Patients get chronic neck pain because they are fearful and catastrophic right? Wrong.

One of the things we try to do at BiM is to help PhD students get their word out. Here is an example, but before you click out because it is ‘just a PhD thesis’ take a look at it because this is one impressive PhD AND it throws up a really interesting finding. DR Esther Williamson asked 599 people a range of questions just after they hurt their neck. Then she followed them for a year. I won’t steal her thunder but I will suggest to you that you should read this post.

Beliefs about recovery and injury and their role in the development of late whiplash syndrome


My PhD research investigated the role of patient held beliefs about injury and recovery in the development of late whiplash syndrome (LWS) following an acute whiplash injury. LWS refers to ongoing symptoms sufficient to interfere with normal activities 6 months after injury. It was hoped that by gaining greater understanding into how patient beliefs influence outcome we can improve physiotherapy management of acute whiplash injuries. Participants taking part in this research also took part in a randomised controlled trial of physiotherapy treatments for whiplash injuries[1].

Studies and results

Following a systematic literature review [2], a prospective cohort study (n=599) was conducted to investigate if beliefs assessed soon after injury were risk factors for poor outcome. Patients who thought they would take longer than 6 months to recover or were unsure they would recover were more likely to develop LWS. Lower expectations of benefit from the treatments that patients’ received as part of the RCT were also a risk factor for LWS. Patients with less confidence in their ability to cope with their neck problems (self-efficacy) were at risk of poor outcome in the short term (at 4 months) but not longer term follow up (8 and 12 months). Interestingly, fear avoidance beliefs and catastrophic thinking were not related to outcome even though these factors are often considered risk factors for poor outcome. Potential confounders such as baseline pain intensity, disability, distress and age were controlled for in the analyses.

I also conducted a qualitative (interview) study (n=20). This study highlighted that although being optimistic was helpful it was important to help patients form realistic expectations of recovery. Patients emphasised the value of reassurance from health professionals. A feature of patients recovering slowly from their injury was a lack of confidence to self-manage their injury and they often expressed unhelpful beliefs about pain. Fear avoidance early on after injury was common but it became problematic when patients did not overcome their fears and return to activity. This helps to explain why baseline fear avoidance beliefs do not predict outcome.

Implications for practice

These findings suggest several potential ways we can help patients to recover from whiplash injuries. Provide reassurance to patients. It is very important and valued by patients. Encourage patients to be optimistic but realistic about their recovery. Ensure patients have the confidence (and skills) to manage their own condition. Address unhelpful pain beliefs that could be potential barriers to recovery.

About Esther Williamson

Dr Esther WilliamsonEsther is originally from Brisbane, Australia where she did her physiotherapy degree at the University of Queensland. Esther has lived in the UK for the past 15 years and recently completed her PhD at the University of Warwick where she worked part-time as a research fellow in the Clinical Trials Unit. Esther’s main research interest is finding out ways that physiotherapists can improve the management of acute injuries and prevent long term disability. Her other job is looking after her two kids, Fraser and Annie.


[1] Williamson E, Williams M, Hansen Z, Joseph S, & Lamb SE (2009). Development and delivery of a physiotherapy intervention for the early management of whiplash injuries: the Managing Injuries of Neck Trial (MINT) Intervention. Physiotherapy, 95 (1), 15-23 PMID: 19627681

[2] Williamson E, Williams M, Gates S, & Lamb SE (2008). A systematic literature review of psychological factors and the development of late whiplash syndrome. Pain, 135 (1-2), 20-30 PMID: 17570588


  1. Frédéric Wellens,pht says

    Hi Esther,

    Great study. More and more, expectation comes up as a potential prognostic tool. Could it be that expectation really affect outcomes or rather, that patient are good judges at analysing the true gravity of their painful condition thus the association between outcome and expectation at baseline? That does seem somewhat important to me as it would definitively moderate the impact of reassurance by the therapist in the latter example.

  2. Yeah well … I both seem to exaggerate if not catastrophize pain experiences and more often, suppress them – both successfully.

    My phantom pains are usually ridiculously painful but I do let my brain drive pain perception down to very low levels and after 3 years of successfully doing that I can confirm that it works VERY well. Well, in fact, to the point where I had a two segment Herpes Zoster (Shingles) on my left hip in February 2011 after some serious overtraining (too much sports) and smoking off some parts of my immune system, contracting Herpes Zoster and H1N1 flu within a week or so – and it took me about 4-5 days into the fully blown Shingles rash to actually *consider* it, and to not that the skin there was slightly tender. I slept and worked without noticeable distraction and certainly without pain killers all through the shingles rash time which took maybe 2-3 weeks all in all. The skin stayed a tiny bit “tender” until April or so but I cannot say when it returned to normal as I did not care to check or follow up on it, so little did it hurt there. And I gave that no attention. – Conversely, if a threshold of pain is exceeded and maintained over a period of time, I will at one point in time, then rapidly and over maybe 30 to 90 minutes, get seriously nauseous, and then pain is not even too painful, but definitely exhaustive. I will suddenly feel sick and completely drained. I would say that then, pain levels cause significant distress and adrenaline to get pumped out even though I may suppress pain perception per se. But I cannot suppress waste of adrenaline – and if the adrenaline and endorphine machine collapses, I then really get pain flush and exhaustion, nausea and a system collapse at once in big waves. And then the world around me really collapses, and I plunge into a very deep valley of pain, tiredness, nausea and depression. All within 1-2 hours, it is pretty dramatic. Then I *must* address pain issues and maybe causes. To do that it helps to take it a bit serious. As no one can see that this is like that with me, invisible from the outside, they mostly think I exaggerate. That would be what people call catastrophizing I would assume. But really, it’s a pain machine pain generation pain suppression hormone roller coaster, it has its moments of hell.

  3. Andrew Claus says

    Thanks Esther, BIM and other contributors to this thread.
    I really appreciate sitting in on the constructive discussion.

  4. Hi John. I think it is important to be careful to distinguish between risk factors and mechanisms. My study has identified risk factors for poor outcome or factors that are not risk factors as in the case of fear avoidance or catastrophising. From my study this means that catastrohising measured by the pain catastrophising scale and fear avoidance as measured by the fear avoidance beliefs questionnaire cannot be used to identify patients at risk of poor outcome. These findings are in agreement with other studies such as the work by Michele sterling. This does not mean they do not have a role to play in the development or maintenance of disability (I am sure they do). It means they cannot be used to screen patients.
    It is an important distinction and was probably lost in the summary of my thesis but I was trying to condense 60,000 words into 400. Sorry!

    David Reply:

    I think Esther is providing further evidence of the need for primary care providers to address the psychosocial issues which perpetuate disability. As Geoff highlights above this is far too frequently avoided in acute setting’s with the result that there are some fairly well-established behaviors, beliefs, fears etc. by the time integrated care is instituted.

    I’ve observed this debate over at least a decade regarding interdisciplinary boundaries, scope of practice and preferred provider to deliver this holistic approach. I think this allowed the waters to be sufficiently muddied to avoid the necessary change in primary care practice. The decades of the flags conference 2007 ( I think was a significant watershed in establishing an individual duty of care for primary care provider’s to address these issues.

    As David rightly points out this is difficult, challenging and can take many clinicians out of the comfort zone

    Regarding the specifics of catastrophisation, fear and perceived disability (and even phobias as John Ware alludes to above as it has been suggested that the criteria for defining fear of pain and phobias are remarkably similar I wonder if in reality are they such distinct phenomena as the well-established rating scales would suggest?

    Far be it for me to dispute their validity but I wonder if there is much difference in the therapeutic management of somebody who is fearful of the pain messages they percieve and someone who’s just “hacked -off “because they do not function at a level they expect /demand of themselves?
    We may choose to define psychosocial factors by particular scales /category but ultimately I would suggest that if the brain output is one of a predominant “negative emotion” reflecting the summation of input and processing then the glass is “half empty” and will be perceived as such. The clinical challenge is to create a summation of input which will ultimately tip the balance into a positive emotion – the glass “half full”, via all the associated neurobiological correlates involved.

    Empathy, believing, non- judgmental communication and caring word’s may all sound a bit “new agey” but I suggest can go a very long way.

    To this end I think styles of communication as well as educational content is likely to be highly relevant. In terms of the “How” Geoff, I think all the factors which contribute to therapist credibility are uncomfortable aspects we must confront and develop necessary competence.


  5. Thank you for that, Esther.

    What may be contributing to my confusion is the title of this particular thread, which rather boldly states that fear and catastrophizing do not lead to chronic neck pain after a whiplash injury. I don’t think we can draw that conclusion yet based on just one-albeit a very large and well-designed-study. The terms we use to define these states are murky and the effects that the cognitive and affective components of the neuromatrix have on the pain experience are exceedingly complex. So, I think it’s going to be a while before we can draw definitive conclusions about how factors like fear, expectations, catastrophizing, distress- however those are operationally defined and measured- will impact the ultimate outcome.

    Certainly more studies like this that steer away from the “trees” of peripheral events and more towards what’s going on in the “forest” of the patient’s brain- where the pain experience is constructed- will help.

  6. I also included a measure of post traumatic distress (Impact of Events Scale) in my study but my results did differ somewhat from Michele Sterling’s work as I found that it was predictive of non-recovery in the short term (4 months follow up) but not the longer term (8 and 12 months follow up). I definately agree it is something that therapists should be aware of and should refer individuals on for specialist managment if they appear to have PTSD as there is clear evidence that is related to poor outcome. My study found that general psychological distress meausre by the GHQ-12 was a more consistent predictor of outcome than the IES. However, you may expect there to be some correlation between the two meausures so if the GHQ was not included the IES may have been significantly associated with outcome at the later follow up points. Due to the nature of statistical modeling the results you get will depend on the measures that you use and what you include. Also, I was not able to do any physical meausres of pain hypersensitivity (such as pain pressure thresholds) that were included in Sterling’s models as we did not have the equipment available so the findings may have been different if these were included. My study was in agreement with Michele Sterling’s work that fear avoidance and catastrophising were not predictive of outcome. However, you make a very good point that these things are probably related and will impact on each other. PTSD may impact on catastrophising and fear avoidance but my study did not actually investigate the relationship between these variables. It is also possible that catastrophic thinking or fear avoidance may moderate patient’s expectations of outcome but I did not examine this in my study. I may look at this data in the future. We still have a lot to learn about how all of these constructs interact.

  7. David Nolan says


    I think you make a great point in your last answer. I feel too that acute injuries are often dealt with within the medical model until they don’t get better and then it flips on its head and is dealt with purely psychologically. The beauty and the art is between the two, and as I keep telling my Dr colleagues, physiotherapy is really really hard. One day I hope to get better!

    Great research.

  8. I’m curious what the author and others think about how the findings from this study regarding expectations and beliefs jive with Michele Sterling’s work showing the relationship between post-traumatic stress disorder (PTSD) and poor outcome from whiplash (particularly this one: Sterling M, Hendrikz J, Kenardy J. Similar factors predict disability and posttraumatic stress disorder trajectories after whiplash injury. Pain. 2011Jun;152(6):1272-8. Epub 2011 Mar 10. PubMed PMID: 21396780).

    It seens logical that PTSD would have a negative impact on beliefs and expectations following any injury. It may also contribute to catastrophizing and fear-avoidance behavior, but, according to this study, not necessarily.

    I’m going to have to read the study, but I’m not sure how we separate out the impact fear has on beliefs and expectations. Doesn’t the limbic system impact cognitions?

  9. Hi David,
    Thanks for your comments. I have just given you the head lines of my thesis due to the word limit but you ask a good question. Reassurance about what? It is largely what you have suggested. Reassurance on a whole range of things including: the injury is not serious (e.g. no broken bones) and further scane or x-rays are not needed, reassurance that it is ok to exercises/be active and not do further damage, reassurance that it is not unusual to experience pain at this point in their recovery etc. Addressing any unhelpful beliefs about pain and injury may be part of providing this reassurance. Prognosis is a tricky one and so many patients want to know when they will be better. Generally, I think we need to encourage optimism but this should be balanced with the whole clinical picture and it maybe a realisitic expectation that recovery happens over a number of months rather than weeks. I agree completely, we can’t expect a generic management approach to fit everyone but I think there are some key aspects that should be considered with patients and addressed appropriately. For example: their expectations of treatment, their expectations of recovery, their ability to cope with their injury and what sorts of strategies are they using, how do they interpret pain, are there other potential barrier to recovery (my interview study threw up a few other important factors to consider such as priority – how much of a priority is attending for treatment?). You are obviously very much on board with this type of approach so it might look to you like I am just stating the obvious but unfortunately not everyone is like you and there a still an aweful lot of therapists who work within a medical model. From dealing with therapists during my research, I would also say that there are quite a lot who struggle to draw in the psychosocial aspects when it is an acute injury and see that as being reserved for chronic pain patients. You described it as a clinical challenge and that is exactly what it is. I think it actually takes a very skilled clinician to integrate these skills into practice and we are still learning what is important in regards to whiplash. You are right, there is still a lack of consensus on what is the best way forward but hopefully this research is starting to work towards gaining an understanding on why some patients recover and some don’t so we can develop effective treatment strategies.

    Geoff Bostick Reply:

    Hi Esther and David,

    Great thoughts – I have more questions than answers …

    I agree optimism should be endorsed. However, what happens in the context when one’s positive outcome expectations become incompatible with an experience of persistent pain? In other words, if the dominant pain narrative after injury is restitution, a desire to be cured or fixed, (borrowed from Arthur Frank) what happens when this story does not fit as pain persists? Does a therapist for an acute injury speak a language congruent with restitution (do A, B and C and your facet injury will be fixed up) and then speak another language related to management and acceptance of pain when the tides turn chronic? How do you foster optimism, yet suggest conceptualizing pain as fixable is not so good either?
    Most people, in my research anyway, appear to expect a good outcome – or is it desire a good outcome. So what is the best expectation? A desire (reflecting motivation) for a good outcome, or an informed positive expectation?
    Arrrgh, my head hurts. David might be on to something though – we often consider patients expectations and beliefs in prescribing treatments like manipulation, IMS, etc., perhaps we should do the same with education? After all, who am I to say that your beliefs are wrong and mine are right!


    Esther Reply:

    It is really difficult. I definately think we should encourage optimism as these patients were in the acute/sub-acute phase of recovery and you would expect that natural recovery was still occuring. Maybe the therapists role in this situation is to make sure that beliefs, expectations etc are not getting in the way of the natural healing process. Cognitions have the ability to modulate pain so we need to try and make sure that happens in a helpful way and not an unhelpful way. Expectations drive behaviours so positive expectations of treatment will mean patients are more likely to engage with treatment/advice given. When a patient does not recovery as we hope it does become more difficult to help them adapt their expectations. In terms of beliefs, I think there are some beliefs that we know based on pain physiology are not helpful so eduaction is warranted on these factors to remove potential barriers to recovery. The more I think about it, there more there is to still figure out.

  10. Esther
    I’m reminded of a former research supervisor who advised me only to work with cadavers!

    I took his advice at that time but now see the error of my ways to (I don’t treat many cadavers now! ) – and I certainly admire your tenacity / perseverance for the whiplash cohort you followed.

    On a practical note regarding the implications for clinical practice there are a number of regular issues which present conflicting clinical challenges and I think would it be interesting to hear your thoughts.

    I would think “providing reassurance” would be now fairly widely accepted. However the devil is in the detail. Reassurance about what?

    That they haven’t sustained catastrophic damage?

    That they do not need to see an orthopedic consultant?

    That the role of radiological investigations is of minimal the clinical utility (and potentially harmful, see other BIM posts on this).

    The predominant question is get asked is” will it get better and how long will it take”. I think it’s a fairly reasonable question? Reassurance on recovery trajectory in the acute whiplash phase is not easily predictable (other than quoting statistics). In the chronic or LWS scenario the facts already speak for themselves – I would suggest?

    Depending on the clinical signs, the duration and severity of symptoms and the cluster of psychometric profiles regarding beliefs, attitudes, fear avoidance, coping strategies etc. determine how answer to this ubiquitous question, I believe.
    In general I would exercise caution about prognostic discussions in a case demonstrating catastrophic thinking. This is not trying to be evasive but more a strategy of leading towards a more definitive prognosis over the course of three or four contacts where physical signs can be monitored, psychosocial elements be evaluated and practical benefits of cognitive functional interventions (borrowed from Peter O’Sullivan’s cognitive functional therapy model) can be evaluated.
    I don’t see this in terms of avoiding uncomfortable discussions but more a reflection of the dynamic nature of communication and the multitude of variables which can be utilized to address the cortical matrix so commonly referred to in these pages.

    Conversely some patients prefer more direct style of information delivery, an analytical evaluation of their symptoms and a structured management plan. Whether that information is in the form of detailed anatomical explanation, neurobiological mechanisms or more vague generalities I think depends more on the patient’s preference rather than necessarily what research tells us – and by implication dictates variable therapist strategies .
    This may well be a simplistic view in a forum of accomplished researchers but nonetheless I think still warrants discussion /recognition.

    So as the saying goes “one man’s meat is another man’s poison”. The type of information, the timing, the prognosis, the components of the management plan, the coping strategies and indeed the weaning from intervention to a self-efficacy program in reality is not a one-size-fits-all approach.

    The clinical challenge is to weigh up all of these variables in a continuous manner and importantly to be able to adapt strategies as necessary. In uncomfortable truth is that a significant proportion of persistent pain patients like to remain within a system of support even when acknowledgment has been reached regarding the prognosis. It then becomes an issue of whether this support is deliverable or outside the remit of the healthcare provider.

    Is it realistic to expect generic management approaches to be effectively delivered and accepted? If we can’t reach consensus on physical interventions the mind boggles at the array cognitive/ neurobiological options.


  11. Good luck Geoff with your PhD viva. Feel free to email me: if you want to discuss anything further. I would be very interested in finding out more about your study.

    Lorimer, the key to recruitment is mostly chocolate. Maybe you could send out virtual chocolate to get people to take part.

  12. On-line quiz can be found top right just underneath the photograph box…

  13. I might add my congrat’s Esther. Perhaps you could use whatever it was to GET MORE PEOPLE TO PARTICIPATE IN OUR CURRENT ON-LINE QUIZ!!!

  14. Dan Aldrige Flores says

    I am a HILOT/ CMT/Health Educator. I’m a Pain & Chronic pain Management Therapist. I have work with a lot of clients with neck pain due to whiplash. Unknown to many, stress make our neck muscles tight especially when your using computer at work, you barely move. after 5 or 10 your neck muscles start to tightens up., and a car accident will strain that muscles, & become inflame. When inflamation subsidea or heal it leave a mark , a lump or knots. Those with whiplash has knots on the back of neck near the base of the skull.Whenever you move it cause arterial tension resulting in pain. Right now I have a client(nurse) who calls me when in pain, after I work pain go away. Call again after month or two for another session. It works she get relief. She dont want surgery, so she do it this way. Well I’m just trying to tell you what I do and thought about these subject.Thanks

  15. Geoff Bostick says

    Hi Esther,
    Congratulations on your success! I am defending in 52 hours and my project is remarkably similar to yours (save the impressive recruitment). I also found baseline future beliefs (Permanence on the PBPI) as well as Medical Cure beliefs (SOPA) explained more variance in pain intensity then Catastrophizing. Since my study suffered from a poor response and significant attrition I am cautious about my interpretations, so it is nice to see some similar themes. I also had a qualitative piece examining factors why people believed what they believed. I’d love to find time to chat with about similarities and differences in our work.


  16. This is a very interesting study. It clearly demonstrates the importance of really thinking about how to approach a problem such as whiplash—what is the patient doing and thinking that could possibly affect outcomes? How can we influence these with reassurance, explanations, our words, the treating environment and working with expectations? Seeing what could be prohibitive is important but reinforcing ‘useful’ actions can be powerful.
    In the early stages I help patients to develop their own skills to solve problems and deal with the pain effectively. These are learning opportunities and we become more skilled with practice. This is true for re-learning movement as much as for thoughts that could be underpinned by strong beliefs. Simple explanations about pain, analogies, examples of other cases, raising awareness of how the body responds to threat, tackling the influences upon pain and altered movement, considering other body systems and ‘the story of the nervous system thus far’—wired from previous?—I find helpful in an integrated manner.

  17. Thanks for your kind comments. My PhD was carried out alongside a large RCT (Managing Injuries of the Neck Trial (MINT)) that ran in the Emergency Departments (ED) of 12 NHS trusts (16 EDs) and we recruited for nearly 2 years. The study had two steps. In the first step we compared two types of ED advice. In this step of the trial we recruited 3800 patients and managed to follow up 70% at one year. The patients in my PhD studies were from step 2 where we looked at physio treatment for individuals with ongoing problems 3 weeks post injury. Step one was cluster randomised so the EDs were randomised to deliver either the usual advice or an active management approach based on The Whiplash Book (as opposed to the individual being randomised). We had a great team of researchers who worked incredibly hard to recruit that number of patients. We visited each ED 2-3 times a week, we designed a proforma that ED clinicians used to assess patients which doubled as a referal to the study, we attended as many doctor training sessions as possible and generally hung around the ED making sure that everyone knew about the study. We also gave out large amount of mints and chocolates and ran competitions to encourage staff to tell patients about the trial and refer them to us. As the trial was cluster randomised we did not have to officially gain consent from patients in the ED. The ED clinicians asked the patients if they were happy for them to pass their details onto the university. We then wrote to the patients and asked them to take part. We wrote to around 7000 patients through out the course of the study. They were not officially part of the study until they returned their baseline questionnaire and by doing this they gave their consent to take part. Patients were also informed they could self-refer to step two at this point and 599 patients then went on to be individually randomised in Step two. We have just submitted the study findings for publication but if you are interested in more info about the study protocol have a read of this:
    Lamb S, Gates S, Underwood M, Cooke M, Ashby D, Szczepura A, et al. Managing Injuries of the Neck Trial (MINT): design of a randomised controlled trial of treatments for whiplash associated disorders. BMC Musculoskelet Disord. 2007;8:7. I am also busy writing papers from my PhD so hopefully there will be plenty to read about the MINT study soon.

  18. Steve Kamper says

    Aside from agreeing with James (although to be honest I’ve tried to erase those recruitment nightmares from my memory) I’d add my congratulations on getting through such an impressive study.
    One thing in particular that I like is the fact that you’ve (by the looks of your description) looked at the predictive relationships of these beliefs with outcome rather than the cross-sectional associations that are most commonly reported. Great work, looking forward to reading the report when it comes out!

  19. Esther.. this sounds like an incredibly impressive body of work. Recruiting (almost) 600 patients with acute whiplash is, well, amazing.

    We have conduced some WAD trials and cohort studies in Sydney and found recruitment very slow, painful even. Care to share how you managed this? Or are drivers just really really bad in the Midlands??