“What in heaven’s name is going on?” It was a question that nagged at me every night. (For reasons I never did understand, my neck pain always got worse around midnight, just when I wanted to go to sleep.) The pain made no sense. I had had no injury. No car wreck. Yeah, I worked all day on a laptop in ergonomically disastrous postures, but so does everybody else. Physically, I had never been stronger. I had taken up competitive US Masters swimming in my late 50s and turned out to be pretty good at it, qualifying regularly for Nationals—and even the World Championships—in several events.
I was also happier than I had been in years. After suffering through my late husband’s 11-year battle with two kinds of cancer, I did a lot of heavy grieving after he died, then pulled myself together and signed up for Match.com. There, to my astonishment and delight, I met a wonderful new man. That relationship was going well (we are married now). Work was going well, too. So I could not imagine what was wrong with my neck. Why did it always hurt on the left, never the right? Why did I get a burning, shooting pain in a straight line from C-7, the lowest cervical (neck) joint, down to my shoulder? Why did my muscles spasm so much my head would get locked into a crooked position (cervical dystonia) for long periods of time?
When this sudden immersion into the world of chronic pain began five years ago, I set out on a journey to understand my own pain and to unravel the secrets of pain in general. It has been a fascinating trip—with a happy ending, at least so far—that has taken me to the outer frontiers of molecular biology and the inner reaches of the brain and mind. In this chapter, I will share the highlights of what I have learned so you can better understand your own pain, while sparing you, hopefully, excessive details and impenetrable jargon. I’ll describe briefly the different kinds of chronic pain and how the nervous system works, particularly the ways that pain signals travel in the body up to the brain—where we actually “feel” pain—and the ways nerves running back down to the body help damp pain down.
Most important, we’ll talk about crucial new insights into chronic pain. That it physically changes the nervous system, literally causing the loss of brain tissue. That chronic pain (pain lasting three months or more) is not just a symptom of something else, but is often a disease of the nervous system in its own right. That it’s not just the nervous system that gets revved up in chronic pain, but some parts of the immune system, too—specifically, microglial cells. That, with luck, brain scanning technology may soon allow doctors to “prove” that a person is in pain, a vast improvement over just listening to (and often, not believing) his or her story.
For the record, throughout this book I will use the official definition of pain: “an unpleasant sensory and emotional experience associated with actual or potential tissue damage, or described in terms of such damage.”1 This comes from the International Association for the Study of Pain (IASP), the world’s top pain research group. The IASP adds— and this is crucial—that “the inability to communicate verbally does not negate the possibility that an individual is experiencing pain and is in need of appropriate pain-relieving treatment.”
To my surprise, I learned that in many pain syndromes, an obvious, physical injury or cause of chronic pain can never be found, a puzzle that is a source of enormous frustration to people with pain and doctors alike. Doctors call these real, but mysterious, conditions functional pain syndromes. In other words, although pain often is triggered by an initial, sensory experience, that isn’t always the case. The acute pain of labor, for instance, has an obvious cause; the chronic pain of fibromyalgia, which is associated with changes in the central nervous system, often does not.2
Legendary Canadian pain researcher Ronald Melzack was one of the first to realize how wickedly complex chronic pain is. He developed the neuromatrix theory, the idea that pain is the result of output from a widely distributed network of nerves—not a straightforward stimulus-response to sensory input.3 Nowadays, researchers know it’s even more complicated than that. Chronic pain is now seen as a biopsychosocial phenomenon, not just a linear response to an unpleasant stimulus, but an ever-changing, almost living thing, an intricately interwoven tapestry that includes not just physical sensations but also our emotional responses to pain and our emotional responses to other people’s emotional responses to our pain. No wonder it’s so hard to understand someone else’s pain. Given this complexity, it’s not surprising that different regions of the brain.—the prefrontal cortex, amygdala, hypothalamus, hippocampus, primary and secondary somatosensory cortex, to name just a few—all contribute to the overall experience of pain.4 Indeed, precisely how active each given region is at any moment can change dramatically, as high-tech, moment-by-moment images of the brain show.
Things can be so malleable, in fact, that some people with chronic pain are able to look at real-time brain scan images of their own brains in pain and learn, via biofeedback, to change their perception of pain. For the better.5 But usually, it’s not that easy. Author Melanie Thernstrom captured this beautifully in her 2010 book, The Pain Chronicles.6
“To be in physical pain,” she wrote, “is to find yourself in a different realm—a state of being unlike any other, a magic mountain as far removed from the familiar world as a dreamscape.” And the longer the pain persists, she continued, “the more excruciating the exile becomes. Will you ever go home? you begin to wonder, home to your normal body, thoughts, life?”
Harvard professor Elaine Scarry put it eloquently in her 1985 book, too. “To have great pain,” she wrote, “is to have certainty; to hear that another person has pain is to have doubt. (The doubt of other persons, here as elsewhere, amplifies the suffering of those already in pain).”7
One of my hopes for this book is that, with the growing understanding of pain, some of those doubts—and some of that suffering—can be dispelled.
Chronic pain comes in four basic “flavors”—nociceptive, inflammatory, dysfunctional, and neuropathic, which describe different ways of understanding how pain signals travel in the nervous system. Sometimes, chronic pain can be a combination of these types. Cancer pain, for instance, is often a combination of nociceptive, inflammatory, and neuropathic, depending on how big the tumor is and where it lodges. Historically, scientists thought about pain in terms of its etiology—that is, what caused it: an injury, a tumor, an infection, and so on. Increasingly, they are turning to a different method, says Clifford Woolf, a pain researcher at Boston Children’s Hospital.8 It’s the pain phenotype that counts, he says, particularly when it comes to classifying the various subtypes of neuropathic pain. The pain phenotype refers to the unique collection of symptoms of each person with pain; the idea then is to link each phenotype to specific, underlying neural mechanisms.9, 10
So here we go.
Nociceptive pain is that instant, intense reaction you get when you hit your thumb hard with a hammer or cut your finger with a knife. The word nociception simply means the perception of a noxious, or unpleasant, stimulus. Compared to other types of pain, nociceptive pain is relatively simple—essentially an on–off switch. To trigger nociceptive pain, it takes a pretty big (that is, a high-intensity) wallop. If you just touch your hand lightly with a hammer, you won’t feel pain. The usual stimuli for nociceptive pain are mechanical forces, like that hammer, too much heat or cold, and chemicals, including acids.11, 12 Less obviously, nociceptive pain also happens when some part of the body is subject to unusual mechanical force, like the grinding of bone-on-bone in osteoarthritis of the knee, or injury to an organ like the heart, when it is deprived of oxygen. The way neurologists see it, the bone grinding in osteoarthritis and the oxygen deprivation in the heart are pretty similar to the external stimulus of hitting yourself with a hammer in the sense that they don’t come from any problem intrinsic to the nervous system itself.
Nociceptive pain is basically “good,” or adaptive, because it serves an important, biological purpose: It alerts you to danger and lets you know about your now-damaged tissue, which motivates you to be gentle with the damaged area until it heals. It’s comparatively straightforward, too. You thwack your thumb with the hammer, pain signals travel along a nerve, or, more often, a number of nerves, from your thumb to the spinal cord, get handed off to another set of nerves that run up through the brainstem to the cortex, or thinking part of the brain. End of story. When your thumb heals, it’s all over—pain nerves quiet down and you forget all about it.
Inflammatory pain is a bit different. While it takes a big wallop to trigger nociceptive pain, both big wallops and smaller ones can trigger inflammatory pain; then, once that sore thumb swells up and gets all red, even the slightest touch can hurt. Inflammatory pain persists as long as the tissue remains damaged and swollen. In a sense, inflammatory pain, like nociceptive pain, is somewhat “good,” or adaptive, in the sense that inflammation does good things, including the secretion of natural chemicals called cytokines that act to promote healing. But inflammatory pain can also be “bad,” or nonadaptive, because pro-inflammatory cytokines amplify, or rev up, the transmission of pain signals along nerves.13 Indeed, with inflammatory pain, nerves rev up in both the central nervous system (the brain and spinal cord) and in the periphery—our arms, legs, and everything else—becoming overactive and hypersensitive, to the point that you can end up with more pain than you started with.
The third type of pain, dysfunctional pain, is really nasty, and, as its name implies, it serves no purpose at all—it’s totally useless, maladaptive, and devoid of any redeeming virtues. In dysfunctional pain conditions like fibromyalgia, irritable bowel syndrome, and some types of headache, pain can be triggered without any external pain stimulus at all, at least as far as scientists can tell. With dysfunctional pain, there’s no damage to the nervous system itself and no inflammation. As with inflammatory pain, there is sensory amplification, or revving up, of pain signals in both the periphery and central nervous systems.14
Worst of all, in many ways, is the fourth type of pain, neuropathic pain, the most complicated of the pain problems.15, 16 Neuropathic pain is caused by damage to the nervous system itself. In a sense, neuropathic pain is to the nervous system what AIDS is to the immune system. The AIDS virus attacks the immune system, the very system that is supposed to deal with viruses. Neuropathic pain attacks the nervous system itself by altering the way nerves function—an insult to the very system that is supposed to deal with pain. This, of courses, makes treatment extraordinarily tricky. While nociceptive pain, that simple kind, is all over when it’s over, neuropathic pain goes on and on, long after the initial trigger, if there was an obvious one, is history.
Like dysfunctional pain, neuropathic pain can occur without any obvious external pain stimulus, and it, too, gets amplified and revved up in both the peripheral and the central nervous systems. Neuropathic pain occurs in conjunction with many diseases and different types of damage to the nervous system, including trauma to nerves (as in surgery), pressure on nerves (as from a herniated disc in the neck or back), injury from toxic chemicals landing on nerves (as in chemotherapy), infection from neurotropic viruses like herpes zoster, and so on. A truly dramatic demonstration of the differences between neuropathic and inflammatory pain was reported in late spring 2012 by University of California, San Francisco pain researchers Allan Basbaum and Joao Braz.17 Their team transplanted fetal cells that make gamma-aminobutyric acid (GABA), a chemical pain inhibitor, into mice with neuropathic pain. The cells dampened neuropathic pain but did not affect inflammatory pain.
Researchers and physicians use a lot of jargon when they talk about the nervous system, but the concepts are fairly simple. The central nervous system consists of the brain and spinal cord and is chock full of nerve cells. Interestingly, scientists have recently discovered that other cells called microglial cells—which are part of the immune system—also live amidst the nerve cells and contribute mightily to the processing of pain signals. (In fact, the overlap between the nervous and immune systems in pain processing is so important I’ve given it a whole separate chapter—Chapter 9.) Here’s the snapshot version of how nervous system works: Electrical signals travel along sensory nerves from the periphery—our limbs—to a specific structure in the spinal cord called the dorsal horn. There, the nerves release chemicals that communicate with a second nerve cell that runs all the way to the brain, where a third nerve cell takes the stimulatory message to various regions of the brain where we feel what we interpret as pain.
Perhaps the most impressive thing about nerve cells is their sheer number. In the adult human brain alone, Harvard Medical School neuroscientist Gary Brenner told me, there are an estimated 100 billion neurons with perhaps 100 trillion connections.18 Put differently, there are more brain cells at a dinner for 10 people than all the stars in the Milky Way galaxy.19 And then there’s the peripheral nervous system— all the nerves outside the brain and spinal cord.
The peripheral nervous system is divided into the somatic nervous system, which means the nerves just under the skin, and the autonomic nervous system, nerves everywhere else. The autonomic nervous system is further divided into two branches: sympathetic and parasympathetic. (Scientists sometimes also refer to a third division, the enteric nervous system, which controls the gastrointestinal tract and is often called “the second brain” because it can operate on its own and has direct connections to the central nervous system.) The sympathetic system is in charge of the famous fight-or-flight response, the body’s instantaneous reaction to stress via the hormone adrenaline. Well-known signs of the sympathetic response are increased heart rate and blood pressure, constricted blood vessels, and sweating. The parasympathetic system is the quieter partner, dominating things while the body is at rest. Normally, the two systems act in concert to produce homeostasis.
The raison d’être of the entire nervous system, of course, is to convey information from the outside world to the brain or from one part of the body to the brain, all of which is accomplished by translating a huge variety of incoming information into bite-sized electrical and chemical signals. Nerve cells, also called neurons, are electrically excitable and come in three basic types: sensory neurons, which respond to light, touch, sound, chemicals, and other stimuli and are the most important nerve cells for pain; motor neurons, which act on signals from the brain and spinal cord to make muscles contract; and interneurons, neurons in the central nervous system that “talk” only to other nearby neurons. Scientists also use the word nociceptor—this just means a subcategory of sensory nerves that responds to the noxious, tissue-damaging stimuli that cause pain. Each nerve cell consists of a cell body, where the nucleus of the cell resides, one axon, and a bunch of filaments called dendrites that look like the messy split ends of a strand of hair and function as receiving centers for incoming information.
Axons are incredibly long filaments that can extend a meter or more. When bundled together into fibers and further bundled into cables, they’re called nerves. The sciatic nerve, the longest nerve in the body (and famous for the leg pain it causes), has one long axon that runs from the base of the spine down to the foot. Interestingly, nerve fibers carry information at different speeds. The so-called A-delta fibers, which are relatively large in diameter and are covered with myelin, convey information very fast and are major players in pain transmission. C fibers, which are skinnier, conduct pain information 15 times more slowly because they are not coated with myelin. A-beta fibers, which carry information from the skin about touch and pressure but not pain, are the biggest and fastest fibers. But the most amazing thing about these nerves is how specialized they are for receiving different types of information—from mechanical, chemical, thermal, or other stimuli.20, 21 In fact, the specialization of these nerves is so fine-tuned that there are different thermal receptors for responding to fairly small differences in temperature.
For instance, a receptor called TRPV1, located on the tip of certain nerves, is programmed to detect heat above 109 degrees Fahrenheit. It also responds to “hot” substances like capsaicin, the ingredient in chili peppers, and to acid, as well as to an endogenous fatty substance called anandamide, which is somewhat like marijuana. TRPV2, a closely related receptor, detects hotter temperatures—above 126 degrees Fahrenheit. Still another receptor, TRPA1, detects cold temperatures—less than 63 degrees Fahrenheit—and responds to mustard oil and wasabi as well. (Not surprisingly, targeting drugs to these receptors, and fiddling with the genes that make the receptors, is a hot area of research.22) Acid-sensing receptors are also exquisitely sensitive. By studying the highly painful and toxic venom of the Texas coral snake, researchers have shown that acid-sensing channels in sensory nerves play a much larger role in pain than previously thought.23, 24
Once all this this incoming sensory information is collected by the receptors, the next task of the nervous system is to transport it to the brain, a problem for which evolution has come up with a remarkably elegant solution: pass information along a nerve cell electrically and between one nerve and the next, chemically.
Incoming information enters the system through dendrites (those “split ends”) and exits at the other end of the nerve cell through an axon. When information reaches this far end of the axon, the axon releases chemical messengers that float into a narrow gap called a synapse. In the synapse, waiting dendrites from the next nerve cell in the chain pick up the chemical signal, convert it to an electrical message that travels through the second nerve to its axon, which in turn pumps out its chemical messengers into the synapse for the dendrites from the third nerve in the chain to pick up, and so on. Electrically, the passage of information works through impulses called action potentials that act like on–off switches, giving a nerve a simple message: to fire or not. It is an all-or-nothing, yes–no, binary system.
Like any other cell, nerve cells are encased in a two-layer, insulating membrane. It is in this fatty membrane that the receptors, also known as ion channels, lie. Some ion channels are voltage-gated, which means they are activated by electrically charged ions that flow back and forth across the membrane in and out of the cell. Others are chemically gated, which means they are activated by chemicals—neurotransmitters— floating around outside the cell. In its normal, resting state, the interior of the nerve cell has a negative charge compared to the outside. It’s a big difference. There are 10 times more sodium ions outside the cell than inside. (With other charged particles, such as potassium, it’s the other way around—the concentration of potassium ions is 20 times higher inside the cell than outside.)
When the nerve receives a signal—say, mechanical pressure on your thumb or an acid on an acid-sensitive dendrite—sodium from the outside rushes into the cell through sodium channels. The sudden influx of sodium depolarizes, or reverses, the electrical charge. Now the outside of the cell has the negative charge and the inside, the positive. The cell “thinks” of this as an unnatural state and immediately seeks to correct it by having potassium ions rush outside. Once this happens, the electrical balance returns to normal.
Sodium channels, in other words, act as molecular amplifiers, turning small electrical signals into action potentials that can conduct for long distances along an axon.25 The short-lived change in the electrical charge—just milliseconds long—is passed step by step from one little section of the axon to the next. In many axons, these sections are covered with myelin, but there are tiny gaps (called nodes) between the sections, and it’s actually in these gaps that the sodium channels lie. So when sodium rushes in through these channels, the change in electrical charge “jumps” across the gap with enough electrical energy to depolarize the next section of the axon, then the next and the next, all along the axon. In technical jargon, this is called a wave of depolarization. (In nerve cells that don’t have a myelin sheath, there’s no gap for energy to jump across, so transmission of the signal is slower.) With potassium channels, researchers have recently shown that two channels in particular (called K2P) may be especially important in pain sensation and chronic pain.26, 27 And a new family of ion channels that respond to mechanical force also appears to play a role in pain—and touch—sensation.28, 29, 30 Specific proteins convert mechanical stimuli into electrical signals.
When the wave of depolarization reaches the end of the axon, that’s when the mode of information transfer changes from electrical to chemical, with the release of a chemical—a neurotransmitter—that lands on receptors in the dendrites on the other side of the synapse to keep the signal going, nerve after nerve. But the chemical—neurotransmitter— signals can carry different messages. If the neurotransmitter is excitatory, as happens with the chemical glutamate, the result is an increase in nerve activity and, ultimately, continued transmission of the pain signal. (Obviously, one way to stop transmission of a pain signal would be to block the glutamate receptor; there are drugs, such as the anesthetic ketamine, that can do this. But these drugs have too many side effects for chronic, widespread use.) On the other hand, if the neurotransmitter is inhibitory, or calming, like the neurotransmitter GABA, the result is a dampening of pain.
Granted, this is pretty esoteric stuff, but it matters. Think of this— with gratitude—the next time you go to the dentist. It’s only because scientists have unraveled these basic mechanisms that they were able to come up with excellent local anesthetics such as lidocaine that work by blocking sodium channels. With sodium channels blocked, the wave of depolarization never happens and pain signals—from a dentist’s drill— never get going.
But the story, of course, doesn’t end here. It ends in the brain. In order to actually “feel” pain, signals must travel upward from the dorsal horn in the spinal cord to the brain, specifically, to the brainstem and also to the thalamus (the spinothalamic tract) and finally to the rest of the brain, where perception of the pain finally occurs.
As pain signals pass through the spinal cord, cells in the dorsal horn act like a gate, either sending the signal straight on up or modifying it. Some designated nerve cells (the thick, myelinated ones that detect touch as opposed to pain) also kick in at this point. This is a good thing because it means that if you thwack your thumb with a hammer and then rub your thumb hard, the touch nerves can overpower the pain nerves, somewhat decreasing pain. Once pain signals get to the brainstem, the brainstem begins sending electrochemical signals downward (called descending modulation) to try to block incoming pain signals.31 In fact, some of the drugs that help control pain—anticonvulsants, opioids, and antidepressants—work in part by enhancing this descending modulation. In the dorsal horn, that all-important relay station in the spinal cord, the downward-moving, pain-blocking signals do help somewhat, and they act in several ways. One is by slowing the release of pain signals coming in from the periphery. Another is by blocking the receptors into which the chemical pain signals land in nerve cells across the synapse.
But even as the brainstem “tries” its best to dampen pain, some pain signals manage to keep going upward, eventually reaching the thalamus. Here, they get passed on to three main areas of the brain: the somatosensory cortex in the parietal lobe, which tries to figure out where in the body the pain is coming from; the limbic system, which adds emotional importance to the pain; and the frontal cortex, the part of the brain behind the forehead that governs thinking and gives meaning to the pain. (And meaning really counts. For instance, severely injured soldiers, brimming with feelings of heroism and nobility, have long been known to report significantly less pain than similarly injured civilians, whose pain has no lofty meaning.32 Similarly, childbirth hurts like hell— but we see it as beautiful, at least afterward, because we get a lovely, new baby for our efforts.)
The somatosensory cortex—and a fascinating little site inside it called the homunculus (Latin for “little man”)—have been the site of some of the most profound discoveries in science.33 In the 1940s, pioneering neurosurgeon Wilder Penfield operated on epileptic patients who were awake and could talk. (The brain itself does not have receptors to detect noxious stimuli, so an acute injury to the brain, such as surgery, is not painful.34) He applied mild electrical currents to different areas on the surface of the brain and asked patients where in their bodies they felt a tingling or movement. From this information, he was able to create a map that showed where sensations from different parts of the body are processed in the somatosensory cortex.35, 36 The homunculus is a grotesque-looking but kind of adorable little thing, essentially a very distorted map of the body.
The first thing you notice about this map is that a hugely disproportionate amount of brain tissue is devoted to sensations coming from the mouth, tongue, lips, face, and hand. Obviously, this suggests that information coming from these areas is extremely important for survival. Like other parts of the nervous system, the homunculus is also quite plastic, or changeable. For instance, the homunculus hand of a concert pianist would look quite different from that of a newborn baby. Rats have a ratunculus, and the rat’s little body map exaggerates information coming from the whiskers. The homunculus also plays a role in the mysterious problem of phantom limb pain. Many, though not all, people who by accident or surgery, wind up with an amputated arm, leg, breast, or other body part often feel pain in the missing body part, and sometimes it’s excruciating. Someone with a missing arm, for instance, may feel as though the fist on that limb is tightly clenched, with the fingernails digging painfully into the palm. For years, as often happens with things doctors can’t readily explain, it was assumed that phantom pain was psychological.
But it’s not psychological, as some ingenious experiments have shown. For instance, in the homunculus, the area that maps sensations coming in from the face is located very close to the area for incoming information from the hands, as University of California, San Diego neurologist V. S. Ramachandran has dramatically demonstrated with a person whose left arm was lost in a car crash. As an experiment, Ramachandran touched the patient’s cheek with a Q-tip, then asked him what he felt. The man said he felt his cheek being touched—but his phantom thumb as well. Perhaps, speculated Ramachandran, the somatosensory cortex noticed that it was not getting any more information from the left arm, because it was missing, and somehow the space in the homunculus that was allocated to the arm got taken over by nerves from the face.37
Ramachandran figured out a way to help some people with phantom limb pain. Using a box with two armholes cut in the side and a mirror placed inside, he would have the person place his good arm and his stump through the holes, then look inside. What the person “saw,” because of the mirror, was the illusion that he had both arms. He then asked the person to clench and unclench his good fist. Because of the mirror, the person actually “saw” both fists clenching and unclenching. By unclenching the “good” fist, it felt to him as if both fists opened, which, over time, relieved the pain in the phantom.
In a sizable percentage of (unfortunate) people, a series of unhappy events conspires to turn acute pain into chronic pain. In fact, one of the puzzles in pain research today is to figure out why this happens in some people and not others, a question that falls to scientists who study pain genes. (See Chapter 3 on pain genetics.)
Chronic pain—usually defined as pain lasting three months or longer—is not just acute pain that doesn’t go away. It is the result of fundamental changes in the nervous system itself. When pain becomes chronic, it is no longer just a symptom of something else, but can become a disease in its own right. It quite literally changes the brain, in some cases causing a loss of gray matter equivalent to 20 years of aging, as Northwestern University neuroscientist Vania Apkarian showed dramatically with brain scans in 2004.38
So, how does this change occur? The transformation of pain involves neural plasticity, or the changeability of the nervous system, and sensitization, which means nerve cells become more and more responsive to weaker and weaker pain signals, as if, like good students, they “learned” to get better and better at transmitting pain. This results in a kind of runaway hyperarousal of the nervous system that some call “wind-up.” In fact, the properties of nerve cells can be altered so much that the pain is no longer coupled to the presence, intensity, or duration of noxious stimuli.39 It is now physically transformed from short-term, acute pain into a long-term, self-perpetuating phenomenon.40, 41, 42, 43, 44 This “learned” hypersensitivity occurs in both peripheral and central nerves.
In the periphery, several important things happen, and two in particular. Nerves that used to be responsive only to nasty—noxious— stimuli “learn” to react to benign substances as if they were noxious, too. And nerves become hyperresponsive—extra sensitive—to the original noxious stimuli, too. It’s almost as if the system gets addicted to all the excitement, craving more and more stimulation. Scientists call this ugly, new state of affairs allodynia. The system is so overrevved that the brain now responds to the most benign of stimuli—like a feather stroking the skin—as if it were a burning blowtorch. Mere touch now triggers excruciating pain.
Consider what happens in inflammatory pain. In inflammation, immune cells—with the best of protective “intentions”—begin secreting chemicals called cytokines, among them TNF-a, interleukin 1B, specific pain messengers like bradykinin and prostaglandin E2, and even the normally benign nerve growth factor (NGF). (In embryos, NGF guides the development of new nerves in embryos, but it can also rev up the pain response.) Together, these chemical messengers and others act on the tips of nerve cells in the periphery, making them increasingly sensitive to pain signals. Inflammatory diseases such as rheumatoid arthritis are prime examples of the cycle of misery caused by this revved-up process.
But sensitization doesn’t just occur in the periphery: It happens in the central nervous system, too, starting in the first signal relay station in the spinal cord, the dorsal horn. That’s the place where axons from peripheral nerves spurt out their chemical signals to waiting dendrites of nerve cells across the synapse, spurring the transmission of pain signals up to the brain. A key player in this process is the excitatory neurotransmitter glutamate. When glutamate is released by the axon of a C fiber, it floats across the synapse and lands on several different types of receptors: most important, receptors called NMDA. When triggered by glutamate, ion channels—in this case, for calcium—open up in the receiving cell. This triggers a cascade of chemical steps inside the cell, the net result of which is that the cell puts even more NMDA receptors on its surface. This, of course, makes the cell extra sensitive to pain, allowing even more pain signals to get through.
While glutamate and other chemical messengers like substance P and BDNF (brain-derived neurotrophic factor) rev up short-term pain fast through calcium channels, other transmitters land on different receptors in the receiving cell and start a slower revving up process that keeps the pain signals going longer term by actually acting on genes. Once this hypersensitivity process gets started in the central nervous system, it takes less and less stimulation from peripheral nerves to keep it going.
This has huge potential implications. To keep central sensitization to a minimum, it helps, as nurses often put it, to “keep ahead of the pain.” In one study, when doctors gave prostate surgery patients spinal injections of pain relievers before surgery, the patients had less pain afterward than those treated conventionally.45 If further research confirms this, one of the most straightforward ways to reduce the amount of chronic pain in this country would be to provide better control of acute pain after surgery.
But it’s not just that the nervous system learns to react to a benign stimulus like the stroke of a feather as if it were a blast from a blow torch, or that the nervous system becomes hyper-responsive to genuinely noxious stimuli. In some kinds of pain, particularly neuropathic pain (damage to the nervous system itself), what happens is that once a nerve is injured and gets revved up, the firing of electrical impulses can keep going without any trigger at all from the periphery.
It’s as if the nerves get on a roll, turned on by having been turned on before, eventually becoming free of the need for a triggering event. The nerves learn to fire spontaneous, or ectopic, signals all by themselves. Worse yet, it’s not just the nerves that once were triggered that gallop away with all these spontaneous, ectopic signals. Nearby nerve cells get caught up in the process as well, just like the flu spreading from person to person. At the most basic, neurological level, pain, too, becomes contagious, spreading from one nerve to the next. (I remember it well. That killer-burning feeling in my left shoulder and those fiery jolts of electricity shooting from my neck to my shoulder took on a life of their own. Now I understand why.) Adding significantly to the changeover from acute to chronic pain is the fact that nerves that are injured can actually change the activity of their genes, in many cases, increasing the activity of genes that pump out substance P, BDNF, and another pain molecule called neuropeptide Y.
And the body plays yet another nasty trick. The “good” neurotransmitters, like GABA, and another called glycine, chemicals that normally quiet the nervous system down, now get deranged themselves. During the process of sensitization, the beneficial inhibitory function of GABA and glycine gets turned on its head—these neurotransmitters literally change sides in the battle, revving up pain signals instead of quieting them down. In other words, with sensitization, not only do nerves increase transmission of pain signals, they also decrease their normal methods of damping down pain. After a while, some of the GABA-producing nerves in the spinal cord even die off altogether, perhaps from sheer overwork and exhaustion: They just can’t cope anymore with the onslaught of incoming pain signals.
And even all this is, unfortunately, only half the story. Making matters even worse is that immune cells also get into the act. Years ago, early neuroanatomists dubbed these cells glia, Greek and Latin for “glue,” though there are other, less lofty, translations as well, including “slime,” or “snot,” says University of Colorado neuroscientist and glial cell researcher Linda Watkins.46 Shocking as it may seem, glial cells— there are three types: astrocytes, oligodendrocytes, and microglia— outnumber nerve cells in the central nervous system by 10 to 1. In the developing brain, they do good work, guiding wandering neurons to the right destinations and helping them form synapses.
But, like Jekyll and Hyde, glial cells also have a dark side. If they become activated by chemical pain signals from nerves, they send out their own chemical signals that increase chronic pain. As if adding insult to injury, if a person is taking morphine to control pain, glial cells actually steal some of it to further rev up pain signals. Glial cells spring into action whenever the body senses that it is in some kind of physiological distress—from pain, but also from things as diverse as physical trauma, chemotherapy, diabetes, direct nerve damage, inflammation, and even bits of blood leaking out from blood vessels. Linda Watkins calls these chemical distress signals “alarmins.”
The minute an alarmin lands on a receptor called TLR-4 (toll-like receptor number 4) on the surface of a glial cell, the glial cell begins pumping out huge numbers of cytokines called IL-1 (interleukin 1), IL-6 (interleukin 6), and TNF (tumor necrosis factor). These cytokines, like most molecules in biology, are both “good” and “bad.” IL-6, for instance, is great at its “day job” as an immune stimulant. If you get an infection, it’s IL-6 that signals the brain to produce fever, thus raising body temperature, which in turn helps kill bacteria. IL-1 is a “good” cytokine, too, in its normal, immune-boosting role: It helps white blood cells flock to the site of an infection to fight bacteria. But these pro-inflammatory cytokines, particularly Il-1, are devils as well as angels. They are neuroexcitatory, meaning they rev up nerve cells to carry pain signals faster and faster.
Specifically, the cytokines pumped out by glial cells land on sensory nerves that carry pain signals up to the brain. This amplifies the original pain signal, making nerve cells fire faster and faster, generating ever more pain signals headed for the brain. The relatively recent discovery that glial cells play a major role in turning acute into chronic pain has revolutionary implications. If drugs can be developed to stop glial cells from amplifying pain signals—and scientists are working on this now—that would be a breakthrough in pain treatment because it would provide an alternative, nonnarcotic approach to pain control. Granted, once again, this is pretty intense biochemistry. But understanding what’s going on chemically validates the fact that chronic pain is not “all in one’s head.”
For the first time, scientists using modern brain scanning techniques, particularly functional magnetic resonance imaging (fMRI), can now look noninvasively inside the brains of people in chronic pain—in real time. This is a major step forward, not just because it adds significantly to understanding how chronic pain affects different parts of the brain, but because it makes real, explicit, and visible a phenomenon that, until now, has been totally subjective. As researchers Irene Tracey and Catherine Bushnell put it, fMRIs are at long last providing objective proof that the physical, emotional, and cognitive suffering of people with chronic pain is real.47
In 2011, Stanford University researchers led by neuroscientist Sean Mackey showed that they could use fMRI scans along with fancy computer algorithms to detect specific patterns of brain activity and, in essence, to tentatively diagnose pain—in the research lab.48 If this technology pans out, doctors may no longer need to rely solely on a person’s self-report of pain because they would be able to “see” it on a scan. (Of course, because no test is perfect, it’s possible that people in real pain might be dismissed as faking if they “flunked” such a test.) In the study, the researchers put volunteers in the brain scanner and then applied heat to their forearms, producing moderate pain.49 The scanner recorded brain patterns with and without pain and then analyzed the patterns to create a computerized model of what experimental, thermal pain looks like. Researchers then had the computer analyze the brain scans of other volunteers to see whether it could detect brain activity suggestive of thermal pain. The computer got it right 81 percent of the time.
We’re still a long, long way away from using fMRI as a diagnostic test for pain, Mackey told me.50 But it is a major first step—with enormous legal, as well as clinical, implications.51
Hundreds of thousands of legal cases every year depend on “proving” the existence of pain, though the technology could cut both ways.52 Some people may be able to use fMRIs to prove to skeptical insurers and employers that their pain is real, that they are not malingering and they merit compensation. But no scientific test, including fMRI, is perfect. Some insurers and employers may be able to argue that fMRIs are not yet sophisticated enough—or that the tests may indicate some kind of general brain arousal, like anxiety or distress, but not pain specifically, which could cast doubt on patients’ claims.
The most dramatic finding from MRI scans that measure brain structures (as opposed to function) in people in chronic pain is how much that pain causes loss of brain tissue, though no one really knows the exact cause of this loss—an important unresolved question. The first evidence of this came in 2004 with studies by A. Vania Apkarian of Northwestern University Feinberg School of Medicine. Using brain scans, he showed that people with chronic back pain have 5 to 11 percent less gray matter in their brains than healthy people.53 (Gray matter consists primarily of the cell bodies of neurons in the brain; white matter consists of the axons of these cells, which look white because of the fatty, myelin sheath that surrounds them.)
Normally, it would take 20 years of aging to lose this many brain cells, Apkarian found. Significantly, his team also showed that the decrease in the prefrontal cortex and thalamus was linked to the length of time a person had chronic pain. Since then, other researchers have found similar losses of brain tissue in people with fibromyalgia, irritable bowel syndrome, tension headaches, the facial pain of trigeminal neuralgia, and other chronic pain problems.54, 55, 56, 57 Even phantom pain and spinal cord injury can trigger losses of gray matter.58, 59 (Interestingly, when given to people with pain, opioids—formerly called narcotics— can also produce changes in the volume of 13 specific regions of the brain’s “reward” circuitry, though whether these changes are beneficial or harmful is not clear.60)
Changes in the brain due to pain constitute one of nature’s cruelest tricks. Catherine Bushnell and Irene Tracey have shown that chronic pain damages the thinking centers of the brain,61 as Greg Scherr, a 50-year-old California stockbroker now disabled by chronic back pain, discovered to his dismay.62 Scherr somehow fractured 11 vertebrae in his back, and no one can figure out why. “My vertebrae are fracturing like tempered glass,” he says. A seemingly endless series of surgeries and medications has helped only minimally. Typically, when he wakes up in the morning, his pain is a 5 on a 0-to-10 scale. By the end of the day, it’s often 8 or 9, depending on how much sitting or standing he has done. And that is despite three opioid medications. Often, he spends an entire day lying down in the fetal position. Recently, he’s been getting relief with thrice-weekly deep-tissue massage to relieve his cramped muscles and learning to live “in the sandbox that I now have to play in and NOT doing things outside that box to cause me more pain and additional fractures.”63
But the worst part of his long ordeal, for both Sherr and his wife of more than 25 years, is his memory loss. On the days when he is able to drive the short distance to the store, he announces that he’s going for coffee and the paper, to which his wife usually adds, “Get milk.” He often forgets the milk and has no recollection of her asking him. It wasn’t until he was filling out a pain questionnaire from his insurance company that asked about memory problems that he put it together. “Son of a bitch,” he said to himself. “That’s what it’s from. You’d never think a bad back would affect memory.”
Nobody knows exactly how chronic pain wipes out gray matter. One unproved idea is that the cells simply poop out and die from metabolic exhaustion; in other words, they burn out from overwork. A provocative study from McGill University researchers might support this idea. Women who have had the painful genital condition called vulvodynia for only a few years have increased gray matter, as if their brains were scrambling to cope with all the incoming pain signals. But women who have had vulvodynia for more than a few years have decreased gray matter.64
But there’s a conundrum here. If brain cells really are dying because of chronic pain, it’s hard to explain the emerging good news that some brain damage caused by chronic pain may be reversible! Germany’s Arne May, from the University Medical Center in Hamburg, studied 32 people with osteoarthritis in one hip who were scheduled for total hip replacement surgery. (Osteoarthritis of the hip, which can cause debilitating pain, is an appealing problem for pain researchers to study because, unlike most other types of pain, surgery can truly fix the problem, eliminating pain 88 percent of the time.65) Before the surgery, the researchers documented decreases in gray matter in several parts of the brain including the anterior cingulate cortex, the right insular cortex, the operculum, the dorsolateral prefrontal cortex, the amygdala, and the brainstem. They then looked at 10 of the patients after surgery. In all 10, hip pain had disappeared and there was an increase in gray matter in several of the previously affected areas.66 Seven months later, Irene Tracey of Oxford University published similar conclusions on 16 patients undergoing hip surgery.67 And in 2011, McGill University researchers similarly showed that effective treatment of patients’ low back pain restored normal brain function.68
Brain scanning has revealed some other remarkably specific changes accompanying chronic pain. One study in people with fibromyalgia, for instance, showed that “catastrophizing” about pain (imagining the worst) is a mental phenomenon separate from depression and linked to very specific brain regions—those associated with attention and anticipation.69 In a different study, researchers showed that a very specific area of the medial prefrontal cortex—right behind the forehead— was extremely active and tightly linked to the intensity of chronic back pain.70
A number of specific brain regions now appear to be changed by chronic pain, including the insular cortex, thalamus, cingulate cortex, somatosensory cortex I and II, and the prefrontal cortex.71, 72 It’s good news that scientists have been able to document such changes in the emotional and cognitive parts of the brain.73 These discoveries make clear and visible the subjective experience of many people with pain— that when pain is chronic and intense, emotions are intensified and it’s difficult to think straight.
Lastly, fMRIs are beginning to document precisely where in the brain opioid drugs like morphine—and drugs that block morphine—work. You can take a human brain, give it a drug, and see patterns of activation, David Borsook, a Harvard Medical School neurologist, told me.74 Just as one would expect, when he maps the actions of morphine and the opioid blocker naloxone, they show opposite activation patterns. This suggests that brain scanning will increasingly be useful not just to document chronic pain but to test which drugs might be most effective in which people with pain.75, 76, 77
My mother died in my arms a number of years ago—my mother, her two poodles, and me all snuggled up together in her bed at home. She had been in a semi-coma for several days—quite peaceful, actually. During those long days, I sat nearby and watched, amazingly peaceful myself. I became a student of her face, a face that was almost as familiar to me as my own. I learned to watch for the telltale signs that, despite her apparent unconsciousness, the pain from her terminal leukemia might be breaking through the drugs she was taking.
When that happened, her eyebrows would draw together, and the muscles around her eyes would contract, creating crow’s feet where, despite her 79 years, she had barely had them otherwise. The muscles in the middle of her face would contract, too, wrinkling her nose a bit and slightly raising her upper lip. (Interestingly, mice in pain do much the same thing, as researchers from McGill University and elsewhere have shown: Their eyes squeeze shut, their noses and cheeks bulge, and their whiskers stand on end.78 Rabbits express their pain similarly with facial changes: Their whiskers move, their noses bulge, and their eyes narrow.79)
Sometimes, it wasn’t her face that I noticed so much as a general restlessness—she would move her legs and wiggle around on the bed as if trying to get rid of something, or trying to get comfortable. Whenever I noticed what seemed to be signs of increasing pain, I asked the hospice worker and nurse, both of whom were there with us, whether it was time to increase her pain medications. Often, we did. I was utterly sure that I was accurately assessing her pain by these cues.
But was I? It turns out that assessing someone else’s pain is trickier than you might think. Functional MRIs, as we’ve just seen, may be one way. But even if they were infallible, objective tests for pain, very few people in pain have an fMRI machine in their bedrooms.
So, as a practical matter, that leaves other methods: looking at the faces of people in chronic pain, for example. Asking people to rate their pain on numerical, linear, or pictorial scales. Asking them to keep track of pain using electronic scales on iPads or other personal digital devices. Or, as doctors have done for generations, asking people to describe their pain verbally. And combining a few well-chosen questions with a very focused physical exam—in under 15 minutes. All of these methods— even simply rating pain on a scale of 0 to 10—can be helpful. But they all have serious drawbacks, too, partly because chronic pain is intrinsically such a subjective phenomenon.
For instance, looking at a person’s face as he or she grimaces in pain turns out to be pretty accurate—if the observer is a layman. (Accurate meaning that the observer’s rating of the pain correlates fairly well with the patient’s, and, to some extent, with fMRIs as well.) But if the observer is a healthcare professional, watch out. Research shows that the professionals—the very people you would hope would be most attuned to facial pain cues—routinely underestimate their patients’ pain. And if the doctor in question has reason to suspect—as doctors are often trained to do—that the person is not really in pain but is just seeking drugs, the tendency to misread and underestimate facial cues becomes even worse.80, 81, 82
The attempt to capture pain in some sort of orderly fashion really began in 1975 when Canadian psychologist Ronald Melzack developed what has become known as the McGill Pain Questionnaire.83 The questionnaire divides pain into different types by the words people in pain use to describe it. For instance, to specify the temporal qualities of pain, the questionnaire asks people to pick one of the following words: “flickering,” “quivering,” “pulsing,” “throbbing,” “beating,” or “pounding.” To capture the spatial aspects of pain, the questionnaire asks people whether their pain is “jumping,” “flashing,” or “shooting.” The questionnaire also asks about so-called punctate pressure—that is, whether the pain feels “pricking” or more severe, like “stabbing.” It asks about thermal features of the pain, from “hot” to “searing”; about the “dullness” of pain, from “dull” to “heavy”; and about the overall pain, from “annoying” to “unbearable.”
Although some pain clinics still use the McGill Questionnaire, many now use other methods. For young children, many doctors use the “faces” scale, in which the child picks the one picture of a face out of six that most closely matches his or her pain. These scales show a range of faces: from a happy face showing no pain to a scrunched-up face grimacing in agony. Partly because the faces scales have so few gradations, they’re pretty awful from a scientific point of view, Donald Price, a University of Florida neuroscientist, explained to me.84 A bit better is the simple 0 (no pain) to 10 (the worst pain imaginable) numerical scale, and even better is a numerical scale with finer gradations, from 0 to 100. Personally, I found the 0–10 scale helpful because it allowed me to take heart from even tiny decrements in pain. There were many days—and nights—when I ranked my pain at 10++++ on the scale. Deep breathing, lying still, meditating as best I could, even just trying to focus on a TV show, could sometimes drop the pain a tiny notch or two. This simple scale also helped me keep track of what activities or physical positions made my pain better or worse, and which treatments knocked it down a number or two.
But numerical scales have a flaw, at least for researchers trying to assess the efficacy of pain medications.85 The problem is that the scales are linear and can’t easily reveal ratios, such as the amount of improvement a certain drug or treatment provides. For that, a ratio scale such as the visual analog scale (VAS) is better. In general, the VAS scales correlate quite well with fMRI assessments of pain. What’s really important, however, for both patients and doctors, is to know how pain is affecting a person’s ability to function and participate in the activities of daily life. Even better, says Harvard Medical School professor of anaesthesia Robert N. Jamison, are electronic versions of this kind of scale.86 A number of commercial pain tracking programs and e-diaries are available online.
And even better than simple e-diaries are pain-tracking systems that use what psychologists call ecological momentary assessment. A handheld device beeps at random times during the day, prompting a person to enter data on his or her current pain rating and mood. This is more reliable than asking people to remember—at the end of the day, or in a doctor’s office a month later—what their pain has been like. The take-home message for people with chronic pain is that there’s growing evidence that if they present these factual e-diaries to their doctors, doctors are more likely to change medications to improve pain control.87, 88
But there’s an even more promising pain assessment method in the works—combining better-designed questions with highly focused physical exams in the doctor’s office. The idea, says pain researcher Clifford Woolf of Boston Children’s Hospital, is to get the phenotype of a person’s pain. This means using the person’s own verbal description of pain to figure out precisely what kind of pain the person has in order to get at the underlying neurological mechanisms. Pain described as “burning,” for instance, may be different from pain that makes a benign experience like taking a shower excruciating. The verbal descriptions often point to different mechanisms.
Woolf’s team has now developed a program called the Standardized Evaluation of Pain (StEP) that involves just six questions and 10 physical tests. It’s cheap, low-tech, takes only 15 minutes, and is helpful for figuring out, for instance, whether a person’s back pain is neuropathic— that is, caused by damage to the nervous system itself, or not. In fact, the StEP system seems able to predict with 90 percent accuracy (even better than a brain scan) what kind of pain a person has.89 In one study of 137 people in pain, STeP was able to distinguish axial back pain, that is, nonspecific pain that does not travel to the buttocks, legs, or feet and is not associated with a specific anatomical problem, from radicular back pain (also known as sciatica), which is neuropathic pain caused by inflammation of the nerve root and bulging discs or bone spurs in the lower region of the spine.
This difference matters. If the pain is axial, the best treatment may simply be nonsteroidal anti-inflammatory drugs (NSAIDs). But if it’s radicular, the best choices may be gabapentin (Neurontin) or duloxetine (Cymbalta).
And, finally, what about the old-fashioned technique that I used with my mother, simply looking at a person’s face and trying to gauge the degree of pain? Since my mother couldn’t talk and I knew her very well and had no trouble believing that her terminal cancer was causing pain, I suspect that my assessment of her face and body language probably was pretty accurate. It did give me a sense of how much pain she was in. Besides, as a practical matter, my family and I had already agreed that if there was any doubt, we would err on the side of overmedicating her, since it was clear to all of us that she was dying.
But in less extreme situations, important questions remain: How accurate are facial cues in conveying pain? How well do observers read these signals? Why do doctors do worse at this than other people? Can a person fake pain expression? How much does an observer’s own history of pain, or lack thereof, influence how he or she rates someone else’s facial cues? And what happens, as it often does with health professionals, if the observer’s training is geared more toward spotting potential drug abuse than toward treating pain?90
Facial cues, it turns out, are a reasonably accurate way for the person to communicate pain, says psychologist Kenneth Prkachin of the University of Northern British Columbia. But from the observer’s side, things can get muddied, with a big confounder being how much pain the observer has encountered in his or her own life. Prkachin has found that observers who have had chronic pain in their own or their families’ lives rate patients’ facial expressions of pain much higher than observers who haven’t.91 This is important. If an older person experiencing pain is seen by a young doctor who has never yet had significant chronic pain, that doctor might underestimate that person’s pain.
But doctors, and some other healthcare professionals, routinely underestimate patients’ pain, regardless of their own personal experiences with pain. Study after study going back to the late 1990s shows that healthcare professionals minimize patients’ pain far more than nonhealthcare professionals do.92 If an ordinary person downgrades a patient’s facial expression of pain by 10 percent, healthcare professionals downgrade it by 15 to 20 percent, Prkachin told me. All of which, of course, contributes to dismissing a person’s pain as “all in his head.”
In a clever experiment, Prkachin and European colleagues videotaped people with shoulder pain being asked to move their shoulders in a painful way. The videotapes were then shown to 120 healthcare professionals, who were divided into three different groups and told to rate the patients’ pain.93 One group saw only the videotaped faces. The second group saw the videotaped faces and was also given the patients’ numerical pain ratings. The third group was given the same information as the second group, but was also told that the patients were probably cheating and were faking their expressions of pain to seek narcotics.
The healthcare professionals all routinely underestimated the patients’ pain. But those who got the patients’ own pain ratings as well gave estimates closer to the patients’ own rating. And the healthcare professionals who were told the patients might be cheating underestimated pain just as badly as the group that saw only the faces. (Interestingly, a 2011 study from McGill University suggested that observers are more likely to think men are faking pain than women.94)
If doctors really want to make more accurate assessments of facial pain cues, psychologist Kenneth Craig has a suggestion: Pay attention to the timing of factual cues. Facial cues, after all, are subject to both voluntary and involuntary muscle movements. When a person is faking, he or she often gets the timing of winces and grimaces wrong. The timing of facial cues during faked pain are out of sync and typically exaggerated, Craig explained, almost like a caricature of pain expression.95
But I have a suggestion, too. Obviously, healthcare professionals see so much pain every day that they may become numb to it and fail to read facial cues right. But when in doubt, they could err on the side of believing the patient.