(Again, just going for writing, not for quality.)

The New York Times published an AP article with the scandalous headline of “1 in 5 Young Adults Has Personality Disorder“. The article refers to the paper “Mental Health of College Students and Their Non–College-Attending Peers” in the Archives of General Psychiatry.

I am still not the incisive reader of journal articles that I would like to be, though, nonetheless, here are my thoughts. (I will begin with the disclosure that I sighed with annoyance when I saw the article because (1) personality disorders are considered rare (see below) and (2) college students (and their “non-college-attending peers”) are just getting out of adolescence and should not be considered paragons of virtue, consistency, and stability. Furthermore, long time-readers are already acquainted with my annoyance with several aspects of psychiatric diagnosis… and with my appreciation and ardor for alliteration.)

First of all, to build the case for the paper, the authors open with reminders about recent violent events on school campuses (Virginia Tech last year and Northern Illinois this year). This primes the reader to link mental disorders with violence, though the two are often not related.

The data for this paper was drawn from an epidemiologic study called the National Epidemiologic Survey on Alcohol and Related Conditions (NESARC), which surveyed over 43,000 people for, specifically, alcohol and related conditions (with the caveat that the survey also examined psychiatric disorders). I can’t express too much skepticism about the population because the NESARC population is, ostensibly, representative of the nation (versus specifically focused on a certain group). Thus, it’s not like the NESARC researchers were specifically looking for people who use alcohol, etc. So, using this group for this college study seems okay, though something about that still seems suspicious to me.

The paper actually does not report who conducted the NESARC survey (only that the interview was built so that non-clinicians could administer it with valid results: “All of the diagnoses were made according to DSM-IV criteria using the National Institute on Alcohol Abuse and Alcoholism Alcohol Use Disorder and Associated Disabilities Interview Schedule–DSM-IV version, a valid and reliable fully structured diagnostic interview designed for use by professional interviewers who are not clinicians.”), though the AP articles explicitly states that non-clinicians gave the interview. (Some researchers will administer an interview through a computer to reduce the “social” aspects of an interview and thus hopefully increase the truthfulness of the interviewee. Having never seen the above interview schedule, I don’t know if this interview is amenable to a computer administration.)

81% of people responded to this survey and the following groups were over-represented: “African American individuals, Latino individuals, and young adults (aged 18-24 years)”. This introduces bias into the NESARC data because the population is skewed.

The researchers for this paper then focused on college-aged people within the NESARC study, which summed to about 5,000 people.

That interview schedule listed above focuses on the following “Axis I” conditions (considered “major mental disorders”): substance use, mood disorders, and anxiety disorders. Notably, psychotic disorders are missing. (Why?) Presumably, the focus on “Axis II” conditions (”personality disorders”) is just as rigorous. Do note, however, that this paper actually does not survey for all of the personality disorders that are listed in DSM-IV. (Why not?)

The time frame that the paper covers is 12 months (which lends itself to inflammatory statements like, “Didya know that one in two college students and their non-college-attending peers experienced a psychiatric condition in the past year???”).

The paper also inquired into who sought out mental health services in the past 12 months (included MDs, PhDs, emergency rooms, crisis centers, self-help groups, etc.).

The paper then discusses the statistics, which I do not fully understand, which consistently leaves me feeling like I am hugely vulnerable to being duped.

The paper then goes through the findings; the “jackpot” is Table 2, which I cannot reproduce here because I don’t want to get a letter from a lawyer asking me to “cease and desist”. Basically, though, the media can report things like this as a result of this paper:

• “Nearly 1 in 2 college-aged people have experienced suffered experienced any psychiatric disorder in the past year!”
• “About 2 in 5 college-aged people have experienced a MAJOR mental disorder in the past year!”
• “Nearly one-third of college-aged people have substance-use disorders!”
• “1 in 10 college students have experienced a mood or anxiety disorder in the past twelve months!”

The AP already jumped on the 1 in 5 have a personality disorder. Again, I’m wondering why there’s a highlight on personality disorders.

Let’s take a look at DSM-IV in regards to personality disorders. You know, the text upon which the interview above was based.

• “The prevalence of Avoidant Personality Disorder in the general population is between 0.5% and 1.0%.”
• “Dependent Personality Disorder is among the most frequently reported Personality Disorders encountered in mental health clinics.” (No prevalence is reported.)
• “Studies that have used systematic assessment suggest prevalence estimates of Obsessive-Compulsive Personality Disorder of about 1% in community samples…”
• “The prevalence of Paranoid Personality Disorder has been reported to be 0.5%-2.5% in the general population…”
• “Schizoid Personality Disorder is uncommon in clinical settings.” (No prevalence is reported.)
• “Limited data from general population studies suggest a prevalence of Histrionic Personality Disorder of about 2%-3%.”
• “The overall prevalence of Antisocial Personality Disorder in community samples is about 3% in males and about 1% in females.”

If I assign 1% (which is very generous) to the disorders where no prevalence is given and take the highest percentages that are given, that gives a sum of 12.5%. That’s significantly less than the “1 in 5″ reported in the article.

In fact, all of the percentages listed in the article are much higher than that listed in the DSM. The only one that comes close is, to my surprise, histrionic personality disorder (the paper cited 3.5% of college students and 4.4% of non-college-attending peers meet criteria for it). DSM underwent a “text revision” in 2000. Does this mean that we have had an explosion in the prevalence of personality disorders in the past eight years? If so, what do we attribute this to? The water? the air? the President?

How do we relate the surge in the use of antipsychotic medications (and psychotropic medications in general) to this data? Who (what?) is the force driving this seeming pandemic of psychiatric conditions in the nation?

It’s not that I take issue specifically with the paper. It is an interesting study and included a lot of people. My issue is with the definition of “psychiatric condition”.

Instead of expressing umbrage about the high percentage of people with psychiatric conditions, why not reassess if the definition of “psychiatric conditions” means anything? If 50% of college-aged people “have something wrong with them”, perhaps we should scrutinize the words we use to define “wrong”. We should question why there is this huge discrepancy between DSM prevalences and those listed in the paper. (For what it is worth, personality disorders are generally diagnosed over time and cannot be diagnosed if someone is under the age of 18. This was a cross-sectional study—that is, a study that looked at a thin slice of time and did not follow people over the course of a year or two or ten. And, by the way, I’m also not warmly endorsing the DSM; I’m highlighting discrepancies.)

I’m not against the idea of mental health and hygiene. I’d rather focus and build upon strengths that are individually tailored. Focusing on general pathology that is ill-defined, however, is hugely problematic. One of the purposes of diagnosis is to guide treatment. If the process of diagnosis is unclear, the process of treatment is unclear. If the diagnosis itself is vague, the treatment is also hazy (hence the rash of psychotropic prescriptions). If the diagnosis encompasses too much of the variation of normal, we’re doing more harm than good. Whatever happened to “Primum non nocere“?

DSM-V is in the works (and cloistered, at that). I don’t know what anyone has to gain if 50% of the population has a mental disorder. Sure, the cynics will remark that shrinks and pharmaceutical companies will have lots of financial gain (because that’s all that matters, right?), but really, what pride is there is believing that 50% of the population has a mental disorder? And doesn’t that really say more about the culture than the “patients”?