Martin McDonagh & Amy Parkinson
School of Sport and Exercise Sciences, University of Birmingham, Birmingham, UK

https://doi.org/10.36866/pn.73.12

It’s early evening after a long Christmas day and the kids are fractious. What on earth do you do next? Time for dad to produce a party trick. How about The mysterious rising arms? Tell your volunteer to stand in a doorway and then, with straight arms, press out as hard as possible against the door frame for a full minute. At the end of the minute the volunteer stops pushing and, after a couple of seconds, the arms rise sideways like the wings of a bird in flight (malt-fuelled grandads with dodgy hearts should not try this!). Your volunteer has just experienced the postural aftercontraction or, more poetically, the Kohnstamm phenomenon (Fig. 1). Unravelling the physiology behind this strange phenomenon has revealed fresh insights into the control of movement and posture.

The aftercontraction was first described by a neurologist, Alberto Salmon, at the 1914 meeting of the Italian Neurological Society in Florence (Salmon, 1914). The following year, an ‘internist’ turned psychiatrist, Oscar Felix Kohnstamm, demonstrated the same thing at the Ärztlichen Verein in Frankfurt (Kohnstamm, 1915). Kohnstamm ran a sanatorium in the Taunus Hills just outside the spa town of Königstein not far from Frankfurt. The sanatorium became fashionable amongst artists with depression, including the conductor Otto Klemperer and the expressionist painter Ernst Ludwig Kirchner. Kohnstamm’s 1915 paper drew the attention of several German physiologists who began calling the aftercontraction the Kohnstamm phenomenon. This irritated Salmon who complained bitterly that his primacy had been poached (scientists don’t change much it seems!).

The Salmon-Kohnstamm phenomenon attracted much research interest between the wars. Even Harvard professor Alexander Forbes, a collaborator of Lord Adrian, wrote a paper on what he termed’this rather baffling phenomenon’ (Forbes et al. 1926). Forbes experimentally rebutted the claim made by the Brazilian clinician Jayme Pereira that the aftercontraction took place without ‘action currents’. Forbes noted that the electromyograph (EMG) amplitude did not remain constant once the arm moved, but grew with the movement. However, if the arm was blocked in mid movement, the EMG levelled off. Forbes took this effect as evidence of a proprioceptive influence on the motor drive.

Two years earlier, Rupprecht Matthaei at the University of Bonn found that aftercontractions were strong in axial and proximal muscles, such as the deltoid, but weak or absent in the hand and other distal muscles (Matthaei,1924). I (MM) took Matthaei’s observations as a clue that postural control systems might be involved, including the vestibular apparatus. To investigate a possible vestibular influence on aftercontractions in the deltoid muscle we strapped our subjects on their backs to a rotating board, which could be fixed at any inclination. At each inclination we suspended the subject’s arms from slings. These tricky but enjoyable (at least for the experimenter) experiments revealed a linear relationship between the body inclination and the amplitude of the aftercontraction EMG. This clear result was pleasing but not decisive, as the gravitational load on the arms also varies with body inclination. Could load be a more important signal than vestibular input in driving the phenomenon? We decided to go back to where Forbes had left off and to ask the questions : What is the proprioceptive influence? How does it work? And which receptors are involved?

In the next experiments we varied the load on the arm whilst holding the vestibular signal constant. The subject sat on a chair with the head supported and to simplify the procedures we used only one arm. The gravitational load on the muscle increases as the limb moves towards the horizontal, so we used a counter lever to nullify this effect (Fig. 2)(Parkinson & McDonagh, 2006). In later experiments we added the experimental loads via a pulley wheel so that muscle loading remained the same at each joint angle. (Parkinson & McDonagh, unpublished). The results of these experiments are illustrated in Fig. 3. We found that the strength of the aftercontraction depends on both load and joint angle. Earlier experiments revealed that the size of the response increased with muscle shortening and decreased with muscle lengthening( Adamson & McDonagh, 2004) .This is very reminiscent of Sherrington’s shortening and lengthening reactions and also to the release of muscle cramps by stretching. Perhaps they share a common mechanism?

Next, we asked: Does joint angle directly modulate the EMG, or is the source some other factor which is varying with angle, such as muscle spindle length? In our more recent experiments we have tried to separate these two variables. The subject performs the 60s pre-contraction (see beginning of article) at a variety of joint angles. Activation of gamma, as well as alpha, motorneurones during the 60s precontraction will cause the sensory region of the muscle spindle to be tightened by gamma action. This tightening should result in a similar spindle output at a range of initial joint angles. If the spindle signal modulates the aftercontraction EMG, then the EMG will have a closer relationship to the presumed spindle length than to the absolute joint angle. This is exactly what we found (Fig. 4; McDonagh & Parkinson, unpublished); the tight relationship of emg amplitude to joint angle seen in Fig. 3 broke down. A future step is to test the spindle hypothesis by direct microelectrode recording from the spindle afferents in sensory nerves during aftercontractions.

To summarise so far: the aftercontraction is driven by the nervous system. It is not an intra-muscular contracture. Loading and muscle shortening strengthen it and forced lengthening weakens it. The proposed sensory receptors are Golgi tendon organ load receptors and muscle spindle length receptors.However, to fit with the data this would require positive force feedback from the tendon organs and negative length feedback from the spindles, the exact opposite of what is produced by the known spinal circuits. This suggests control from supraspinal centres capable of more complex processing. Which areas of the CNS could be generating the motor drive?

Over the years, the brain stem, basal ganglia, cerebellum and motor cortex have all had their advocates, who have based their guesses on argument and weak circumstantial evidence. Some have regarded it as a purely spinal phenomenon with no involvement of higher centres. Recently we took a direct approach by taking fMR images of the brain during aftercontractions. We found (Parkinson & McDonagh, unpublished) that the motor cortices are indeed involved in the phenomenon, as is the anterior cingulate cortex (ACC), an area of brain which has been implicated in error processing. In voluntary movement, predicted sensory outcome and actual sensory outcome match. In the after-contraction movement there is no motor command and so no predicted sensory outcome to match the actual one. Perhaps the ACC is processing this error signal.

It is now clear that the aftercontraction has both proprioceptive and central components – so how does the complete response come about? Our current hypothesis is as follows. During the 60s voluntary isometric pre-contraction, the system controlling the arm adapts to a strong opposing force. It changes its stored value of motor drive for this position. Once the pre-contraction ceases, this adaptation is still present. When this drive is released, as the aftercontraction, it produces more force than needed to maintain arm position against gravity, so the arm moves. Once the arm starts to move the motor drive is increased by two positive proprioceptive influences, muscle shortening and muscle load.

Most of our motor acts are carried out unconsciously. The after-contraction fMRI vividly illustrates that even the most simple of motor acts involves large areas of the cerebral cortex. Furthermore, the proprioceptive mechanisms involved in aftercontractions almost certainly underlie the everyday control of limb orientation in a gravitational field. These positive force feedback, and negative length feedback, mechanisms are fruitful areas for future research in movement neuroscience. In conclusion, the aftercontraction is an intriguing phenomenon in itself, but its study can also make a valuable contribution to our general understanding of motor control. Not bad for a party trick.

Visit http://www.metacafe.com/watch/323026/make_your_arms_levitate/ to watch a video of arm levitation.

References

Adamson G & McDonagh M (2004). Human involuntary postural aftercontractions are strongly modulated by limb position. Eur J Appl Physiol 92, 343–351.

Forbes A, Baird PC & Hopkins AM (1926). The involuntary contraction following isometric contraction of skeletal muscle in man. Am J Physiol 78, 81–103.

Kohnstamm O (1915). Demonstration einer katatoneartigen erscheinung beim gesunden (Katatonuersuch). Neurol Centrbl 34, 290–291.

Matthaei R (1924). Nachbewegungen beim Menschen (Untersuchungen uber das sog Kohnstammsche Phanomen). Pflugers Arch f.d ges Physiol 202, 88–111.

Parkinson A & McDonagh M (2006). Evidence for positive force feedback during involuntary aftercontractions. Exp Brain Res 171, 516–523.

Salmon A (1914). Nuove osservazioni sui movimenti automatici che si compiono dopo gli sorzi muscolari e del loro valore in neuropatologia. Atti della Accademia Medico -Fisica Fiorentina, 78–91.