Fig 1.
Illustration of “Bolster compensation” and “Hold-back upregulation” hypotheses.
A–B, neural activity that supports age-related neural compensation and neural reserve. C–D, Hypothetical change in neural activities and brain-behavior correlation pattern due to the interplay between neural reserve and neural compensation.
Fig 2.
Behavioral performance of the three groups under three SNR conditions.
Violin plots and individual data points of behavioral performance. OMs performed worse than YNMs under all SNRs, but better than ONMs under SNR 8 and SNR 0. Data are available on OSF (https://osf.io/89hbn/). *** < 0.001, **
< 0.01. OMs, older musicians; ONMs, older non-musicians; YNMs, young non-musicians.
Fig 3.
TiFC of the three groups, and brain-behavior correlations.
A, left and B, right ROIs used in PPI analysis. Bilateral posterior superior temporal gyri (pSTG) were the seed ROI. Bilateral auditory dorsal stream regions, including supramarginal gyrus (SMG), supplementary motor area (SMA), superior part of precentral gyrus (PrCGsup), and speech motor areas (SM) were the target regions. C, ONMs showed upregulated task-induced functional connectivity (TiFC) in bilateral dorsal streams compared to YNMs, while OMs exhibited TiFC resembling YNMs. Violin and point plots show the individual TiFC in specific target regions. D–E, TiFC strength of OMs was negatively correlated to the behavioral performance after controlling for resting-state functional connectivity (RSFC). ROIs were mapped onto the brain surface using the BrainNet Viewer toolbox [42]. Data are available on OSF (https://osf.io/89hbn/). *** < 0.001, **
< 0.01, *
< 0.05, †
< 0.1; OMs, older musicians; ONMs, older non-musicians; YNMs, young non-musicians; L, left; R, right.
Fig 4.
Spatial alignment of TiFC to YNMs and within group.
A, Illustration of the calculation of intersubject spatial correlation; see “Methods” for details. B–C, OMs showed greater spatial alignment to YNMs than ONMs in left PrCGsup (B), and greater spatial alignment to YNMs predicted lower TiFC strength in OMs (C). D–E, Mean 3D coordinates of the voxels that showed top 10% TiFC were plotted in D. Voxels of ONMs exhibited significantly different positions in Z axis compared to YNMs, while no difference was found between OMs and YNMs (E). F, ONMs showed greater spatial correlation within the group than OMs and YNMs. Bar plots show the group mean of specific Z-axis index in L PrCGsup. Error bars indicate the SEM. Data are available on OSF (https://osf.io/89hbn/). Brain icon in panel A was sourced from: https://upload.wikimedia.org/wikipedia/commons/0/08/Wikimedia_Brand_Guidelines_Update_2022_-Brain.svg. * < 0.05, **
< 0.01; OMs, older musicians; ONMs, older non-musicians; YNMs, young non-musicians; L PrCGsup, left superior part of precentral gyrus.
Fig 5.
Bilateral RSFC of the three groups.
ONMs and OMs showed upregulated RSFC in bilateral dorsal streams compared to YNMs. Violin and point plots show the individual RSFC in specific target regions. Data are available on OSF (https://osf.io/89hbn/). ** < 0.01, *
< 0.05, †
< 0.1, OMs, older musicians; ONMs, older non-musicians; YNMs, young non-musicians; pSTG, posterior part of superior temporal gyrus; SMG, supramarginal gyrus; SMA, supplementary areas; SM, speech motor areas; PrCGsup, superior part of precentral gyrus; L, left; R, right.
Table 1.
The group mean (standard deviation) values and statistics of age, education, pure tone average (PTA) at 250–4,000 Hz, MOCA score, age of training onset, and years of music training in each group.