The entire field of psychiatric epidemiology has a systematic bias that leads it to misleadingly report what are highly inflated rates of psychiatric disorder. Psychiatric symptoms in mild form are widely distributed in the general population—from time to time, almost everyone will have some depression or anxiety. But symptoms alone do not define psychiatric disorder. As indicated in an earlier blog, the symptoms must also cause clinically significant distress or impairment. Epidemiologic studies all routinely ignore this crucial requirement. They mistakenly diagnose as psychiatric disorder symptoms that are mild, transient, and lacking in clinical significance.
How can an entire field of scientific endeavor go so far astray? The reason for the error comes down to simple dollar and cent cost considerations. Epidemiologic studies sampling the general population require very large n's with the performance of many thousands of interviews. It would be too expensive to employ clinicians in so extensive an endeavor—so the studies rely on the cheap labor provided by lay interviewers who have no clinical experience and no discretion in judging whether symptoms are clinically meaningful. They make their diagnoses of psychiatric disorders based on symptom counts alone with no consideration of whether the symptoms are severe enough to warrant diagnosis or treatment. Results generated in this rough and ready way should be considered as no more than an upper limit on the rate of disorder. They should never be taken at face value as a true reflection of the real rates of mental disorder in the community.
But the exaggerated rates are always reported without proper caveat and are accepted at face value as if they are an accurate reflection of the real prevalence of psychiatric disorder. This is ludicrous, especially since the methods used have become much looser with time. Psychiatric epidemiology is a young field that emerged only 30 years ago with the development of diagnostic criteria. The first large, systematic study was (the Epidemiologic Catchment Area) found fairly high rates of disorder, but in the next generation these bubbled further (sometimes almost doubling) due to the use of even less rigorous methodology. The reported diagnostic inflation did not reflect any real change in the mental health of America—just greater carelessness in the application of the tools used to measure it. We are not suffering from an epidemic of psychiatric disorder—instead we are being misinformed by the results of misleading science.
Were one to believe NIMH, almost 25% of the general population of the US has a mental disorder in any given year. And this would have to be considered an underestimate since it is being measured retrospectively. Prospective studies suggest even higher rates of lifetime disorder. But all this is based on the fantasy that the lay interviewers can make accurate diagnoses using simple symptom counts. They can't.
So my advice is never to take seriously the rates reported by any epidemiological study. Be aware that they are always overestimates including people who may have the required symptom pattern but do not necessarily have sufficient severity, duration, distress, or impairment to be considered mentally ill. The dirty little secret of epidemiological research is that it can report only upper limits, not true rates—which would almost certainly be much lower especially for the milder conditions at the populous boundary with normality. Epidemiological research is necessary and valuable, but its limitations must be acknowledged and its results taken with more than a grain of salt.