How do digital CBT-I effects on daytime functioning compare with effects on nighttime sleep measures?

Nighttime sleep improves clearly against no treatment

The strongest evidence concerns sleep itself. A meta-analysis of fifteen randomized trials of digital CBT-I in adults with chronic insomnia found meaningful improvements over inactive controls: total sleep time rose by about 0.21 hours (roughly 13 minutes), the time taken to fall asleep fell by about 15.5 minutes, sleep efficiency improved by about 10.5 percentage points, time spent awake after falling asleep dropped by about 16 minutes, and the number of awakenings fell by about half an awakening per night [1]. These are self-reported diary measures, not lab recordings; none of the included trials supplied polysomnography data [1].

The same meta-analysis adds an important qualifier. When digital CBT-I was compared with an active control, such as in-person CBT-I or telehealth-delivered CBT-I, no significant differences appeared [1]. In other words, the large nighttime gains are relative to doing nothing or receiving minimal care, and the digital format appears roughly on par with therapist-delivered treatment rather than superior to it.

Daytime functioning is measured less often and less directly

Daytime outcomes are the weaker half of the comparison. The meta-analysis that anchors the nighttime numbers did not examine daytime functioning at all, so it cannot tell us whether the sleep gains carried over into waking hours [1]. That gap matters because a person can sleep longer by the diary and still feel foggy, irritable, or unfocused during the day.

What evidence exists is narrower. In poor sleepers with insomnia symptoms, digital CBT-I reduced attentional vigilance toward sleep-related cues, a marker of the mental preoccupation with sleep that can spill into daytime attention [2]. That is a specific cognitive measure rather than a broad picture of daytime energy, mood, or work performance, so it supports a daytime benefit without quantifying how large or how general it is.

Why daytime effects are harder to pin down

Part of the reason is measurement. Nighttime sleep can be captured with a diary or a sensor and summarized in minutes and percentages, which is why the meta-analysis could pool five sleep parameters across trials [1]. Daytime functioning is a looser bundle, including alertness, mood, attention, and daily performance, and different studies pick different pieces of it. The attentional-bias result [2] and the broader outcome sets in guided digital programs for university students [3] are not measuring the same thing, which makes them hard to add up into a single daytime effect size.

A related point is that daytime change may lag behind sleep change. If attention and mood improve only after sleep has stabilized, a short trial could show clear nighttime gains while daytime measures look flat. The available studies do not settle this timing question, so the daytime picture should be read as promising but incomplete rather than as evidence that daytime functioning does not improve.

What this means if you are considering a digital program

If your main concern is nighttime sleep, the case is reasonably strong: digital CBT-I outperforms no treatment on how long you sleep, how quickly you fall asleep, and how consolidated that sleep is, and it appears comparable to seeing a therapist [1]. If your main concern is daytime functioning, the evidence is encouraging but less complete. A reduction in sleep-related attentional bias is a real, measured change [2], and guided digital programs have been studied with daytime outcomes among them [3], but the size of the daytime benefit and how broadly it extends to energy, mood, and performance remain less firmly established than the nighttime numbers.

It is also worth keeping the comparison in perspective. Studies of other sleep interventions, such as sleep hygiene education, report improvements in both sleep quality and daytime functioning [5], which suggests daytime benefits are plausible for sleep treatments generally. But those results come from different interventions and populations, so they cannot be transferred directly to digital CBT-I. For now, the fair summary is that digital CBT-I reliably improves measured nighttime sleep and probably helps daytime functioning, with the daytime side of the ledger still being filled in.

Sources used in this answer

Related research questions