Slide Slam O4
A systematic review and meta-analysis of speech-in-noise perception in musicians
Elisabeth Maillard1,2,3, Marilyne Joyal1, Micah Murray3,4,5, Pascale Tremblay1,2; 1Centre de recherche CERVO, Quebec city, 2Université Laval, Faculté de Médecine, Département de Réadaptation, Quebec City, 3The Laboratory for Investigative Neurophysiology (The LINE), Department of Radiology, Lausanne University Hospital and University of Lausanne, Lausanne, 4CIBM Center for Biomedical Imaging, Lausanne, 5Department of Hearing and Speech Sciences, Vanderbilt University, Nashville
Introduction. The ability to process speech in noise (SPiN) is essential for everyday communication. The literature shows that SPiN is particularly challenging in energetic maskers, speech maskers with various talkers, especially with 2-3 talkers, and when the signal-to-noise ratio (SNR) is negative. Furthermore, a negative effect of ageing on SPiN performance, particularly in speech maskers, is well documented. Indeed, SPiN perception declines with age, which negatively impacts communication-mediated activities, including social participation. It is therefore crucial to find strategies to mitigate this decline. Because it involves cognitive and auditory training and is associated with brain plasticity across multiple functional systems, musicianship has raised interest as a potential prevention/rehabilitation tool for SPiN performance decline. Yet, the literature on the potential benefits of musicianship on SPiN performance is mitigated, with some studies showing a musician advantage and others not. We therefore conducted a systematic review of the literature and performed the first meta-analyses of the effect of musicianship on SPiN performance, including an evaluation of the effect of age. We hypothesized that musicians would outperform non-musicians in conditions known to be challenging (energetic maskers, speech maskers with various talkers, negative SNR). We also expected to observe a greater advantage in older than in younger musicians, especially in speech maskers. Method. 49 articles comparing the performance of musicians and non-musicians in ≥1 SPiN behavioural task were selected after a literature search on PubMed and PsycNet. Study quality was assessed using the QualSyst tool (Kmet et al., 2004). The participants’ demographic information, their musical experience and the characteristics of the SPiN tasks were extracted. Eight separate random-effect meta-analyses were performed for the following SPiN conditions: speech in energetic maskers, speech in speech maskers (1-talker, 2-talker and 4-talker masker), speech in spatially separated masker and different SNR levels (< 0 dB, 0 dB > 0 dB). The participants’ mean age was used as a continuous moderator. Results. Significant effect sizes in favour of musicians were found for speech in energetic maskers, speech in 2- and 4-talkers and for stimuli at SNR < 0 db and 0 dB. No significant effects were found with 1-talker maskers, spatially separated maskers and SNRs > 0 dB. A significant positive effect of age was found in the energetic and SNR = 0 dB analyses but could not be robustly and consistently assessed due to the paucity of studies on adults aged > 55 years (n=3). Conclusion. This is the first systematic review and meta-analysis on the effect of musicianship on SPiN performance. Our results based on the existing literature are compelling - musicianship appears to have a significant effect in contexts producing significant masking in non-musicians. This pattern of results supports an advantage for musicians on SPiN performance. Further studies, especially on older participants, are needed to confirm and extend the present conclusions on the effectiveness of training methods based on musical skills for SPiN perception.