The historical DSIP report describes different sleep-response timelines in age groups that also differed in initial sleep disturbance. That makes a simple age-effect interpretation unsafe. Primary report
What was reported
Eighteen people with chronic psychophysiological insomnia were divided into middle-aged and older groups. The sleep-laboratory investigation included one treatment week and one follow-up week. The abstract reports improvement, with older participants starting with more severe disturbance and reaching the reported normal pattern later. It also links response magnitude with baseline disturbance. Primary report
Our appraisal: recover the comparator before making a causal claim
The abstract does not provide enough detail to reconstruct allocation, the comparator sequence, or all outcome definitions. Database indexing alone should not fill those gaps. Full-text methods are necessary before describing the study as a particular randomized placebo-controlled design.
Age and baseline severity can also point in similar directions. A difference between age groups may reflect one, the other, or a combination. A small regression analysis does not automatically settle that attribution, especially without its model assumptions and uncertainty.
Finally, “normal” requires an operational definition: which sleep measure, what reference range, and at which night? A short laboratory observation is not equivalent to durable relief in daily life, and daytime function may require separate measurement.
This record preserves the favorable historical observation while leaving design uncertainty visible. It does not establish that DSIP reliably treats insomnia or provides a safe long-term alternative to clinical care.
Frequently asked questions
How many participants were described?
Eighteen.
Did the age groups start with comparable sleep disturbance?
No. The older group had more severe initial disturbance.
Was follow-up long-term?
No. The abstract describes a short laboratory timeline.
Does the abstract fully reconstruct the controlled design?
No. Complete methods are needed.
Does “normal sleep” necessarily mean lasting symptom resolution?
No. The measurement and duration must be specified.
Can the age-group comparison isolate an age effect?
Not without adequately addressing baseline differences.
What remains before release?
Comparator, allocation, outcome-definition, safety, and publication-history review.