Part 4: The role of operant conditioning and pain

We’ve been talking about pain. In Part 3 of this article series we introduced the the idea that the brain can “learn to hurt”. Through Classical Conditioning the brain has the potential to create fear-avoidance beliefs and behaviors. But before we leave this topic we also need to also look at how these fears and phobias may be reinforced and maintained over time. In some cases long after they provide any purpose or benefit to the patient. For this we need to review the work of yet another prominent psychologist, B.F. Skinner.

According to Skinner, behavior is a function of its consequences. In other words, behaviors are more likely to occur when followed by pleasant outcomes (positive reinforcement) or by the removal of unpleasant ones (negative reinforcement).

The inverse is also true. Behaviors are less likely to occur when they are followed by unpleasant and undesirable outcomes (positive punishment) or when positive results fail to occur (negative punishment).

In even simpler terms, reinforcement of behaviors leads to their maintenance. Punishment or nonreinforcement of behaviors leads to their extinction. Of course these principles are easily understood. We see them in practice when training the family dog or putting a gold sticker next to our kids name on the chore chart. But what is less obvious is how reinforcement and punishment relates to pain.

Operant Conditioning and Pain

Think of the individual who experiences an episode of acute low back pain after lifting or bending. Such an episode will result in bending being painful. This pain is a form of positive punishment. According to Skinner, this reduces the likelihood of the behavior being repeated.

Alternatively, not bending avoids the pain. In operant speak, this is negative reinforcement. The removal of an undesirable outcome. Again, according to Skinner, this reinforces the behavior and increases the likelihood of reoccurrence. (In this case the behavior is to stay tall and upright instead of bending).

But there’s a critical point to make here before moving on from this section to the stuff that’s actually important. (Like how all this rambling on about the brain and our ancient ancestors matter. And how these boring theories from long-gone psychologists that you thought you would never have to hear about again after you passed your first year psychology class changes how we treat patients in pain.) It’s the reason that we needed to include the seminal work of Burrhus Frederic Skinner here in the first place. 

Reinforcement comes in many forms

When it comes to reinforcement, it’s not just personal experience that matters. What a person observes or is told by others is another powerful player. Especially when this involves figures of authority or those held in high regard.

So the pain a patient feels with one activity (or the lack of pain with another) is a form of reinforcement. But so is the message delivered by a clinician or therapist. A doctor telling their patient that flexing their lower back when bending or lifting can be harmful to their spine and damage their discs can also serve to reinforce unhelpful or maladaptive pain behaviors.

Here’s the point…

When treating patients, it is not just about specific treatment techniques or which stretches or exercises we prescribe. It’s also about creating a dialogue with patients about the source and meaning of their pain. Here we can see the power of educating patients on the modern understanding of pain. About how chronic and recurrent pain is more about tissue sensitivity and an overactive pain system as opposed to tissue damage. And that imaging findings like building discs and arthritic joints don’t have to mean that there is no hope.

But be careful. Not all patients fear and avoid movement

The fear-avoidance model has had a significant impact on treating patients in pain. Encouraging patients to be more active and get back to activities despite their pain is often a critical component of care in these cases.

But before moving forward a word of caution is needed here. Not all pain patients will hold these fear-avoidance beliefs. It is also common to see patients who continue pushing through offending tasks and activities despite ongoing or worsening pain. This seems particularly common when certain tasks are closely tied to a person’s identity or self-worth, or if the activity is a significant source of joy or well-being. Athletes and musicians come to mind here.

In fact, these endurance-related pain responses have been shown to occur as often or even more frequently as fear-avoidance responses (1,2). And as unhealthy and maladaptive as fear-avoidance responses can be, these endurance-related pain responses can also be detrimental to health and recovery.

So pain education needs to be tailored to the unique needs and beliefs of each individual. Messaging to help patients understand that hurt does not always equate to harm, along with encouragement to increase exposure to feared or sensitive movements may be helpful to patients exhibiting fear-avoidance beliefs. But this advice is less suitable with endurance copers. These patients may do better with a temporary activity reduction then building back to desired activity levels. More on this in the next article.

References

1. Hasenbring MI, Verbunt JA. Fear avoidance and endurance-related responses to pain: New models of behavior and their consequences for clinical practice. Clin J Pain. 2010; 26(9).
2. Smeets RJ et al. The association of physical deconditioning and chronic low back pain: A hypothesis-oriented systematic review. Disabil Rehabil. 2006; 28(11).

PART 3: Fear, learning, and pain

From an evolutionary perspective, pain is a meant to protect us. It’s a tool your brain can call upon to help promote behaviors deemed most beneficial to survival of the organism. These were the key topics we covered in Part 2 of this article series.

But viewed through this same evolutionary lens we see another critical role for pain. That of a powerful teacher. 

Again, this links back to survival. Your ancient ancestor needed to be burnt just once to learn not to put his hand into the flame again. Likewise, a painful rash or searing stomach pain after eating a poisonous plant will teach this same ancestor to choose a more desirable flaura the next time he is out gathering or forging for his next meal.

Of course, it is no surprise that past experiences heavily influence our actions and behaviors. Calling on past experiences is a shortcut our brains use to predict potential outcomes of a given situation. The brain uses these predictions to reduce response time when encountering similar stimuli in the future, and to modify behavior in an effort to increase the odds of achieving a desirable result. 

Again, this ties back to the powerful influence pain has in driving behavior. But here we see that this influence occurs not only in the moment pain is felt (like the hand in the flame), but also when it’s remembered (like getting sick from the poisonous plant).

In other words, pain from the past shapes behavior in the future. This is learning. And we now know this plays a major role in pain. Both in terms of how the brain interprets and processes sensory information, as well as how we interact with the world around us.

The role of learning in pain

We usually think of learning in an academic sense or with acquiring certain skill sets. Like learning arithmetic or to tie our shoes. But learning in psychology relates long-term changes in behavior based on experience.

Before going too far I should also point here that behavior is not just an observable action. In the scientific community a behavior is simply a mechanistic response to a given stimulus. So here behavior would include things like being shy at a party or biting your nails when you’re anxious. But behavior also includes the physiological response of sweating or increased muscle tension when you’re scared or faced with a stressful situation.

In psychology, there are two key models that have come to dominate our understanding of how learning shapes behavior. Classical and Operant Conditioning.

Pavlov’s Dogs

The Classical Conditioning model was first put forth in the 1890’s by now famed psychologist Ivan Pavlov. You know, the guy with the dogs. Most of us are familiar with Pavlov’s research where he recognized his dogs would salivate when presented with food. Of course, the salivation was a naturally occurring physiological response. It did not have to be learned. It was an innate response tied to the upcoming task of eating. 

But in his now famous experiments, Pavlov paired that automated response with another stimulus. The sound of a bell. Pavlov rang the bell as the food was presented to the dogs. After doing this a few times he noticed the dogs salivated when they heard the bell, even when no food was given. The dogs learned to associate the ding of the bell with eating. As such, simply hearing the bell triggered the salivation. The bell became the conditioned stimulus and salivation had become the conditioned response. 

This is learning, and it provides a powerful illustration of how the brain loves to create patterns that allow connections between various situations and behavioral responses. As stated above, these patterns allow your brain to more quickly and effectively generate an appropriate behavioral response when encountering the same stimulus. 

This groundbreaking discovery by Pavlov had a major impact on our understanding of how learning takes place. It also has significant implications in our modern understanding of pain. But before we get there, we need to look at another classic psychology experiment. A study that extends the ideas of Pavlov and starts to encapsulate the role of classical conditioning with respect to pain. 

Little Albert

Extending on the ideas first brought forth by Pavlov, John Watson and graduate student Rosalie Rayner conducted another study in the early 1900’s that has become famously known as the “Little Albert Experiment”.

The study followed the behavioral response of a 9-month-old infant whom Watson and Rayner called “Albert B.” When presented with various stimuli, Albert showed no signs of fear or anxiety of any of the the objects he was exposed to. Including a white rate used by the researchers. 

But the next time Albert was presented with the white rat, Watson paired the exposure with the load band of a hammer against a metal pipe. As expected, Albert was startled and frightened by the bang. But interestingly, after repeatedly pairing the bang with the rat Albert began to cry when he saw the white rat. Even in the absence of the bang. As described by Watson and Rayner: “The instant the rat was shown, the baby began to cry. Almost instantly he turned sharply to the left, fell over on [his] left side, raised himself on all fours and began to crawl away so rapidly that he was caught with difficulty before reaching the edge of the table.”

So like Pavlov’s dog, Albert learned to associate the white rat with the bang. But here we see classical conditioning create a phobia. An irrational fear that is out of proportion to the danger posed by the presented stimulus. 

Stimulus Generalization

But even more remarkable, Albert’s phobia didn’t just end with white rats. It extended to other objects that shared similar characteristics. For example, in addition to the rat Little Albert demonstrated a phobia of the family dog, a fur coat, and even a Santa mask with the characteristic white beard and mustache. 

So Little Albert not only shows us that classical conditioning can lead to unhelpful and irrational fears, but that these fears and their conditioned responses can extend to other similar objects or situations. A process that has been termed generalization. It is here where  we can easily see the role of classical conditioning with respect to pain and pain behavior.

Not All Learning is Helpful

Remember, this type of behavioral learning is essentially a shortcut for your brain to help it more quickly respond to a given situation or a similar stimulus.

Think of it this way. The conscious brain represents a neurological bottleneck. It can only attend to so much information at any one moment. So like the CEO of a big fortune 500 company, it must delegate “less critical” tasks to other members of the organization to keep basic day-to-day operations running smoothly. Like so many of your routine habits and behaviors you perform throughout the day, in most cases the CEO isn’t even consciously aware of these ongoing tasks. They simply occur under the radar. 

This has clear advantages. When your ancient ancestor comes across that lion there is no time to stop and ponder his various behavioral options. He just needs to run. Immediately. 

Learning to fear movement

But as adaptive and necessary as this mechanism of learning is, it can also have a dark side. It can lead to the development of inappropriate fears and behavioral responses. Just like Little Albert, the experience of pain can cause some patients to develop phobias of certain movements or tasks. They learn to associate certain activities with pain or the potential for injury even when there is no actual danger.

Think of the individual who experiences an episode of acute low back pain after lifting or bending. Such an episode would likely result in bending to be painful and sensitive. This in turn will cause the patient to be careful with this and related movements for a few days. While this may be an appropriate response in the short term, the danger lies in the patient learning to associate these types of postures and movements with pain and injury. The result is fear and avoidance of these activities. 

Seen through Pavlov’s eyes, bending or lifting becomes the bell, while fear, avoidance and increased pain (fear and anticipation of pain can sensitize the pain system – see Part 2) becomes the conditioned response. The salivation. 

The Behavioral Smoke Detector

Remember modern science holds pain as an output generated by your brain that is meant to protect you. This is what we saw with our discussion of a hyper-sensitive pain system. Remember the smoke detector principle ? Here an over-exaggerated response that turns out to be a false alarm is a better “error” to make than failing to sound the alarm if there was an actual fire. 

Building on this modern view it is easy to see how these conditioning mechanisms and their behavioral consequences can be viewed the same way. As another type of over-exaggerated response meant to protect you. The difference here is the brain is making a decision to avoid a task or situation that it believes may be harmful or damaging.

Said another way, the brain does not trust it can safely navigate a pending activity or movement. So the behavioral response is fear and avoidance. A pre-emptive strike orchestrated by the brain as a strategy to stop the perceived threat before it occurs.  

If the brain does not trust it can safely navigate a pending activity or movement the behavioral response is fear and avoidance. A pre-emptive strike orchestrated by the brain as a strategy to stop the perceived threat before it occurs

Some behaviors matter more than others

Of course, as unhelpful as a dog’s inappropriate salivation or the unwarranted fear of a little white mouse may be, these behaviors are unlikely to significantly interfere with normal day to day existence. But this is not the case with the fear-avoidance beliefs demonstrated with many chronic pain patients. Fear-avoidance related to activities such as bending, lifting, walking, and sitting can have a massive impact on a patient’s life. Inability to participate in such tasks can interfere compromise the ability to earn a living and complete the physical tasks needed for normal daily living.

This is pain-related disability. This is why these theories and mechanisms of learning are so relevant to our modern understanding of pain. And this is why our treatment approaches must account for these beliefs when they are part of the patient presentation. 

But before we get to that, we need to understand one more critical aspect of how learning affects behavior. For this we need to dust off our first year psychology textbook one last time, and review the work of Burrhus Frederic Skinner.

Part 1: Re-thinking The Traditional View of Pain

At first glance pain seems like a straightforward process. You stub your toe or break a bone and it hurts. This is how most people understand pain. It’s consistent with the patho-anatomical view that holds tissue damage or pathology as the source of pain.

Of course, few would dispute that tissue injury can and typically does result in pain, but a big problem occurs when we try to apply this model too broadly and make the assumption that all pain is related to tissue pathology, or alternatively, that all tissue pathology leads to pain. This is simply not supported in the clinical or pain science literature.

It’s Not Always About Injury or Damage

We see countless examples where imaging studies demonstrate tissue damage and pathology such as disc protrusions, annular tears, and arthritic joints in the absence of pain or other aberrant symptoms. More extreme examples can be found in reports in which soldiers experience gunshot wounds in battle yet feel no symptoms.

These are examples in which tissue injury or pathology is present in the absence of symptoms. But we also see many examples where patients experience pain with no discernable tissue damage or injury. Phantom limb pain is an example. Here we have individuals who have lost a limb due to injury or amputation, yet feel what is often described as excruciating and debilitating symptoms in the absent limb.

And of course as clinicians and therapists we have all seen pain that extends long past known healing times for tissue injury. In these cases tissue damage may be the initial instigator of pain, but the patho-anatomical model has trouble explaining why these symptoms remain long after the damage has healed.

Are Faulty Movement Patterns to Blame?

As the role of structural pathology with respect to pain and support for the patho-anatomical model began to wane, the focus shifted from a structural to a functional paradigm. Instead of how things looked on an x-ray or MRI, the focus turned to how things worked. The central tenet here was the idea that many common musculoskeletal conditions and their associated symptoms were the result of poor or suboptimal movement patterns.

Clinicians subscribing to this view (the kinesio-pathological-movement model, or KPM) – hold that improving and correcting these so-called faulty or compensatory movement patterns is a critical and necessary part of care.

This sounds enticing and is an easy sell on the surface, but the KPM also has it’s problems. Perhaps the biggest and most glaring issue is knowing and agreeing on what these so-called “faulty” movements are. Proponents of this model suggest there is an optimal way in which the body is designed to move to prevent injury and maximize performance. But is there actually a best/right way to move? Or is the variability of movement patterns and strategies seen across patients simply reflective of varying adaptive patterns of the complex, dynamic, biological system that is our body?

“Faulty” Movements Don’t Always Lead to Pain

The other elephant in the room is the fact that if these faulty patterns are in fact a cause of pain we would therefore expect to see a predictable relationship between these undesirable patterns and presenting symptoms. But this relationship is a difficult one to establish both in clinical practice as well as in the injury literature. 

Take hyper-pronation for example. The image below depicts what appears to be a horrific case of hyperpronation. But this foot belongs to elite distance runner Joseph Cheptegei. Instead of leading to a life of injury and agony this foot has propelled Cheptegi to an illustrious running career, breaking 3 world records in 2020 alone. 

• 

Of course, this is an anecdotal report. A sample of one from an elite athlete with superhuman abilities. An outlier. Surely these observations do not apply to the general population. What about a more in depth analysis of the literature? What does the research have to say about the relationship between how we move and how we feel?

Let’s look at knee pain as a specific example. 

Biomechanics and Knee Pain?

Patellofemoral pain syndrome (PFPS) is typically thought to be associated with a valgus malalignment of the knee. This “faulty’ alignment is believed to result from inadequate strength or motor activity of the posterior-lateral hip musculature (specifically the gluteal muscle group). So treatment is often focused around correcting knee mechanics by strengthening the hip. 

But is this view supported in the literature?

At first glance it may appear so. With a quick article search we see numerous studies that consistently demonstrate hip weakness in the presence of PFPS (1-7). But does this weakness actually lead to faulty movement at the knee as has been suggested? Here the findings are less clear.

For example, Bolga et al. (2008) found no differences in lower extremity kinematics in subjects with PFPS despite differences in hip strength.

Other studies do report altered kinematics of the lower extremity with hip muscle weakness (2,19), but these changes are not consistent across groups. For example, both Souza et al. (2009) and Nakagawa et al. (2008) found females with hip muscle weakness also demonstrated greater magnitudes of hip internal rotation. However, these same results were not found in male subjects. Furthermore, while Nakagawa reported increases in hip adduction, Souza found no differences in hip adduction between PFPS and control groups. 

So while patients with PFPS may demonstrate hip muscle weakness, exactly how this weakness affects pain and movement is not as clear as conventional beliefs suggest. 

Establishing Cause Vs. Effect

But the tenuous relationship between strength and kinematics is not even the main issue. The more significant elephant in the room is that all these studies compare subjects who currently have PFPS with healthy controls. These cross-sectional designs provide no way to determine causation. They cannot tell us if the weak hip (or the “faulty movement that results) is the cause or the effect of knee pain. 

It’s the chicken versus the egg debate.

We know that pain affects sensory-motor reflex arcs associated with the execution and control of movement. So it is plausible that increased nociceptive drive in the presence of knee pain leads to hip muscle weakness through neurologic motor inhibition. 

Support for this can be found in a study by Park et al. (2017) who studied the effects of experimentally induced knee pain on hip muscle activation. Following injection of a hypertonic solution into the infrapatellar fat pad the authors reported a reduction in gluteus medius activation during a landing task, concluding that experimental knee pain leads to hip muscle inhibition.

Determining Cause Vs Effect

To determine whether gluteal weakness/knee valgus is a cause of PFPS we need to look at prospective/cohort studies. Unlike cross-sectional studies, prospective studies seek to measure variables of interest in a population before they develop a given condition. They seek to determine prospective risk factors for developing a certain condition of interest. 

So if hip muscle weakness is in fact a significant causative factor in PFPS we would expect these prospective studies to bear this out. But this is not what we see. 

Thijis et al. (2011) measured hip strength in healthy female athletes prior to the initiation of a running program. Of the 77 subjects who took part in the study 16 developed PFPS. However there were no differences in initial hip strength between these subjects and those who remained healthy. 

Other prospective studies have not only failed to demonstrate hip weakness as a risk factor in knee pain, but in some cases have actually shown that individuals with greater hip strength are more likely to develop PFPS (10,11). Two recent in-depth systematic reviews also echo these conclusions, suggesting there is little evidence to support the notion that hip strength is a risk factor for PFPS (13,14).

So the commonly held belief that hip strength plays a pivotal role in lower extremity alignment, and that this alignment is a significant contributor to pain and injury at the knee appears to be much less clear than when studied on a deeper level.

Strengthening the Hip Reduces Knee Pain

But here’s where things get really interesting… 

When we look at clinical studies in which the intervention focuses on exercises to strengthen the hip and improve dynamic alignment of the knee, these studies seem to consistently show improvements in PFPS (1-4). 

For example, Tyler er al. (2006) and Khayambashi et al. (2012) studied the effects of exercise programs aimed at improving hip abduction and external rotation in subjects diagnosed with PFPS . Both studies concluded that following the program subjects reported significant reductions in pain. Furthermore, these reduced pain scores were accompanied by increases in hip muscle strength. 

So on the surface it appears that increases in hip strength do in fact lead to improvement in PFPS. But the problem with this conclusion is, again, there is no way to infer causation. Yes we can say the intervention was effective in reducing knee pain, but to say that increases in hip strength lead to improvements in pain is beyond what these studies can tell us. Perhaps it was the other way around. Maybe the reduction in pain was responsible for the improvement in hip function? 

This confusing relationship between hip strength and knee pain is further seen in studies by Dolak et al. (2011) and Nakagawa et al. (2008). Here we again see improvements in knee pain following a hip exercise program. In fact, both studies concluded subjects who performed hip exercises reported far greater improvements compared to subjects who performed exercises that were solely targeted at the knee.

But the interesting finding here is that despite improvements in knee pain, there were no improvements in actual hip strength. So if hip strengthening exercises reduce knee pain but don’t improve hip strength, what are the mechanisms or mitigating factors that are driving improvement?

What About Kinematics?

What about kinematics? Maybe these hip exercises have a bigger effect on the movement pattern. Maybe it is a reduced valgus orientation of the knee that is responsible for the improvements in knee pain?

Here we can look at an article by Saad et al. (2018) who compared the effects of an exercise program aimed at strengthening the hip against exercises aimed at the knee. Here we do see an improvement in hip strength in the hip exercise group that is not seen in the knee group.  But when we look at knee kinematics (measured here as the number of subjects who demonstrated a knee valgus pattern with a step down task) we do see improvements in the hip group, but these improvements are no better than those seen in the knee group. In fact, there was a trend towards greater improvement in the knee exercise group.

So here we see improvements in both pain and kinematics, but these changes are not specific to the hip exercise group or do not seem to be driven by changes in hip strength. 

So again this begs the question, what is driving these changes?  

Re-thinking Pain in Terms of Modern Pain Science

So pain is not a straightforward process. The failures of the patho-anatomical and the KPM models to explain why we hurt clearly demonstrates pain does not follow a straightforward cause-effect relationship. There must be other factors at play that influence the quality, intensity, location, and duration of pain beyond the classical view that pain is synonymous with tissue injury or mechanical overload.

While we still don’t have all the answers, a significant body of work has emerged over the past few decades that has helped provide valuable insight into at least a few of these mysteries. In this article series we are going to explore some of these findings in an effort to provide a more modern and comprehensive view of pain.

To be clear, this does not necessarily mean completely re-working our treatment approach. Stretching and exercise prescription (and yes, manual therapy) still play a valuable role in the treatment of pain. But some changes are needed.

Coming Up Next…

Here in Part 1 we began by discussing the shortcomings of the traditional view of pain. Up next we’ll shift our attention to the critical role the brain plays in pain experience. Including an evolutionary perspective of the importance of pain why our ancient ancestors would need a pain system that is prone to hyper-sensitivity.

Here’s the link to Part 2…

References

1. Tyler et al. (2006). The role of hip muscle function in the treatment of patellofemoral pain syndrome. Am J Sports Med, May.
2. Nakagawa et al (2008). The effect of additional strengthening of hip abductor and lateral rotator muscles in patellofemoral pain syndrome: a randomized controlled pilot study. Clin Rehabil, 2008 Dec;22(12):1051-60.
3. Khayambashi et al. (2012). The effects of isoloated hip abductor and external rotator strength on pain, health status, and hip strength in females with paterllofemoral pain: A randomized controlled trial. JOSPT, 2012; 42 (1).
4. Dolak et al. (2011). Hip strengthening prior to functional exercises reduces pain sooner that quadriceps strengthening in females with patellofemoral pain syndrome: A randomized controlled trial. JOSPT, 2011; 41 (8).
5. Saad et al. (2018). Is hip strengthening the best treatment option for females with patellofemoral pain? A randomized controlled trial of three different types of exercise. Brazilian J Physical Therapy, 2018; 22 (5): 408-416.
6. Bolga et al. (2008). Hip Strength and knee kinematics during stair descent in females with and without patellofemoral pain syndrome. JOSPT, 2008; 38 (1).
7. Robinson and Nee. (2007). Analysis of Hip Strength in Females Seeking Physical Therapy Treatment for Unilateral Patellofemoral Pain Syndrome. JOSPT, 2007; 37 (5): 232-8.
8. Souza et al. (2009). Differences in hip kinematics, muscle strength, and muscle activation between subjects with and without patellofemoral pain. JOSPT, 2009, 39
9. Park et al. (2017). Effects of Experimental Anterior Knee Pain on Muscle Activation During Landing and Jumping Performed at Various Intensities. J Sport Rehabil, 2017 Jan;26(1):78-93
10. Finnoff et al. (2011). Hip strength and knee pain in high school runners: A prospective study. PMR, 2011; 3 (9): 792-801.
11. Herbst et al. (2015). Hip Strength Is Greater in Athletes Who Subsequently Develop Patellofemoral Pain. Am J Sports Med. 2015 November ; 43(11): 2747–2752
12. Thijis et al. (2011). Is hip muscle weakness a predisposing factor for patellofemoral pain in female novice runners? A prospective study. Am J Sports Med, 2011; 39(9): 1877-82.
13. Rathleff et al. (2014). Is hip strength a risk factor for patellofemoral pain? A systemic review and meta-analysis. Br J Sports Med, 2014, 4; 48(14): 1088.
14. Neal et al. (2018). Risk factors for patellofemoral pain: A systematic review and meta-analysis. Br j Sports Med, 2018.