What Is This Article About?

A recent version of the NYT carried this article.  (It's a blogsafe link, so it should still work even after the article goes into the archive.)

Who's Mentally Ill? Deciding Is Often All in the Mind
By BENEDICT CAREY
Published: June 12, 2005

THE release last week of a government-sponsored survey, the most comprehensive to date, suggests that more than half of Americans will develop a mental disorder in their lives.

The study was the third, beginning in 1984, to suggest a significant increase in mental illness since the middle of the 20th century, when estimates of lifetime prevalence ranged closer 20 or 30 percent. [...]

That evolving understanding can have implications for diagnoses. For example, in 1973, the American Psychiatric Association dropped homosexuality from its manual of mental disorders, amid a growing realization that no evidence linked homosexuality to any mental impairment. Overnight, an estimated four to five million "sick" people became well.

More common, however, is for psychiatrists to add conditions and syndromes: The association's first diagnostic manual, published in 1952, included some 60 disorders, while the current edition now has about 300, including everything from sexual arousal disorders to kleptomania to hyposomnia (oversleeping) and several shades of bipolar disorder.

"The idea has been not to expand the number of people with mental conditions but to develop a more fine-grained understanding of those who do," said Dr. Ronald Kessler, a professor of health care policy at Harvard Medical School and lead author of the latest mental health survey.

Naturally, I was curious about this, so I looked up the study, meanwhile feeling irritated by the lack of a direct reference to the study.  I know the NYT thinks hyperlinks are beneath their dignity, but even a printed reference would be nice.  Anyway, this appears to be the study:

Prevalence and Treatment of Mental Disorders, 1990 to 2003
Ronald C. Kessler, Ph.D., et. al.
NEJM Volume 352:2515-2523

[...] Results The prevalence of mental disorders did not change during the decade (29.4 percent between 1990 and 1992 and 30.5 percent between 2001 and 2003, P=0.52) [...]

In the NYT article's second paragraph, the author stated "The study was the third, beginning in 1984, to suggest a significant increase in mental illness..."  Yet the first sentence in the Results section of the abstract states no such thing.  In fact, the study shows that the small increase from 1990-2 to 2001-3 is as likely as not (P=0.52) to be a mere artifact of the study design or implementation.  The study is not about any possible change in lifetime incidence from the middle of the 20th century to the present time.  

The aim of our study was to present more comprehensive data on national trends with regard to the prevalence and rate of treatment of 12-month mental disorders based on the NCS, conducted from 1990 to 1992, and the NCS Replication (NCS-R), conducted from 2001 to 2003. 

One potential point of confusion is that the NYT author, Mr. Carey, emphasizes discussion of lifetime prevalence, which is a different statistic than the 12-month prevalence numbers emphasized in the NEJM article.  That is not necessarily a problem, although it would have been good for him to take a moment to clarify the point.  Even allowing for that, the guy made a mistake.  Earlier studies may have indicated a lifetime prevalence of 20-30%, but the NEJM study does not address that.

The mistake, however, actually is not what this post is about, although I do take pains to point it out.  Rather, if you read the NYT article, then read the original study, you would not have any idea that the former had anything to do with the latter.  So what, exactly, is the article about?

It appears that Mr. Carey used the occasion of the publication of the NEMJ article to blather on about his own ideas about mental health diagnosis and treatment.  Snarky readers, no doubt, realize that what Carey did is exactly what I often do in this blog, so who am I to be critical?

Good question.  The answer is that this is a blog, but the New York Times is a newspaper.  If someone is going to write an article that starts out with: "THE release last week of a government-sponsored survey, the most comprehensive to date, suggests that more than half of Americans will develop a mental disorder in their lives...", then it seems that the article should be about the study.  

If the author wants to take the opportunity to express his varied opinions, and connect all kinds of things that are not directly connected, fine.  Just do it on the op-ed page; or better yet, get a blog and do it there.  

Although I disagree with a lot of what he says, it still is an interesting article, and is worth reading.   It would make a great blog post.  The problem is that, if the reader is not careful, it would be possible for the reader to get the impression that the opinions expressed by the author are backed up by the scientific article he has cited, or at least by some of the scientists he quotes.  

To be fair, Carey does not make any declarative statements of opinion; rather, he implies them, as in this section:

But what does it mean when more than half of a society may suffer "mental illness"? Is it an indictment of modern life or a sign of greater willingness to deal openly with a once-taboo subject? Or is it another example of the American mania to give every problem a name, a set of symptoms and a treatment - a trend, medical historians say, accentuated by drug marketing to doctors and patients?

The NEJM article says nothing about drug marketing.  None of the other experts, of those quoted in the article, are quoted as having said anything about drug marketing.  Carey does not cite, specifically, any studies about drug marketing.  (There have been such studies, but none is cited.)  I agree with the implication, that drug marketing is an interesting topic; I've probably posted on that subject at least a dozen times, and cited studies, but he offers this question, with no attempt to answer it.  

Is it true that drug marketing accentuates the American mania to give every problem a name?  Is it even true that it is an American mania, or even a mania at all?  If so, what name should we give to a mania about giving things a name? (Diagnosomania, American type, severe, with obsessive features?)

Note that I do not mean to impugn the entire newspaper, nor do I intend to imply that any of Mr. Carey's other articles are suspect in any way.  I just want to point out that readers need to be careful when reading these kinds of articles, regardless of who wrote them, or where they are published.

category: science, mental health, media, rants