DESIGN: This was a cross-sectional study. Three hundred forty-two members of the public with tinnitus volunteered to complete a survey comprising a series of questionnaires and subscales of questionnaires measuring each of the constructs contained within the Cognitive Behavioral Model of Tinnitus Distress. The optimum factor structure of each measure for the study population was established, and the resulting factors were used to construct a series of path models based on the theoretical model. Path analysis was conducted for each of these, and the goodness of fit of the models was assessed using established fit criteria.
RESULTS: Five of the six path models tested reached the threshold for adequate fit, and further modifications improved the fit of the three most parsimonious of these. The two best-fitting models had comparable fit indices which approached the criteria for good fit (Root Mean Square Error of Approximation = 0.061, Comparative Fit Index = 0.984, Tucker Lewis Index = 0.970 and Root Mean Square Error of Approximation = 0.055, Comparative Fit Index = 0.993, Tucker Lewis Index = 0.982). They differed principally in the placement of tinnitus magnitude and the inclusion/noninclusion of control beliefs.
CONCLUSIONS: There are theoretical arguments to support both a beliefs-driven and a loudness-driven model, and it may be that different configurations of the Cognitive Behavioral Model of Tinnitus Distress are more appropriate to different groups of people with tinnitus. Further investigation of this is needed. This notwithstanding, the present study provides empirical support for a model of tinnitus distress which provides a clinical framework for the development of more effective psychological therapy.