Persistent mirror movements: force and timing of "mirroring" are task-dependent.

A simple isometric motor task was used to quantify intended and unintended finger movements in two subjects (father and son) with persistent mirror movements. One hand voluntarily changed grip force between thumb and index finger at different amplitudes and frequencies, while the other hand was to maintain a constant force. During all experimental conditions the "steady" hand showed insuppressible, highly cross-correlated contractions, compatible with bilateral distribution of a single motor command to the spinal cord. However, these associated movements were not strictly mirror images, nor did they show a fixed relationship to the voluntary movements across experimental conditions. The ratio of mirror to voluntary movement ranged from 1.4 to 19.1% and from 3.4 to 78.4% in the two subjects and was directly related to voluntary strength and speed. At maximum speed, mirror activity tended to precede voluntary activity, while it was delayed in slow force changes. Comparable time lags were not found in control subjects instructed to simulate mirror movements. We conclude that neuronal mechanisms in addition to bilateral corticomotoneuronal connections are at work in persistent mirror movements.