Four independent tests. Here's what they measured.

Sandia National Laboratories, Cybernetic Systems Integration, Org. 6633. SAND No. 2005-5045 P. Michael Lyons, Ty Pankretz, Jason Wheeler.

Abstract. This report summarizes findings from a gait comparison study that was performed at Sandia National Laboratories in Albuquerque, NM in July of 2005. Test subjects were outfitted with two kinds of shoes, Z-Coil and a leading competitor. They walked and jogged while acceleration data was collected. The results showed significantly lower mean peak accelerations in the Z-Coils when walking (13.34%, P value <0.0001) and jogging (17.2%, P value <0.0001). It is proposed that these reduced accelerations may be a reason why wearers of the Z-Coil shoes report reduced joint pain compared to traditional elastomer-based footwear.

Key finding. The accelerations in Z-Coil shoes were significantly lower for all individuals who did not bottom out the spring. Accelerations averaged 13% lower while walking and 17% lower while running.

Introduction. Many wearers of Z-Coil footwear cite a pain reduction when wearing Z-Coil brand shoes vs. traditional footwear. While there are many testimonials expressing this viewpoint, there is little scientific evidence to explain it. Al Gallegos, owner of Z-Coil, hoped to use the expertise of organization 6633, Cybernetic Systems Integration, through a small business technology transfer project to quantify the advantage that his shoes offer over traditional footwear. Given a somewhat limited budget, it was clear that a traditional gait analysis was out of the question, as this typically involves force plates that are capable of withstanding high loads while having a very short time response as well as high-fidelity cameras and motion tracking hardware. Such equipment was prohibitively expensive for this project, but a suitable and more portable alternative was found in the form of a three-axis accelerometer. An accelerometer works by having a tiny arm that flexes in response to accelerations, in much the same way that a person's head will lean back as his/her car accelerates. When this arm flexes, it presses on a piezoelectric material that releases charge when compressed. The internal circuitry of the accelerometer converts the charge to a voltage proportional to the current acceleration. The accelerometer used in this study contains three of these arms all mounted perpendicularly, measuring accelerations in all three spatial dimensions. In an effort to quantify the benefits of the Z-Coil footwear, we sought to validate our hypothesis that the reduced pain individuals experience is due to reduced foot impacts that occur while performing everyday tasks such as walking or jogging.

Methodology. In a typical gait analysis, data is gathered from a force plate (similar to a bathroom scale) which registers forces as a person walks across it. The force increases as the person's heel strikes the force plate. The largest force is a combination of two separable quantities: the weight of the person, and the force caused by the mass of the person being accelerated upward. The force plate gait analysis is useful for many reasons, but it is quite costly, and actually does not show the data of interest to this study. While the impact of the heel striking the force plate does show up, the force associated is only a fraction of the person's body weight. By measuring accelerations, we are measuring the speed with which the person's foot comes to a halt, not what portion of their body weight is pushing down on the sole of the shoe. It is our hypothesis that the accelerations the foot undergoes are transmitted up through the body and could be a contributor to, or a cause of, joint pain. After receiving Human Subject Testing Board approval, an ad was placed in the Sandia Daily News bulletin, and the first three females and males to reply were accepted to be study participants. One female participant dropped out of the study for health reasons, and an alternate took her place. After reading and signing documents of informed consent, each participant performed the test outlined in the Experimental Setup section.

Figure 2: force plate response of a normal gait cycle

Theory. We did not use a force sensor to gather gait data, but instead an accelerometer. This approach, while not the traditional means of gait assessment, does give useful data. The similarity of the two approaches is intuitive when one considers Newton's law of motion, F = ma, where F represents force, m mass, and a acceleration. The relationship between force and acceleration is directly proportional and varies only by the multiplier of mass, which is constant for each individual. In the case of traditional force-plate assessment, the equation governing the vertical forces is F = m(a + g), where g is the acceleration due to gravity, which when multiplied by mass is commonly referred to as weight. It was hypothesized that the accelerations should be lower in the Z-Coil shoes because of the very definition of acceleration: a = Δv / Δt. The velocity with which an individual's foot approaches the ground should be the same, regardless of shoe type. By landing on a spring, the time over which the foot comes to a stop is increased. As Δt increases, the resulting acceleration decreases. This is the advantage that the Z-Coils offer over traditional shoes.

Figure 3: typical plot of summed accelerations in normal gait

Experimental setup. The accelerometer (a Crossbow Technologies 3-axis LP model) was attached via a steel clip to the outside of the right shoe. When the test participant was comfortable with the shoes, the attached data logger was triggered to gather data at 512 Hertz. Each participant walked on a flat surface for approximately 60 paces while data were collected for all three accelerometer axes. After data were collected for both types of shoes with the participant walking at a self-selected normal walking pace, the accelerometer was moved to an elastic band attached with Velcro just below the participant's knee. Fitted like this, the participant jogged, again at a normal jogging pace, for three recorded segments of approximately 20 paces each in each type of shoe. The accelerometer was moved to the knee for the running portion of the test to ensure accurate data recording: the accelerometer is accurate to ±10 g, and preliminary tests had shown that heel-based accelerations during jogging could exceed 10 g.

Figure 1: accelerometer mounting

Data. By using a three-axis accelerometer, our results were three channels of data that represent vectors in the x, y, and z directions. With the participant's foot flat on the floor, the positive x vector points forwards, the positive y is upwards, and the z vector points medially through the ankle. These were vectorially summed to produce a total acceleration vector at each moment in time. The summed data were analyzed, and the highest peaks (one per gait cycle, or stride) were averaged to produce the statistic used for comparison, called here the mean peak acceleration (MPA). The significance of the difference between the statistics for Z-Coils and the other shoes was found using the Student t-test; associated P values are reported in the Appendix.

Figure 4: component and summed accelerations

Figures 5 and 6: mean peak acceleration by participant, walking and jogging

Results. In general, the accelerations recorded while individuals were wearing Z-Coils were significantly smaller than those in the others. In the walking trial, all individuals recorded significantly lower accelerations while wearing Z-Coil shoes except participant number 3. In the jogging trial, participant number 3 again had higher accelerations in the Z-Coils, as did another participant, and one individual had the same mean peak acceleration in both pairs of shoes. During the jogging trial, participant number 6 noted that it felt like the shoe was bottoming out, but had not felt that during the walking trial. This is consistent with the data for number 6, who had a lower MPA in Z-Coils while walking and a lower MPA in the competitor's shoes while jogging. Our hypothesis for these results was that the individuals who had higher accelerations in Z-Coils were bottoming out the spring at heel strike. Without the support of the spring, the participants' heels were, in effect, hitting the running surface, in this case a concrete floor. This hypothesis seemed reasonable, especially when the weight of participant number 3 (250 lbs) was considered. This was verified by numerically integrating the y-axis acceleration data twice to get vertical heel displacement. In the case of the one individual who experienced higher accelerations in the Z-Coils while walking, the vertical heel displacement was found to be 1.1 inches. This very nearly matches the actual height of the bottom of the Z-Coil shoe above the floor. In this individual's case, it is clear that the spring is becoming fully compressed, and the bottom of the shoe is striking the floor. In every other case, the vertical displacement is within the useful travel range of the spring, resulting in reduced accelerations.

Table 2: vertical heel displacement and participant weight

Travel range diagram

Conclusions. For most participants, the Z-Coil shoes had significantly lower accelerations while walking (5 out of 6) and jogging (4 out of 6). It appears that the use of a stiffer spring in shoes that will be worn by heavier individuals would be beneficial, and should result in reduced accelerations for all wearers of Z-Coil shoes. The analysis technique used in this study is somewhat limited in its utility, and the possibility of gait variation among subjects was not researched. While the conclusion that some individuals are bottoming out the spring is plausible, and mathematically verified, it is possible that some portion of the abnormal gait accelerations are due to another cause, including gait type or a lack of familiarity with the Z-Coil shoes. This work proved that individuals who are properly fitted in their Z-Coil shoes with the correct spring have reduced accelerations when walking and jogging. This finding allows the exclusion of the data from individuals who suffered from inadequate springs. With the exclusion of the data for those subjects who bottomed out the springs, the reduction in walking MPA is changed from 8.7% to 13.4%. The reduction in jogging MPA is changed from 5.1% to 17.2%.

Figure 7: properly fitted results

Further work. This study showed a statistically significant reduction in acceleration for most wearers of Z-Coil shoes. The link between this finding and joint pain is suggested, but not researched. This study could be used as a pilot to future work where matched cohorts of individuals with joint pain could be observed for several months or years to determine long-term benefits of reduced accelerations during everyday activities.

Appendix: participant data and P values

Los Alamos National Laboratory. Evaluation of Z-Tech (DBA Z-CoiL) running shoes. Hopkins, 1995.

Conclusions. The pulse curves for the Recoil running shoes were typically of longer total duration than the curves for the elastomer cushioned running shoe. Pulse durations were observed to be as much as 50% longer for the Recoil shoes in some cases. This was particularly true for the lower energy impacts that would be typical of relatively light weight runners (100–150 lbs.) or heavier runners with relatively little vertical displacement in their running style.

Summary. The data obtained in this project seems to indicate one area in which the design of the Recoil shoe may be an improvement over that of the typical elastomer cushioned running shoe. The initial impact forces appear to be less abrupt in the Z-CoiL shoes, resulting in a reduction of the jarring effect to the foot and the lower leg of a runner as the heel impacts the ground. The fact that the peak forces are equivalent to those of the traditional running shoe may be offset by the longer period of time required to reach these peaks after initial impact.

"Recoil" was the working name of the shoe at the time of the evaluation. Only the conclusions and summary of this evaluation survive in our files.

New Mexico Highlands University, Department of Human Performance and Sport. Ergonomic Implementation to Reduce Plantar Pressure. Wayne Scheler, ATC. Master's field project, December 2008. IRB approved.

Abstract. Industrial employees are on their feet eight to ten hours a day, and when external weight is carried, plantar pressure multiplies. This study evaluated whether Z-CoiL footwear, worn as occupational footwear, reduces plantar pressure. Nineteen male participants from industrial occupations were fitted with Z-CoiL shoes by a certified athletic trainer and, carrying 40-pound bags, moved from station to station on a simulated production line in their own work shoes and again in Z-CoiL. The Novel Pedar in-shoe system (99 sensors per insole, 50 Hz) recorded pressure across the whole foot, heel, midfoot and forefoot. Paired t-tests compared mean peak pressure between conditions.

Results. Mean peak pressure was significantly lower in Z-CoiL across the whole foot (right 13.1%, left 12.3%), at the heel (right 17.0%, left 14.1%) and at the midfoot (right 15.7%, left 11.8%). The forefoot was essentially unchanged (right about 1%, left not significant). The reduction was statistically significant in 87.5% of the regions measured. Two participants showed increased pressure in one or more regions.

Figure: mean peak pressure by participant, whole foot

Figure: mean peak pressure by participant, by region

Conclusion. With Z-CoiL shoes worn as occupational footwear, the plantar surface of the foot experienced a significant decrease in mean peak pressure at the heel and midfoot, where the coil and built-in orthotic act, and no meaningful change at the forefoot. The study measured pressure; it did not measure pain or injury outcomes.

Read the full 91-page study (PDF) — replace this link with the file URL once it's uploaded to Shopify Files.