Fractal regulation and incident Alzheimer's disease in elderly individuals

Figures

• 

  Fig. 1.. Fractal correlations in motor activity fluctuations. (A-B) Representative activity recordings of two ~ 80-year-old female subjects. Subject corresponding to (B) developed Alzheimer’s and the other corresponding to (A) did not. Gray shaded area indicates 9 PM-7 AM. (C) The corresponding detrended fluctuation analysis results of the signals in (A) and (B). The fluctuation function F(n) is shown versus time scale, n, in log-log scale, where F(n) is log-linearly fitted in the region of 1.25–90 min. The slope of the fitting line is defined as FR metric α. The value of α was lower for the signal in (B), indicating more random activity fluctuations. Abbreviation: FR, fractal regulation.

• 

  Fig. 2.. The distribution of FR metric α. The 10th percentile (“low”), median, and the 90th percentile (“high”) are highlighted in red line, black dash line, and blue dotted line, respectively. Abbreviation: FR, fractal regulation.

• 

  Fig. 3.. Predicted cumulative hazards of developing Alzheimer’s dementia and MCI. Panel (A) shows the risk of Alzheimer’s dementia for three representative participants with same age (81 years old), sex (females), and education (15 years), but with high (the 90th percentile at α = 0.98), median (α = 0.92), and low (the 10th percentile at α = 0.85) α levels. Panel (B) shows the risk of MCI for the same three illustrative participants. Abbreviation: MCI, mild cognitive impairment.

• 

  Fig. 4.. Predicted trajectory of global cognitive score. Shown the model-derived global cognitive decline for three illustrative participants with low α (the 10th percentile), median α, and high α (the 90th percentile).