Does CBT-I improve objective cognitive test performance or only self-reported cognitive difficulties?

CBT-I fixes the sleep problem, not necessarily the thinking problem

CBT-I is a structured talk therapy for insomnia: it changes sleep habits, schedules, and the anxious thoughts that keep people awake. The strongest and most consistent effect is on how people experience their sleep. In the 62-person randomized trial, insomnia severity dropped and self-reported sleep satisfaction rose after eight weekly group sessions, while measured sleep latency and duration did not change [1]. That split is the key to the cognition question: CBT-I works largely on the subjective layer of sleep, so it is reasonable to expect its cognitive benefits to land on the subjective layer too.

The same trial found that the gap between how long people thought they slept and how long they actually slept narrowed after treatment, suggesting a more accurate perception of sleep [1]. That is a real gain, but it is a gain in perception, not in raw mental horsepower. When the researchers then tested attention and working memory directly, and asked participants about cognitive difficulties with a questionnaire, neither measure improved [1]. The therapy changed the person's relationship to their sleep without changing their test scores.

On timed lab tasks, the scores mostly hold steady

Objective cognitive testing means timed, scored tasks, such as pressing a button when a target appears or holding digits in mind. Across the studies that measured these outcomes, the pattern is flat. The randomized trial found no significant change in attention and working-memory performance three months after CBT-I compared with a wait-list group [1]. A trial in older adults measured speed, executive function, and memory after CBT-I, but its abstract does not report a clear cognitive improvement [3].

A trial in cancer survivors with insomnia compared acupuncture with CBT-I and assessed objective memory with the Buschke Selective Reminding Test alongside a self-report attention scale [4]. And in patients who had both insomnia and sleep apnea, objective cognitive performance showed no meaningful change after CBT-I, and the title itself frames the outcome as worse rather than better [5]. These are different populations and different tests, yet they point the same way: clearing insomnia does not reliably sharpen performance on a stopwatch.

Feeling sharper is not the same as testing sharper, and even that can fail

It would be tidy if CBT-I improved self-reported cognition while leaving objective scores alone. The evidence is messier. The 62-person trial measured self-reported cognitive impairment with a dedicated questionnaire and found no improvement there either, even though sleep perception improved [1]. So the subjective benefit is well documented for sleep itself, but it does not automatically spill over into feeling mentally sharper.

That matters for anyone hoping CBT-I will fix brain fog. The therapy's demonstrated strength is in sleep perception and insomnia severity [1]. A systematic review set out to examine CBT-I's effect on cognitive performance using both questionnaires and cognitive tests, noting that the impact of CBT-I on cognition had yet to be properly evaluated [2]. That framing is a caution: the field's evidence base on cognition is thin and mixed, so a confident claim in either direction outruns what has been shown.

What to expect if you are doing CBT-I for brain fog

If your goal is to sleep better and feel less distressed about sleep, CBT-I has good support: insomnia severity and sleep satisfaction improved in a randomized trial [1]. If your goal is a measurable jump in attention or memory, the current evidence does not promise one. Objective attention and working-memory scores did not improve after treatment [1], and trials in older adults and in people with insomnia plus sleep apnea likewise did not show meaningful cognitive gains [3][5].

The practical reading is that CBT-I treats insomnia, and better sleep may still matter for how you feel day to day, but the cognitive payoff on formal tests is not established. One open question is whether the flat results reflect a true absence of effect or tests that are not sensitive to the subtle, everyday changes people notice. The trial that measured both objective tasks and a self-report questionnaire and found neither improved [1] leans toward a genuine null, but the systematic review's observation that this question had barely been studied [2] means the book is not closed.

Sources used in this answer

Related research questions