Spinal cord and brain tissue impairments as long-term effects of rugby practice? An exploratory study based on T1 and ihMTsat measures

Abstract

Rugby players are subject to multiple impacts to their head and neck that could have adverse neurological effects and put them at increased risk of neurodegeneration. Previous studies demonstrated altered default mode network and diffusion metrics on brain, as well as more foraminal stenosis, disc protrusion and neck pain among players of contact sports as compared to healthy controls. However, the long-term effects of practice and repetitive impacts on brain and cervical spinal cord (cSC) of the rugby players have never been systematically investigated. In this study, 15 retired professional and amateur rugby players (R) and 15 age-matched healthy controls (HC) (all males; mean age R: 46.8 ± 7.6; and HC: 48.6 ± 9.5) were recruited both to investigate cord impairments and further characterize brain structure damage. Medical questionnaires including modified Japanese Orthopedic Association scale (mJOA) and Neck Disability Index (NDI) were filled by all participants. A 3 T multi-parametric MR protocol including conventional qualitative techniques such as T₁-, T₂-, and T₂*-weighted sequences, as well as state-of-the art quantitative techniques including MP2RAGE T₁ mapping and 3D ihMTRAGE, was used on both brain and cSC. Normalized brain WM and GM volumes, spine Overall Stenosis Score, cord cross-sectional area and regional T₁ and ihMT metrics were derived from these acquisitions. Rugby players showed significantly higher NDI scores, as well as a faster decline of normalized brain GM volume with age as compared to HC. Moreover, higher T₁ values on cSC suggestive of structural degeneration, together with higher T₁ and lower ihMTsat on brain WM suggestive of demyelination, were observed in retired rugby players as compared to age-matched controls, which may suggest cumulative effects of long-term impacts on the tissues. Metrics also suggest early aging and different aging processes on brain tissue in the players. These preliminary observations provide new insights in the domain, which should now be further investigated on larger cohorts and multicentric longitudinal studies, and further correlated to the likelihood of neurodegenerative diseases and risk factors.

Keywords: Brain; Cervical spinal cord; Inhomogeneous magnetization transfer; Neurodegeneration; Rugby; T(1) MP2RAGE.

Fig. 1

Main post-processing steps (motion correction, B₁⁺ and/or T₁ bias-correction, registration and segmentation, ROI labeling) for MP2RAGE and ihMT images on both a) SC and b) brain. (ROIs on SC: GM ant-int: anterior and intermediate; WM CST: corticospinal tracts; LST: lateral sensory tracts; RST: rubro/reticulospinal tracts; PST: posterior sensory tracts).

Fig. 2

Regression plots for (a) Normalized GM and (b) Normalized WM (on brain) volumes vs age, for Rugby players (R, blue) and Healthy Controls, (HC, orange). Age had a significant effect on the normalized GM volume in the rugby players group (p = 0.003) that was not observed in HC, nor in WM.

Fig. 3

Representative SC and brain images acquired on one retired rugby player (top: sagittal UNI MP2RAGE showing both brain and cervical cord; axial quantitative T₁ map, axial MT weighted image obtained with a dual-offset saturation (MT_dual) and corresponding ihMTsat acquired mid-brain; bottom: axial T₂*-weighted MGE, T₁ map, MT_dual and ihMTsat acquired at C2, C4 and C6 levels).

Fig. 4

(a) Mean R₁ (1/T₁) maps on SC, (b) mean ihMTsat maps on SC, (c) mean R₁ maps on brain, (d) mean ihMTsat maps on brain, for the 15 rugby players and 15 HC (sagittal and axial planes on SC, presented in the PAM50 space and axial planes on brain presented in MNI-152 template).

Fig. 5

Identification of the WM tracts where significant clusters from the PALM multi-variate analysis of R₁ and ihMTsat are located. The atlases used for cluster localization are JHU white-matter tractography (Wakana et al., 2007) and ICBM-DTI-81 white-matter labels atlases (Mori et al., 2008, Hua et al., 2008). The ROIs illustrated here are: ACR L: Anterior Corona Radiata L; PCR R/L: Posterior Corona Radiata Right/Left; RLIC R: Retrolenticular Limb of Internal Capsule; PTR R/L: Posterior Thalamic Radiation Right/Left; SLF R/L: Superior Longitudinal Fasciculus Right/left; SS R: Sagittal Stratum R; ILF R: Inferior Longitudinal Fasciculus Right; and Forceps minor.

Fig. 6

Linear regression plots and equations (along with MANOVA p-value if < 0.05) for evolution of T₁ and ihMTsat with age in GM and WM of brain and SC in R (blue) and HC (orange). The T₁ in brain and SC GM decreased for rugby players with age, contrary to HC (a, b). In brain WM of rugby players, a decrease of T₁ with age together with a moderate decrease of ihMTsat as compared to HC can also be observed (d, h). Finally, values of ihMTsat in brain and SC GM remained fairly stable with age for rugby players whereas they tend to decrease for HC. GM in brain corresponds here to the whole cortical GM; WM in SC includes PST, CST, LST, and RST regions.