Fatigue-associated electromyographic (EMG) reactions of intrinsic hand muscles were studied during maintained isometric voluntary contractions of normal subjects. Most measurements concerned actions of the first dorsal interosseus (FDI). In a smaller number of subjects, complementary measurements were obtained for adductor pollicis (AP). 2. Measurements were made of isometric force (thumb adduction, index finger abduction and flexion) and of surface EMG amplitudes (AP and FDI) after rectification and smoothing (rsEMG). 3. In the analysis of fatigue, the subjects were required to maintain a steady isometric force (index finger abduction or thumb adduction) of half their maximum voluntary contraction (1/2MVC test) for as long as possible. Average endurance times were 88 +/- 19 s (mean +/- S.D.) for FDI and 119 +/- 29 s for AP (Student's t test, P < 0.02). 4. Pronounced differences in fatigue-associated EMG behaviour were observed between AP and FDI. In AP the reaction was as expected: a rise of EMG during maintained force (mean rsEMG at end of fatigue test/mean rsEMG at start of test (rsEMG-FI): 181 +/- 64%). In FDI this reaction was seen in half of the recorded cases, the remainder displaying bidirectional changes or a more or less marked decrease of EMG during the endurance task (mean for all cases together: rsEMG-FI, 103 +/- 15%; difference between AP vs. FDI significant, P < 0.01). 5. The unexpected EMG variability of the FDI reactions was further analysed with multiple bipolar recordings of surface EMG. For all the four thoroughly studied subjects, recordings were obtained which showed simultaneously occurring EMG changes in opposite directions (decrease and increase) at different sites of FDI while force was kept constant at 50% of the maximum voluntary contraction (MVC). 6. Further observations on FDI showed that EMGs simultaneously obtained from different recording sites could show dramatic differences in their responses depending on 'synergistic context' (e.g. in relation to changes in index finger extension force during maintained abduction at 50% MVC). Evidence for 'task switching' (shift in rsEMG distribution, shift in hand muscle synergy) was frequently observed during the performance of the 1/2MVC test. 7. The results indicate that FDI is not handled in a topographically homogeneous manner during the execution of an isometric constant force endurance test. Furthermore, the results suggest that this seemingly simple motor performance can be executed in several alternative manners associated with the activation of different muscle synergies and with different distributions of activity within the FDI.
