TY - JOUR
T1 - Regularity and randomness in ageing
T2 - Differences in resting-state EEG complexity measured by largest Lyapunov exponent
AU - Ma, King Hang Matthew
AU - Fong, Manson Cheuk-Man
AU - Xie, Chenwei
AU - Lee, Tan
AU - Chen, Guanrong
AU - Wang, William Shiyuan
PY - 2021/12
Y1 - 2021/12
N2 - The loss of complexity in ageing hypothesis (LOCH) has found support from EEG studies, most of which adopted signal-domain complexity measures. The present study adopted the largest Lyapunov exponent (LLE) to measure complexity from a nonlinear dynamical systems perspective. A total of 144 participants were included and divided into young, young-old and old-old groups. Both sensor-space and source-space results showed significantly lower LLE for older than younger adults. The age-related differences were region-dependent, being most prominent in the frontal region, followed by bilateral temporal regions. The occipital region showed non-significant differences. Significant reduction of LLE in the posterior cingulate was also observed by virtue of the source-space analysis. We also evaluated the relationships between LLE and other complexity measures. The most intriguing result was the negative correlation between LLE and Lempel-Ziv complexity (LZC). The age-related decrease in LLE indicated a higher regularity in dynamics, while the higher LZC indicated a higher randomness in the signal domain. The new findings support the LOCH by demonstrating the simultaneous increase in regularity and randomness.
AB - The loss of complexity in ageing hypothesis (LOCH) has found support from EEG studies, most of which adopted signal-domain complexity measures. The present study adopted the largest Lyapunov exponent (LLE) to measure complexity from a nonlinear dynamical systems perspective. A total of 144 participants were included and divided into young, young-old and old-old groups. Both sensor-space and source-space results showed significantly lower LLE for older than younger adults. The age-related differences were region-dependent, being most prominent in the frontal region, followed by bilateral temporal regions. The occipital region showed non-significant differences. Significant reduction of LLE in the posterior cingulate was also observed by virtue of the source-space analysis. We also evaluated the relationships between LLE and other complexity measures. The most intriguing result was the negative correlation between LLE and Lempel-Ziv complexity (LZC). The age-related decrease in LLE indicated a higher regularity in dynamics, while the higher LZC indicated a higher randomness in the signal domain. The new findings support the LOCH by demonstrating the simultaneous increase in regularity and randomness.
KW - complexity
KW - ageing
KW - EEG
KW - largest Lyapunov exponent
KW - resting state
U2 - 10.1016/j.ynirp.2021.100054
DO - 10.1016/j.ynirp.2021.100054
M3 - Journal article
SN - 2666-9560
VL - 1
JO - Neuroimage: Reports
JF - Neuroimage: Reports
M1 - 100054
ER -