Cognitive Training for Adults With Bothersome Tinnitus: A Randomized Clinical Trial

Key Points

Question  Can patients with severe, bothersome tinnitus benefit from a computer-based, cognitive training rehabilitation program?

Findings  Of the 40 individuals with tinnitus enrolled in this randomized clinical trial, patients who used the training program endorsed improvements in domains of tinnitus perception, attention, memory, and concentration, but no significant differences in behavioral measures were observed between these patients and individuals who did not use the training program. Neuroimaging changes in brain systems responsible for attention and cognitive control were observed in patients who used the training program.

Meaning  Cognitive training programs, which purport to exploit the brain’s capacity for neuroplasticity, might have a role in the future treatment of patients with tinnitus.

Importance  Individuals with tinnitus have poorer working memory, slower processing speeds and reaction times, and deficiencies in selective attention, all of which interfere with readiness and performance. Brain Fitness Program–Tinnitus (BFP-T) is a cognitive training program specially designed to exploit neuroplasticity for preservation and expansion of cognitive health in adults with tinnitus.

Objective  To evaluate the effect of the BFP-T on tinnitus.

Design, Setting, and Participants  This open-label, intention-to-treat randomized clinical trial prescreened 191 patients with tinnitus and 64 healthy controls (HCs) from June 1, 2012, through October 31, 2013. Participants were 40 adults with bothersome tinnitus for more than 6 months and 20 age-matched HCs. Patients with tinnitus were randomized to a BFP-T or non–BFP-T control group. The BFP-T was completed online, and assessments were completed at Washington University School of Medicine.

Interventions  Participants in the intervention group were required to complete the BFP-T online 1 hour per day 5 days per week for 8 weeks. Tinnitus assessment, neuroimaging, and cognitive testing were completed at baseline and 8 weeks later. The HCs underwent neuroimaging and cognitive assessments.

Main Outcomes and Measures  The primary outcome measure was the change in Tinnitus Handicap Inventory (THI) score. Behavioral measures, neuroimaging, and cognitive tests were performed before and after the intervention.

Results  A total of 40 patients with tinnitus and 20 HCs participated in the study (median [range] age, 56 [35-64] years in the BFP-T group, 52 [24-64] years in the non–BFP-T group, and 50 [30-64] years in the HC group; 13 [65%] in the BFP-T group, 14 [70%] in the non–BFP-T group, and 13 [65%] in the HC group were males; and 16 [80%] in the BFP-T group, 16 [80%] in the non–BFP-T group, and 15 [75%] in the HC group were white). There was a reduction in the THI score in the BFP-T group (median, 7; range, −16 to 64) and non–BFP-T group (median, 11; range, −6 to 26), but this reduction was not significantly different between the 2 groups (median difference, 0; 95% CI, −10 to 8). There was no difference in cognitive test scores and other behavioral measures. There was a significant difference between baseline and follow-up in functional connectivity in cognitive control regions in the BFP-T group but not in HCs or individuals with untreated tinnitus. Of the 20 patients in the BFP-T group, 10 (50%) self-reported improvement attributable to the intervention, and 6 (30%) reported to be much improved in the domains of tinnitus, memory, attention, and concentration.

Conclusions and Relevance  These findings suggest that the computer-based cognitive training program is associated with self-reported changes in attention, memory, and perception of tinnitus. A possible mechanistic explanation for these changes could be neuroplastic changes in key brain systems involved in cognitive control. Cognitive training programs might have a role in the future treatment of patients with tinnitus.

Trial Registration  clinicaltrials.gov Identifier: NCT01458821

An estimated 14% of the firefighter community has tinnitus.^1,2 Firefighters and policemen have an increased risk of developing tinnitus because of exposure to hazardous levels of intermittent noise.³ Tinnitus may interfere with multiple facets of daily living, including sleep and activities dependent on seeing and hearing.⁴ Previous research⁵ revealed that individuals with tinnitus had deficits in verbal learning, auditory attention, and phonemic verbal fluency that depend on attention resources, indicating that tinnitus interferes with attention in some patients. Individuals with tinnitus have poorer working memory, slower processing speeds and reaction times, and deficiencies in selective attention.^6,7 All of these deficiencies may interfere with firefighter and first-responder readiness and performance.

Neuroplasticity refers to the brain’s ability, both after injury and during typical exogenous and endogenous experience, to change its functional and structural architecture. Neuroplasticity has been the foundation for the creation of several cognitive enhancement programs intended to slow normal aging,^8-10 improve internal locus of control,¹¹ and potentially ameliorate disorders, such as attention deficits,¹² dyslexia,¹³ stroke,¹⁴ traumatic brain injury,¹⁵ schizophrenia,¹⁶ mild cognitive impairment,¹⁷ and Alzheimer disease.¹⁸ Brain Fitness Program–Tinnitus (BFP-T) is a cognitive training program specially designed by Posit Science to exploit neuroplasticity for preservation and expansion of cognitive health in adults with tinnitus. The objective of our study was to evaluate the effect of BFP-T on tinnitus. We hypothesized that BFP-T would also aid in the recovery of cognitive function affected by the tinnitus.

This is a single-institution, open-label, intention-to-treat randomized clinical trial of 40 patients with tinnitus and 20 healthy controls (HCs) recruited from June 1, 2012, through October 31, 2013. The trial protocol can be found in Supplement 1. The CONSORT flow diagram of study participation is shown in Figure 1. Men and women between the ages of 20 and 65 years were recruited from the greater St Louis, Missouri, area fire departments and the public at large. A total of 17 firefighters and first responders completed the study protocol. Because insufficient firefighters enrolled in the trial to meet recruitment goals, we used the Volunteers for Health at Washington University and solicited referrals via the Washington University otolaryngology clinic for the remainder of the study participants. The study was approved by Washington University’s Human Protection Research Office and the Federal Emergency Management Agency before the conduct of the study. All participants provided written informed consent, and data were deidentified at study closure.

The patients with tinnitus were required to have nonpulsatile, idiopathic, subjective tinnitus for a duration of 6 months or greater and report being extremely bothered, bothered a lot, or bothered more than a little but not a lot on the overall Global Bother Score assessment of tinnitus.¹⁹ Uninterrupted use of a computer to complete the BFP-T for approximately 1 hour per day 5 days per week for the 8-week duration of the study was required. To avoid the confounding effect of age in cognitive and neuroimaging analyses, HCs were age matched within 5 years to patients with tinnitus. All participants provided audiograms performed within 18 months of the start of the study.

Patients with a history of head trauma, brain lesions or brain surgery, or active depression or patients who had previously used a cognitive training program were excluded. In addition, we excluded patients with tinnitus if their tinnitus was related to cochlear implantation, retrocochlear lesion, Meniere disease, or other known anatomical lesions of the ear or temporal bone or if they had an active workers’ compensation claim or pending litigation-related event.

The Posit Science BFP, a cognitive training program that contains 6 interactive training exercises, has been used in a variety of illnesses and conditions.^8,16,20-23 The BFP-T is a proprietary modification of the BFP and contains 11 interactive training exercises (simple acoustic stimuli, continuous speech, and visual stimuli) in an attempt to address the attentional effect of tinnitus. Each exercise focused on 1 of the following: (1) auditory processing speed, (2) discriminating sounds, (3) sound precision, (4) sound sequencing, (5) working memory, and (6) narrative memory. Exercise measures were calibrated to individual performance at the onset of training and adapted in difficulty to maintain an 85% correct response rate and to provide constant progress feedback.

A permuted-block sequence was used for randomization. Study participants randomized to the BFP-T group were required to complete approximately 1 hour of uninterrupted training per day over 5 days per week for a total of 8 weeks. Participants who had not completed the program were sent reminders to do so. The study participants randomized to the non–BFP-T group did not perform any additional tasks.

Assessments were performed at the beginning of the study and at the end of 8 weeks. Information on tinnitus description and history and medical status was collected using the Oregon Hearing Research Center forms (http://www.tinnitusarchive.org/forms). Tinnitus Handicap Inventory (THI),^24,25 Tinnitus Functional Index (TFI),²⁶ and the overall Global Bother Score were used to assess severity and bother from tinnitus. All 3 measures of tinnitus bother are self-reported measures. Changes in scores between baseline and postintervention of 17 on the THI and 13 on the TFI were considered clinically meaningful differences. The Global Bother Score measures the global bother from tinnitus in an ordinal scale in response to the following request: “All persons have their own unique problems and attach different importance to these problems. Please indicate the overall amount of disturbance or ‘bother’ that you experience in your life as a result of your tinnitus.”

All participants completed the Patient Health Questionnaire 9²⁷ to evaluate depression and the Cognitive Failures Questionnaire²⁸ to measure difficulties with perception, memory, and motor function. Participants completed the Brief Symptom Inventory 18²⁹ to measure somatization, depression, and anxiety; the Whiteley-7 scale³⁰ to evaluate hypochondriac traits; and the Adult Attention Deficit Hyperactivity Disorder Self-report Scale, version 1.1³¹ to screen for symptoms of adult attention-deficit/hyperactivity disorder.

All participants completed several different cognitive tests both before and after intervention (eMaterial in Supplement 2).

All patients with tinnitus were asked at the end of the study to qualitatively report how much their tinnitus and their memory or attention had changed compared with the start of the study. In addition, patients were asked to rate the degree of change using a 7-point Likert scale.

Imaging was performed on a Siemens 3-T Tim Trio magnetic resonance imaging (MRI) scanner at Washington University. Resting-state functional connectivity MRI and anatomical images were collected during the same imaging session. A short description of the acquisition sequence is described here; a more complete description can be found elsewhere.^32,33 Three 164-frame (6 minutes; repetition time, 2200 milliseconds), blood oxygen level–dependent runs recorded spontaneous brain activity while participants were awake, performed no task, and remained with their eyes closed in a darkened room.

Standard blood oxygen level–dependent image preprocessing was performed as in the study by Wolf et al.³³ A more complete description is provided in the eMaterial in Supplement 2.

To calculate functional connectivity, we used 10-mm-diameter spherical regions—a group of 264 regions distributed across the entire brain³⁴ and a subset of 77 of these regions that represented 5 cognitive control systems (ie, cingular opercular, frontal parietal, salience, dorsal attention, and ventral attention). A time series of blood oxygen level–dependent signal intensity was calculated in each of the regions of interest for each participant, within-participant Pearson correlation coefficients were calculated between each pair of regions, and then the Fisher R-Z transformation was performed to ensure normal distribution of data.³⁵ Object oriented data analysis³⁶ was used to compare functional connectivity across the brain between days within a participant group and between-participant groups. Object oriented data analysis uses an iterative approach and comparison to the Gibbs distribution to assess the significance of differences found in a multiple dimensional approach.

Descriptive statistics were used to summarize the distribution of demographic and clinical characteristics, as well as the scores of the assessments for each study group. Because the continuous-level variables were not normally distributed, the nonparametric Kruskal-Wallis and Mann-Whitney tests were used to test for significant differences among groups. The χ² test or Fisher exact test was used to compare distribution of categorical-level variables.

The primary outcome measure for this study was the change before intervention to after intervention in the THI score. Using a 2-sided t test at an α level of .05, we estimated that 17 individuals per group would allow us to be able to detect, with 85% power, a difference of 17 points or more on the THI in the BFP-T group compared with the non–BFP-T control group. This sample size is reasonable for analysis of groupwise differences in neuroimaging data.

Differences between preintervention and postintervention outcome measures were calculated. The median of the difference and the 95% CI for the difference were calculated using the methods of Altman et al.³⁷ Frequency and relative frequency of patients reaching a clinically meaningful reduction in tinnitus bother as measured by a 17-point reduction in THI score or a 13-point reduction in TFI score were calculated and reported. A Proc Mixed procedure was used to test for significant pre-post changes in either of the outcome measures and to compare the changes between the randomization groups. All statistical tests were 2-sided and were evaluated at the α level of .05. Bonferroni correction was used when needed to adjust the α level for multiple comparisons. SPSS statistical software, version 20.0 (SPSS Inc) and SAS statistical software, version 9.3 were used for statistical analysis.

A total of 40 patients with tinnitus and 20 HCs (median [range] age, 56 [35-64] years in the BFP-T group, 52 [24-64] years in the non–BFP-T group, and 50 [30-64] years in the HC group; male, 13 [65%] in the BFP-T group, 14 [70%] in the non–BFP-T group, and 13 [65%] in the HC group; white, 16 [80%] in the BFP-T group, 16 [80%] in the non–BFP-T group, and 15 [75%] in the HC group) were recruited for the study. Seven patients randomized to the BFP-T study arm withdrew from the study for nonadherence with the program (n = 6) or for personal reasons (n = 1).

As indicated in Table 1, no differences were found in the distribution of main demographic characteristics between the groups. A total of 17 firefighters and first responders completed the study protocol. The number of firefighters in the HC group was higher (10 [50%]) than in the BFP-T group (2 [10%]) and the non–BFP-T group (5 [25%]). Most HC participants (14 [70%]) had no hearing loss, whereas in the groups of participants with tinnitus, most participants (18 [90%] in the BFP-T group and 16 [80%] in the non–BFP-T group) had sensorineural type hearing loss. Of the 20 participants in the BFP-T arm, adherence data were collected for 19 participants. Of these 19 participants, adherence was 86.9% of expected, which is equivalent to completing a mean of 34.8 training sessions daily of the 40 prescribed.

Participants with tinnitus randomized to the BFP-T group were similar to those randomized to the non–BFP-T group (Table 1). Changes in each of the behavioral measures and the THI, TFI, Brief Symptom Inventory 18, Patient Health Questionnaire 9, Whiteley-7 scale, Cognitive Failures Questionnaire, and Adult Attention Deficit Hyperactivity Disorder Self-report Scale scores during the 8 weeks of the intervention were calculated for all patients with tinnitus as preintervention measure minus postintervention measure (Table 2). The mixed within-between participant analyses revealed that, although there is significant reduction after intervention compared with before intervention, this reduction was not significantly different between the study groups (P = .51).

A clinically meaningful reduction of 17 points on the THI was observed in 7 patients (35%) randomized to the BFP-T group and 3 patients (15%) randomized to the non–BFP-T group for a difference of 20% (95% CI, −0.6% to 46%). A clinically meaningful reduction of at least 13 points on the TFI was observed in 6 patients (30%) randomized to the BFP-T group and 5 patients (25%) randomized to the non–BFP-T group for a difference of 5% (95% CI, −0.23% to 33%).

Distribution of pre-post intervention change in the scores of each of the neurocognitive tests used for each of the study groups is given in eTable 1 in Supplement 2.

There were 27 patients (13 in the BFP-T group and 14 in the non–BFP-T group) and 16 HCs who had enough quality frames of MRI data to be included in the neuroimaging analysis. Table 3 lists the Bonferroni-corrected P values for comparisons of differences on many groups of regions. At baseline, no significant difference was found between HCs and patients with tinnitus and between patients with tinnitus randomized to the BFP-T group and to the non–BFP-T group (P = .64) in functional connectivity across 264 regions distributed across the entire brain.

Comparing participants between baseline and 8-week follow-up (Table 3) reveals that, although there is no difference for the HCs (P = .19 for baseline and P = .21 for 8-week follow-up) and non–BFP-T tinnitus group (P = .63 for baseline and P = .84 for 8-week follow-up), there are statistically significant differences for the BFP-T group after the intervention for the control, cingular opercular and salience, and cingular opercular alone groups of regions. At the 8-week follow-up, the P values for the difference in connectivity between the non–BFP-T and BFP-T groups were .28 in the cingular opercular and salience group and .10 for the cingular opercular alone group. The eFigure in Supplement 2 shows that network connections are stronger after intervention for the patients in the BFP-T group but not for the patients in the non–BFP-T group or the HCs.

When asked, “Compared to the start of the study, how has your tinnitus changed?” 10 participants (50%) randomized to the BFP-T group reported improvement in their tinnitus compared with the start of the study (Figure 2). Most patients (15 [75%]) randomized to the non–BFP-T group reported no change. Compared with the start of the study, improvement in memory and attention was reported by most patients (14 [70%]) in the BFP-T group, whereas no one in the non–BFP-T group reported improvement, and most (19 [95%]) reported no change (Figure 2).

Testimonials of patients randomized to the BFP-T group were explored to better understand the effect of BFP-T on tinnitus and are presented in eTable 2 in Supplement 2. Of the 20 patients in the BFP-T group, 10 (50%) self-reported improvement because of the intervention, including 6 (30%) who reported being much improved. The improvements endorsed by the participants were mainly in the domains of tinnitus, memory, attention, and concentration. For each participant, there is wide variability in response to intervention as measured by the self-reported global response question, change in THI score, and change in TFI score.

Two patients in the BFP-T group and 3 patients in the non–BFP-T group reported changes in health status, and 5 reported medication changes during the study period, but none were associated with the study. Thirty-nine patients (98%) would recommend this program to a friend.

In this study, we found that patients with tinnitus who engaged in a computer-based cognitive training program endorsed improvements in tinnitus perception, memory, attention, and concentration compared with patients who did not participate. However, no difference was observed in any of the behavioral measures, which we believe may be in part attributable to the limitations of the existing tinnitus instruments. In addition, we found increased functional connectivity strength in cognitive control neural networks among the intervention group, particularly the cingular opercular network, which is thought to set and maintain task objectives.^38,39 Although this finding suggests a neuroplastic mechanism for the observed changes, we were unable to identify consistent changes in cognitive testing associated with the neuroimaging findings. On the basis of our broad recruitment and enrollment strategies, we believe the results of this study are applicable to most patients with tinnitus who seek medical attention.

The BFP exploits the mechanisms of neuroplasticity,^9,40 reflecting a change in connectional strength at the synapse and multisynaptic circuit levels. An increasing body of literature now supports the contention that our observed behaviors are a product of complex, multilevel recurrent networks rather than changes at the level of a single neuron.⁴¹ Therefore, the BFP focuses on improving the many deficits across the various system levels that contribute to the degradation of the neurologic representation of information that the patient is struggling to record. The program’s training tasks are constructed with highly salient training targets held in working memory.^8,40,42 Merzenich et al^9,40,42 studied the parametric dimensions of progressive stepwise change and the dose-response (repetitive stimulus trial) conditions required for progressively driving enduring change. The training strategies used in the cognitive training programs created at Posit Science are informed by these studies in an attempt to optimize achievable rates and magnitudes of learning-driven change. There is considerable debate^43,44 regarding the true benefit of cognitive training programs, including the generalizability of the trained task to the targeted domain, whether positive changes are noted in everyday activities, and duration and durability of the gains.

A previous double-blind, randomized clinical trial⁴⁵ explored the effect of a neuroplasticity-enhancing drug, d-cycloserine, to facilitate learning and tinnitus response after the use of a computer-assisted program for 34 patients with bothersome tinnitus. The d-cycloserine and placebo groups had a significant improvement in median TFI score and self-reported cognitive deficits after participation in the program. However, patients in the d-cycloserine group had a significantly greater improvement in self-reported cognitive deficits compared with the placebo group.

Previous research⁴⁶ found significant abnormalities in key cortical neural networks among patients with bothersome tinnitus. The previous study⁴⁶ also found significant changes in functional connectivity that appeared to be normalized within the frontal parietal and cingular opercular networks among a group of patients with tinnitus who completed an 8-week, mindfulness-based stress reduction program.⁴⁷ We were unable to identify abnormalities in these same networks among patients without bothersome tinnitus⁴⁸ and patients who had no change in tinnitus after repetitive transcranial magnetic stimulation to the tempoparietal junction.⁴⁹ In a previous study,⁵⁰ attention to motion artifact was not as strict as in this current study. Because inclusion of motion artifact induces differences between groups, caution should be exercised when interpreting the results of studies that do not remove higher-motion MRI frames. The key findings of significant change in functional connectivity that involves regions in the cingular opercular and salience networks is consistent with the hypothesis that neuroplasticity is the mechanism by which the cognitive training program appears to treat tinnitus and comorbidities.

The improvements endorsed by the participants in the current study are mainly reported for tinnitus, memory, attention, and concentration. Of interest, some patients who reported no change or minimal improvement changes in their tinnitus when asked to compare how the tinnitus currently feels with the start of the study reported to have learned a few strategies to ignore tinnitus. Reading our participants’ poststudy comments, we believe that we could have defined the questions and clarified the expectations better. For some patients, being able to ignore their tinnitus or learn how to better cope with tinnitus was defined as much improved, whereas for others, the same effect was defined as no change. Related to this ambiguity in the patients’ global response is the question of the ability of the THI and TFI to reliably and accurately capture the experience of patients with tinnitus. The failure of retrospective, patient-reported outcome measures to accurately capture the physical, functional, and emotional problems related to a particular condition, which has significant moment-to-moment variability, has been previously described.⁵¹ A former study⁵² found that these reports are prone to recall bias and errors in summarizing prior events and often emphasize the patient’s current state and environment as well as the most recent and extreme events.

With a 1-time assessment, the exact time of measurement may not reflect the typical burden of tinnitus symptoms. In fact, given the large amount of moment-to-moment variability in tinnitus,⁵³ it is unclear how to define what typical means. Thus, we believe that the assessment of the preintervention to postintervention tinnitus burden using the THI and TFI was a significant limitation in the present study (eTable 2 in Supplement 2). In addition, many participants’ qualitative descriptions of intervention effect were not corroborated by the patient-reported outcome measures. We believe the assessment of the effect of the cognitive training program is complicated by the failure to include outcome measures that accurately reflect the fluctuating and variable nature of tinnitus.

The generalizability of our findings might be limited because of the high percentage of firefighters and first responders in our sample. It is well known that firefighters’ working conditions affect cognition, memory, and attention; therefore, firefighters may derive more benefit from cognitive enhancement programs than the general population. Other limitations of the current research program include a large percentage of withdrawals from the intervention group, use of a new and never-tested modification of the BFP, failure to include an active control, and lack of participant masking to the intervention. The BFP-T was a modification of the BFP with the goal of addressing the persistent attentional effect of tinnitus. Numerous computer problems with the program were identified. These problems were eventually corrected, but the investigators suspect that they could have affected the efficacy of the intervention. We also believe that a longer intervention exposure time would result in greater behavioral and resting-state functional connectivity MRI changes. The identification of an adequate placebo for cognitive training rehabilitation programs is challenging.⁵⁴ The main challenge with selecting a rehabilitation training protocol as a control is that the various training protocols differ in so many ways that it is hard to ascribe the cause of training-induced differences between the groups to the training program. Thus, we used a wait list–type, no intervention control group program. Another limitation of this study was that the participants were not masked to the intervention. Thus, those participants randomized to the intervention group could have reported falsely elevated improvements because of expectation bias and belief in the efficacy of the BFP-T. We believe that future research into the role of brain training rehabilitation protocols for tinnitus should include an active control that reflects standard of care or the best current treatment.

Participants enrolled in this randomized clinical trial endorsed improvements in domains of tinnitus perception, attention, memory, and concentration. However, no changes in behavioral measures were observed between the 2 tinnitus study groups. Neuroimaging changes in brain systems responsible for attention and cognitive control were observed in patients who used the BFP-T. Several specific limitations of the trial prevent definitive conclusions about the role of this particular cognitive training rehabilitation program, or such programs in general, for the treatment of tinnitus. Nevertheless, we believe that continued research into the role of cognitive training rehabilitation programs is supported by the findings of this study, and the role of neuroplasticity seems to hold a prominent place in the future treatments for tinnitus.

Corresponding Author: Jay F. Piccirillo, MD, Department of Otolaryngology–Head and Neck Surgery, Washington University School of Medicine in St Louis, 660 S Euclid Ave, PO Box 8115, St Louis, MO 63110 (piccirilloj@ent.wustl.edu).

Accepted for Publication: October 8, 2016.

Published Online: January 19, 2017. doi:10.1001/jamaoto.2016.3779

Authors Contributions: Dr Piccirillo had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Kallogjeri, Piccirillo, Spitznagel, Nicklaus.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: Kallogjeri, Piccirillo, Spitznagel, Hardin, Coalson.

Critical revision of the manuscript for important intellectual content: Piccirillo, Spitznagel, Hale, Nicklaus, Shimony, Coalson, Schlaggar.

Obtained funding: Piccirillo.

Administrative, technical, or material support: Nicklaus, Hardin, Shimony.

Study supervision: Piccirillo, Nicklaus, Schlaggar.

Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest and none were reported.

Funding/Support: This trial was supported by grant EMW-2010-FP-00601 from the Federal Emergency Management Agency and the Department of Homeland Security Assistance to Firefighters Grant Program–Fire Prevention and Safety Grants (Dr Piccirillo).

Role of the Funder/Sponsor: The funding source had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and the decision to submit the manuscript for publication.

Disclaimer: Dr Piccirillo is the editor and Dr Kallogjeri is the statistics editor of JAMA Otolaryngology–Head & Neck Surgery, but neither was involved in the editorial review nor the decision to accept the manuscript for publication.

Additional Contributions: The Brain Fitness Program–Tinnitus was supplied by Posit Science, San Francisco, California.