Assessment of skin blood flow following spinal manual therapy: a systematic review.

Manual Therapy 20 (2015) 228e249

Contents lists available at ScienceDirect

Manual Therapy journal homepage: www.elsevier.com/math

Systematic review

Assessment of skin blood flow following spinal manual therapy: A systematic review* Rafael Zegarra-Parodi a, *, Peter Yong Soo Park b, Deborah M. Heath c, Inder Raj S. Makin c, Brian F. Degenhardt a, Matthieu Roustit d, e a

A.T. Still Research Institute, A.T. Still University, Kirksville, MO, USA A.T. Still University, Kirksville College of Osteopathic Medicine, Kirksville, MO, USA A.T. Still University, School of Osteopathic Medicine, Mesa, AZ, USA d Clinical Pharmacology Unit, Inserm CIC1406, Grenoble University Hospital, Grenoble, France e UMR 1042 e HP2, Inserm and University of Grenoble-Alpes, Grenoble, France b c

a r t i c l e i n f o

a b s t r a c t

Article history: Received 11 June 2014 Received in revised form 22 August 2014 Accepted 28 August 2014

Skin blood flow (SBF) indexes have been used to describe physiological mechanisms associated with spinal manual therapy (SMT). The aims of the current review were to assess methods for data collection, assess how investigators interpreted SBF changes, and formulate recommendations to advance manual medicine research. A database search was performed in PubMed, Cochrane Library, the Physiotherapy Evidence Database, and the Cumulative Index to Nursing and Allied Health Literature through April 2014. Articles were included if at least 1 outcome measure was changes in 1 SBF index following SMT. The database search yielded 344 records. Two independent authors applied the inclusion criteria. Twenty studies met the inclusion criteria. Selected studies used heterogeneous methods to assess short-term post-SMT changes in SBF, usually vasoconstriction, which was interpreted as a general sympathoexcitatory effect through central mechanisms. However, this conclusion might be challenged by the current understanding of skin sympathetic nervous activity over local endothelial mechanisms that are specifically controlling SBF. Evaluation of SBF measurements in peripheral tissues following SMT may document physiological responses that are beyond peripheral sympathetic function. Based on the current use of SBF indexes in clinical and physiological research, 14 recommendations for advancing manual medicine research using laser Doppler flowmetry are presented. © 2014 Elsevier Ltd. All rights reserved.

Keywords: Spinal manual therapy Skin blood flow index Skin microcirculation Systematic review

1. Introduction Spinal manual therapy (SMT) is a therapeutic procedure commonly provided by clinicians from several healthcare disciplines for the management of musculoskeletal and nonmusculoskeletal conditions (Bronfort et al., 2010). Several underlying physiological mechanisms have been proposed for clinical outcomes associated with SMT, the most common being a possible influence on segmental and suprasegmental reflexes with a prominent role given to the peripheral sympathetic nervous system (PSNS) (Beal, 1985; Van Buskirk, 1990; Pickar, 2002;

* Sources of support: Mr. Park was funded through an internal A.T. Still University grant (#501-415). * Corresponding author. A.T. Still Research Institute, A.T. Still University, 800 West Jefferson Street, Kirksville, MO 63501, USA. Tel.: þ1 660 626 2267; fax: þ1 660 626 2099. E-mail address: [email protected] (R. Zegarra-Parodi).

http://dx.doi.org/10.1016/j.math.2014.08.011 1356-689X/© 2014 Elsevier Ltd. All rights reserved.

Bialosky et al., 2009). This neurophysiological hypothesis relies mainly on the interpretation of significant changes in skin blood flow (SBF) indexes following SMT (Chu et al., 2014; Kingston et al., 2014). Denslow, Korr, and Wright developed the use of SBF indexes in manual therapy (Wright, 1956) and considered these outcome measurements as potential markers for PSNS function and evaluation of vasomotor reactivity. SBF changes are usually assessed through skin temperature (ST), skin conductance (SC), pulse plethysmography (PPG), and laser Doppler flowmetry (LDF) (Bolton and Budgell, 2012). These tools are valuable for manual medicine researchers because of their noninvasive nature and ease of use in clinical research. Further, outcomes may reflect changes in the skin and in deeper tissues, but these conclusions are now considered overly simplistic with regard to the complex regulation of the skin microcirculation (Bolton and Budgell, 2012). In thermoneutral environments, many local mediators are involved in skin microvascular reactivity (Roustit and Cracowski, 2013) with a low and transient involvement of

R. Zegarra-Parodi et al. / Manual Therapy 20 (2015) 228e249

sympathetic vasoconstrictor response at rest (Krupatkin, 2008; Charkoudian, 2010). These specificities of skin sympathetic nervous activity (SSNA) and their influence on endothelial pathways regulating SBF have not been explicitly described in previous manual therapy publications (Bolton and Budgell, 2012; Chu et al., 2014; Kingston et al., 2014). Further, previous publications have used heterogeneous methods to assess SBF (Bolton and Budgell, 2012). All these physiological and methodological issues may challenge previous interpretations of SBF changes associated with SMT. Local human cutaneous circulation has been proposed as a marker of systemic microvascular function. It has been studied in many diseases and with therapeutic interventions assessing microvascular function (Roustit and Cracowski, 2013). Recent advances include simple and noninvasive methods using LDF to assess skin microvascular function when coupled with reactivity tests. Although these methods provide an overall assessment of microvascular function, some are considered more specific for assessing different physiological pathways. Using LDF equipment, reactivity tests evaluating SSNA influence on SBF induce a vasoconstriction (Wilder-Smith et al., 2005), whereas tests evaluating microvascular endothelial function induce vasodilation with limited involvement of SSNA (Cracowski et al., 2006). However, these methods commonly used in physiological research have not been used in manual medicine research (Bolton and Budgell, 2012). Further, SBF combined with other outcome measures evaluating symptoms (e.g., pain scales) and musculoskeletal activity (e.g.,

229

electromyography) associated with SMT may provide useful information for clinicians (Sterling et al., 2001; Chu et al., 2014). The objective of the current systematic review was to critically review the different methods used to evaluate changes in SBF indexes following SMT. We focus on how SBF measurements have been performed and interpreted, and discuss which methodological approaches could be incorporated to advance manual medicine research. 2. Methods 2.1. Search strategy Guidelines from the PRISMA statement (Moher et al., 2009) were followed in the current review. A search for articles was performed without date and language limitations in PubMed, Cochrane Library, the Physiotherapy Evidence Database (PEDro), and the Cumulative Index to Nursing and Allied Health Literature (CINAHL) through April 2014 using the following search terms: “spinal manipulation” and “spinal mobilization” for SMT in combination with “blood”, “cardiovascular”, “conductance”, “microcirculation”, “sympathetic”, and “temperature” for SBF indexes (Appendix). “Sympathetic”, “conductance”, and “temperature” were used in our initial search because ST and SC were originally used as markers of PSNS activity and the SBF wording was not always in the titles of those studies. The search was limited to human studies. Articles were included in the current review if at least 1

Fig. 1. PRISMA flowchart of the literature search for the current review.

Article

Setting

Eshleman et al. Unknown (1971)

Design

Demographics

Nonrandomized 59 healthy, young participants parallel cohort study with 1 control group Group 1 (30)

Group 2 (15)

230

Table 1 Overview of studies included in the current systematic review. Stimulation

Control

SBF index

Other outcome measure

Myofascial stretching to lower cervical and upper thoracic regions

Control (no manipulation)

None Electrical impedance plethysmography (blood volume)

Spinal muscle mass derived from the same myotome

Supine position

Effect on SBF

Statistics

Clinical significance

17.06% decrease at end of 5 min stimulation period (Group 1)

Standard errors

Relationship between changes in PSNS activity and stimulation

17.61% decrease after 1 min of 5 min stimulation period (Group 2) Small effect during rest (Group 3, control)

No comparison between 3 groups

No consistent effect of manipulation on ST when the spine is considered as a whole Manipulations in the sympathetic regions of the spine (T1-L3) caused a 0.25 F significant decrease in ST (P < .001)

Student's paired t-test to see if change in ST before and after manipulation different than zero Chi-square analysis to see if the frequency of positive and negative responses different from expected by chance

Group 3 (14) Harris and Wagnon (1987)

Student clinic at a Noncontrolled cohort study chiropractic college

None

ST

None

Manipulations in the non-sympathetic regions of the spine (C1eC7, L4, L5) caused a significant 0.42 F increase in ST (P < .001) Ellestad et al. (1988)

Osteopathic college

26 men and 14 Randomized, controlled study women with 2 groups: low-back pain and pain-free

22e36 years

HVLAT to the entire axial skeleton and pelvis

OMT to any of the following dysfunctions diagnosed: (1) muscle energy to long restrictors of the hip, (2) muscle energy to the pelvis, (3) sacroiliac mobilization and ischial tuberosity spread, (4) HVLAT to L5/S1, (5) HVLAT to T12/L1, (6) HVLAT to the thoracic spine and ribs (7) HVLAT to the thoracic inlet (C7/T1/R1), and (8) HVLAT for occipital/ C1, C1/C2, and C3eC7

Only medication, structural examination (palpation of spine, motor testing) but no OMT

Skin resistance (SC)

EMG

Unpaired t-test 77.67 (41.21) decrease in SC following OMT in the LBP group compared with control (P < .001)

Lumbar lordosis

Nonsignificant changes in the non-LBP group following OMT

2-way ANOVA

Small and significant changes probably have no physiological significance Influence of PSNS on ST evaluated since changes measured over a 10 s time interval following spinal manipulation and no humoral factors could have been responsible Spinal manipulations can cause measurable changes in the physiology of distant tissues

Effectiveness of HVLAT OMT for symptomatic and asymptomatic participants as measured by absolute EMG potentials and SC in the lumbar area Decrease in SC associated with a variation in PSNS nerves leading to a decreased peripheral vascular resistance resulting in better blood flow in the skin


Cervical, thoracic, and 196 patients lumbar HVLAT attending a student clinic at a chiropractic college Participants' position not described

20 LBP participants (>2 weeks but 3 months Symptoms primarily originating from the C5/C6 segment

ANOVA with post hoc analyses (Newman eKeuls test)

Forearm SBF not affected by a single HVLAT following 20 min induced cutaneous inflammatory reactions (continued on next page)

233

Article

Moulson and Watson (2006)

Laboratory

Laboratory

Jowsey and Laboratory Perry (2010)

Design

Demographics

Stimulation

Order of 2 procedures randomized then cross-over

14 healthy men

Participants in various postures

21e37 years; mean age, 27 years

HVLAT applied to areas of altered spinal function

11 healthy women

Mulligan's SNAGs to the C5/C6 intervertebral joint while participant simultaneously turned head to the right

Placebo (touch over the same spinal area without practitioner pressure but with active participant movement)

SC

5 healthy men

Repeated 3 times

Control (no touch, no movement)

ST

Mean (SD) age, 23.06 (5.35) years

Seated position

Double-blind, independent (matched) group, betweensubjects study

45 healthy, physiotherapeutically naïve, asymptomatic, non-smoking males

Unilaterally applied grade III oscillatory mobilization, at a rate of 2 Hz, to the left L4/L5 facet joint

Placebo (touch over the same spinal area without application of any oscillatory movement)

SC

3 groups

18e25 years; mean (SD) again, 21.5 (1.85) years

Prone position


Double-blind, randomized, placebocontrolled, independent group study

23 healthy, asymptomatic women

Grade III rotatory posteroanterior intervertebral mobilization, at a rate of 0.5 Hz, to the T4 vertebral segment

Placebo (touch over the same spinal area without application of any oscillatory movement)

Randomized, single-blind, within-subject, repeatedmeasures study

Control

SBF index


Effect on SBF

Statistics

Hypoalgesic effects of a single HVLAT due to central rather than peripheral mechanisms

Measurement No significant effects of the HVLAT or the placebo of spontaneous (P ¼ .28) pain with VAS

SC

None


No statistical differences 2-way ANOVA with repeated measures with for ST and SC between post hoc Bonferroni left and right for treatment, placebo, or control conditions at any phase of the intervention

SNAGs may contribute to manipulationinduced analgesia via a centrally mediated phenomenon rather than a local mechanism although this needs to be considered in light of the asymptomatic population tested

Significant increase in SC during the SNAGs (P < .0005) and for a 2min period after the SNAGs (P ¼ .001) compared with control Significant increase in SC for the placebo condition during SNAGs (P ¼ .015) and after SNAGs (P ¼ .011) compared with control Non-significant decrease in ST for the SNAGs and placebo compared with control None

None

13.5% (20.25) increase (P ¼ .002) specific to the side treated for the treatment group during the stimulation compared with placebo and control conditions

Side-specific increase greater than placebo (P ¼ .034) in the right hand

Multivariate analysis using a mixed betweenwithin subject design involving analysis of 1 between-subject factor (3 levels: control, placebo, and treatment groups), and 2 withinsubject factors: time (from baseline to intervention and from baseline to final rest period) and leg (right or left) Tukey post hoc analysis

1-way ANOVA

Unilaterally applied mobilization resulted in homolateral significant sidespecific changes in lower limb PSNS activity during the intervention period

This response was greater than those of the contralateral limb and of both the placebo and control conditions Mobilization of T4 has a sympathoexcitatory effect in the upper limbs


Perry and Green (2008)

Setting

234

Table 1 (continued )

13 healthy, asymptomatic men

Prone position

Spinal mobilization is an appropriate treatment choice to aim to improve higher center-mediated modulation of pain Supports a theoretical link between PSNS and T4 syndrome

Right hand: 5.74% increase in the treatment group compared with placebo during the intervention period Right hand: 16.85% increase in the treatment group compared with placebo during the postintervention period

18e35 years; mean (SD) age, 22.7 (5.2) years

Naïve to spinal manipulative therapy Roy et al. (2010)

Unknown

Randomized, parallel cohort study with 2 groups: treatment and sham

12 women with acute LBP

HVLAT, traditional lumbar roll, with a pisiform contact on the ipsilateral mamillary of L5

Treatment group (n ¼ 10); mean (SD) age, 35.7 (11.73) years

Sham group (n ¼ 10); mean (SD) age, 44.7 (9.8) years

ST

None

Significant changes in ST in the treatment group on the treatment side compared with the nontreatment side

Factorial groups sides time ANOVA model with repeated measures

Treatment group: Post hoc Tukey analysis treated side cooler by 0.46 F immediately after the manipulation and later warmed by 0.49 F 10 min after the manipulation compared with the control period; contralateral side cooled down for the entire recording period and was 0.17 F cooler 10 min after manipulation compared with the control period Sham group: similar changes between treated and nontreated sides (initial rise followed by a drop and rise in ST) Increase of ST in the treatment and sham groups immediately after the spinal manipulation Opposite trend on the initial ST measurements compared with asymptomatic participants

Multiphasic warming induced by spinal manipulation could be regulated by a neurologic supraspinal control, a physiologic cellular reaction from the cutaneous or deep tissue blood vessels, or the immunologic systems Valuable information for clinicians about tissue response after spinal manipulation (initial reactive circulation and tendency of the manipulated area to continue warming)


8 men with acute LBP Prone position

5-s pressure with the clinician's hand without the thrust

(continued on next page)

235

Table 1 (continued ) Setting

Desmarais et al. Unknown (2011)

Demographics

Stimulation

Control

SBF index


Effect on SBF

Statistics


Cohort study

10 healthy men

T4 HVLAT with participants in a prone position

Session 1: 35 electrical stimuli

SC

Numerical rating scale to rate pain and pain-related anxiety induced by electrical stimulation and noxious heat Mean respiration frequency

Palmar SC: spinal manipulation decreased the amplitude of shockevoked SC, significantly different from the control session for conditioning versus baseline (P ¼ .041) and post-conditioning versus baseline (P ¼ .010)

2-way repeatedmeasures ANOVA followed by planned contrasts

Somatic stimulation of the thoracic spine may modulate specific sympathetic pathways, partly through segmental processes

None

74.6 mohm or 76.3% increase in SC from baseline for the manipulation technique (P ¼ .0005)

Each participant 7 healthy women underwent 4 experimental sessions

Perry et al. (2011)

University laboratory

Prospective, quasiexperimental, randomized, independent subjects study

Session 2: 35 electrical stimuli þ T4 spinal manipulation

23 participants recruited but 4 dropped out and 2 had lost data

Session 3: 35 electrical stimuli þ noxious heat (9 cm2 contact heat thermode applied on the midline of T3eT5 area)

Mean (SD) age, 25.0 (1.1) years

Session 4: 35 electrical stimuli þ spinal manipulation þ noxious heat

59 healthy, physiotherapeutically naïve, asymptomatic, nonsmoking participants

None Localized HVLAT grade V manipulation segmental rotation technique applied to the L4/L5 segment in either right or left side-lying (random allocation)

Spinal manipulation technique group (n ¼ 25); 11 male and 14 female; mean (SD) age, 36.9 (8.27) years

Localized central posteroanterior technique statically applied to the spinous process of the L4/L5 segment (over-pressure) while the participant actively performed 3 sets of 10 repetitions of a lumbar extension maneuver in prone lying with 1 min rest between the 3 sets

SC

35.5 mohm or 35.7% increase in SC from baseline for the extension exercises (P ¼ .0005)

Side-lying position

Outpatient rehabilitation clinic

84 patients with Prospective, cervical spondylotic randomized, radiculopathy parallel cohort study with 2 groups: cervical

None Manual steady traction performed with participant in the seated and supine positions with a pillow under the neck (3 different

Reliable for measurements of PSNS response with any SC measurements in excess of 0.3154 mohms (4.633%) Manipulation Independent t-test technique can cause performed between the 2 groups during the twice the magnitude intervention periods and of response when compared with the between the final rest extension exercises periods to test for any differences in magnitude and has an effect that continues into the and longevity of effect final rest period between the 2 techniques

Paired t-tests performed for each intervention between the baseline, experimental, and the final rest periods

Only the manipulation technique had a lasting effect that was carried out into the final rest period (12.9% increase from baseline, P ¼ .012) Manipulation technique had greater effect (P ¼ .012) No difference between the sides of the manipulation technique (P ¼ .76)

Extension exercises group (n ¼ 25); 10 male and 15 female; mean (SD) age, 37.7 (8.28) years

Jiang et al. (2012) [in Chinese]

Conditioning stimulations (noxious heat and spinal manipulation) appear to rely on different processes Modulation of SC by thoracic manipulation caused by modulation of spinal networks involving interneurons and preganglionic sympathetic fibers

Plantar SC: spinal manipulation did not significantly affect shockevoked SC

ST as recorded with infrared thermal imaging system

Mean (SD) VAS for Significant increase radiculopathy in ST between normal and abnormal limbs in the manual traction group (P < .01) and in the

SC response was not a side-specific phenomenon

Better efficacy with cervical manual fixedpoint traction manipulation than cervical computer


Design

236

Article

computer traction and traditional Chinese osteopathic manipulation 18e70 years

Manual treatment group (n ¼ 42)

sizes of pillow forming 3 different angles with the table for traction depending on pain location, i.e., 10 for upper neck, 20 for middle neck, and 30 for lower neck) Computer traction with cervical fixed-point traction performed only in seated position with 3 possibilities of a forward angle for traction depending on pain location, i.e., 10 for upper neck, 20 for middle neck, and 30 for lower neck Weight of traction started at 3e5 kg and was augmented within the participant's pain-free zone

traction in treating patients with cervical spondylotic radiculopathy

computer traction group (P < .05) before and after treatment

ST difference between normal and abnormal limbs more obvious in manual treatment group compared with computer traction group (P < .01)

Paired t-tests performed for each intervention between the pre- and post-treatment periods

10.60% (7.5) increase for the right limb between treatment and control (P ¼ .044)

Analysis of SC performed using the “integral measurement” for baseline, intervention, and final rest period Intervention and final rest period converted to a percentage of change from baseline

Computer traction group (n ¼ 42) Prospective, single-blind, randomized, parallel-group, 3-arm study 3 equally numbered groups

45 healthy, Mulligan's SNAGs to the physiotherapeutically L4 spinous process while naïve participants participant simultaneously performed 6 repetitions of full active lumbar flexion in sitting position Treatment group (n ¼ 15): 6 men and 9 women; 18e46 years; mean (SD) age, 25 (8) years Sham group (n ¼ 15): 6 men and 9 women; 19e43 years; mean (SD) age, 27 (8) years

Placebo (touch over the SC same spinal area without practitioner pressure but with active participant movement)

None


11.19% (7.85) increase for the left limb between treatment and control (P ¼ .004) No statistical difference between touch procedures and control

Randomized, double-blind, placebocontrolled study

Touch and movement component of the shame technique appeared to affect PSNS

1-way ANOVA with post hoc Bonferroni to compare the percentage change from baseline between groups

No statistical difference between left and right side for treatment, placebo, or control

Control group (n ¼ 15): 15 women; 18e45 years; mean (SD) age, 27 (10) years La Touche et al. Unknown (2013)

Stimulation demonstrated a sympathoexcitatory response in both lower limbs

32 patients with chronic craniofacial pain of myofascial origin

Anterior-posterior upper cervical mobilization, at a rate of 0.5 Hz, to the 3 upper cervical segments (occiput-C3)

Placebo (touch over the SC same spinal area without application of any oscillatory movement)

21 women and 11 men

3 treatment sessions

3 treatment sessions

ST

Self-reported variables: Beck Depression Inventory, State-Trait Anxiety Inventory, and Neck Disability Index Pain intensity: VAS

83.75% increase in SC (P < .001)

2 3 repeated measures ANOVA; factors analyzed were time (pre, post 1, and post 2) and group (treatment and sham)

Stimulation reduces pain intensity, increases pain pressure threshold in the cervical and craniofacial regions and causes sympathoexcitation

ST changes not significant (P ¼ .071) after application of the technique compared with the placebo

Bonferroni post hoc analysis

Indicate an influence of the technique on the central nervous system (medullar or supramedullar effect)


Moutzouri et al. University (2012) laboratory


237

1-way ANOVA to analyze the percent change in group factor and time factor between sessions Percent change of the total of the means of the 3 sessions in the treatment and placebo groups analyzed with a Student t-test

Statistics

Changes in PSNS: heart rate, breathing rate

Pressure pain threshold

Effect on SBF Other outcome measure SBF index Control

Mean (SD) age, 33.19 (9.49) years for the treatment group (n ¼ 16) Mean (SD) age, 34.56 (7.84) years for the sham group (n ¼ 16)

Stimulation Demographics Design Setting Article

Table 1 (continued )

Abbreviations: ANOVA, analysis of variance; AUC, area under the curve; BMI, body mass index; dPAG, dorsal periaqueductal gray matter; EMG, electromyography; HVLAT, high-velocity, low-amplitude thrust; LBP, low back pain; LDF, laser Doppler flowmetry; MAX, maximum; MIN, minimum; OMT, osteopathic manipulative treatment; PPG, pulse plethysmography; PSNS, peripheral sympathetic nervous system; SBF, skin blood flow; SC, skin conductance; SD, standard deviation; SNAG, sustained natural apophyseal glide; ST, skin temperature; VAS, visual analog scale.



238

outcome measure was changes in 1 index of SBF following manual SMT, whatever the study design. 2.2. Review process Selection of articles was made independently by 2 authors. Abstracts of identified papers were reviewed and full texts of the original publication were obtained when studies met our criteria. A final search of additional articles was performed on the references lists of all retrieved articles (Fig. 1). Each article included was independently reviewed by the same 2 authors, and relevant information was extracted from all sections. At the end of each step, disagreements were resolved by 2 additional authors who independently reviewed specific articles. Consensus was then reached following a meeting of all 4 authors. 3. Results 3.1. Search strategy The computer-assisted search of PubMed (n ¼ 185), Cochrane Library (n ¼ 14), PEDro (n ¼ 126), and CINAHL (n ¼ 19) yielded 344 records (Appendix). Twenty papers were included in our review (Fig. 1). Published abstracts were not included because few details were available. 3.2. Study characteristics With the exception of the study by Harris and Wagnon (1987) with 196 patients, sample sizes in the retrieved studies were relatively small, ranging from 16 to 84 participants (Table 1). Investigators used 4 different mechanical stimulations: (1) myofascial technique applied to the cervicothoracic junction or to the suboccipital triangle; (2) steady traction applied to the cervical spine; (3) spinal manipulation applied to the T4/T5 facet, to the lumbosacral junction, or to the cervical, thoracic, and lumbar spine; and (4) spinal mobilization applied to the cervical spine, to the T4/T5 facet, or to the L4/L5 facet (Table 1). Investigators used a single SBF index as an outcome measurement: SC in 6 studies (Ellestad et al., 1988; Perry and Green, 2008; Jowsey and Perry, 2010; Desmarais et al., 2011; Perry et al., 2011; Moutzouri et al., 2012), ST in 3 studies (Harris and Wagnon, 1987; Roy et al., 2010; Jiang et al., 2012), PPG in 2 studies (Eshleman et al., 1971; Purdy et al., 1996), and LDF in 2 studies (Karason and Drysdale, 2003; Mohammadian et al., 2004). Investigators also used 2 different combinations of SBF indexes: SC and ST in 6 studies (Petersen et al., 1993; Chiu and Wright, 1996, 1998; Sterling et al., 2001; Moulson and Watson, 2006; La Touche et al., 2013) or SC, ST, and LDF in 1 study (Vicenzino et al., 1998). No manual medicine study using more recent techniques such as laser Doppler imaging or laser speckle contrast imaging (Roustit and Cracowski, 2013) was found. Sixteen of 20 investigators used SBF indexes as markers of PSNS activity. Other investigators used SBF indexes to evaluate vasomotor changes and PSNS activity (Karason and Drysdale, 2003), only vasomotor changes (Eshleman et al., 1971; Mohammadian et al., 2004), or changes in local ST (Roy et al., 2010) following SMT (Table 2). 3.3. Synthesis of results Due to the different methodologies, SMTs, populations, and statistical analyses, a meta-analysis of the reviewed studies was not appropriate (Table 2). Despite the small sample sizes, some authors showed significant SBF changes in studies with a control group. In

Table 2 Methodology of studies included in the current systematic review. Article

Purpose of SBF measurements

Controlled parameters of the stimulation

Location of probes

Controlled internal and external parameters

Length of recordings

Data extraction

Data expression

Eshleman et al. (1971)

Marker of peripheral vasomotion

All stimulations performed by same operator

Right forearm (unilateral)

Emotional state, temperature, exercise, and disease (no details)

10 min baseline

Averaged percent change from baseline

Time domain

Length of stimulation 5 min

Marker of PSNS activity

Ellestad et al. (1988)

Marker of PSNS activity (that might coincide with relief of LBP)

Not specified

ST probe over the right index finger

Not specified

10 s stabilization period ST measured within 10 s after the spinal manipulation

Pre- and postmanipulation readings

Time domain

SC probes placed 2 cm laterally on each side of the L2 spinous process

Room temperature maintained between 70 F and 74 F

Not specified

Subtracting the initial value after an equilibration period from the value taken 7 days following OMT

Time domain

SC and ST values noted at 15 s intervals throughout each session


Time domain

10 min stabilization period

MAX and MIN measured effects

Time domain

2 min baseline period

Mean and AUC of data points over time

Several student practitioners

All evaluations and treatments performed by 1 of 3 OMM fellows (students)

All LBP participants were prescribed a combination of chlorzoxazone and acetaminophen: 2 tablets 4 times a day for 5 days for the control group and 2 tablets 4 times a day if they continued to have severe pain for the OMT group (compliance not verified) Petersen et al. (1993)


All stimulations performed by same experienced operator

SC probes to the thumb and index fingers of the right hand

C5 spinous process marked with ink

ST probe to the tip of the right thumb

Manual palpation examination to determine the force required to produce a grade III mobilization Length of stimulation 5 min: 3 1min applications of stimulation with a 1-min rest interval in between

Participants asked to refrain from consuming alcohol on the days of the study and to avoid consuming food, drink, caffeine, or nicotine during the 30 min prior to each session Low noise, temperature- and humidity-controlled room 3 sessions on 3 consecutive days at the same time each day Normal breath

10 min in initial rest period 2 min of baseline rest period


Harris and Wagnon (1987)

5 min stimulation 16 min poststimulation

5 min of treatment period

5 min of final rest period (poststimulation) Chiu and Wright (1996)

Marker of PSNS activity used (to investigate the effects of 2 different rates of a spinal mobilization technique)

SC probes over the index and ring fingers of the right hand

Previous reliability test for the operator, frequency of mobilization recorded with a vibration sensor over the C5 spinous process: mean

ST probe over the tip of the right thumb

Instructed to avoid consuming alcohol for 24 h before the sessions, to avoid consuming caffeine 2 h preceding each session, and to avoid eating and drinking for 1 h before each session Participants and operator blinded to the recordings

239

To maintain a constant and rhythmic rate of mobilization (0.5 Hz or 2 Hz), operator listened to a tape player which had previously recorded metronome beats


Article

240

Table 2 (continued ) Purpose of SBF measurements


Location of probes

value for the rate of 2 Hz was 122 (1.872) with a coefficient of variation of 1.533 and mean value for the rate of 0.5 Hz was 30.4 (0.548) with a coefficient of variation of 1.802 Prone positioning during 27 min

Length of stimulation 5 min: 3 1min applications of stimulation with a 1-min rest interval in between


Length of stimulation 2 min

Distal phalanx of left index finger (unilateral)


Data extraction

Participants instructed to relax, remain quiet, not to go to sleep, cough, or sneeze

5 min intervention period

Room temperature and humidity

10 min final rest period

MAX and MIN values converted to a percentage of the baseline mean values Baseline period, experimental period, final rest period, and AUC normalized to a standardized time period of 5 min Treatment AUC values converted to a percentage of the normalized baseline AUC values Treatment and final rest period AUC values combined to provide a total AUC measure, then converted to a percentage of the baseline

Room temperature (25 C)

10 min baseline

Same operator

2 min stimulation or sham then cross-over to the second protocol

10 min baseline and stimulation or sham Chiu and Wright (1998)


Length of stimulation 5 min: 3 1min applications of stimulation with a 1-min rest interval in between

SC probes over the tip of the index and middle fingers of the right hand

Room temperature and humidity


C5 spinous process marked on the participant's skin

ST over the tip of the right thumb

Instructed to avoid consuming alcohol for 24 h before the sessions, to avoid consuming caffeine 2 h preceding each session, and to avoid eating and drinking for 1 h before each session


Data expression

Averaging 5 representative pulse contours from each interval Analyzed changes from baseline in total pulse amplitude (Y) and distance from dicrotic notch to peak amplitude (X) X/Y ratio measured with comfort level

N/A

7 data segments, 1 for the baseline period, 5 for the experimental period and 1 for the final rest period, grouped to determine the MAX, MIN, Mean and AUC values Baseline period, experimental period, final rest period and AUC normalized to a standardized time period of 5 min

Time domain


Purdy et al. (1996)


All stimulations performed by same operator Amplitudes of mobilizations applied between the operator and an experimented operator recorded onto a pressure sensor ICC value of 0.99 and 0.96 for the amplitudes of the C5 posteroanterior unilateral mobilization and the C5 transverse vertebral pressure mobilization, respectively Vicenzino et al. (1998)

4 sessions on 4 consecutive days, at the same time each day Participants and operator blinded to the recordings

5 min intervention period 10 min final rest period

ST and LDF of the glabrous skin over the thumb and pileous skin over the lateral epicondyle on the affected side SC of the glabrous skin over the index and middle fingers

Room temperature and noise

2 min baseline


3 30-s periods with intervening 1-min rest period 3 different days with at least 48 h between sessions

Maximum effect (maximum increase or decrease)

Length of stimulation 5 min: 3 1min applications of stimulation with a 1-min rest interval in between All stimulations performed by same experienced operator

SC probes bilaterally over the distal palmar surfaces of the index and middle fingers ST probes bilaterally over the palmar surface of the thumb

Noise attenuated, temperatureand humidity-controlled laboratory

2 min of baseline rest period

AUC for SC and ST expressed as percentages of the prestimulation measures Maximum effect (maximum increase or decrease) expressed as a percentage of the prestimulation mean level of SC and ST Minimum effect (minimum increase or decrease) expressed as a percentage of the prestimulation mean level of SC and ST

Time domain


Adjust to side-lying position on table

LDF probes bilaterally on dorsum of foot over L5 dermatome

Minimized movements of participants

5 min baseline

Measured APU

Time domain


Cavitation as benchmark for HVLAT intervention to standardize stimulation Only 1 attempt for HVLAT

Room monitored for temperature, noise, odors, light, drafts, and distractions

5 min between sham and HVLAT

APU averaged by 30 s periods

5 min after HVLAT


2 min control

APU averaged by 2-min periods


Length of stimulation 3 30-sec periods with intervening 1-min rest period

Double-blind status assessed by post-experiment questionnaire

Karason and Drysdale (2003)

Mohammadian et al. (2004)



Length of stimulation: 15 min of spinal manipulative treatment following a 20 min application of capsaicin cream on the forearm inducing local cutaneous inflammatory reactions

2 LDF probes: 1 placed in the center of the capsaicin application on the forearm and 1 positioned 2 cm from the edge of the application site, i.e., 4 cm from the center

3 sessions on 3 different days with at least 24 h between sessions

Participants instructed not to take drugs 7 days prior to sessions, not to drink caffeine or alcohol-containing beverages within 8 h prior to session, and not to receive chiropractic treatment 30 days prior to sessions


Sterling et al. (2001)

Time domain

Time domain

(continued on next page) 241

Article

242



Location of probes


All stimulations performed by same operator


Data extraction

Data expression

Mean differences between the intervention period compared with the preintervention period (Diff A) for SC and ST data Mean differences between the postintervention period compared with the preintervention period (Diff B) for SC and ST data

Time domain

Intervention period and final rest period values converted into percentage change from baseline

Time domain

Percentage change from baseline normalized to the time period

Time domain

2 min pre-stimulation 2 min post-stimulation

Moulson and Watson (2006)


ST probes bilaterally to the palmar surface of the distal phalanx of the ring fingers

Participants instructed to refrain from smoking, participating in strenuous exercise, and consuming alcohol and caffeine for 1 h prior to sessions


Intervertebral joint level of C5/C6 marked on the participant's skin

SC probes bilaterally to the palmar surface of the distal phalanx of the thumb and index fingers

Recordings of temperature and humidity before and after each session

2 min baseline for SC and ST

Noise and discussion kept to a minimum 3 sessions on 3 different days at the same time

2 min post-stimulation

All-male group used to negate the effects on variance that the female hormone progesterone has on electrodermal response


Sound-proofed, temperaturecontrolled laboratory


Inclusion and exclusion criteria as described in previous studies


3 groups matched on age, weight, and height Post-treatment questionnaire: significant differences in the perceptions of the participants as to whether they had received the treatment, placebo, or control condition Post-treatment questionnaire: no significant difference (P ¼ 0.388) between the treatment and placebo groups


Exclusion criteria as described in previous studies to control for factors known to influence the PSNS Temperature-controlled laboratory


Mean (SD) length of stimulation, 22 (3.6) s

Perry and Green (2008)


Mechanical mid-to-end range mobilization technique

SC probes bilaterally applied to the dorsum of the 2nd and 3rd toes of both feet

ICC value of 0.96 for the frequency of oscillation of the technique (2 Hz) in a pilot study ICC value of 0.80 for repeatability of depth of mobilization in a pilot study Participants in a prone standardized position All stimulations performed by same operator

Length of stimulation 5 min: 3 1min applications of stimulation with a 1-min rest interval in between Jowsey and Perry (2010)


Participants in a prone position, arms by their side, cervical spine in neutral Same experienced practitioner, a physiotherapist with 12 years postqualification Standardized hand positioning for the stimulation

SC probes bilaterally applied to the index and middle fingers of both hands

2 groups matched on age, weight, and height




Participants positioned in a chair in a standardized position and instructed to look at a marked spot on the wall

Length of stimulation 5 min: 3 1min applications of stimulation with a 1-min rest interval in between ICC value of 0.61 for the frequency of oscillation of the technique (0.5 Hz) in a pilot study Spinous process of T4 marked with ink Exit questionnaire Roy et al. (2010)

Measurement of local changes in paraspinal ST

Participants in a prone position wearing a cotton gown with an opened slit in the back for ST recordings

5 min postintervention period

Participants free of any underlying pathologic conditions (acute or chronic diseases, cold, and/or any thermogenic disease)

8 min resting period

Mean (SD)

Instructed not to drink any coffee or any other beverages containing caffeine (e.g., caffeinated soft drinks, tea) and to abstain from smoking or chewing tobacco at least 2 h before the recording sessions Women asked to present at a later time if they were having their menses

Total recording session lasted 10 min; ST measured at 6 specific time points

Differential ST in relation to the control period over the entire recording session

Time domain


Participants in side-lying posture for treatment or sham (less than 30 s to return to the prone position from the side posture)

Wooden sticks secured to the side of each infrared camera casing to ensure a constant 0.5 inch distance between the infrared camera lens and the skin surface L5 spinous process marked with ink

Subject naivety exit questionnaire

Control period (2 min before the spinal manipulation); immediately after the spinal manipulation; 1, 3, 5, and 10 min after the spinal manipulation

Sessions rescheduled if not complying with inclusion criteria Significant differences in age (P ¼ .04), weight (P ¼ .0004), and BMI (P ¼ .002) but no significant difference in height Desmarais et al. (2011)

Marker of PSNS activity (segmental response following cutaneous noxious heat and spinal manipulation of the thoracic spine associated with transcutaneous electrical stimulation of the sural nerve)

All stimulations performed by same experienced operator

Palmar SC probes over the thenar and hypothenar eminences of the left hand

4 experimental sessions lasting 60 min each, on 4 separate days

Cavitation as benchmark for HVLAT intervention to standardize stimulation

Plantar SC probes over the medial part of the left foot sole and under the 5th metatarsal head

Rating of pain and pain-related anxiety at the end of each session

Standardization of participant's positioning and operator's hands positioning

Excluded if any acute or chronic illness or any medication

10.5 min: 7 30-sec blocks of 5 electrical stimuli (6 s interstimulus interval) separated by 60 s of rest Baseline SC recorded during the 2nd and 3rd blocks (1st block served as familiarization period and excluded in the analyses) SC conditioning by thoracic stimulation recorded during the 4th and 5th blocks

30 s means

Time domain

Onset-to-peak amplitude of shockevoked SC

(continued on next page) 243

Article

244



Location of probes

Stimulation applied immediately before the conditioning period

Perry et al. (2011)

Marker of PSNS activity (to compare the magnitude of changes elicited by 2 specific manual therapy techniques)

1 single session

SC probes bilaterally over the 2nd and 3rd toes of each foot

Detailed protocol for the manipulation technique described in Maitland et al. and by Herzog


Instructed to breathe as regularly as possible and to refrain from moving for the duration of the experimental sessions

Post-conditioning SC recorded during the 6th and 7th blocks

Temperature and humidity of the room recorded


Participants instructed not to sleep, breathe deeply, cough or sneeze, talk, fidget with the sensors, or move unless otherwise instructed to do so by the investigator


Detailed protocol for the extension exercises described by McKenzie

Jiang et al. (2012) [in Chinese]

Marker of PSNS activity Marker of peripheral ST

Manual, steady cervical traction applied once every other day; total of 7 treatments during 14 days Manual traction performed for 5 e10 min in a seated position then in a supine position


Standardized participant positioning, instructions received

Length of stimulation 3 min: 3 sets of 6 repetitions (participant's active flexion in a sitting position) lasting 30 s alternating with a 1-min rest period between each set

Data expression

Calculation of the “integral measurement” (mohms) for baseline, experimental, and final rest periods Experimental and final rest period values converted into percentage change from baseline

Time domain

Mean (SD)

Time domain

Time domain

Application of 1 of the 2 techniques, intervention period; last 2 min selected for comparison with baseline 10 min final rest period, last 2 min selected for comparison with baseline Images taken from 1.5 m distance towards a fixedpoint of upper limb skin

Computer steady cervical traction applied during 30 min with participants in a seated position, once a day during 14 days; total of 14 treatments Forward angle for traction: 10 for upper neck (C2eC3), 20 for middle neck (C3eC5), and 30 for lower neck (C5eC7) Moutzouri et al. (2012)

Data extraction

Temperature (23.5 C-26.5 C) and moisture (41%e60%) controlled treatment room 24 h before intervention: no medication, alcohol, smoking, and overnight activities

Stabilization period 10e15 min Measurements taken before treatment and after treatment (14 days)

Before intervention: no clothes on the upper body

SC probes bilaterally over the plantar surface of the 2nd and 3rd toes of both feet

Temperature-controlled and sound-proof laboratory


Mean (SD)

Participants instructed to avoid exercise, caffeine intake, nicotine, and alcohol consumption for at least 6 h before measurements No baseline differences between the 3 intervention groups regarding age, weight,


Percent change from baseline




Length of stimulation 7 min: 3 2min applications of stimulation with a 30-sec rest interval in between

Rate of application of the technique (0.5 Hz) controlled with a metronome

Patient in a supine position, with a neutral position of the cervical spine Standardization of hands positioning, direction of the force, and the rate of application of the technique

Abbreviations: APU, arbitrary perfusion unit; AUC, area under the curve; BMI, body mass index; HVLAT, high-velocity, low-amplitude thrust; ICC, intraclass correlation coefficient; LBP, low back pain; LDF, laser Doppler flowmetry; MAX, maximum; MIN, minimum; OMM, osteopathic manipulation medicine; OMT, osteopathic manipulative treatment; PSNS, peripheral sympathetic nervous system; SBF, skin blood flow; SC, skin conductance; SD, standard deviation; ST, skin temperature.

Length of recordings not specified

Percent change from baseline 2 groups matched on age and clinical data (pain duration and self-reported variables) Population selected on 7 clinical criteria (diagnosis criteria, pain location, pain duration, pain intensity, pain provocation, pain description, and presence of trigger points) All participants received the same explanations about the intervention before each session 3 sessions over 2 weeks; entire experiment lasted 8 months

1st recordings after 10 min stabilization period 2nd recordings 5 min after the treatment Room temperature controlled at 25 C

ST probe taped to the tip of the 4th finger of the left hand SC probes taped to the tip of the index and middle fingers of the left hand All stimulations performed by same experienced operator Marker of PSNS activity La Touche et al. (2013)

and height but only females in the control group Participants instructed to relax and breathe normally


Mean (SD)

Time domain


245

healthy populations, SMT was associated with significant increases in SC (Petersen et al., 1993; Chiu and Wright, 1996; Moulson and Watson, 2006; Perry and Green, 2008; Jowsey and Perry, 2010; Moutzouri et al., 2012), decreases in ST (Petersen et al., 1993; Vicenzino et al., 1998), and opposite changes in LDF measurements with a decrease among smokers and an increase among nonsmokers (Karason and Drysdale, 2003). In symptomatic populations, SMT was associated with significant and opposite changes in SC with a decrease in low back pain (Ellestad et al., 1988) and an increase in cervical pain (Sterling et al., 2001) or craniofacial pain (La Touche et al., 2013), in ST with an increase in low back pain (Roy et al., 2010) or a decrease in cervical pain (Sterling et al., 2001) or epicondylalgia (Vicenzino et al., 1998), and LDF measurements with a decrease in epicondylalgia (Vicenzino et al., 1998). All investigators attributed these changes to a short-term sympathoexcitatory effect. The similarity of SBF outcomes between sham-SMT and SMT has been attributed to the nonspecific touch effect that may elicit a response greater than control through other neurophysiological or psychoemotional pathways (Moulson and Watson, 2006). 4. Discussion 4.1. Summary of findings The purpose of the current review was to assess how SBF measurements have been performed and interpreted, and discuss which methodological approaches could be included in future manual medicine research. Eighteen of 20 studies investigated remote SBF changes in limbs following SMT; only 2 investigated local SBF changes over the paraspinal tissues following SMT (Ellestad et al., 1988; Roy et al., 2010) where underlying physiological mechanisms might be different. Bilateral SBF recordings allowed investigators to discuss possible side-specific effects of SMT applied unilaterally to a facet joint (Petersen et al., 1993; Chiu and Wright, 1996; Vicenzino et al., 1998; Sterling et al., 2001; Jowsey and Perry, 2010): this design is appropriate to investigate SBF changes that authors have associated with short-term sympathoexcitatory effects mediated through central neurophysiological mechanisms (Kingston et al., 2014). All investigators used only noninvasive tools assessing changes in superficial tissues, but the extent to which they reflect PSNS changes in deeper tissues is still unknown. Further, these methods suffer from a lack of homogeneity, limiting the comparability of the different studies. Consequently, we propose several methodological recommendations to standardize measurements for future studies based on current understanding of SBF and use of LDF techniques. 4.2. Physiological mechanisms, microvascular reactivity, and spinal manual therapy Despite conflicting results, the initial response to SMT was frequently described as a peripheral vasoconstriction of the arterioles within the dermis and a decrease in peripheral SBF leading to a decrease in ST and an increase in SC (Sterling et al., 2001) with changes in SC of greater magnitude than changes in ST (Petersen et al., 1993). Short-term peripheral SBF changes were attributed to changes in PSNS function depending on the excitatory involvement of the ventral (vasoconstriction) or dorsal (vasodilation) periaqueductal gray (Perry and Green, 2008). SC has been proposed as a reliable tool to investigate PSNS response for outcomes greater than 4.6% change from baseline (Perry et al., 2011). Compared to other SBF recordings, SC captures an additional mechanism, the activation of sweat glands (Chu et al.,

246


2014) reflecting the SSNA, the specific sympathetic pathway controlling SBF. ST recordings in reviewed studies should be interpreted cautiously regarding PSNS function. ST is considered a more sensitive and reliable measure to evaluate vasomotor function rather than PSNS function (Moulson and Watson, 2006). Local paraspinal ST warming following SMT may result from a physiological reaction to mechanical pressure, a neurological reflex reaction, or renewed muscular activity resulting in a return to normal paraspinal ST (Roy et al., 2010). Pressure-induced vasodilation has been recently described as a mechanism delaying the decrease in SBF produced by local application of low pressure to the skin (Fromy et al., 2012) and has been overlooked in SMT publications (Bolton and Budgell, 2012). The SSNA could be noninvasively assessed with pulse rate variability, a frequency-domain analysis of a PPG signal that is considered an accurate estimate of heart rate variability in quiet and controlled conditions (Schafer and Vagedes, 2013). Other nonsympathetic factors are involved in SBF regulation, especially in a normothermal environment. Endotheliumdependent mediators such as nitric oxide, a potent vasodilator, and various eicosanoids, with complex cross-talk between these pathways and neurovascular regulation, are described mechanisms (Roustit and Cracowski, 2013) that may challenge interpretations of previous studies. Based on the current understanding of SSNA and SBF physiology and the methods for their assessment, different methodological approaches could be included in future manual medicine research. 4.3. Reactivity tests, microvascular reactivity, and spinal manual therapy Skin microcirculation is an accessible vascular bed proposed as a model for generalized microvascular endothelial and neurovascular function (Roustit and Cracowski, 2013). It could be used to noninvasively study superficial changes following SMT that may reflect changes in deeper tissues in specific disease populations, or it may be associated with clinical outcome measurements. The use of various reactivity tests and pharmacological tools, coupled with devices that assess SBF (e.g., LDF or laser speckle contrast imaging), has been proposed to explore mechanisms of SBF regulation in greater detail (Roustit and Cracowski, 2012, 2013). Such methods may increase the clinical relevance of studies by assessing physiological pathways involved in the response to SMT, when combined with primary clinical outcome measurements (Chu et al., 2014) and spinal palpatory signs. For example, ST has been used as a palpatory finding to determine areas of spinal dysfunction requiring SMT (Triano et al., 2013). However, little is known about normal SBF values for pain-free and symptomatic participants, and an increase or decrease cannot be proven as beneficial or detrimental (Roy et al., 2010). A recent systematic review suggests cervical spinal mobilization improved SC outcomes by approximately 20% relative to control (Schmid et al., 2008), but the minimal clinically important difference for SBF has yet to be determined (Chu et al., 2014). SBF indexes could be used as markers of specific (SSNA) or nonspecific (vasoreactivity or ST) physiological mechanisms with the potential to correlate post-SMT changes with SMT efficacy. Methods for SBF collection should be standardized for that purpose. 4.4. Methodological recommendations 4.4.1. Population study In healthy and cervical pain populations, SMT was associated with the same pattern: significant increase in SC and significant decrease in ST. In low back pain populations, investigators found

the opposite pattern (Table 1). A recent meta-analysis of studies evaluating changes in SC and ST in the upper limbs following cervical and thoracic SMT described similar results: the effect size (95% confidence interval) of SC was 0.94 (0.47, 1.41) and 0.48 (0.83, 0.12) for ST (Chu et al., 2014). These results suggest that SBF changes in asymptomatic participants may be appropriate to study SBF changes in cervical pain populations since a similar pattern has been observed. A limitation might be the absence of evaluation of the spine of participants to determine areas of spinal dysfunction requiring SMT (Triano et al., 2013), as routinely performed by clinicians. Healthy participants may have asymptomatic dysfunctional components of their neuromusculoskeletal system that could affect physiological mechanisms (Pickar, 2002; Bialosky et al., 2009). Future studies should document spinal palpatory findings where SMT is applied to improve the applicability of results in clinical practice. 4.4.2. Controlled parameters One challenge of measuring SBF is variability of measurements, which suggests that investigators should pay close attention to methodological concerns. The control of behavioral and environmental factors before and during testing sessions is vital to properly record SBF data (Roustit and Cracowski, 2013). Constitutional factors, such as age, gender, pigmentation, skin type, or smoking habits, may also influence SBF recordings (Sandby-Moller et al., 2003). Paungmali et al. (2003) provided the most detailed description of variables controlled during their study. For instance, participants were requested to be caffeine-, nicotine-, and analgesic-free 6 h prior to testing and exercise-free 4 h prior, confirmed by questionnaire; and asked to not talk, cough, or sneeze during testing in a noise-, temperature-, and humidity-controlled environment. In addition, variability of the SBF signal was documented at baseline from repeated measurements obtained by 1 investigator. The proposed potential influence of gender and estrogen contraceptives for female participants is conflicting, so menstrual cycle and contraceptive use should be considered in clinical studies but no definitive methods have been proposed to manage this influence (Roustit and Cracowski, 2013). Four investigators used male-only populations (Table 2) to avoid this influence. Additional variables, such as respiratory rate set by a metronome at 0.25 Hz or blood pressure and heart rate values taken before and after recording sessions, have been noted because they can affect the variability of recordings (Budgell and Polus, 2006). 4.4.3. Mechanical stimulations Loading characteristics on the musculoskeletal system during clinical application of SMT have been studied. Joint position, direction, velocity, duration, and force amplitude are biomechanical features of SMT, but the impact of this variability on biological mechanisms that may contribute to the clinical effects of SMT remain unknown (Pickar, 2002; Cambridge et al., 2012). In the reviewed papers, investigators controlled several parameters to standardize the biomechanical component of the SMT: participant and operator positioning, operator's experience, localization of stimulation, length of stimulation, rate of application, and cavitation as a benchmark for spinal manipulation (Table 2). Other components, such as pressure used, magnitude of applied force, and direction, are inherently difficult to measure and have not been part of most study designs although use of a noncalibrated stimulus for SMT has been reported as a limitation when interpreting SBF data (Desmarais et al., 2011). Further, 3 sets of 1-min spinal mobilization applied at a rate of 0.5 Hz (Jowsey and Perry, 2010) or 2 Hz (Chiu and Wright, 1996) with 2 intervening periods of 1-min rest have been associated with significant short-term changes in SBF


data (Jowsey and Perry, 2010). The cervical region of the spine has been the most investigated (Schmid et al., 2008; Bolton and Budgell, 2012), and upper limbs appear to be the preferred location for monitoring bilateral SBF changes. 4.4.4. Probe positioning Most SMT involve movement of participants, so investigators must manually select a portion of the signals free from motion artifacts (Bolton and Budgell, 2012). The positioning of the probes can influence SBF variability (Roustit and Cracowski, 2012). Reproducibility of LDF on fingertip skin is higher compared with the forearm, possibly due to a greater proportion of arteriovenous anastomoses in finger pads (Roustit and Cracowski, 2012). Taping probes by using an anatomical landmark for repeated measurements on the same participant may be useful. The positioning of probes over a specific dermatome to investigate segmentally related changes following SMT (Karason and Drysdale, 2003) appears less relevant due to the complex regulation of SBF. 4.4.5. Technique LDF equipment is appropriate to correlate SBF changes with successful outcomes of SMT and more affordable than laser speckle contrast imaging. LDF coupled with a reactivity test (either mechanical, thermal, electrical, or pharmacological) has been extensively used to investigate peripheral microvascular disorders in diabetes, artherosclerosis, kidney dysfunction, hypertension, and heart disease (Roustit and Cracowski, 2012). Although it requires

247

the control of several parameters, LDF is simple to implement (Table 3). 4.4.6. Reactivity tests Tests evaluating SSNA influence on SBF (vasoconstrictor reflex) include respiratory tests such as the inspiratory gasp test, thermal tests such as the cold pressor test, mental tests such as the Stroop test, and noxious stimuli (Wilder-Smith et al., 2005). Tests evaluating microvascular endothelial function include local thermal hyperemia and postocclusive hyperemia tests, inducing vasodilatory effects with limited involvement of SSNA (Cracowski et al., 2006). Common pharmacological tools include invasive tests with injection of acetylcholine or sodium nitroprusside to explore endothelium or nonendothelium dependent vasodilation. Capsaicin cream has also been used since it induces a transient receptor potential vanilloid type 1-dependent vasodilation. Mohammadian et al. (2004) used this test to compare changes following SMT with those elicited by this positive control that induces vasodilation. 4.4.7. Length of recordings Studies should use an acclimatization period before baseline recordings to allow stabilization of the SBF signal. Of the reviewed studies, the length for baseline recordings varied from 2 min (Vicenzino et al., 1998) to 10 min (Purdy et al., 1996), which are below the 20e30 min acclimatization phase suggested in a pharmacology review (Roustit and Cracowski, 2013). Ideal periods to record paraspinal ST using digitized infrared segmental

Table 3 Fourteen recommendations for future studies with laser Doppler flowmetry. Methodological aspects

Recommendations for evaluating changes in SBF following SMT

Purpose

Design

Randomized cross-over trial when evaluating changes in a single cohort Use of control and sham/light touch interventions

To reduce variability of SBF recordings

Sample size calculation Participants Experimental conditions Participant positioning Probe positioning with LDF Probe positioning in the upper limbs

Other outcome measures

Type of SMT (spinal mobilization) Number of SMT Control of the biomechanical components of the SMT Evaluation of magnitude of changes in SBF Length of LDF recordings

Data expression

Minimum of 20% change from baseline at 5% significance level with 80% power Adults with report on normal/abnormal palpatory findings of spinal tissues; naïve to SMT Control of behavioral and environmental factors (e.g., noise, distraction) Allowing a reduction in motion artifacts during sessions, especially during application of SMT Probe location on glabrous skin versus pileous skin (higher proportion of arteriovenous anastomoses in fingertips) Bilateral measurements on index, middle, or ring fingers (most commonly used sites in previous studies) On the forearm, use anatomical landmarks for repeated measures Standardize skin temperature (e.g. 33 C) SBF measurements combined with another marker of PSNS function (e.g., heart rate variability or pulse rate variability) Primary clinical outcome measurements 3 sets of 1-min spinal mobilization applied at a 0.5 Hz or a 2 Hz rate with 2 intervening periods of 1-min rest 1 SMT applied to 1 spinal segment Report on localization of hands, amount of pressure, orientation of force, and rate and length of application Use of reactivity test 20 min for baseline 5 min for SMT (3 sets of 1-min spinal mobilization with 2 intervening 1-min rest periods) 5 min after SMT Time-domain analysis by averaging SBF values and evaluating percent changes from baseline Express SBF values as percentage of maximal vasodilation

To evaluate SBF recordings without any mechanical stimulation (control) and influence of experimental conditions (sham/light touch) To answer a clinically oriented research question To document physiological responses in participants who would require SMT in a clinical setting To reduce variability of SBF recordings To evaluate SBF recordings during and after SMT without manual selection of SBF signals free from motion artifacts To improve reproducibility of LDF recordings To evaluate changes in SBF with underlying theoretical suprasegmental reflex involvement To minimize variability of locations over different sessions To distinguish the influence of PSNS versus local factors in overall SBF changes To determine minimal clinically important difference To use SMT that have shown the highest significant changes in PSNS outcome measurements in previous studies To reduce potential dose effect of several SMT applied at different spinal levels on SBF To minimize variability of somatic stimulation applied to participants To evaluate microvascular function To compare magnitude of possible changes in SBF To stabilize LDF recordings To compare initial changes with other published results

To compare the initial changes with other published results To improve reproducibility

Abbreviations: APU, arbitrary perfusion unit; LDF, laser Doppler flowmetry; SBF, skin blood flow; SMT, spinal manual therapy.

248


thermometry have been extensively investigated, and 2 periods for recording stable data have been proposed for a participant in a prone resting position. The first period of paraspinal ST stabilization occurs between 8 and 16 min while core temperature is still adapting to its environment, and the second period occurs at 30 min when core temperature is more stable (Roy et al., 2008). 4.4.8. Data extraction, analysis, and expression The “maximal effect method” (maximum increase or decrease) was used by early investigators, but this method might not be accurate due to the variability of SBF. Area under the curve has been used to quantify the effect of a treatment condition because it demonstrates the magnitude effect over time (Chiu and Wright, 1996), but its value has been questioned (Moulson and Watson, 2006). A more appropriate method for data evaluation may be determining the percent change of the signal from baseline. Currently, a related time-domain method for investigating SBF changes is the “integral measurement”, which is the summation of the total value of a physiologic parameter over time (Perry and Green, 2008; Jowsey and Perry, 2010). LDF data is expressed as arbitrary perfusion units (APU) and can be used in a time domain by averaging APU values over a specific time and then expressing it as a percent change from baseline. Spectral analysis can provide information related to various aspects of SSNA in the frequency-domain analysis. The time over which SBF is averaged should be carefully chosen because it will influence data expression (Stefanovska et al., 2011), but this time has yet to be defined by investigators in manual medicine. The use of a frequency-domain or a time-domain analysis might depend on the research question of the study, where overall changes in SBF require a time-domain analysis and changes in SSNA require a frequency-domain analysis. Because specific changes occurring in one spectrum might be missed with a time-domain analysis, no consensus on the best way to express LDF data has been reached (Stefanovska et al., 2011). 4.5. Limitations We cannot verify that all relevant articles were retrieved. However, it is unlikely that an article was missed since the databases were searched using numerous keywords and researchers in this area were contacted. Older articles using technologies that have not been replicated might have been missed, especially for the ST assessment, but most were unblinded, noncontrolled cohort studies with limited scientific relevance to inform current knowledge (Plaugher, 1992). Only 20 papers were included in our review, which prevents any definitive interpretation of the influence of SMT on SBF indexes. 5. Conclusion The current systematic review of the literature allowed us to assess published studies and propose several methodological recommendations for future investigators to properly collect, extract, and interpret SBF measurements that could be linked with clinical outcomes. Noninvasive SBF indexes have been used mainly as markers of PSNS function, but this practice may be challenged due to overlooked nonsympathetic mechanisms regulating local SBF. SC may be an appropriate marker for SSNA, but LDF should be used in combination with reactivity tests for the evaluation of local vasomotor changes. The use of SBF indexes as secondary outcome measurements in combination with primary clinical measurements, such as a pain intensity scale or other validated clinical tools, may help researchers to determine minimal clinically important differences and clinicians to understand the physiological component of SMT.

Conflict of interest One author has received research grants from Pfizer, Actelion Pharmaceuticals, Switzerland, GlaxoSmithKline, and Bioproject for other studies. The remaining authors have no conflicts of interest to report. Acknowledgments The authors thank Qunying Yang, MS, research coordinator in Research Support at A.T. Still University, Kirksville, Missouri, for extracting data from the paper by Jiang et al., written in Chinese; and Deborah Goggin, MA, scientific writer in Research Support at A.T. Still University, Kirksville, Missouri, for her editorial assistance. Appendix. Search process in databases with species (human) and language (all selected) limits for the current systematic review

Keywords

Number found CINAHL Cochrane PEDro PubMed Library

“Spinal manipulation” and “blood flow” 6 “Spinal manipulation” and cardiovascular 0 “Spinal manipulation” and conductance 0 “Spinal manipulation” and microcirculation 0 “Spinal manipulation” and sympathetic 5 “Spinal manipulation” and temperature 1 “Spinal mobilisation” or “spinal 0 mobilization” and “blood flow” “Spinal mobilisation” or “spinal 0 mobilization” and cardiovascular “Spinal mobilisation” or “spinal 2 mobilization” and conductance “Spinal mobilisation” or “spinal 0 mobilization” and microcirculation “Spinal mobilisation” or “spinal 4 mobilization” and sympathetic “Spinal mobilisation” or “spinal 1 mobilization” and temperature 0.5 Total 19

1 0 0 1 4 4 0

18 57 4 2 26 19 N/Aa

38 80 7 3 30 25 0

0

N/Aa

0

2

a

N/A

1

0

N/Aa

0

2

N/Aa

1

0

a

N/A

0

14

126

185

Abbreviations: CINAHL, Cumulative Index to Nursing and Allied Health Literature; PEDro, Physiotherapy Evidence Database. a The “therapy” search term in the PEDro database includes “stretching, manipulation, mobilisation, massage”.

References Beal MC. Viscerosomatic reflexes: a review. J Am Osteopath Assoc 1985;85(12): 786e801. Bialosky JE, Bishop MD, Price DD, Robinson ME, George SZ. The mechanisms of manual therapy in the treatment of musculoskeletal pain: a comprehensive model. Man Ther 2009;14(5):531e8. Bolton PS, Budgell B. Visceral responses to spinal manipulation. J Electromyogr Kinesiol 2012;22(5):777e84. Bronfort G, Haas M, Evans R, Leininger B, Triano J. Effectiveness of manual therapies: the UK evidence report. Chiropr Osteopathy 2010;18:3. Budgell B, Polus B. The effects of thoracic manipulation on heart rate variability: a controlled crossover trial. J Manip Physiol Ther 2006;29(8):603e10. Cambridge ED, Triano JJ, Ross JK, Abbott MS. Comparison of force development strategies of spinal manipulation used for thoracic pain. Man Ther 2012;17(3): 241e5. Charkoudian N. Mechanisms and modifiers of reflex induced cutaneous vasodilation and vasoconstriction in humans. J Appl Physiol 2010;109(4):1221e8 (1985). Chiu TTW, Wright A. Comparing the effects of two cervical mobilization techniques on sympathetic outflow to the upper limb in normal subjects. Hong Kong Physiother J 1998;16:13e8.

R. Zegarra-Parodi et al. / Manual Therapy 20 (2015) 228e249 Chiu TW, Wright A. To compare the effects of different rates of application of a cervical mobilisation technique on sympathetic outflow to the upper limb in normal subjects. Man Ther 1996;1(4):198e203. Chu J, Allen DD, Pawlowsky S, Smoot B. Peripheral response to cervical or thoracic spinal manual therapy: an evidence-based review with meta analysis. J Man Manip Ther 2014. Cracowski JL, Minson CT, Salvat-Melis M, Halliwill JR. Methodological issues in the assessment of skin microvascular endothelial function in humans. Trends Pharmacol Sci 2006;27(9):503e8. Desmarais A, Descarreaux M, Houle S, Piche M. Tuning the gain of somatosympathetic reflexes by stimulation of the thoracic spine in humans. Neurosci Lett 2011;490(2):107e11. Ellestad SM, Nagle RV, Boesler DR, Kilmore MA. Electromyographic and skin resistance responses to osteopathic manipulative treatment for low-back pain. J Am Osteopath Assoc 1988;88(8):991e7. Eshleman J, Myers S, Pantle P. Measurement of changes in blood volume as a result of osteopathic manipulation. J Am Osteopath Assoc 1971;70(10):1073e9. Fromy B, Lingueglia E, Sigaudo-Roussel D, Saumet JL, Lazdunski M. Asic3 is a neuronal mechanosensor for pressure-induced vasodilation that protects against pressure ulcers. Nat Med 2012;18(8):1205e7. Harris W, Wagnon RJ. The effects of chiropractic adjustments on distal skin temperature. J Manip Physiol Ther 1987;10(2):57e60. Jiang CB, Wang J, Zheng ZX, Hou JS, Ma L, Sun T. Efficacy of cervical fixed-point traction manipulation for cervical spondylotic radiculopathy: a randomized controlled trial. J Chin Integr Med/Zhong Xi Yi Jie He Xue Bao 2012;10(1):54e8 [in Chinese]. Jowsey P, Perry J. Sympathetic nervous system effects in the hands following a grade III postero-anterior rotatory mobilisation technique applied to T4: a randomised, placebo-controlled trial. Man Ther 2010;15(3):248e53. Karason AB, Drysdale IP. Somatovisceral response following osteopathic HVLAT: a pilot study on the effect of unilateral lumbosacral high-velocity low-amplitude thrust technique on the cutaneous blood flow in the lower limb. J Manipul Physiol Ther 2003;26(4):220e5. Kingston L, Claydon L, Tumilty S. The effects of spinal mobilizations on the sympathetic nervous system: a systematic review. Man Ther 2014;19(4):281e7. Krupatkin AI. Cardiac and respiratory oscillations of the blood flow in microvessels of the human skin. Hum Physiol 2008;34(3):323e9. La Touche R, Paris-Alemany A, Mannheimer JS, Angulo-Diaz-Parreno S, Bishop MD, Lopez-Valverde-Centeno A, et al. Does mobilization of the upper cervical spine affect pain sensitivity and autonomic nervous system function in patients with cervico-craniofacial pain?: a randomized-controlled trial. Clin J Pain 2013;29(3):205e15. Mohammadian P, Gonsalves A, Tsai C, Hummel T, Carpenter T. Areas of capsaicininduced secondary hyperalgesia and allodynia are reduced by a single chiropractic adjustment: a preliminary study. J Manipul Physiol Ther 2004;27(6): 381e7. Moher D, Liberati A, Tetzlaff J, Altman DG. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. PLoS Med 2009;6(7): e1000097. Moulson A, Watson T. A preliminary investigation into the relationship between cervical snags and sympathetic nervous system activity in the upper limbs of an asymptomatic population. Man Ther 2006;11(3):214e24. Moutzouri M, Perry J, Billis E. Investigation of the effects of a centrally applied lumbar sustained natural apophyseal glide mobilization on lower limb sympathetic nervous system activity in asymptomatic subjects. J Manipul Physiol Ther 2012;35(4):286e94.

249

Paungmali A, O'Leary S, Souvlis T, Vicenzino B. Hypoalgesic and sympathoexcitatory effects of mobilization with movement for lateral epicondylalgia. Phys Ther 2003;83(4):374e83. Perry J, Green A. An investigation into the effects of a unilaterally applied lumbar mobilisation technique on peripheral sympathetic nervous system activity in the lower limbs. Man Ther 2008;13(6):492e9. Perry J, Green A, Singh S, Watson P. A preliminary investigation into the magnitude of effect of lumbar extension exercises and a segmental rotatory manipulation on sympathetic nervous system activity. Man Ther 2011;16(2):190e5. Petersen N, Vicenzino B, Wright A. The effects of cervical mobilisation technique on sympathetic outflow to the upper limb in normal subjects. Physiother Theory Pract 1993;9(3):149e56. Pickar JG. Neurophysiological effects of spinal manipulation. Spine J 2002;2(5): 357e71. Plaugher G. Skin temperature assessment for neuromusculoskeletal abnormalities of the spinal column. J Manipul Physiol Ther 1992;15(6):365e81. Purdy WR, Frank JJ, Oliver B. Suboccipital dermatomyotomic stimulation and digital blood flow. J Am Osteopath Assoc 1996;96(5):285e9. Roustit M, Cracowski JL. Non-invasive assessment of skin microvascular function in humans: an insight into methods. Microcirculation 2012;19(1):47e64. Roustit M, Cracowski JL. Assessment of endothelial and neurovascular function in human skin microcirculation. Trends Pharmacol Sci 2013;34(7):373e84. Roy RA, Boucher JP, Comtois AS. Effects of a manually assisted mechanical force on cutaneous temperature. J Manipul Physiol Ther 2008;31(3):230e6. Roy RA, Boucher JP, Comtois AS. Paraspinal cutaneous temperature modification after spinal manipulation at L5. J Manipul Physiol Ther 2010;33(4):308e14. Sandby-Moller J, Poulsen T, Wulf HC. Epidermal thickness at different body sites: relationship to age, gender, pigmentation, blood content, skin type and smoking habits. Acta Dermato-Venereol 2003;83(6):410e3. Schafer A, Vagedes J. How accurate is pulse rate variability as an estimate of heart rate variability? A review on studies comparing photoplethysmographic technology with an electrocardiogram. Int J Cardiol 2013;166(1):15e29. Schmid A, Brunner F, Wright A, Bachmann LM. Paradigm shift in manual therapy? Evidence for a central nervous system component in the response to passive cervical joint mobilisation. Man Ther 2008;13(5):387e96. Stefanovska A, Sheppard LW, Stankovski T, McClintock PV. Reproducibility of LDF blood flow measurements: dynamical characterization versus averaging. Microvasc Res 2011;82(3):274e6. Sterling M, Jull G, Wright A. Cervical mobilisation: concurrent effects on pain, sympathetic nervous system activity and motor activity. Man Ther 2001;6(2): 72e81. Triano JJ, Budgell B, Bagnulo A, Roffey B, Bergmann T, Cooperstein R, et al. Review of methods used by chiropractors to determine the site for applying manipulation. Chiropr Man Ther 2013;21(1):36. Van Buskirk RL. Nociceptive reflexes and the somatic dysfunction: a model. J Am Osteopath Assoc 1990;90(9):792e4. 797e809. Vicenzino B, Collins D, Benson H, Wright A. An investigation of the interrelationship between manipulative therapy-induced hypoalgesia and sympathoexcitation. J Manipul Physiol Ther 1998;21(7):448e53. Wilder-Smith EP, Fook-Chong S, Liurong L. Reflex vasoconstrictor responses of the healthy human fingertip skin. Normal range, repeatability, and influencing factors. Microvasc Res 2005;69(1e2):101e5. Wright HM. The origins and manifestations of local vasomotor disturbances. J Am Osteopath Assoc 1956;56(4):217e24.

Psychosocial correlates of depression following spinal injury: A systematic review.

Pulmonary Complications following Thoracic Spinal Surgery: A Systematic Review.

Patient Reported Outcome Assessment of Periodontal Therapy: A Systematic Review.

Contrast enhanced ultrasound imaging for assessment of spinal cord blood flow in experimental spinal cord injury.

The efficacy of manual therapy for rotator cuff tendinopathy: a systematic review and meta-analysis.

Thoracic manual therapy in the management of non-specific shoulder pain: a systematic review.

Blood pressure targets for vasopressor therapy: a systematic review.

Criterion validity of manual assessment of spinal stiffness.

[Cerebral blood flow assessment of preterm infants during respiratory therapy with the expiratory flow increase technique].

Autoregulation of spinal cord blood flow.

Neural control of human skin blood flow.

Measurement of liver blood flow: a review.

Effect of fusion following decompression for lumbar spinal stenosis: a meta-analysis and systematic review.

Regional differences in skin blood flow and temperature during total spinal anaesthesia.

A systematic review of visual image theory, assessment, and use in skin cancer and tanning research.

Economic evaluation of manual therapy for musculoskeletal diseases: a protocol for a systematic review and narrative synthesis of evidence.

Changes of Local Blood Flow in Response to Acupuncture Stimulation: A Systematic Review.

Impact of Atrial Fibrillation on Coronary Blood Flow: A Systematic Review.

Skin necrosis following warfarin therapy.

Modifiable and non-modifiable factors associated with employment outcomes following spinal cord injury: A systematic review.

Regional spinal cord blood flow in primates.

Manual therapy: a critical assessment of role in the profession of physical therapy.

Physical therapy after spinal cord injury: a systematic review of treatments focused on participation.

Heart blood flow simulation: a perspective review.