Startle-triggered responses indicate reticulospinal drive is larger for voluntary shoulder versus finger movements

Abstract

Recent primate studies have implicated a substantial role of reticulospinal pathways in the production of various voluntary movements. A novel way to assess the relative reticulospinal contributions in humans is through the use of a "StartReact" paradigm where a startling acoustic stimulus (SAS) is presented during a simple reaction time (RT) task. The StartReact response is characterized by short-latency triggering of a prepared response, which is attributed to increased reticulospinal drive associated with startle reflex activation. The current study used a StartReact protocol to examine differences in reticulospinal contributions between proximal and distal effectors by examining EMG onset latencies in lateral deltoid and first dorsal interosseous during bilateral shoulder or finger abduction. The magnitude of the StartReact effect, and thus relative reticulospinal drive, was quantified as the difference in RT between startle trials in which startle-reflex related EMG activation in the sternocleidomastoid (SCM) was present (SCM +) versus absent (SCM -). A significantly larger StartReact effect was observed for bilateral shoulder abduction versus bimanual finger abduction and a higher incidence of SCM + trials occurred in the proximal task. Additionally, both startle reflex and response-related EMG measures were larger on SCM + trials for the shoulder versus finger task. These results provide compelling novel evidence for increased reticulospinal activation in bilateral proximal upper-limb movements.

Figure 1

Premotor reaction time (RT; ms) shown for control trials, and startle trials in which activation in the sternocleidomastoid (SCM) was either absent (SCM −) or present (SCM +). Note the analysis only involved a comparison of SCM + /− conditions (n = 890 trials), whereby a significant interaction effect (p = 0.031) was found due to a larger decrease in RT for SCM + versus SCM − trials for the shoulder versus the finger response. Error bars represent 95% CI.

Figure 2

Task-related EMG variables (n = 3793 data points), separated by control trials and startle trials in which activation in the sternocleidomastoid (SCM) was either absent (SCM −) or present (SCM +). Panel (A) shows peak agonist amplitude (normalized to control values) whereas panel (B) shows normalized agonist Q₃₀. Significant interaction effects were found for both variables (p < 0.001), due to significant differences between all conditions except for SCM + versus SCM − for the finger task and between control conditions. Error bars represent 95% CI.

Figure 3

Startle-related EMG variables for trials in which sternocleidomastoid activation was present (SCM +) for the shoulder and finger responses. Panel (A) shows the proportion of SCM + trials, which was significantly higher for the shoulder task (p = 0.003) (n = 25 participants). Panel (B) shows the SCM onset latency in ms (n = 554 data points), which was not significantly different between movement types (p = 0.058). Panels (C) and (D) (n = 554 data points each) show SCM peak amplitude (mV) and SCM Q₃₀ (mV*ms), which were both larger for the shoulder versus the finger task (p < 0.001 for peak amplitude, p = 0.016 for Q₃₀). Error bars in panel (A) represent SD and grey dots show individual participant means; for all other panels error bars represent 95% CI.