This project will examine the development of sleep state organization and physiological measures that define sleep states in normal infants and infants at risk for the sudden infant death syndrome (SIDS). Two groups at risk for SIDS, infant who later succumbed to SIDS and siblings of SIDS victims, will be compared with age-matched controls using data already collected. The objectives are based on the assumptions 1) that SIDS victims succumb from a failure of mechanisms that normally allow a state transition when the infant is exposed to a life-threatening challenge, and 2) that physiologic signs measured during different states will provide suggestions of the mechanisms that lead to failure of cardiac and respiratory systems. Twelve-hour portions of twenty-two 24-h ambulatory recordings of SIDS victims and 66 recordings of control infants will be subjected to machine classification of sleep state. The temporal characteristics of sleep-waking states, including distribution of state epoch duration, state transition probabilities, periodic organization, and relationship of state epochs to time of night and feeding will be compared between these two groups and with 12-h observer-classified recordings of SIDS siblings and age-matched controls. The temporal organization of physiological variables that define sleep states will be assessed using linear time domain filtering procedures and frequency domain spectral techniques, together with nonlinear dynamic procedures including phase-phane plots and measures of dimensionality, to determine if development in moment-to-moment activity and interactions between variables are altered in infants at risk. Relationships of moment-to- moment alterations in activity to the 3-4 h """"""""feeding rhythm"""""""" and to the circadian cycle will be quantified for each group. Assessment of phase- plane plots will make use of quadrant occupancy histogram Chi-square tests. Multivariate statistical procedures, including stepwise logistic regression and discriminant analysis, will also be used to assess differences among risk groups.
Showing the most recent 10 out of 66 publications
