Aberrant Auditory Prediction Patterns Robustly Characterize Tinnitus

Reviewer #2 (Public review):

Summary:

This study aimed to test experimentally a theoretical framework that aims to explain the perception of tinnitus, i.e., the perception of a phantom sound in the absence of external stimuli, through differences in auditory predictive coding patterns. To this aim, the researchers compared the neural activity preceding and following the perception of a sound using MEG in two different studies. The sounds could be highly predictable or random, depending on the experimental condition. They revealed that individuals with tinnitus and controls had different anticipatory predictions. This finding is a major step in characterizing the top-down mechanisms underlying sound perception in individuals with tinnitus.

Strengths:

This article uses an elegant, well-constructed paradigm to assess the neural dynamics underlying auditory prediction. The findings presented in the first experiment were partially replicated in the second experiment, which included 80 participants. This large number of participants for an MEG study ensures very good statistical power and a strong level of evidence. The authors used advanced analysis techniques - Multivariate Pattern Analysis (MVPA) and classifier weights projection - to determine the neural patterns underlying the anticipation and perception of a sound for individuals with or without tinnitus. The authors evidenced different auditory prediction patterns associated with tinnitus. Overall, the conclusions of this paper are well supported, and the limitations of the study are clearly addressed and discussed.

Weaknesses:

Even though the authors took care of matching the participants in age and sex, the control could be more precise. Tinnitus is associated with various comorbidities, such as hearing loss, anxiety, depression, or sleep disorders. The authors assessed individuals' hearing thresholds with a pure tone audiogram, but they did not take into account the high frequencies (6 kHz to 16 kHz) in the patient/control matching. Moreover, other hearing dysfunctions, such as speech-in-noise deficits or hyperacusis, could have been taken into account to reinforce their claim that the observed predictive pattern was not linked to hearing deficits. Mental health and sleep disorders could also have been considered more precisely, as they were accounted for only indirectly with the score of the 10-item mini-TQ questionnaire evaluating tinnitus distress. Lastly, testing the links between the individuals' scores in auditory prediction and tinnitus characteristics, such as pitch, loudness, duration, and occurrence (how often it is perceived during the day), would have been highly informative.

Comments on revisions:

Thank you for your responses. There are a few remaining points that, if addressed, could further enhance the manuscript:

- While the manuscript acknowledges the limitation of not matching groups on hearing thresholds in Study 1, a deeper analysis of participants' hearing abilities and their impact on MEG results, similar to that conducted in Study 2, would be valuable. Specifically, including a linear model that considers all frequencies, group membership, and their interactions could highlight differences across groups. Additionally, examining the effect of high-frequency hearing loss on prediction scores, as performed in Study 2, would strengthen the analysis, particularly given the trend noted (line 719). Such an addition could make a significant contribution to the literature by exploring how hearing abilities may influence prediction patterns.

- The connection with the hippocampal regions (line 864) remains somewhat unclear. While the inclusion of the Paquette reference appropriately links temporal region activity with tinnitus, it does not fully support the statement: "An increased focus on hippocampal regions, e.g., in fMRI, patient, or animal studies, could be a worthwhile complement to our MEG work, given the outstanding relevance of medial temporal areas in the formation of associations in statistical learning paradigms"

- Authors should add a comparison of participants mini-TQ scores on both studies
- Authors should add significant level on Fig 6.B as in Fig 3.C, and a n.s on Fig 6.D

Author response:

The following is the authors’ response to the original reviews.

Public Reviews:

Reviewer #1 (Public Review):

This work presents a replicable difference in predictive processing between subjects with and without tinnitus. In two independent MEG studies and using a passive listening paradigm, the authors identify an enhanced prediction score in tinnitus subjects compared to control subjects. In the second study, individuals with and without tinnitus were carefully matched for hearing levels (next to age and sex), increasing the probability that the identified differences could truly be attributed to the presence of tinnitus. Results from the first study could successfully be replicated in the second, although the effect size was notably smaller.

Throughout the manuscript, the authors provide a thoughtful interpretation of their key findings and offer several interesting directions for future studies. Their conclusions are fully supported by their findings. Moreover, the authors are sufficiently aware of the inherent limitations of cross-sectional studies.

Strengths:

The robustness of the identified differences in prediction scores between individuals with and without tinnitus is remarkable, especially as successful replication studies are rare in the tinnitus field. Moreover, the authors provide several plausible explanations for the decline of the effect size observed in the second study.

The rigorous matching for hearing loss, in addition to age and sex, in the second study is an important strength. This ensures that the identified differences cannot be attributed to differences in hearing levels between the groups.

The used methodology is explained clearly and in detail, ensuring that the used paradigms may be employed by other researchers in future studies. Moreover, the registering of the data collection and analysis methods for Study 2 as a Registered Report should be commended, as the authors have clearly adhered to the methods as registered.

Weaknesses:

Although the authors have been careful to match their experimental groups for age, sex, and hearing loss, there are other factors that may confound the current results. For example, subjects with tinnitus might present with psychological comorbidities such as anxiety and depression. The authors' exclusion of distress as a candidate for explaining the found effects is based solely on an assessment of tinnitus-related distress, while it is currently not possible to exclude the effects of elevated anxiety or depression levels on the results. Additionally, as the authors address in the discussion, the presence of hyperacusis may also play a role in predictive processing in this population.

The authors write that sound intensity was individually determined by presenting a short audio sequence to the participants and adjusting the loudness according to an individual pleasant volume. Neural measurements made during listening paradigms might be influenced by sound intensity levels. The intensity levels chosen by the participants might therefore also have an effect on the outcomes. The authors currently do not provide information on the sound intensity levels in the experimental groups, making it impossible to assess whether sound intensity levels might have played a role.

Thank you very much for your favorable and constructive evaluation of our manuscript. We agree with you on various additional confounds that we did not consider and included a section in our discussion. It is also correct that we did not include the sound intensity levels in our analysis, which is also a potential confound. Unfortunately, we do not have the data on the individual sound intensity levels but we included a section regarding this issue in our discussion as well.

Line 937-949:

“In both studies, tinnitus distress was not correlated with the reported prediction effects. Nevertheless, tinnitus can also be characterized by other features such as its loudness, pitch or duration which were not included in the experimental assessment. Additionally, we solely used a short version of the Mini-TQ (Goebel and Hiller, 1992) in Study 2, which did not allow us to relate prediction scores to subscales like sleep disturbances which potentially influence cognitive functioning and thus predictive processing. Next to sleeping disorders and distress, tinnitus is often also accompanied by psychological comorbidities such as depression or anxiety (Langguth, 2011) which are potential confounds of the results. For the work described in this manuscript the replicability of the core finding was of main importance. More studies are needed taking into account to assess relate the prediction patterns in more detail to aspects of tinnitus sensation and distress.”

Reviewer #2 (Public Review):

Summary:

This study aimed to test experimentally a theoretical framework that aims to explain the perception of tinnitus, i.e., the perception of a phantom sound in the absence of external stimuli, through differences in auditory predictive coding patterns. To this aim, the researchers compared the neural activity preceding and following the perception of a sound using MEG in two different studies. The sounds could be highly predictable or random, depending on the experimental condition. They revealed that individuals with tinnitus and controls had different anticipatory predictions. This finding is a major step in characterizing the top-down mechanisms underlying sound perception in individuals with tinnitus.

Strengths:

This article uses an elegant, well-constructed paradigm to assess the neural dynamics underlying auditory prediction. The findings presented in the first experiment were partially replicated in the second experiment, which included 80 participants. This large number of participants for an MEG study ensures very good statistical power and a strong level of evidence. The authors used advanced analysis techniques - Multivariate Pattern Analysis (MVPA) and classifier weights projection - to determine the neural patterns underlying the anticipation and perception of a sound for individuals with or without tinnitus. The authors evidenced different auditory prediction patterns associated with tinnitus. Overall, the conclusions of this paper are well supported, and the limitations of the study are clearly addressed and discussed.

Weaknesses:

Even though the authors took care of matching the participants in age and sex, the control could be more precise. Tinnitus is associated with various comorbidities, such as hearing loss, anxiety, depression, or sleep disorders. The authors assessed individuals' hearing thresholds with a pure tone audiogram, but they did not take into account the high frequencies (6 kHz to 16 kHz) in the patient/control matching. Moreover, other hearing dysfunctions, such as speech-in-noise deficits or hyperacusis, could have been taken into account to reinforce their claim that the observed predictive pattern was not linked to hearing deficits. Mental health and sleep disorders could also have been considered more precisely, as they were accounted for only indirectly with the score of the 10-item mini-TQ questionnaire evaluating tinnitus distress. Lastly, testing the links between the individuals' scores in auditory prediction and tinnitus characteristics, such as pitch, loudness, duration, and occurrence (how often it is perceived during the day), would have been highly informative.

Thank you very much for your careful and constructive evaluation. We agree with the weaknesses stated in our manuscript and aimed to highlight these aspects more in our analyses and discussion, so future studies can take them into account (see e.g., line 937949).

Recommendations for the authors:

Reviewer #1 (Recommendations For The Authors):

I would strongly recommend the inclusion of data on the used sound intensity levels. It would be very useful to assess whether there are any group differences regarding sound intensity of the stimuli, to exclude any effects of sound intensity on the results.

We agree with you that - next to experimental aspects like the stimulus frequencies and the number of trials - the sound intensity levels potentially influence the effects as well. Unfortunately, this data was not saved during the experimental procedure and we are not able to include this as a variable in our analyses. As we, however, acknowledge this issue and want to provide guidelines for future research, we added a section to our discussion targeting sound intensity levels.

Line 902-913:

“Thirdly, both studies used individual sound intensity levels to ensure a comfortable listening situation for the participants. These differences in sound intensity levels are, however, a potential confound in the experimental design as well since sound intensity can have an impact on neural responses (Thaerig et al., 2008). Although in this design, we expect the intensity levels balanced equally to the hearing loss of the participants (which did not differ between groups), and basic decoding of sound frequency did not differ in both studies, we are not able to ultimately exclude the sound intensity level as a driver of our effects. Future studies should include a perceived loudness matching for each frequency and should compare the adapted sound intensity values between each group or integrate them into the analysis (e.g., using the logistic regression approach in Fig. 8).”

Reviewer #2 (Recommendations For The Authors):

Major comments

Introduction

• The authors wrote: "Overall, this situation calls for the pursuit of alternative or complementary models that place less emphasis on the hearing status of the individual." They clearly demonstrated that the altered-gain model focuses on hearing loss and does not overcome the three described limitations. However, they mentioned other models focusing on brain activity outside of the auditive pathway (noise cancellation, map reorganization, specific neural networks. The authors should better explain the novelty of their approach compared to the existing ones.

Thank you for your input. The inconclusive results and open questions about the altered-gain framework let us search for a different theoretical foundation for this work. We agree with you, that there are other models such as the map reorganization theory or neural network models next to the altered gain model and recent literature showed results supporting these frameworks (see e.g., a review from our group discussing tinnitus research in MEG over the last 10 years, Reisinger et al. (2023)). Nevertheless, as we focus on prediction processes, the Bayesian inference framework in tinnitus (Sedley et al., 2016) fits best for our approach. As we stated in line 113-116 “The Bayesian inference framework could, therefore, explain the experience of tinnitus in lieu of any increase in neural activity in the auditory system, or indicate an additional alteration, on top of hearing loss, for tinnitus to be perceived”, this framework differs from the other models and demonstrate a novel approach in tinnitus research. The novelty in this work is our methodological approach, which allows for explicit analyses of predictive patterns, irrespective of the exact location in the brain. This is a first step towards our actual underlying question whether aberrant auditory prediction patterns act as a neural correlate of tinnitus or rather as a risk factor or disposition. In our opinion, this question is of crucial relevance for understanding tinnitus processes on a neural level and our robust effects highlight the necessity to investigate these predictive processes in a longitudinal manner. We included a paragraph in our manuscript to make this more apparent for the reader.

Line 128-137:

“We utilized a powerful, recently established experimental approach (Demarchi et al., 2019) showing anticipatory activations of tonotopically specific auditory templates for regular tone sequences. This method allows us to explicitly investigate predictive patterns in line with the Bayesian inference framework (Sedley et al., 2016), leading towards the overall question whether alterations in predictive coding can be interpreted as a neural correlate of tinnitus or rather as a risk factor. Since this question can solely be targeted in a longitudinal manner, we aimed in a first step to investigate prediction patterns in tinnitus over two independent samples, deriving robust effects that should be considered in future research.”

• "This conceptual model bridges several explanatory gaps: for example, the inconsistent findings in humans regarding the "altered gain" view which states enhanced neural activity in the auditory pathway". What are "the inconsistent findings in humans regarding the 'altered gain'"? It would be helpful if the authors were more explicit about their idea here and added reference(s) to support it.

Thank you for pointing that out. We agree with you that this section lacks clarity and we aimed to be more precise.

Line 108-116:

“This conceptual model bridges several explanatory gaps: for example, the inconsistent findings in humans regarding the “altered gain” view which states altered neural activity in the auditory pathway. Recent findings vary in both the targeted frequency bands and the direction of the reported power changes which impede consistent conclusions (Eggermont and Roberts, 2015; Elgohyen et al., 2015, Reisinger et al., 2023). The Bayesian inference framework could, therefore, explain the experience of tinnitus in lieu of any increase in neural activity in the auditory system, or indicate an additional alteration, on top of hearing loss, for tinnitus to be perceived.”

• I suggest moving this part to the discussion:

"However, alternative explanations cannot be excluded with certainty, such as tinnitus being the cause of altered prediction tendencies or that there is a third variable being responsible for predictions and tinnitus development. Furthermore, even if altered predictive tendencies were to be found, there could be various possibilities of exactly how they could be altered to contribute to the onset or persistence of tinnitus. Some further clarity might then be gained through longitudinal studies in humans or animals."

Thank you for your suggestion, we moved this part to the corresponding section in the discussion.

Line 742-756:

“Distinct predictive processing patterns could e.g., either develop within an individual in contributing to chronification of tinnitus (e.g., shift of “default prediction” from silence to sound; Sedley, 2019). Alternatively, they could be conceived as sensory processing style, making certain individuals more vulnerable to develop tinnitus under certain conditions (e.g., hearing loss, aging), a notion reminiscent of the “strong prior” hypothesis of hallucinations (Corlett et al., 2019). Hence, the direction of the effect remains unclear and alternative explanations, such as a third variable being responsible for predictions and tinnitus development, cannot be excluded with certainty. Furthermore, even if altered predictive tendencies were to be found, there could be various possibilities of exactly how they could be altered to contribute to the onset or persistence of tinnitus. In any case, any more conclusive claims would require longitudinal data, ideally with a tinnitus-free baseline. As such research is challenging to implement, especially in humans, we first focused in this work on finding cross-sectional group differences between individuals with and without tinnitus.”

Methods

Participants

• "We calculated the individual mean hearing ability based on the values for 500, 1000, 2000, and 4000 Hz, which is a common approach for averaging results of pure-tone audiometry". Even if this method has been used multiple times in the literature, I would not recommend it as it can hide differences. Hearing loss is usually larger at high frequencies (starting at 6 000 Hz). An average threshold calculated with those central frequencies is more relevant for clinical use than in research. I strongly recommend performing a linear model with the factors Frequency (including all tested frequencies), Group, Ear side, and their interactions to precisely test the group differences in hearing thresholds.

Thank you for pointing that out. We agree with you that higher frequencies are of potential interest as well when analyzing hearing loss. We included your suggested linear model in our methods section and the results were in line with our assumption that the groups did not differ substantially. Additionally, we included another logistic regression model in our exploratory analyses when investigating the influence of hearing loss on the prediction scores. Once more, the addition of higher frequencies did not substantially influence the effects.

Line 194-203:

“We calculated the individual mean hearing ability based on the values for 500, 1000, 2000, and 4000 Hz, which is a common approach for averaging results of pure-tone audiometry (i.e., PTA-4, see for example Lin et al. (2011); Ozdek et al. (2010)). Using independent t-tests, we found no differences in hearing status over frequencies between groups for the left(t=-1.19, p=.238) and right ear (t=-1.72, p=.09). An additional linear regression including all frequencies from 125 Hz to 8000 Hz also showed that hearing thresholds did not differ between ears (b=0.311, SE=1.600, p=.846) and groups (b=1.702, SE=1.553, p=.273), but solely between frequencies (b=0.003, SE=0.000, p<.001). Interactions were not significant as well.”

Line 712-725:

“As these logistic regression models were computed using an average hearing score computed over the frequencies 500, 1000, 2000, and 4000 Hz (i.e., PTA-4, see for example Lin et al. (2011); Ozdek et al. (2010)), we questioned whether hearing loss in higher frequencies influenced our effects. We therefore computed an additional logistic regression including also the PTA values of 6000 and 8000 Hz. In this analysis, hearing loss was not a significant predictor of tinnitus but rather showed a trend with b=0.211, SE=0.111, p=.062. Prediction scores, however, remained a significant predictor of tinnitus even after including high-frequency hearing loss (b=0.232, SE=0.111, p=.040). In this analysis, odds ratios indicated an increase of 26% in the odds of having tinnitus with a one standard deviation increase in the prediction score. Overall, this analysis strongly supports the notion that the main effect genuinely reflects a process related to the experience or statistical risk of experiencing tinnitus.”

Stimuli and experimental procedure

• Can you explain the use of movies during sound listening? And not an active listening task with oddball events, for example, to ensure that the subject attention is directed to the sounds?

Thank you for your comment. We agree with you that attention is a relevant factor and with our design we cannot exclude potential attention effects on our findings. We chose this paradigm since previous research in our group including this exact experimental design (Demarchi et al., 2019) impressively demonstrated the formation of feature-specific auditory predictions in the brain and we aimed to investigate to what extent this can be detected in the tinnitus brain.

We acknowledged this issue in our discussion (see line 916-919): “In the current work, we used passive listening tasks including a movie to reduce attentional focus on the presented stimuli. Therefore, we cannot draw conclusions whether differences in attention had an influence on the effects. Future studies should include more manipulations of attention to investigate its relevance”.

Results

Pre-stimulus effects are not related to hearing loss and tinnitus-related features

• How was the hearing loss calculated for this analysis? I recommend a PCA on the hearing levels, to get individual scores with a data-driven approach. Usually, the first dimension will be an average of all the frequencies. The second should be a difference between low and high frequencies. The same comment applies to study 2.

Thank you for pointing that out. In the first study, participant groups were not controlled for hearing loss and pure-tone audiograms were solely averaged over all frequencies and both ears. As we marked out throughout the manuscript, insufficient control for hearing loss was the key issue in study 1 which led to the implementation of study 2. Further, we do not have data about the hearing status of every participant in study 1 and we do therefore not believe that a more complex approach for calculating hearing loss will increase interpretability in study 1. Nevertheless, we agree with you that it is not apparent how hearing loss was calculated in study 1. The results of the pure-tone audiometry were averaged over all frequencies and both ears, but no cut-off values were defined to characterize hearing loss. We therefore highly appreciate your detailed revision of our manuscript and adjusted the phrasing in the corresponding section. With our approach, it is not justifiable to talk about hearing loss but rather hearing thresholds. As for study 2, the methodological approach was reviewed and accepted as a Registered Report and we therefore do not want to deviate drastically from our pre-registered approach.

Line 162-165:

“Standardized pure-tone audiometric testing for frequencies from 125Hz to 8kHz was performed in 31 out of 34 tinnitus participants using Interacoustic AS608 audiometer.

Averages were computed over all frequencies and both ears.”

Line 356-362:

“In the whole sample of participants with tinnitus (n=34) we performed a Spearman correlation of the β-coefficient values corresponding to the time-point of the maximum and the minimum t-value in intergroup analysis (comprised of positive and negative significant clusters emerging in group comparison for sound trials) with hearing thresholds (averaged audiogram for both ears), tinnitus loudness (10-point scale) and tinnitus distress scores (TQ).”

Line 463-464:

See as well Line 471-481.

Line 491-495:

“Our main findings are: 1) basic processing of carrier frequencies are not altered in tinnitus; 2) with increasing regularity of the sequence, individuals with tinnitus show relatively enhanced predictions of frequency information; 3) the effect is not related to hearing thresholds and tinnitus distress or loudness in this sample.”

• In the methods, the authors indicated that the volume was adjusted individually at a pleasant volume. Can authors test if the volume was related to the individual's accuracy? Did they test that all frequencies were audible for all participants?

Thank you for your feedback. We agree with you that it would be interesting to see whether sound intensity levels were related to the accuracy. Unfortunately, data regarding the volume was not saved during the experimental procedure and we are not able to include this as a variable in our analyses. We acknowledge this issue and added a section to our discussion targeting sound intensity levels. As for the second question, the individual volume adjustment was also meant to guarantee that all frequencies were audible for the participant. We clarified this in the methods section. Overall, it is important to mention that we did not find any differences between groups in the decoding of random tones (see Fig. 2 and Fig. 6C), indicating that the volume did not substantially have an influence on one group compared to the other.

Line 232-234:

“Sound intensity was individually determined by presenting a short audio sequence to the participants and adjusting the loudness according to an individual pleasant volume with all four frequencies audible for the participant.”

Line 902-913:

“Thirdly, both studies used individual sound intensity levels to ensure a comfortable listening situation for the participants. These differences in sound intensity levels are, however, a potential confound in the experimental design as well since sound intensity can have an impact on neural responses (Thaerig et al., 2008). Although in this design, we expect the intensity levels balanced equally to the hearing loss of the participants (which did not differ between groups), and basic decoding of sound frequency did not differ in both studies, we are not able to ultimately exclude the sound intensity level as a driver of our effects. Future studies should include a perceived loudness matching for each frequency and should compare the adapted sound intensity values between each group or integrate them into the analysis (e.g., using the logistic regression approach in Fig. 8).”

Pre-stimulus differences in ordered and random tone sequences are not related to tinnitus distress • Accuracy was not correlated with tinnitus distress. Could the authors test if the accuracy was related to other clinical data, such as tinnitus pitch, duration, and loudness? And at the subscales of the mini-TQ?

We appreciate your constructive feedback and agree with you that other tinnitus features such as pitch, duration, or loudness are also interesting in this regard. Unfortunately, these features were not assessed in study 2 and we are therefore not able to provide this information. Additionally, we solely used a short version of the Mini-TQ in this study and did not assess all subscales but rather used all available items for calculating tinnitus distress. This is a limitation of our study design and we included it in the discussion.

Line 937-949:

“In both studies, tinnitus distress was not correlated with the reported prediction effects. Nevertheless, tinnitus can also be characterized by other features such as its loudness, pitch or duration which were not included in the experimental assessment. Additionally, we solely used a short version of the Mini-TQ (Goebel and Hiller, 1992) in Study 2, which did not allow us to relate prediction scores to subscales like sleep disturbances which potentially influence cognitive functioning and thus predictive processing. [...] More studies are needed taking into account to assess relate the prediction patterns in more detail to aspects of tinnitus sensation and distress.”

The strength of group effects differs between the two studies

• This section should be in the discussion, not the results

Thank you for your valuable input. In this section, we show comparisons between the two studies and report Bayes factors over time for the differences in decoding accuracy (see Figure 7A). We introduce novel results and believe therefore that this section should remain in the results and is discussed later in the manuscript.

Discussion

• Globally, the discussion is very long and a bit speculative. I recommend the authors shorten the discussion (especially the speculations), and delete the repetition.

Thank you very much for your constructive feedback. We aimed to shorten our discussion and delete repetitions to increase clarity and readability.

• The effect of hearing loss has been tested in this study, evaluated as the mean hearing threshold of 4 central frequencies. However, hearing abilities cannot be limited to a central audiogram. High frequencies, speech-in-noise abilities, or other hidden hearing loss can be impacted, even for individuals without hearing loss on 500Hz- 4000Hz. The conclusion on the prediction effect being independent of hearing loss should include this limitation.

Thank you for pointing that out. We added this limitation to the discussion.

Line 781-794:

“In a complementary analysis, we used our prediction score in addition to hearing loss magnitudes as predictors of tinnitus in a logistic regression. Prediction related pre-activation levels were informative whether participants perceived tinnitus, also when statistically controlling for hearing loss. However, it has to be mentioned that we calculated hearing loss based on the PTA results of the frequencies between 500 and 4000 Hz. This does not reflect hearing impairments like high frequency hearing loss or hidden hearing loss (i.e., hearing difficulties despite a normal audiogram, Liberman (2015)). As for hidden hearing loss, we were not able to draw conclusions regarding our effects since this concept of hearing damage is difficult to measure objectively, especially in humans. However, we included an additional logistic regression expanding the frequency range up to 8000 Hz and again, hearing loss did not substantially impact the prediction score as an informative tinnitus predictor.”

Line 712-723:

“As these logistic regression models were computed using an average hearing score computed over the frequencies 500, 1000, 2000, and 4000 Hz (i.e., PTA-4, see for example Lin et al. (2011); Ozdek et al. (2010)), we questioned whether hearing loss in higher frequencies influenced our effects. We therefore computed an additional logistic regression including also the PTA values of 6000 and 8000 Hz. In this analysis, hearing loss was not a significant predictor of tinnitus but rather showed a trend with b=0.211, SE=0.111, p=.062. Prediction scores, however, remained a significant predictor of tinnitus even after including high-frequency hearing loss (b=0.232, SE=0.111, p=.040). In this analysis, odds ratios indicated an increase of 26% in the odds of having tinnitus with a one standard deviation increase in the prediction score.”

• "An increased focus on hippocampal regions, e.g., in fMRI, patient, or animal studies, could be a worthwhile complement to our MEG work, given the outstanding relevance of medial temporal areas in the formation of associations in statistical learning paradigms (see e.g., Covington et al., (2018); Schapiro et al., (2016)).".

in the opinion of this reviewer, this claim is not well introduced and should be removed.

Thank you for pointing that out. In our opinion, an increased focus on hippocampal regions is an important consideration for future research and we decided to keep this part in the manuscript. However, we added a third reference highlighting the relevance of temporal areas in tinnitus to strengthen our claim.

Line 866-868:

“... given the outstanding relevance of medial temporal areas in the formation of associations in statistical learning paradigms (see e.g., Covington et al., (2018); Paquette et al., (2017); Schapiro et al., (2016)).”

References:

Paquette, S., Fournier, P., Dupont, S., de Edelenyi, F. S., Galan, P., & Samson, S. (2017). Risk of tinnitus after medial temporal lobe surgery. JAMA neurology, 74(11), 1376-1377. https://doi.org/10.1001/jamaneurol.2017.2718.

• "Overall, our work clearly underlines the true presence of differences, in terms of predictive processing, between individuals with and without tinnitus. At the same time, distinct design choices impact the strength of the effects which is not only apparent in the present work but was also reported recently by Yukhnovich and colleagues (2024). Further to controlling for basic variables (age, sex, hearing loss), future studies using our paradigm and analysis approach should opt for a broad frequency spacing (>2 octaves) and ideally more than 2000 trials per carrier frequency in the random sequence. These recommendations are likely even more important for efforts of testing this paradigm using EEG, which normally comes with inferior data quality as compared to MEG."

This reviewer considers that the entire paragraph should be deleted, as the effects are already covered in the previous paragraph.

Thank you very much for your feedback, however, we believe that this paragraph acts as a brief and accurate summary for our guidelines to improve future research in this field. This section therefore remained in the manuscript.

Minor comments

Introduction

• "The onsets of tinnitus and hearing loss often do not occur at the same time ". This sentence should have a reference.

We appreciate your careful evaluation of our manuscript and included a reference to the sentence pointing out hearing loss as a precursor of tinnitus.

Line 95f.:

“2) The onsets of tinnitus and hearing loss often do not occur at the same time (Roberts et al., 2010).”

Methods

Participants

• Participants' laterality needs to be mentioned.

Thank you for your input. We agree with you that laterality is an interesting aspect that should be taken into account. Unfortunately, however, we did not assess this in the current design. We mentioned the lack of this information in the methods section.

Line 158:

“Laterality of the participants was not assessed.”

176-177:

“No participants with psychiatric or neurological diseases were included in the sample. Laterality of the participants was not assessed.”

"Four individuals with tinnitus did not show any audiometric abnormality; four of the participants showed unilateral hearing impairments; 26 volunteers had high-frequency hearing loss; and six individuals were hearing impaired over most frequencies (i.e. hearing thresholds higher than 30 dB)."

This part is not precise enough. "Unilateral hearing impairment": is it on one or multiple frequencies? "26 volunteers had high-frequency hearing loss". What is considered as highfrequency here? The precision "(i.e. hearing thresholds higher than 30 dB)" can be dropped as it was defined in the sentence just before.

We appreciate your constructive feedback and added information to clarify the audiometric characteristics of our participants.

Line 186-190:

“Four individuals with tinnitus did not show any audiometric abnormality; four of the participants showed unilateral hearing impairments on at least one frequency; 26 volunteers had high-frequency hearing loss (i.e. hearing thresholds higher than 30 dB); and six individuals were hearing impaired over most frequencies (i.e. hearing thresholds higher than 30 dB).”

Results

• Figure 3C: are those group differences significant? It should be noted on the graphs.

• Figure 6D: I would suggest to remove this figure, as the correlation is not significant.

• Figure 7A: It would be useful to precise the number of trials for each study, in parenthesis.

• Figure 8 is unnecessary.

Thank you for your careful assessment of our figures. We agree with you that significance should be indicated in Figure 3C and that the precise number of trials is relevant information in Figure 7A. We corrected the figures accordingly. However, the Figures 6D and 8 remained in the manuscript since they were already part of our Registered Report and we do not want to remove graphical information that was reviewed and accepted already.