A dynamical disease model
of stereotyped movement disorder:
measuring complexity and regularity in motor behavior
M.H. Lewis1, KM. Newell2 and J.M. Bodfish3
1Psychiatry, University of Florida, Gainesville, FL, U.S.A.
2Pennsylvania State University, State College, PN, U.S.A.
3University of North Carolina, Chapel Hill, NC, U.S.A.
In a dynamical disease model, illness can be characterized by strikingly periodic or predictable dynamics in relevant physiological systems. Healthy systems, however, are complex, variable and difficult to predict. We have sought to apply a dynamical disease model to motor and behavioral systems in individuals with developmental disorders. Our target has been abnormal repetitive behavior, typical of individuals with autism and severe mental retardation. In testing the utility of this model, we have employed both automated and observational methods, coupled with data analysis strategies designed to provide an index of complexity or regularity.
We have assessed repetitive behavior (stereotyped body rocking) by live coding using standard observational methods, as well as kinematic analysis of videotaped body rocking motions. We have also assessed body rocking using a force platform, which measures forces and moments using a 100 Hz sampling rate. We have also measured aspects of other motor behavior in individuals with stereotyped body rocking, including spontaneous eye blink rate, as well as the dynamics of postural control involved in standing and sitting. Blink rate was assessed using standard observational techniques, whereas the dynamics of standing and sitting were assessed on the force platform.
Combining repeated measurements of behavior over time with non-linear dynamical analyses (approximate entropy, or ApEn) has yielded several findings. First, kinematic and force platform analyses of body rocking both showed that this behavior, although highly periodic, is in fact less regular or more variable than body rocking exhibited by healthy volunteers (higher ApEn). Second, compared to matched controls, individuals who engaged in stereotyped body rocking exhibited significantly lower rates of spontaneous blinking with less variable inter-blink intervals. Third, compared to matched controls, individuals who engaged in stereotyped body rocking displayed postural control deficits when standing on the force platform. These differences included significantly increased amplitude and velocity of postural sway, as well as markedly reduced complexity (lower ApEn) in the distribution of forces around a centre of pressure. Viewed as a time series, these moment-to-moment forces were highly periodic compared to control series. Finally, compared to healthy volunteers, individuals who engaged in stereotyped behavior exhibited less complexity in the dynamics of sitting (lower ApEn).
These findings suggest that the notion of dynamical diseases might be extended to behavioral disorders which, as a consequence of CNS insult, represent the loss of variability or complexity in behavior, and their replacement by highly repetitive, periodic actions. Evidence of a shift from complexity to regularity can be observed not only in the target behavior (i.e. stereotypy), but also in other motor behaviors.
Paper presented at Measuring Behavior 2002 , 4th International Conference on Methods and Techniques in Behavioral Research, 27-30 August 2002, Amsterdam, The Netherlands