To the Editor: Dr Morin and colleagues1 carried out a difficult study that adds to our
knowledge and our management of insomnia. The authors discuss certain limitations
inherent in drawing their total sample from community volunteers, but they
fail to address the serious effects of this cohort choice on our ability to
compare and evaluate the therapeutic interventions.
In the general population, most people with insomnia who seek medical
care are exposed to a hypnotic, whereas few have experienced the detailed,
time-consuming cognitive-behavioral therapy (CBT). Most practitioners have
a number of stable, effectively treated patients who safely and comfortably
remain taking the same low dose of a hypnotic for decades. The huge majority
of these pharmacological "successes" in the community would be highly unlikely
to volunteer for a new approach. With this background, the volunteer cohort
of Morin and coworkers' study likely included almost exclusively (1) those
who had previously "failed" or are presently dissatisfied with pharmacological
management or (2) those who had never been so exposed. The study cohort, therefore,
was composed of about 77% who were sufficiently dissatisfied with pharmacological
therapy and chose to seek other help (most or all of the 60 of 78 subjects
with past or current use of "sleep aid" as shown in Table 1). Probably close
to 100% had never been exposed to CBT and would have no adverse predisposition
to such therapy. Such a sample mix would be destined to skew the results toward
a higher success rate for any reasonable intervention other than drugs.