Activity/rest cycle and disturbances of structural backbone of cerebral networks in aging

Highlights

• Sleep fragmentation is related to WM integrity of the aging brain.

• A disturbed 24-h cycle is associated with alterations of WM integrity.

• Activity/rest cycle is not associated with age-related GM atrophy.

• Activity/rest cycle alterations may contribute to age-related WM cerebral frailty.

Abstract

Objective

Although aging is associated with alterations of both activity/rest cycle and brain structure, few studies have evaluated associations between these processes. The aim of this study was to examine relationship between activity/rest cycle quality and brain structural integrity in aging subjects by exploring both grey and white matter compartments.

Material and methods

Fifty-eight elderly subjects (76±0.5 years; 41% female) without dementia, sleep disorders and medications were included in the analysis. Actigraphy was used to measure parameters of activity/rest cycle (24-h amplitude, 24-h fragmentation and 24-h stability) and sleep (total sleep time and sleep fragmentation) over a minimal period of 5 days. Whole brain linear regression analyses were performed on grey matter volumes maps using voxel based morphometry and on white matter integrity using tract based statistics analyses.

Results

A lower 24-h amplitude and a higher sleep fragmentation were independently associated with a reduction of white matter integrity in models including age and gender as covariates. The association between 24-h amplitude and white matter integrity decreased but remained significant in a model accounted for sleep fragmentation, indicating a specific effect of 24-h cycle disturbances. No association with grey matter volumes was observed.

Conclusion

In elderly, not only sleep but also 24-h cycle disturbances were associated with altered structural connectivity. This alteration of structural backbone networks related to activity/rest cycle disturbances in aging might constitute a cerebral frailty factor for the development of cognitive impairment.

Introduction

Knowledge on activity/rest cycle and sleep disturbances in aging mainly results from subjective self-reported questionnaires (Buysse et al., 1989, Johns, 1991), easily applied in large epidemiologic or cohort-based investigations. However, these questionnaires are subject to the influence of perception and several studies have illustrated their relative inaccuracy compared to objective measures (polysomnography, PSG), in middle age and older individuals (Buysse et al., 2008, Buysse et al., 1991). More recently, technological advances have led to the development of mobile activity monitoring systems (actigraphy), a non-invasive technique easy to apply in large samples. Actigraphy has been used to objectively assess global activity/rest cycle including measures of daytime activity and sleep over a long-term period in a natural environment. Previous actigraphic studies have demonstrated that aging is not only related to sleep alterations but also to activity/rest cycle disturbances including fragmentation and a reduced amplitude of the cycle (Huang et al., 2002, van Someren et al., 1993). Evidence suggest that activity/rest cycle and sleep disruptions are related to cognitive deficits (Lim et al., 2013, Lim et al., 2012, Luik et al., 2015, Oosterman et al., 2009, Tranah et al., 2011, Walsh et al., 2014) and may represent earliest signs of neurodegenerative diseases (Wulff et al., 2010).

Although post-mortem and magnetic resonance imaging (MRI) investigations have reported widespread age-related changes in brain structure within grey (GM) and white matter (WM; Lockhart and DeCarli, 2014), few studies have evaluated associations between activity/rest cycle and brain modifications. GM atrophy, either widespread (Sexton et al., 2014) or circumscribed to the inferior prefrontal cortex (Branger et al., 2016, Lim et al., 2016), has been associated with disturbed sleep quality assessed through self-reported and actigraphic measures. Concerning the structural state of WM, research have shown that activity/rest cycle disturbances (fragmentation and reduction of amplitude) were related to WM integrity alterations assessed through WM hyperintensities (WMH; Oosterman et al., 2008; Zuurbier et al., 2015), suggesting an association between activity/rest cycle and cerebrovascular changes.

Diffusion Tensor Imaging (DTI) enables indirect investigation of WM microstructure via water diffusion within brain tissues and appears as a sensitive technique to describe age-related modifications of WM (Sullivan and Pfefferbaum, 2006). During aging, a consistent decrease of fractional anisotropy of water diffusion was described in the major WM tracts indicating a decrease of their anatomical coherence (Sullivan and Pfefferbaum, 2006). Moreover, recent research indicate that DTI parameters are modified before the appearance of WMH and therefore would reflect the early phase of WM changes (Maillard et al., 2013, Pelletier et al., 2015). Whereas age-related modifications of WM have been extensively investigated with DTI, relationship between DTI modifications and actigraphic parameters in aging have not been explored yet.

The present study aims to investigate associations of activity/rest cycle assessed through actigraphy with GM atrophy and WM integrity in a population of elderly. Whole brain analyses will be conducted on GM volumes using voxel based morphometry pipeline and on WM integrity using tract based statistics analyses.