DTI Analysis of White Matter Integrity and Cognitive Brain Reserve in Lifelong Musicians and Controls

,


Introduction
Neuroimaging studies of musicians and musicianship is an important focus of inquiry that began to emerge robustly in the early 2000s [2,3,4,5,6].Recent work on musicians, especially diffusion tensor imaging (DTI) analysis of white matter tracts, has contextualized imaging analysis within the framework of aging and building cognitive reserve [e.g.7,8,9,1].
Neuroimaging research focusing on professional musicians can be categorized into the following main topics: (1) the relationship of musicianship and neuroplasticity [5,3], (2) neuroanatomical correlates of musicianship [2,10], (3) enhanced encoding of vowels and speech in professional musicians [11], (4) timbre-specific auditory cortical representations in musicians [12], (5) shared networks for auditory and motor processing in professional musicians [4], and (6) brainstem recordings of speech sounds in musicians [10].The above studies have been investigated using various functional and structural magnetic resonance imaging techniques (MRI).Resting-state functional MRI (rfMRI) focusing on musicians began to emerge more recently and have been limited to studies of improvisation [13,14,15,16,17 for some examples), different types of musical training [18] and more general musical creativity [19].These studies apply regions of interest (ROI) seed-based analysis approaches.
and genetically based, and includes different definitions, including: (1) "The neurological brain reserve hypothesis proposes that individuals generally differ in the numbers of neurons and synapses available to be lost before clinical symptoms emerge" [21].
(2) "Brain reserve refers to 'passive' factors (e.g., brain volume, synapse count) that confer a particular capacity to endure neuropathological processes until a critical threshold is reached, after which cognitive and functional impairments are expressed" [22].
In contrast, cognitive reserve, also called behavioral brain reserve, is acquired through specific sensory-motor activities that span across the life cycle (including but not restricted to musicianship and bilingualism).This type of cognitive brain reserve is considered "resilience to neural insult" and strategies that strengthen alternative FNs across the life cycle can improve tolerance of atrophy [23].Moreover, higher cognitive reserve should require more structural decline in order for symptoms to manifest [24].For more discussion on cognitive reserve, see [25,26,27,28,9,1].
In the current work, we will examine white matter (WM) integrity and in particular fractional anisotropy (FA) values in healthy subjects who are high proficiency musicians and in age and educational matched non-musician controls.This is a direct extension of our original analysis [1] where only high proficiency, lifelong musicians were analyzed.The mechanisms for the different types of brain and cognitive reserve are poorly understood, but may include enhanced generation of neuronal, dendritic and synaptic connections and functional reorganization of neural networks across multimodalities.For example, [29] argue in favor of a shift to subcortical/posterior region-based networks in bilinguals, as opposed to more frontal regions in monolinguals [30].focuses on increased gray matter (GM) in bilingualism as an example of cognitive reserve and potential protective effects in aging.Thus, while cognitive brain reserve may play an important role in delaying the appearance of symptoms of certain pathologies, it also may play a role in healthy subjects.And that is the focus of our current study -lifelong musicians and potential effects of musicianship on subcortical WM matter fiber tracts.While we do not consider GM in this paper, we do not exclude the importance of cognitive reserve in GM volume changes [31].

Research question and hypotheses
The published research on cognitive reserve, the potential interactions between bilingualism and musicianship, and studies specifically focusing on musicianship provide the foundation for our research question and hypotheses.It has been shown in multiple studies that WM integrity decreases in normal aging [32,33,34,35,36].
Research focusing on cognitive reserve has noted that the process of WM integrity loss can be slowed down or changed by lifelong bilingualism and musicianship.Our first paper published on DTI study of lifelong musicianship [1] hypothesized that there will be an increase in WM integrity in certain subcortical fiber tracts in lifelong musicians, and that this increase will be reflected bilaterally in FA values for the bilateral superior longitudinal fasciculus (SLF) and the uncinate fasciculus (UF) -2 specific WM tracts shown to be relevant in musicianship.We also hypothesized that the bilateral SS, which includes the IFOF (a tract important in bilingualism), may not show the same effects.The current study considers specifically lifelong musicianship with agematched non-musician controls and the age span included is from 20 to 67 years and hypothesizes that the age-matched non-musician controls will not show the increases in WM integrity across the age span found in lifelong musicians for bilateral SLF and bilateral UF.

Participants
Sixteens subjects (eight musicians and eight non-musician controls) gave written consent in accord with the guidelines of Duke Health Institutional Review Board and were scanned in 2019-2021 at Duke University Hospital at the Duke Brain Imaging and Analysis Center: Eight musicians (five females and three males between the ages of 20 and 67 years, with a mean age of 44.1 years) and eight non-musician healthy controls (five females and three males between the ages of 20 and 63 years, with a mean age of 43 years).Both groups (musicians and controls) have college degrees that span the baccalaureate (completed or in progress) through the MA/MS and PhD, and 1 Associate's degree.
All eight musicians began musical training between the ages of three and twelve years, at an average of 6.4 years.The number of years the subjects had been musicians was a correlate of chronological age since all were currently active musicians, performing regularly with a mean of 38 years of active musicianship.Musicians reported an average of three hours per day rehearsing/practicing and an average of nine hours per day in peak times (i.e. during periods of performance).Additional credentials included college degrees in music (including undergraduate minor, master's and doctorate) and multiple affiliations with orchestra, symphony or operatic theatre in the United States and abroad.Five of the professional musicians have extensive experience in the United States and abroad.All subjects played at least the piano or violin.The eight non-musicians control subjects do not play any instruments and do not sing in any groups.Four of the eight have never played a musical instrument, three played in high school between 14 to 18 years of age, and one played for six months at the age of 14.This yields a compare the slope difference between the two groups for each white matter tract of interest using a test for parallelism [43], followed by application of the Fisher's method for pooling p-values, a statistical method for combining results from multiple tests into one test statistic.
Fisher (1025) showed that the test statistic has the chisquared distribution with 2k degrees of freedom, where k is the number of p-values being pooled.The intuition is that if one has several independent experiments, each of which modestly supports the alternative hypothesis, then by combining the weak signals one can increase power for rejecting the null.Previous studies of lifelong musicianship showed that relevant subcortical white matter fiber tracts show an increase in FA values.According to our research hypothesis, WM integrity may increase over the life span in musicians in bilateral SLF and UF, but not in others.

Results
mean of 35 years since any of the controls were involved in musical activity.

Data Behavioral data
All 16 participants (8 musicians and 8 non-musicians) completed an extensive background questionnaire on their education, musicianship and knowledge of languages.

Image acquisition
All subjects were scanned in the same scanner at Duke University Hospital, using a GE 3T Discovery 750 MRI (General Electric, Milwaukee, WI) with an 8-channel head coil.Diffusion weighed images (DWIs) were collected using an oblique single-shot spin-echo echo-planar imaging (EPI) sequence with a repetition time (TR) = 9,000 ms, echo time (TE) = 93 ms, having a 256 × 256 image matrix over a 25.0 cm2 FOV, covering 69 interleaved axial slices each 2 mm thick.25 diffusion directions were uniformly distributed in 3D space with a b = 1000 s/mm2, together with an acquisition with b = 0 s/mm2.A T2 and proton density weighted sequence (FRFSE-XL) with a TR = 3,000 ms, TE = 90 ms, 256x256 matrix, 24.0 cm2 FOV and 5 mm slice thickness was used to acquire 28 interleaved axial slices.

DTI processing
All images were processed using Tortoise 3.1.4[37,38] and FMRIB Software Library [39,40,41].The raw DTI were corrected for distortion, eddy currents and motion, using the DIFFPREP command in Tortoise and subject specific T2 weighted image.A brain mask was then generated from the distortion corrected B0 using ExtractImage and bet2.All Tortoise quality control outputs were inspected for signal dropouts, excessive motion, and B0 and T2 registration accuracy.Then, the Tortoise commands EstimateTensorNLLSRESTORE, with the RESTORE option and ComputeFAMap generated the fractional anisotropy maps.Next, FSL was used to normalize the individual FA maps to MNI space, using the FLIRT and FNIRT commands.Finally, the FA for each ROI in each subject were averaged, using the commands fslmaths and fslstats.Using the ICBM-81 atlas [42] FA values in the following regions of interests were computed: left and right superior longitudinal fasciculus (LSLF and RSLF), left and right sagittal striatum (LSS and RSS), and left and right uncinate fasciculus (LUF and RUF).

Statistical methods
In order to test our hypotheses concerning changes in WM integrity in aging musicians in the original 2021 study, we present linear regression calculations of the six regions of interest (LSLF, RSLF, LUF, RUF, LSS, RSS).Additionally, we present linear regression calculations of the same six regions of interest for both musicians and non-musician controls (Figure 2), and all slope regression analysis was conducted in SAS (SAS Institute Inc., Cary, NC, USA).We A simple linear regression shows the following FA values for bilateral SLF and UF in musicians.As shown below, there is greater WM integrity in aging musicians in the bilateral SLF and bilateral UF.
Figure 2 shows the linear regression of the following FA values for bilateral SLF, UF and SS in musicians and age and education matched non-musician controls.First, we used the test for parallelism [43] to calculate the significance of how much higher the musicians' slopes were than the controls' slopes for each of the four white matter tracts (LSLF, RSLF, LUF and RUF).Then we pooled the tracts, using Fisher's method, resulting in a chi-squared distribution with 8 degrees of freedom (-2 Σ ki=1,4 ln(pi) = 21.255), the resulting pooled p-value is 0.0065.
Based on previous research on cognitive brain reserve using DTI, we hypothesized that we would find increased WM integrity in FA values of two important tracts, SLF and UF, in lifelong musicians, but not in non-musician controls.We also hypothesized that the bilateral SS, a tract important for bilingualism, would not be impacted by musicianship and show loss as seen in healthy aging.Following our hypothesis, the superior longitudinal fasciculus (SLF) and uncinate fasciculus (UF) tracts show a positive correlation between FA and age in subjects with high musical proficiency, while FA decreases with age in the sagittal strata (SS) in the same subjects and, thus, the SS may be unrelated to musical proficiency.Therefore, we focus on the 4 WM tracts from our original hypothesis (bilateral SLF and UF).Using Fisher's method, resulting in a chi-squared distribution with 8 degrees of freedom (-2 Σ ki=1,4 ln(pi) = 21.255), the resulting pooled p-value is 0.0065.The results are in keeping with our hypothesis.In future analysis, we will explore the WM integrity and FA values of bilateral SS and IFOF in bilinguals and multilinguals, as well as possible cross-modality benefits between musicianship and multilingualism.

Discussion
The bilateral SLF and bilateral UF have been noted in the literature on cognitive reserve and bilingualism [44,45,46,301]while only the FA values of the bilateral SLF has been noted in the literature on cognitive reserve in musicianship, specifically with regard to relative and absolute pitch [8].Halwani et al. [7] focuses primarily on increased FA values in bilateral AF in musicians, including instrumentalists and singers.In terms of the function of these white matter tracts in processing language, the dorsal-ventral pathway is often mentioned, but without further differentiation or explanation [31].In studies of cognitive reserve and healthy aging in musicians, there are no studies that focus specifically on the UF.The lesion-deficit DTI literature includes research involving musicians and music-based treatments identify the UF as important in pitch perception in healthy subjects, and also important in music therapy.There is a larger body of lesion-deficit studies that consider the bilateral UF and FA values, that address visual memory delay [47,48,49] naming impairment [50,51,52] and psychopathy [53].One possible point of interest emerges from the work of [54] on the possible role of the IFOF and UF in language processing with patients undergoing surgery (left anterior temporal lobe or orbito-frontal region).They were not able to confirm or reject that the UF is essential in language processing.While our scans did not show any systematic scanner-related errors, there is some research suggesting that all diffusion imaging may require additional correction [55,56].Duke University BIAC is part of the Biomedical Informatics Research Network (BIRN) that examined potential system biases across the vendors, and have taken measures to address any concerns about systematic errors in DTI acquisition in DUHS scanners [57].The present study focuses on how musicianship may affect changes to WM integrity across the life cycle, and change the trend of reduction in FA structures that are relevant in musicianship, specifically the SLF and UF.Previous research has shown that FA values in the SLF and UF show a significant decrease in FA values with age [32,58,59,60,33,33,34,35,36].
Westlye et al. [34] examined age changes in WM integrity in 430 healthy subjects between the ages of 8 and 85 years.The SLF and UF are included bilaterally in their analysis of 7 major WM tracts.Their findings show a significant (p < 0.0001) decrease in FA values across whole brain WM fiber tracts and the maximum FA values were found at 29.1 years of age (2010: 2058-60).The age of maxima for SLF and UF were 28.8 and 28.6 respectively (2010: 2061).Rathee et al. [36] also examined FA values for 177 healthy subjects and divided the subjects into three age groups: 20-40, 41-60, 61-85.Whole brain values show a consistent decrease in FA values across each of the three age groups, and voxel-wise FA values include a significant decrease in 22 regions (mid to oldest group), 26 regions (youngest to oldest group), and 4 regions (youngest to mid group).Furthermore, all FA values were lower in all regions between the oldest and youngest groups, and explicit reference to the SLF and SS showed significantly lower FA values in the oldest to youngest groups (2016: 14-15).As mentioned earlier, the research focusing on cognitive reserve has shown that the process of WM integrity loss in normal aging may be slowed or changed by lifelong bilingualism and musicianship.Our study is unique in its requirement for active musicianship across the life cycle.Our hypotheses looked specifically at the bilateral SLF and the bilateral UF in lifelong musicians and age-matched/ education-matched non-musician controls.While previous studies on musicianship did not focus on the UF, given that it is one of the tracts with a later period of maturation and connects temporal and frontal lobe structures, it was included in our analysis.Our results support the research on cognitive reserve and show that the FA values for bilateral SLF and bilateral UF were greater in older musicians, but not in agematched non-musician controls.As shown in the application of Fisher's method, the resulting chi-squared distribution yields a pooled p-value = 0.0065.These findings suggest that musicianship across the life cycle may change the expected decrease in FA values in these two subcortical tracts.A third tract, the bilateral SS, which includes the IFOF -a tract important in bi-and multilingualism, did not show the same effects.However, the decrease in FA values in bilateral SS was less in lifelong musicians than that found in the nonmusician controls.

Limitations and Future Directions
Future directions include replication of results across a larger data set, investigation of the functional activations and connectivity through fMRI, and other direct comparisons of musicians and bi/multilinguals.

Conclusion
The present study has focused on lifelong professional musicians and age and education-matched non-musician controls.Our findings suggest that lifelong musicianship may contribute to WM integrity in the bilateral SLF and bilateral UF in aging.While earlier research identified higher FA values in the right SLF [64], our findings show increases in bilateral SLF.The identification of higher FA values in bilateral UF is also interesting and expands the examination of subcortical WM tracts that may be affected by lifelong musicianship.This study is an additional contribution to the growing body of research on behavioral cognitive reserve and suggests additional evidence that highly proficient musicianship may produce similar effects of higher FA values in certain WM tracts, while differing in others.

Figure 1 :
Figure 1: The superior longitudinal fasciculi (SLF) and uncinate fasciculi (UF) are related to musical proficiency and age.The sagittal striata (SS) are also related with age, but not musical proficiency.The SLF (red), UF (magenta) and SS (blue) are depicted in accordance with the ICBM-81 atlas [42].The tracts are superimposed on a T1 weighted image in the Montreal Neurological Institute 152 space.The figure uses the left-posterior-inferior convention.Legend: L, left.

Figure 2 :
Figure 2: Scatter plots and linear fits between FA values and subjects' ages for musicians (in red) and controls (in blue).The difference between the musician and non-musicians' slopes by region have the following p-values: LSLF 0.094, RSLF 0.101, LUF 0.023, RUF 0.111, LSS 0.212, RSS 0.034.The superior longitudinal fasciculus (SLF) and uncinate fasciculus (UF) tracts (top two rows) show a positive correlation between FA and age in musicians, but a negative correlation between FA and age in non-musicians.

Citation:
Edna Andrews, Cyrus Eierud, David Banks, Todd Harshbarger, Andrew Michael and Charlotte Rammell.DTI Analysis of White Matter Integrity and Cognitive Brain Reserve in Lifelong Musicians and Controls.Journal of Psychiatry and Psychiatric Disorders.7 (2023): 80-88.

Table 1 :
Mean fractional anisotropy (FA) values for each subject (controls in blue, musicians in red) in specific ICBM-81 brain regions *.