10 2519jospt 2017 7348

Neuro-endocrine Response Following a Thoracic Spinal Manipulation in
Healthy Men
Kesava Kovanur Sampath, PT, MOst1
Erik Botnmark, PT1
Ramakrishnan Mani, PT, PhD1
James David Cotter, PhD
Rajesh Katare, PhD
Pujika Emani Munasinghe, BSc
Downloaded from www.jospt.org at University of Hong Kong Libraries on July 14, 2017. For personal use only. No other uses without permission.
Steve Tumilty, PT, PhD1.
Centre for Health, Activity, and Rehabilitation Research, School of Physiotherapy,

University of Otago, Dunedin, New Zealand.
School of Physical Education, Sport and Exercise Sciences, University of Otago,

Dunedin, New Zealand.
Department of Physiology-Heart Otago, University of Otago, Dunedin, New

Zealand.
Copyright ${year} Journal of Orthopaedic & Sports Physical Therapy. All rights reserved.
The study protocol was approved by the University of Otago Human Ethics
Committee (H14/150), Dunedin, New Zealand. The study was funded by a grant
from the New Zealand Manipulative Physiotherapist Association (NZMPA). The
authors declare that they have no affiliations with or financial involvement in any
organization or entity with a direct financial interest in the subject matter or materials
discussed in the article.
We sincerely thank Dr Andrew Gray for his expert statistical advice for this project.
Journal of Orthopaedic & Sports Physical Therapy
Address correspondence to Kesava Kovanur Sampath, Centre for Health, Activity,

and Rehabilitation Research, School of Physiotherapy, University of Otago, Dunedin,
New Zealand. Email: kesava.kovanur-sampath@otago.ac.nz.
1 Neuroendocrine Response Following a Thoracic Spinal Manipulation in Healthy Men
2 Abstract
3 Study design: Randomized controlled trial.
4 Background: Spinal manipulation (SM) can trigger a cascade of responses
5 involving multiple systems including the sympathetic nervous system (SNS) and the
6 endocrine system, specifically the hypothalamic-pituitary (HP) axis. However, no

7 manual therapy study has investigated the neuroendocrine response to SM (i.e.,
8 SNS-HP axis) in the same trial.
9 Objective: To determine short-term changes in SNS activity; heart rate variability
10 (HRV) and endocrine activity (cortisol, testosterone and testosterone/cortisol (T/C)
11 ratio) following a thoracic SM.

12 Methods: Twenty four healthy males aged between 18 and 45 years were
13 randomized into two groups: thoracic spinal manipulation (n=12) and sham (n=12).
14 Outcome measures were salivary cortisol (g/dL), salivary testosterone (pg/mL), T/C
15 ratio, HRV and changes in oxy-haemoglobin (O2Hb) concentration of the right calf
16 muscle (mol/L). Measurements were done before and at 5 minutes, 30 minutes and
17 approximately 6 hours after intervention.
18 Results: A statistically significant group by time interaction was noted for T/C ratio
19 (p<0.05) and salivary cortisol (p<0.01) concentrations. Significant between group
20 differences were noted for salivary cortisol concentration at 5 min (mean difference
21 (MD), 0.35; 95% CI: 0.12to0.6; interaction: p < .001) and T/C ratio at 6 hours post-
22 intervention (MD), -0.91; 95% CI: -1.69to-0.04, p < 0.05). However, SM did not
23 differentially alter O2Hb, testosterone or HRV relative to responses in the sham
24 group.
1
25 Conclusion: Thoracic SM resulted in an immediate decrease in salivary cortisol
26 concentration and reduced T/C ratio 6 hours after intervention. A pattern of
27 immediate sympathetic excitation was also observed in the SM group.
28 Key Words: autonomic nervous system, cortisol, spinal manipulation, sympathetic
29 nervous system, testosterone.
30
31
2
32 Spinal manipulation (SM) is a specific hands-on approach commonly used by
33 several different therapeutic professions to treat spinal pain4,35. Though the exact
34 mechanism through which SM operates are still unclear4; various mechanisms have
35 been proposed including changes in the autonomic nervous system (ANS),
36 sympathetic nervous system (SNS) and the endocrine system (hypothalamic-
37 pituitary-adrenal (HPA) axis)24.

38 Several studies have been undertaken to investigate the effects of SM on the
39 SNS8,9,76,87,89. However, there is no gold standard of non-invasive measurement of
40 changes in activity93. One reliable, non-invasive technique to measure SNS activity is
41 the use of Near Infrared Spectroscopy (NIRS)12,26,37. This method uses light in the
42 near infrared spectrum (600 1000 nm) from an optode attached to the skin.
43 Different wavelengths of this light are absorbed by either oxygenated (O2Hb) or
44 deoxygenated (HHb) hemoglobin and changes in the absorption of light can be used
45 to calculate changes in levels of O2Hb and HHb27,36. This in turn can be used to
46 measure blood flow (vasoconstriction/dilation) of aterioles in skeletal muscles. As the
47 autonomic regulation of skeletal muscle blood flow is dominated by the SNS 29,56,
48 OHB changes in skeletal muscles as measured by NIRS may provide a reliable
49 estimate of SNS activity.
50 Most tissues innervated by the SNS also receive innervation by the
51 parasympathetic nervous system (PNS)20; therefore the interplay between these two
52 systems effects how the tissues respond to stress/damage. Heart rate variability
53 (HRV), the beat-to-beat variation in heart rate is a well-established method of
54 assessing autonomic function82. Analysis of HRV in the frequency domain reveals
55 three oscillatory components in the spectral profile: (a) a very low-frequency (VLF)
56 band (< 0.04 Hz); (b) a low-frequency (LF) band (0.04 to 0.15 Hz) and (c) a high-
3
57 frequency (HF) band (0.15 to 0.4 Hz)55,82. While the HF component is related to the
58 PNS efferent activity, the LF component corresponds to both SNS and PNS efferent
59 activity55,60. Therefore the LF/HF ratio is interpreted as an indicator of cardiac
60 autonomic neural activity55. The use of LF/HF ratio has been criticized32,45 as it
61 oversimplifies the complex non-linear interactions between the SNS and the PNS
62 divisions of the ANS5. Despite this, the LF/HF ratio still remains a widely used tool to
63 assess autonomic cardiovascular regulation49,60,65,72. Previous investigations into the
64 effects of thoracic SM on HRV have produced conflicting results9,87. While Budgell
65 and Polus (2006) reported that a thoracic SM resulted in changes in HRV, Ward et
66 al. (2012) did not report any changes in ANS function following a thoracic SM.
67 Cortisol is released by the activation of HPA axis14. The use of salivary

68 sampling as a non-invasive tool for the measurement of free cortisol and thereby the
69 HPA axis activity has been well established46. Being a hormone, cortisol has a
70 known circadian rhythm that could be influenced by a number of confounding factors
71 including timing of sampling; number and nature of study days; day of data collection
72 (weekend vs. week days); drinking habits (Example: coffee and alcohol); eating
73 habits (food with high sugar content); dental hygiene/tooth brushing; smoking; and
74 intense physical activity before data collection15,80. Therefore, adequate control of
75 these factors is crucial for reliably measuring changes in cortisol before and after an
76 intervention. It is unclear how these confounding factors were addressed in previous
77 studies that have investigated the effects of SM on the HPA axis64,66,90 making it
78 difficult to interpret the findings.
79 It has been proposed that models need to include multiple measurements of
80 stress-related biological processes in order to understand the role of an individuals
81 response to stress48. For example, the hypothalamic-pituitary-gonadal (HPG) axis
4
82 and the HPA axis are known to interact with testosterone and cortisol mutually
83 inhibiting each other3,52. The balance between testosterone and cortisol (T/C ratio)
84 may therefore provide a better estimation of hypothalamic-pituitary (HP) axis
85 activity28 and has been used as a hormonal biomarker to determine over training in
86 athletes41,50 and susceptibility to certain diseases31,77.
87 When an individual is presented with a stressor, such as a painful injury, the

88 hypothalamus coordinates the stress response by activation of the HP axis and/or a
89 fast-acting neural component involving the SNS15. Therefore the hypothalamus
90 maintains homeostasis by the activation of the neuroendocrine system (SNS-HP
91 axis)15. It has been shown that SM results in rapid hypoalgesia with concurrent SNS
92 activation81,86. Given their integrated function, it could be hypothesised that SNS

93 changes following SM would be accompanied by changes in the HP axis 70. To our
94 knowledge, no manual therapy study has investigated the neuroendocrine response
95 to SM (i.e., SNS-HP axis) in the same trial. Moreover, studies have only explored
96 immediate changes (30 minutes to 2 hours) in the HPA axis activity. Therefore,
97 changes in HPA axis greater than 2 hours following SM is still unknown.

98 We hypothesised that a thoracic SM would result in neuroendocrine response
99 (SNS-HP axis). Activation of SNS will be demonstrated by an immediate
100 vasoconstriction (reduced O2Hb) of skeletal muscle in the calf measured by NIRS.
101 By measuring salivary cortisol and testosterone, it will be possible to understand
102 whether changes in the SNS activity are accompanied by changes in the HP axis
103 activity (T/C ratio). This may provide important mechanistic information that could be
104 of interest to physical therapists for two important reasons: (1) it has been well
105 established that the neuroendocrine mechanisms (HPA-SNS) play a crucial role in
106 the modulation of pain and inflammation15. Hence, neuroendocrine response can be
5
107 a potential mechanism of pain inhibition and tissue healing following SM (2) As
108 argued previously4, identification of SM mechanisms may increase the acceptance of
109 these techniques by health care providers. Therefore, the objectives(s) of this
110 randomized controlled trial (RCT) were to determine the short-term changes in SNS
111 activity (using NIRS); SNS/PNS balance (using HRV); changes in cortisol,
112 testosterone and the T/C ratio following a thoracic SM.

113 MATERIALS AND METHODS
114 Study Design
115 This was a randomised controlled trial using a repeated measures study
116 design. The ethical approval for this study was obtained from the University Human
117 Ethics Committee. All participants signed an informed consent prior to participation in
118 the study.
119 Participants
120
121 Participants were recruited from the wider community through flyers, posters
122 and e-mail. Healthy males between 18 and 45 years of age with no history of
123 significant pain or illness were eligible to participate in the study. Exclusion criteria
124 were: history of serious pathologic or psychiatric disorder; previous spinal (or other
125 relevant) surgery; any bone weakness or thoracic spine conditions such as
126 osteoporosis, osteopenia, Scheuermanns disease and ankylosing spondylitis;
127 currently on any medications; any contraindication to SM such as rib/spine fractures.
128 Randomisation and blinding
129 The randomisation schedule was prepared by a research administrator using
130 a computer generated random numbers table69. Participants were then block
6
131 randomized to either the intervention (SM) group or a sham intervention group. This
132 was done by the use of sealed opaque envelopes. A research assistant not
133 otherwise involved in the study did initial screening and randomisation. This ensured
134 that group allocation was concealed, and that the outcome assessor and the
135 participant were blinded. However, due to the nature of the intervention blinding of
136 the manual therapist delivering the intervention was not possible.
137 Procedure
138
139 Participants were invited to attend a screening session with a research
140 assistant not otherwise involved in the study. Participants who met the inclusion
141 criteria were given salivary collection devices, an instruction sheet about saliva
142 collection and the perceived stress scale (PSS-10)17. The PSS-10 is a self-reported
143 instrument with high reliability and validity designed to measure ones level of
144 perceived stress and has been widely used in research78. The PSS-10 was used in
145 this study as perceived stress is known to influence both the HPA axis and the ANS
146 activity15,17. Scores of 20 or higher are considered high stress. The participants were
147 also given a sheet to self-report (a) number of hours slept during the two days of
148 data collection (b) time of saliva sample collection and (c) any alcohol or medication
149 usage which was not anticipated.
150 The RCT took place in a laboratory where lighting, temperature and humidity
151 (24C, 40%) were kept constant as environmental factors have been shown to
152 influence HRV recordings42. Upon arrival, the participants height and weight were
153 recorded and any contra-indications for SM were screened. Participants returned the
154 completed PSS-10 scale and were given a 5-minute rest period following which pre-
7
155 intervention salivary samples (B2) were collected. The participant lay supine upon a
156 treatment plinth with their legs extended. During the data recording, participants
157 were asked to lie down as still as possible to avoid interference with data recording 79.
158 Salivary protocol
159
160 Saliva sample were collected via unstimulated passive drool as cotton based
161 sampling devices may introduce bias into the measurement75. The time of saliva
162 sample collection was standardised in a way that participants collected saliva at
163 approximately the same time on day-1 and day-2. For complete collection procedure
164 and protocol, please refer supplementary file-1.
165 Intervention
166
167 The participant was asked to cross the arms in front of the body placing each
168 hand on the opposite shoulder so that elbows were on top of each other in front of
169 the lower chest (figure 1). The physical therapist then rolled the participant over on
170 their side and placed the fixating hand, curled into a fist with the thumb and index
171 finger extended (commonly called the pistol grip)18, at the level of the fifth thoracic
172 vertebra, an area known to contain preganglionic neurons of the SNS 8,34,87. The
173 therapist then rolled the participant back to the supine position. High velocity low
174 amplitude thrust was delivered through the participants upper extremity and thorax
175 upon expiration18. A single thrust was used to standardise the intervention.
176 For the sham intervention the same setup was used, but the therapist did not
177 place a fixating hand against the thoracic spine and no thrust was given. Both the
178 intervention and sham intervention took approximately 10 seconds.
8
179 *********INSERT FIGURE 1 HERE*************************
180 Outcome measures
181
182 Salivary cortisol and testosterone
183
184 Salivary cortisol and testosterone have been shown to be a valid and reliable
185 surrogate measures of the activities of the HPA and HPG axes respectively23,40.
186 From these measurements the T/C ratio can be calculated.
187 O2Hb changes using NIRS
188
189 NIRS has been shown to demonstrate high degree of validity and intra-subject
190 reproducibility51. The standard error of measurement for O2Hb in skeletal muscle by
191 NIRS was calculated at 8.53%51. For this study, the point of largest circumference of
192 the right calf was identified and measured. The NIRS-optode was placed one third of
193 this distance medially from the anterior border of the tibia along the line of largest
194 circumference. This placed the optode over the medial belly of gastrocnemius, which
195 has previously been shown to be a reliable placement of a NRIS optode 79 (figure 1).
196 The leg was supported on pillows under the knee and ankle to just suspend the
197 NIRS-optode from the table to eliminate any movement of the optode in relation to
198 the skin. The NIRS-optode was held in place by double sided adhesive tape, and an
199 elastic bandage was loosely wrapped around the calf to eliminate disturbances from
200 any outside light.
201 HRV measurement
202
9
203 For changes in HRV, the LF/HF ratio was the outcome measure as this has
204 been suggested to best represent changes in autonomic activity82. HRV was
205 recorded using single use disposable electrodes (Blue Sensor SP, Ambu,
206 Denmark), and a lead 2 ECG setup was used. The ECG signal was fed through a
207 BioAmp and PowerLab 16/30 (ADInstruments, New Zealand).
208 Data recording

209
210 Data were recorded using LabChart 8 (ADInstuments, Dunedin, New Zealand,
211 www.adinstruments.com). NIRS and ECG data were continuously recorded from 5
212 minutes pre-intervention to 30 minutes post-intervention. Salivary samples were
213 collected at 8 different time points on two consecutive week days to reduce the
214 within subject variation of situational factors such as daily schedules15. On day 1,
215 participants collected baseline salivary samples at home on 3 different time points
216 (morning (AM) afternoon and night (PM)). On day 2, salivary samples were collected
217 at pre-intervention (B - 5 minutes before intervention) and post-intervention at 3 time

218 points (P1- 5 minutes, P2 30 minutes and P3 night (PM), approximately 6 hours
219 post-intervention). The experimental timeline is given in figure 2.
220 *********INSERT FIGURE 2 HERE*************************
221 Data processing
222
223 For NIRS data, mean 30s blocks of OHB before intervention were calculated,
224 as this has have been recommended to achieve accurate readings36. This was set to
225 zero, as the equipment used in this study calculates change from baseline. Change
226 scores from baseline were then calculated immediately at 1 minute, 5 minutes and
10
227 30 minutes after the intervention. For HRV data, 5 minutes before, 5 minutes and 30-
228 minutes after the intervention were analysed, as analysis of 5 minute blocks is
229 recommended for short term ECG recordings82.
230 Salivary biomarker analysis
231
232 The level of testosterone and cortisol in all the test samples were measured
233 using commercially available ELISA kit (Salimetrics, USA) according to the
234 manufacturers instruction. Supplementary file-1 provides a brief description of lab
235 procedure used.
236 Power calculations
237
238 Of the 3 outcome measures (the minimum clinically important difference of
239 which are currently unknown), cortisol was considered primary. From previous
240 literature13,64,66,90, an effect size of 0.25 was considered practical in studies involving
241 measuring cortisol in normal individuals. Based on the repeated measures mixed
242 design of the current study, a power calculation was performed using G power
243 (version 3.0.1.0). A sample size of n = 24 was required to provide 80% power at the
244 5% level of significance for an effect size of 0.25.
245 Statistical analysis
246
247 Data for continuous variables were expressed as mean SD. Raw cortisol
248 values were log transformed to address non-normality and skewness43. Any outliers
249 were handled by winsorising the data point54. The Shapiro-Wilk and Levene tests
250 were performed to assess normality and homoscedasticity67. A two-way mixed model
11
251 with absolute agreement was used to calculate the intra-class correlation coefficient
252 (ICC (3, 1)) for the inter-day reliability (day 1 and day 2) of morning and afternoon
253 salivary cortisol and testosterone values. An ICC 0.20 was defined as poor, an ICC
254 between 0.21 to 0.40 was defined as fair, an ICC between 0.41 and 0.60 was
255 defined as satisfactory, an ICC between 0.61 and 0.80 was defined as good, and an
256 ICC 0.81 was defined as an excellent agreement between both evaluations 88. A
257 repeated measures analysis of variance (RM-ANOVA) was performed to test the
258 effect of the between-group factor (sham or SM) and within group factor (time) on the
259 dependant variables (salivary cortisol, salivary testosterone, T/C ratio, HRV and
260 OHB)67. The hypothesis of interest was the group-by-time interaction. A post-hoc
261 pairwise comparison between groups at different time points was also carried out. A
262 Bonferroni correction was used for adjusting for multiple post hoc comparisons 1.
263 Eta-square was used for measuring effect sizes. Management and analysis of data
264 were performed using the statistical package SPSS for windows version 22.0 (SPSS
265 Inc, Chicago, IL). The level of statistical significance was set at p<0.05.
266 RESULTS
267
268 The flow of participants is shown in figure 3. The data collection for the study
269 happened between March and June 2015. The data collection was stopped after the
270 24th participant as no loss to follow-up was anticipated given the one-off nature of the
271 intervention.
272 *********************INSERT FIGURE 3 HERE********************
273 There was no missing data. Participants did not vary in baseline
274 characteristics, expect for height (table 1). No difference was found in the number of
12
275 hours slept during the two days of data collection; alcohol or medication usage
276 between groups.
277 *********************INSERT TABLE 1 HERE********************
278 The ICC calculated between day 1 and day 2 for salivary cortisol samples
279 indicated strong agreement (0.71) between the morning measures; a satisfactory
280 agreement (0.47) between the afternoon measures. The ICC calculated for salivary
281 testosterone indicated a fair agreement (0.4) between the morning measures; and a
282 strong agreement in the afternoon samples (0.78).
283 T/C ratio
284 There was a statistically significant interaction of group by time for T/C ratio (p
285 < 0.05), with an effect size of 0.11. Between groups analysis (table 2) showed
286 statistically significant difference between the sham and SM groups at only one post-
287 intervention time point, night (PM samples, mean difference (MD), -0.91; 95% CI: -
288 1.69to-0.04; p=0.02). Figure 4 shows group changes in T/C ratio.
**************INSERT FIGURE 4 HERE************************

289
290 Salivary cortisol
291 There was a statistically significant interaction of group by time for salivary
292 cortisol concentration (p < .001), with an effect size of 0.28. Pairwise comparison
293 between groups (table 2) at each post-intervention time points revealed a statistically
294 significant difference between the sham group and SM group at 5 minutes post-
295 intervention (mean difference (MD), 0.35; 95% CI: 0.12 to 0.6). However, no
296 statistically significant interaction of group by time was found for salivary cortisol at
297 30-minutes or night (6 hours) post-intervention. Within group comparison showed
13
298 that salivary cortisol levels were lower (significant) post-intervention (5 and 30
299 minutes) compared to pre-intervention levels in the SM group (p < .05).
300 Salivary testosterone
301 No statistically significant interaction of group by time was found for salivary
302 testosterone levels (p=0.33), with an effect size of 0.05. No statistically significant
303 difference was found between or within groups at any of the post-intervention time
304 points (table 2).
305 ************INSERT TABLE 2 HERE**************
306 Heart rate variability
307 No statistically significant interaction of group by time was found for HRV
308 (p=0.75), with an effect size of 0.13. No statistically significant difference (between or
309 within groups) was found for HRV at any of the post-intervention time points (table
310 3).
311 Oxy-haemoglobin
312 No statistically significant interaction of group by time was found for OHB
313 (p=0.32), with an effect size of 0.05. No statistically significant difference (between
314 groups) was found in OHB concentration at any post-intervention time points (table
315 3). However in the SM group, a statistically significant increase in O2HB
316 concentration was found at 30 minutes post-intervention (p < .05) compared to
317 O2HB values at baseline, 1-minute and 5-minutes post-intervention.
318 ************INSERT TABLE 3 HERE**************
319
14
320 DISCUSSION
321
322 We hypothesised that a thoracic SM will result in a neuroendocrine response,
323 change in HP axis activity, as observed by the T/C ratio, salivary cortisol and OHB
324 concentration. We found statistically significant interaction of group by time in T/C
325 ratio approximately 6 hours post-intervention and salivary cortisol levels at 5-minutes
326 post-intervention. Though statistical significance between groups was not achieved
327 for OHB or HRV, decreases in OHB levels suggest an immediate vasoconstriction
328 in the SM group.
329 T/C ratio

330 Our study found that the T/C ratio was significantly lower in the SM group
331 compared to the sham group 6 hours post-intervention. This may be because of
332 changes in cortisol levels as the testosterone levels were similar in both groups 6
333 hours following intervention. When compared with the sham group (Mean Difference
334 (MD)- 0.58), a lesser drop in salivary cortisol levels was evidenced from 30-minutes
335 post-intervention to 6 hours post-intervention in the SM group (MD- 0.14). This may
336 indicate activation of the HPA axis following SM.
337 Cortisol has widespread effects on glucose, fat and protein metabolism and
338 can stimulate gluconeogenesis that can be used as building blocks of tissue repair25;
339 thereby playing an important role in tissue healing. It is important to note that tissue
340 healing is a complex process and may be influenced by other hormones produced by
341 the pituitary such as growth hormone and thyroid hormone63. However, to measure
342 these were beyond the scope of this study. On the other hand, elevated cortisol level
343 has been associated with dysregulated insulin, impaired glucose metabolism,
15
344 abdominal obesity, osteoporosis, hyperlipidemia and hypertension 7,53,91. Therefore,
345 our findings should be interpreted with caution and need to be verified by future
346 studies using a series of interventions replicating clinical practice.
347 The T/C ratio had been widely used in sports and exercise science research
348 as an indicator of overtraining and recovery. To our knowledge, this is the first study
349 to investigate the effects of thoracic SM on T/C ratio. Previously, studies have used
350 only cortisol as an indicator of HPA axis activity13,64,66,90. However, it is now clear
351 that the HPA axis can be influenced by other endocrine mechanisms (HPG axis
352 specifically). Hence, it has been advocated that the balance between these two
353 hormones may provide a better estimation of HPA axis activity3,19. The exact clinical
354 utility of changes in T/C ratio noted in our study is unclear. Despite this, our findings
355 may provide a platform for future studies to explore further on this area of manual
356 therapy research.
357 Salivary cortisol and testosterone
358 There was a significant decrease in salivary cortisol levels at 5-minutes post-
359 intervention in the SM group compared to sham group. Cortisol is regulated by the
360 HPA axis through feedforward and feedback loops11,92. The glucocorticoid mediated
361 feedback loop operates in at least three time domains: fast (within seconds to
362 minutes), intermediate (2 to 10 hours) and slow (over hours to days)44. Animal
363 models further demonstrate that an innocuous stimulus (such as SM) applied to the
364 spine can have inhibitory effects on adrenal gland activity10. Therefore it could be
365 argued that immediate (5 minutes) drop in cortisol level was through reflexive
366 inhibition in adrenal activity via SM. This is in contrast with findings from previous
367 studies13,90 that reported no changes in cortisol levels immediately following a SM.
16
368 While our intervention was a thoracic SM; Whelan et al (2002) used an upper
369 cervical SM; it is unclear whether Christian et al (1998) used a cervical or a thoracic
370 SM or both. Our data collection happened in the afternoon when cortisol levels are
371 shown to be more stable; whereas, the salivary/plasma cortisol were collected during
372 late morning in the other studies13,90, which may partly explain the difference in
373 findings.
374 The drop in cortisol at 5 minutes post-intervention may then trigger feedback
375 loop mechanisms resulting in an increase in cortisol concentration observed 6 hours
376 following SM. However, it is to be noted that many factors including the circadian
377 rhythm, ultradian rhythms and stress levels (to name a few) may influence blood
378 corticosterone concentrations58. The PSS-10 scores were similar in both groups.
379 Hence, the difference in cortisol levels may not be due to difference in baseline
380 perceived stress levels. Most participants in our study received SM for the first time.
381 It could be argued that the participants perceived the act of SM as stressful. If this
382 was true, then an increase in cortisol levels should have been noticed immediately
383 rather than many hours following SM. However, we found no significant changes in
384 salivary cortisol levels between the two groups at 30-minutes post-intervention. This
385 is in agreement with previous findings66. Therefore our finding strengthens the fact
386 that the act of SM itself may not induce a state of stress as noted previously85.
387 Alternatively it could be argued that SM had an influence on the opioid system;
388 thereby, influencing the HPA axis6. However, this can only be speculated as we did
389 not measure the opioid system as part of this study.
390 The baseline salivary cortisol and testosterone levels reported in our study are
391 consistent with the literature38. Both cortisol and testosterone exhibit circadian
392 rhythmicity with peak concentrations in the morning and reduced concentrations in
17
393 the evening and overnight84. As mentioned previously, the circadian rhythmicity of
394 hormones can be affected by a number of confounding factors15,80. It is unclear how
395 these confounding factors were addressed in previous studies13,64,90 which may also
396 partly explain the conflicting results reported in the studies. Therefore, participants in
397 our study had to follow a strict protocol before saliva collection (refer supplement).
398 Also, the data collection happened only on Wednesday and Thursday as the day of
399 data collection (weekend vs. week days) has been shown to influence stress and
400 hence cortisol levels46. Calculating the inter-day reliability (i.e., ICC) for both the
401 hormones therefore was a crucial aspect of our design. The ICC calculated for
402 morning and afternoon cortisol and testosterone concentration in this study are
403 consistent with the literature39,90. Establishing the ICC enabled reliable analysis of
404 pre and post-intervention cortisol and testosterone levels on day-2.

405 Oxy-haemoglobin
406 Findings from our study indicated that within SM group, OHB reduced
407 immediately (one minute) after thoracic SM and steadily rose with changes at 30-
408 minutes being statistically different from baseline, 1-minute and 5-minute values. It is
409 to be noted that between group changes did not reach statistical significance.
410 Previously studies investigating the effects of SM have routinely used skin blood flow
411 (SBF) as a measure of SNS activity. However, as SBF is not only mediated by the
412 SNS, but is also dependant on many other factors20, interpretation of changes have
413 recently been questioned93. In that systematic review by Zegarra-Parodi et al.
414 (2014), not only were the included studies heterogeneous in nature, but the extent to
415 which SBF changes reflect SNS changes in deeper tissues still remains unclear 93.
416 Therefore, we used NIRS to measure OHB changes in calf muscle as an indicator
417 of SNS activity. NIRS has been shown to reliably measure changes in skeletal
18
418 muscle blood flow when compared to established methods such as Doppler
419 ultrasound26 and functional magnetic resonance imaging (f-MRI)21.
420 Our study did not find significant between group changes in OHB of calf
421 muscle. Research has shown that SM elicits extra segmental effects and changes in
422 SNS activity has been suggested as one possible explanation for these effects. By
423 measuring blood flow in calf muscle, which has no segmental connection to the
424 thoracic spine, extra segmental effects of thoracic SM via the SNS could be
425 assessed. Absence of significant between group changes in OHB found in our study
426 may reflect ongoing challenges with regards to measuring autonomic activity 93.
427 Further, the precise location and extent of vasoconstriction within skeletal muscle
428 following SNS activation can be difficult to ascertain83. Alternatively, other reliable
429 markers of sympathetic activity such as salivary--amylase68 may be considered for
430 future studies exploring these changes following SM.
431 Heart rate variability
432 No significant difference in HRV (LF/HF ratio) between the sham and the SM groups
433 was found. A thoracic SM could have a stimulatory action on the ANS (especially the
434 SNS); considering the anatomical and physiological relationship of the thoracic spine
435 and the SNS. Similar to previous findings76,87 we found no changes in HRV values
436 following a thoracic SM. However this is in contrast to Budgell and Polus 9 who
437 reported that a prone HVLA of thoracic spine may influence the autonomic output to
438 the heart that is not duplicated by a sham procedure. It is important to note that the
439 SM technique used in our study (supine HVLA thrust) is different from that used in
440 the other study (prone HVLA thrust). Therefore, the results may not be comparable.
441 There have been disagreements over the accuracy of LF/HF ratio as a marker of
19
442 cardiac autonomic activity5,57,73. For instance, some authors dispute the direct
443 contribution of sympathetic activity to LF components of the HRV and attribute it to
444 the modulating baroreflex activity33,61. The inaccuracy of the test may also be a
445 reason why the findings are different between studies. However ratios of spectral
446 power (LF/HF ratio) continue to be used in research to provide better information
447 about sympathovagal balance, especially in short term recordings60.

448 In summary, our findings indicate a quick response from the SNS followed by
449 a longer lasting slow response from neuroendocrine system following SM. This may
450 provide justification for the use of SM in the early phases of injury to enhance
451 physiological processes either before the patient can start exercising or as an
452 adjunct to exercise, regardless of the site of injury. Previous manual therapy studies
453 have explored individual biomechanical30 and neurophysiological effects86 with the
454 potential interaction of these effects often being overlooked4. In chronic pain
455 population however, dysregulation of both the ANS and the HPA axis have been
456 reported74. Therefore, our findings may provide preliminary evidence for the
457 combined effects (i.e. neuroendocrine) of thoracic SM, the therapeutic benefits of
458 which needs further investigation in symptomatic population.
459
460 Strength and limitations
461 One of the major strength of the study is its design. We used a repeated
462 measures design collecting salivary samples at different time points as indicated by
463 a review2. We were able to successfully address various methodological factors that
464 could confound hormonal measurements. The atmospheric chamber used during the
465 experiment allowed us to control various parameters such as the temperature and
20
466 humidity, which are known to have an effect on endocrine as well as ANS functions.
467 Further, measuring testosterone to enable calculation of the T/C ratio, an indicator of
468 HPA activity is also unique to our study. This may improve our understanding of
469 complex interactive hormonal response to SM. The use of NIRS to measure SNS
470 activity is also novel to our study. Because of the one-off nature of the intervention
471 used in this study, there was no missing data or attrition of participants.
472 The study is not without its limitations. The effect size for T/C ratio (0.11) and
473 cortisol (0.28) changes noted in our study can be considered as medium16,22. This
474 could be because the response of HP axis is known to be minimal in healthy
475 individuals and has been shown to be amplified in painful population 59,62. Therefore,
476 the magnitude of effects should be bigger in symptomatic population and needs to
477 be verified by future research. Females were not included in the current study
478 because it is well known that the HPA and HPG axis response differs between men
479 and women, especially during menstrual cycle and contraceptive usage 19,47. Future
480 research would either stratify participants by gender or adjust the hormonal data
481 (females) for known confounders. We found changes in HP axis activity following SM
482 and measured only two hormones.
483 An apriori power calculation to determine the sample size was undertaken to
484 achieve an effect size of 0.25. This effect size can be considered as medium effect
485 size22. Further, a study71 indicated that a minimum of 24 participants are required for
486 studies involving HRV. Hence a sample size of 24 was considered adequate for this
487 mechanistic study. However, it is to be noted that the power calculations were not
488 based on the minimally clinically important difference (MCID) of the primary outcome
489 measures which are currently unknown. Therefore the possibility of a type-II error
490 cannot be ruled out. We expect minimal chance bias as baseline difference between
21
491 groups were not substantial and we consider our randomisation process as
492 successful.
493 CONCLUSION
494
495 Thoracic SM has an effect on HPA axis activity as indicated by changes in
496 salivary cortisol immediately and T/C ratio many hours following SM. A thoracic SM
497 may also have immediate effects on the SNS as indicated by OHB levels. The
498 clinical implication of these changes are however unclear. More research is therefore
499 warranted in this area.
500 KEY POINTS
501 Findings: Thoracic spinal manipulation can reduce salivary cortisol levels
502 immediately and T/C ratio many hours following intervention.
503 Implications: This study may provide preliminary evidence that thoracic spinal
504 manipulation may influence tissue healing via modulation of the neuroendocrine
505 system.
506 Cautions: this studys findings are based on healthy male participants. Therefore, the
507 nature of neuroendocrine response in females is still unknown. Future research on
508 symptomatic population is required to ascertain the clinical utility of neuroendocrine
509 response following SM noted in this study.
510
22
511 REFERENCES
512 1. Abdi H. Bonferroni and idk corrections for multiple comparisons. Encyclopedia of
513 Measurement and Statistics. 2007;3:103-107.
514 2. Adam EK, Kumari M. Assessing salivary cortisol in large-scale, epidemiological research.
515 Psychoneuroendocrinology. 2009;34(10):1423-
516 1436.http://dx.doi.org/10.1016/j.psyneuen.2009.06.011
517 3. Bedgood D, Boggiano MM, Turan B. Testosterone and social evaluative stress: The
518 moderating role of basal cortisol. Psychoneuroendocrinology. 2014;47:107-
519 115.http://dx.doi.org./j.psyneuen.2014.05.007.
520 4. Bialosky JE, Bishop MD, Price DD, Robinson ME, George SZ. The mechanisms of manual
521 therapy in the treatment of musculoskeletal pain: a comprehensive model. Man Ther.
522 2009;14:531-538.http://dx.doi.org/10.1016/j.math.2008.09.001.
523 5. Billman G. The effect of heart rate on the heart rate variability response to autonomic
524 interventions. Front Physiol. 2013;4(222):130-
525 138.http://dx.doi.org/10.3389/fphys.2013.00222.
526 6. Brennan PC, Kokjohn K, Kaltinger CJ, et al. Enhanced phagocytic cell respiratory burst
527 induced by spinal manipulation: potential role of substance P. J Manipulative Physiol Ther.
528 1991;14(7):399-408.
529 7. Brown ES, Varghese FP, McEwen BS. Association of depression with medical illness: does
530 cortisol play a role? Biol Psychiatry. 2004;55(1):1-9.http://dx.doi.org/10.1016/S0006-
531 3223(03)00473-6.
532 8. Budgell B, Hirano F. Innocuous mechanical stimulation of the neck and alterations in heart-
533 rate variability in healthy young adults. Auton Neurosci. 8/13/ 2001;91(12):96-
534 99.http://dx.doi.org/10.1016/S1566-0702(01)00306-X.
535 9. Budgell B, Polus B. The effects of thoracic manipulation on heart rate variability: a controlled
536 crossover trial. J Manipulative Physiol Ther. 2006;29(8):603-
537 610.http://dx.doi.org/10.1016/j.jmpt.2006.08.011.
538 10. Budgell B, Sato A, Suzuki A, Uchida S. Responses of adrenal function to stimulation of lumbar
539 and thoracic interspinous tissues in the rat. Neurosci Res. 1997;28(1):33-
540 40.http://dx.doi.org/10.1016/S0168-0102(97)01173-5.
541 11. Calogero AE, Gallucci WT, Gold PW, Chrousos GP. Multiple feedback regulatory loops upon
542 rat hypothalamic corticotropin-releasing hormone secretion. Potential clinical implications. J

543 Clin Invest. 1988;82(3):767.http://dx.doi.org/10.1172/FJCI113677.
544 12. Chavoshan B, Sander M, Sybert TE, Hansen J, Victor RG, Thomas GD. Nitric oxide dependent
545 modulation of sympathetic neural control of oxygenation in exercising human skeletal
546 muscle. J Physiol. 2002;540(1):377-386.http://dx.doi.org/10.1113/jphysiol.2001.013153.
547 13. Christian GF, Stanton GJ, Sissons D, et al. Immunoreactive ACTH,[beta]-Endorphin, and
548 Cortisol Levels in Plasma following Spinal Manipulative Therapy. Spine. 1988;13(12):1411-
549 1417.
550 14. Chrousos GP. Stress and disorders of the stress system. Nat Rev Endocrinol. 2009;5(7):374-
551 381.http://dx.doi.org/10.1038/nrendo.2009.106.
552 15. Chrousos GP, Gold PW. The concepts of stress and stress system disorders: overview of
553 physical and behavioral homeostasis. JAMA. 1992;267(9):1244-
554 1252.http://dx.doi.org/10.1001/jama.1992.03480090092034.
555 16. Cohen J. Statistical power analysis for the behavioral sciences: Routledge Academic.
556 Statistical power analysis for the behavioral sciences: Routledge Academic. 2013.
557 17. Cohen S, Kamarck T, Mermelstein R. A global measure of perceived stress. J Health Soc
558 Behav. 1983:385-396.
559 18. Cook C. Orthopedic Manual Therapy. New York: Pearson Higher Ed; 2011.
23
560 19. Craddock TJ, Fritsch P, Rice Jr MA, et al. A Role for Homeostatic Drive in the Perpetuation of
561 Complex Chronic Illness: Gulf War Illness and Chronic Fatigue Syndrome. PLoS One.
562 2014;9(1):e84839.http://dx.doi.org/10.1371/journal.pone.0084839.
563 20. Cramer GD, Darby SA. Clinical anatomy of the spine, spinal cord, and ANS. 3rd ed. St Louis,
564 MO: Elsevier Health Sciences; 2013.
565 21. Cui X, Bray S, Bryant DM, Glover GH, Reiss AL. A quantitative comparison of NIRS and fMRI
566 across multiple cognitive tasks. Neuroimage. 2011;54(4):2808-
567 2821.http://dx.doi.org/j.neuroimage.2010.10.069.
568 22. Cunningham JB, McCrum-Gardner E. Power, effect and sample size using GPower: practical
569 issues for researchers and members of research ethics committees. Evidence-Based
570 Midwifery. 2007;5(4):132-137.
571 23. Dabbs JM. Salivary testosterone measurements: Reliability across hours, days, and weeks.
572 Physiol Behav. 1990;48(1):83-86.http://dx.doi.org/10.1016/0031-9384(90)90265-6.
573 24. Evans DW. Mechanisms and effects of spinal high-velocity, low-amplitude thrust
574 manipulation: previous theories. J Manipulative Physiol Ther. 2002;25(4):251-
575 262.http://dx.doi.org/10.1067/mmt.2002.123166.
576 25. Evans KD, Douglas B, Bruce N, Drummond PD. An exploratory study of changes in salivary
577 cortisol, depression, and pain intensity after treatment for chronic pain. Pain Med.
578 2008;9(6):752-758.http://dx.doi.org/10.1111/j.1526-4637.2006.00285.x.
579 26. Fadel PJ, Keller DM, Watanabe H, Raven PB, Thomas GD. Noninvasive assessment of
580 sympathetic vasoconstriction in human and rodent skeletal muscle using near-infrared
581 spectroscopy and Doppler ultrasound. J Appl Physiol. 2004;96(4):1323-
582 1330.http://dx.doi.org/10.1152/japplphysiol.01041.2003.
583 27. Ferrari M, Mottola L, Quaresima V. Principles, techniques, and limitations of near infrared
584 spectroscopy. Can J Appl Physiol. 2004;29(4):463-487.

585 28. Filaire E, Bernain X, Sagnol M, Lac G. Preliminary results on mood state, salivary
586 testosterone: cortisol ratio and team performance in a professional soccer team. Eur J Appl
587 Physiol. 2001;86(2):179-184.http://dx.doi.org/10.1007/s004210100512.
588 29. Fleming BP, Gibbins IL, Morris JL, Gannon BJ. Noradrenergic and peptidergic innervation of
589 the extrinsic vessels and microcirculation of the rat cremaster muscle. Microvasc Res.
590 1989;38(3):255-268.
591 30. Gal J, Herzog W, Kawchuk G, Conway PJ, Zhang Y-T. Movements of vertebrae during
592 manipulative thrusts to unembalmed human cadavers. J Manipulative Physiol Ther.

593 1997;20(1):30-40.
594 31. Glenn AL, Raine A, Schug RA, Gao Y, Granger DA. Increased testosterone-to-cortisol ratio in
595 psychopathy. J Abnorm Psychol.
596 2011;120(2):389.http://psycnet.apa.org/doi/10.1037/a0021407.
597 32. Goldberger JJ, Ahmed MW, Parker MA, Kadish AH. Dissociation of heart rate variability from
598 parasympathetic tone. Am J Physiol. 1994;266(5 Pt 2):H2152-2157.
599 33. Goldstein DS, Bentho O, Park MY, Sharabi Y. Lowfrequency power of heart rate variability is
600 not a measure of cardiac sympathetic tone but may be a measure of modulation of cardiac
601 autonomic outflows by baroreflexes. Exp Physiol. 2011;96(12):1255-
602 1261.http://dx.doi.org/10.1113/expphysiol.2010.056259.
603 34. Groen GJ, Baljet B, Drukker J. Nerves and nerve plexuses of the human vertebral column.
604 American Journal of Anatomy. 1990;188(3):282-296.
605 35. Gross A, Miller J, DSylva J, et al. Manipulation or mobilisation for neck pain: a Cochrane
606 review. Man Ther. 2010;15(4):315-333.http://dx.doi.org/10.1016/j.math.2010.04.002.
607 36. Hachiya T, Blaber A, Saito M. Nearinfrared spectroscopy provides an index of blood flow
608 and vasoconstriction in calf skeletal muscle during lower body negative pressure. Acta
609 Physiologica. 2008;193(2):117-127.http://dx.doi.org/10.1111/j.1748-1716.2007.01827.x.
24
610 37. Hansen J, Thomas GD, Harris SA, Parsons WJ, Victor RG. Differential sympathetic neural
611 control of oxygenation in resting and exercising human skeletal muscle. J Clin Invest.
612 1996;98(2):584.http://dx.doi.org/10.1172/JCI118826.
613 38. Hayes LD, Bickerstaff GF, Baker JS. Interactions of cortisol, testosterone, and resistance
614 training: influence of circadian rhythms. Chronobiol Int. 2010;27(4):675-
615 705.http://dx.doi.org/10.3109/07420521003778773.
616 39. Hayes LD, Grace FM, Kilgore J, Young J, Baker J. Diurnal variation of cortisol, testosterone,
617 and their ratio in apparently healthy males. Sport SPA. 2012;9(1):5-13.
618 40. Hellhammer DH, Wst S, Kudielka BM. Salivary cortisol as a biomarker in stress research.
619 Psychoneuroendocrinology. 2009;34(2):163-
620 171.http://dx.doi.org/10.1016/j.psyneuen.2008.10.026.
621 41. Hug M, Mullis PE, Vogt M, Ventura N, Hoppeler H. Training modalities: over-reaching and
622 over-training in athletes, including a study of the role of hormones. Best Pract Res Clin
623 Endocrinol Metab. 2003;17(2):191-209.http://dx.doi.org/10.1016/S1521-690X(02)00104-5.

624 42. Karim N, Hasan JA, Ali SS. Heart rate variabilitya review. Australian Journal of Basic Applied
625 Science. 2011;7(1):71-77.
626 43. Keene ON. The log transformation is special. Statistics in medicine. 1995;14(8):811-819.
627 44. Keller-Wood ME, Dallman MF. Corticosteroid Inhibition of ACTH Secretion*. Endocr Rev.
628 1984;5(1):1-24.
629 45. Kingwell BA, Thompson JM, Kaye DM, McPherson G, Jennings GL, Esler MD. Heart rate
630 spectral analysis, cardiac norepinephrine spillover, and muscle sympathetic nerve activity
631 during human sympathetic nervous activation and failure. Circulation. 1994;90(1):234-
632 240.http://dx.doi.org/10.1161/01.CIR.90.1.234.
633 46. Kirschbaum C, Hellhammer DH. Salivary cortisol in psychoneuroendocrine research: recent
634 developments and applications. Psychoneuroendocrinology. 1994;19(4):313-

635 333.http://dx.doi.org/10.1016/0306-4530(94)90013-2.
636 47. Kirschbaum C, Kudielka BM, Gaab J, Schommer NC, Hellhammer DH. Impact of gender,
637 menstrual cycle phase, and oral contraceptives on the activity of the hypothalamus-
638 pituitary-adrenal axis. Psychosom Med. 1999;61(2):154-162.
639 48. Kivlighan KT, Granger DA. Salivary -amylase response to competition: Relation to gender,
640 previous experience, and attitudes. Psychoneuroendocrinology. 2006;31(6):703-
642 49. Kleiger RE, Stein PK, Bigger JT. Heart rate variability: measurement and clinical utility. Ann
643 Noninvasive Electrocardiol. 2005;10(1):88-101.http://dx.doi.org/10.1111/j.1542-
644 474X.2005.10101.x.
645 50. Kreher JB, Schwartz JB. Overtraining syndrome a practical guide. Sports Health: A
646 Multidisciplinary Approach. 2012;4(2):128-
647 138.http://dx.doi.org/10.1177/1941738111434406.
648 51. Lacroix S, Gayda M, Gremeaux V, Juneau M, Tardif J-C, Nigam A. Reproducibility of near-
649 infrared spectroscopy parameters measured during brachial artery occlusion and reactive
650 hyperemia in healthy men. J Biomed Opt. 2012;17(7):0770101-
651 0770105.http://dx.doi.org/10.1117/1.JBO.17.7.077010.
652 52. Lennartsson A-K, Kushnir MM, Bergquist J, Billig H, Jonsdottir IH. Sex steroid levels
653 temporarily increase in response to acute psychosocial stress in healthy men and women. Int
654 J Psychophysiol. 2012;84(3):246-253.http://dx.doi.org/10.1016/j.ijpsycho.2012.03.001.
655 53. Levitt NS, Lambert EV, Woods D, Hales CN, Andrew R, Seckl JR. Impaired glucose tolerance
656 and elevated blood pressure in low birth weight, nonobese, young South African adults:
657 Early programming of cortisol axis 1. J Clin Endocrinol Metab. 2000;85(12):4611-
658 4618.http://dx.doi.org/10.1210/jcem.85.12.7039.
25
659 54. Lien D, Balakrishnan N. On regression analysis with data cleaning via trimming,
660 winsorization, and dichotomization. Communications in StatisticsSimulation and
661 Computation. 2005;34(4):839-849.http://dx.doi.org/10.1080/03610910500307695.
662 55. Malliani A, Pagani M, Montano N, Mela GS. Sympathovagal balance: a reappraisal.
663 Circulation. 1998;98(23):2640a-2643.http://dx.doi.org/10.1161/01.CIR.98.23.2640.a.
664 56. Marshall JM. The influence of the sympathetic nervous system on individual vessels of the
665 microcirculation of skeletal muscle of the rat. J Physiol. 1982;332:169.
666 57. Martelli D, Silvani A, McAllen RM, May CN, Ramchandra R. The low frequency power of heart
667 rate variability is neither a measure of cardiac sympathetic tone nor of baroreflex sensitivity.
668 Am J Physiol Heart Circ Physiol. 2014;307(7):H1005-
669 H1012.http://dx.doi.org/10.1152/ajpheart.00361.2014.
670 58. McMaster A, Jangani M, Sommer P, et al. Ultradian cortisol pulsatility encodes a distinct,
671 biologically important signal. PLoS One.
672 2011;6(1):e15766.http://dx.doi.org/10.1371/journal.pone.0015766.
673 59. Melzack R, Coderre TJ, Katz J, Vaccarino AL. Central neuroplasticity and pathological pain.
674 Ann N Y Acad Sci. 2001;933(1):157-174.http://dx.doi.org/10.1111/j.1749-
675 6632.2001.tb05822.x.
676 60. Metelka R. Heart rate variability-current diagnosis of the cardiac autonomic neuropathy. A
677 review. Biomedical Papers. 2014;158(3):327-338.
678 61. Moak JP, Goldstein DS, Eldadah BA, et al. Supine low-frequency power of heart rate
679 variability reflects baroreflex function, not cardiac sympathetic innervation. Heart Rhythm.
680 2007;4(12):1523-1529.http://dx.doi.org/10.1016/j.hrthm.2007.07.019.
681 62. Muhtz C, RodriguezRaecke R, Hinkelmann K, et al. Cortisol response to experimental pain in
682 patients with chronic low back pain and patients with major depression. Pain Med.
683 2013;14(4):498-503.http://dx.doi.org/10.1111/j.1526-4637.2012.01514.x.
684 63. Oliva F, Piccirilli E, Berardi AC, Frizziero A, Tarantino U, Maffulli N. Hormones and
685 tendinopathies: the current evidence. British medical bulletin. 2016;117(1):39-58.
686 64. Padayachy K, Vawda GHM, Shaik J, McCarthy PW. The immediate effect of low back
687 manipulation on serum cortisol levels in adult males with mechanical low back pain. Clinical
688 Chiropractic. 2010;13(4):246-252.http://dx.doi.org/10.1016/j.clch.2010.05.002.
689 65. Pagani M, Lucini D, Porta A. Sympathovagal balance from heart rate variability: time for a
690 second round? Exp Physiol. 2012;97(10):1141-
691 1142.http://dx.doi.org/10.1113/expphysiol.2012.066977.
692 66. Plaza-Manzano G, Molina F, Lomas-Vega R, Martnez-Amat A, Achalandabaso A, Hita-
693 Contreras F. Changes in biochemical markers of pain perception and stress response after
694 spinal manipulation. J Orthop Sports Phys Ther. 2014;44(4):231-
695 239.http://dx.doi.org/10.2519/jospt.2014.4996.
696 67. Portney LG, Watkins MP. Foundations of clinical research: applications to practice. Vol 2:
697 Prentice Hall Upper Saddle River, NJ; 2000.
698 68. Rohleder N, Nater UM, Wolf JM, Ehlert U, Kirschbaum C. Psychosocial stressinduced
699 activation of salivary alphaamylase: an indicator of sympathetic activity? Ann N Y Acad Sci.
700 2004;1032(1):258-263.http://dx.doi.org/10.1196/annals.1314.033.
701 69. Saghaei M. Random allocation software for parallel group randomized trials. BMC Med Res
702 Methodol. 2004;4(1):26.http://dx.doi.org/10.1186/1471-2288-4-26.
703 70. Sampath KK, Mani R, Cotter JD, Tumilty S. Measureable changes in the neuro-endocrinal
704 mechanism following spinal manipulation. Med Hypotheses. 2015;85(6):819-
705 824.http://dx.doi.org/10.1016/j.mehy.2015.10.003.
706 71. Sandercock G, Pinna GD, Maestri R, et al. Heart rate variability measures: a fresh look at
707 reliability. Clinical Science. 2007;113(3-4).
708 72. Sassi R, Cerutti S, Lombardi F, et al. Advances in heart rate variability signal analysis: joint
709 position statement by the e-Cardiology ESC Working Group and the European Heart Rhythm
26
710 Association co-endorsed by the Asia Pacific Heart Rhythm Society. Europace.
711 2015;17(9):1341-1353.http://dx.doi.org/10.1093/europace/euv015.
712 73. Shaffer F, McCraty R, Zerr CL. A healthy heart is not a metronome: an integrative review of
713 the heart's anatomy and heart rate variability. Front Psychol.
714 2014;5:1040.http://dx.doi.org/10.3389/fpsyg.2014.01040.
715 74. Shahidi B, Sannes T, Laudenslager M, Maluf KS. Cardiovascular responses to an acute
716 psychological stressor are associated with the cortisol awakening response in individuals
717 with chronic neck pain. Physiol Behav. 2015;150:93-
718 98.http://dx.doi.org/10.1016/j.physbeh.2015.02.010.
719 75. Shirtcliff EA, Granger DA, Schwartz E, Curran MJ. Use of salivary biomarkers in biobehavioral
720 research: cotton-based sample collection methods can interfere with salivary immunoassay
721 results. Psychoneuroendocrinology. 2001;26(2):165-173.http://dx.doi.org/10.1016/S0306-
722 4530(00)00042-1.
723 76. Sillevis R, Cleland J, Hellman M, Beekhuizen K. Immediate effects of a thoracic spine thrust
724 manipulation on the autonomic nervous system: a randomized clinical trial. J Man Manip
725 Ther. 2010;18(4):181-190.http://dx.doi.org/10.1179/106698110X12804993427126.
726 77. Smith GD, Ben-Shlomo Y, Beswick A, Yarnell J, Lightman S, Elwood P. Cortisol, testosterone,
727 and coronary heart disease prospective evidence from the Caerphilly Study. Circulation.
728 2005;112(3):332-340.http://dx.doi.org/10.1161/CIRCULATIONAHA.104.489088.
729 78. Smith KJ, Rosenberg DL, Timothy Haight G. An assessment of the psychometric properties of
730 the Perceived Stress Scale10 (PSS10) with business and accounting students. Accounting
731 Perspectives. 2014;13(1):29-59.http://dx.doi.org/10.1111/1911-3838.12023.
732 79. Southern WM, Ryan TE, Reynolds MA, McCully K. Reproducibility of near-infrared
733 spectroscopy measurements of oxidative function and postexercise recovery kinetics in the
734 medial gastrocnemius muscle. Appl Physiol Nutr Metab. 2013;39(5):521-529.

735 80. Stalder T, Kirschbaum C, Kudielka BM, et al. Assessment of the cortisol awakening response:
736 expert consensus guidelines. Psychoneuroendocrinology. 2016;63:414-
738 81. Sterling M, Jull G, Wright A. Cervical mobilisation: concurrent effects on pain, sympathetic
739 nervous system activity and motor activity. Man Ther. 2001;6(2):72-
740 81.http://dx.doi.org/10.1054/math.2000.0378.
741 82. Task-Force. Heart rate variability standards of measurement, physiological interpretation,
742 and clinical use. Eur Heart J. 1996;17:354-381.

743 83. Thomas GD, Segal SS. Neural control of muscle blood flow during exercise. J Appl Physiol.
744 2004;97(2):731-738.http://dx.doi.org/10.1152/japplphysiol.00076.2004.
745 84. Touitou Y, Haus E. Alterations with aging of the endocrine and neuroendocrine circadian
746 system in humans. Chronobiol Int. 2000;17(3):369-390.http://dx.doi.org/10.1081/CBI-
747 100101052.
748 85. Vicenzino B, Cartwright T, Collins D, Wright A. An investigation of stress and pain perception
749 during manual therapy in asymptomatic subjects. Eur J Pain. 1999;3(1):13-18.
750 86. Vicenzino B, Collins D, Benson H, Wright A. An investigation of the interrelationship between
751 manipulative therapy-induced hypoalgesia and sympathoexcitation. J Manipulative Physiol
752 Ther. 1998;21(7):448-453.
753 87. Ward J, Coats J, Tyer K, Weigand S, Williams G. Immediate Effects of Anterior Upper Thoracic
754 Spine Manipulation on Cardiovascular Response. J Manipulative Physiol Ther.
755 2013;36(2):101-110.http://dx.doi.org/10.1016/j.jmpt.2013.01.003.
756 88. Weir JP. Quantifying test-retest reliability using the intraclass correlation coefficient and the
757 SEM. J Strength Cond Res. 2005;19(1):231-240.
758 89. Welch A, Boone R. Sympathetic and parasympathetic responses to specific diversified
759 adjustments to chiropractic vertebral subluxations of the cervical and thoracic spine. J
760 Chiropr Med. 2008;7(3):86-93.http://dx.doi.org/10.1016/j.jcm.2008.04.001.
27
761 90. Whelan TL, Dishman JD, Burke J, Levine S, Sciotti V. The effect of chiropractic manipulation
762 on salivary cortisol levels. J Manipulative Physiol Ther. 2002;25(3):149-
763 153.http://dx.doi.org/10.1067/mmt.2002.122328.
764 91. Whitworth JA, Williamson PM, Mangos G, Kelly JJ. Cardiovascular consequences of cortisol
765 excess. Vascular Health and Risk Management. 2005;1(4):291.
766 92. Yates F, Leeman S, Glenister D, Dallman M. Interaction between plasma corticosterone
767 concentration and adrenocorticotopin releasing stimuli in the rat: evidence for the reset of
768 endocrine feedback control 1 2. Endocrinology. 1961;69(1):67-
769 80.http://dx.doi.org/10.1210/endo-69-1-67.
770 93. Zegarra-Parodi R, Park PYS, Heath DM, Makin IRS, Degenhardt BF, Roustit M. Assessment of
771 skin blood flow following spinal manual therapy: A systematic review. Man Ther.
772 2015;20(2):228-249.http://dx.doi.org/10.1016/j.math.2014.08.011.
773
28
774 Figure Legends
775 1) FIGURE 1. Illustration of SM technique (large photo) with hand position (top
776 insert) and NIRS optode placement (bottom insert)

777
778
779 2) FIGURE 2. Study design and data collection
780 Abbreviations: B, 5 minutes pre-intervention; A1, 1 minute post-intervention; P1,
781 5 minutes post-intervention; P2, 30 minutes post-intervention; P3, 6 hours post-
782 intervention; HRV, Heart rate variability; OHB, Oxy-haemoglobin.*Salivary

783 samples were also collected at morning similar to day-1.
784
785
786
29
793
792
791
790
789
788
787
3) FIGURE 3. Flow of participants
30
4) FIGURE 4. Changes in T/C ratio between groups following intervention
TABLE 1. Baseline characteristics of participants
Characteristics Sham (n = 12) Spinal Manipulation (n = 12)

Mean SD Mean SD
Age 27.336.8 28.335.8
Weight 78.447.73 74.7915.2
Height 1.790.06 1.720.07
PSS-10 15.06.20 13.924.2

T/C ratio 0.360.03 0.290.03
Cortisol -0.9870.25 -0.7830.24
Testosterone 14968.6 11545.4
HRV 1.320.33 1.250.29
Abbreviations: HRV, Heart rate variability; PSS, Perceived stress scale; T/C,
Testosterone/Cortisol.
* p<0.05
Downloaded from www.jospt.org at University of Hong Kong Libraries on July 14, 2017. For personal use only. No other uses without permiss
TABLE 2. Hormonal changes after intervention
Outcome Groups (mean and SD) Interaction Difference between groups: SM - sham
(MD and CI; p-value*)
B P1 P2 P3 P1 P2 P3

SM Sham SM Sham SM Sham SM Sham p-value Effect
(n = 12) (n = 12) (n = 12) (n = 12) (n = 12) (n = 12) (n = 12) (n = 12) Size
Cortisol -0.78(.24) -0.99(.25) -0.99(.27) -0.64(.29) -0.93(.29) -0.73 (.32) -1.07(.26) -1.31(.42) <0.01 0.28 -0.35(-.59 to -12 -0.20(-.46to 0.23(-.06to0.53;0.11)
;0.005) .05;0.11)
115(45.4) 149(68.6) 151(58.7) 141(67.6) 157(67.4) 164 (59.3) 81(25.6) 94(51.8) 0.33 0.05 10.08(-43.5to63.8; -6.29(-60 -12.87(-47.5to
Testosterone
0.70) to47.5;0.8 21.8;0.45)
1)
T/C ratio .29(.036) .36(.036) .34(.032) .30(.032) .35(.033) .35(.033) .30(.026) .39(.026) <0.05 0.11 0.04(-0.05to0.13; 0.00(- 0.09(-0.16to-
0.38) 0.09to0.0 0.09;0.02)
9; 1.0)
Abbreviations: CI, Confidence Interval; MD, Mean Difference; B, Pre-intervention samples; P1, Post-intervention samples (5 minutes); P2,
Post-intervention samples (30 minutes); P3, Post-intervention samples (6 hours); SD, Standard Deviation; SM, Spinal Manipulation; ,
group by time; *, between groups; , log-transformed values.

Downloaded from www.jospt.org at University of Hong Kong Libraries on July 14, 2017. For personal use only. No other uses without permiss
TABLE 3. Changes in OHB and HRV after intervention
Outcome Groups (mean and SD) Interaction Difference between groups: SM - sham (MD and CI; p-
value*)
A1 P1 P2 A1 P1 P2
SM Sham SM Sham SM Sham p- Effect

(n = 12) (n = 12) (n = 12) (n = 12) (n = 12) (n = 12) value Size
OHB -12.7(25.6) -2.54(19.5) 4.9(20.8) 4.1(13.2) 46.8(59) -12.7(25.6) 0.32 0.05 -10.2 (29.5to9;0.28) 0.89(13.9to15.6;0.9) 15.2(-30.5to60.9;0.49)
B P1 P2 P1 P2
SM Sham SM Sham SM Sham

(n = 12) (n = 12) (n = 12) (n = 12) (n = 12) (n = 12)
HRV 1.250.33 1.320.29 1.230.52 1.340.31 1.350.36 1.340.3 0.73 0.13 -0.11(-0.47to0.25;0.54) 0.01(-0.26to0.28;0.95)
Abbreviations: CI, Confidence Interval; HRV, Heart rate variability; OHB, Oxy-Haemoglobin; SD, Standard Deviation; SM, Spinal
Manipulation; B, Pre-intervention samples; A1, Post-intervention (O2Hb) samples (1 minute); P1, Post-intervention samples (5 minutes); P2,
Post-intervention samples (30 minutes); , group by time; *, between groups.

10 2519jospt 2017 7348

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

10 2519jospt 2017 7348

Uploaded by

Copyright:

Available Formats

Neuro-endocrine Response Following a Thoracic Spinal Manipulation in

Steve Tumilty, PT, PhD1.

Centre for Health, Activity, and Rehabilitation Research, School of Physiotherapy,

School of Physical Education, Sport and Exercise Sciences, University of Otago,

Department of Physiology-Heart Otago, University of Otago, Dunedin, New

Address correspondence to Kesava Kovanur Sampath, Centre for Health, Activity,

3 Study design: Randomized controlled trial.

4 Background: Spinal manipulation (SM) can trigger a cascade of responses

6 endocrine system, specifically the hypothalamic-pituitary (HP) axis. However, no

7 manual therapy study has investigated the neuroendocrine response to SM (i.e.,

8 SNS-HP axis) in the same trial.

9 Objective: To determine short-term changes in SNS activity; heart rate variability

10 (HRV) and endocrine activity (cortisol, testosterone and testosterone/cortisol (T/C)

11 ratio) following a thoracic SM.

17 approximately 6 hours after intervention.

19 (p<0.05) and salivary cortisol (p<0.01) concentrations. Significant between group

23 differentially alter O2Hb, testosterone or HRV relative to responses in the sham

26 concentration and reduced T/C ratio 6 hours after intervention. A pattern of

27 immediate sympathetic excitation was also observed in the SM group.

28 Key Words: autonomic nervous system, cortisol, spinal manipulation, sympathetic

29 nervous system, testosterone.

35 been proposed including changes in the autonomic nervous system (ANS),

36 sympathetic nervous system (SNS) and the endocrine system (hypothalamic-

37 pituitary-adrenal (HPA) axis)24.

38 Several studies have been undertaken to investigate the effects of SM on the

39 SNS8,9,76,87,89. However, there is no gold standard of non-invasive measurement of

40 changes in activity93. One reliable, non-invasive technique to measure SNS activity is

43 Different wavelengths of this light are absorbed by either oxygenated (O2Hb) or

46 measure blood flow (vasoconstriction/dilation) of aterioles in skeletal muscles. As the

48 OHB changes in skeletal muscles as measured by NIRS may provide a reliable

49 estimate of SNS activity.

50 Most tissues innervated by the SNS also receive innervation by the

53 (HRV), the beat-to-beat variation in heart rate is a well-established method of

54 assessing autonomic function82. Analysis of HRV in the frequency domain reveals

59 activity55,60. Therefore the LF/HF ratio is interpreted as an indicator of cardiac

63 assess autonomic cardiovascular regulation49,60,65,72. Previous investigations into the

64 effects of thoracic SM on HRV have produced conflicting results9,87. While Budgell

67 Cortisol is released by the activation of HPA axis14. The use of salivary

70 known circadian rhythm that could be influenced by a number of confounding factors

74 intense physical activity before data collection15,80. Therefore, adequate control of

76 intervention. It is unclear how these confounding factors were addressed in previous

78 difficult to interpret the findings.

79 It has been proposed that models need to include multiple measurements of

80 stress-related biological processes in order to understand the role of an individuals

81 response to stress48. For example, the hypothalamic-pituitary-gonadal (HPG) axis

84 may therefore provide a better estimation of hypothalamic-pituitary (HP) axis

86 athletes41,50 and susceptibility to certain diseases31,77.

87 When an individual is presented with a stressor, such as a painful injury, the

88 hypothalamus coordinates the stress response by activation of the HP axis and/or a

89 fast-acting neural component involving the SNS15. Therefore the hypothalamus

90 maintains homeostasis by the activation of the neuroendocrine system (SNS-HP

92 activation81,86. Given their integrated function, it could be hypothesised that SNS

93 changes following SM would be accompanied by changes in the HP axis 70. To our

94 knowledge, no manual therapy study has investigated the neuroendocrine response

97 changes in HPA axis greater than 2 hours following SM is still unknown.

98 We hypothesised that a thoracic SM would result in neuroendocrine response

99 (SNS-HP axis). Activation of SNS will be demonstrated by an immediate

101 By measuring salivary cortisol and testosterone, it will be possible to understand

108 argued previously4, identification of SM mechanisms may increase the acceptance of

112 testosterone and the T/C ratio following a thoracic SM.

113 MATERIALS AND METHODS

114 Study Design

118 the study.

220 *INSERT FIGURE 2 HERE*****************

272 *INSERT FIGURE 3 HERE

277 *INSERT TABLE 1 HERE

INSERT FIGURE 4 HERE**********

305 INSERT TABLE 2 HERE**

318 INSERT TABLE 3 HERE**