Patient-Reported Auditory Functions After Stroke of the Central Auditory Pathway
Background and Purpose—Auditory functional limitations experienced by patients after stroke of the central auditory pathways remain underinvestigated.
Purpose—To measure patient-reported hearing difficulties in everyday life in nonaphasic patients with stroke of the auditory brain versus normal control subjects. To examine how hearing difficulties correlate with auditory tests and site of lesion in individual cases.
Methods—We recruited 21 individuals with auditory brain stroke (excluding those with aphasia) diagnosed on the basis of a brain MRI conducted 1 to 2 weeks after the stroke and assessed in the chronic stage of stroke. Twenty-three controls matched for age and hearing were also recruited. All subjects completed the Amsterdam Inventory for Auditory Disability (consisting of subscales of sound detection, recognition, localization, speech in quiet, speech in noise) and underwent baseline audiometry and central auditory processing tests (dichotic digits, frequency and duration patterns, gaps in noise).
Results—Sound recognition and localization subscores of the inventory were significantly worse in case subjects versus control subjects, with severe and significant functional limitation (z score >3) reported by 9 out of 21 case subjects. None of the inventory subscales correlated with audiometric thresholds, but localization and recognition subscales showed a moderate to strong correlation with dichotic digits (left ear) and pattern tests.
Conclusions—A substantial proportion of patients may experience and report severe auditory functional limitations not limited to speech sounds after stroke of the auditory brain. A hearing questionnaire may help identify patients who require more extensive assessment to inform rehabilitation plans.
Hearing is a complex function underpinned by analysis of sounds in temporal, spectral, and spatial domains. Anatomically, this requires transmission of the auditory signal from the ear to the auditory cortex and further processing to facilitate sound perception and recognition, attention, memory, and learning, which are all integral components of auditory cognition.1 Hearing impairment has a profound effect on an individual's ability to function at a personal, social, and professional level.2
Stroke is the commonest neurological disorder and may cause both physical and cognitive impairment. Disruption of hearing attributable to stroke pathology within the auditory pathways is a largely unexplored aspect of poststroke impairment. Although acute loss of hearing acuity attributable to stroke pathology is unusual,1 elderly stroke patients have an increased risk of hearing loss compared with the general population.3 Furthermore, stroke may also result in disordered auditory processing.4 Forty-nine percent of patients with unilateral cerebrovascular auditory structures lesions report auditory perceptual problems with sound localization or in situations involving simultaneous speakers, with the majority of these reporting intact hearing to the treating medical team unless specific detailed questions were asked.5 A screening study of 51 acute stroke unit patients found that 41% of those failed the hearing screening and reported that 81% of those who failed would have remained unidentified without screening.6
Hearing facilitates good communication between patients and health professionals, which is essential to deliver appropriate care.7 Hearing loss increases the risk (OR, 1.83) of physical decline of stroke patients after discharge to the community, because it may restrict the patient's participation in physical rehabilitation programs.8
Despite their high prevalence and potentially important functional impact, auditory processing deficits attributable to stroke of the central auditory pathway remain unrecognized and underinvestigated. Implementation of language-related training activities9 with a high cognitive and linguistic load10 or even simply listening to music or audiotapes11 after stroke may lead to broad functional improvements in communication, but improvements may not generalize to untrained auditory stimuli or cognitive tasks.12,13 Moreover, everyday life communication is more demanding than laboratory tests.14 Thus, as well as identifying auditory processing deficits after stroke, there is a need to assess the patient-reported difficulties in a range of everyday life situations, including persistent nonspeech-based difficulties with recognition and localization of sounds.
The aims of our study were to measure patient-reported hearing difficulties in everyday listening situations in nonaphasic patients with stroke of central auditory pathways compared with normal control subjects who were age-matched and hearing-matched, and to examine how questionnaire-reported hearing deficits correlate with results of auditory processing tests, audiometric thresholds, and site of the lesion in individual cases.
Materials and Methods
Ethics approval was obtained. All subjects gave written informed consent. All procedures adhered to institutional guidelines.
Consecutive patients with an acute history of ischemic or hemorrhagic cerebral stroke affecting the auditory brain who had been admitted to the Acute Stroke and Brain Injury Unit at the National Hospital for Neurology and Neurosurgery were identified on the basis of their brain MRI, recruited, and tested 4 weeks to 12 months after their stroke. Inclusion criteria for case subjects were acute stroke of the central auditory pathway (±adjacent areas) on MRI and audiometric thresholds better than 40 dBHL at 1 kHz. Exclusion criteria were the presence of aphasia, dementia, and psychiatric disorders. Normal control subjects were matched at group level to the case groups for gender, handedness, and average audiometric thresholds. All subjects underwent baseline audiometric tests, clinical central auditory test battery, Amsterdam Inventory for Auditory Disability, and brain MRI.
Baseline audiometric tests, including tympanometry to assess middle ear function and pure-tone audiometry (PTA) with a calibrated GSI 61 audiometer, to assess hearing thresholds were performed. Outcome measures for the pure-tone audiometry were right and left ear averages at 250, 500, 1000, 2000, 4000, and 8000 Hz for comparisons in the 2 groups, and the better ear threshold for comparison with questionnaire scores.
A clinical central auditory test battery, as recommended by the American Academy of Audiology,15 of tests that are minimally affected by elevated hearing thresholds was performed. These included the dichotic digits, frequency pattern and duration pattern, and gaps in noise tests (online-only Supplemental Methods, http://stroke.ahajournals.org). Outcome measures were the proportion of abnormal results compared with normative data for each test to compare results in cases versus controls and separate left and right ear dichotic digits scores (because these reflect ear performance in competition), better ear score for duration pattern and frequency pattern, and better ear threshold for gaps in noise to assess correlation between test results and questionnaire scores.
The Amsterdam Inventory for Auditory Disability was performed.16 This is a 28-item patient-reported questionnaire that assesses everyday listening ability with 5 main subscales: detection, distinction/recognition, localization of sounds, intelligibility of speech in noise, and intelligibility of speech in quiet. Answers range from “almost never” (score of 3 points) to “almost always” (0 points), scored on a 4-point scale, with a higher score denoting higher disability The questionnaire was administered to subjects before testing. Outcome measures were subscores calculated as the sum of scores for questions answered divided by their number for each subscale.
Case subjects but not control subjects underwent brain MRI conducted 1 to 2 weeks after the clinical presentation of the original neurological lesion and evaluated by an experienced neuroradiologist (J.S.). MRI were performed on a Sigma 1.5-T system (Echo Speed +; General Electric). All subjects underwent an axial T2-weighted fast spin echo sequence (repetition time/echo time=6000/102 ms), an axial T2*-weighted gradient echo sequence (repetition time/echo time=300/40 ms), and a coronal FLAIR sequence (repetition time/inversion time/echo time=9895/2473/140 ms) as per the vascular MRI protocol. The precise anatomic location of each lesion was reported by an author (J.S., who was blinded to the auditory test and questionnaire results).
Definition of the Central Auditory Pathway
The central auditory nervous system, consisting of cortical and subcortical structures and interhemispheric connections, was defined on the basis of current scientific understanding4,17,18 (see online-only Supplemental Methods).
The statistical package for the social sciences (SPSS 17) was used. After Bonferroni adjustment, P≤0.003 was deemed significant. Mann-Whitney U test was used to assess differences in descriptive variables. The CI for differences in means were reported. For subscales that showed significant differences, individual patient scores were converted into z scores relative to the mean and the SD of the controls to assess results versus lesion on a case-by-case basis, with a z score of ≥3 deemed as significant.
The χ2 tests were conducted to assess difference in proportions of abnormal test results. A Spearman rho correlation analysis was performed to assess correlations between the questionnaire's 5 subscores and test results. Correlations were classified as moderate (0.4<Spearman rho <0.6) and strong (rho≥0.6) for P<0.003.
We recruited 44 subjects (21 case subjects and 23 control subjects). Age, sex, and pure tone averages were not significantly different in case subjects versus control subjects (Table 1). The lesion description of case subjects is provided in Table 2.
Case subjects had a higher proportion of abnormal results than control subjects in all the auditory processing tests. After Bonferroni adjustment, this remained significant for dichotic digits, frequency pattern tests, and duration pattern tests (P<0.003; online-only Supplemental Table I).
The 5 Amsterdam Inventory subscores were higher (indicating more disability) in case subjects than in control subjects. After Bonferroni correction, this remained significant for sound recognition and localization (Table 3). The z scores for sound recognition and localization subscales are shown in Table 4.
Correlation of Amsterdam Inventory Subscores and Central Auditory Test Better Ear Results
The Spearman rho correlation matrix (online-only Supplemental Table II) showed strong (rho≥0.6; P<0.001) correlations between all questionnaire subscales and the left ear dichotic digits and localization.
Correlations were moderate (rho>0.4; P<0.003) between the left ear dichotic digits score and sound detection, recognition, and speech in quiet. Correlation was also moderate with speech in noise, albeit at P=0.005. They were also moderate between the frequency pattern better ear score and sound recognition and localization, and between the duration pattern better ear score and localization.
This is one of few studies to address patient-reported auditory function in the chronic phase after stroke. We found that patients with stroke affecting the central auditory pathways reported significantly greater functional difficulties in everyday tasks that require recognition and localization of sound than control subjects: 9 out of 21 case subjects (43%) reported significant limitations, as judged by z scores. Deficits in behavioral tests in these 2 key domains of sound processing have been reported in the chronic stage after stroke.4 Such deficits are attributed to stroke-related damage of overlapping but largely separate auditory processing networks that consist at the cortical level of the “what” and the “where” auditory streams,19 with a similar functional organization postulated at the subcortical brain stem level.20
Visual inspection of individual case results did not show any obvious relation between extent of the lesion and z scores in sound recognition/localization: in 3 out of 9 cases with high z scores, the lesion was subcortical only. Others7 similarly report an absence of simple relation between extent of lesion and severity of behavioral test deficits.
On a single case study basis, there were some interesting correlations between abnormal findings in the 2 subscales (Table 4) and site of lesion (Table 2). Case subjects 3 and 5 both had localization deficits in the presence of right insular involvement. This is consistent with previous findings of prominent sound movement detection deficits in a stroke patient with a right insula lesion.21 Case subjects 8 and 14 had a lesion in the (left and right, respectively) posterior limb of the internal capsule and reported recognition and localization difficulties, indicating that this structure may be necessary for both functions. Of interest, disuse-related anisotropy of the internal capsule has been reported in early deafness.22 Case subject 10, with a lesion of the right brain stem, including the superior olive, and cerebellum, had a high z score for localization. Within the human brain stem, the superior olivary complex compares binaural sound cues (interaural intensity and time differences) and is an integral component for the computation of auditory space.23 Localization of sounds may also depend on integration of audiovisual cues, with combined audiovisual motion leading to functional MRI activation clusters in the cerebellum.24 Case subject 20, who had a right medial frontal gyrus stroke, reported both recognition and localization difficulties. Westerhausen et al25 identified significant interaction of functional MRI activation in this structure by interaural intensity difference and attention demand, whereas it has been proposed that the right frontal cortex is an important working memory station for both “what” and “where” auditory functions.26
In contrast to previous studies,5 we did not find a significant difference between patient reported difficulties with speech in quiet or in noise. This may be because of the exclusion of aphasics from our study, or because of false-negative type errors attributable to the Bonferroni adjustment. Patient-reported disability did not correlate with better ear pure-tone average. The original validation study of the questionnaire similarly found a poor correlation with thresholds.27 Our findings may be consistent with reports that in patients with auditory neglect after stroke, manipulation of the sound volume has no influence on the neglect, consistent with the hypothesis that sound volume is represented in the brain as encoded information rather than as a linear representation.28
The left ear dichotic digit test score was a good marker for difficulties with localization, sound detection, recognition, as well as speech. The dichotic digits relies in interhemispheric callosal transfer that is sensory-driven, context-dependent, and strongly modulated by attention, and is necessary for auditory scene analysis tasks such as figure/background separation and sound localization.29 The frequency pattern correlated with patient-reported difficulties with sound recognition. Some of the items of this subscale include recognition of melodies, distinguishing intonation and inflections in people's voices, and hearing rhythm in a song, which may well depend on the pitch pattern content and changes of the perceived sounds.30 Higher-order top-down effects also play a significant role in such functions.31 Both pattern tests correlated with localization. Both spectral and temporal cues are required for this function; however, localization of sounds in real life is a more complex process that may encompass multiple sensory input.32 The correlation of tests results with questionnaire subscales thus could be attributed to a range of auditory and cognitive factors.
Limitations of the study include lack of information regarding cognitive aspects such as memory and attention that might also have an impact on everyday life auditory function in the sample, lack of imaging data for the normal control subjects, and lack of a stroke control group with lesions not affecting the auditory pathway. Nevertheless, the findings of this questionnaire study are broadly consistent with either case or group studies that included psychophysical and imaging tests,4,21 and they provide evidence that some case subjects after stroke of the auditory brain report severe auditory functional limitation, not limited to speech, in everyday life. Although the study was limited to 44 subjects, it is the largest used for this topic to date and we corrected for the sample size using the Bonferroni adjustment.
Our findings have implications for the clinical management of stroke patients. Hearing loss may increase the risk of functional decline in older patients with stroke living in the community.8 There are no data about the effect of “central” hearing impairments on functional decline, but such impairments could limit the ability of the patient to reintegrate in the community and to participate in the rehabilitation programs after stroke. The neurological model posits that “auditory cognition” disorders require a detailed diagnostic interview and individually tailored testing.1 Exhaustive testing of every single patient may be time-consuming and impractical; however, our preliminary results suggest that a minimum assessment with some basic tests (eg, pure tone audiometry and dichotic digits testing) in addition to a questionnaire may identify those patients with high levels of disability that require additional investigation and input. Such an approach would facilitate identification of specific deficits and functional limitations that should be remediated by a range of approaches. Future studies should combine administration of a questionnaire in addition to psychoacoustic tests that are more specific to the patient-reported difficulties and comparison of questionnaire results (ideally on question basis) to imaging data, neurocognitive function, and psychoacoustic results. Further research is required to replicate and extend these findings. The sensitivity of any screening test for auditory processing deficits will need validation in larger stroke cohorts.
Sources of Funding
This study was funded by a small grant by the Hearing Conservation Council and Mercer's Charity.
Drs Bamiou and Werring are funded by a Higher Education Funding Council for England/Department of Health Clinical Senior Lecturer Award. Dr Brown's Chair in Stroke Medicine at University College London, is supported by the Reta Lila Weston Trust for Medical Research. Part of this work was undertaken at University College London Hospital/University College London, which received a proportion of funding from the Department of Health's National Institutes for Health Research Biomedical Research Centre's funding scheme.
The online-only Data Supplement is available with this article at http://stroke.ahajournals.org/lookup/suppl/doi:10.1161/STROKEAHA.111.644039/-/DC1.
- Received November 4, 2011.
- Accepted December 22, 2011.
- © 2012 American Heart Association, Inc.
- Rees A,
- Palmer AR
- Griffiths TD,
- Bamiou DE,
- Warren JD
Royal College of Physicians. Hearing and Balance Disorders. London, UK: Royal College of Physicians; 2008.
- Gopinath B,
- Schneider J,
- Rochtchina E,
- Leeder SR,
- Mitchell P
- Blaettner U,
- Scherg M,
- von Cramon D
- Edwards DF,
- Hahn MG,
- Baum CM,
- Perlmutter MS,
- Sheedy C,
- Dromerick AW
- Azouvi P,
- Samuel C,
- Louis-Dreyfus A,
- Bernati T,
- Bartolomeo P,
- Beis JM,
- et al
American Academy of Audiology. Clinical practice guidelines. Diagnosis, treatment, and management of children and adults with central auditory processing disorder. Available at: http://www.audiology.org/resources/documentlibrary/Documents/CAPD%20Guidelines%208-2010.pdf. August 2010. Accessed January 6, 2012.
- Roeser R,
- Valente M,
- Hosford-Dunn H
- Musiek FE,
- Oxholm VB
- Griffiths TD,
- Rees A,
- Witton C,
- Cross PM,
- Shakir RA,
- Green GGR
- Kim DJ,
- Park SY,
- Kim J,
- Lee DH,
- Park HJ
- Grothe B,
- Pecka M,
- McAlpine D
- Baumann O,
- Greenlee MW
- Alain C,
- McDonald KL,
- Kovacevic N,
- McIntosh AR