The other two categories, neuropathic and idiopathic pain, seem not to serve any purpose. Neuropathic pain results when a nerve is injured by surgery, trauma or a disease such as shingles or diabetes. This eventually triggers abnormal functioning of the nervous system, with neurons firing erratically and forming irregular circuits. Neuropathic pain may be spontaneous or caused by contact with something that normally wouldn’t be painful. Idiopathic pain is similar to neuropathic pain but has no known trigger. It may occur all over the body—a condition known as fibromyalgia—or can be localized, such as with TMJD. Physicians treat neuropathic and idiopathic pain with the same arsenal of drugs they prescribe for nociceptive and inflammatory pain, as well as with stronger anticonvulsants and antidepressants that may provide some relief. But the pain rarely goes away.

Woolf likens the entire pain system to a fire alarm. “With nociceptive pain, there’s a fire, the alarm goes off and you do something about it,” he says. “With inflammatory pain, just the threat of a fire is enough to trigger the alarm while someone is healing. Neuropathic and idiopathic pain are like false alarms. The alarm goes off all the time, but there’s no fire and nothing to heal. The alarm system is broken.”

Unlike nociceptive and inflammatory pain, neuropathic and idiopathic pain don’t happen to everyone. Around 10% to 50% of those who undergo such procedures as hernia repair and coronary bypass suffer chronic neuropathic pain, and the pain is severe for some 2% to 10%. Recent studies suggest that genetic makeup determines whether a patient will develop chronic pain, but much remains unknown.

Because the genes involved in pain sensitivity have just recently been discovered, preliminary findings have yet to be replicated in large populations. Further complicating the search for answers is the dauntingly complex nature of pain genetics. Each type of pain—hot, cold, pinch, stab—may well be transmitted to the brain in a completely different manner, involving different genes and neurons. Yet these complexities may ultimately be a good thing, providing highly individualized targets for therapies, says Zubieta.

If that happens, it may be thanks in part to a study published in the journal Science in 2003 that ignited the field of pain genetics. The paper investigated whether a gene that codes for the manufacture of an enzyme called catechol-O-methyltransferase, or COMT, is involved in pain sensitivity. Researchers already had determined that COMT regulates such molecules as dopamine, epinephrine (adrenaline) and norepinephrine, all components of the endogenous opioid system—the pain-relief mechanism activated in response to a burn, stab or pinch. They also knew that a particular variant of the COMT gene, called met158, encodes a version of the COMT enzyme that works three to four times less well than the regular version in breaking down dopamine and epinephrine. (Met158 is a single nucleotide polymorphism, or SNP. Pronounced “snip,” it’s a small change in the DNA code that can alter a gene’s function.) Zubieta was curious to know whether people with met158 experienced pain differently than those with the regular version of the gene. So he tested the DNA of 29 subjects to determine which version of COMT they had. Then he injected hypertonic saline into their jaw muscles—causing deep, sustained muscle pain—and ran them through a PET scanner to observe their brain activity. Zubieta found that the brains of one in four subjects who had two copies of the met158 SNP—one inherited from each parent—showed less activation of the opioid system, and the subjects reported more pain.

Two years later Luda Diatchenko, a geneticist at the University of North Carolina at Chapel Hill, built on Zubieta’s work. Diatchenko and William Maixner, head of the UNC Center for Neurosensory Disorders, recruited 202 healthy female volunteers for pain-sensitivity experiments. Researchers pressed the skin of their cheeks, forearms and feet with a thermal cylinder resembling a car’s cigarette lighter, measuring how long it took the subjects to say the cylinder was hot. The researchers also applied pressure to different muscles, to measure the kilograms of force required for subjects to register pain. They tested how painful the hot cylinder became over time with 15 heat pulses at 53°C (127°F) to the same part of the hand. And they examined deep-muscle pain, using an arm cuff similar to those for measuring blood pressure.

After combining the subjects’ responses to all these stimuli into a single pain-sensitivity measurement, the researchers sampled the subjects’ DNA, testing not only for met158 (as Zubieta did) but also for five other SNPs from other parts of the COMT gene. Diatchenko found that certain groupings of these six SNPs—combinations known as haplotypes—were linked more closely to pain sensitivity than was any single gene variant. Almost four in 10 subjects had a haplotype strongly associated with low pain sensitivity (LPS), 49% had a haplotype linked to average sensitivity (APS) and 11% had the HPS haplotype, for high sensitivity. Women with the LPS haplotype also had higher levels of the COMT enzyme (encoded by the COMT gene) and were 2.3 times less likely to develop TMJD.

In a follow-up study, Diatchenko found that although people with the LPS haplotype felt less pain in general than did those with other combinations of SNPs, the difference was greatest for thermal pain. In addition, she discovered that Zubieta’s met158 SNP was, by itself, strongly associated with pain that increased over time—those 15 pulses of heat became successively more painful for people with met158. (That was in line with Zubieta’s findings, in which subjects with met158 felt more pain only if it was sustained for 10 to 20 minutes.)