J Manipulative Physiol Ther 2005 (Jul); 28 (6): 414–422 ~ FULL TEXT
Christopher J. Colloca, DC, Tony S. Keller, PhD, Pierre Black, MSc,
Martin C. Normand, PhD, DC, Deed E. Harrison, DC, Donald D. Harrison, PhD, DC
Neuromechanical Innovations, L.L.C, USA.
OBJECTIVE: To quantify the force-time and force-delivery characteristics of six commonly used handheld chiropractic adjusting devices.
METHODS: Four spring-loaded instruments, the Activator Adjusting Instrument; Activator II Adjusting Instrument, Activator III Adjusting Instrument, and Activator IV Adjusting Instrument, and two electromechanical devices, the Harrison Handheld Adjusting Instrument and Neuromechanical Impulse Adjusting Instrument, were applied to a dynamic load cell. A total of 10 force-time histories were obtained at each of three force excursion settings (minimum to maximum) for each of the six adjusting instruments at preload of approximately 20 N.
RESULTS: The minimum-to-maximum force excursion settings for the spring-loaded mechanical adjusting instruments produced similar minimum-to-maximum peak forces that were not appreciably different for most excursion settings. The electromechanical adjusting instruments produced short duration ( approximately 2–4 ms), with more linear minimum-to-maximum peak forces. The force-time profile of the electromechanical devices resulted in a more uniform and greater energy dynamic frequency response in comparison to the spring-loaded mechanical adjusting instruments.
CONCLUSIONS: The handheld, electromechanical instruments produced substantially larger peak forces and ranges of forces in comparison to the handheld, spring-loaded mechanical devices. The electromechanical instruments produced greater dynamic frequency area ratios than their mechanical counterparts. Knowledge of the force-time history and force-frequency response characteristics of spinal manipulative instruments may provide basic benchmarks and may assist in understanding mechanical responses in the clinical setting.
From the Full-Text Article:
To understand the biomechanical consequences of chiropractic adjustment/spinal manipulation more fully, chiropractic researchers are currently focusing on quantifying the applied forces associated with spinal manipulation and the mechanical response of the spine to these forces. [2, 17, 18, 21, 23, 26, 27] Basic experiments to quantify the forces transmitted during MFMA spinal manipulation as presented in the current study are important first steps in understanding the mechanics of spinal manipulation. In comparison to manual spinal manipulation (without the use of instruments), larger magnitude forces have been reported to be used by clinicians when treating the sacroiliac joint or lumbar spine  as opposed to the cervical spine. [23, 24] In this study, the electromechanical devices were found to produce larger peak forces and ranges of force in comparison to the mechanical instrument and, thus, may offer clinicians a wider selection and range of peak forces in the delivery of chiropractic manipulation.
Peak forces transmitted with the HAI and NMI devices at the maximum setting averaged 275 and 380 N, respectively, which is higher than the Activator devices (121, 154, 149, and 211 N) for the AAI, AAI 2, AAI 3, and AAI 4, respectively. It is hypothesized that higher peak forces may cause a greater magnitude vertebral displacements during chiropractic adjustments.  Previous biomechanical comparisons of MFMA and HVLA spinal manipulation have raised the issue of effective transmitted force distribution locally to the spine. Specifically, global measures of loading have been found to overestimate the local effective forces at the target site.  Herzog et al  reported average peak forces of 238.2 N for reinforced hypothenar contact HVLA spinal manipulation applied to the thoracic spine. In this work, the average peak local force was found to act over a target area of 25 mm2. When comparing these data with MFMA spinal manipulation, the cross-sectional area of the styli attached to MFMA devices ranges from 100 to 27 mm2. Thus, it is possible that the local forces applied with the AAI normalized to a 25–mm2 area may be the same as those observed here for HVLA hypothenar contact spinal manipulation,  whereas the HAI and NMI device acting over the same contact area may deliver higher forces. It should be noted, however, that each of the MFMA devices delivers forces over a very short time interval (<5 ms) as opposed to HVLA spinal manipulation (˜150 ms), which may result in much lower force impulse imparted to the spine. These differences, together with distinctions of articular cavitation responses, vertebral movements, and spinal reflex activities, all reflect possible considerations when studying different forms of chiropractic adjustment/spinal manipulation. [16, 29–35]
The force-time and frequency-response parameters determined for the HAI and AAI 2 instruments did not correlate linearly with the shuttlecock experiments. Rather, shuttlecock flight height showed a nonlinear dependency on force and frequency parameters, wherein the flight height increased less in comparison to the peak force or frequency parameters. Shuttlecock flight height correlated with the respective impulses of the two devices, however. The shuttlecock experiment, although novel, possesses limitations because of the coefficients of drag on the shuttlecock during its flight among other factors related to indirect measurements of transmitted force. In addition, any deviation of the shuttlecock flight path from 90° of its origin results in experimental error from geometry. Although attempts were made throughout the experiment to ensure a plumb shuttlecock flight path along the line of the background ruler, it was inherently not possible to maintain an exact 90° flight path, which subsequently affected the results.
Questions may arise whether the results from our bench tests on a table-mounted transducer can be extrapolated to data obtained in actual patients. A difference in stiffness response would be expected from a load cell mounted to a table compared to that obtained in patients; we believe that controlling the testing material by using a standard bench is appropriate for this study design. We have reported the force-time profiles of the Activator devices both from tests on a steel beam18 as well as thrusts delivered to normal subjects and actual patients. [26, 27] A review of these data shows little difference in the imparted force-time profiles to patients or rigid structures. In addition, the sampling frequency was chosen to ensure that the primary peak force-time profile of the various instruments was accurately captured, which in the case of the NMI device was only approximately 2 ms in duration. Fifty samples over a 2–ms duration (25 kHz) was deemed more than adequate to characterize the primary peak force-time response of this device, and 32768 samples per second was chosen as this was the next power of 2 integer above 25 kHz. Subsequent Fourier transforms of the adjusting instrument force vectors indicated that there was little or no frequency content above 2 kHz, which is over an order of magnitude lower than the sampling frequency. The results of this study suggest that a sampling frequency of 4 kHz or higher should be used to characterize the force-time response of the chiropractic adjusting instruments examined in this study.
Because the spinal column is a viscoelastic structure, increased mobility (motion response) will occur when the manipulation or mobilization therapy is applied at certain loading rates and frequencies. The relative stiffness of different regions of the thoracolumbar spine may vary with the mechanical stimulus frequency. [26, 36] Other important considerations in studying the biomechanics of spinal manipulation include the nonlinear, load-deformation behavior of the human spine. Inherent nonlinearities in the load-deformation characteristics of the spine result in variations in the measured posterior to anterior displacement and stiffness that are dependent on the magnitude of the applied force. For example, posterior to anterior mobilization studies have reported an increase in posterior to anterior stiffness when the peak force applied is increased. [37, 38] Greater forces, thus, may result in greater intersegmental and segmental motion responses of functional spinal units. [28, 39, 40] A structural model of the lumbar spine has been developed to characterize the sagittal plane static, sinusoidal, and impulsive motion response of lumbar spine segments.  The model provides data on segmental and intersegmental motion patterns that are otherwise difficult to obtain experimentally. Knowledge of the transmitted forces during chiropractic adjustment/spinal manipulation as presented in the current study and others, thus, can be modeled to contribute to the understanding of the motion response of the vertebral column. Such information is important in assessing the characteristics of chiropractic treatments.
Each of the chiropractic adjusting instruments examined in this study produced relatively large amplitude (maximum setting) force-time histories with primarily peak pulse durations less than 5 ms. Forces that are relatively large in magnitude, but act for a very short time (less than the natural period of oscillation of the structure), are called “impulsive.”18 Impulsive forces acting on a mass will result in a sudden change in velocity, but are typically associated with smaller amplitude displacements in comparison to longer duration forces. However, the manner in which the structure is mechanically excited will depend on the frequency content of the instrument's force-time history, and significant displacements can be produced provided that the force-time history contains frequency components at or near the natural frequencies of oscillation of the structure. In this study, the frequency area ratio of each device was computed to estimate the overall frequency content or relative frequency distribution of the impulsive force within a frequency range that was consistent with the first few natural frequencies of vibration of the spine subjected to posterior-anterior forces.  We found that the HAI and NMI produced a higher frequency area ratio (more uniform frequency distribution) in comparison to the Activator adjusting instruments examined. The frequency area ratio results reported herein differ from those previously reported for the AAI 3 and AAI 4. Namely, the results of the current study indicate that the mean frequency area ratio of the AAI 3 is lower than the original Activator 3 design, which was reportedly developed to improve the force-frequency spectrum of the Activator line of instruments.  Likewise, the dynamic frequency area ratio of the AAI 4 has not appreciably improved over the original AAI. A possible explanation for this discrepancy is that the data cited by Fuhr and Menke  were obtained by us using a prototype of the AAI 3 device, and not the commercial instrument ultimately manufactured. The present study presents the first comprehensive force-time and force-frequency data for several impulsive force chiropractic adjusting instruments that are currently being manufactured.
Of potential clinical interest is the finding that the motion response of the spine is closely coupled to the frequency or the time history of the applied force. External mechanical forces applied at or near the natural frequency of the structure are associated with appreciably greater displacements (over twofold) in comparison to external forces that are static or quasi-static.  Thus, it may be possible to achieve comparable posterior-anterior segmental motion responses for lower applied forces during spinal manipulation, provided that the forces are delivered over time intervals at or near the period corresponding to the natural frequency. We propose, because of the more uniform frequency response (haversine force-time profile) of the electromechanical devices, a testable hypothesis arising from the current study involves measuring the mechanical and physiological response of the spine among different MFMA devices at the same force settings but different frequencies. Further research into the force-time and force-frequency inputs of chiropractic adjustment/spinal manipulation on mechanical, physiological, and clinical responses in patients may help to optimize chiropractic interventions and treatment regimens.
In this study, the handheld, electromechanical HAI and NMI instruments produced a greater peak force and larger range of forces in comparison to the handheld, spring-loaded Activator devices. The electromechanical instruments were faster and produced greater dynamic frequency range (area ratios) than the spring-activated Activator instruments. Knowledge of the force-time history and force-frequency response characteristics of spinal manipulative instruments may provide basic benchmarks and may assist in understanding mechanical responses in the clinical setting.