This is no longer a contested area of clinical practice. The following organizations have issued explicit recommendations placing CBT-I above pharmacotherapy as initial treatment for chronic insomnia disorder in adults, including older adults:
| Organization | Recommendation | Strength | Source |
|---|---|---|---|
| American College of Physicians | All adult patients should receive CBT-I as the initial treatment for chronic insomnia disorder. Pharmacotherapy should be added only via shared decision-making in patients where CBT-I alone was unsuccessful. | Strong — moderate-quality evidence | Annals of Internal Medicine, 2016 |
| American Academy of Sleep Medicine | Multicomponent CBT-I is the only treatment receiving a "strong" recommendation. All individual pharmacological agents receive only "weak" recommendations. Medication should be considered mainly in patients unable to participate in CBT-I or who remain symptomatic after it. | Strong for CBT-I; weak for all medications | AASM CPG, 2021 |
| American Geriatrics Society (Beers Criteria) | Benzodiazepines and Z-drugs (zolpidem, eszopiclone, zaleplon) should be avoided for insomnia in older adults. All increase risk of cognitive impairment, delirium, falls, fractures, and motor vehicle accidents. CBT-I is the indicated first-line approach. | Avoid — explicit Beers listing for all Z-drugs and BZDs in elderly | AGS Beers Criteria, 2023 update |
| World Sleep Society | Endorsed AASM strong recommendation for multicomponent CBT-I as the treatment of choice for insomnia disorder. Strong recommendation applies to patients with or without comorbid psychiatric and medical conditions. | Strong — international endorsement | Sleep Medicine, 2023 |
| European Insomnia Guideline | CBT-I recommended as first-line treatment. Pharmacotherapy only as adjunct or second-line. Guideline updated 2023 to reflect growing digital CBT-I evidence. | First-line recommendation | European Insomnia Guideline, 2023 |
The clinical consensus is unambiguous. CBT-I is the treatment of choice for chronic insomnia in older adults. The question is not whether to recommend it, but how to make it accessible to patients who need it.
CBT-I is among the most thoroughly investigated non-pharmacological treatments in sleep medicine. The evidence base now spans hundreds of randomized controlled trials and multiple large-scale meta-analyses, several published in 2024 and 2025.
A 2024 component network meta-analysis in JAMA Psychiatry synthesized 241 trials and 31,452 participants. Multicomponent CBT-I had the highest likelihood of remission at OR 2.50 (95% CI 1.93–3.24) compared to psychoeducation, outperforming all individual components and all passive controls. A December 2025 meta-analysis in Cognitive Behaviour Therapy — covering 32 real-world effectiveness studies and 5,231 participants — reported very large within-group effect sizes for insomnia severity at post-treatment (Hedges' g = 1.87) and follow-up (Hedges' g = 1.68, mean 12 months post-treatment), with remission rates of 45% at post-treatment and 51% at follow-up. These are real-world clinical care settings, not efficacy-optimized RCT conditions.
A November 2025 meta-analysis in JAMA Internal Medicine specifically examined CBT-I in 67 RCTs of patients with chronic disease comorbidities (cancer, chronic pain, cardiovascular disease, IBS, stroke) — a population directly relevant to older adults. CBT-I was associated with insomnia severity improvement of g = 0.98 (95% CI, 0.81–1.16), sleep efficiency g = 0.77, and sleep onset latency g = 0.64, with a mean dropout rate of 13.3% — comparable to pharmacological trials. Effect sizes were consistent across disease subgroups.
A July 2025 RCT in npj Digital Medicine, conducted at the University of Virginia, enrolled 311 participants aged 55–95 in a tailored digital CBT-I program. Both the self-guided and coach-supported arms showed significant improvements in insomnia severity, sleep onset latency, wake after sleep onset, sleep efficiency, and fatigue compared to controls at post-treatment, 6-month, and 12-month follow-up. This is among the most age-inclusive CBT-I trials conducted to date.
A 2025 secondary analysis in Aging & Mental Health pooled three digital CBT-I RCTs in adults aged 65 and older (N = 315). Digital CBT-I significantly improved insomnia, anxiety, and depression scores at post-treatment and at 24-week follow-up. These secondary mental health benefits are particularly relevant given the high comorbidity burden in this population.
A 2024 study in the Journal of the American Geriatrics Society evaluated CBT-I in community-dwelling older veterans (mean age 72.1 years) with chronic insomnia and found improvement in insomnia severity, sleep quality, fatigue, and daytime sleepiness — with patients experiencing higher chronic pain showing greater insomnia improvement, not less.
For prescribers evaluating program quality, the component-level evidence matters. A 2024 network meta-analysis in Clinical Psychology Review (80 studies, 15,351 participants) found that sleep restriction therapy was the most effective single component for insomnia severity (d = −0.45; 95% CI −0.63 to −0.36), improving subjective sleep continuity, sleep onset latency, WASO, and sleep efficiency. Stimulus control therapy was the most effective for improving total sleep time. Relaxation therapy, psychoeducation, and sleep hygiene alone showed no significant effects on insomnia severity as stand-alone components.
The clinical implication: a program that relies primarily on sleep hygiene education or relaxation techniques — regardless of how it is branded — is not delivering the treatment components with demonstrated efficacy. Effective CBT-I must include sleep restriction and stimulus control, both of which require patient adherence to protocols that are difficult to maintain without structured guidance.
| Dimension | CBT-I | Pharmacotherapy (Z-drugs, BZDs) | Pharmacotherapy (DORAs) |
|---|---|---|---|
| Onset of benefit | 2–3 weeks for initial improvement; full benefit 6–8 weeks | Days | Days |
| Long-term durability | Sustained and often improving at 12-month follow-up; one RCT showed sustained benefit at 10-year follow-up | Tolerance develops; insomnia returns on discontinuation; rebound worsening | Sustained while on medication; discontinuation outcomes not well established long-term |
| Fall / dizziness risk in older adults | None — no adverse event signal in RCTs | High: 3–4× increased fall risk; 1.92× relative risk of hip fracture; on AGS Beers list | Lower than Z-drugs; narcolepsy-like symptoms reported; somnolence dose-dependent |
| Cognitive effects | Improves cognitive performance via improved sleep quality | 5× increased memory loss, confusion, and disorientation in older adults; associated with Alzheimer's risk at high cumulative doses | Better cognitive profile than Z-drugs; less residual next-day impairment |
| Dependency risk | None | High — 68% of zolpidem users take it long-term despite short-term-only approval; median 192 days of supply in one cohort study | Low (Schedule IV); misuse potential present |
| Comorbidity benefit | Secondary improvements in depression (SMD −0.42) and anxiety (SMD −0.29); effective across chronic disease populations | May exacerbate depression; linked to increased suicidal ideation at OR 2.08 in one meta-analysis | Emerging data on anxiety benefits; early research on dementia populations |
| Supports deprescribing | Yes — concurrent CBT-I significantly increases successful hypnotic discontinuation rates; AASM guidelines specifically note CBT-I facilitates medication tapering | N/A | N/A |
The primary barrier to CBT-I in clinical practice is not evidence — it is access. Behavioral sleep medicine specialists are scarce; waitlists commonly run 3–6 months. Primary care physicians frequently lack the training time to deliver CBT-I themselves, and referring patients to sleep psychologists in many regions is simply not practical.
The AASM explicitly recommends digital CBT-I as a valuable resource when in-person care is unavailable or waitlists are prohibitive — followed by therapist-delivered CBT-I when access becomes available. A March 2025 meta-analysis in npj Digital Medicine covering 29 RCTs and 9,475 participants found fully automated digital CBT-I produces moderate to large effect sizes on insomnia severity — though it is less effective than therapist-assisted CBT-I, reinforcing the value of human guidance in the protocol.
The clinical distinction between digital CBT-I programs matters. Automated platforms deliver structured content but cannot adapt to individual patient presentations in real time. Coached platforms — where a human reviews the patient's sleep diary and adjusts the protocol week by week — more closely replicate the adaptive individualization of in-person treatment. This is particularly relevant for older adults, who often present with comorbidities, polypharmacy, and a concurrent medication taper that requires more nuanced management than a fixed algorithm can provide.
Sleep Reset is an appropriate referral for older patients presenting with chronic insomnia who need a structured, evidence-based behavioral treatment and where in-person CBT-I is unavailable, impractical, or delayed by waitlists. It is also well-suited for patients currently on sleep medication for whom deprescribing is a clinical goal.
Several features distinguish it from general digital wellness apps or basic automated CBT-I platforms:
| Feature | Sleep Reset | Typical automated digital CBT-I (Sleepio, etc.) | Meditation / wellness apps |
|---|---|---|---|
| CBT-I protocol (sleep restriction + stimulus control) | ✓ Yes — full multicomponent | ✓ Yes | ✗ No — relaxation only |
| Dedicated human sleep coach | ✓ Yes — assigned, reviews daily sleep diary | ✗ No — algorithm-driven | ✗ No |
| Board-certified sleep physician oversight | ✓ Yes — licensed sleep medicine clinicians | ✗ No | ✗ No |
| Home sleep testing for apnea | ✓ Yes — physician-interpreted | ✗ No | ✗ No |
| Supports concurrent medication taper | ✓ Yes — designed for concurrent use with prescribing physician | Partially — no clinical oversight | ✗ No |
| Access | Immediate — no waitlist | Immediate — no waitlist | Immediate |
| Cost / coverage | CBT-I program: out-of-pocket, HSA/FSA eligible. Clinical services (clinician visits, home sleep testing, apnea care): covered by select insurers including Aetna, BCBS, Anthem. Check eligibility. | Employer plans; SleepioRx via CMS DMHT codes (Jan 2025) | Subscription; not covered |
Sleep restriction is the most effective and most demanding CBT-I component. In older adults, the standard protocol requires modification: the minimum prescribed sleep window should not fall below 5.5 hours (versus the 5-hour floor used in younger adults), and titration should proceed more gradually — typically in 15-minute increments every 5–7 days rather than weekly. A human coach who understands the patient's baseline, comorbidities, and medication status is essential for calibrating this safely. The Sleep Reset approach to sleep restriction accounts for these age-related adjustments.
Sleep apnea is significantly underdiagnosed in older adults and frequently presents as insomnia — particularly sleep maintenance insomnia with daytime fatigue. Initiating CBT-I without ruling out obstructive sleep apnea in patients who present with relevant risk factors (male sex, obesity, witnessed apneas, hypertension, large neck circumference, or refractory insomnia) risks treating the wrong condition. Sleep Reset's integrated home sleep testing allows this to be evaluated concurrently rather than as a prerequisite that delays behavioral treatment.
The AASM clinical guidelines on insomnia management explicitly state that "medication tapering and discontinuation are facilitated by CBT-I" and recommend that "whenever possible, patients should receive an adequate trial of cognitive behavioral treatment during long-term pharmacotherapy." Starting CBT-I before or at the time of initiating a taper is superior to stopping medication first and hoping sleep improves. The prescribing physician manages the taper rate; Sleep Reset's coach manages the behavioral protocol — the two run in parallel under a shared patient understanding.
Standard CBT-I requires sufficient cognitive capacity to engage with sleep diaries, follow behavioral instructions, and process psychoeducation. Patients with mild cognitive impairment (MCI) can often complete modified CBT-I with appropriate support, but those with moderate or severe dementia require alternative approaches. A January 2026 pilot study in Brain Sciences developed and evaluated a clinician-supported mobile CBT-I app specifically designed for older adults with MCI, demonstrating feasibility and early efficacy signals. Patients with more significant cognitive decline should be evaluated individually; CBT-I is not contraindicated in MCI but may require clinical adaptation.
For many older patients, "therapy" for sleep raises immediate questions. The framing that tends to work clinically:
Patients can take a free sleep assessment at Sleep Reset to identify their specific insomnia profile and be matched with an appropriate coach before committing to the program.
Yes. The November 2025 JAMA Internal Medicine meta-analysis of 67 RCTs in chronic disease populations found CBT-I effect sizes for insomnia severity of g = 0.98 — comparable to results in otherwise healthy adults. Cancer, chronic pain, cardiovascular disease, IBS, and stroke were all represented. Efficacy was consistent across disease subgroups. Dropout rates were low (mean 13.3%). The recommendation against CBT-I in patients with comorbidities is not supported by the current evidence base.
CBT-I improves insomnia and has secondary benefits on depression and anxiety — a meta-analysis of digital CBT-I in patients with both insomnia and mood symptoms found SMD = −0.42 for depressive symptoms and SMD = −0.29 for anxiety symptoms at post-treatment, alongside a large effect on insomnia (SMD = −0.76). Treating insomnia directly may alleviate pressure on concurrent pharmacological management of mood disorders. In patients where the psychiatric presentation is primary and severe, concurrent specialist involvement is appropriate.
Sleep restriction can temporarily increase daytime sleepiness in the first 1–2 weeks. This is expected and is the mechanism of the treatment. In older adults, a minimum sleep window of 5.5 hours is recommended rather than 5 hours, and titration should be more gradual. There is no evidence of serious adverse events from CBT-I in elderly populations across the RCT literature — a significant contrast to the adverse event profile of pharmacotherapy in this group.
Multiple studies have demonstrated that concurrent CBT-I significantly increases successful hypnotic discontinuation rates compared to taper alone. The AASM and European Insomnia Guidelines both explicitly state that CBT-I facilitates medication tapering. A 2025 Lancet microsimulation study modeled the outcomes of eliminating prescription sleep medication in 15.3 million US adults over 50 — finding an 8.5% reduction in lifetime falls, 2.1% reduction in cognitive impairment, 0.11 additional life years, and $6,600 per-person net savings ($101 billion aggregate). CBT-I is the clinical mechanism by which that transition is made sustainable.
Dr. Areti Vassilopoulos | Psychologist | Sleep Medicine Expert
Dr. Vassilopoulos is the Clinical Content Lead for Sleep Reset and Assistant Professor at Yale School of Medicine. She has co-authored peer-reviewed research articles, provides expert consultation to national nonprofit organizations, and chairs clinical committees in pediatric health psychology for the American Psychological Association. She lives in New England with her partner and takes full advantage of the beautiful hiking trails.
Sleep Reset © 2025 All Rights Reserved
.jpeg)
.jpeg)
.jpeg)
.jpeg)
