A Feasibility Study to Assess Vibration and Sound from Zygapophyseal Joints During Motion Before and After Spinal Manipulation

Abstract

Objective

This feasibility study used novel accelerometry (vibration) and microphone (sound) methods to assess crepitus originating from the lumbar spine before and after side-posture spinal manipulation (SMT).

Methods

This study included 5 healthy and 5 low back pain (LBP) subjects. Nine accelerometers and 1 specialized directional microphone were applied to the lumbar region, allowing assessment of crepitus. Each subject underwent full lumbar ranges of motion (ROM), bilateral lumbar SMT, and repeated full ROM. Following full ROMs the subjects received side-posture lumbar SMT on both sides by a licensed doctor of chiropractic. Accelerometer and microphone recordings were made during all pre- and post-SMT ROMs. Primary outcome was a descriptive report of crepitus prevalence (average number of crepitus events/subject). Subjects were also divided into 3 age groups for comparisons (18–25, 26–45, and 46–65 years).

Results

Overall, crepitus prevalence decreased pre-post SMT (average pre= 1.4 crepitus/subject vs. post= 0.9). Prevalence progressively increased from the youngest to oldest age groups (pre-SMT= 0.0, 1.67, and 2.0, respectively; and post-SMT= 0.5, 0.83, and 1.5). Prevalence was higher in LBP subjects compared to healthy (pre-SMT-LBP= 2.0, vs. pre-SMT-healthy= 0.8; post-SMT-LBP= 1.0 vs. post-SMT-healthy= 0.8), even though healthy subjects were older than LBP subjects (40.8 years vs. 27.8 years); accounting for age: pre-SMT-LBP= 2.0 vs. pre-SMT-healthy= 0.0; post-SMT-LBP= 1.0 vs. post-SMT-healthy= 0.3.

Conclusions

Our findings showed that a larger study is feasible. Other findings included that crepitus prevalence increased with age, was higher in LBP than healthy subjects, and overall decreased following SMT. This study showed that crepitus assessment using accelerometers has the potential of being an outcome measure/biomarker for assessing spinal joint (facet/Z joint) function during movement and the effects of LBP treatments (eg, SMT) on Z joint function.

Introduction

Zygapophyseal Joint Vibrations and Sounds

Crepitus, audible sound arising from tissues during normal motion, originating from the vertebral column has been associated with degeneration of the posterior joints of the spine [facet joints/zygapophyseal (Z) joints].^1–3 More subtle crepitus (fine crepitus) may possibly be related to joint hypomobility (increased joint stiffness) and/or loss of joint lubrication.³ Protapapas and Cymet (2002) made an interesting observation: “The noises of normal and abnormal [Z] joints are built into the nature of the structures. To ignore these noises would be foolish, as they might be used to help determine the effectiveness of treatment.”⁴ Cavitations are audible sounds commonly associated with spinal manipulative therapy (SMT),^5–10 and are theoretically associated with gas (probably carbon dioxide)^11,12 entering joints as they gap during manipulation.^6,7,10,13

One theory of the mechanism of the anatomical/biomechanical beneficial effects of SMT is depicted in Fig. 1. Briefly: 1) connective tissue adhesions develop in hypomobile Z joints,^14–16 which would be accompanied by increase joint sounds and vibrations during motion (e.g., crepitus); 2) the Z joint facet surfaces separate (gap) during SMT,¹⁷ which is sometimes accompanied by cavitation;^5–10 and 3) the gapping is thought to break-up the connective tissue adhesions, allowing restoration of joint mobility.^15,16 Over time (weeks-months) this would lead to decreased joint vibrations and sounds during motion (decreased crepitus).

Figure 1.

Flowchart showing a model of putative beneficial anatomical/biomechanical effects of spinal manipulation, including the theoretical relationships of crepitus and cavitation to the model.

In addition, we hypothesize that LBP subjects would have more crepitus than healthy subjects because of increased muscle tightness of LBP subjects leading to relative compression of the articular surfaces making up the Z joints. Such compression could conceivably interrupt the normal smooth gliding motion of the Z joints resulting in crepitus vibrations. We also hypothesize that older subjects would have more crepitus than younger ones, because of decreased joint lubrication and increased joint degeneration in older Z joints resulting in more crepitus vibrations/sounds.

A few studies have been conducted to explore the mechanism(s) and consequences of cavitation during SMT.^9,10,18–21 However, there have been no studies assessing lumbar crepitus. In addition, the peer-reviewed literature has no record of a study assessing tissue characteristics of the spine using acoustics or a study comparing accelerometers with microphones in assessing acoustic data originating from the spine (including data associated with Z joint crepitus).

Such sounds have been studied in the knee.^22–24 In addition, analysis of vibration and sound waves recorded from accelerometers and microphones have been used rather extensively to assess sounds as a type of biomarker associated with temporomandibular joint dysfunction,^25–32 as compared with temporomandibular joints functioning normally.^33,34 Interestingly, Brown even identified unique sound patterns in muscles of patients with Parkinson’s Disease.³⁵

Purpose

Our previous work used accelerometers to assess Z joint cavitations during SMT.^19–21 This feasibility study (n=10) was designed to continue the process of addressing the current gaps of knowledge by further developing previous methods and evaluating additional methods to assess and localize Z joint crepitus recorded during lumbar motion both before and after SMT. More specifically, this study evaluated recordings made simultaneously from accelerometers and a specialized directional microphone. Recordings were made from both healthy (n=5) and low back pain (LBP, n=5) subjects of different age groups. The reason for using both accelerometers and a microphone is that each may produce unique and complementary information.³⁶

If successful, joint crepitus identified during lumbar motion could be used as an outcome measure (biomarker) to assess the mechanisms of SMT.

Methods

Subject Screening and Enrollment

This project was approved by the National University of Health Sciences IRB. Five (5) healthy and 5 low back pain subjects (n=10) were sought from three age ranges: 18–25, 26–45, and 46–65 years. More subjects (n=8) were sought from the older two age groups to better assess crepitus. In addition, an equal distribution of males and females was also sought. Subjects were recruited from the student, faculty and staff population of a complementary and integrative health institution through posters, flyers, and email announcements. Those who passed the initial screening were scheduled for an examination where a research chiropractic physician conducted a thorough history and physical examination that included orthopedic and neurological tests to ensure subjects were selected who met stringent inclusion and had none of the exclusion criteria. Separate criteria that had been previously developed and implemented for healthy^19,20 and LBP subjects³⁷ were used in this study. Generally, healthy subjects had no current LBP, known spinal pathology, or spinal surgery; and no previous history of LBP lasting more than 2 weeks, or no more than 3 episodes of LBP of brief duration (one week) in any given year. LBP subjects had current LBP and a history of an episode of LBP lasting for more than 2 weeks, or more than 3 episodes of back pain of brief duration (one week) in any given year. Subjects were excluded if they reported having had spinal surgery, or having been previously been diagnosed with disc degeneration, significant osteoarthritis, scoliosis of greater than 5 degrees (Cobb’s angle, scoliosis decreases ability to adequately image the Z joints), or other known significant pathology. Current pain was assessed using a 100 mm visual analog scale and scores were converted to a 0–10 numeric pain rating with 0 being “no pain” and 10 being “worst pain imaginable.”^38–40 Subjects completing the study received a $10 gift card or (if preferred by a student) University Service credit equivalent to the number of hours of participation.

Placement of Accelerometers and Microphone

After completing the informed consent process, enrolled subjects were scheduled for an “Accelerometry and Microphone Visit.” During this visit, patients were initially asked if they had received SMT in the previous 24 hours or if their health had changed significantly after the examination visit (all subjects answered “no” to both questions). Seven (7) accelerometers (Bruel & Kjaer 2693, Naerum, Denmark) were then taped, using elastic kinesiology tape (Rock Tape, Campbell, California, USA), in position in the midline over the L1-S2 spinous processes and sacral tubercles (Fig. 2–color versions of Figs. 2–7 can be viewed in supplemental electronic materials). An additional 2 accelerometers (total of 9) were taped lateral to the midline accelerometers: one 3 cm to the left and one 3 cm to the right of the mid-distance between the L4 and L5 midline accelerometers. A microphone (PCB Piezotronics 130A40, Depew, NY, USA) was then taped 4 cm to the left of the T12 spinous process. Data were collected using a LabView program (National Instruments, Austin, Texas) written for a previous project¹⁹ and modified to record from an additional channel to accommodate the microphone. Recordings were displayed on and recorded from a computer-based oscilloscope.

Figure 2.

Placement of accelerometers and specialized directional microphone; A = subject with the 9 accelerometers and microphone taped in place for recording; B = close-up of one of the accelerometers; C = close-up of the microphone; D = illustration showing placement of 9 accelerometers (solid circles) and the microphone. See supplemental electronic material for color figure.

Figure 7.

Distribution of cavitation during ranges of motion. See supplemental electronic material for color figure.

Recordings During Initial Ranges of Motion

Recordings (320,000 Hz sampling rate)^19–21 were then conducted during full ranges of motion (ROMs, Fig. 3). The motions were: forward flexion, extension, left and right lateral flexion, and left and right axial rotation. Subjects were asked to stop the motion if/when s/he felt pain. After each motion subjects were asked if any sounds/vibrations related to joint movement were heard and/or felt in the spine during the motion.

Figure 3.

Accelerometer and microphone recordings were taken during 6 ranges of motion, including: flexion (A), extension, right and left (B) rotation, and right and left (C) lateral flexion. See supplemental electronic material for color figure.

Spinal Manipulation

Following full ROMs the subjects received side-posture lumbar SMT on both sides. Healthy subjects began with the left side up and LBP subjects began with the most painful side up. The accelerometers and microphone remained in place during the SMTs (Fig. 4). A slightly modified general SMT (hypothenar ilium technique)⁴¹ directed at the L3/L4–L5/S1 segmental levels was used. The modification was to the clinician’s “contact hand” (thrusting hand), which was moved from the usual up-side posterior superior iliac spine to the infero-lateral sacrum. The modification was to avoid contact with accelerometer wires and had been developed for previous studies that used accelerometers to assess cavitation during SMT.^19,21

Figure 4.

Spinal manipulation was performed with the accelerometers and microphone in place. Ultrasound of the right L3/4 Z joint is shown being conducted simultaneously for another component of the project that will be reported separately.

Recordings During Final ROMs

Recordings were then conducted (post-SMT) during full ROMs. Again, after each motion subjects were asked if s/he heard and/or felt any sounds/vibrations related to joint movement in his/her spine during the motion.

Accelerometer placement and taping were examined after each set of ROMs and after SMT. If contact between the skin and an accelerometer appeared to have loosened during the motions/SMT the accelerometer tape was adjusted until proper contact was made. Occasionally new tape was used to ensure accelerometer skin contact.

Assessment of Recordings

The data collected from the accelerometers were assessed using methods previously found to be reliable and valid for cavitation.^21,42,43 In this study, two observers, who each had over two years of experience reading accelerometer and microphone recordings in our lab, assessed each recording and reached consensus to identify crepitus and to determine the Z joint of origin for each instance of crepitus. Amplitude (measured as the change of voltage generated in the piezoelectric accelerometers) from the two primary accelerometers involved in a crepitus event (e.g., L2 and L3 accelerometers for either a left or right L2/3 crepitus) was also recorded, and the average of the two values was used as the amplitude for each crepitus. The accelerometer results were qualitatively compared with simultaneous recordings from the microphone to determine similarities and differences between the two acoustic methods and to determine the usefulness of the microphone in assessing crepitus.

Data Analysis

Data were analyzed descriptively to determine: distribution of crepitus, i.e., the specific Z joints from which crepitus originated; crepitus prevalence (primary outcome, average number of crepitus events/subject); comparison of crepitus prevalence from subjects of different age groups; comparison of crepitus prevalence in healthy and LBP subjects; crepitus amplitude (overall and for healthy and LBP subjects); crepitus recorded during each range of motion to determine the motions that produced the most and least crepitus; and agreement between subject report of crepitus and accelerometer recordings. In addition, comparison of crepitus before and after SMT was made for all of the groups of subjects (i.e., age groups, LBP subjects, and healthy subjects).

Results

Subjects

Ten subjects were enrolled. Distribution of subjects by age and sex is shown in Table 1. The difference between the ages of healthy vs. LBP subjects (healthy > LBP) was 13.0, ±7.01 (p=0.0504). Pain was rated as 3 (±1) for LBP subjects and 0 (±0) for healthy subjects. No adverse events were reported during the study.

Table 1.

Demographics of subjects

Age Group (in years)	Healthy1	Low Back Pain	Total
18–25	1	1	2
26–45	2	4	6
46–65	2	0	2
Total	5 (3 female, 2 male)	5 (2 female, 3 male)	10
Average Age (SD)	40.8 (15.1)	27.8 (4.2)
Age Range	23–61	24–35

¹Values are numbers of subjects

Accelerometry and Acoustic Methods

The accelerometry and acoustic methods were successfully implemented on all subjects, with simultaneous recordings from the accelerometers and the microphone being accomplished. Crepitus was identified as simultaneous deviations from baseline of multiple accelerometers (Fig. 5A and 4B). As in previous work, artifacts had a distinct appearance (Fig. 5C) and were easily distinguished from crepitus.⁵ The microphone was somewhat useful in helping to differentiate crepitus from artifacts; however, it was erratic in recording crepitus.

Figure 5.

Example of crepitus detected by accelerometers and microphone (see supplemental electronic material for color figure). Fig. 5A is a LabView oscilloscope recording of crepitus taken during left lateral flexion in a 55 year old male. The flat baseline can be seen at the far left, followed by the sharp response of several accelerometers and the distinct, broad waveform of the microphone. Fig. 5B is an expanded (by ~100X) timeline of the same crepitus. Again, notice that the accelerometers are at their individual baselines at the far left, followed by a dramatic simultaneous shift from the baseline of several accelerometers. The L2 (red) and L3 (green) accelerometers leave their baselines first, followed by L4 (gray) and the left accelerometer (dark blue). This was crepitus at the Left L2/3 Z joint. The microphone (light blue) has a very broad wave pattern that is best seen with the more contracted timeline (Fig. 5A). This distinct wave pattern from the microphone was rare. Fig. 5C is a recording of an artifact of the L1 and S2 accelerometers, which may have been caused by the subject’s gown bumping the accelerometers.

Distribution of Crepitus

Fig. 6 summarizes the incidence and distribution of crepitus before and after SMT. Overall, 23 instances of crepitus (Fig. 6B + 5D) were recorded from 15 different Z joints (Fig. 6A + 5C). Pre-SMT there were 14 instances of crepitus (Fig. 6B) recorded from 7 joints (Fig. 6A). Post-SMT there were 9 instances of crepitus (Fig. 6D) recorded from 8 joints (Fig. 6C). Crepitus could be recorded from the same joint of the same subject during different motions. For example, if crepitus was recorded from a subject’s right (R) L4/5 Z joint during flexion and left lateral flexion, this would be counted as 2 instances of crepitus (shown in Fig. 6B and 5D) from one Z joint (shown in Fig. 6A and 5C). Such multiple instances from the same Z joint but during different motions occurred from 2 Z joints recorded from 2 different subjects, one pre-SMT healthy, and one post-SMT LBP (1 Z joint for each subject).

Figure 6.

Figs. 6A and B report crepitus taken during motions before spinal manipulation (Pre-SMT) and Figs. 6C and D report crepitus taken during motions after spinal manipulation (Post-SMT). Figs. 6A and C report the individual zygapophyseal (Z) joints from which crepitus were recorded during full ranges of motion. Joints from which crepitus was recorded during more than one movement or multiple times during the same movement were counted only once in 6A and C. In 6B and D, if crepitus was recorded during multiple movements of the same joint, or multiple times during the same movement, each incidence of crepitus was counted. Results for healthy subjects are reported in bold typeface and those for low back pain (LBP) subjects are reported in italics. See supplemental electronic material for color figure.

Also, multiple instances of crepitus were counted for Fig. 6B and 5D if crepitus was recorded from the same Z joint during different times of the same motion (e.g., flexion) with all recordings reaching the baseline between instances. This occurred for 2 joints (2 different LBP subjects, both pre-SMT).

Generally, crepitus was recorded from different joints pre-SMT vs. post-SMT. There were only two exceptions to this, one in a healthy and one in a LBP subject when crepitus was recorded from the same joint before and after SMT. The amplitude was the same before and after SMT for the Z joint (L L4/5) of the healthy subject (0.01 mV). However, the exception in the LBP subject was unique in that 6 instances of crepitus were recorded from the R L4/5 Z joint during flexion pre-SMT (accelerometer recordings went to baseline between instances of crepitus) and only one was recorded post-SMT. The average amplitude of the six pre-SMT recordings was 0.10 mV and the amplitude of the single post-SMT R L4/5 recording was 0.04 mV.

Crepitus Prevalence

Table 2 presents the prevalence of crepitus. Overall crepitus prevalence (last column in Table 2) decreased pre-post SMT (average pre= 1.4 crepitus/subject, range 0–7 crepitus/subject vs. post= 0.9, range 0–2).

Table 2.

Total crepitus (and crepitus prevalence) by age group pre- and post SMT

Age Range Years (n)	18–251 Total Crepitus2 (Crepitus/Subject)3	26–45 Total Crepitus	46–65 Total Crepitus	Totals Total Crepitus
Pre-SMT
Healthy (n=5)	0 (0.0)3	0 (0.0)	4 (2.0)	4 (0.8)
LBP (n=5)	0 (0.0)	10 (2.5)	0 (0.0)	10 (2.0)
All (n=10)	0 (0.0)	10 (1.67)	4 (2.0)	14 (1.4)
Post-SMT
Healthy (n=5)	1 (1.0)	0 (0.0)	3 (1.5)	4 (0.8)
LBP (n=5)	0 (0.0)	5 (1.25)	0 (0.0)	5 (1.0)
All (n=10)	1 (0.5)	5 (0.83)	3 (1.5)	9 (0.9)

¹See Table 1 for numbers of subjects in each age group.

²All instances of crepitus, including the same joint producing crepitus in different motions

³Crepitus prevalence, i.e., crepitus/subject in first set of parentheses throughout

Prevalence progressively increased from the youngest to oldest age groups (pre-SMT= 0.0, 1.67, and 2.0, respectively, range 0–10; and post-SMT= 0.5, 0.83, and 1.5, range 1–5).

Prevalence was higher in LBP subjects than healthy subjects (pre-SMT-LBP= 2.0, range 0–7, vs. pre-SMT-healthy= 0.8, range 0–4; post-SMT-LBP= 1.0, range 0–2 vs. post-SMT-healthy= 0.8, range 0–2), even though healthy subjects were older than LBP (40.8 years vs. 27.8 years). Accounting for differences in age of the healthy vs. LBP subject groups by removing healthy 46–65 year-old subjects (n=2, no LBP subjects this age group), prevalence in LBP remained the same while that of healthy subjects decreased (pre-SMT-LBP= 2.0, range 0–7, vs. pre-SMT-healthy= 0.0, no range; post-SMT-LBP= 1.0, range 0–2 vs. post-SMT-healthy= 0.3, range 0–1).

Amplitude of Crepitus

With the exception of one LBP post-SMT crepitus of the left SI joint that had the amplitude of a cavitation (1.66 mV), crepitus amplitude was <0.16 mV. Average pre-SMT crepitus amplitude was 0.08 mV, which was greater than post-SMT crepitus, which was 0.03 mV, excluding outlier mentioned above. Excluding the one outlier, the pre- vs. post-SMT crepitus difference was greater in LBP (pre-SMT 0.09 mV, post-SMT 0.04) than healthy subjects (pre-SMT 0.04 mV, post-SMT 0.03 mV).

Crepitus During ROMs

Fig. 7 reports pre-SMT, post-SMT, and overall crepitus recorded during each motion. The most crepitus was recorded during flexion (pre-SMT eight, 57.1%, post-SMT six, 66.7%, and overall fourteen, 60.9%). For pre-SMT, flexion was followed by right lateral flexion (three, 21.4%) with extension, left lateral flexion and left rotation all producing the same amount of crepitus (one each, 7.1%). Post-SMT flexion was followed by extension (two, 22.2%) and right rotation (one, 11.1%), with no crepitus recorded in the other motions.

Agreement Between Subject Report of Crepitus and Recording of Crepitus

Subjects were more frequently in agreement with the recordings; however, this is largely accounted for by 94 instances of a possible 120 when the subjects heard no crepitus and none were recorded (78.3%). There were 3 instances (2.5%) when crepitus was heard/felt by the subject and recorded by the accelerometers. There were 23 (19.2%) instances when crepitus was either recorded but not heard/felt by the subject or heard/felt by the subject but not recorded.

Discussion

Summary of Main Findings

This feasibility study successfully collected crepitus data from the lumbar region of 10 subjects (5 healthy, 5 LBP) from three different age ranges. Overall, crepitus prevalence decreased following SMT. In addition, prevalence progressively increased from the youngest to oldest age groups, and was higher in LBP compared to healthy subjects. The majority of crepitus was recorded during flexion and subjects generally did not hear or feel the recorded crepitus.

The research extended previous work on cavitation during SMT^19–21 by evaluating recordings of crepitus made from both accelerometers and from a specialized directional microphone during full ROMs. These methods are novel to the assessment of spinal crepitus and the peer-reviewed literature has no record of a study assessing spinal crepitus by any means.

Accelerometers vs. Microphone

Even though the amplitude of crepitus was approximately one-tenth that of cavitation, the accelerometers performed well in this study and crepitus was identified as distinct events of multiple accelerometers responding in such a way that the specific Z joints of origin could be identified.

Previous reports using both accelerometers and a microphone indicated that each produced unique and complementary information in the assessment of other tissues,^25–29,36 and in the assessment of cavitation of the lumbar region associated with SMT.⁵ However in this study, even though the microphone recordings did move from the baseline during crepitus, the response was erratic. The amplitude of the microphone signal was close to baseline in all artifacts, which contrasted with the movement from baseline seen in crepitus. The artifacts in this study were very similar to those identified in previous research assessing cavitation and were usually the result of a subject’s gown or shorts bumping one or more accelerometers. This created a distinct appearance on the accelerometer recordings (Fig. 5C).⁵ As mentioned, the artifacts could have been identified without the use of a microphone. The microphone may not have been useful in this crepitus study because the amplitude of crepitus was markedly less than that previously recorded for cavitations.¹⁹ The low amplitude combined with the microphone located a considerable distance from most of the recorded crepitus, likely precluded adequate assessment of crepitus by the microphone. The microphone was positioned at T12 to avoid contact with any of the 9 accelerometers. Future studies of crepitus by our group will probably not include microphone recordings.

Z Joints as the Origin of Crepitus

Although other anatomic structures (e.g., muscles, ligaments) cannot be completely ruled out as the possible source of the recorded vibrations (crepitus), the Z joints are the most likely source of the recorded crepitus for two primary reasons. First, similar waveforms were identified from artificially produced crepitus conducted during validity testing of the methods⁴³(and unpublished data). These experiments were conducted on a spine embedded in silicone and vibrations originating directly from the Z joints were produced and recorded. Second, recordings not matching the types of waveforms previously seen from cavitations (only with lower amplitudes)^5,19–21,43 and artificial crepitus (unpublished data), produced in reliability testing, were not included as crepitus in this study. Consequently, crepitus may have potentially been underreported in this study; however, the methods increased the confidence that reported crepitus was of Z joint origin.

Age

Crepitus increased with age. This is consistent with the theory that crepitus increases with arthritic changes^2,3 and Z joint arthritic changes increase with age.^44–46 Although 8 subjects were enrolled from the older two age groups, as was the goal, no LBP subjects were enrolled from the oldest age group.

Crepitus in Healthy and LBP Subjects Pre- and Post-SMT

The finding that more crepitus was recorded in pre-SMT LBP than in healthy subjects in spite of the younger age of the LBP subjects was interesting. Following SMT (post-SMT), the number of crepitus events (both total crepitus and joints producing crepitus) in LBP and healthy subjects was almost identical (Fig. 6C and D). One hypothesis to explain the difference in pre-SMT crepitus is increased muscle tightness of the LBP subjects leading to relative compression of the articular surfaces making up the Z joints. Such compression could conceivably cause the normal smooth gliding motion of the Z joints to change to a series of small rapid motions, which could result in crepitus vibrations. The results of one LBP subject in the study may provide an example that supports this hypothesis. In this subject, crepitus was recorded from the same joint (Right L4/5) 6 times throughout the 4 seconds of pre-SMT flexion. The accelerometer recordings went completely to baseline between each instance of crepitus. Following SMT there was only one crepitus recorded from this joint during flexion. The average amplitude of the 6 pre-SMT recordings was 0.10 mV and the amplitude of the single post-SMT R L4/5 recording was 0.04 mV. This would suggest the joint was moving in a more regular (smooth) pattern following SMT. The small number of subjects in this feasibility study prevents further speculation on this issue.

Subject Report of Crepitus vs. Recorded Crepitus

Overall, subject report of crepitus was not an indicator of whether or not a crepitus was recorded. This finding is quite different from previous studies on cavitation during SMT where subject (and clinician) report of cavitation was closely related to recorded cavitation.^10,19,21 As discussed previously, the wave magnitude of crepitus was markedly less that that of a typical cavitation. The more subtle crepitus vibrations detected by the accelerometers likely went undetected by the subjects in this study. There were also 12 instances when the subject heard or felt a sound or vibration during a motion, but none was recorded. At first this seemed more difficult to explain. However, in a previous study there was an instance where a subject responded that s/he heard a cavitation, but none was recorded.¹⁹ We hypothesized that the cavitation in this instance occurred above the lumbar level and, therefore, was not recorded by the lumbar accelerometers. The same phenomenon may very well be occurring in this study. One difference being that during the full ROM recordings, the thoracic region was also going through full ROMs along with the lumbar region, whereas during the SMT cavitation study the SMT force was directed to the lumbar region, resulting primarily in lumbar segmental motion.^19–21,41 Consequently, much more motion occurred in the thoracic region in this study than in previous ones and we hypothesize that some of the crepitus heard/felt by the subjects was originating from the thoracic region and the lumbar-placed accelerometers did not respond to the subtle thoracic crepitus. Subjects were asked about sounds/vibrations in the spine, and most likely responded “yes” if any sound/vibration was heard/felt in the thoracic region. To further assess this issue, future studies should use procedures to minimize thoracic motion during ROM recordings and ask specific questions to verify the location of the reported sound/vibration.

Ranges of Motion

Flexion was without question the motion that produced the most crepitus (60.9% overall). This result is probably due to the lumbar region having considerably more motion in flexion (60^o) than any other motion.⁴⁷ Future studies using crepitus as a secondary outcome measure could perhaps be limited to recordings taken during flexion only in order to use time most efficiently. However, because crepitus was recorded during all 6 motions (Fig. 7), research using crepitus as a primary outcome should continue to record during all ROMs.

SMT and Crepitus

SMT had an effect on crepitus in this study, both the Z joints producing crepitus and the amplitude of crepitus (overall less amplitude post-SMT). However, the results indicate that the relationship between crepitus and SMT may be complex. Total crepitus was reduced following SMT in this study; however 9 instances of crepitus from 8 Z joints were recorded post-SMT (Fig. 6). Interestingly, with only two exceptions, the Z joints producing crepitus post-SMT were different from those that produced crepitus pre-SMT. This would seem to indicate a change beyond that of mere chance.

There may be several reasons for the pre- post-SMT differences. The Z joint movement (gapping) produced by SMT^17,37 may have added sufficient mobility to the pre-SMT crepitus-producing joints that they moved more freely and did not produce crepitus post-SMT. Contrariwise, the SMT force on pre-SMT non-crepitus producing joints that were somewhat hypomobile, may have induced enough motion that they began to move into a range that had not been used immediately pre-SMT, consequently the joint surfaces may have been contacting in facetal areas that were not as well lubricated (i.e., less synovial fluid in these regions) as the more frequently used facetal regions, thus creating crepitus as the joints entered the previously hypomobile zones. Much further work is needed to assess the veracity of this speculation.

Future research on larger numbers of healthy and LBP subjects from different ages would help to clarify the effects of SMT on crepitus. Combining studies assessing crepitus with imaging of the Z joints could potentially help to study the relationship between Z joint degeneration and crepitus that has been alluded to by other authors.^1–3 Diagnostic ultrasound (US), which is a cost-effective, low-risk (i.e., non-ionizing radiation) imaging modality,^48–50 would be an appealing candidate for complementary use with the accelerometry outcomes in future clinical mechanistic research. We are currently assessing the feasibility of adding US to future studies.

LIMITATIONS AND FUTURE RESEARCH

This was a feasibility study on a small number of subjects. Therefore, the results cannot be generalized to a larger population, but may be used to inform future studies. In addition, largely due to the convenience sample of faculty, staff, and students from a complementary and integrative health university, there were no LBP subjects in the 46–65 age group; consequently, the LBP subjects were younger than the healthy subjects. In addition, they and had only mild LBP (3 of 10). Even with these limitations, differences in crepitus were found for age, LBP vs. healthy subjects, the effects of SMT on crepitus, and the motions that produced crepitus.

Although reliability and validity studies have been completed for the phenomenon of cavitation during SMT,^19–21,43 further validity studies of the subtler phenomenon of crepitus are currently underway.

The findings of this small feasibility study indicate that further research in this area is merited. Crepitus in healthy and LBP subjects and in subjects from different age ranges and with Z joints of different levels of degeneration⁵¹ should be evaluated in studies fully powered to allow for generalizability of data to larger populations. Such studies should recruit from a larger population, as has been done in previous, larger clinical trials conducted by our research group.^37,52–55 The broader recruiting methods will likely lead to more LBP subjects from the older age group. In addition, the larger studies would allow for further assessment of SMT on crepitus, possibly shedding further light on the complex relationships between these two variables.

A power analysis was conducted using PROC POWER in SAS 9.4 (SAS Institute, Inc.; Cary, North Carolina) to determine the number of subjects needed for future clinical investigation. The sample size estimates were based on means and standard deviations from the descriptive statistics of this feasibility study. All power analyses conducted assumed two-sided tests, β=0.80, α=0.05, and a normal distribution of the data. Power analysis to compare pre- and post-SMT measures of crepitus, which would be the primary study objective, that was done using a paired t-test (SD=1.8; assuming r=0.5) showed that 28 subjects would be needed to complete the study to detect a mean difference of 1, which is hypothesized to demonstrate a meaningful change. Secondary objectives would be to determine if there are differences in mean crepitus between healthy and LBP subjects, as well as in younger subjects and older subjects (two sample t-test assuming equal variances; mean difference=1 crepitus/subject – large effect). To compare crepitus pre-SMT in healthy versus LBP subjects (SD=2.5), 100 subjects would need to be in each group. To compare pre-SMT crepitus in adults aged 26–45 years with adults aged 46–65 years (SD=2.7), 116 subjects would be needed in each group. Consequently, the ideal study would be 232 subjects with 116 LBP and 116 healthy subjects and the same numbers (i.e., 116) between the ages of 26–45 and between the ages of 46–65. Studies assessing purely the effects of SMT on crepitus could be conducted on many fewer (i.e., 28) total subjects.

In addition, the accelerometry methods could be further enhanced through the process of automation. The methods for analyzing each recording were extremely time consuming. Each recording took approximately one hour for a well-trained researcher to assess, and there were 12 recordings per subject, one for each of 6 motions both pre- and post-SMT. In addition, to increase reliability, 2 observers assessed each recording independently and then came to consensus. Therefore, approximately 240 hours were required to assess the recordings. Automation of the methods is feasible (personal communications with P.K. Raju, PhD and D. Marghitu, PhD, professors of acoustics at Auburn University)⁵ and accomplishing such automation would allow for more rapid assessment and quantification of the data. Increasing the speed of crepitus analysis would aid in future clinical studies.

Biomarkers are essential to the advancement of research into complementary and integrative health therapies, including manual therapies.⁵⁶ If the accelerometry analysis automation and studies outlined above are successful, joint vibrations (crepitus) identified during lumbar motion could eventually be used as an outcome measure (biomarker) to assess Z joint function, and to assess the effects of LBP treatments, including SMT, on the Z joints. Such new information could potentially help with the important efforts of identifying subpopulations of LBP patients who respond best to various types of treatments.^57,58

Conclusions

In this feasibility study, data collected from accelerometers provided useful information showing the prevalence of crepitus increased with age, was higher in LBP than healthy subjects, and overall decreased following SMT. Consequently, crepitus assessment using accelerometers has the potential of being an outcome measure/biomarker for providing meaningful information on Z joint function during movement and for providing information on the effects of LBP treatments (e.g., SMT) on Z joint function. Further research on larger numbers of subjects is warranted to assess crepitus in healthy and LBP subjects of different age ranges and different severities of degeneration. In addition, assessment of the effects of SMT on crepitus is also warranted.