Quantifying fractal dynamics of human respiration: age and gender effects

Ann Biomed Eng. 2002 May;30(5):683-92. doi: 10.1114/1.1481053.


We sought to quantify the fractal scaling properties of human respiratory dynamics and determine whether they are altered with healthy aging and gender. Continuous respiratory datasets (obtained by inductive plethysmography) were collected from 40 healthy adults (10 young men, 10 young women, 10 elderly men, and 10 elderly women) during 120 min of spontaneous breathing. The interbreath interval (IBI) time series were extracted by a new algorithm and fractal scaling exponents that quantify power-law correlations were computed using detrended fluctuation analysis. Under supine, resting, and spontaneous breathing conditions, both healthy young and elderly subjects had scaling exponents for the IBI time series that indicate long-range (fractal) correlations across multiple time scales. Furthermore, the scaling exponents (mean +/- SD) for the IBI time series were significantly (p < 0.03) lower (indicating decreased correlations) in the healthy elderly male (0.60 +/- 0.08) compared to the young male (0.68 +/- 0.07), young female (0.70 +/- 0.07), and elderly female (0.67 +/- 0.06) subjects. These results provide evidence for fractal organization in physiologic human breathing cycle dynamics, and for their degradation in elderly men. These findings may have implications for modeling integrated respiratory control mechanisms, quantifying their changes in aging or disease, and assessing the outcome of interventions aimed toward restoring normal physiologic respiratory dynamics.

Publication types

  • Comparative Study
  • Research Support, Non-U.S. Gov't
  • Research Support, U.S. Gov't, P.H.S.

MeSH terms

  • Adult
  • Age Factors
  • Aged
  • Aged, 80 and over
  • Algorithms*
  • Emotions / physiology
  • Female
  • Fractals*
  • Humans
  • Lung Volume Measurements / methods
  • Male
  • Models, Biological*
  • Models, Statistical*
  • Nonlinear Dynamics
  • Respiration
  • Respiratory Mechanics / physiology*
  • Sex Factors
  • Signal Processing, Computer-Assisted*
  • Statistics as Topic