The Role of Canes



The extent of locomotion assistance gained through cane use by those with hip disorders is assayed. Employing as a standard the propulsive impulse delivered by each lower limb of a healthy young male, the cane is shown to supply about one-fifth the equivalent impulse, aside from other possible benefits. Full test values are given for nine handicapped subjects.


What does a cane do? What good is it? We know of four uses for a cane: The cane acts to

  1. Increase stability
  2. Supply sensory feedback of position and force
  3. Reduce the loading of internal body structures
  4. Assist propulsion.

We will review the first three of these applications briefly and then concentrate on the last function, propulsion.


Pressing a cane against the ground creates a tripod consisting of two legs and the cane ( Figure 1 ). The tripod base area is quite large compared with the base area supplied by the legs alone. As long as the action line of the cane user falls within the bounds of the base area, that person is geometrically stable. The increase in stability obtained with a cane is simply a matter of comparing the large area, bounded by the heavy line, with the no-cane base area, shown by the light lines. In this particular case, the stable base is roughly doubled in size when a cane is used.

Sensory Feedback

Concerning sensory feedback, even small cane loadings can supply sensory clues to a user. A cane may load against intact neuromuscular areas, bypassing some reflexes lost in certain types of handicaps. For example, those who lack voluntary control of forcible plantar flexion of the ankle can receive knowledge of body sway through upper-limb deflections produced by guiding a cane.

* Text of a paper presented at the Eighth Annual Prosthetics and Orthotics Symposium, Kingsbrook Jewish Medical Center, Brooklyn, New York, October 24, 1979. ** Program Manager, Clinical Evaluation Programs. Formerly of the Office of Technology Transfer.


The cane is also used to unload portions of the body, such as the hip joint. This action is accomplished by using the upper limb, plus a cane, as a long lever arm so that even a modest force can produce a large mechanical moment at the hip-joint center. This moment reduces the amount of force required from hip abductor musculature which normally contracts during single-limb support to control the descent of the pelvis on the unsupported side. In this way, even a modest push can produce a large moment about the center of rotation. In the case of arthritic hip joints, the load relief obtained in this way can make walking tolerable to many who would otherwise-without canes-be forced into wheelchairs. The cane's effect in reducing the load on the contralateral limb is suggested by a set of schematic drawings ( Figure 2 ). Note in (B) and (C) the manner in which the hip abductor muscle normally plays an important role in controlling the descent of the pelvis on the unsupported side. When a cane supports one end of the pelvic lever (via the arm and the structure of the trunk), the cane's support allows the pelvis to be maintained essentially horizontal, even if the hip abductor contribution is small or absent.

Figure 3 illustrates an instrumented cane design by Ali Seireg of the Wood VAMC, Wood, Wisconsin. This cane was used to develop data for our study.


Now consider the cane as an aid to propulsion. Perhaps the simplest index to propulsion is the average momentum maintained in gait. In Figure 4 each light dot represents the average forward velocity and weight of one of our cane-using subjects. Each of these subjects suffered unilateral hip arthritis secondary to degenerative joint disease and/or avascular necrosis. All subjects exhibited the kind of limp that is associated with hip pain. The graph has contour lines representing fixed momentum values. The top contour line, 21 Ib.-sec., is the typical value Tor a healthy man of average weight and speed. You can see that our cane users develop relatively smaller amounts of mean momentum.

Wc have used the mean in dealing with momentum. Actually there arc variations in momenium over each step, superimposed on the mean value. The full story of momentum variation is made complex by the phased aspects of the legs, For our purposes it is sufficient to consider the change in momentum owing to the action of a single leg. Figure 5 shows the variation in momentum of a single, normal leg. The forward momentum of a normal subject in the stance phase of one limb is shown. The decrease from heel contact to foot flat is the result of one limbsworth of brake impulse; increase from foot flat to toe off is the result of one limbsworth of impulse during the acceleration phase. At heel contact (extreme left) braking starts and momentum is lost, becoming a minimum at the foot-flat condition. Push-off adds momentum until at toe-off (extreme right) the entire initial momentum is restored. We can picture this process as a roughly sinusoidal variation in momentum about some mean value.

We shall call the amplitude of the changing momentum a limbsworth. Making use of the classic impulse-momentum relationship, we can equate a limbsworth, or the change in momentum owing to the action of a single, healthy leg, to the impulse, either braking or accelerating.

Figure 6 shows some actual data (drawn from the classic 1946 University of California-Berkeley work) of the shear force, or the force and aft force, versus time, of a healthy subject. By performing a piece-wise strip integration as shown by the dotted lines, we can compute the value of a limbsworth. If the subject is walking at a truly constant mean speed (neither speeding up nor slowing down), one can see that the momentum lost in braking must precisely equal that gained in accelerating. The area below the time axis must equal the area above the time axis; and each area in turn equals a limbsworth-the amplitude of the momentum change.

Scanning the Berkeley data ( Figure 7 ), we found an average of 4.3 lb.-sees, for the braking portion of the step and 4.8 lb.-sees, for the accelerating portion. This means that the subjects are accelerating slightly despite their efforts to produce a constant-speed gait. As a practical measure, we will consider a rough average value, 4.5 lb.-secs., to represent the single-leg impulse developed by either leg of a healthy young male while walking at constant velocity.

We may apply the same technique to the cane user. Figure 8 shows raw data from one of our hip-disability subjects. At the top of the figure we have the cane load as a function of time. The lower curve gives the cane angle from the vertical, also as a function of time. By working along these curves in a piece-wise fashion, we can compute the change in horizontal momentum, just as we did for a leg. As the cane makes initial contact, we have a braking phase; and then the cane passes over to an accelerating phase. These phases correspond exactly to the action of a human leg, so much so that we can treat the cane as a kind of third leg.

However, unlike the normal person, each of whose legs must generate precisely equal amounts of acceleration and brake impulse in a constant-velocity gait, the cane user has a perfectly free choice regarding the sign and magnitude of the cane impulse. Thus the positive and negative portions of this curve need not balance. This freedom exists because the cane is redundant from the mechanics point of view; either leg is potentially capable of generating an appropriate counter impulse to the cane, if required. Thus, in principle, the subject can use the cane to generate an accelerating impulse only or a braking impulse only or any combination of such impulses and still maintain a constant-velocity gait.

In practice the cane user trades off many variables-pain, energy cost, stability, sensory feedback-to arrive at some combination he deems optimal.

Figure 9 shows the results arranged in a bar graph, in terms of the acceleration impulse developed at the cane. A limbsworth of impulse is also shown to supply a sense of scale. The average cane user chooses to develop about 0.14 limbsworth of braking impulse (the lower bars) and 0.31 limbsworth of accelerating impulse (the upper bars). Thus the boosting aspect of cane use is employed more heavily than the braking aspect, by roughly two to one. The average impulse developed is about one-fifth of a limbsworth. The nine subjects are middle aged or elderly. It may be noted that the drawing of one-fifth of a limbsworth from their arms represents great muscular effort. The cane propulsive effort for these subjects may be likened to the output of a third leg, a leg that happens to be about one-fifth as powerful as that of a healthy young man.

Still, is it not possible that the propulsion benefits are fringe benefits, and that the basic motivation of the cane user is elsewhere? That is, perhaps in the course of relieving pain or increasing stability, the canes are inadvertently used in a fashion that is helpful toward propulsion. Nothing in this work permits any response to this issue. We can only say that cane users suffering hip pain use canes in a manner that aids propulsion, and that the personal cost in terms of muscular effort is high.

In summary, as used by those with severe hip pain, the cane delivers about one-fifth the propulsive impulse obtained from either leg of a healthy young male. The maximum value of propulsive impulse appears related to the time of floor contact with the cane.


Parts of this material are from the article, "Locomotion Assistance Through Cane Impulse," by Leon Bennett, Mary Patricia Murray, Eugene F. Murphy, and Tamara T. Sowell, published in BPR 10-31, Spring 1979, pp. 38-47.