The epistemology of lower-back pain (pt. 2)

Bad backs, healthy minds, and the art of not paying attention.

About nine months ago, I wrote an article about the putative causes of (and possible treatments for) lower back pain. The article was inspired by my own experiences: I was dealing with an intense bout of sciatica, and one way I navigated the pain was by trying to learn more about the condition. I was surprised, and honestly quite moved, by the response to the article—a number of readers reached out directly to share their own experiences, offer advice, or simply express their hopes that I’d recover soon.

Of course, I shouldn’t have been surprised. Lower back pain is extremely prevalent¹, and many people who’ve suffered from some kind of chronic pain feel a sense of recognition when they encounter echoes of what they’ve gone through²; moreover, many people are quite kind and empathetic, so even those who haven’t personally experienced a certain flavor of pain are compelled to reach out in sympathy. Nonetheless, I was both grateful and moved, so I’d like to say thank you to those who reached out or even just read the article.

---

Getting to the bottom of lower back pain is an extremely difficult epistemological challenge.

Pain, after all, is an experiential phenomenon. Even if the causes are physical, it’s hard (perhaps impossible) to establish the precise mechanistic causes of subjective experience. While we can point to potential causes, the thing we’re trying to explain—the pain itself—is a property of an individual’s experience. Here, I think a quote from Elaine Scarry’s The Body in Pain (which I included in a previous post on ineffability) might help illustrate the problem:

Thus when one speaks about “one’s own physical pain” and about “another person’s physical pain”, one might almost appear to be speaking about two wholly distinct orders of events. For the person whose pain it is, it is “effortlessly” grasped (that is, even with the most heroic effort it cannot not be grasped); while for the person outside the sufferer’s body, what is “effortless” is not grasping it (it is easy to remain wholly unaware of its existence; even with effort, one may remain in doubt about its existence or may retain the astonishing freedom of denying its existence; and finally, if with the best effort of sustained attention one successfully apprehends it, the aversiveness of the “it” one apprehends will only be a shadowy fraction of the actual “it”). (Pg. 4)

In the last nine months or so, I’ve read a variety of sources from the literature on back pain, some of which was suggested by readers in email exchanges; I’ve also experienced more ups and downs myself (more on that below). Together, these input streams have gone some way towards constructing a kind of rudimentary mental model of lower back pain—both in terms of an appreciation for the diversity of potential mechanisms underlying the thing itself, and a better understanding of why our epistemology is so fraught on this topic. Change is the only constant, but this post is intended to give a snapshot of that mental model as per October, 2025.

Where I’m at now

Things are, in general, much better but not totally normal. This is typically what I say when asked about my back: compared to January of 2025 (when the pain was at its worst), I’m really grateful for the progress.

These days, I feel functional: I go to work, teach, play with (and take care of) my daughter, and even do some light weight lifting (more on that below). Pain, or the possibility of pain, is sometimes present and occupies some part of my attention some of the time, but to a lesser degree than earlier this year.

Another question I sometimes get is what’s worked for me. I’ll discuss various treatment approaches throughout the course of the article, but part of what makes this so challenging is that time is the great confounder: were the treatments I tried early on less effective than the ones I tried later on, or is it simply that healing follows a slow but inexorable path that would’ve looked exactly the same had the order of operations been reversed? Nevertheless, the routine I’ve settled on is a combination of over-the-counter medications (calibrated to minimize other side effects), careful weight training and mobility work (under the guidance of a great trainer at Kinetic Impact; more on that below), and, honestly, trying not to think about it too much. Other treatments I’ve tried or been prescribed (including acupuncture, massage, physical therapy, believing the pain is only in my head, osteopathic manipulative treatment, dry needling, epidural steroid injections) have, for me, had little to no efficacy. That’s not to say they don’t work for anyone or even most people; back pain is complicated and you can’t generalize from an N of 1!

Of course, I recognize that most people don’t follow this newsletter to read updates about my back pain. I mention all this to provide additional context for what’s below: my limited experience has in some sense run the gamut of theoretical views about the underlying mechanistic causes of back pain, from the purely “bottom-up” approach to what I think of as a kind of modern “mind-cure” movement. I sit now at a tentative synthesis between these views, which may well change.

The bottom-up view

Perhaps the most straightforward account of lower-back pain—and the one I covered in a post nine months ago—is what I think of as the “bottom-up” view. In this account, an injury in the lower back (such as a herniated disc) causes pain signals to be sent from the affected area through the nervous system to the brain, which is where the “experience” of pain takes place. One concrete mechanism by which this could take place is compression of the sciatic nerve (e.g., by a herniated disc).

The reason I call this “bottom-up” is that the brain’s role in the process is primarily as a recipient of neural signals (sometimes called afferent), which have themselves been issued from (roughly³) the site of injury. A further assumption, of course, is that the brain is the seat of consciousness: it’s where these electrochemical signals are somehow (mysteriously, magically) converted into experience. I’m not going to take issue with the latter assumption here, but alternatives to the former assumption are explored in the sections below.

Now, even though the bottom-up mechanisms are relatively straightforward, the prognosis is still complicated. Herniated discs can heal on their own, but it can take a while; moreover, nerve tissue itself can be damaged, and nerves are complicated beasts with unpredictable recovery rates. As I wrote before, doctors may suggest a variety of interventions to facilitate recovery, including: surgery (cutting out the offending disc tissue); epidural steroidal injections (intended to reduce inflammation in the area); or simply emphasizing careful movement patterns and exercises (like McGill’s “Big 3”) that reinforce core stability and avoid further aggravating the tissue.

There’s a lot to like about this bottom-up view: notably, it’s quite parsimonious, and the posited mechanism clearly fits with broader physiological theories. Nothing mysterious—save the mystery of conscious experience itself, but that’s a problem with any theory—need be posited. The view might well be right, and should probably be viewed as the default hypothesis. Nevertheless, there are a few inconvenient observations that complicate the reality of the situation.

A modern mind-cure

Many people who’ve experienced back pain have probably heard of John Sarno. Sarno was a professor and physician who treated a number of patients with back pain. At some point, he came to believe that the cause of the pain for a nontrivial percentage of these people was, effectively, psychological: repressed emotions (especially anger and stress) caused the brain to “generate” the pain (e.g., by decreasing blood flow to certain muscles or nerves), possibly as a way to distract itself from those unpleasant emotions. He called this Tension Myositis Syndrome (or TMS). In Sarno’s view—presented in books like Healing Back Pain (1991) and in other writing—the best treatment for TMS was simply learning about the syndrome and accepting that the pain was, in some sense, “in your head”. Once you learned that the pain was a distraction from these unpleasant emotions, it no longer functioned as effectively as a distraction—so the pain would go away.

To some people, Sarno’s advice sounds like magical thinking. Yet many former sufferers of chronic pain swear by it: in fact, there’s a website called “Thank you, Dr. Sarno” dedicated to testimonials from people who suffered (often for years) from various forms of chronic pain (in their backs, their wrists, their knees, etc.) and who, upon reading the book, experienced a rapid and complete recovery that years of treatment in the medical system could not accomplish. I would advise any skeptical readers to check out some of these testimonials: not because I think it will convince you that Sarno is right, but because I think it’s useful for inculcating some epistemic humility about our ability to understand what’s really going on here.

I’ll be honest: I struggled quite a bit with Sarno’s view when I read his book, and I think that experience is not uncommon.

One common source of resistance comes from a preconception about what, exactly, Sarno is saying. These days, there’s a very understandable (and I think correct) stigma against asserting that pain is “in your head”. Attributing pain to deficits in a person’s psychology has historically been used to dismiss the very real medical complaints of many people. It’s pretty natural, I think, to equate the argument that “the pain is psychological in nature” with the argument that “the pain is not real”. This is aggravating to hear, especially when you’re in pain, and especially when you’ve already been told by a doctor that the cause is a herniated disc (which you have MRI evidence for)!

Of course, these two statements do not mean the same thing: the former is a claim about the underlying cause of an unpleasant experience, while the latter is, essentially, a claim of astonishing doubt (to paraphrase Elaine Scarry) about the pain’s existence. Sarno is saying the former and not the latter. His writing, in fact, is extremely compassionate towards those experiencing back pain, and he is careful to point out that TMS can and does afflict anyone.

It’s also worth noting—and this is me editorializing, not a quote from Sarno—that unless one is a dualist (which some may well be!), one probably thinks the mind is part of the body. Therefore, saying that the pain is “psychological” is, again, a statement about the causal origin of the pain, not necessarily a dismissal of the pain’s reality. This account could thus be nicely contrasted with the “bottom-up” account I described earlier: in the top-down view, the relevant signals originate (somehow) in the brain and are issued towards the muscles and nerves in the lower back (presumably via efferent pathways).

Another source of my resistance, however, is that the mechanism Sarno describes is (in my view) rather vague. The constructs Sarno invokes—e.g., the unconscious mind “needing” to distract itself from repressed anger—are all fairly high-level, and it’s not clear to me how they connect to the proposed mechanism (i.e., the brain “cutting off” blood flow to certain muscles or nerves). Unlike the bottom-up account I described earlier, the causal story doesn’t fit clearly with my understanding of physiology. That doesn’t mean it’s not right! Paradigms shift all the time, and I’d be among the first to say our understanding of physiology (especially as it relates to the mind) is woefully incomplete. But coherence with established mechanisms is still an important factor when it comes to evaluating a new theory.⁴

Still, efficacy is another crucial factor, and the success of Sarno’s work can’t easily be ignored. (Again, I urge skeptics to investigate websites such as Thank You, Dr. Sarno, or any of the many online testimonials by those whom Sarno’s advice has helped.) That’s certainly the most convincing aspect to me.⁵ Moreover, there are other presentations of the “top-down” view that I find more intuitively and mechanistically appealing, such as the idea of central sensitization. I’ll discuss those in a moment, but first, I want to briefly touch on the connection between Sarno’s view and what the psychologist William James called the mind-cure movement.

James associates “mind-cure” with a broader quasi-religious movement of the 19th century, which he dubs the religion of healthy-mindedness.⁶ The religion of healthy-mindedness takes as doctrine that to be human is to be made in God’s image, and thus to be whole and healthy is a natural state of affairs, and thus much of what we experience as “affliction” can be remedied by, more or less, believing (or realizing, depending on one’s perspective) that we are all part of some unified, divine presence—no part of which can really be unhealthy. James writes (bolding mine):

But the most characteristic feature of the mind-cure movement is an inspiration much more direct. The leaders in this faith have had an intuitive belief in the all-saving power of healthy-minded attitudes as such, in the conquering efficacy of courage, hope, and trust, and a correlative contempt for doubt, fear, worry, and all nervously precautionary states of mind. Their belief has in a general way been corroborated by the practical experience of their disciples; and this experience forms to-day a mass imposing in amount. (pg. 94-95)

In The Varieties of Religious Experience, James quotes at length from various individuals who suffered for years from various afflictions (digestive issues, chronic fatigue, aching joints, etc.), and who experienced almost immediate relief once they accepted the gospel of mind-cure (or “New Thought”, as it was called). The parallels to Sarnos’s account of back pain are clear: in both cases, the majority of medical practitioners failed to alleviate the suffering of these individuals; and in both cases, the thing that did alleviate their suffering was the power of belief.

I draw this connection in order to illustrate that this way of thinking is, in fact, quite old. It also helps illustrate the epistemological puzzle at play: towards the end of the chapter on healthy-mindedness, James points out that the mind-cure movement inverts the standard scientific worldview of the time (and, indeed, of the current age), allowing for the possibility that consciousness can, contrary to the dictates of scientific thought, influence the world around us (bolding mine):

Now science, on the other hand, these positivists say, has proved that personality, so far from being an elementary force in nature, is but a passive resultant of the really elementary forces, physical, chemical, physiological, and psycho-physical, which are all impersonal and general in character…Follow out science’s conceptions practically, they will say, the conceptions that ignore personality altogether, and you will always be corroborated. The world is so made that all your expectations will be experientially verified so long, and only so long, as you keep the terms from which you infer them impersonal and universal.

But here we have mind-cure, with her diametrically opposite philosophy, setting up an exactly identical claim. Live as if I were true, she says, and every day will practically prove you right. That the controlling energies of nature are personal, that your own personal thoughts are forces, that the powers of the universe will directly respond to your individual appeals and needs, are propositions which your whole bodily and mental experience will verify. (pg. 119)

And as with Sarno’s advice for back pain, mind-cure enjoyed actual successes; which, as James points out, is its own source of “verification”.

Central sensitization and the role of attention

Above, I mentioned that there are other flavors of the top-down view. These approaches are broadly consonant with Sarno’s presentation, but tend to tone down the role of repressed emotions—instead emphasizing the twin roles of predictive processing and attention.

For instance, the YouTube channel “Pain Free You” (run by Dan Buglio) contains a number of videos describing what Buglio calls Perceived Danger Pain (or “PDP”). The notion of PDP begins with the observation that pain is, in some sense, the body’s response to perceived danger. In most cases, this danger is real, and the pain serves as a useful, often life-saving, signal to avoid some action that could cause harm to the body—this is why congenital analgesia, a genetic insensitivity to pain, can be so dangerous. But in some cases, the brain is effectively “misfiring”; it’s perceiving danger where there is, in fact, none. There are all sorts of reasons why this could happen: maybe bending in precisely this way in the past caused a flare-up, and the body has in some sense “remembered” that association, thus leading the brain to (unhelpfully, in this case) “generate” a pain signal in the absence of real danger.

Rachel Zoffness, a psychologist and professor at Stanford, makes a similar case. Her Pain Management Workbook opens with a “tale of two nails”: two stories that serve to highlight the role of belief—and in particular, predictive processing—in the creation of pain signals. Here’s the first of those stories excerpted from her article on Psychology Today:

In 1995, the British Medical Journal reported on a 29-year-old construction worker who’d suffered an accident: after jumping onto a plank, a 7-inch nail pierced his boot clear through to the other side (Fisher et al, 1995). In terrible pain, he was carted off to the ER and sedated with opioids. When the doctors removed his boot, they discovered a miracle: the nail had passed between his toes without penetrating his skin! There was zero damage to his foot: no blood, no puncture wound, not even a scratch. But make no mistake: despite the absence of injury, the pain was real. What happened?

Zoffness argues (I think convincingly) that this man’s brain, having processed the visual signal of a nail piercing his boot, perceived a threat to his safety, setting off a “cascade of biological and neurochemical processes”. That is, pain was generated as a response to the prediction that this event was causing the man harm.

She follows this up with another tale:

On the flip side, another construction worker (dangerous job, that!) was using a nail gun when it unexpectedly discharged, clocking him in the face (Dimsdale & Dantzer, 2007). Other than a mild toothache and a bruise under his jaw, he thought he’d escaped relatively unscathed. Six days later—six days of eating, sleeping, and going to work—he went to the dentist. Much to his surprise, an X-ray revealed a 4-inch nail that was embedded in his head! Indeed, the nail had pierced his cerebral cortex, putting him in potentially grave danger. However, because contextual cues failed to put his brain on high alert, his pain system remained quiet—despite actual bodily harm and the need for medical intervention (#fail).

She draws several lessons from this pair of stories. The first is that the experience of pain is not 100% reliable as an indicator of tissue damage. To borrow terminology from machine learning, we might cast this in terms of precision and recall: there are cases in which tissue damage can occur in the absence of a pain signal (failed “recall”), and there are cases in which a pain signal can occur in the absence of tissue damage (failed “precision”). This doesn’t entail that pain is a poor indicator of damage—precision and recall might well be high overall—but merely that the connection is not 1-1.

The second lesson is that our experience of pain is very much affected by context: our emotional state, who we’re with, what we’re thinking about. Attention to pain can amplify the signal; attention to something else can “turn down the volume” of pain, so to speak—or even reinterpret the signal as something more positively valenced. This is why Zoffness emphasizes the metaphor of a “pain dial”. Some activities can increase a person’s attention to pain (e.g., lying in bed and thinking about every little twinge), while others can decrease it (e.g., talking with a good friend). Her pain management workbook encourages readers to identify the parameters of their own pain dial: which activities turn up your pain, and which turn it down?

The idea that people can develop a heightened sensitivity to pain, and thus experience pain signals in the absence of direct tissue damage, is sometimes called central sensitization. Central sensitization is one of the proposed mechanisms underlying “nociplastic pain”. Nociplastic pain can be contrasted with nociceptive pain (caused by direct stimulation of pain receptors) and neuropathic pain (caused by some kind of lesion or disorder in the nervous system). We still don’t understand exactly how central sensitization works, but it’s sometimes presented as a form of “dysregulation” in how the brain creates and processes pain signals: somewhere along the line, something has gone haywire in the nervous system, and signals no longer mean quite what they used to. In this way it seems almost akin to the dysregulation observed in autoimmune diseases, where a system evolved to protect the body somehow—for hard-to-understand reasons—responds in ways that, paradoxically, hurt the body.

At this point, a couple questions naturally arise. First, is any of this actually right? And second, if it is right, where do you go from here?

A tentative synthesis

Personally, I think there’s a lot to like about each of the views I’ve presented here. As a scientist, I tend to be attracted to views that emphasize concrete, physical mechanisms; that’s the appeal of the straightforward “bottom-up” account I started with. But I’m also generally predisposed towards the idea that things are rarely as simple as we believe them to be, and that learning how things “really” are is quite challenging—that’s a large part of my ongoing interest in epistemology. That part of me is most compelled by Zoffness’s presentation: namely, that attention and nervous system dysregulation can play some nontrivial role in the experience of back pain. This is what I mean when I say I’ve arrived at a “tentative synthesis” between these views.

Of course, this does not tell us anything about the relative balance of these causes. Is it mostly bottom-up, with some minor role for central sensitization? Is it 50/50? Or is it mostly in our heads? My sense is that the balance varies considerably across individuals, and also varies considerably across time within an individual. For example, a plausible scenario might be that the initial stages of pain are caused by actual tissue damage (say, a herniated disc compressing a nerve); in some individuals, this takes a while to heal, and the body’s adjustments to the pain—both in terms of neural reorganization and in terms of actual changes to posture or gait—can “lock in” the pain, turning it from something acute into something chronic; this is when, according to Zoffness, the signal (pain) is decoupled from the thing it’s taken to indicate (damage).

How do you know which situation you’re in?

You don’t, of course! That’s partly why this is so hard. My current sense is that have to take it gradually, using a process of trial-and-error.⁷ For example, I’ve been working for months now with an excellent rehab trainer (Donald Mull) at Kinetic Impact, and much of that training revolves around incrementally reintroducing movement patterns under safe conditions (with an emphasis on functional movements, such as deadlifts), then observing the effect on the body over the next few days. The logic is so simple it’s profound: if something makes the pain worse, back off; if it doesn’t make it worse, or it makes it better, explore it further. Reintroducing movement patterns can be scary for someone in pain, which is one reason why it’s so helpful to do it under the guidance of a trainer who’s developed hands-on expertise in shepherding people through this process and, importantly, can calibrate their advice to each individual person. It also involves learning how to listen to your body—something that can be difficult if you’ve been experiencing chronic pain.

One perspective I’ve appreciated on the problem of central sensitization comes from Brendan Backstrom, the creator of the YouTube channel (and rehab program) Low-Back Ability. Backstrom argues that chronic back pain comes in part from a vicious cycle: we injure our backs, so we avoid engaging our backs at all; this makes our backs weak and extremely sensitive, which leads to triggering injuries more easily in the future. He suggests that we need to gradually “build evidence” that our backs can work again. Opinions will likely differ on the best way to do this (his approach is to begin with isometric holds on a back extension machine), but the underlying philosophy resonates deeply with me: the brain and the rest of the body needs to be gradually convinced that it can move normally without fear of injury, and that requires striking the right balance between reintroducing movement patterns (as I wrote above) and not pushing through the pain (lest you inadvertently reinforce a particular “story” of injury).

The argument that back pain can sometimes be caused by protective mechanisms in the absence of actual tissue damage can be pretty frustrating to hear, or even hard to believe. Why would the body have evolved mechanisms to cause pain when it doesn’t need to? To be clear, I’m not asserting that this is true, or even that it accounts for the majority of back pain (including my own). But I don’t think the retort that it’s implausible holds water: the body, unfortunately, “misfires” in all sorts of ways. You can think of this as the misapplication of a rule that was formed, at some point, for good reasons—it’s just hard to be at the receiving end of that misapplication.

The epistemological challenge

I started this post with the observation that getting to the root cause of chronic back pain is a challenging epistemological problem. The main reason for this difficulty is that pain is experiential and we don’t have a good theory of how conscious experience arises from physical mechanisms. The body is a very complicated system, and the way in which it relates to the mind is even more complicated.

The bottom-up view of back pain does seem like a good starting point, and I still think there’s a lot to be said for it: it’s entirely possible that the vast majority of lower back pain cases can be explained by the fact that herniated discs sometimes compress the sciatic nerve, and that this, unfortunately, can take quite some time to heal. But it’s also possible that there’s some kind of interaction with how the nervous system responds to this: that the brain does, in some cases, turn up (or down) the “pain dial”, and this can affect recovery. Certainly, this insight seems to help some people, and that’s worth something.

This challenge connects more generally to my interest in the limits of what we can know about ourselves and about the world around us. Sitting as it does at the intersection of the experiential and the physical, pain represents a particularly interesting and important epistemological niche.

I’ll close with what I think is a fitting quote from William James, again from that chapter on healthy-mindedness (bolding mine):

I believe that the claims of the sectarian scientist are, to say the least, premature. The experiences which we have been studying during this hour (and a great many other kinds of religious experiences are like them) plainly show the universe to be a more many-sided affair than any sect, even the scientific sect, allows for. What, in the end, are all our verifications but experiences that agree with more or less isolated systems of ideas (conceptual systems) that our minds have framed? But why in the name of common sense need we assume that only one such system of ideas can be true? The obvious outcome of our total experience is that the world can be handled according to many systems of ideas, and is so handled by different men, and will each time give some characteristic kind of profit, for which he cares, to the handler, while at the same time some other kind of profit has to be omitted or postponed. Science gives to all of us telegraphy, electric lighting, and diagnosis, and succeeds in preventing and curing a certain amount of disease. Religion in the shape of mind-cure gives to some of us serenity, moral poise, and happiness, and prevents certain forms of disease as well as science does, or even better in a certain class of persons. Evidently, then, the science and the religion are both of them genuine keys for unlocking the world’s treasure-house to him who can use either of them practically. Just as evidently neither is exhaustive or exclusive of the other’s simultaneous use. And why, after all, may not the world be so complex as to consist of many interpenetrating spheres of reality, which we can thus approach in alternation by using different conceptions and assuming different attitudes, just as mathematicians handle the same numerical and spatial facts by geometry, by analytical geometry, by algebra, by the calculus, or by quaternions, and each time come out right? (pg. 122-123)

1. According to this article, it’s considered the number one cause of disability globally. Some estimates for lifetime prevalence (i.e., the rate of people experiencing it at least once) are as high as 80%.

2. One need only look as far as the comments in nearly any YouTube video about exercises for dealing with chronic back pain to see evidence for this claim.

3. With sciatica, the pain radiates down the leg, so technically extends beyond the original site of injury.

4. There’s also the problem that, like many Freudian theories, TMS is challenging to falsify. If someone says they tried to follow the “prescription” (i.e., believing their back pain is caused by TMS, not by a disc herniation) but it didn’t affect their back pain, it’s always possible they didn’t really believe sufficiently. Of course, many back pain treatments have mixed efficacy, but at least with something like surgery, we can agree whether the prescription was followed and thus evaluate the outcome as a function of the prescription.

5. Other arguments, like the fact that herniations don’t always cause pain and therefore the pain is not attributable to the herniation, are less convincing: it seems intuitively quite plausible to me that herniations would vary in size and scope, and further, that individuals would vary in anatomy (e.g., the precise location of the sciatic nerve), which seems like more than enough to reconcile the bottom-up view with the observation that many people have herniations without back pain.

6. It’s worth noting that James follows this discussion with an alternative route to conversion, namely the sick soul.

7. Sarno’s recommendation is, effectively, to stop worrying about it and get back as soon as you can to your old activities. But if you’re in the throes of back pain, it’s hard to imagine going out and playing tennis. Moreover, if your pain is related to tissue damage, you might simply injure yourself further.