Abstract

Trips and falls are a major concern for older adults. The resulting injury and loss of mobility can have a significant impact on quality of life. An emerging field of study, known as Perturbation Training, has been shown to reduce injury rates associated with trips and falls in older adults. Treadmills traditionally used for Perturbation Training are large, expensive, and immobile, forcing users to travel long distances to receive care. A portable treadmill would serve a larger portion of the at-risk population than current methods. We developed a portable, low-cost, twin-belt perturbation treadmill capable of high-intensity Perturbation Training. Belt speeds are controlled by a custom mechanical and software interface, allowing operators with no programming experience to control the device. The treadmill can accommodate users up to 118 kg and provides a maximum acceleration and speed of 12 m/s2 and 3.3 m/s, respectively, under full load. The total weight is 180 kg, and the treadmill can be moved like a wheelbarrow, with handles in the back and wheels in the front. The prototype was validated with mechanical and human participant testing, showing it as a viable device for Perturbation Training. In this paper, we will go over the design, fabrication, and validation processes used to create the Portable Perturbation Treadmill.

1 Introduction

Falls are a common and significant health threat among older adults [1]. Nearly one out of every three people over 65 years old will fall annually, in half of these cases, the falls are recurrent [2]. The most severe nonfatal injury from falling is hip fracture (and its comorbidities) which can double the average mortality risk for an individual for up to 12 years, with the most significant risk in the first year [3]. Therefore, falls are a substantial factor influencing the reduction in quality of life of older adults and present a significant burden on the healthcare system.

While numerous factors contribute to falls, few are modifiable, e.g., inconsistencies in sidewalks, weather conditions (snow and ice), or the use of graduated lenses in eyewear, all contribute to fall occurrence but are not feasible to modify across the breadth of the population. However, physical function (coordinated muscle activity, strength, and power) is a readily modifiable factor that declines with age [4], increasing fall risk. The idea of improving physical function to reduce fall risk has given rise to multiple fall-prevention exercise interventions for older adults. Accordingly, fall prevention programs such as Stepping On, Stay Active, and many other Exercise-based balance training (EBT) programs have been introduced across the country [57]. EBT programs can reduce fall incidence by 23% on average and require a relatively high training dose. If training totals less than 50 h or 3×/week or does not provide an adequate challenge to balance, the fall rate reduction is estimated to be 6–7% [8]. Perturbation-based balance training (PBT) is a newer balance training method intended to reduce falls by improving reactive balance performance. Unlike conventional EBT programs, which target the ability to maintain balance and deal with known balance disruptions, PBT targets reactive balance control by simulating unexpected balance disruptions, in a safe and controlled environment. PBT has been shown to improve stability control and reduce fall risk, with a 50–57% reduction in fall rate on average, at a substantially lower training dose than EBT [911]. Therefore, perturbation-based training is an effective and efficient form of exercise-related balance training.

Perturbation-based balance training is characterized by unexpected motions to simulate real-world trip or slip conditions. During a training session, the user will stand or walk on a moveable surface and will be exposed to sudden and unexpected accelerations (Fig. 1). A typical training program consists of multiple sessions spread out over several weeks and each session consists of multiple trials with varying intensities. Because the perturbation to balance must be rapid, perturbation training is typically done on specialized treadmills where the belts are accelerated under the user to induce a slip or trip. The advantage of using a treadmill is that perturbations can be triggered at any phase of the natural gait cycle due to its potential for continuous movement [1217]. The intensity of treadmill-based perturbations is varied by two factors, the acceleration of the belts and the duration of the perturbation with acceleration values range from 3 to 18 m/s2 for a duration of 0.2–0.5 s [1214].

Fig. 1
A participant experiencing: (a) forward loss of balance (backwards translation of treadmill surface) and (b) backward loss of balance (forwards translation of treadmill surface)
Fig. 1
A participant experiencing: (a) forward loss of balance (backwards translation of treadmill surface) and (b) backward loss of balance (forwards translation of treadmill surface)
Close modal

Although the efficacy of PBT is well documented, there is a general lack of PBT services in the commercial market, particularly in rural states. In our location, Montana, United States, only a single facility provides this service, with over 200,000 individuals 60 years or older living in this area [18], making the market for PBT currently underserved. Standard, exercise-grade treadmills are not suitable for PBT because it requires accelerations larger than standard treadmills can achieve [19]. While other PBT treadmills exist from commercial manufacturers such as MeditouchUSA, Treadmetrix, and Bertec, their cost precludes them from overall market adaptation, limiting their use to dedicated research facilities. In addition, PBT treadmills are not designed to be portable. This presents a problem for perturbation training because it requires the user to come to the treadmill, causing a barrier to implementing PBT with older adults, particularly those with limited mobility and in rural or remote areas.

Perturbation-based Balance Training is an effective training modality to improve balance recovery performance thus reducing the likelihood of experiencing an injurious fall. However, the widespread application of PBT has been limited in part because PBT requires large, expensive, and inaccessible training devices. This presents a clear need to develop a low-cost and accessible training solution. The purpose of this study was to design, fabricate, and validate a low-cost and portable PBT treadmill capable of inducing high-intensity perturbations, equivalent to those performed by commercial devices.

2 Methods

The design of the portable perturbation treadmill (PPT) was guided by the need to create a functional, portable device capable of inducing postural responses similar to falls and slips.

2.1 Functional Requirements.

A dual-belt system was specified to allow independent perturbations to each lower limb. The gap between belts should be obstruction free to allow for crossover gait and be no greater than half a foot's width, 4 cm [20], to ensure proper foot contact with the belt. Each belt should be capable of supporting the full load of the user on a single lower limb. The PPT must accommodate a 118-kg user, encompassing 95% of people over 65 years old [21]. The system was designed to achieve high-intensity slip or trip perturbations with a maximum belt speed of 3 m/s and acceleration of 12 m/s2 [1214] on the heaviest user. The belt speeds must be regulated so that various perturbation profiles can be achieved. For use in a later clinical trial, the treadmill is needed for approximately 5 h per day for 28 days and must not suffer from any mechanical failures during that time.

The PPT should be operated by a single individual and require no coding or specialty computer experience. The operator needs to control belt speed and perturbation intensity from an external Human-Machine-Interface (HMI). To be accessible for clinical and academic teams, the total cost of the system must be less than existing research units, which range in price from approximately $25,000–$75,000 [2224].

Bench testing of functional requirements utilized an experimental protocol. A human analog, constructed from a 132-kg weighted bag, was used to simulate both the rotational and translational momentum of the heaviest participant during perturbation. A step input to maximum speed, then a hard stop was implemented to create the highest stress condition. Starting at zero initial velocity, the belt was accelerated at 12 m/s2 up to 3 m/s and then immediately decelerated back to zero at the same rate. The test was performed ten times, and slow-motion video was recorded to investigate relative motion between the belt and drum, which was used to determine if adequate pretension had been applied to the belts. After the test the PPT was disassembled, and components were inspected for premature wear and damage. Specifically, the critical loading components were examined; shafts were inspected for hairline fractures using a liquid die penetrant system (Magnaflux, Glenview, IL), belts were visually examined for tears and delamination of the splice, and bearings were examined for excessive runout and resistance-free rotation.

2.2 Portability.

The PPT is intended as a portable research and clinical tool that travels to multiple locations and can be setup by two operators. A pickup truck is readily available to most research and clinical teams in the United States. The PPT must fit between the wheel arches with the gate closed, requiring a maximum size of 1.1 m × 1.9 m [25,26]. The device must be safely transported from the truck to the testing location. The maximum safe lifting force for a two-person lift is 80 kg [27].

The PPT must run on standard household power within the United States without tripping a breaker or creating a fire risk. For 120- and 240-volt systems, the maximum current draw cannot exceed 15 and 50 amps, respectively [28].

2.3 Perturbation Response Testing.

An experimental human study was performed to verify that the PPT produces postural responses analogous to trips and slips. While the protocol is intended for an older adult population a single, young participant (22 y/o male, 84 kg, 1.75 m tall) was used to improve safety of initial testing and consented to the Montana State University Institutional Review Board approved protocol (#2022-204). Full-body kinematics were measured using an Xsens MVN system (Xsens, El Segundo, CA) with a single sensor attached to each body segment totaling 17 MTx IMU sensors. Data were recorded at 60 Hz with kinematic joint angles calculated in the MVN Studio software utilizing the Xsens proprietary algorithms [29] that adhere to ISB standard for calculations and orientations of joint angles [30,31]. The protocol tested five different perturbation profiles of varying intensities (Fig. 2). Each profile was performed ten times, and trunk flection angle in reference to the ground plane was compared to known postural responses affiliated with falling [3234].

Fig. 2
Perturbation profiles tested. All profiles start from rest, max speed is indicated at the top and ramp, up and down times are given next to the respective arrows.
Fig. 2
Perturbation profiles tested. All profiles start from rest, max speed is indicated at the top and ramp, up and down times are given next to the respective arrows.
Close modal

2.3.1 Statistical Analysis.

A one-way ANOVA and Tukey's Honestly Significant Differences [35] procedure was used to determine pairwise differences in trunk lean angles for varying perturbation levels with significance set at p < 0.05. Following the postural testing, the treadmill was again disassembled and examined for signs of excessive wear similar to the human analog test procedure.

3 Results

3.1 Functional Requirements.

Each belt of the split-belt treadmill (Fig. 3) is connected to a dedicated motor (Fig. 4(a)) (CPM-MCVC-N0563P-RLN, Teknic, Victor, NJ) via a timing belt and pulley with a 5.6:1 gear ratio. Kinematic calculations indicated that 30 N-m of motor torque was required to induce the highest intensity perturbation when the heaviest user is standing on a single belt of the dual belt system. The selected motor outputs a peak torque of 32.9 N-m. Each treadmill belt is 0.46 m wide, and the gap between the belts is 1 cm. The shafts connecting the pulley and drum were sized based on the breaking strength of the belts. A shaft with a 45 mm diameter connects the driven pulley to a 150 mm diameter treadmill belt drum (Fig. 5). To prevent motor shutdowns during deceleration periods, a braking resistor (RES 255, Teknic, Victor, NJ) was installed to convert excess inertial energy into heat (Fig. 4(b)).

Fig. 3
Final treadmill assembly. (a) Fully assembled treadmill, rear view. (b) Fully assembled treadmill, front view. (c) Treadmill with cover and belts removed to show motor and drum assemblies.
Fig. 3
Final treadmill assembly. (a) Fully assembled treadmill, rear view. (b) Fully assembled treadmill, front view. (c) Treadmill with cover and belts removed to show motor and drum assemblies.
Close modal
Fig. 4
Front of treadmill (cover removed) showing electrical components. (a) Motors used to drive the treadmill belts. (b)Breaking resistors which are activated during hard deceleration periods to transform excesses inertial energy into heat. (c)Power and electrical connection junction box, which is where the treadmill connects to a power source and the HMI.
Fig. 4
Front of treadmill (cover removed) showing electrical components. (a) Motors used to drive the treadmill belts. (b)Breaking resistors which are activated during hard deceleration periods to transform excesses inertial energy into heat. (c)Power and electrical connection junction box, which is where the treadmill connects to a power source and the HMI.
Close modal
Fig. 5
Cross sectional view of shaft components
Fig. 5
Cross sectional view of shaft components
Close modal

Operation of the PPT by an individual with no coding or specialty computing experience is facilitated via the HMI (Fig. 6). A five-position rotary switch dictates the desired perturbation intensity from predefined values. A center detent potentiometer controls the steady-state walking speed for walking perturbations and provides tactile feedback when the belt speed is set to zero. A three-position switch determines which belt will be active (left, right, or both). A pushbutton switch initiates the perturbation, providing tactile and auditory feedback when engaged. A green trigger status Light Emitting Diode (LED) indicates when a perturbation is being performed. A large Emergency Stop button shuts down the treadmill in case of an emergency with a red LED to indicate activation. A green LED on the right indicates when the system is active and awaiting operator input. All LEDs blink in unison if the software detects a dangerous situation, such as when the belt speed is at a nonzero value on initial power-up or when the Emergency Stop is disengaged. The treadmill is controlled with an Arduino Mega 2560 microcontroller (Arduino, Somerville, MA), which monitors HMI button states and regulates motor speed via pulse width modulated (PWM) signals. Predetermined perturbation profiles are stored in memory via an array of speeds (PWM values) and incremental timing values. Custom perturbation profiles are created via a Matlab (Mathworks, Natick, MA) graphical user interface, where profiles of belt speed versus time are indicated with mouse click events by dragging points to desired locations. The software interpolates the line between points and calculates the appropriate PWM and timing values. The resulting array is saved as a CSV file that the Arduino can read. This approach requires no coding experience, as contextual menus and popups are used to define initial parameters, while also simplifying the creation of large, complex profiles. The system's total cost, including all accessories, was approximately $11,000.

Fig. 6
The human machine interface (HMI). The emergency stop button (EStop) is used to stop the belts in case of an emergency. The Mode switch selects the perturbation intensity from five predefined values. The Belt Select switch selects the belt the perturbation will be performed on (left, right or both). The Speed dial is used to set the belt velocity before and after a perturbation. The Trigger button is used to initiate a perturbation.
Fig. 6
The human machine interface (HMI). The emergency stop button (EStop) is used to stop the belts in case of an emergency. The Mode switch selects the perturbation intensity from five predefined values. The Belt Select switch selects the belt the perturbation will be performed on (left, right or both). The Speed dial is used to set the belt velocity before and after a perturbation. The Trigger button is used to initiate a perturbation.
Close modal

Reviewing the slow-motion video recorded during the human analog test revealed no noticeable signs of slippage between the belt and drum. No abnormalities were detected when disassembling the treadmill and examining critical loading components. No hairline fractures were observed in the shafts, the belt splice was intact, and the bearings showed no signs of excessive wear.

3.2 Portability.

Transport of the treadmill is facilitated by wheels in the front and handles in the back (Fig. 7). Its outer dimensions are 1 m × 1.8 m. The usable walking area is 1.67 m × 0.92 m. The treadmill weighs approximately 181 kg when fully assembled. The frame was made from aluminum T-Slot sections to facilitate portability and reduce overall mass. A two-ply fabric treadmill belt was selected for its strength and low weight [36].

Fig. 7
Manual treadmill transport
Fig. 7
Manual treadmill transport
Close modal

The motors can operate with either 120/240-volt input, which enables a top speed of 1.6/3.3 m/s. Because many of the intended test locations do not have an available 240-volt source, a portable generator was selected (XP13000EH, DuroPower, Covina, CA), facilitating maximum perturbation intensity anywhere. The generator can output 50 amps on the 240-volt line and can run on gasoline or propane [37]. A 15-m, 6 AWG extension cord delivers power from the generator to the treadmill, allowing the generator to be placed outside, reducing noise and noxious fume concerns. The extension cord connects to a power input box on the treadmill (Fig. 4(c)), which also serves as a connection point for the motor control wiring harness. A 3 m data cable connects the HMI to the treadmill, with connectors at each end to aid in transport and disassembly. This allows for remote operation by a single operator from a central location.

3.3 Perturbation Response Testing.

Motion capture data from the human participant test (Fig. 8) showed a maximum trunk lean angle of 31.2 ± 0.9 deg (mean±SE) with varying outcomes based on peak speed and deceleration time (Fig. 9). Following the postural testing, disassembly of the treadmill found no signs of excessive wear.

Fig. 8
Setup for human participant test
Fig. 8
Setup for human participant test
Close modal
Fig. 9
Results from human participant testing (n = 1) for various profiles (P1–P5) with 10 trials per profile. Bars show mean trank lean angle±SE. Perturbation Profiles 1 and 2 were similar (P1 and P2 = Group A). Profiles 3 and 5 were similar (P3 and P5 = Group B). Profile 5 elicited the highest trunk lean angle and was different from the other four. Outcomes from the ANOVA and Tukey's HSD are shown on inset table.
Fig. 9
Results from human participant testing (n = 1) for various profiles (P1–P5) with 10 trials per profile. Bars show mean trank lean angle±SE. Perturbation Profiles 1 and 2 were similar (P1 and P2 = Group A). Profiles 3 and 5 were similar (P3 and P5 = Group B). Profile 5 elicited the highest trunk lean angle and was different from the other four. Outcomes from the ANOVA and Tukey's HSD are shown on inset table.
Close modal

4 Discussion

Perturbation Balance Training is an effective method to reduce fall rates in older adults. Current devices are large, expensive, and generally inaccessible. This presents a barrier to implementing PBT, especially for older adults with limited mobility or those in rural/remote areas. Therefore, there is a clear need to develop a low-cost and accessible training solution. The objective of this project was to design, fabricate, and validate a Portable Perturbation Treadmill. The project was split into three main requirements: functionality, portability, and postural response verification.

4.1 Functional Requirements.

The treadmill needed to be dual belt, accommodate a user of up to 118 kg and provide belt accelerations up to 12 m/s2. Initial calculations indicated that a belt 0.318 m wide would be sufficient; however, 0.46 m is the recommended minimum width for a traditional, single belt treadmill [3840]. Since each side of the dual belt treadmill was designed to be redundant (each belt, drivetrain, and motor can handle 100% of the load), it is possible to fabricate two single-belt treadmills from existing components with little additional cost.

A single individual with no programming or specialty computing experience can operate the treadmill. The HMI's buttons and switches are used to dictate perturbation parameters, and the Drag & Drop Perturbation Profile Generator uses contextual menus and prompts that facilitate the creation of new perturbation profiles. The Drag & Drop Profile Generator has a practical maximum of 20 segments due to the size of each point on the screen and the lack of a scrollable graph. The motor control software was initially designed to calculate belt speed profiles independently whenever an HMI button or switch changed state without an external computer. Initial device testing showed that the Arduino lacked sufficient memory for internally calculated profiles longer than 2 s because the profile creation function requires more memory than was available. The limitations of the Arduino could be overcome with a more powerful microcontroller, such as a Teensy or Raspberry Pi [41,42]. Current perturbation training devices range in price from $25,000–$75,000. The total cost of the PPT was less than $11,000, representing a 50% cost savings. Therefore, clinics could potentially purchase this for their use, and the cost requirements were satisfied. Further cost reductions could be achieved if the design were modified to use a single belt, ordering components in larger quantities, and eliminating the generator for locations with adequate power sources. With these modifications, the total material cost would be approximately $4,000 per unit.

The results of the human analog test show that the treadmill is safe to use with a user of up to 118 kg, and a maximum belt speed and acceleration of 3 m/s and 12 m/s2. However, higher accelerations are possible for users of lower weight. Therefore, the functional requirements of the treadmill were satisfied.

4.2 Portability.

With outer dimensions of 1 m × 1.8 m, the treadmill fits within a standard pickup truck bed. Therefore, the dimensional requirements were satisfied; however, the weight of the system exceeded initial expectations. The treadmill weighs 181 kg with a centrally located center of mass, meaning that two operators need to lift approximately 91 kg each when moving the treadmill. This does not meet OSHA standards for a two-person lift; therefore, we recommend that a team of at least three people is used to move the prototype device. Half of the overall weight is from drivetrain components such as the pulleys, shafts, and drums (Fig. 5), which are made from solid steel or cast iron. However, the design could be modified to use lightweight materials with an estimated weight reduction of approximately 50 kg. Additionally, if a dual belt system was not needed, the total weight would be approximately 70 kg which would be in line with OSHA regulations.

When connected to 120 volts, the treadmill draws up to 17 amps, which is greater than the 15-amp specification. Operating the treadmill on household 110-volt power limits the motor's top speed and can potentially trip a standard residential breaker or create a fire risk. A slow blow, type C, circuit breaker [28] could be used since motor inrush current is only sustained for a short time, but this would need to be installed at each training location. Since this was not feasible, a portable generator was used, allowing high-intensity perturbations to be performed at any training site. The generator is capable of outputting 50 amps on the 240-volt line and is connected to the treadmill by a 15 m extension cord, reducing noise and noxious fume concerns.

4.3 Perturbation Response Testing.

The primary goal of this project was creating the perturbation treadmill, which included testing that it produces perturbations simulating trips and slips. Previous studies have indicated that a trunk flection angle of approximately 25 deg simulates real-world slips and trips [3234], we achieved a maximum of 31.2 ± 0.9 deg during Profile 4 which started from rest and accelerated to 3.3 m/s in 0.3 s and decelerated to a stop in 0.3 s; therefore, the postural response requirements have been satisfied. It is interesting to note that the maximum speed and time to decelerate influences the postural response. Profiles 3 and 5 produced similar responses of 23.7 ± 0.9 deg and 26.9 ± 0.9 deg, p = 0.10, both similar to the real-world slip value, but were achieved with different peak speeds and deceleration times. This information will be used to inform appropriate profiles for the treadmill during a future clinical trial. As a measure of safety, the Perturbation Response Test was performed on a healthy young adult. Further testing is needed to identify the varying ranges of postural responses in the target population of older adults.

Following safety and functional validation, the treadmill was part of an efficacy trial to delivery deliver PBT to community dwelling older adults. It was used for 28 days at multiple test sites for approximately 5 h each day on nineteen older adults (aged 69.6 ± 6.6 years, 17 women, 2 men) [43]. During the trials, no component failures were observed, reinforcing the mechanical integrity of the device, and satisfying the design requirements. On each testing day, the treadmill was transported to and from the testing site, indicating that even though weight requirements were not met, portability was possible. Anecdotally, many of the older adults tested were not able to successfully recover from the high-intensity profile used in the perturbation response test of this work, requiring the perturbation levels to be reduced, demonstrating that the treadmill exceeded the postural response requirements for individuals in the target demographic. Full results from the clinical trial will be detailed in a future publication.

5 Conclusion

This work outlined the design and validation of a portable, split-belt perturbation treadmill. High-Intensity perturbations can be performed on a user who weighs up to 118 kg, with a maximum belt speed and acceleration of 3 m/s and 15 m/s2, and with independent belt control. It weighs 181 kg and is moved like a wheelbarrow with wheels in the front and handles in the back. Training can be performed in any location and the system requires no coding or specialty computing experience to operate. It costs 50% less than similar commercial devices, is capable of producing postural responses similar to slips and trips, and was used in multiple locations in a clinical trial without failure or degradation.

Acknowledgment

The authors of this paper would like to thank Joe Eldring and Glenn Foster at the Montana State University Machine Shop for manufacturing guidance and for providing space to build and test the device. Additional thanks need to be given to the gracious technical support staff at Teknic for help with debugging motor connectivity issues. This work is primarily derived from the Master's Thesis of the first author, Robert Knutson [44]. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Funding Data

  • National Institute of General Medical Sciences of the National Institutes of Health (Award No. P20GM103474; Funder ID: 10.13039/100000002).

Data Availability Statement

The datasets generated and supporting the findings of this article are obtainable from the corresponding author upon reasonable request.

References

1.
Rubenstein
,
L. Z.
,
2006
, “
Falls in Older People: Epidemiology, Risk Factors and Strategies for Prevention
,”
Age Ageing
,
35
(
SUPPL.2
), pp.
37
41
.10.1093/ageing/afl084
2.
Moylan
,
K. C.
, and
Binder
,
E. F.
,
2007
, “
Falls in Older Adults: Risk Assessment, Management and Prevention
,”
Am. J. Med.
,
120
(
6
), pp.
493.e1
493.e6
.10.1016/j.amjmed.2006.07.022
3.
Katsoulis
,
M.
,
Benetou
,
V.
,
Karapetyan
,
T.
,
Feskanich
,
D.
,
Grodstein
,
F.
,
Pettersson-Kymmer
,
U.
,
Eriksson
,
S.
, et al.,
2017
, “
Excess Mortality After Hip Fracture in Elderly Persons From Europe and the USA: The CHANCES Project
,”
J. Intern. Med.
,
281
(
3
), pp.
300
310
.10.1111/joim.12586
4.
Landi
,
F.
,
Liperoti
,
R.
,
Russo
,
A.
,
Giovannini
,
S.
,
Tosato
,
M.
,
Capoluongo
,
E.
,
Bernabei
,
R.
, and
Onder
,
G.
,
2012
, “
Sarcopenia as a Risk Factor for Falls in Elderly Individuals: Results From the IlSIRENTE Study
,”
Clin. Nutr.
,
31
(
5
), pp.
652
658
.10.1016/j.clnu.2012.02.007
5.
Ory
,
M. G.
,
Smith
,
M. L.
,
Jiang
,
L.
,
Lee
,
R.
,
Chen
,
S.
,
Wilson
,
A. D.
,
Stevens
,
J. A.
, and
Parker
,
E. M.
,
2015
, “
Fall Prevention in Community Settings: Results From Implementing Stepping on in Three States
,”
Front. Public Health
,
2
, p.
116688
.10.3389/fpubh.2014.00232
6.
York
,
S. C.
,
Shumway-Cook
,
A.
,
Silver
,
I. F.
, and
Morrison
,
A. C.
,
2011
, “
A Translational Research Evaluation of the Stay Active and Independent for Life (SAIL) Community-Based Fall Prevention Exercise and Education Program
,”
Health Promot. Pract.
,
12
(
6
), pp.
832
839
.10.1177/1524839910375026
7.
Thomas
,
E.
,
Battaglia
,
G.
,
Patti
,
A.
,
Brusa
,
J.
,
Leonardi
,
V.
,
Palma
,
A.
, and
Bellafiore
,
M.
,
2019
, “
Physical Activity Programs for Balance and Fall Prevention in Elderly
,”
Medicine
,
98
(
27
), p.
e16218
.10.1097/MD.0000000000016218
8.
Sherrington
,
C.
,
Whitney
,
J. C.
,
Lord
,
S. R.
,
Herbert
,
R. D.
,
Cumming
,
R. G.
, and
Close
,
J. C. T.
,
2008
, “
Effective Exercise for the Prevention of Falls: A Systematic Review and Meta-Analysis
,”
J. Am. Geriatr. Soc.
,
56
(
12
), pp.
2234
2243
.10.1111/j.1532-5415.2008.02014.x
9.
Pai
,
Y. C.
,
Bhatt
,
T.
,
Yang
,
F.
, and
Wang
,
E.
,
2014
, “
Perturbation Training Can Reduce Community-Dwelling Older Adults' Annual Fall Risk: A Randomized Controlled Trial
,”
J. Gerontol., Ser. A
,
69
(
12
), pp.
1586
1594
.10.1093/gerona/glu087
10.
McCrum
,
C.
,
Gerards
,
M. H. G.
,
Karamanidis
,
K.
,
Zijlstra
,
W.
, and
Meijer
,
K.
,
2017
, “
A Systematic Review of Gait Perturbation Paradigms for Improving Reactive Stepping Responses and Falls Risk Among Healthy Older Adults
,”
Eur. Rev. Aging Phys. Act.
,
14
(
1
), pp.
1
11
.10.1186/s11556-017-0173-7
11.
Lurie
,
J. D.
,
Zagaria
,
A. B.
,
Pidgeon
,
D. M.
,
Forman
,
J. L.
, and
Spratt
,
K. F.
,
2013
, “
Pilot Comparative Effectiveness Study of Surface Perturbation Treadmill Training to Prevent Falls in Older Adults
,”
BMC Geriatr.
,
13
(
1
), pp.
1
8
.10.1186/1471-2318-13-49
12.
McCrum
,
C.
,
Karamanidis
,
K.
,
Grevendonk
,
L.
,
Zijlstra
,
W.
, and
Meijer
,
K.
,
2020
, “
Older Adults Demonstrate Interlimb Transfer of Reactive Gait Adaptations to Repeated Unpredictable Gait Perturbations
,”
Geroscience
,
42
(
1
), pp.
39
49
.10.1007/s11357-019-00130-x
13.
Golyski
,
P. R.
,
Vazquez
,
E.
,
Leestma
,
J. K.
, and
Sawicki
,
G. S.
,
2022
, “
Onset Timing of Treadmill Belt Perturbations Influences Stability During Walking
,”
J. Biomech.
,
130
, p.
110800
.10.1016/j.jbiomech.2021.110800
14.
Sessoms
,
P. H.
,
Wyatt
,
M.
,
Grabiner
,
M.
,
Collins
,
J. D.
,
Kingsbury
,
T.
,
Thesing
,
N.
, and
Kaufman
,
K.
,
2014
, “
Method for Evoking a Trip-Like Response Using a Treadmill-Based Perturbation During Locomotion
,”
J. Biomech.
,
47
(
1
), pp.
277
280
.10.1016/j.jbiomech.2013.10.035
15.
Lurie
,
J. D.
,
Zagaria
,
A. B.
,
Ellis
,
L.
,
Pidgeon
,
D.
,
Gill-Body
,
K. M.
,
Burke
,
C.
,
Armbrust
,
K.
,
Cass
,
S.
,
Spratt
,
K. F.
, and
Mcdonough
,
C. M.
,
2020
, “
Surface Perturbation Training to Prevent Falls in Older Adults: A Highly Pragmatic, Randomized Controlled Trial Background. Falls Are the Leading Cause of Injuries Among Older Adults, and Trips And
,”
Phys. Ther.
,
100
(
7
), pp.
1153
1162
.10.1093/ptj/pzaa023
16.
Gerards
,
M. H. G.
,
Marcellis
,
R. G. J.
,
Poeze
,
M.
,
Lenssen
,
A. F.
,
Meijer
,
K.
, and
de Bie
,
R. A.
,
2021
, “
Perturbation-Based Balance Training to Improve Balance Control and Reduce Falls in Older Adults – Study Protocol for a Randomized Controlled Trial
,”
BMC Geriatr.
,
21
(
1
), pp.
1
12
.10.1186/s12877-020-01944-7
17.
Fitzgerald
,
G. K.
,
Axe
,
M. J.
, and
Snyder-Mackler
,
L.
,
2000
, “
The Efficacy of Perturbation Training in Nonoperative Anterior Cruciate Ligament Rehabilitation Programs for Physically Active Individuals
,”
Phys. Ther.
,
80
(
2
), pp.
128
140
.10.1093/ptj/80.2.128
18.
U.S. Census Bureau
,
2021
, “
American Community Survey 5-Year Estimates
,” U.S. Census Bureau, Suitland, MD, accessed Nov. 7, 2023, http://censusreporter.org/profiles/04000US30-montana/
19.
Whitten
,
J.
,
Oleary
,
B.
,
Tarabochia
,
D.
, and
Graham
,
D.
,
2022
, “
Does Perturbation-Based Balance Training on Commercial Treadmills Improve Balance Recovery in Older Adults?
,”
Innov. Aging
, 6(Suppl 1), p.
38
.10.1093/geroni/igac059.144
20.
White
,
R. M.
,
1982
,
Comparative Anthropometry of the Foot
, United States Army Research & Development Laboratories,
Natick, MA
.
21.
Fryar
,
C. D.
,
Carrol
,
M. D.
,
Gu
,
Q.
,
Afful
,
J.
, and
Ogden
,
C. L.
,
2021
, “
Anthropometric Reference Data for Children and Adults: United States, 2015–2018
,”
Natl. Cent. Health Stat. Vital Health Stat.
,
3
(
46
).https://stacks.cdc.gov/view/cdc/100478
22.
Rehab Store
,
2023
, “
Woodway Bariatric Medpro Treadmill
,” accessed May 17, 2024, https://www.rehab-store.com/p-woodway-bariatric-medpro-tread-ergometer.html
23.
Burns
,
B.
,
2023
, “
Bertec Treadmill FIT GEN 5
,” Bertec Corporation (Quote).
24.
Brasher
,
S.
,
2023
, “
Woodway SPLIT-BELT
,” Woodway USA (Quote).
25.
Tailored Trucks
,
2023
, “
Size of a Pickup Truck Bed & Styles
,” accessed May 17, 2024, https://tailoredtrucks.com/blog/size-of-a-pickup-truck-bed/
26.
Hearst Auto Research: Car and Driver
,
2023
, “
How Long is a Truck Bed
,” accessed May 17, 2024, https://www.caranddriver.com/research/a32811097/how-long-is-a-long-bed-truck/
27.
Waters
,
T. R.
,
Putz-Anderson
,
V.
, and
Garg
,
A.
,
2021
,
Applications Manual for the Revised NIOSH Lifting Equation
,
Cincinnati, OH
.
28.
National Fire Protection Association
,
2023
, “
NFPA 70: National Electric Code
,” National Fire Protection Association, Boston, MA.
29.
Schepers
,
M.
,
Giuberti
,
M.
, and
Bellusci
,
G.
,
2018
, “
Xsens MVN: Consistent Tracking of Human Motion Using Inertial Sensing
,” White Paper, Xsens Technologies B.V., Enchede, The Netherlands.10.13140/RG.2.2.22099.07205
30.
Wu
,
G.
, and
Cavanagh
,
P. R.
,
1995
, “
ISB Recommendations for Standardization in the Reporting of Kinematic Data
,”
J. Biomech.
,
28
(
10
), pp.
1257
1261
.10.1016/0021-9290(95)00017-C
31.
Wu
,
G.
,
Siegler
,
S.
,
Allard
,
P.
,
Kirtley
,
C.
,
Leardini
,
A.
,
Rosenbaum
,
D.
,
Whittle
,
M.
, et al.,
2002
, “
ISB Recommendation on Definitions of Joint Coordinate System of Various Joints for the Reporting of Human Joint Motion - Part I: Ankle, Hip, and Spine
,”
J. Biomech.
,
35
, pp.
543
548
.10.1016/S0021-9290(01)00222-6
32.
Aviles
,
J.
,
Wright
,
D. L.
,
Allin
,
L. J.
,
Alexander
,
N. B.
, and
Madigan
,
M. L.
,
2020
, “
Improvement in Trunk Kinematics After Treadmill-Based Reactive Balance Training Among Older Adults is Strongly Associated With Trunk Kinematics Before Training
,”
J. Biomech.
,
113
, p.
110112
.10.1016/j.jbiomech.2020.110112
33.
Feldman
,
A. G.
,
Krasovsky
,
T.
,
Baniña
,
M. C.
,
Lamontagne
,
A.
, and
Levin
,
M. F.
,
2011
, “
Changes in the Referent Body Location and Configuration May Underlie Human Gait, as Confirmed by Findings of Multi-Muscle Activity Minimizations and Phase Resetting
,”
Exp. Brain Res.
,
210
(
1
), pp.
91
115
.10.1007/s00221-011-2608-0
34.
Bieryla
,
K. A.
,
Madigan
,
M. L.
, and
Nussbaum
,
M. A.
,
2007
, “
Practicing Recovery From a Simulated Trip Improves Recovery Kinematics After an Actual Trip
,”
Gait Posture
,
26
(
2
), pp.
208
213
.10.1016/j.gaitpost.2006.09.010
35.
Tukey
,
J. W.
,
1949
, “Comparing Individual Means in the Analysis of Variance,”
Biometrics
, 5(2), pp.
99
114
.10.2307/3001913
36.
Mol Belting Systems
,
2023
, “
2AW5-0BL-SB Product Specifications
,” Mol Belting Systems, Grand Rapids, MI, accessed May 17, 2024, https://www.molbelting.com/belts/2AW5-0BL-SB
37.
DuroMax
,
2023
, “
DuroMax XP13000EH 13,000 Watt Dual Fuel Portable Generator
,” accessed May 17, 2024, https://www.duromaxpower.com/products/duromax-xp13000eh-13000-watt-portable-hybrid-gas-propane-generator
39.
Bye
,
D.
,
2023
, “
Proper Treadmill Size
,” accessed, https://www.livestrong.com/article/286794-proper-treadmill-size/
40.
Roeleveld
,
M.
,
2023
, “
Which Size Treadmill Belt Do You Need?
,” accessed May 17, 2024, https://homegymresource.com/which-size-treadmill-do-you-need-treadmill-belt-size/
41.
PJRC
,
2023
, “
Teensy® 3.6 Development Board
,” accessed May 17, 2024, https://www.pjrc.com/store/teensy36.html
42.
Raspberry Pi
,
2023
, “
Raspberry Pi 4
,” accessed May 17, 2024, https://www.raspberrypi.com/products/raspberry-pi-4-model-b/
43.
Whitten
,
J.
,
O'Leary
,
B.
,
Graham
,
D.
,
Grocke-Dewey
,
M.
,
Riley
,
J.
,
Harper
,
D.
, and
Tarabochia
,
D.
,
2024
, “
The Acceptability of a Community-Based Perturbation-Based Balance Training to Older Adults and Healthcare Professionals
,”
Gerontol. Geriatr. Med.
,
10
, p.
23337214241246843
.10.1177/23337214241246843
44.
Knutson
,
R. G.
,
2022
, “
Design, Fabrication, and Validation of a Portable Perturbation Treadmill for Balance Recovery Research
,”
M.Sc. thesis
,
Montana State University, Bozeman, MT
.https://scholarworks.montana.edu/handle/1/17592