Evidence for cervical cord tissue disorganisation with aging by diffusion tensor MRI

Abstract

This study investigated the influence of normal aging on cervical cord volumetry and diffusivity changes and assessed whether magnetic resonance imaging (MRI) abnormalities of the aging cervical cord and brain are associated. Conventional and diffusion tensor (DT) MRI of the brain and cervical cord were acquired from 96 healthy subjects (age range = 13–70 years). Cross-sectional area, mean diffusivity (MD) and fractional anisotropy (FA) of the cervical cord were measured. Volumetry and diffusivity metrics were also obtained for the brain white matter (WM) and grey matter (GM) (overall and cortical). No cervical cord lesions were seen on conventional MR images from all subjects. Degenerative vertebral column changes (not associated to cord compression) were found in 41 subjects (43%). Average FA of the cervical cord, but not average MD and cross-sectional area, was correlated with age (r = − 0.70, p < 0.001). Additionally, T2 brain lesion volume, normalised brain volume (NBV), normalised global and cortical brain GM volumes and average MD of the brain GM and WM also correlated with age (r values ranging from − 0.83 to 0.62). Only brain WM average FA was weakly correlated with cervical cord average FA (r = 0.25, p = 0.02). The final multivariate model retained cord average FA (r = − 0.37, p < 0.001), normalised cortical GM volume (r = − 0.56, p < 0.001) and NBV (r = − 0.22, p = 0.04) as independent correlates of age (r² = 0.76). Cervical cord is vulnerable to aging. The decrease of FA, in the absence of atrophy and MD changes, suggests gliosis as the most likely pathological feature of the aging cord.

Introduction

Normal aging of the brain is accompanied by a set of structural, metabolic and functional abnormalities, involving both the grey matter (GM) and the white matter (WM) (Kemper, 1994, Raz, 2000). Incidental WM hyperintensities (WMHs) are visible on conventional magnetic resonance (MR) images of the brain in about 30% of healthy subjects older than 60 years, and their prevalence rises steadily with increasing age (de Leeuw et al., 2001). In addition, post mortem (Terry et al., 1987, Meier-Ruge et al., 1992, Tang et al., 1997, Marner et al., 2003) and in vivo MR imaging [MRI] (Salat et al., 1999, Good et al., 2001, Jernigan et al., 2001, Ge et al., 2002a, Liu et al., 2003, Resnick et al., 2003, Rovaris et al., 2003, Walhovd et al., 2005, Benedetti et al., 2006a) studies have provided convincing evidence of age-related loss of brain total (Resnick et al., 2003, Rovaris et al., 2003, Benedetti et al., 2006a), GM (Good et al., 2001, Jernigan et al., 2001, Ge et al., 2002a, Liu et al., 2003, Resnick et al., 2003, Walhovd et al., 2005, Benedetti et al., 2006a) and WM (Salat et al., 1999, Good et al., 2001, Jernigan et al., 2001, Ge et al., 2002a, Liu et al., 2003, Resnick et al., 2003, Walhovd et al., 2005, Benedetti et al., 2006a) volumes. Finally, the application of modern quantitative MR techniques allowed to grade the extent of tissue damage associated to WMHs (Wong et al., 1995, Tanabe et al., 1997, Fazekas et al., 2005) and to show that also the normal-appearing (NA) GM and WM undergo progressive structural changes with aging, which would go undetected when using conventional MRI (Nusbaum et al., 2001, Abe et al., 2002, Ge et al., 2002b, Rovaris et al., 2003, Salat et al., 2005, Benedetti et al., 2006a).

Post mortem studies have shown that, in addition to the brain, the spinal cord tissue is also not spared by age-related pathological changes (Morrison, 1959, McComas et al., 1973, Tomlinson and Irving, 1977, Terao et al., 1996, Cruz-Sanchez et al., 1998), including neuronal loss (Morrison, 1959, McComas et al., 1973, Tomlinson and Irving, 1977, Terao et al., 1996, Cruz-Sanchez et al., 1998), neuronal inclusions and spheroids (Cruz-Sanchez et al., 1998), WM fibre bundle degeneration (Morrison, 1959, Cruz-Sanchez et al., 1998) and a proliferation and hypertrophy of astrocytes (Cruz-Sanchez et al., 1998). Despite this, the assessment of in vivo age-related changes of the cord using MRI is still limited. In detail, whereas there is a general agreement that spinal cord hyperintense lesions develop rarely with aging (Thorpe et al., 1993, Lycklama à Nijeholt et al., 1997), it is still unclear whether aging is associated to loss of cord tissue since cord atrophy in the elderly was reported by some (Ishikawa et al., 2003), but not by other (Sherman et al., 1990, Yanase et al., 2006) investigators. Furthermore, to the best of our knowledge, no study has assessed whether aging is also accompanied by intrinsic structural changes of the cord.

Diffusion tensor (DT) MRI enables the random diffusional motion of water molecules to be measured, thus providing information about the integrity and structural arrangement of the brain tissues in patients with various neurological conditions (Mori and Zhang, 2006). DT MRI has also been used to study the elderly brain (Nusbaum et al., 2001, Abe et al., 2002, Rovaris et al., 2003, Salat et al., 2005, Benedetti et al., 2006a). More recently, technical advancement has made possible to acquire good quality DT MRI of the cervical cord, which was shown to be able to grade the extent of cord damage in patients with demyelinating (Agosta et al., 2005, Valsasina et al., 2005, Benedetti et al., 2006b) and neurodegenerative (Valsasina et al., 2007) diseases.

Against this background, we obtained conventional and DT MRI of the cervical cord from a large cohort of healthy individuals, spanning seven decades of life, in order to investigate the influence of normal aging on cervical cord cross-sectional area and diffusivity metrics. We also quantified volumes and diffusivity properties of the brain tissue to assess whether age-related changes of the cervical cord correlate with the concomitant development of brain damage.