Manual Therapy xxx (2015) 1e6

Contents lists available at ScienceDirect

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

Original article

Pro-nociceptive and anti-nociceptive effects of a conditioned pain modulation protocol in participants with chronic low back pain and healthy control subjects Martin Rabey*, Cheryl Poon, Jonathan Wray, Chutiporn Thamajaree, Ryan East, Helen Slater School of Physiotherapy and Exercise Science, Faculty of Health Sciences, Curtin University, GPO Box U1987, Perth, WA, 6845, Australia

a r t i c l e i n f o

a b s t r a c t

Article history: Received 26 September 2014 Received in revised form 22 January 2015 Accepted 27 February 2015

Background: People with chronic pain may exhibit pro-nociceptive phenotypes characterised partly by reduced conditioned pain modulation (CPM). Characterising variability in CPM in people with chronic low back pain (CLBP) may inform management. Objectives: To investigate pro/anti-nociceptive effects of a CPM protocol in age/sex-matched healthy controls (HCs) and people with CLBP. Design: Case-controlled trial (64 participants/group). Method: The CPM protocol involved: test stimulus (TS) (noxious pressure applied by algometer to lumbar region); conditioning stimulus (CS) (noxious heat applied by thermode to dorsal hand). CPM recruitment was measured by the change in pain intensity (rated on a numeric rating scale (NRS)) of the TS in the presence and absence of the CS. Results: Responses to this CPM protocol were variable for both groups with measures consistent with either inhibitory or facilitatory effects. A significantly greater proportion of facilitatory responses were seen in the CLBP cohort compared to HCs (73% versus 31%). In response to the CS, participants with CLBP demonstrated a mean increase in NRS scores (mean 1.3 points; p < 0.001), while HCs did not (mean
0.2 points; p ¼ 0.35) and the between-group difference in change scores was significant (mean 1.4 points; p < 0.001; effect size (Hedges' g): 1.03). Conclusion: In HCs and participants with CLBP this CPM protocol elicited responses consistent with varying pro/anti-nociceptive effects. The higher proportion of participants with CLBP demonstrating a facilitatory response suggests a pro-nociceptive phenotype may characterise this cohort. © 2015 Elsevier Ltd. All rights reserved.

Keywords: Conditioned pain modulation Chronic low back pain Pro-nociceptive Palpation

1. Background The understanding that pain in humans can be modulated by counter-irritant, noxious stimuli has formed the basis for many therapies. A variety of terms have been used to describe pain inhibition by a second noxious stimulus. The term conditioned pain modulation (CPM) is typically used when relating to humans (Yarnitsky et al., 2010). Assessment of CPM may provide a surrogate measure of a person's capacity for endogenous pain modulation,

* Corresponding author. Tel.: þ61 48 700 7116; fax: þ61 8 9266 3699. E-mail addresses: [email protected] (M. Rabey), cheryl_poon@ymail. com (C. Poon), [email protected] (J. Wray), [email protected] (C. Thamajaree), [email protected] (R. East), [email protected] (H. Slater).

and can be evaluated experimentally by applying a noxious test stimulus (TS) in the presence/absence of a noxious conditioning stimulus (CS) (Yarnitsky et al., 2010). If pain experienced from the TS is altered during the CS, CPM has occurred (Lewis et al., 2012). There is considerable variability regarding methodology for inducing CPM (Pud et al., 2009; van Wijk and Veldhuijzen, 2010; Lewis et al., 2012). Insufficient data exist to support one optimal protocol (Yarnitsky et al., 2010). TS include cold water immersion (cold pressor test), heat, pressure and chemical or electrical stimulation (Pud et al., 2005; King et al., 2009; Graven-Nielsen et al., 2012). Common CS include cold water immersion and noxious heat using a contact thermode or water bath (Pud et al., 2009; Lewis et al., 2012). The mechanism by which CPM is achieved has been investigated in animals and is thought to be due to activation of brainstem

http://dx.doi.org/10.1016/j.math.2015.02.011 1356-689X/© 2015 Elsevier Ltd. All rights reserved.

Please cite this article in press as: Rabey M, et al., Pro-nociceptive and anti-nociceptive effects of a conditioned pain modulation protocol in participants with chronic low back pain and healthy control subjects, Manual Therapy (2015), http://dx.doi.org/10.1016/j.math.2015.02.011

2

M. Rabey et al. / Manual Therapy xxx (2015) 1e6

structures modulating nociception via descending inhibitory serotonergic/noradrenergic systems (Bouhassira et al., 1992; Le Bars, 2002; Nir et al., 2011). Animal and human studies have shown pain modulation following manual therapy may involve similar pathways (Skyba et al., 2003), however, multiple pain modulatory mechanisms may be involved in CPM and manual therapy-induced analgesia, involving spinal and supra-spinal networks (Bialosky et al., 2009; Yarnitsky, 2010; Knudsen et al., 2011). Of clinical significance, CPM may be sub-optimal, compared to healthy controls (HCs), in people with knee osteoarthritis (Kosek and Ordeberg, 2000; Arendt-Nielsen et al., 2010), fibromyalgia (Kosek and Hansson, 1997), irritable bowel syndrome, temporomandibular disorder (King et al., 2009), migraine and tension-type headaches (Sandrini et al., 2006). No inhibitory CPM is evoked in 70% of people with chronic pain (Lewis et al., 2012). A recent hypothesis proposes CPM responses may reflect pro/anti-nociceptive phenotypes, with pro-nociceptive phenotypes characterised by reduced inhibitory CPM and enhanced temporal summation; and anti-nociceptive phenotypes by inhibitory CPM and a lack of enhanced temporal summation (Yarnitsky et al., 2014). Variation in this pro/anti-nociceptive balance may occur in HCs. Those with a pro-nociceptive phenotype may be predisposed to chronic pain (Yarnitsky et al., 2014). Three studies suggest a lack of inhibitory CPM pre-operatively predicts chronic post-surgical pain (Yarnitsky et al., 2008; Landau et al., 2010; Wilder-Smith et al., 2010). Those with a pro-nociceptive phenotype may also respond differently to pharmacotherapy (Yarnitsky et al., 2012). Therefore, characterising CPM in people with CLBP may inform management. Despite its significant contribution to the burden of musculoskeletal pain (Vos et al., 2012) only one study has examined CPM in participants with CLBP using a contemporary protocol (O'Neill et al., 2014). The effect of the CS (cold pressor test, hand) on the TS (pressure, contralateral thumb) varied from inhibitory to facilitatory, both in HCs and participants with CLBP; with a greater number of participants with a facilitatory effect in the CLBP cohort, possibly indicating a pro-nociceptive phenotype. Given that manual therapists apply mechanical pressure at the lumbar region during assessment and treatment (Hengeveld and Banks, 2013), examining CPM using a protocol involving pressure applied in the lumbar region as the TS could be of greater clinical relevance. Therefore, the objectives of this study were: (i) investigate potential pro/anti-nociceptive responses to CPM in HCs, where the TS was mechanical pressure applied to the low back; (ii) compare these HC data with data from age/sex-matched participants with CLBP using the same protocol. 2. Materials and methods A case-controlled trial compared the effects of CPM on HCs and age/sex-matched participants with CLBP. Ethical approval was obtained from Curtin University Human Research Ethics Committee (HR112/2012). Participants gave written, informed consent. 2.1. Participants Seventy-one HCs were recruited via advertisements. Exclusion criteria for HCs were: (i) diagnosed neurological disorders including those that may impede sensation (e.g. diabetic neuropathy); (ii) any pain disorder at the time of testing; (iii) taking analgesic medication; (iv) diagnosed mental health conditions. Following collection of HC data, a researcher (RE), blinded to outcome data, age/sex-matched HCs to participants with CLBP from a larger concurrent study using the same CPM protocol. Inclusion criteria for participants with CLBP were: (i) aged 18e70 years old; (ii) suffering CLBP >3-months; (iii) 2 points on a

Numerical Rating Scale (NRS) for pain intensity (0 “no pain”, 10 “worst pain imaginable”); (iv) 5-points on the Roland Morris Disability Questionnaire (RMDQ) examining effects of LBP on activities of daily living (Roland and Morris, 1983); (v) minimum of 60% LBP on the following question: “Which situation describes your pain over the past four weeks the best? 80% of the pain in the low back and 20% in the leg(s)” etc.; which reliably differentiates participants with dominant leg pain or dominant LBP (Wai et al., 2009), facilitating exclusion of radicular symptoms. Exclusion criteria were: (i) diagnosed neurological disorders including those that may impede sensation; (ii) serious spinal pathology. Exclusion criteria for both groups included pregnancy and inability to understand English. Participants were allowed to continue all medications as prescribed. 2.2. CPM protocol Participants lay prone. The stimuli were chosen to reduce participant burden in the larger study, and align with use of mechanical pressure in manual therapy. These stimuli have been utilized in previous CPM research (Nahman-Averbuch et al., 2013). A review of TS revealed pain ratings or thresholds are commonly examined (Pud et al., 2009). We utilised subjective ratings of supra-threshold noxious stimuli as the review's conclusion questioned use of pain thresholds due to high variability in the studies included. Application of a TS of an intensity rated as 6/10 on an NRS also reduces response variability (Granot et al., 2006). In participants with CLBP the TS was applied at the most painful lumbar region (indicated by the participant). To avoid inherent differences in tissue sensitivity TS were applied over myofascial rather than bony structures. In HCs testing was conducted over paraspinal muscles adjacent to the L5 spinous process (Giesbrecht and Battie, 2005). Pressure was applied using an algometer (probe size 1 cm2; Somedic AB, Sweden). Pressure increased at 50 kPa/s until the participant indicated the pressure was equal to a pain intensity of 6/10 (Granot et al., 2006) using a switch. This pressure was recorded. For the CS, participants positioned their arm above their head, their hand resting on the plinth. In HCs this was the non-dominant arm. In participants with CLBP, this was an arm with no hand pain, or the non-dominant arm if both were pain-free. The CS was noxious heat applied to the dorsal hand using a thermode (Thermotest, Somedic AB, Sweden). Uniformity of contact was ensured by examination of skin post-application. In participants with CLBP, heat pain thresholds may be as low as 39.5  C (Neziri et al., 2012). Therefore the CS commenced at 40  C. The participant was asked whether this was tolerable. If so, the CS was increased by 1  C at 1  C/s, and the question repeated. The duration of each increment was guided by the participant's response to the question of whether they could tolerate the stimulus. Incremental increases continued until the participant indicated the stimulus was intolerable. It was then reduced by 1  C. Despite a previous study utilising a thermode to apply a CS at 46  C (Nahman-Averbuch et al., 2013), during pilot testing one researcher developed a burn at this temperature. Therefore, to satisfy institutional ethical requirements 45  C was the maximum temperature administered. The participants then received these instructions: “I am going to apply three further pressure tests to your back. You rated the last pressure as 6/10. Now I want you to give me a score out of 10 compared to that first rating for each of these next three pressures.” No indication was given as to whether the pressures to be applied would vary from the initial pressure. This wording allowed determination of whether the participant perceived whether baseline pain intensity was altered in the presence of the CS.

Please cite this article in press as: Rabey M, et al., Pro-nociceptive and anti-nociceptive effects of a conditioned pain modulation protocol in participants with chronic low back pain and healthy control subjects, Manual Therapy (2015), http://dx.doi.org/10.1016/j.math.2015.02.011

M. Rabey et al. / Manual Therapy xxx (2015) 1e6

3

Table 1 Demographic data for healthy controls and participants with chronic low back pain. Demographic characteristics

HCs (n ¼ 64)

CLBP (n ¼ 64)

Female e n (%) Age e mean (SD) Baseline CLBP NRS rating e mean (SD) RMDQ e mean (SD) Baseline pressure to achieve 6/10 pain intensity on NRS (kPa) e mean (SD)

35 (55%) 33.5 (11.0) n/a n/a 625.3 (239.4)

35 (55%) 34.6 (10.6) 5.6 (2.0) 9.1 (4.0) 524.4 (364.3)

Statistical significance p ¼ 0.59

p ¼ 0.07

Healthy controls (HCs), participants with chronic low back pain (CLBP), pain intensity numeric rating scale (0e10) (NRS), not applicable (n/a), Roland Morris Disability Questionnaire score (RMDQ), kilopascals (kPa).

The TS was re-applied, at the same application rate, to the same point, to the predetermined pressure indicated by the participant (þ/
10 kPa). The participant then indicated another NRS score for the intensity of the TS while the CS was ongoing (Pud et al., 2009). The TS was tested after the CS had been ongoing for 30, 60 and 90 s. The mean of these three NRS ratings was analysed. 2.3. Statistical analyses Analysis was undertaken using Stata 13.1 (Statacorp, Texas, USA). Based on a power of 0.8 and a ¼ 0.05 (two tailed), a sample of n ¼ 45 per group was required to determine an effect size of 0.6 (Cohen, 1988). Change in NRS scores between baseline TS NRS (6/ 10) and mean NRS rating for the TS while the CS was applied, were grouped as defined by O'Neill et al. (2014): 1) Inhibitory effect of CS on TS (Change score decreased 0.5 points); 2) No effect of CS on TS (Change score decreased 0.5 point increase in TS pain intensity),

Table 2 Conditioned pain modulation in healthy controls and in participants with chronic low back pain. CPM effect

HCs n (%) CLBP n (%) Statistical significance

Inhibitory No Change Facilitatory

23 (35.9) 7 (10.9) 21 (32.8) 10 (15.6) 20 (31.2) 47 (73.4)

2 Point Decrease 8 (12.5) 1 (1.6)