Impaired thresholds at extended high frequencies (EHF) are tightly linked to the prevalence of tinnitus, but little is known about how EHF status relates to tinnitus characteristics. In the present study, 93 individuals with tinnitus underwent standard (from 0.125 to 8 kHz) and EHF (from 10 to 16 kHz) audiometry and indicated their degree of tinnitus distress by completing the tinnitus functional index and their perceived tinnitus loudness by using a numeric rating scale. Partial correlation analyses indicated that the magnitude of EHF loss was significantly associated with degree of auditory related tinnitus distress (r = 0.343, p < 0.001) when controlling for pure tone average at standard frequencies and compensating for multiple testing. It is concluded that EHF status is related specifically to auditory related tinnitus distress, but not to intrusive-, sense of control-, cognitive-, sleep-, relaxation-, quality of life-, emotional-related tinnitus distress, total tinnitus distress, or perceived tinnitus loudness.
I. INTRODUCTION
Tinnitus and hearing status seem to be closely related. Tinnitus is suspected to be an effect of the brain's compensatory neuroplastic response to reductions in peripheral auditory activity, which can result in increased spontaneous auditory activity.1 While there are possible non-auditory underlying causes for tinnitus (e.g., sounds from blood flow caused by vascular anomalies2) the previously mentioned cause seems to be more common, as hearing impairment has been reported to be the main risk factor for tinnitus.3,4 In addition, tinnitus remission seems related to remission of hearing loss.5 Hearing impairment does not only appear to be related to the cause of tinnitus but also to how tinnitus manifests in the individual patient. Tinnitus patients tend to associate hearing related difficulties with their tinnitus rather than their hearing impairment,6 and among the many factors contributing to higher degrees of tinnitus distress,7 poorer hearing thresholds seem to be one.8,9
However, since not all tinnitus patients have clinical hearing loss as defined by pure tone average (PTA) at 0.5, 1, 2, and 4 kHz, tinnitus patients with “normal hearing” [i.e., PTA better than 25 dB hearing level (HL)] do exist.10,11 However, a standard hearing evaluation with apparently normal findings at standard audiometric frequencies should not be taken as definitive evidence that the individual's tinnitus is not related to hearing impairment. The current definition of hearing loss lacks precision, and a recent review concluded that the WHO definition should be nuanced so that it reflects the possibility that subtler hearing deviations may cause clinically relevant difficulties.12 There are good reasons for these suggestions, as a significant portion of individuals seeking health care due to hearing difficulties have “normal” hearing thresholds according to the current definitions13,14 and these individuals may benefit from hearing aid amplification.15
There is evidence suggesting that tinnitus may be triggered even by subtle hearing impairments which are not always detected with conventional pure tone-audiometry from 0.125 to 8 kHz (which is currently the gold standard for evaluating hearing status). The mechanical structures of the inner ear seem to be especially sensitive at the base of the cochlea, which corresponds to the highest audible frequencies. Indications of this have been reported in animal research, where the maximal cochlear damage occurs at very high frequencies regardless of the frequency of the sound causing the damage.16–18 Consistent with these findings and the previously mentioned link between hearing impairment and tinnitus, several recent studies have reported a significant association between impaired hearing thresholds at extended hearing frequencies (EHF; frequencies above 8 kHz) and tinnitus in humans.
The growing interest in the relationship between EHF status and tinnitus was summarized in a recent systematic review and meta-analysis by Jafari and colleagues,19 focusing on studies comparing EHF thresholds in subjects with and without tinnitus having normal hearing thresholds (≤25 dB HL) from 0.25 to 8 kHz. They screened 261 documents and found nine studies meeting their inclusion criteria. Jafari et al.19 conducted separate analyses for 10, 12.5, 14, 16, and 18 kHz, and the trend was consistent across all frequencies and all nine included studies in their meta-analysis, where individuals with tinnitus were showing poorer hearing thresholds compared to control subjects without tinnitus. When visualizing weighted mean hearing thresholds in both the standard audiometric frequency region and the EHF region, the authors revealed that control subjects and tinnitus subjects tended to differ minimally at standard audiometric frequencies, in contrast to the EHF range where the difference was clear and consistently to the tinnitus subjects' disadvantage. From the data obtained in the nine included studies, the authors concluded that tinnitus in most cases seems to be related to cochlear dysfunction manifested as impaired EHF thresholds. Jafari et al.19 also encouraged future studies to explore how impaired EHF hearing relates to tinnitus characteristics, as very little is known about such aspects from present literature. Illuminating such aspects is the focus of the present article.
One tinnitus characteristic of particular interest is perceived tinnitus loudness, as reducing tinnitus loudness is how most tinnitus patients define successful tinnitus treatment.20 Tinnitus can be associated with a negative impact on a wide range of aspects, such as sleep, concentration, anxiety, depression, and social life. Understanding what factors could potentially contribute to such subtypes of tinnitus distress is a step toward establishing effective evidence-based management of tinnitus patients. Tinnitus distress is currently most commonly examined using the Tinnitus Functional Index (TFI).21 The TFI consists of 25 items (statements) regarding how the respondent has perceived their tinnitus during the past week. These items are categorized into eight different subscales, based on what domain they address. Specifically, those are subscales for intrusiveness (items relating to how intense the tinnitus has been), sense of control (items relating to how difficult the tinnitus has been to ignore), cognitive difficulties (items relating to how much the tinnitus has interfered with the respondent's concentration), sleep difficulties (items relating to how much the tinnitus has interfered with the respondent's sleep), auditory difficulties (items relating to how much the tinnitus has interfered with the respondent's ability to follow conversations), relaxation (items relating to how much the tinnitus has interfered with the respondent's ability to feel calm), quality of life (items relating to how much the tinnitus has negatively affected the respondent's appreciation of every-day life), and emotional (items relating to how much the tinnitus has negatively affected the respondent's mood) domains. The mean score of all subscales gives a total TFI score, which makes the TFI a valuable tool both when estimating how severely affected the respondent overall is by their tinnitus, as well as what specific aspects of the respondent's life are impacted by their tinnitus.
The aim of the present study is to explore the possible relationships between EHF loss and tinnitus characteristics, specifically perceived tinnitus loudness and tinnitus distress including subcategories of tinnitus distress. We hypothesize that the magnitude of EHF loss would be related to perceived tinnitus loudness. The reason for this hypothesis is that hearing loss, which often seems to contribute to tinnitus, most commonly seems to start at the cochlear base (which corresponds to the highest frequencies we can hear).16–18
In addition, we hypothesize that EHF loss may also be related to total tinnitus distress (total TFI score). The reason for this is that previous research has indicated that greater progression of hearing impairment at standard audiometric frequencies is associated with a greater degree of tinnitus distress.7–9 We suspect that this trend could also be seen when examining early hearing impairment. Furthermore, we expect the degree of EHF loss to be related to how much the individual perceives that their tinnitus interferes with their ability to follow conversations (i.e., TFI auditory related distress subscale). This will hereinafter be referred to as auditory related tinnitus distress. While our research group has previously reported on the relationship between EHF loss and poorer cognitive performance in individuals with and without tinnitus,22–25 we do not expect any significant relationship between EHF loss and cognitive related tinnitus distress, as tinnitus patients' objective cognitive performance and perceived cognitive abilities seem to be discrepant aspects of the condition.25,26
II. METHODS
A. Participants
Ninety-three individuals experiencing chronic tinnitus were recruited via Audiology and Ear Nose Throat clinics, as well as through public advertising, in southern Sweden. See Table I for demographic data.
Demographic statistics. NRS, Numeric Rating Scale.
n . | 93 . |
---|---|
Sex (female n[%]/male n[%]) | 43[46.2%]/50[53.8%] |
Age, years (mean, range ± standard deviation, SD) | 53.4, 18.3 to 78.6 ± 13.4 |
Time since tinnitus debut, years (mean, range ±SD) | 14.2, 0.25 to 60.1 ± 12.5 |
Tinnitus distress, TFI (mean, range ±SD) | 43.8, 7.2 to 84.4 ± 18.1 |
Perceived tinnitus loudness, NRS (mean, range ±SD) | 59.8, 5 to 100 ± 22.1 |
Tinnitus laterality (unilateral n[%]/bilateral n[%]/experienced as a sound inside the head n[%]) | 17[18.3%]/68[73.1%]/8[8.6%] |
Tinnitus sound (tonal tinnitus n[%]/combined tone-and-noise tinnitus n[%]/noise only n[%]) | 61[65.6%]/30[32.3%]/2[2.2%] |
n . | 93 . |
---|---|
Sex (female n[%]/male n[%]) | 43[46.2%]/50[53.8%] |
Age, years (mean, range ± standard deviation, SD) | 53.4, 18.3 to 78.6 ± 13.4 |
Time since tinnitus debut, years (mean, range ±SD) | 14.2, 0.25 to 60.1 ± 12.5 |
Tinnitus distress, TFI (mean, range ±SD) | 43.8, 7.2 to 84.4 ± 18.1 |
Perceived tinnitus loudness, NRS (mean, range ±SD) | 59.8, 5 to 100 ± 22.1 |
Tinnitus laterality (unilateral n[%]/bilateral n[%]/experienced as a sound inside the head n[%]) | 17[18.3%]/68[73.1%]/8[8.6%] |
Tinnitus sound (tonal tinnitus n[%]/combined tone-and-noise tinnitus n[%]/noise only n[%]) | 61[65.6%]/30[32.3%]/2[2.2%] |
B. Procedure
Participation in the present study consisted of two parts: a physical visit consisting of a hearing examination at Skånes University Hospital (SUS) in Lund, Sweden, and an online survey delivered via Lund University's digital tool for online surveys, Sunet Survey. The online survey was answered shortly before or after the physical visit.
The physical visit consisted of otoscopy, tympanometry, and audiometry, and took place in a sound treated room adhering to the acoustic requirements stated in the international standard for pure tone audiometry.27 Tympanometry was conducted using a Madsen Zodiac Tympanometer (Natus, Middleton, WI), producing a probe tone of 226 Hz and sweeping the ear canal air pressure from +200 to −400 daPa. Audiometry was conducted using a Madsen Astera2 (Natus, Middleton, WI) audiometer, presenting stimuli via HDA 200 (Sennheiser, Wedemark, Germany) earphones and a bone conductor. The equipment was calibrated using a Brüel & Kjaer type 2209 sound level meter and type 4153 artificial ear (Brüel & Kjaer, Virum, Denmark), calibration was performed in accordance with international standards.28,29 A two-down/one-up (–10/+5 dB) adaptive method version of the international standard for audiometry27 was used to obtain pure tone hearing thresholds at 0.125, 0.25, 0.5, 1, 1.5, 2, 3, 4, 6, 8, 10, 12.5, 14, and 16 kHz for each ear. Participants were given one button in each hand and were asked to indicate whether they had heard a tone by pressing their right hand button if the tone was heard in their right ear and vice versa. Pure sine tones were used as stimuli for all participants and frequencies, however, in rare cases where a participant started responding extremely inconsistently at a particular frequency, warble tones were used instead which made it possible to establish a reliable hearing threshold. All participants underwent otoscopy and audiometry, while all but three participants underwent tympanometry. Tympanometry was skipped for those three participants as the tympanometer was not available at the time of testing.
When answering the online survey, participants indicated general background factors (age, sex, and educational level, and whether they were a current or former hearing aid user) and tinnitus related background factors (time since tinnitus debut, whether their tinnitus was constant, whether their tinnitus fluctuated in loudness and/or pitch, whether their tinnitus was pulsatile, whether external sounds could sometimes mask their tinnitus, and whether their tinnitus sounded like a tone/beep, a noise or a combination of both), filled out the TFI and indicated their perceived tinnitus loudness when not wearing a hearing aid or masker using a NRS ranging from 0 to 100, where 0 represented “inaudible” and 100 represented “loudest imaginable sound.”
C. Data analysis
In cases where a hearing threshold could not be obtained because the participant could not hear the pure tone stimulus at the audiometers highest presentation level at a given frequency, a hearing threshold 5 dB poorer than the audiometers highest presentation level at that given frequency was used in order to calculate a pure tone average.
Extended hearing frequencies pure tone average (EHF-PTA) was computed by adding the dB HL-values of measured hearing thresholds at 10, 12.5, 14, and 16 kHz, and dividing the sum by four. Standard pure tone average (S-PTA) was computed by adding the dB HL-values of measured hearing thresholds at 0.5, 1, 2, and 4 kHz, and dividing the sum by four. In order to take into account that some individuals experience unilateral tinnitus typically with poorer hearing in the tinnitus ear, we decided to use tinnitus sided PTAs for our statistical analyses. This meant that PTAs based on hearing thresholds in the tinnitus ear were used for individuals experiencing unilateral tinnitus and that PTAs based on bilateral average hearing thresholds were used for individuals experiencing bilateral tinnitus or experiencing tinnitus as a sound inside the head.
Data were analyzed using R version 4.2.2. We conducted partial correlations between tinnitus sided EHF-PTA and perceived tinnitus loudness (NRS) as well as total TFI score, and score on each TFI subscale (intrusive, sense of control, cognitive, sleep, auditory, relaxation, quality of life, and emotional). S-PTA was controlled for in each analysis, in order to assure that any detected relationship was in fact a relationship with hearing ability at extended hearing frequencies specifically, rather than poor hearing in general. Bonferroni correction was applied to compensate for multiple testing. As ten correlation analyses were conducted, the corrected α -level is given by: .
D. Ethics
All participants were informed regarding the purposes and conditions of the study, and signed an informed consent form prior to participation. The methods and design used in the present study has been examined and approved by the Swedish Ethical Review Authority (approval number: 2022-00727-01).
III. RESULTS
Otoscopy indicated normal ear canal- and tympanic membrane status in all participants but one, who had a perforated tympanic membrane. Minor abnormalities (such as lightly scarred tympanic membrane) were present in a few cases.
Tympanometry indicated intact tympanic membranes in all ears (except for the participant whose otoscopy procedure indicated perforation), as well as normal air pressure (or corrected to normal air pressure by Valsalva or Toynbee's maneuver) in the middle ear in 87 out of the 90 participants who underwent tympanometry. In two participants it was not possible to perform due to anatomical abnormalities, in one participant who was unable to correct the air pressure with Valsalva or Toynbee's maneuver a slight deviation of air pressure in the middle ear (<150 daPa) was found. The results that will be presented were not altered by excluding patients who required Valsalva or Toynbee's maneuver.
Audiometry indicated that 81 out of 93 participants had a hearing impairment (>20 dB HL in at least one frequency) in both the standard frequency range (0.125 to 8 kHz) and the EHF range (10 to 16 kHz). Seventy-nine of those had sensorineural hearing impairment only, while two had a combined conductive and sensorineural hearing impairment. The remaining 12 participants had no hearing impairment in the standard frequency range, out of which nine had a hearing impairment in the EHF range while three had no hearing impairment regardless of frequency range. See tinnitus sided average hearing thresholds for all participants in Fig. 1.
Tinnitus sided average hearing thresholds for all participants and all tested frequencies. Error bars display one SD.
Tinnitus sided average hearing thresholds for all participants and all tested frequencies. Error bars display one SD.
Partial correlation analyses indicated a statistically significant correlation only between tinnitus sided EHF-PTA auditory related tinnitus distress (auditory TFI subscale; r = 0.343, p < 0.001) when controlling for S-PTA and compensating for multiple testing. This correlation remained significant after adjusting the α-level for multiple comparisons (αcorrected = 0.005). No significant correlations were found between the magnitude of tinnitus sided EHF loss and perceived tinnitus loudness (NRS), total tinnitus distress (TFI score), or the intrusiveness, sense of control, cognitive difficulties, sleep difficulties, relaxation, quality of life, or emotional TFI subscales. See all correlations, including statistically insignificant, in Table II.
Partial correlations between tinnitus sided EHF-PTA and tinnitus characteristics when controlling for tinnitus sided S-PTA, in all participants (n = 93). NRS, Numeric Rating Scale; TFI total, total score on the Tinnitus Functionality Index (i.e., all subscales combined); TFI-I, Tinnitus Functional Index Intrusive subscale; TFI-SoC, Tinnitus Functional Index sense of Control subscale, TFI-C = Tinnitus Functional Index Cognitive subscale; TFI-S, Tinnitus Functional Index Sleep subscale; TFI-A, Tinnitus Functional Index Auditory subscale; TFI-R, Tinnitus Functional Index Relaxation subscale; TFI-QoL, Tinnitus Functional Index Quality of Life subscale; TFI-E, Tinnitus Functional Index Emotional subscale; *, statistically significant after Bonferroni correction for multiple testing (p < 0.005).
Tinnitus characteristic . | r . | p . |
---|---|---|
Perceived tinnitus loudness (NRS) | 0.091 | 0.386 |
TFI total | 0.031 | 0.770 |
TFI-I | −0.011 | 0.916 |
TFI-SoC | −0.112 | 0.287 |
TFI-C | −0.068 | 0.517 |
TFI-S | −0.101 | 0.337 |
TFI-A* | 0.343 | <0.001 |
TFI-R | −0.039 | 0.713 |
TFI-QoL | 0.193 | 0.065 |
TFI-E | −0.113 | 0.285 |
Tinnitus characteristic . | r . | p . |
---|---|---|
Perceived tinnitus loudness (NRS) | 0.091 | 0.386 |
TFI total | 0.031 | 0.770 |
TFI-I | −0.011 | 0.916 |
TFI-SoC | −0.112 | 0.287 |
TFI-C | −0.068 | 0.517 |
TFI-S | −0.101 | 0.337 |
TFI-A* | 0.343 | <0.001 |
TFI-R | −0.039 | 0.713 |
TFI-QoL | 0.193 | 0.065 |
TFI-E | −0.113 | 0.285 |
See a visual presentation of the significant correlation between tinnitus sided EHF-PTA and auditory related tinnitus distress in Fig. 2.
Significant relationship between tinnitus sided extended high frequency pure tone average (EHF-PTA) and auditory related tinnitus distress as measured with the tinnitus functional index auditory subscale (TFI-A), in all participants. The line depicts a line of best fit and the gray area depicts a 95%-confidence interval.
Significant relationship between tinnitus sided extended high frequency pure tone average (EHF-PTA) and auditory related tinnitus distress as measured with the tinnitus functional index auditory subscale (TFI-A), in all participants. The line depicts a line of best fit and the gray area depicts a 95%-confidence interval.
IV. DISCUSSION
Our results indicate that there is a significant positive relationship between tinnitus sided EHF loss and auditory related tinnitus distress, when controlling for hearing thresholds at standard frequencies, i.e., the greater the magnitude of the individual's tinnitus sided EHF loss, the greater the auditory related tinnitus distress is perceived.
Specifically, the auditory subscale of the TFI consists of three questions: “Over the past week how much has your tinnitus interfered with your ability to hear clearly?,” “Over the past week how much has your tinnitus interfered with your ability to understand people who are talking?,” and “Over the past week how much has your tinnitus interfered with your ability to follow conversations in a group or at meetings?” The responder indicates their answer by picking a number on a scale from zero (0) to ten (10), where zero represents “did not interfere” and ten represents “completely interfered.” Hence, the TFI auditory subscale reflects subjective hearing difficulties due to tinnitus, mainly focusing on speech intelligibility. We argue that our finding, that individuals with greater tinnitus sided EHF impairment experienced their tinnitus to interfere to a greater extent with their abilities to hear clearly and follow conversations, even when controlling for hearing impairment at standard audiometric frequencies, could be viewed in a broader context. It may have implications for the interpretation of previous studies reporting poorer speech intelligibility in individuals with tinnitus compared to control subjects without tinnitus even when both groups have normal hearing thresholds within standard audiometric frequencies, when not taking EHF status into account.30 Several recent studies have found that impaired EHF thresholds are associated with poorer speech intelligibility in noise, when not taking tinnitus into account.31–33 Tinnitus sufferers tend to associate negative experiences caused by their hearing impairment with their tinnitus rather than their hearing,6 and our present finding indicate that this may also be the case when it comes to speech intelligibility difficulties among tinnitus patients with apparently normal hearing status according to standard audiometric tests. It may be that tinnitus patients experience difficulties following conversations due to their EHF loss (which is tightly linked to tinnitus19) not due to their tinnitus per se.
Partially in line with our hypothesis, the significant relationship between the magnitude of tinnitus sided EHF loss and auditory related tinnitus distress was accompanied by a weaker relationship between tinnitus sided EHF loss and quality of life related tinnitus distress which approached statistical significance before correcting the α-level for multiple comparisons (r = 0.193, p = 0.065). We suspect that this is a secondary effect of the significant relationship between EHF loss and tinnitus related auditory distress, as a degree of perceived hearing difficulties predicts a decline in quality of life.34
Contrary to our hypothesis regarding perceived tinnitus loudness, there was no significant relationship with tinnitus sided EHF loss when controlling for hearing status at standard audiometric frequencies. This finding could be interpreted as tinnitus being a “warning signal” indicating that we are exposing ourselves to sound levels that threaten our cochlear homeostasis35,36 may be false. Experts within the field of tinnitus research have described tinnitus as a “warning signal,”35,36 making the affected individual aware that the environment they are currently in may cause hearing loss. Sanchez et al. reported a study in line with this hypothesis, demonstrating that almost half of tinnitus patients with normal hearing at baseline had developed a hearing impairment at approximately 3.5 years follow-up.37 Evolutionary, this may make sense. Tinnitus may arise when the peripheral auditory activity decreases, regardless of the cause for decrease (e.g., noise induced hearing loss,38 otosclerosis,39 otitis media,40 ototoxic drugs,41 prolonged earplug use42) However, when the auditory stimulation is restored, the tinnitus is likely to remit, as can be seen in cases of successful stapedotomy surgery43 or cochlear implantation.44 Throughout evolution, humans have likely been far less exposed to loud sounds compared to today's society where we put ourselves at risk for developing hearing loss due to our leisure listening habits,45 jobs,46 and traffic.47 Hence, in prehistoric times, a warning signal in the form of a ringing tone in our ears may have been helpful, making us eager to physically distance ourselves from the loud sound source causing a temporary threshold shift, just as the pain felt on our skin make us eager to avoid direct contact with fire. In other words, there may have been an evolutionary benefit of experiencing temporary tinnitus (decreasing the risk for hearing loss), but that may not be as much of a benefit in today's noise polluted society. This hypothesis is, however, challenged by our present findings, as we could not find a significant relationship between greater early hearing impairment (i.e., poorer EHF thresholds) and perceiving a stronger “warning signal” (i.e., louder tinnitus perception). However, it could also be that limitations of the present study may have clouded any significant relationships. The conclusions from the present study are limited by participants having unheard hearing thresholds. Our solution to this problem was to use the highest presented stimulus level + 5 dB when calculating PTAs for participants with one or more unheard hearing thresholds. As PTAs calculated with this method reflect the participant's hearing status with poorer precision compared to a PTA calculated based on heard frequencies only, the presence of unheard frequencies may have weakened or strengthened any observed relationship. Having at least one unheard threshold in the EHF range was seen in 71 out of 93 participants, in a majority of these individuals, 1–2 thresholds exceeded the upper limit of the audiometer in the EHF range. In addition, we were using subjective rating of perceived tinnitus as a variable. It is possible that two individuals hearing equally loud tinnitus sounds could rate their loudness experience differently, possibly due to their ability to hear external sounds or due to personality type. Future studies could investigate this by adopting behavioral loudness matching procedures instead of subjective ratings of tinnitus loudness.
The expected correlation between tinnitus sided EHF loss and total tinnitus distress was not to be found in the present study. Our interpretation is that EHF loss likely has little contribution to total tinnitus distress, at least compared to the contribution of hearing impairment at standard audiometric frequencies which was statistically controlled for and has previously been shown to be associated with more severe degrees of general tinnitus distress.7–9 As expected, there was no significant relationship between the magnitude of tinnitus sided EHF loss and cognitive related tinnitus distress. Expecting an absence of this relationship was based on our previous reports of a significant relationship between EHF loss and cognitive performance in tinnitus sufferers,23–25 but poor conformity between tinnitus sufferers perceived cognitive abilities and actual cognitive performances.25,26
A secondary finding was that among tinnitus patients with no signs of hearing impairment at standard audiometric frequencies (i.e., no hearing threshold poorer than 20 dB HL from 0.125 to 8 kHz), 10 out of 12 had signs of hearing impairment at EHF (i.e., at least one hearing threshold poorer than 20 dB from 10 to 16 kHz). This is consistent with the figures previously reported by Vielsmeier et al.,48 and adds to the existing literature indicating that standard audiometry alone is not sufficient to judge whether an individual's tinnitus could be related to auditory decline or not.19
The present study highlights that EHF loss is not only related to the presence of tinnitus but also specifically related to perceived hearing difficulties due to tinnitus. With the anticipated development of hearing aids with abilities to amplify sounds in the EHF region, we recommend investigating whether auditory stimulation at EHF frequencies could mitigate hearing difficulties in tinnitus patients.
ACKNOWLEDGMENTS
The present study was funded by a postdoc grant from Afa Försäkring awarded to Sebastian Waechter. The data that support the findings of this study are available from the corresponding author upon reasonable request. The authors report no conflict of interest. The methods and design used in the present study has been examined and approved by the Swedish Ethical Review Authority (Approval No. 2022-00727-01).