to determine the number of participants who failed any DWLT index. This would represent the hit rate for the DWLT as a whole. In the MilTBI group, 24/40 (60%) failed on any one score; ModTBI group 18/28 (64%) and in the SevTBI group 25/28 (89%) performed defectively on at least one score. Therefore, 67/96 failed some aspect of DWLT performance, yielding an overall hit rate of 70% in a sample of closed head injury patients. The adequate DWLT performances of the controls yielded a specificity of 100%, as none of the controls failed any DWLT index. Mean level of Token Test performance (Spreen & Strauss, 1998) was not significantly different between the RCVA and LCVA groups. This indicated that at the point in the recovery period when they were evaluated, most of the LCVA patients had presumably adequate aural comprehension of verbal instructions. In addition, all LCVA patients manifested adequate single-word repetition on the monaural DWLT practice items. Although LCVA subjects were tested at a slightly longer mean time interval following the acute stroke (months: LCVA = S.D. = 22.67; RCVA = 9.18, S.D. = 14.0), this difference was not significant [t(40) =.91, P >.05]; therefore, it seems unlikely that the mean difference in lateralized performances between the two CVA groups could be attributed to differential, generalized, cognitive recovery. Within the LCVA sample, there was a significant correlation between Token Test performance and the overall dichotic performance measure, Both r =.56, P =.004. In the LCVA group 10/27 (37%) failed the Left Index, 11/27 (40%) failed the Right Index, and 15/27 (55%) failed the Both Index. The number of individuals who failed any score on DWLT was 15/27 (55%). In the RCVA group 14/17 (82%) failed in the Left, 5/17 (29%) in

10 88 J.E. Meyers et al. / Archives of Clinical Neuropsychology 17 (2002) the Right, and 15/17 (88%) failed in the Both Indices. The number of individuals who failed in any one score on DWLT was 15/17 (88%). 3. General discussion Considered as a whole, the findings from the present investigation provide general support for the proposition that dichotic listening tasks are likely to be useful adjuncts to more traditional measures in the neuropsychological assessment of adults (cf. Richardson et al., 1994). More specifically, with expanded norms, the DWLT has a wide degree of applicability across the adult life span. The present results confirmed that the DWLT is a relatively nondemanding and straightforward task for neurologically normal adults to perform, replicating the conclusions of Roberts et al. (1994). Before age 40, healthy adults in the present study manifested the expected mean right ear advantage for verbal auditory processing. Although the absolute size of this processing preference was relatively small, it was statistically significant. After age 40, however, there was very little difference in performance between the right and left channels for processing the simple words used in the DWLT. This would appear to be associated with the normal aging process. With regard to brain dysfunction due to focal lesions, the study found that DWLT performance in the left channel was extremely sensitive to the presence of structural lesions in the contralateral right hemisphere; however, DWLT performance in the ipsilateral left channel was most often within normal limits. On the other hand, with regard to left hemisphere strokes, the present findings are fairly consistent with previous research that indicates that performance of either or both the right or left channels may be adversely affected depending on the placement of the lesion, consistent with the paradoxical effects described earlier. Recently, Grote et al. (1995) also demonstrated similar variability in the laterality of unilateral ear suppressions in the context of focal partial seizure syndromes with clear left hemisphere foci. Perhaps even more interestingly, Lee et al. (1994), who also studied patterns of dichotic listening failure in the context of focal partial epilepsy, concluded that these patients need not have a clearly imageable lesion on CT or MRI to perform defectively on a similar dichotic task. Their data supported previous speculation that electrophysiological dysfunction (in the absence of a visualized structural lesion) may be sufficient to cause a dichotic listening suppression (Levin et al., 1989). Concerning the relatively high frequencies of DWLT failure in the context of closed head trauma (including mild cases), several neurobehavioral mechanisms may be operating. First, as is true with more focal disease processes such as strokes, tumors, and plaque formation due to multiple sclerosis, a strategically placed structural lesion associated with head trauma could directly disrupt or impinge upon primary pathways in the auditory processing system. Second, Levin et al. (1989) have proposed that diffuse white matter injuries could be responsible for dichotic listening failure following more severe traumatic brain injuries. Third, Roberts et al. (1990) and Springer et al. (1991) have suggested that, even in the absence of traditional epileptic syndromes or structural lesions on CT or MRI, subcortical neural noise due to underlying electrophysiological dysfunction associated with multiple, partial seizure-like, episodic symptoms could impair dichotic listening

11 J.E. Meyers et al. / Archives of Clinical Neuropsychology 17 (2002) performance. Finally, previous findings with both adults and children manifesting persistent, neurobehavioral dysfunction following so-called minor head injuries are consistent with this electrophysiologic conceptualization (Verduyn et al., 1992). The surviving brain acts in concert as a whole, the summative effects of various combinations of the above mechanisms may account for the abysmal performance of the severely injured group in the present study. To use a crude analogy to a telephone wiring system, whether the phone line is dissected, extremely frayed, or full of static, the phone call may not go through or the message may not be properly decoded at the receiver s end of the line, no matter how clearly the signal was transmitted at the point of origin. To summarize, the good news from the present findings is that the DWLT may well be an extremely useful, cheap, and diagnostically efficient adjunct to more traditional forms of clinical neuropsychological assessment. Several factors contribute to this conclusion. First, the DWLT has a larger normative base and is therefore applicable to a larger clinical population than previous dichotic tasks that were developed for purely experimental purposes. Second, DWLT calls upon the patient to process more than one channel of information in real time in order to perform successfully and is likely to be sensitive to transient as well as static cognitive impairment. The more complex news from the present discussion is that there appears to be multiple routes to DWLT failure after brain damage or disease. Different patients may fail DWLT for different reasons depending on the neuropathologic process involved or defect of linguistic competence. Hence, the clinician will be called upon: (1) to think critically about a variety of mechanisms that could underlie DWLT failure in a given patient; and (2) to correlate impaired DWLT performances with other neurobehavioral findings and sources of ancillary neurodiagnostic information in order to draw valid conclusions regarding the specific cause(s) of task failure. References Auditec of St. Louis. (1991). Dichotic word listening test (DWLT). St. Louis, MO: Auditec. Benton, A., Hamsher, K., Varney, N., & Spreen, O. (1983). Contributions to neuropsychological assessment: a clinical manual. New York: Oxford Univ. Press. Bouma, A., & Ansink, B. (1988). Different mechanisms of ipsilateral and contralateral ear extinction in aphasia. Journal of Clinical and Experimental Neuropsychology, 10, Damasio, H., & Damasio, A. (1979). Paradoxic ear extinction in dichotic listening: possible anatomic significance. Neurology, 29, Dikman, S. F., Machamer, J. E., Winn, H. R., & Temkin, N. R. (1995). Neuropsychological outcome at one year post head injury. Neuropsychology, 9, Grote, C., Pierre-Louis, S., Smith, M., Roberts, R., & Varney, N. (1995). Significance of unilateral ear extinction on the Dichotic listening test. Journal of Clinical and Experimental Neuropsychology, 17 (1), 1 8. Koenig, E. (1957). Pitch discrimination and age. Acta Oto-Laryngology, 48, Lee, G., Loring, D., Varney, N., Roberts, R., Newell, J., Martin, J., Smith, J., King, D., Meador, K., & Murro, A. (1994). Do dichotic word listening asymmetries predict side of temporal lobe seizure onset? Epilepsy Research, 19, Levin, H., High, W., Williams, D., Eisenberg, H., Amparo, E., Guinto, F., & Ewert, J. (1989). Dichotic listening and manual performance in relation to magnetic resonance imaging after closed head injury. Journal of Neurology, Neurosurgery and Psychiatry, 52,

12 90 J.E. Meyers et al. / Archives of Clinical Neuropsychology 17 (2002) Mohr, E., Cox, C., Williams, J., Chase, T., & Fedio, P. (1990). Impairment of central auditory function in Alzheimer s disease. Journal of Clinical and Experimental Neuropsychology, 12, Rao, S., Bernadin, L., Leo, G., Ellington, L., Ryan, S., & Bung, L. (1989). Cerebral disconnection in multiple sclerosis: relationship to atrophy of the corpus callosum. Archives of Neurology, 46, Richardson, E., Springer, J., Varney, N., Struchen, M., & Roberts, R. (1994). Dichotic listening in the clinic: new neuropsychological applications. Clinical Neuropsychologist, 8 (4), Roberts, M., Persinger, M., Grote, C., Evertowski, L., Springer, J., Tuten, T., Moulden, D., Franzer, K., Roberts, C., & Baglio, C. (1994). The Dichotic word listening test: preliminary observations in American and Canadian samples. Applied Neuropsychology, 1, Roberts, R., Varney, N., Paulsen, J., & Richardson, E. (1990). Dichotic listening and complex seizures. Journal of Clinical and Experimental Neuropsychology, 12 (4), Rubens, A. B., Froehling, B., Slater, G., & Anderson, D. (1985). Left ear suppression on verbal dichotic tests in patients with multiple sclerosis. Annals of Neurology, 18 (4), Rubens, A. B., Johnson, M. G., & Speaks, C. (1978). Location of lesions responsible for the paradoxical ipsilateral ear effect with dichotic listening tests in patients with aphasia due to stroke. (Paper presented at the American Academy of Neurology) Neurology, 28, 396. Sparks, R., Goodglass, H., & Nickel, B. (1970). Ipsilateral vs. contralateral extinction in dichotic listening resulting from hemisphere lesions. Cortex, 6, Spreen, S., & Strauss, E. (1998). A compendium of neuropsychological tests: administration, norms, and commentary (2nd ed.). New York: Oxford Univ. Press. Springer, J., Garvey, M., Varney, N., & Roberts, R. (1991). Dichotic listening failure in dysphoric neuropsychiatric patients who endorse multiple seizure-like symptoms. Journal of Nervous and Mental Disorders, 179 (8), Verduyn, W., Hilt, J., Roberts, M., & Roberts, R. (1992). Multiple partial seizure-like symptoms following minor closed head injury. Brain Injury, 6, Volbrecht, M., Meyers, J., & Kaster-Bundgaard, J. (2000). Neuropsychological outcome of head injury using a short battery. Archives of Clinical Neuropsychology, 15 (3),