Several researchers have reported that tinnitus
and hyperacusis are associated with depression (Kehrle et al. 2016; Ooms et al. 2012; Andersson et
al. 2004; Hu et al. 2015; Tyler & Baker 1983).
However, only a few studies have examined the factors that predict depression
in patients with tinnitus and hyperacusis.
Exploration of the factors predicting
depression in patients with tinnitus and hyperacusis is important as it can
guide psychiatric management of the co-morbid depression in this population.
Although audiology departments play a major role in offering rehabilitative
interventions and support for patients experiencing tinnitus and hyperacusis (Thompson et al. 2017; Gander et al. 2011), in the
UK and many other countries the psychological co-morbidities are assessed and
treated by mental health services (Department of Health 2009; McKenna et al. 1991).
Dr. Aazh’s tinnitus team developed the first mathematical model which explained over 60% of the variance of depression severity as measured via
Hospital Anxiety and
Depression Scale (HADS; Zigmond & Snaith 1983) in over 600 patients with tinnitus
and/or hyperacusis (Aazh & Moore 2017). The model is presented in the table
Table : Variables
included in the final version of the stepwise linear regression model for
predicting depression score measured with the HADS-D, together with
regression coefficients, p values, and 95% CI values. Variables are
listed according to the value of the regression coefficient (highest first).
p = 0.001
Note: Hospital Anxiety and Depression Scale-Depression subscale (HADS-D), Hospital Anxiety and Depression Scale-Anxiety subscale (HADS-A), Visual Analogue Scale (VAS) of
tinnitus loudness, Tinnitus Handicap Inventory (THI), Hyperacusis Questionnaire (HQ), and Insomnia Severity Index (ISI).
method used to explore the process in which tinnitus loudness and hyperacusis
may relate to depression
Aazh’s tinnitus team used mediation analysis to assess the direct and
indirect effects of tinnitus loudness (independent variable) on the depression
score (dependent variable). In addition, they assessed the direct and indirect
effects of reduced ULLs (independent variable) on the depression score (dependent
variable). This was achieved by determining whether the relationship between a
given independent variable and the dependent variable changed when other
independent variables were added. If the other variables changed this
relationship, and if they themselves were related to the dependent variable,
they were defined as mediator variables (Baron & Kennedy 1986). The total effect of each independent
variable on the dependent variable was assessed by calculating the regression
coefficient (b) between them. A
relationship was indicated by a regression coefficient that was significantly
different from zero. If the relationship between a given independent variable,
X, and the dependent variable became insignificant after including the effect of
the mediator variable(s), this was defined as full mediation. Full mediation
implies that the effect of X on the dependent variable is determined entirely
through the mediator(s). A remaining significant effect of X after including
the effect of the mediator(s) is called a direct or residual effect, and it is
assumed to reflect the direct influence of X on the dependent variable (Chen & Hung 2016).
The indirect effect of a specific
independent variable, X, was calculated by multiplication of the regression
coefficient between X and the mediator variable, also defined as the first
stage effect, and the regression coefficient between the mediator variable and
the dependent variable, also defined as the second stage effect (Brown 1997). The method described by Preacher and Hayes (2008) for assessing
indirect effects in multiple mediator models was used. All of the necessary
coefficients were calculated via “seemingly unrelated regression” (Zellner 1962). The individual indirect effects were
calculated via the “nonlinear combinations of estimators command” in STATA (Preacher & Hayes 2008). The total indirect effect was
calculated by summing up the individual indirect effects (Brown 1997).
Two mediation models were developed. The
assessed the relationship between tinnitus loudness as measured via the visual
analogue scale (VAS) of tinnitus loudness and Hospital Anxiety and Depression
Scale-Depression subscale (HADS-D) scores. The potential mediator variables were
scores for the Hospital Anxiety and Depression Scale-Anxiety
subscale (HADS-A), Tinnitus Handicap Inventory (THI), Hyperacusis Questionnaire
(HQ), and Insomnia Severity Index (ISI). The second model assessed the relationship
between ULLmin values (average ULL at
0.25, 0.5, 1, 2, 4 and 8 kHz for the ear with the lower average ULL) and the
HADS-D score.The potential mediator variables were scores for the HQ and
Mathematical model explaining the underlying mechanism that relate tinnitus
loudness and depression
regression analysis showed a statistically significant relationship between
tinnitus loudness (b = 0.74, 95% CI: 0.55 to 0.94) (path “a” in Figure 1).
1:Multiple mediation model for the relationship between depression as measured
via the HADS-D and tinnitus loudness as measured by the VAS. Numbers in
parentheses are regression coefficients; see text for details.
analysis showed that the regression coefficients for the indirect effects of
tinnitus loudness on depression were as follows: via tinnitus handicap (path
“b”), b = 0.20 (95% CI: 0.09 to 0.32); via insomnia (path “c”), b= 0.17 (95% CI: 0.09 to 0.25); via hyperacusis (path “e”), b = 0.05 (95%
CI: 0.006 to 0.085); and via anxiety (path “f”), b = 0.31 (95% CI: 0.19
to 0.42). The regression coefficient for the total indirect effect was b= 0.73 (95% CI: 0.55 to 0.92). The direct effect of tinnitus loudness on
depression (path “d”) was not statistically significant (b = -0.04, 95% CI: -0.2 to 0.12). In summary, the relationship between tinnitus loudness and
depression was fully mediated via tinnitus handicap, insomnia, hyperacusis, and
mediation model revealed that the effect of tinnitus loudness on depression,
while significant, was fully mediated through scores for the THI, HQ, HADS-A
and ISI. This is consistent with the idea that high tinnitus loudness is
associated with tinnitus handicap, hyperacusis handicap, anxiety, and insomnia,
and these in turn lead to depression. The clinical implication for audiologists
is that for patients who suffer from tinnitus, depressive symptoms may be
alleviated if tinnitus-induced anxiety, tinnitus handicap and hyperacusis are
managed adequately, even if the self-perceived tinnitus loudness remains
unchanged. Past research has shown that although tinnitus loudness as measured
via the VAS is only minimally reduced following various forms of tinnitus
rehabilitation, THI scores typically improve (Aazh
et al. 2008; Aazh et al. 2013; Aazh & Moore 2016). This improvement
may be sufficient to reduce the severity of depression.
Mathematical model explaining the underlying mechanism that relates hyperacusis
regression analysis showed a small but statistically significant negative
relationship between the ULLmin and depression (b = -0.06, 95% CI: -0.1 to -0.03) (path “a” in Figure 2).
Figure 2: Multiple mediation model for
the relationship between depression as measured via the HADS-D and ULLmin
analysis showed that the regression coefficients for the indirect effects of
ULLmin on depression were as follows: via anxiety (path “b”), b = -0.03 (95% CI: -0.05 to -0.013); via hyperacusis
handicap (path “d”), b = -0.03 (95% CI: -0.04 to -0.014). The regression coefficient for the total
indirect effect was b = -0.06 (95% CI: -0.09 to -0.03). The direct effect of the ULLmin on depression
(path “c”) was not statistically significant (b = 0.02, 95% CI: -0.007 to 0.04). In summary, the relationship between
ULLmin and depression was fully mediated via hyperacusis handicap and
The second mediation
model assessed the relationship between ULLmin and depression. The total effect
of ULLmin values on HADS-D scores was minimal. Furthermore, the mediation model
revealed that the relationship between ULLmin and HADS-D scores was mediated by
hyperacusis handicap and anxiety rather than being a direct relationship.
Although it has been reported that people with hyperacusis often have lower
than normal ULLs in one or both ears (Tyler et
al. 2014) and 42-47% of them may also suffer from depression (Goebel & Floezinger 2008; Aazh & Moore 2017),
the mechanism that produces depression in patients with hyperacusis does not
seem to be explained by reduced ULLs. Therefore, future research should focus
on factors that might lead to depression in patients with hyperacusis.
Aazh, H., &
Moore, B. C. J. (2017). Factors Associated with Depression in Patients with
Tinnitus and Hyperacusis. American
Journal of Audiology, 26,
Andersson, G., Carlbring, P., Kaldo, V., et al. (2004). Screening of
psychiatric disorders via the Internet. A pilot study with tinnitus patients. Nord J Psychiatry, 58, 287-91.
Baron, R. M., & Kennedy, D. A. (1986). The moderator-mediator
variable distinction in social psychological research: Conceptual, strategic,
and statistical considerations. Journal
of Personality and Social Psychology 51, 1173-1182.
Brown, R. L. (1997). Assessing specific mediational effects in
complex theoretical models. Structural
Equation Modeling 4, 142-156.
Chen, L.-J., & Hung, H.-C. (2016). The indirect effect in
multiple mediators model by structural equation modeling. European Journal of Business, Economics and Accountancy 4, 36-43.
Department of Health (2009). Provision
of Services for Adults with Tinnitus: A Good Practice Guide. London UK
Department of Health.
Gander, P. E., Hoare, D. J., Collins, L., et al. (2011). Tinnitus
referral pathways within the National Health Service in England: a survey of
their perceived effectiveness among audiology staff. BMC Health Serv Res, 11,
Hu, J., Xu, J., Streelman, M., et al. (2015). The correlation of the
tinnitus handicap inventory with depression and anxiety in veterans with
tinnitus. . Int J Otolaryngol, 2015, 689375.
Kehrle, H. M., Sampaio, A. L., Granjeiro, R. C., et al. (2016).
Tinnitus annoyance in normal-hearing individuals: correlation with depression
and anxiety. Ann Otol Rhinol Laryngol,
McKenna, L., Hallam, R. S., & Hinchcliffe, R. (1991). The
prevalence of psychological disturbance in neuro-otology outpatients. Clinical Otolaryngology & Allied
Sciences 16, 452-456.
Ooms, E., Vanheule, S., Meganck, R., et al. (2012). Tinnitus
severity and its association with cognitive and somatic anxiety: a critical
study. Eur Arch Otorhinolaryngol, 269, 2327-33.
Preacher, K. J., & Hayes, A. F. (2008). Asymptotic and
resampling strategies for assessing and comparing indirect effects in multiple
mediator models. Behavioral Research
Methods 40, 879-891.
Thompson, D. M., Hall, D. A., Walker, D. M., et al. (2017).
Psychological therapy for people with tinnitus: a scoping review of treatment
components. Ear Hear, 38, 149-158.
Tyler, R. S., & Baker, J. L. (1983). Difficulties experienced by
tinnitus sufferers. Journal of Speech and
Hearing Disorders, 48, 150-154.
Zellner, A. (1962). An efficient method of estimating seemingly
unrelated regressions and tests for aggregation bias Journal of the American Statistical Association 57, 348-368.
Zigmond, A. S., & Snaith, R. P. (1983). The hospital anxiety and
depression scale. Acta Psychiatrica Scandinavica,