A post hoc analysis did not find significant differences in SHP465 treatment effects between baseline sleep-quality groups. Some subgroup comparisons were significant and others were not, but that pattern alone does not demonstrate that sleep modified the treatment effect.
The analysis used a seven-week randomized, placebo-controlled, dose-optimization trial in adults with ADHD. The intent-to-treat population included 268 participants. Baseline Pittsburgh Sleep Quality Index scores defined impaired and non-impaired sleep groups, including classifications based on individual sleep components.
Outcomes were ADHD symptom ratings and Brown scale executive-function ratings. This was secondary analysis of controlled human treatment data, rather than an experiment assigning people to sleep deprivation.
Most active-versus-placebo comparisons nominally favored SHP465 within the sleep groups. Exceptions included symptom ratings in the impaired sleep-efficiency subgroup, and executive ratings in impaired sleep-duration and sleep-efficiency subgroups.
However, treatment-effect differences between the sleep groups were not statistically significant. A smaller subgroup can fail to reach significance because its estimate is less precise. Comparing its p value with another subgroup’s p value does not replace a direct interaction test.
Sleep was self-reported and analyzed after the trial, with multiple subgroup definitions. The analysis does not establish that sleep problems are unimportant clinically, that treatment cures them, or that objectively measured sleep would produce the same findings.
The selected population, attrition and limited follow-up constrain generalization. The study was industry funded. Its contribution is a qualified negative moderation result, not a validated method for choosing a stimulant or predicting individual impairment from sleep complaints.
See why treatment schedules and cognitive endpoints matter.
Post-Hoc Analyses of the Effects of Baseline Sleep Quality on SHP465 Mixed Amphetamine Salts Extended-Release Treatment Response in Adults with Attention-Deficit/Hyperactivity Disorder. doi:10.1007/s40263-019-00645-z.