Background: The construct of Hidden Hearing Loss (HHL) proposes a link between patient-reported hearing difficulties and underlying neural deficits not captured by the standard audiogram. However, the heterogeneity of this population challenges the utility of HHL as a unitary diagnosis. This study presents an exploratory analysis aimed at deconstructing the HHL symptom complex. Methods: In 30 participants with a range of hearing abilities and complaints, we measured binaural unmasking using the Binaural Intelligibility Level Difference (BILD). We employed a two-stage analysis. First, a "lumping" analysis tested whether participants could be grouped into a unitary "HHL profile" that predicted a BILD deficit, using both theory-driven classification and data-driven clustering. Second, after this approach failed, a pre-planned exploratory "splitting" analysis used a Linear Mixed-Effects Model (LMM) to investigate whether individual clinical markers (tinnitus, self-reported speech difficulty) were independently associated with the BILD. Results: The "lumping" analyses failed to find a significant difference in the BILD between subgroups, questioning the utility of a unitary HHL profile. In contrast, the exploratory "splitting" analysis found a significant interaction between tinnitus and listening condition (β = 1.57, p = 0.009), suggesting that participants with tinnitus exhibited a smaller BILD. The complaint of speech perception difficulty was not significantly associated with a BILD deficit (p = 0.086) but was associated with lower scores on a test of short-term memory (forward digit span, p = 0.046). Conclusion: Our findings challenge the value of a unitary HHL profile for predicting this specific binaural deficit. Instead, our exploratory analysis generated a specific, testable hypothesis of a sensory-cognitive dissociation: in our sample, tinnitus was associated with a reduced capacity for binaural unmasking, while the complaint of speech difficulty was associated with poorer short-term memory. These preliminary findings, derived from post-hoc analysis of an underpowered study, require rigorous validation in larger, pre-registered studies.
Bleeck et al. (Fri,) studied this question.