Manufacturing depression (49 page)

Despite the prominence of clinical significance
in diagnostic criteria, there is currently no consensus as to how it should be defined or operationalized. In large epidemiological surveys, direct clinical judgment is rarely used because of the high cost of clinical time and the large number of subjects.

Doctors, in other words, are too busy diagnosing patients to worry about whether or not they are really sick.

Regier thinks that the data are there to assess clinical significance. Researchers—epidemiologists and clinicians alike—do ask patients how much a given symptom interferes with their lives or whether it ever prompted them to visit a doctor. But the answers aren’t necessarily factored into the final results, so, as I discovered
at Mass General, a person who gives little or no indication of significant impairment but who has five symptoms of depression is still depressed. The data remain, however, and Regier and his team extracted them to reassess the outcomes of the major epidemiological studies of depression, the ones that lead to those dire estimates. And it turns out that, once people are ruled out who have the symptoms but aren’t impaired or distressed by them,
the prevalence of depression is cut nearly in half
.

This problem has led some psychiatrists to suggest that
the current categorical approach
to diagnosis should be replaced with a dimensional approach, in which only people at the extreme end of the symptom spectrum, the ones with the most symptoms that rise to the highest levels of clinical significance, would receive diagnoses. Regier’s position as vice chair of the DSM revision committee has led some psychiatrists to worry that if this happens, the disorders at the mild end of the spectrum will soon go the way of homosexuality and neurosis.
These doctors have already struck back
with some statistics of their own, showing that the mildly symptomatic will eventually get worse. We don’t wait for people with high cholesterol and blood pressure to have a stroke before we diagnose and treat them, they argue, so why should we wait for the mildly depressed to cross the threshold?

The dimensionalists suggest that such people ought to get treatment and that any unmet need this creates “
should be addressed
by developing comprehensive triage rules that allocate available resources based on evidence-based assessments of the cost-effectiveness of available treatments.” Interestingly (and speaking of a “guild attempt” to increase business) this proposal not only casts the net wide; it also creates an argument for early assessment (which can only increase the numbers of diagnoses) and preemptive treatment—much of which will no doubt be with medication administered by psychiatrists, usually the cheapest of the available resources.

Whichever way this dispute breaks, one thing is certain: you are
very unlikely to hear much about the dispute over mild disorders. The squabbles over neurosis and homosexuality taught the American Psychiatric Association a lesson about airing the family linen in public, so it has made people serving on the DSM-V committees sign a confidentiality agreement as a condition of participation. That’s too bad, not only because we’ll miss a debate bound to be as absurd as it is enlightening, but also because from what I can gather from
A Research Agenda for DSM-V
, a book the APA brought out at the beginning of the planning process for DSM-V in 1999 (a mere five years after DSM-IV was released; you would think that they’d realize how all this dithering looks to the rest of us), the guild is going to make the zero validity problem part of its proceedings as it fashions a new diagnostic manual:

The major problem for mental disorders
as currently defined is that their causes and pathophysiological mechanisms remain largely unknown. It is expected that, at some point in the future (perhaps decades from now), the pathophysiological states predisposing or contributing to major mental disorders will be identified… Once it is possible to define a mental disorder based on the identification of its underlying pathology, then it would surely make sense to follow the course of other medical conditions and have the presence of the disorder be based solely on pathology and not on the effect this pathology exerts on the individual’s functioning.

It would be nice to hear psychiatrists acknowledge in public that even though they’ve been telling people for two decades that they know what the underlying pathology of depression is, they really don’t. But with
the pathophysiologically based classification system
that the book says will solve this problem decades away, it’s no wonder that the APA wants to keep a tight lid on the proceedings. They don’t want us to know that they’re still working off that promissory note until they’re ready to put paid to it.

*      *      *

There is some evidence that reform-minded doctors should be careful what they wish for. The transformation of psychiatry into clinical neuroscience may hold some unpleasant surprises for them. Consider what psychiatrist Max Fink has been saying recently.

Fink thinks
that the APA took a disastrous wrong turn when it resurrected Kraepelin’s categorical approach but left buried one of his most important categories: melancholia, a diagnosis that Fink thinks fits the subgroup of the depressed who feel anxious and despondent for no particular reason, who wring their hands and sleep all day, who are delusionally guilty and self-reproachful—patients like the woman I’ve called Ann. This, Fink points out, is the cluster of symptoms that Hippocrates originally observed, that has been reported by Kraepelin, Freud, Meyer, and Kuhn (and nearly anyone else who has bothered to look). The disease, as Fink sees it, has none of the Chinese menu fuzziness of the DSM’s major depressive disorder, but instead has four symptoms,
all
of which must be present for the diagnosis. And, best of all, the fourth criterion is a lab test—either a sleep study that shows irregular brain wave activity or an endocrine panel that reveals abnormalities in cortisol, a hormone whose levels increase with stress. Fink’s melancholia, in other words, is a disease in the modern sense—a form of suffering with a specific biochemical signature. To Fink, this means that it would behoove his profession to restore it as a diagnosis in the DSM-V. It could even be the flagship disease in that longed-for pathophysiologically based classification system.

Fink is not the first doctor to propose cortisol tests to verify depression, and they are compatible with current neurochemical theories, which see depression as a stress reaction gone amok. But studies have shown that fewer than 50 percent of the subjects who meet DSM criteria for depression turn up positive on the bioassays. On the other hand, when researchers weed out the nonmelancholics from the subject pool, then that number goes up to 70 percent
or higher. Fink also points out that studies showing homogeneous brain chemistry or structure tend to be strongest when the subjects are more severely depressed, which suggests, he says, “
that what is now considered
the pathophysiology of major depression is best restricted to melancholia.” Make the category less heterogeneous, in other words, and it may actually start to have some scientific integrity.
*

But not without a cost—market share. Melancholics in Fink’s sense make up only a small portion of people who meet the DSM criteria for depression. But that’s not the only reason that the depression industry is not beating a path to Max Fink’s door. It’s also because Fink, who in his late eighties, is one of the world’s leading proponents (and practitioners) of electroconvulsive therapy, which is
a highly effective treatment for melancholia
—as doctors have known since the 1940s. But while doctors continue to provide ECT, very quietly, it’s hard to imagine who is going to pay for clinical trials for a device that lost its patent protection long ago, and which has such a terrible reputation.

Nor are the drug companies likely to get behind Fink’s proposal anytime soon. They do have a cure—tricyclic antidepressants, whose effectiveness with melancholia approaches that of ECT and far outpaces the SSRIs. But they’re also off-patent; there’s not much money to be made there either. Some electricity-based therapies less dramatic than ECT—deep brain stimulation, transcranial magnetic stimulation, and vagus nerve stimulation—have shown promise for melancholia, but these are hardly blockbuster treatments, especially if the patient pool is limited to those who fit Fink’s diagnostic scheme. Given the fact that the burial of melancholia was essential to the expansion of the depression diagnosis, it is very
unlikely that without the lure of large profits the industry is going exhume it anytime soon.

Fink’s proposal to restore melancholia to the DSM suggests a reason for the poor performance of SSRIs in clinical trials. Doctors know that the subjects who show the most melancholic symptoms do worse in clinical trials than the other patients. And the HAM-D, the test that measures the performance of SSRIs, was standardized on hospitalized depressives, many of whom would very likely meet Fink’s criteria for melancholia. The drugs, in other words, may be good for something, just not what the doctors are looking for.

Actually, the drug companies already know at least one thing SSRIs are good for. They’re just not sure they want to make a big deal about it.

Twenty to forty percent of all men suffer at one time or another from premature ejaculation (PE). (These numbers may be inaccurate; in keeping with the DSM’s requirement that a disorder be a problem for the patient, the benchmark criterion is ejaculation that occurs “
on, or shortly after, penetration
and before the person wishes it,” so men can have the disease if they think “shortly after” means anytime before they wish it, even if that’s, say, after the four hours promised in the Viagra ads, while men who are just in a hurry to get back to the football game don’t qualify for the diagnosis at all.) As many as 70 percent of people taking SSRIs suffer from sexual dysfunction—which in men often takes the form of delayed ejaculation. Indeed,
repeated studies
, most of them conducted in European countries, show that the intravaginal ejaculation latency time of men suffering from PE increases significantly when they take SSRIs.

You don’t have to be a marketing whiz to see the lemonade-out-of-lemons opportunity here. Our pharmacological Calvinism isn’t quite intense enough to justify an ad campaign promising an anti-depressant whose side effects cancel your guilt. (“Makes you feel
better, but don’t worry. It will ruin your sex life.”) But the endless desire for sex and the inescapable feeling that it’s never quite good enough offer a chance to turn side effects to advantage by creating a new market for the drug.

But the fact that the SSRIs do reliably treat an illness, even if it’s not depression, hasn’t yet translated into drug company riches. The reason is obvious. To advertise SSRIs as a cure for PE, the drug companies would have to get an indication from the FDA, and, as one of those European PE researchers pointed out, “
focusing on the ejaculation-delaying effects
of these drugs would highlight their potent sexual adverse effects and thereby hamper marketing strategies for use of these agents for depressive disorder.” Indeed, currently the industry officially denies, and the FDA officially doesn’t know about, the sexual side effects. According to
the package insert for Paxil
, for instance, only 1.6 percent of men taking it experience abnormal ejaculation. (That’s probably because clinical trials aren’t exactly designed to elicit accurate information about such intimacies. My Mass General doctors sandwiched their questions about my sex life between items about nausea and muscle cramps. It is as if a person’s sexual performance were no more difficult to talk about with a stranger than his headaches. “Oh, yes, doctor,” we must imagine the subject saying, “I have a sore throat and, now that you mention it, I can’t come. Now, about this backache…”) There’s no particular reason to change this. After all, there is nothing to stop the drug makers from having it both ways—whispering to doctors about antidepressants as an off-label treatment for premature ejaculation while shouting from the rooftops that they don’t interfere with your sex life.

That may change. Pfizer recently funded a study in which
researchers gave Viagra
to women suffering from
antidepressant-associated sexual dysfunction
(look for this one in the DSM-V) to see if it undid the orgasm- and libido-inhibiting effects of Zoloft on women. Seventy-two percent of the women reported improvement, a finding that was trumpeted in press releases that resulted in
widespread news coverage. Perhaps when Pfizer finally formulates a combination therapy (Vizoft, anyone?) that promises to enhance both happiness and sexual pleasure—pardon me, to cure major depressive disorder and
female sexual dysfunction
—those package insert numbers will finally change.

But I digress. The real point is that to say SSRIs don’t test well is not to say they don’t work. Any clinician can tell you that they do, that he or she has been humbled by how quickly and effectively the drugs improve the lives of some of their patients; and the success of the drugs in the marketplace indicates that the customers are happy.

The problem with the clinical trials may simply be that they don’t measure what the drugs do, or, to put it another way, that the diagnosis and the treatment aren’t well matched. If researchers want the numbers to break their way more often, then all they need to do is to figure out what the people who actually respond to SSRIs are suffering from, find a way to test for it, and then give it a name.

Here’s my suggestion: Prozac-deficit disorder.

Okay, that’s glib and maybe even unfair. But it’s also honest, at least to the extent that it reflects the way doctors actually prescribe the drug. Just ask Richard Kravitz. He’s the physician who led the study in which standardized patients faked depression and then asked for Paxil. I called him to ask about, among other things, what it meant that doctors seemed willing to prescribe the drugs without rendering a diagnosis. “The pursuit of a more precise diagnosis often hinges on the relative risks and benefits of the available treatment,” he told me. “If the treatment is relatively harmless, then sometimes you give empiric therapy a try, as many of these doctors did.” What matters, he added, are results. “People who are unhappy will get better and people with major depression will get better.” Or, as the head of psychiatry at Stanford University once told an interviewer, “
for the vast majority of… the walking wounded
,
the SSRIs are good drugs.” So when someone walks in the doctor’s door, as Kravitz’s accomplices did, bearing the wounds that the doctor has identified as responsive to the drugs, the patient is likely to leave with a prescription.