Ratings of degree of tinnitus distress using the Klockhoff and Lindblom15 rating scale into 3 levels of distress. Grade I is when tinnitus is audible only in silent environments. Grade II is when tinnitus is audible only in ordinary acoustic environments, but masked by loud environmental sounds; it can disturb falling asleep, but not sleep in general. Grade III is when tinnitus is audible in all acoustic environments, disturbs falling asleep, can disturb sleep in general, and is a dominating problem that affects quality of life. Results at admission and follow-up are shown for the cognitive behavioral therapy group and a comparison group who did not receive cognitive behavioral therapy.
Andersson G, Vretblad P, Larsen HC, Lyttkens L. Longitudinal Follow-up of Tinnitus Complaints. Arch Otolaryngol Head Neck Surg. 2001;127(2):175-179. doi:10.1001/archotol.127.2.175
To investigate the long-term outcome of patients with tinnitus, the long-term effects of cognitive behavioral therapy, and what properties of tinnitus predict distress at follow-up.
A longitudinal follow-up of a consecutive sample of patients with tinnitus initially seen by a clinical psychologist.
Department of Audiology, University Hospital, Uppsala, Sweden.
A consecutive series of 189 patients with tinnitus treated between January 1988 and March 1995 were sent a postal questionnaire booklet. One hundred forty-six (77 women and 69 men) provided usable responses, in all yielding a 77% response rate.
Main Outcome Measures
A questionnaire was derived from a structured interview "Questions About Your Tinnitus." Also included were the Tinnitus Reaction Questionnaire and tinnitus-matching data.
Questionnaire data showed that many patients with tinnitus still experienced distress an average of 4.9 years after admission. Tolerance of tinnitus increased over time overall. For patients who had received cognitive behavioral therapy (59%), there was a reduction in tinnitus-related distress. Further, an open-ended question showed that the benefits from treatment outnumbered the deficits. Multiple regression analysis showed that tinnitus maskability at admission was a significant predictor of distress at follow-up.
Severe tinnitus shows some signs of improvement over time, especially when psychological treatment has been given. Tinnitus maskability is an important prognostic factor of future tinnitus annoyance.
DESPITE INTENSE research, interest during the whole century,1 little is known about how tinnitus develops over time. This is important information when counseling patients with tinnitus. In fact, it is one of the most frequent questions facing the clinician at the tinnitus clinic.
Most of the research that has been conducted on the long-term outcome of tinnitus has been retrospective. For example, Stouffer et al2 concluded that tinnitus loudness and severity increased as a function of years since onset. However, since the study was cross-sectional, no definite conclusions could be drawn and the authors recommended longitudinal studies be conducted. Smith and Coles,3 who, in contrast, concluded that the severity of tinnitus was likely to decrease over time, underscored this need in their study. This was an epidemiological study and not just focusing on clinical tinnitus. However, data collected were retrospective. Since 25 of their subjects actually stated that tinnitus had disappeared, it is clear that tinnitus may be temporary. This concurs with our own clinical experience. Retrospective follow-up studies4,5 of tinnitus have also been published regarding acoustic neuroma (vestibular schwannoma) showing that tinnitus may both occur and disappear following surgery for the condition. However, this origin is rare and results are impossible to extrapolate to individuals with more common causes of tinnitus (eg, noise-induced heating loss). More is known regarding the natural history of tinnitus in elderly subjects.6,7 Rubinstein et al8 found substantial longitudinal fluctuations in tinnitus and a high occurrence of spontaneous remission. The patients were studied at the ages of 70, 75, and 79 years. Results showed that tinnitus had increased in severity in 25%, and decreased in 58% of the women leaving 17% unchanged. For the men tinnitus increased in 8% and decreased in 39%, with a larger proportion unchanged (53%). While the long-term outcome of tinnitus in Meniere disease has been studied,9 gradings of tinnitus have seldom been reported, vertigo being the symptom drawing most attention.
Another aspect of the long-term outcome of tinnitus is treatment response. The literature about the treatment of tinnitus is characterized by a lack of long-term outcomes, the main exception being psychological interventions for which a handful of controlled studies have included follow-up of up to 1 year. In a recent meta-analysis,10 support for the effect of cognitive behavioral therapy (CBT) was found noting that the effects remained at an average of 5.4 months after the treatment.
This investigation was undertaken to study the long-term outcome of tinnitus that had motivated specialist consultation. In particular, we were interested not only in changes in tinnitus distress and tolerance toward the tinnitus but also in the current state of the patients. We also wanted to study whether the audiological characteristics of tinnitus at the first appointment could predict distress at follow-up visits. Finally, since a proportion of our clinic patients had received CBT for their tinnitus, we wanted to investigate the long-term outcome of this therapy.
The subject sample in this study was drawn from a consecutive series of patients with tinnitus seen at the Department of Audiology, University Hospital, Uppsala, Sweden, between January 1988 and March 1995. All were primary referrals for tinnitus complaints. They underwent audiological and medical examinations and were seen and eventually treated by a clinical psychologist who conducted a structured interview.11 In this follow-up a postal questionnaire booklet titled "Questionnaire About Your Tinnitus" was sent to 189 former patients of whom 146 provided usable responses (responders), in all yielding a 77% response rate. No significant differences for demographics and tinnitus characteristics were found between the responders and those who declined participation. There were 77 women (53%) and 69 men (47%), ranging in age from 22 to 83 years (mean [SD] age, 56.4 [13.0] years). Average (SD) duration between the first appointment to follow-up was 4.9 (1.9) years (range, 3 and 10 years). Mean (SD) duration of tinnitus at follow-up was 10.5 (12.5) years (range, 3-50 years). Pretreatment audiological data were obtained from the medical records. Average (SD) hearing loss (pure-tone average calculated for "the better ear" over the frequencies of 0.5, 1, 2, and 3 kHz) was a 21.4 (19.5)-dB hearing loss. Tinnitus matching11 using an audiometer (model OB 822; Madsen Electronic, Taastrup, Denmark) over headphones (model TDH 39; Telephonics Inc, Farmingdale, NY) showed an average (SD) threshold tinnitus loudness of 44.1 (23.7)-dB hearing loss, 22 (18.9)-dB sensation level, and an average pitch of 5291 (3501) Hz. Minimal masking level was a 40.7 (25.6)-dB sound pressure level. From audiograms and medical records, the hearing losses were classified as sensorineural in most cases (92%). Included among these cases were Meniere disease (11%), acoustic neuroma (1%), and sudden deafness (7%). For the rest of the participants, conductive hearing loss (7.5%) and combined sensorineural-conductive hearing loss (1%) were present.
At the time of follow-up 72% of the sample lived with a spouse. Regarding occupational status, 34% were working full-time, 19% part-time, 41% had retired, 2% were unemployed, and 4% were on sick leave. In all, 72% were married or lived with a partner. At the time of the first appointment, CBT was offered to patients fulfilling the criteria of being distressed by their tinnitus (grade II or III, see the "Questionnaire Measures" section for an explanation of the grading system) and suited for psychological treatment. Of the responders in this follow-up, 59% had completed CBT. The average (SD) number of treatment sessions was 6.5 (3.1).
Cognitive behavioral therapy interventions are aimed at decreasing the psychological distress associated with tinnitus and are not targeted toward the loudness of tinnitus, which is usually unaffected by the therapy.12 The treatment package conducted at the Department of Audiology, University Hospital, Uppsala, is presented in 6 to 10 sessions on a weekly basis. Most patients receive the therapy on an individual basis. One characteristic feature of CBT for tinnitus is the use of homework assignments between therapy sessions and that a rationale is presented for each treatment component. In addition, the therapeutic relation between the therapist and the patient is collaborative in the sense that the outline of each session and the treatment as a whole are negotiated.12 Motivation to change habits and to alter behavior is crucial and it is made clear to the patient that work is required for the therapy to have any effect. The main components of the treatment package are as follows: analysis of influencing factors (behavior analysis), practicing and teaching applied relaxation (in 4 stages including positive imagery), the development of coping strategies (eg, sound enrichment and exercise), use of cognitive therapy techniques such as disputing negative beliefs about tinnitus, practicing concentration and distraction skills, and advice regarding sleep.12,13 Follow-up sessions are usually scheduled and advice is given to prevent relapse.
A structured interview protocol was adapted to fit a self-report format ("Questions About Your Tinnitus"). In all, the derived self-report questionnaire consisted of 44 questions. Additional open-ended questions were included but are only partly described in this article. Here the focus is on the number of positive treatment results and the number of drawbacks reported by each participant. This open-ended questionnaire follows procedures recommended in tinnitus research.14 Further, the questionnaire included background information (eg, duration of tinnitus).The included questions are listed in Table 1. Question 8 asked the participant to rate his or her tinnitus according to a 3-point scale of tinnitus distress developed by Klockhoff and Lindblom.15 In this system, grade I is when tinnitus is audible only in silent environments. Grade II is when tinnitus is audible only in ordinary acoustic environments, but masked by loud environmental sounds; it can disturb falling asleep, but not sleep in general. Grade III, finally, is when tinnitus is audible in all acoustic environments, disturbs falling asleep, can disturb sleep in general, and is a dominating problem that affects quality of life.
In addition to the questionnaire derived from the structured interview,11 the Tinnitus reaction questionnaire (TRQ)16 was included. The TRQ consists of 26 items used for assessing tinnitus-related distress. Wilson et al16 reported an internal consistency of 0.96 and a test-retest correlation of r = 0.88. The responses to each of the 26 items are assessed by a 0- to 4-point scale (0 indicates not at all; 4, almost all of the time), which are summed into a total score.
Table 1 lists the results for each of the questions in the "Questions About Your Tinnitus" questionnaire. Data for all participants are reported. Percentages are given for all questions except for questions 29 and 30 where means (SDs) are given. The mean number of positive results of CBT (1.19) outnumbered the number of deficits (0.75), and the difference was significant by means of dependent samples t tests [t51 = 2.7, P<.05]. The mean (SD) total score for the TRQ was 34.5 (3.9), the internal consistency of the scale was Cronbach α 0.97.
Differences between participants who had received and those who did not receive CBT were analyzed. To avoid type I errors, the number of tests was restricted to testing tinnitus grading and tolerance since onset. Nonparametric tests were used since data were not normally distributed.
Results for treated and untreated participants on the Klockhoff and Lindblom15 tinnitus grading at pretreatment and follow-up are shown in Figure 1. There was a significant pretreatment difference between the groups by the Mann-Whitney test (z = − 3.19, P<.005). Patients not included in the CBT program had a lower average grade of tinnitus at the first point of assessment. Within-group differences were tested with the Wilcoxon signed rank test. Patients who did not receive CBT showed no improvement over time (z = − 1.14, P = .26), whereas the treated patients had a significantly lower tinnitus grading at follow-up (z = − 2.33, P<.05). In Figure 1 a trend can be noted that the comparison group did deteriorate, but this was not significant.
Data from ratings of tolerance since onset showed that the participants who received CBT significantly increased their tolerance of tinnitus (z = − 4.69, P<.001). Calculating mean (SD) tolerance scores resulted in a pretreatment score of 2.3 (1.1) and a follow-up score of 3.1 (1.1) for the treated participants. However, the results for the comparison group also indicated an improvement in tolerance since onset (z = − 2.81, P<.01), with corresponding means (SDs) of 2.6 (1.1) and 3.2 (1.1). The between-group differences were not significant at admission or at follow-up.
While the data set allowed several questions to be asked, overall change in distress caused by tinnitus was restricted to a few focused analyses. The previously mentioned variable tinnitus grading did not change significantly when participants were considered as one group. Tolerance did change for the better when analyzing the whole group (z = − 5.47, P<.001).
Tinnitus matching data (maskability and loudness) and pure-tone thresholds at admission were used as independent variables in a multiple regression with the TRQ as the dependent variable at follow-up. The regression was significant (R2 = 0.11, P<.05) with a significant contribution of maskability (β = .30, P<.05), but not pure-tone average (β = − .29) or tinnitus loudness (β = .18). Thus, the higher masking sound needed the more distress at follow-up.
In this article the long-term outcome of tinnitus was studied. To our knowledge, the results presented in Table 1 includes information not previously reported in the tinnitus literature. Usually pretreatment data are used in the description of participants. Data regarding tinnitus characteristics, onset, and other factors have been reported in an earlier study of ours using admission data.11 Overall, the results at follow-up were similar to the admission data for the tinnitus characteristics. It is notable that a substantial proportion of sufferers still experience distress at follow-up. This sample represents only severe tinnitus, ie, individuals seeking help. The natural history of less severe tinnitus is not addressed by this study. Differences between those who report tinnitus and those who do not have been well documented in the literature and include factors such as complexity of the tinnitus sounds, degree of hearing loss, and psychological factors.17,18 By definition, the participants included in this study were persons who reported severe tinnitus, at least initially. Hence, they are not only likely to have more complex tinnitus but also to be more distressed psychologically. However, loudness of tinnitus is not as likely to differ from those who do not report tinnitus and who do not seek treatment.
A complex picture emerged from the data and individual differences were evident. Half of the sample had experienced longer periods with either more or less tinnitus (item 27 in Table 1), indicating that tinnitus may have fluctuated during the period between assessments, an observation in line with studies on elderly patients with tinnitus.8 On the positive side, more than half experienced situations when tinnitus was less problematic (item 16) and that they were able to do something about their problems (item 13).
The longitudinal part of this study set out to answer 2 overall questions: what happens to tinnitus over time and what is the role of CBT? Out of a large set of variables only a few differences were tested for statistical significance. The changes observed for the total sample was that tolerance of tinnitus increased. However, tinnitus grading remained stable from a statistical standpoint. For 21% tinnitus grading did increase and for 26% it decreased, leaving tinnitus grading unaffected for 53% of the total sample. Since the participants had different treatment experiences, it is more informative to consider the long-term outcome in relation to the treatment received. Participants who had completed CBT improved over time for tinnitus grading. This was not found in the patients who did not receive this form of therapy. However, data on tolerance since onset indicated that both groups improved over time, suggesting dicated an overall habituation to tinnitus. Other studies have evaluated the effects of CBT for tinnitus in controlled trials,10 and this long-term follow-up result gives partial support for the notion that CBT has beneficial long-term effects. As an additional way to evaluate treatment outcome, participants were asked to list benefits and deficits of CBT. Although the positive results outnumbered the deficits from a statistical standpoint, the difference in means was not large.
Data in this study allowed us to test what predicts distress at follow-up. Using the TRQ as a dependent measure, we found that tinnitus maskability at admission predicted distress at follow-up (for an average of 5 years following admission). This gives further support to the notion that audiological characteristics of tinnitus are relevant for the distress experienced.19 A rather similar finding was found in the previous analysis of the admission data,11 but the research literature is inconsistent regarding these issues.20 This could be due to the fact that hearing loss is an important factor for tinnitus annoyance and that the practice of reporting tinnitus loudness in sensation level (eg, loudness − hearing loss) results in weaker correlations between loudness and distress.21
Many methodological issues are raised by this study. The long-term outcome of CBT could only be collected in a naturalistic manner since such a long follow-up duration makes it impossible to conduct a controlled trial. Although we were able to use interview data from admission to the program, we did not, at this point, use a tinnitus questionnaire with established psychometric properties. The choice of the TRQ in this follow-up study was made because of its emphasis on psychological distress. Preferably, other measures such as the Tinnitus Handicap Inventory22 could also be included in future studies.
Although these data suggest that CBT can be beneficial in fostering habituation, only a modest improvement was found. Since tinnitus continues to be a source of distress for many patients at follow-up, tailored and effective treatments for diverse forms of tinnitus should be developed.
Accepted for publication November 1, 2000.
This work was sponsored in part by Stiftelsen Tysta Skolan, Stockholm, Sweden.
Corresponding author: Gerhard Andersson, PhD, Department of Psychology, Uppsala University, Box 1225, SE-751 42 Uppsala, Sweden (e-mail: Gerhard.Andersson@psyk.uu.se).