Do subjective sleep quality and objective sleep measures improve to the same degree with CBT-I?

Why does how you feel about sleep improve more than the sleep itself?

CBT-I targets the thoughts and behaviors that surround sleep—worrying about not sleeping, spending too long in bed, and the mental habits that make bedtime stressful. Those targets sit close to the subjective experience of insomnia, so it makes sense that the felt sense of sleep improves first and most reliably. A systematic review of 15 studies found that most reported significant gains in sleep-diary measures such as sleep efficiency, time to fall asleep, and time awake during the night, along with better scores on self-reported sleep questionnaires [1].

The objective picture is different. In that same review, the direction, size, and statistical significance of change on objective sleep measures varied from study to study [1]. A randomized controlled trial with 62 people who had chronic insomnia found that insomnia severity dropped and self-reported sleep satisfaction rose after group CBT-I, yet there was no significant change in objective or subjective measures of sleep timing and duration, such as how long it took to fall asleep or how long people slept [2]. In other words, feeling better about sleep and sleeping measurably better are not the same thing.

The gap between felt sleep and measured sleep can itself shrink

People with insomnia often misjudge their sleep: they may believe they lay awake far longer than the recording shows. This mismatch is called sleep-state misperception. One trial found that the discrepancy between subjective and objective estimates of sleep latency and sleep duration narrowed after CBT-I, suggesting people came to perceive their sleep more accurately even though the underlying measured sleep did not change much [2].

A study of veterans with moderate-to-severe traumatic brain injury who used a computerized CBT-I program tells a similar story with a twist. Agreement between sleep diaries and wrist actigraphy was poor for all four sleep outcomes examined. At baseline, participants over-reported how long they took to fall asleep and how early they woke, while greatly under-reporting time awake during the night, which made their diary-based total sleep time look longer than actigraphy showed. After treatment, the over-reporting of time to fall asleep decreased, and the gap in sleep efficiency actually grew larger—because subjective sleep improved while objective sleep did not [5]. So the two measures can move apart even as the person feels better.

When do objective measures improve too?

Objective sleep is not always stubborn. In a pilot trial of CBT-I for 44 patients with primary brain tumors, sleep diaries showed broad improvements: less time awake after falling asleep, shorter time to fall asleep, higher sleep efficiency, fewer awakenings, and better perceived sleep quality. Actigraphy also showed significant improvements in number of awakenings, time to fall asleep, time in bed, and total sleep time—but not in time awake during the night or sleep efficiency [4]. That pattern suggests objective gains can appear, but they tend to be narrower and to cover fewer measures than the subjective gains. The authors note that larger controlled trials are underway to confirm these preliminary findings [4].

One 2025 study reports that CBT-I improves both subjective and objective sleep parameters and reduces the discrepancy between them, though the available summary of that work is fragmentary [3]. Taken together, the evidence points to a consistent asymmetry rather than a clean rule: subjective sleep quality is the measure most reliably improved by CBT-I, while objective sleep changes are real in some groups and on some measures but inconsistent across studies [1][2][4].

What should count as success in CBT-I?

The practical takeaway is that a person can genuinely benefit from CBT-I—feeling less distressed, more satisfied with sleep, and more accurate about how they slept—without their measured sleep architecture changing much. Because the two kinds of measurement can diverge, and can even move in opposite directions after treatment [5], researchers recommend tracking both subjective and objective sleep when evaluating insomnia treatment rather than relying on either alone [5]. The mismatch itself may be worth studying: the TBI study's authors suggest possible contributors such as cognitive impairment, brain activity patterns, or sleep-pressure regulation, but these remain hypotheses rather than established causes [5].

Sources used in this answer

Related research questions