Does digital cognitive behavioral therapy have enough long-term follow-up evidence?

Does digital CBT for insomnia hold up over the long term?

Yes, the evidence is strongest here. Multiple randomized trials show that digital CBT for insomnia (dCBT-I) produces improvements that last 6 to 12 months after treatment. In one large trial with 238 participants, dCBT-I led to a large reduction in insomnia severity at 8 weeks, and those gains were sustained at both 6- and 12-month follow-ups [10]. Another study using a smartphone app in China found that participants who received dCBT-I had significantly lower insomnia scores than a sleep-education group at 3 months and maintained better sleep efficiency (87.5% vs. 78.1%) and longer total sleep time at 6 months [4].

A retrospective study of over 4,000 patients compared dCBT-I to medication and found that dCBT-I was more effective than medication alone at 6 months, though the authors noted that benefits 'fluctuated' after the first 3 months, suggesting that some people may need a booster or combination approach [5]. Importantly, a pilot study found that a second course of dCBT-I helped 62.5% of patients who hadn't fully responded to the first course achieve remission at 3-month follow-up [13]. This suggests that for insomnia, long-term success is achievable, but it may require persistence or tailored support for those who don't improve initially.

What about depression, OCD, smoking, and other conditions?

For depression, a comprehensive meta-analysis of 15 randomized controlled trials (over 2,000 participants) found that internet-based CBT (iCBT) self-help programs were superior to control conditions at follow-up points ranging from 8 weeks to over a year, with a medium effect size [3][11]. Importantly, within the iCBT groups, symptoms continued to improve slightly from post-treatment to follow-up, meaning the benefits didn't just hold—they got a little better over time [3][11].

For obsessive-compulsive disorder (OCD), a 24-month follow-up of guided iCBT found that remission rates actually increased from 48% right after treatment to 76% at 24 months, suggesting that skills learned during therapy continue to pay off [1]. For smoking cessation, a randomized trial of a clinician-assisted digital CBT program found that 22.6% of participants remained abstinent at 52 weeks, compared to 13.2% in a brief-advice control group—a 71% higher likelihood of staying quit [7]. For atopic dermatitis (eczema), a 12-week internet-delivered CBT program produced moderate-to-large reductions in symptoms that were sustained at 12-month follow-up [6].

However, the evidence is not uniformly strong. A systematic review of digital CBT for obesity found that while initial weight loss was promising (up to -7 kg), user engagement dropped sharply over time, and long-term results were best when digital tools were combined with occasional human coaching [9]. Similarly, a study of digital CBT for chronic pain in primary care found only 'modest but sustained' improvements in pain impact at 12 months, and it did not reduce opioid use [12].

What predicts whether someone will benefit long-term?

Not everyone responds equally, and the research points to a few key factors. First, people with more severe symptoms at the start may need extra support: a study of older adults with insomnia found that those with higher baseline insomnia severity were significantly less likely to make sufficient progress after two months of dCBT-I [8]. Second, engagement matters—participants who completed more of the digital program saw greater improvements [2][9]. For example, in a trial for Black women with insomnia, those who completed the full digital program had a 10.4-point reduction in insomnia severity, compared to 6.2 points for non-completers [2].

Third, the type of delivery influences long-term durability. Blended models—digital programs with some human guidance (e.g., brief therapist check-ins or group sessions)—tend to show stronger and more sustained effects than fully automated programs [1][7][9]. For instance, the OCD study that showed 76% remission at 24 months used guided iCBT with therapist support [1], and the smoking cessation study that found 52-week benefits included clinician assistance [7]. Finally, cultural tailoring can improve engagement: a version of dCBT-I adapted for Black women had a 78% completion rate versus 65% for the standard version, and completion was linked to better outcomes [2].

About These Sources

This answer is built on 13 peer-reviewed studies — published from 2021 to 2026, 4 from 2024 or later, 10 in Q1 journals, collectively cited 378 times — selected as the most relevant from 15 studies that passed quality screening, drawn from 65 papers retrieved from a database of over 500 million.

Sources used in this answer

Related research questions