Wednesday, June 20, 2012
Comparing apples, oranges, and kiwi fruit
Research can be tricky because different people usually ask different questions - which, of course, may lead to finding different answers.
On June 2nd Allan V. Horwitz and Jerome C. Wakefield posted an article titled “Our New Era of Anxiety” at Salon dot com. It was adapted from their new book All We Have to Fear, and subtitled “Anxiety disorder rates have risen twentyfold in 30 years -- largely because psychiatry misunderstands human nature.”
Horwitz and Wakefield say that:
“....Moreover, it might seem as if a startling increase in the number of anxiety disorders has occurred in recent years. Consider that in 1980, the third edition of the authoritative Diagnostic and Statistical Manual of Mental Disorders (DSM) of the American Psychiatric Association stated: “It has been estimated that from 2 to 4% of the general population has at some time had a disorder that this manual would classify as an Anxiety Disorder.”
Then, the first large community study conducted after the publication of this manual, the Epidemiologic Catchment Area Study, found that about one out of ten people had an anxiety disorder in any given year and that roughly 15% of individuals experienced these conditions at some point in their lives.
Just two decades later, a similar and equally rigorous study, the National Comorbidity Study(sic) Replication (NCS-R), yielded the shocking result that almost one out of five people had had an anxiety disorder over the past year, and more than a quarter of the population (28.8%) had had one at some point in their lives.
Even the NCS-R actually underestimates the frequency of anxiety disorders, as measured by psychiatry’s current criteria for diagnosing mental disorder. Because it asks people to remember years later what anxieties they had earlier in life, many respondents forget past episodes. A recent New Zealand study with substantially improved methodology that involved repeated interviews of participants established that in any given year between ages 18 and 32, nearly a quarter of all young adults (22.8%) experience an anxiety disorder and that virtually half (49.5%) report at least one such disorder during the entire period. Obviously, the study would have yielded even higher estimates if it had included disorders emerging after age 32 or before age 18.
How did rates of anxiety disorders rise by as much as twentyfold over the past 30 years to encompass as much as over half the population?”
Now, I don’t think those different rates really are such a big mystery - if you look carefully at what was being measured, how it was done, and where it was done. I’ve summarized those factors in a handy table (which is worth at least a hundred words):
The first paragraph from that quote just described an estimate, which is a polite word for a guess. The DSM defines a yardstick for measuring anxiety considered serious enough to be called a disorder - a phobia. Perhaps the yardstick was the second edition, DSM-II?
Their second paragraph described results from the Epidemiologic Catchment Area (ECA) Study, which surveyed the “watersheds” around five metropolitan areas in the U.S. It was like an apple - a first bite at surveying what went on, and not a random national survey. The yardstick was DSM-III. I’ve blogged about the ECA study.
Their third paragraph describes the NCS-R, which was a national survey of the U.S. using another yardstick, the DSM-IV. So, it’s like an orange, and can’t necessarily be compared with an apple. I’ve blogged about the NCS-R.
Their fourth paragraph described a study done in New Zealand, home of the kiwi fruit. That’s halfway around the world from the U.S., so there’s no obvious reason for finding the same rate of anxiety.
Horwitz and Wakefield also omitted discussing the original National Comorbidity Survey (NCS), which used DSM-III as a yardstick. That survey was repeated a decade later as NCS-2, but those results haven’t been reported yet. I don’t think the survey in New Zealand really had any better methodology than what had been used in the U.S.