The generalisation of predictability of pain to similar hand postures in healthy controls

Kerwin Talbot, Victoria Joy Madden, Sara Jones, G Lorimer Moseley


A study that investigated motor imagery and pain-related fear demonstrated that imagined movements can elicit pain-related fear (Meulders et al. 2015). Using a classical conditioning paradigm, the study taught participants to associated one imagined movement (Conditioned Stimulus, CS) with a painful stimulus (Unconditioned Stimulus, US) and recorded the participant’s predicted fear of the painful imagined movement (Conditioned Response, CR).  The findings demonstrated that 1.) ratings of predicted fear were greater for the imagined movement associated with the painful stimulus than for the imagined movements that were not and 2.) that the predicted fear extended to similar but distinct imagined movements (generalisation). Clinical observations suggest that pain is influenced by contextual factors (Karos, Meulders & Vlaeyen 2015). That is, pain can be influenced by a range of different stimuli, such as a particular image, thought, feeling state, or person, and not solely by movement (Gatchel & Turk 1999; Loggia, Mogil & Bushnell 2008; Meulders & Vlaeyen 2013; Price, Harkins & Baker 1987; Walco 2008). This study will use a classical conditioning paradigm to examine whether predicted likelihood of pain (CR) associated with images of a particular hand posture (CS+) is generalised to images of novel but similar hand postures.


  1. To determine the extent of generalisation of predicted pain experience (CR) to hand posture images that differ from the hand posture image (CS+) that was originally paired with a painful US.


1: To investigate the predicted likelihood of pain to images of hand postures (generalisation stimuli, GSs) that are of graded similarity to the originally painful hand posture (CS+).

2: To investigate whether the predicted likelihood of pain (CR) expressed in response to the CS+ and CS- postures modulates the generalisation of predicted likelihood of pain to hand postures of varying similarity to the originally painful hand posture (CS+).

3: To investigate whether contingency awareness is associated with the difference in predicted likelihood of pain associated with the CS+ and CS-.

4: To investigate whether the strength of the differential conditioned response is related to contextual fear ratings (Ploghaus et al. 2003).




Based on a previous study with similar methods sample size calculation was completed a priori using G*power (Faul, Erdfelder, Buchner & Lang,Institute of psychology, University of Duesseldorf, version 2, Germany) on Windows XP (Windows 7 Enterprise, 2009 Microsoft Corporation, Dell Inc, Australia). To detect an effect size of 0.4, with a power of 80% and alpha of 0.05, it was calculated that this study needs to recruit 16 participants.

All potential participants will be contacted and if they are interested, an information sheet will be emailed to them. If they are willing to participate, the participant will be asked several questions about their general health. The answers will determine their eligibility to participate in the study. The following exclusion criteria will be assessed:

  • Neurological diseases
  • Current chronic pain (pain which has lasted longer than 12 weeks and is present on most days (Green 2013))
  • Previous, recent history of chronic pain (within the last 6 months)
  • Current pregnancy
  • Current pain (at any site)
  • Hand pain or recent hand injury (unless they have had complete resolution and no symptoms for 12 weeks)
  • Inability to communicate effectively using English (read, understand and speak)

The following safety criteria will be assessed to identify any additional risk associated with the use of the Nd:YAP laser.

Exclusion criteria related to use of laser stimuli:

  • Minor or major illnesses – a list of all common illnesses that would alter heat or cold sensitivity was compiled (see Attachment One).
  • Medications that alter heat and cold sensitivity (e.g. opioids) – a list of common medications that would alter thermal sensitivity was compiled (see Attachment One).
  • Altered skin sensation ( e.g. peripheral neuropathy of the hand, significant scars or burns, Raynaud’s disease)
  • Psychiatric illnesses (e.g. schizophrenia, bipolar disorder)
  • Skin conditions on the upper limb and hand (e.g. eczema)

If the participant meets one or more of the above criteria then they will be excluded from the study. It will be more time and cost effective, for both the participant and experimenter, if participants are screened via the telephone. If eligible, a time will be arranged for participants to come in to the testing location and complete the participant informed consent process. After informed consent is obtained, the participant will be re-screened to check their eligibility for inclusion. Volunteers who have recently participated in conditioning experiments will be excluded, except under the following conditions: longer than 6 months since previous participation; the nature of the conditioning experiment was very different to the current experiment (i.e. different stimulus and different delivery of stimulus).


Materials and measures:

Conditioning stimuli: images of certain hand postures

Unconditioned stimulus: Nd:YAP laser

Images of different hand postures will be used as the conditioned stimuli (CSs). The CS+ and CS- will be assigned to two specific hand postures, and the participant will view four different orientations of the CS+ and the CS-, so as to ensure that the CSs are the actual hand postures recognised via visual images, rather than the geometric features of the images.  In fact, the CSs could be considered to be a kind of implicit motor imagery task – the process of recognising a particular hand posture, regardless of the orientation in which it is pictured (Bowering et al. 2013; Moseley 2004). The CS+ and CS- are allocations are at opposite ends of the hand posture range of motion i.e. if the CS+ is fully flexed then the CS- is fully extended. There will also be six generalisation stimuli (GSs) – six different hand postures, chosen for their position in a continuum of posture between the CS+ and the CS-, and each shown in only 1 orientation, and four unrelated hand postures (distractor stimuli, DSs) – see Figure Two.

Figure Two: Depiction of the different visual presentations of the CSs.

Depiction of the different visual presentations of the CSs

Participant and experimental setup:

Participants will be seated in a comfortable armchair, 600mm from the screen and in a dimly lit room.  The blinded researcher will record the participant’s verbal responses. The laser handpiece will be attached to a robotic arm. The robotic arm will be controlled and moved by the software (Solid Edge ST 8, CNC Mac 3 controller program, Australia) that has recorded co-ordinates, and will move the arm by 80mm after each laser stimulus. This will be done to ensure that the laser stimulus is never delivered to the same skin location twice. The images and sequencing of the stimuli will be sequenced using Gentask software (Compumedics Neuroscan, El Paseo, TX). During each block, the Gentask software will control delivery of the laser stimulation when the CS+ is displayed.  A webcam will be attached to the robotic arm to provide the blinded researcher with a good view of the hand at all times. Before the experiment and after every block of 16 trials, an infrared thermometer will be used to check the skin temperature of the hand. This will be done as an extra safety precaution to ensure that the skin temperature has not increased significantly (more than 2° hotter than the skin temperature as measured just before the start of the procedure).

Laser settings:

Neodymium-doped: yttrium-aluminum-perovskite (Nd:YAP) laser (1340nm wavelength), 4 mm spot diameter, 5ms pulse duration.  The distance between the handpiece and the hand will be 150-200mm, depending on the height of the participant’s hand and the clarity of the laser.

Calibration phase:

Participants will be required to report the level of pain evoked by each laser stimulus using an 11 point numeric rating scale, where 0 means “no pain”, and 10 means “the worst pain tolerable”(see Figure Three). The stimulus rated as a 5/10 will then be administered again to different locations on the hand, so as to ensure consistency of the chosen stimulus. The stimulus will be deemed consistent if the participant rates it as 5 for 2 of the 3 different locations. If participants rating of the laser stimulus vary then the stimulus will be adjusted until a consistent rating of 5/10 can be obtained. A stimulus intensity that consistently elicits a reported pain level of 5 (moderate pain) will be established and then assigned as the US (painful) and paired with the CS+ in the acquisition phase.

Figure 3: Numerical pain rating score

Numerical pain rating score

Group assignment:

Participants will be allocated to either Group A or Group B, to allow for the CS+ and CS- to be counterbalanced. For Group A, the CS+ will be a fully flexed hand and the CS- will be a fully extended hand. For Group B, the allocation will be reversed. The order in which the images are presented will be semi-randomised with the restriction that the same conditioned stimulus (e.g. a fully flexed hand) cannot be displayed in the same or different orientations in consecutive trials.

Acquisition phase – 3 blocks of 16 trials (see Appendix One)

The acquisition phase comprises three blocks. Half of all acquisition phase trials are question trials; the remaining half are non-question trials. Blocks 1 and 2 include both question and non-question trials. Block 3 includes question trials only. Stimuli are selected from a pool including CS+, CS- and DSs.

Trial procedure (flow of a single trial- see Figure Four)

  • Participant sits in front of computer screen
  • Fixation cross is displayed (duration 5 seconds) to prompt participant that the experimental block is beginning.
  • Image displayed (CS+, CS- or DS) in the middle of the screen (duration 5 seconds).
    IF trial is ‘question trial’ (see above), question displayed: “How much do you expect the painful stimulus?” Participant rates US-expectancy (expectation of the painful stimulus) using 11-point scale: 0, ‘not at all’ to 10, ‘very much’.

IF trial is ‘non-question trial’, image is displayed alone.

IF image is CS+, US delivered as image disappears (delay timing).

Inter-trial interval (duration 5 seconds).  Next trial.

Figure 4: Trial Procedure

Trial Procedure

Fear rating:

At the end of each experimental block of 16 trials, a visual cue will be displayed in the middle of the screen: ‘how fearful are you?’. Participants will be required to rate the question using a numerical 11 point rating scale: 0- ‘not at all’ to 10 ‘very fearful’.

Generalisation phase- 1 block of 12 trials:

This phase will consist of 1 block of 12 trials. In this experimental block the participant will be presented with six GSs, one original CS+ and CS- as well as two DSs.

Block 1:  

During the generalisation phase, the procedure will be similar to the last block of the acquisition phase, except in this phase six generalisation stimuli (GSs) will be added and all but two DSs will be removed.  The GSs will be of images depicting novel hand postures that vary in similarity between the CS+ and CS-. Restrictions will be placed on the sequencing of the trials, such that no more than two consecutive trials may have the same stimulus type or orientation.

Extinction phase: 1 block of 16 trials

The extinction phase procedure will follow exactly the same outline as the acquisition phase, with the exception that the CS+ and CS- will no longer be followed by the US.

Hand posture order:

After completion of the experiment, subjects will be given all images (six GSs, one CS+ and one CS-) and they will be asked to arrange the images in a sequence that is meaningful to them.

An awareness test will be completed to determine the level of learning that has occurred during the acquisition phase. This will be done by presenting participants with the six GSs and the CS+ and CS-. The participants will then be asked to identify the image that was linked with the laser stimulus and the image that was not linked with the laser stimulus. Participant’s awareness will then be rated depending on how close the chosen image was to the correct image (scale of 8) i.e. a participant will be classified as ‘aware’ if the image they pick is either the correct one (8) or the next similar one (7), if the participant fails to do this, they will be classified as ‘unaware’. they are a 6 or lower.


Participants will be asked to fill out the Positive and Negative Affect Scale (PANAS), Catastrophising Assessment Tool (CAT) and State-trait anxiety inventory (STAI) questionnaires. This will allow us to examine the influence of other variables that may affect the perception of pain. The PANAS was selected because negative affect has shown to increase pain perception and positive affect to decrease pain perception (Meulders, Meulders & Vlaeyen 2014). The CAT is used because catastrophising can lead to over-estimation of the probability of pain (Hood et al. 2012). In this study the probability of pain will be the primary outcome, therefore catastrophising may alter the ratings and therefore will be measured (Turner, Mancl & Aaron 2004). Lastly, trait anxiety can influence an individual’s perception of pain and therefore will be recorded (Ploghaus et al. 2001).

Manipulation checks:

Predicted pain and fear:  participants will be asked to rate the predicted pain throughout the experiment, as well as rating their fearfulness at the end of each experimental block using a numerical 11 point rating scale.

Before the experiment participants will be informed about the importance of concentrating. After each experimental phase, participants will be asked questions (‘Do you feel you were able to pay attention for the whole block?’ & ‘Did you look at each individual image?’) regarding whether they were able to concentrate during the experiment.  If the participant states that they were unable to maintain attention during that phase of the experiment then that participant’s data will be excluded from the study.

Proposed data extraction and statistical analysis:

All data will be recorded in Microsoft Excel (Windows 8, University of South Australia, Adelaide) and then analysed using SPSS (IBM Corp. Released 2012. IBM SPSS Statistics for Windows, Version 21.0. Armonk, NY: IBM Corp)

1:  Generalisation of predicted pain (primary analysis): One-way repeated measures ANOVA with 8 levels will be used to determine the influence of the pictures of the 8 different hand postures on predicted pain – (factors: hand posture – 8 levels, outcome: expected painful stimulus (US-expectancy). We expect an interaction between image 8 (CS+) and 7 (GS 6) and predicted pain.

Exploratory analyses:

2: Contingency and generalisation:  Two-way mixed ANOVA – between-subjects factor of group (aware / unaware) x within-subjects factor of hand posture (8 levels) will be used to determine the influence of contingency on generalisation of pain prediction. We expect steeper generalisation gradients in the aware group than in the unaware group.

3: Contingency and conditioning:  Three-way ANOVA- block (2: First / final acquisition block) x group (2: aware/unaware) x stimulus type (2: CS+/CS-).  We expect a greater 3-way interaction between block and stimulus type for the aware group than for the unaware group.

4:  Fear and conditioning: Two-way ANOVA – block (First/final acquisition block) x group (aware/unaware). We expect a difference in fear ratings between the aware and unaware groups for the final acquisition block, but not for the first acquisition block (Field 2009)

Appendix One– Question trial of CS+

Question trial of CS+