Journal of Applied Physiology
Vol. 81, No. 5, pp. 1884-1890, November 1996
EXERCISE AND MUSCLE

Mechanisms regulating regional cerebral activation during dynamic handgrip in humans

J. W. Williamson, D. B. Friedman, J. H. Mitchell, N. H. Secher, and L. Friberg

Department of Clinical Physiology and Nuclear Medicine, Bispjeberg Hospital, DK-2400 Copenhagen NV; and The Copenhagen Muscle Research Centre, Department of Anesthesia, University of Copenhagen, DK-2100 Copenhagen Ø, Denmark

ABSTRACT

INTRODUCTION

METHODS

RESULTS

DISCUSSION

ACKNOWLEDGEMENTS

FOOTNOTES

REFERENCES

ABSTRACT

Williamson, J. W., D. B. Friedman, J. H. Mitchell, N. H. Secher, and L. Friberg. Mechanisms regulating regional cerebral activation during dynamic handgrip in humans. J. Appl. Physiol. 81(5): 1884-1890, 1996.Dynamic hand movement increases regional cerebral blood flow (rCBF) of the contralateral motor sensory cortex (MS1). This increase is eliminated by regional anesthesia of the working arm, indicating the importance of afferent neural input. The purpose of this study was to determine the specific type of afferent input required for this cerebral activation. The rCBF was measured at +5.0 and +9.0 cm above the orbitomeatal (OM) plane in 13 subjects during 1) rest; 2) dynamic left-hand contractions; 3) postcontraction ischemia (metaboreceptor afferents); and 4) biceps brachii tendon vibration (muscle spindles). The rCBF increased only during dynamic hand contraction; contralateral MS1 (OM +9) by 15% to 64 ± 8.6 ml ^· 100 g^{1 ·} min¹ (P < 0.05); supplementary motor area (OM +9) by 11% to 69 ± 9.8 ml ^· 100 g^{1 ·} min¹ (P < 0.05); and there were also bilateral increases at MS2 (OM +5) [by 16% to 64 ± 8.6 ml ^· 100 g^{1 ·} min¹ (P < 0.05)]. These findings suggest that the rCBF increase during dynamic hand contraction does not require neural input from muscle spindles or metabolically sensitive nerve fibers, although the involvement of mechanoreceptors (group III or Ib) cannot be excluded.

single-photon-emission computerized tomography; ¹³³Xe inhalation; muscle ischemia; tendon vibration; exercise

INTRODUCTION

REGIONAL CEREBRAL BLOOD FLOW (rCBF) increases in the human motor cortex during vigorous hand exercise are proportional to the increase in the cortical oxygen consumption (24). Thus regional brain metabolism and regional cerebral blood supply are closely coupled such that measurements of rCBF can be used as an indication of activation of cortical brain centers during motor activity and sensory stimulation (14). Cerebral activation induced by dynamic hand movement can be blocked by regional anesthesia of the working arm, suggesting that such increases in rCBF are dependent on, at least in part, afferent input from the working muscle (9, 26). Vigorous dynamic hand contractions can activate primary muscle spindles (group Ia) and Golgi tendon organs (group Ib), as well as the smaller groups III and IV muscle afferents. Although any of these fiber types may provide the afferent signal for rCBF increases during contraction, differences in the effect of axillary blockade (using lidocaine) on rCBF in the primary motor sensory cortex [area at orbitomeatal plane +9 cm (OM +9); MS1] and the secondary somatosensory cortex (area at OM +5; MS2) during attempted hand contractions further suggest the possible involvement of different types of afferent fibers (9).

Because lidocaine has a greater effect on the thinly myelinated group III and the unmyelinated group IV muscle afferent fibers than on the larger group Ia, Ib, and II afferents (7), those fibers of importance for the increase in rCBF in the secondary somatosensory area at OM +5, as well as the supplementary motor area (SMA), may be smaller and less myelinated than those responsible for the large contralateral increase in the primary motor sensory cortex at OM +9 (9). The purpose of this study was to determine the specific type of muscle afferent important for the transmission of the signal that activates the human cerebral cortex during dynamic hand contraction. Thus rCBF was measured during four conditions: 1) rest; 2) unilateral dynamic hand contractions, which presumably activated groups Ia, Ib, II, III, and IV afferents; 3) postexercise muscle ischemia, which would primarily stimulate the small groups III and IV metabolically sensitive muscle afferents (12); and 4) vibration of the biceps brachii tendon, which would activate the large group Ia primary muscle spindles (6, 19). On the basis of the results from previous blocking studies (9, 26), it was hypothesized that isolation of groups III and IV metabolically sensitive muscle afferents during postexercise muscle ischemia would produce significant increases in rCBF.

METHODS

Four female and nine male right-handed volunteers [age = 33 ± 6 (SD) yr, weight = 79 ± 18 kg, and height = 181 ± 10 cm] were studied after giving informed consent for the experiment, which was approved by the Ethics Committee of the University of Copenhagen. With eyes covered, subjects rested horizontally in a dark, quiet room with their head positioned on a pillow to avoid tensing of the neck muscles. The head was placed in a tomograph in such a way that the midslice plane of the recorded two slices was positioned at 5.0 and 9.0 cm above the OM plane (Fig. 1).

Each subject was tested four times on the same day in a random order: 1) resting measurements were made with the subjects' left arm relaxed at their side; 2) dynamic hand contractions consisted of repetitive left-handed squeezing by using a handgrip device at a rate of 60 times/min for 5 min; 3) muscle ischemia of the left arm was induced for 5 min (measurement period) by inflation of a small arm cuff (to 200 mmHg) during the last 15 s of a 2-min bout of left-handed dynamic contractions; and 4) vibration of the left biceps brachii tendon at a frequency of 100 Hz for 5 min was accomplished by using a handheld vibrator (Bio-theisometer, Newbury, OH). For all protocols, subjects were given clear instructions not to move or tense muscles other than those directly involved in the hand contractions. Heart rate (HR), systolic blood pressure (SBP), diastolic blood pressure (DBP), and mean arterial pressure (MAP) data were collected each minute by using a sphygmomanometer cuff placed on the right arm.

The rCBF was measured by recording the regional distribution of cerebral radioactivity after inhalation of ¹³³Xe by using a rapidly rotating single-photon tomograph (Tomomatic 232, Medimatic, Denmark) (4, 11, 14). Increases in rCBF lead to a rise in the amount of radioactivity recorded from the respective region (17). One measurement lasted 4.5 min. During the first 90 s, ¹³³Xe was rebreathed from a closed-airway system with a CO₂ absorber and cleared from the brain during the remaining 3 min while the subject breathed normal air. For each measurement, a series of four consecutive samples was recorded at 1.5-, 1.0-, 1.0-, and 1.0-min intervals, respectively. These samples were used to calculate rCBF by deconvolution of the input curve recorded over the apex of the left lung (4, 11). The rCBF of the two slices was displayed on a screen in a 12-step color scale. The Tomomatic 232 has a spacial resolution of 12 mm in the transaxial plane (full width one-half maximum). With a rebreathing volume of 4 liters and a ¹³³Xe concentration of 740 MBq/l, one study produces an exposure of 0.63 mSv calculated as a whole body dose equivalent (2). The method has an error of ~5%, which is a measure of the overall biological and physical intraindividual and day-to-day variability (28). The cortical areas to be investigated were determined a priori such that each region was located in the same area of the rCBF map during all measurements. Twenty-four regions of interest (ROI) were defined (Fig. 2) as previously utilized by Friedman et al. (9). The size of the ROI map could be proportionally altered to better fit the confines of each individual's brain.

During each measurement of rCBF, end-expiratory PCO₂ was measured with an infrared capnograph (CD-101, Datex OY, Helsinki, Finland). On the basis of the baseline value, all rCBF values could be automatically corrected 2% by computer for each milliliter of mercury change in alveolar PCO₂ (28). At the end of each activity, the subject was asked to provide a rating of perceived pain (RPP) by using a standard Borg scale (6-20 units) (3). A score of 6 represented no pain, and 20 signified unbearable pain. These ratings were used to describe the sensation involved in the repetitive squeezing and toleration of the muscle ischemia or tendon vibration.

An analysis of variance with main effects of condition (rest, exercise, vibration, and ischemia), region of interest (24 ROIs), and side (right, left) was employed to establish areas with a significant increase in cerebral activation as indicated by changes in the rCBF (ml ^· 100 g^{1 ·} min¹). When significant F-ratios were present, Tukey's multiple-range post hoc test was used to determine specific mean differences. A P value <0.05 was considered to be significant.

RESULTS

1

1

1

1

1

1

1

1

1

1

Table 1. rCBF across conditions Region and Side rCBF, ml · 100 g1 · min1 Supine rest Dynamic hand contraction Postcontraction ischemia Biceps tendon vibration OM +9 Hemispheric   L 57 ± 9.2  60 ± 9.0  56 ± 9.2  57 ± 8.1    R 56 ± 9.1  64 ± 9.3* 51 ± 9.3  55 ± 7.8  Motor sensory   L 55 ± 7.1  60 ± 7.9  57 ± 8.3  58 ± 7.2    R 54 ± 6.7  64 ± 8.6* 55 ± 8.5  56 ± 7.0  Supplementary motor area 61 ± 8.2  69 ± 9.8* 63 ± 9.6  65 ± 9.2  Frontal   L 57 ± 8.8  58 ± 9.0  56 ± 8.9  56 ± 9.2    R 58 ± 9.1  59 ± 8.8  58 ± 9.2  57 ± 9.3  Parietal   L 53 ± 7.3  55 ± 7.9  52 ± 9.9  52 ± 7.2    R 54 ± 7.2  56 ± 8.3  51 ± 9.7  52 ± 8.1  Posterior mesial 58 ± 9.6  61 ± 9.2  57 ± 9.0  57 ± 8.3  OM +5 Hemispheric   L 56 ± 7.3  58 ± 8.9  53 ± 9.3  55 ± 7.8    R 56 ± 7.7  59 ± 9.2  54 ± 9.3  56 ± 8.1  Posterior white matter 54 ± 8.3  53 ± 7.5  54 ± 8.5  53 ± 7.0  Prefrontal   L 58 ± 9.8  59 ± 9.9  58 ± 10.1  60 ± 10.1    R 59 ± 10.1  61 ± 12.2  56 ± 11.1  61 ± 10.3  Frontal   L 56 ± 10.1  58 ± 10.1  54 ± 11.3  55 ± 10.0    R 56 ± 11.2  58 ± 10.2  53 ± 11.4  57 ± 9.8  Motor sensory   L 57 ± 7.7  66 ± 8.9* 57 ± 8.4  61 ± 8.9    R 57 ± 7.9  68 ± 9.5* 56 ± 8.0  62 ± 8.8  Temporal   L 52 ± 7.1  54 ± 7.7  50 ± 7.7  53 ± 7.1    R 53 ± 7.3  55 ± 7.6  51 ± 7.8  54 ± 7.3  Striatum   L 58 ± 8.3  59 ± 8.1  57 ± 9.6  58 ± 7.3    R 57 ± 8.4  58 ± 8.3  57 ± 9.8  57 ± 7.4  Occipital   L 55 ± 7.4  57 ± 7.7  54 ± 8.4  55 ± 7.1    R 56 ± 7.2  57 ± 7.8  55 ± 8.7  56 ± 7.3 Values are means ± SD for 13 subjects. Regional cerebral blood flow (rCBF) values are shown for rest, during left-handed dynamic contractions, postcontraction muscle ischemia of left arm, and brachial tendon vibration of left arm. OM, orbitomeatal; L, left; R, right. * Significantly different from rest at P < 0.05.

1

1

1

1

1

1

1

1

1

1

1

1

DISCUSSION

Consistent with previous findings during hand contraction (8, 9, 26), contralateral increases in rCBF were noted in the OM +9 brain slice in the motor sensory area (15%) and in the SMA (11%) as well as bilateral increases in the motor sensory area at OM +5 (16%). Despite previous findings implicating muscle afferent input in rCBF increases during hand exercise for identical regions of the brain (9), the primary finding from this study was that neither postexercise muscle ischemia nor biceps tendon vibration in isolation affected rCBF in selected ROIs. RPP values were similar between conditions of hand contraction and postcontraction ischemia. Because rCBF was increased only during contraction, it appears unlikely that a pain response was involved in the observed rCBF increases. Similarly, MAP changes were similar between contraction and ischemia; no changes in rCBF were detected during muscle ischemia. To the contrary, HR was elevated to the greatest extent during dynamic hand contraction. Because HR increases have been linked to increases in central command, there could be contributions from some centrally mediated neural mechanism (16, 18). Taken together, these findings could suggest 1) that some combination of centrally mediated signals (e.g., central command) and muscle afferent input is required for cerebral activation of the motor sensory areas investigated; or 2) that muscle afferents not presently isolated (e.g., group III or Ib mechanoreceptors) may play a role in the observed cerebral activation.

It would appear that isolated sensory input from the working muscle does not result in maximal motor sensory stimulation. Muscle spindle primary endings innervated by the large group Ia muscle afferents would be expected to be activated by tendon vibration. Because vibration had no significant effect on rCBF, these larger fibers alone are not responsible for the increase in motor sensory rCBF described. This is consistent with the findings of Kelley et al. (13), who reported no increase in middle cerebral artery (MCA) flow velocity via transcranial Doppler during tendon vibration. It has been shown that the flow velocity of the specific artery corresponding to the cortical projection of a limb (e.g., MCA for the hand) increases when that limb is exercised (10). Also, Jørgensen et al. (10) found no change in MCA flow velocity during postexercise muscle ischemia. In agreement with this finding, postexercise muscle ischemia, which is thought to selectively activate the smaller groups III and IV metabolically sensitive fibers (1, 29), did not result in activation of the motor sensory cortex.

Alternatively, group Ib or thinly myelinated group III muscle afferents serving as mechanosensitive receptors capable of sensing changes in contractile force or intramuscular pressure, respectively (12, 19, 23), were not presently isolated. During hand contractions, reductions in muscle strength (e.g., decreased force and intramuscular pressure) induced by axillary blockade correlate well with the magnitude of decrease in rCBF in motor sensory areas (9), suggesting that a mechanically sensitive receptor could play a role in this response. Although the group III fibers have been implicated in producing reflex increases in blood pressure (22, 30), we are presently unaware of data directly implicating their involvement in rCBF increases. Although this proposed mechanism may be involved in activities producing a moderate muscular contraction, it would appear to have little influence on brain activation during more complex hand activity such as coordinated finger tapping or writing (15).

The changes in rCBF during hand contraction were similar to those reported by Friedman et al. (9) with use of identical methodology, yet the increases in rCBF were smaller in magnitude than those reported for repetitive squeezing of a rubber ball (20) or more complex movement patterns such as finger tapping or writing (15). Similar to methods employed by Friedman et al. (9), specific regions of interest were selected a priori to minimize investigator bias in determining rCBF changes. However, use of these relatively large predefined areas (Fig. 2) can potentially yield smaller changes in rCBF when related to the relatively small (e.g., 2- × 3-cm) regions where the largest changes in flow occur (20, 21). The smaller rise in flow could potentially be related to cerebral autoregulatory mechanisms evoked in response to the substantial elevations in blood pressure. Of note, the rCBF values obtained under resting conditions were similar to those found by others (8, 9, 15, 27).

In conclusion, rCBF to motor sensory areas investigated was increased during dynamic hand contraction but not by postexercise muscle ischemia of the forearm or biceps brachii tendon vibration. In other words, the isolated afferent stimulation of the large group Ia or the smaller types III and IV (metabolically sensitive) fibers did not result in activation of the motor sensory cortex. However, dynamic hand contraction may have activated group III muscle "mechanoreceptors" to a greater extent than other conditions. Because previous rCBF studies have strongly implicated the involvement of muscle afferent input (5, 8, 9, 25) and have further shown that the increases in rCBF that occur during hand contractions are dependent on afferent input from nerves, it is reasonable to postulate an involvement of group III mechanosensitive fibers or Golgi tendon organs (group Ib) in activation of the motor sensory cortex during dynamic handgrip exercise. Given that the voluntary handgrip exercise will also increase central command, it is possible that some combination of neural input from higher brain centers and muscle afferent input is required for maximal activation of motor sensory regions.

ACKNOWLEDGEMENTS

We thank the subjects for their cheerful cooperation and Bente Dall, Eva Brodgaard, and Julius Lamar, Jr., for their excellent technical assistance.

FOOTNOTES

Address for reprint requests: J. W. Williamson, Univ. of Texas Southwestern Allied Health Sciences School, Dept. of Physical Therapy, 5323 Harry Hines Blvd., Dallas, TX 75235-8876.

Received 3 January 1996; accepted in final form 6 June 1996.

REFERENCES