Department of Exercise Sciences

hand and foot movement

Andrew J. K. McIntyre-Robinson and Winston D. Byblow
Published in J Neurophysiol 110: 1039-1046, 2013


Hand and foot movements are made more reliably when both limbs move in the same direction at the same time (isodirectional) compared with when they are made in opposite directions (anisodirectional), as quantified by relative joint angles between ankle dorsiflexion (DF) plantarflexion (PF) and wrist flexion-extension. To date, the neurophysiological mechanisms underpinning the dynamics of hand-foot coordination are not fully understood. There are no anatomical connections between hand and foot representation areas within primary motor cortex (M1), suggesting common input arising from secondary motor areas rather than horizontal connectivity within M1.

Using dual-coil transcranial magnetic stimulation (TMS), Byblow and his group identified a putative functional network between dorsal premotor cortex (PMd) and M1 subserving the isodirectional coupling of hand and foot movement. Currently it is not known whether the PMd-M1 pathway facilitates the production of isodirectional patterns via a direct facilitatory pathway between PMd and M1, the modulation of inhibitory projections between PMd and M1 circuits mediating short-interval intracortical inhibition (SICI), or both. Facilitation along the PMd-M1 pathway specifies the preferred coordination mode, but so far it not known whether this is the sole pathway through which one mode or the other is selected during interlimb coordination.


The present study was designed to investigate whether isodirectional coupling between ankle and wrist flexors and extensors might be explained by facilitation via specific I-wave facilitation using short-latency intracortical facilitation (sICF). I-wave facilitation identified by paired-pulse TMS is presumed to occur between excitatory interneurons within one, two, or three synapses from M1 corticospinal projection neurons and a recent study showed that I waves may originate within PMd as well as within M1. This led to the idea that late (I2, I3) I waves may originate, at least in part, from within PMd even when paired-pulse TMS is applied over M1 alone.

Our hypothesis was that isodirectional tendencies would reflect, at least in part, facilitation occurring at I-wave periodicity attributable to a (putative) PMd origin.

We conducted behavioural and neurophysiological experiments in order to 1) confirm the preference for isodirectional hand-foot coordination in our sample and 2) test the hypothesis that there would be a correlation between the isodirectional tendency and the modulation of corticomotor excitability of wrist extensors (and flexors) during ankle PF versus DF.

McIntyre-Robinson 2013 - Figure 1 (insert after methods)



  • Participants: 17 healthy young adults with no neurological impairments, no wrist or ankle injury, and no contraindications to TMS participated in two experiments. Participants were seated with their right hand and foot supported in a custom-built manipulandum (Fig.1). Electromyography (EMG) data were recorded from the right extensor carpi radialis (ECR), flexor carpi radialis (FCR), right tibialis anterior (TA), and right soleus (SOL) muscles. 
  • Experiment 1: hand-foot coordination. A behavioural paradigm involving concurrent right wrist flexion and extension and right ankle PF and DF in time with a metronome beat. Movements began in one of two patterns, isodirectional (easy) or anisodirectional (difficult). Phase transitions from the anisodirectional to isodirectional pattern were expected. Potentiometer signals were converted to angle and phase measures and the main dependent measure was relative phase between hand and foot calculated as the phase angle of the ankle subtracted from that of the wrist. 
  • Experiment 2: Neurophysiological basis of isodirectional tendency which was completed with the forearm always at rest and the leg either at rest or during ankle DF or PF in time with a metronome. TMS was delivered as single (nonconditioned, NC) or paired (conditioned, C) pulse in a randomized order. Stimuli were applied 100ms before the fifth, sixth, or seventh metronome beat (pseudorandomized), resulting in the delivery of TMS during peak TA or SOL activation during DF and PF conditions, respectively. ECR and FCR MEP areas were calculated over a 20-ms window from MEP onset determined individually for each muscle and participant. 


Experiment 1: During each trial of hand-foot coordinated movement, transitions occurred from an anisodirectional to an isodirectional pattern. As expected, coordination became unstable in the anisodirectional pattern but not the isodirectional pattern, as confirmed in measures of wrist and ankle relative phase error and stability (both P < 0.001) (Fig. 2).

Experiment 2: short-latency paired-pulse TMS was used to elicit MEPs and produce sICF in right ECR and FCR in the presence and absence of right ankle PF/DF (P < 0.015). An isodirectional preference was confirmed by facilitation of FCR MEPs and TMS-induced wrist flexion during ankle plantarflexion (both P < 0.025) (Fig. 5) but no evidence of modulation of any particular “I wave” during foot movement compared with rest.



In support of our hypothesis, there was evidence of isodirectional tendencies between hand and foot coordination (experiment 1) and for directional tuning of resting upper limb neurophysiological excitability and kinematics obtained during foot movement (experiment 2) as well as an observed association between neurophysiological and behavioural measures obtained in separate experiments. Indeed, participants who had a faster time to transition, indicating a breakdown of the more difficult anisodirectional pattern, exhibited a greater degree of MEP modulation in support of producing the isodirectional pattern. While this correlation is not causal, it is reasonable to suspect that the modulation obtained in measures of directional tuning from TMS may reveal, at least in part, an individual’s preference for isodirectional over anisodirectional hand-foot movements. In addition, the present results support earlier findings that M1 corticospinal excitability is modulated to favour the production of hand and foot movements in the same direction over those in opposite directions.  

Although previous studies have identified PMd as a potential origin for I waves, we found no indication of a specific I wave-mediated intracortical facilitation underlying isodirectional tendencies between hand and foot in the present study.


The authors thank James Coxon, Keith Runnalls, Fred Noten, Phil Lacey, and Terry Corin for technical assistance.