Placebo Effect

It ends up, the oft-derided placebo effect is actually really damn interesting. Studies consistently show that the belief that one is receiving treatment for pain can, in many cases, produce analgesia even in the absence of a “real” intervention. Although initially discounted as bias (which no doubt can play a role), neuroscience research suggests that the (conscious or unconscious) expectation of relief can activate the body’s innate pain-fighting response, with downstream consequences for nervous, endocrine, and immune system functioning (see [i] and [ii] for reviews). In fact, this anticipatory analgesia appears to influence the effectiveness of real pain interventions as well.[iii] People who are unaware they’re receiving treatment (e.g., opioids) are less likely to feel a benefit than those who are aware of it, and those who believe a treatment is powerful are more likely to improve than those who believe it to be weak or ineffective.

To understand how this works (as far as scientists understand it), a brief primer of the brain may be helpful. From the bottom up, just above the spinal cord is the #brainstem (yellow in the image), which is where a lot of the core reflexes that keep us alive take place (e.g., the connections that mediate increased respiration in response to a drop in oxygen). The #cerebellum (green) looks like a weird cauliflower-esque growth behind the brainstem, and it includes a lot of the connections necessary for movement. Above the brainstem (red) are the nuclei involved in initial processing and filtering of exteroceptive (from the external senses) and interoceptive (from the internal body) information, including the nuclei of the #hypothalamus that monitor and regulate the #autonomic nervous system and the release of key #hormones, the #thalamus which (among other things) acts as a switchboard for information from all of the major sensory organs, and the “limbic lobe,” which is conventionally (but overly-simplistically) described as the seat of emotional processing.

The largest and most exterior part of the brain (and the part that most sets us apart from other animals) is the cortex (blue), which is where most abstract thinking and complex sensory and motor processing gets added to the mix. In general, incoming (afferent) signals undergo greater integration the closer you get to the cortex, meaning more filtering and fine-tuning based on input from other sources. But all of the reciprocal connections, short cuts, and emergency override pathways make it far from a neatly hierarchical process. Still, it can be useful to know the basic structure when (foolishly) trying to make sense of the plethora of (imperfect) brain imaging studies that attempt to map cognitive and behavioral functions onto neural circuitry.

In general, research investigating the placebo effect (and expectations-based analgesia more broadly), suggests that one of the major pathways involved in the response starts in a region of the cortex known as the dorsolateral prefrontal cortex (dlPFC). The prefrontal cortex (PFC) is, not surprisingly, located at the front of the cortex. The vast neural connections within it are implicated in a broad array of “executive functioning,” which is the umbrella term used to refer to the cognitive processes necessary to get shit done (goal-directed behavior). This includes things like selectively focusing on the task at hand, maintaining that focus, suppressing internal and external distractions, and inhibiting impulsive or maladaptive behavior.[iv]

One of the prototypical research tasks that relies on the #PFC is the Stroop test. This (rather tedious) task involves a computer screen flashing the names of colors, sometimes with a matching font color (e.g., RED) but more often with a mismatch (e.g., RED). When asked to report the color of the font in these mismatch cases, it’s really damn hard to ignore the written color name. It just is. We process RED = red faster than RED = blue. Our ability to overcome the automatic response and report the color of the font (i.e., to learn the rules of the game and apply them) seems to rely on the #dlPFC. At least, the dlPFC tends to light up in imaging studies when people do this kind of task, and people with PFC damage seem to be way worse at it and often can’t do it at all.

Successfully suppressing an automatic or habitual response in the service of a broader “goal” or rule appears to rely on three main factors. First, the goal has to be strong enough to overrule competing drives. (If you don’t care about getting the correct answers on the silly neuroscience task, then odds are your dlPFC will be minimally taxed and you will perform poorly.) Next, the conditions that result in achieving said goal have to be adequately learned (i.e., that “pay attention to text color / ignore word meaning” = correct). Finally, those conditions have to be implemented successfully to bring about the desired result (i.e., you need to selectively attend to input about color and ignore distracting input re: word meaning). It’s this last step that’s particularly interesting. In order to achieve one’s goal of acing the Stroop task, the dlPFC seems to tweak processing in other brain regions in accordance with the goal, essentially turning the volume down on areas that process semantic meaning and turning it up on areas involved in processing color.

A similar mechanism seems to be at the root of the placebo effect. It ends up, expectations about positive outcomes for a treatment (happy, pain free Julie) are remarkably similar to a desire to respond accurately on a task (happy, overachiever Julie). In both cases, just wanting to be happy Julie is unlikely to be enough, you also need to know (and believe) the rules of the game (paying attention to color will yield the correct answer, getting treatment X will provide pain relief). Once you’ve internalized the rules, you may start to bias processing resources in accordance with them to facilitate the outcome, inhibiting activity that’s unrelated to the end goal and augmenting that which is. In the case of overachiever Julie, that involves altering visual attention and processing to facilitate RED = blue. In the case of pain free Julie, that involves altering attention and processing of pain signals to facilitate treatment X = pain relief. In both cases, the likelihood that the dlPFC will tweak processing depends on keeping the goal in mind and adequately internalizing the conditions necessary to achieve the outcome.

The dlPFC seems to be particularly important for maintaining the goal or rule in “mind” (treatment X = analgesia, treatment X = analgesia, treatment X = analgesia). It’s tightly intertwined with the orbitofrontal cortex (OFC), which is the lowest part of the PFC located above the eyeballs (the “orbits”). The #OFC seems to be important in encoding the expected “hedonic value” of a given stimulus or situation,[v] which is basically sciencespeak for how much something will make you feel good (or bad). So yeah, OFC tells the dlPFC that a treatment is likely to provide relief. The dlPFC keeps that information in “mind” in the face of competing information, and voila, you have the ground work for the placebo effect. (I’m way oversimplifying… there are bajillions of neurons, and even more connections between them, but this seems to be the basic idea.)

Actually making the analgesia happen requires some dlPFC-mediated tweakage of activity in pain-centric areas of the brain. This appears to occur in large part through dlPFC communication with the anterior cingulate cortex (ACC), an area of the brain pretty much always implicated in processing discrepancies between expectations and experience and, as I nerd out about here, involved in encoding the unpleasantness of pain. The #ACC lies between the outer cortex and the limbic core of the brain, and it frequently acts as a sort of bridge between ascending sensory and #homeostatic information (bottom-up processing) and descending cognitive signals (top-down processing), making it well situated to play a role in expectancy-based modulation of pain.[vi] Moreover, it’s highly intertwined with all of the other regions of the brain implicated in pain perception including the #insula (internal well-being), the #amygdala (emotional significance), and the periaqueductal gray #PAG (innate defensive responses, including pain control).

The placebo effect appears to result, at least in part, when the expectations processed in the dlPFC and ACC prompt the release of endogenous #opioids in these pain processing areas.[vii] In fact, opioid antagonists (i.e., drugs that block opioids from binding with opioid receptors and doing their thing) can prevent the placebo response from occurring, which seems a fairly good indication that endogenous opioids are a big part of the process.

Essentially, in anticipation of pending relief, the dlPFC and ACC appear able to facilitate the process by activating our innate pain fighting response, which (like narcotics) decreases the processing and encoding of pain signals.[viii, ix] Since the PAG is the main control center for the descending pain pathway, the placebo prompted release of opioids can even alter pain-related activity in the spine, thereby influencing ascending pain signals even before they reach the brain.[x]

Despite being very cool, it’s unfortunately pretty damn hard to will oneself into a placebo effect. For example, all the research evidence in the world may not be enough for your brain to expect pain relief if prior experience has suggested otherwise. This is at least partially due to the fact that the areas of the prefrontal cortex that process expectations about treatment efficacy do so based on input from a variety of “lower-order” neural connections that can (often unbeknownst to us) profoundly shape our beliefs. For example, among other inputs, the dlPFC and OFC are connected both directly and indirectly to the #dopamine -releasing neurons of the brainstem’s ventral tegmental area #VTA . Dopamine release from the VTA is involved in “encouraging” us to take action to achieve a goal (which you can read more about here). It increases when contextual cues (even ones that are not consciously perceived) suggest a good thing is a-comin’. And when you’re in pain, analgesia definitely qualifies. So if prior experience has linked a doctor to pain relief, then subsequent treatments by that doctor may be more likely to prime the OFC to expect analgesia, and therefore to actually facilitate it, all without our conscious knowledge.

Treatments that seem particularly high-tech or that involve impressive rituals or are somewhat invasive are more likely to result in a placebo effect for this reason, too.[xviii] When we feel we are being well taken care of, our brain biases processing and prompts endogenous activity that helps to ensure this to be true. Prominent cues linked to healing, constant reminders of care, belief in a doctor's authority and expertise, and convincing demonstrations of a treatments' effectiveness all increase the likelihood of a placebo boost. In fact, as I sifted through the literature on this topic, I was amazed by how closely the factors that increase the likelihood of it occurring align with the factors social psychologists have identified as key to persuasion. Basically, if something convinces your unconscious mind that a treatment is going to work, there's a decent chance it's going to facilitate changes that help to make sure it does.

Unfortunately, the reverse is also true, meaning (even non-conscious) negative associations between contextual cues and treatment (e.g., based on past treatment failures) can influence the processing involved in making predictions about treatment efficacy, and therefore the likelihood that the endogenous opioid system will be activated. In fact, negative expectations can even worsen pain outcomes, a phenomenon scientists cleverly call the "nocebo" effect.

But even setting aside expectations, other factors influence whether and to what extent an individual gets a boost from the placebo effect.  This includes neurobiological differences (e.g., variations in mu-opioid receptor density or dopamine turnover) as well as contextual ones. For example, in addition to its strong bidirectional links with the amygdala (emotional salience) and insula (well-being), the ACC is highly interconnected with regions of the brain involved in monitoring homeostatic needs and the #stress response (hypothalamus) and in regulating reward and #motivation (striatum). These and related connections provide ample opportunity for psychological and emotional factors (e.g., #anxiety, stress, and despair)[xi] to influence expectancy-induced analgesia in complicated and not entirely understood ways.[xii, xiii]

So while pain perception is super weird and definitely modifiable by a variety of factors “in your head” (which, after all, is where pain is located too), it is far from a straightforward process. This means that, unfortunately, even true believers in a treatment may not get an analgesic boost from simply thinking it will work. (But it does mean there may be ways to work with our brain to find pain relief more effectively.)