Development and Temporal Dynamics of Repetitive Behavior

Background: Although restricted, repetitive behaviors are diagnostic for autism relatively little is known about the development of such behaviors. Developmental patterns derived from longitudinal studies have not been reported and little is known about the trajectories of repetitive behaviors in atypical versus normative development. Moreover, the temporal structure of repetitive behavior has received scant attention although the temporal dynamics of repetitive behavior and developmental changes in such dynamics may prove critical to understanding neuroadaptations that mediate repetitive behavior. Appropriate animal models could provide a wealth of information about how repetitive behaviors develop including age-dependent changes in temporal organization. Our lab has employed a deer mouse (Peromyscus maniculatus) model that involves spontaneous expression of high rates of repetitive motor behavior.

Objectives: To characterize developmental trajectories of repetitive behavior in deer mice and model its temporal organization. We also sought to determine if temporal structure changed systematically with development.

Methods: We assessed repetitive behavior in deer mice at weekly intervals following weaning and subjected these data to a group-based trajectory modeling procedure (PROC Traj). This allowed us to calculate the number of discrete developmental trajectories that best accounted for individual animal data. We also used a Poisson process to model the distribution of discrete sequential repetitive behaviors at three points in development. This allowed us to determine differences in the temporal organization of repetitive behavior, independent of frequency, across developmental time periods within a trajectory group as well as at the same developmental time point across trajectory groups.

Results: Three discrete developmental trajectories best accounted for individual animal differences. Trajectory 1 represented a small percentage of mice (11%) that exhibited a low, flat trajectory. Trajectory 2 (41%) showed a gradual progression from low levels of stereotypy at week 1 post-weaning to high levels by week 6 post-weaning. Trajectory 3 (48%) exhibited high levels of repetitive behavior by week 1 post-weaning and a relatively flat trajectory. Significant differences both within and across developmental trajectory groups were also found. Trajectory 1 mice showed poorly organized (more random) behavior that did not vary across development. Trajectory 2 mice showed a marked change in temporal organization across development going from least organized at week 1 to most organized (more regular or periodic) at week 6. Trajectory 3 mice showed the highest level of organization (least random) at week 1 with a modest increase in organization across development.

Conclusions: Individual developmental curves of repetitive behavior in deer mice resolved into three discrete trajectory groups. These groups were associated with significantly different temporal dynamics at each developmental time point. In addition, each developmental trajectory group showed a different pattern of change in temporal organization across development. This latter finding suggests the importance of temporal dynamics in driving development and neuroadaptations. These findings also point to the potential importance of developmental trajectory and temporal dynamics in determining variable response to treatment and timing of interventions in individuals with ASD.