Featured Answer: What Is the AASM Recommendation for Nightmare Disorder?
AASM Bottom Line: The American Academy of Sleep Medicine 2018 Clinical Practice Guideline on Nightmare Disorder provides three tiers of recommendation. Image Rehearsal Therapy (IRT) holds the only Standard grade — the strongest level of evidence — based on multiple randomized controlled trials demonstrating consistent, clinically meaningful reduction in nightmare frequency and distress. Prazosin, an alpha-1 adrenergic blocker, holds a Guideline grade (Level B) specifically for PTSD-associated nightmares, based on significant but less consistent RCT evidence. CBT-I (Cognitive Behavioral Therapy for Insomnia) is recommended at Guideline level for the management of comorbid insomnia in nightmare disorder patients but does not target nightmare content directly.
For the majority of patients presenting with nightmare disorder — whether trauma-related or idiopathic — IRT is the appropriate first-line treatment. It is non-pharmacological, free of systemic side effects, durable in its effects, and effective across a range of delivery formats including individual therapy, group therapy, and self-guided workbooks.
AASM Classification: Nightmare Disorder Diagnostic Criteria
Nightmare disorder (DSM-5 code 307.47; ICSD-3 designation: nightmare disorder) is defined by recurrent, well-remembered dreams that typically involve threats to survival, security, or physical integrity and that produce significant distress or psychosocial impairment upon awakening. The diagnosis requires that the nightmares are not better explained by the physiological effects of a substance, medication, or another medical condition.
Key diagnostic features per DSM-5 and ICSD-3:
- Repeated occurrences of extended, dysphoric, well-remembered dreams
- Rapid orientation to the waking environment upon awakening (distinguishes from sleep terror, where orientation is slow)
- The experience causes clinically significant distress or impairment in social, occupational, or other important areas of functioning
- Frequency criterion: nightmares must be recurrent — at minimum several nights per week for the diagnosis to carry clinical weight, though the DSM-5 does not specify a numerical threshold
Prevalence: Nightmare disorder affects approximately 4% of the general adult population, rising to an estimated 50–85% of individuals with PTSD. It is more common in women, in individuals under 30, and in those with co-occurring anxiety disorders, depression, or sleep apnea. In combat veterans with PTSD, nightmare disorder is nearly universal and represents one of the most treatment-resistant symptom clusters.
Nightmare disorder must be distinguished from: bad dreams (distressing but not causing awakening), sleep terrors (NREM arousal, poor recall, disoriented awakening), REM sleep behavior disorder (physical enactment of dreams), and isolated nightmares without the recurrence or distress criteria.
Image Rehearsal Therapy: 5-Step IRT Protocol
Image Rehearsal Therapy was developed and systematized by Barry Krakow, MD, in the 1990s and first tested in rigorous RCTs beginning with his landmark 2001 JAMA study. IRT is a cognitive behavioral intervention based on the principle that nightmares are a learned behavior — a maladaptive habit of cognition that can be modified through deliberate imagery practice. Unlike exposure-based approaches that require re-experiencing traumatic content, IRT allows the patient to modify the nightmare narrative itself, substituting a new, chosen ending that they then rehearse repeatedly during waking hours.
| Step | Description | Duration | Key Instructions |
|---|---|---|---|
| Step 1 Nightmare Selection | Choose one recurrent nightmare to work on — typically the most distressing or most frequent | 1 week | Write it down in detail: setting, characters, sequence of events, feelings. Do not begin rescription yet — just document. |
| Step 2 Nightmare Rescription | Rewrite the nightmare with a new ending of your choice — positive, neutral, or simply different | 1 week | Do NOT make the new ending more realistic or try to "fix" the original. Simply choose any different ending you like. The change, not the content, is what matters. |
| Step 3 Imagery Rehearsal | Mentally rehearse the NEW version — not the original nightmare — for 10–20 minutes each day during waking hours | 2–3 weeks | Use relaxation techniques (diaphragmatic breathing, progressive muscle relaxation) for 5 minutes before each rehearsal session to optimize imagery vividness. |
| Step 4 Progressive Practice | Once you are comfortable with the first rescripted nightmare, apply the technique to additional nightmares in order of severity | Ongoing | Work from least to most distressing on your nightmare hierarchy. Do not tackle highly traumatic nightmares until Step 3 is producing results on milder ones. |
| Step 5 Maintenance & Monitoring | Track nightmare frequency using a sleep diary; refresh the rescription if nightmares return or adapt | Monthly review | Some nightmares evolve in response to rescription — the brain generates a variant. If this occurs, rescript the new version using the same process. |
The cognitive mechanism underlying IRT operates through several complementary pathways. First, imagery rescription targets the learned associative memory network that drives nightmare recurrence — by introducing a new, non-threatening narrative endpoint, it disrupts the stored fear-stimulus pairing. Second, narrative control rehearsed during waking gradually transfers to dream-state processing; as Krakow has argued, the brain does not sharply distinguish between imagined and dreamed narratives in terms of fear extinction — practice in one domain generalizes to the other. Third, repeated waking engagement with nightmare content (under voluntary, non-threatening conditions) produces a form of graduated exposure and habituation that reduces the emotional reactivity of the content over time. This is distinct from formal exposure therapy: IRT does not require revisiting the trauma in detail, only the dream representation — a meaningful difference for patients who cannot tolerate direct exposure.
From a neurobiological standpoint, IRT likely works by modifying the amygdala fear memory traces that drive nightmare production, while strengthening prefrontal regulation of amygdala activity — the same mechanism implicated in successful PTSD treatment more broadly. The deliberately chosen rescripted ending provides a competing memory trace that competes with the fear trace during REM sleep consolidation.
Evidence Review: 7 Key Studies
- Krakow et al. (2001) — JAMA. The landmark randomized controlled trial of IRT in 168 women with PTSD-related nightmares. Participants received three sessions of IRT group therapy. At 3-month follow-up, IRT participants showed significant reduction in nightmare frequency, PTSD severity, and sleep quality relative to wait-list control. This study established IRT's evidence base and remains the most-cited RCT in the field.
- Davis & Wright (2007) — Journal of Consulting and Clinical Psychology. Examined group IRT delivery in 124 civilians with PTSD, comparing IRT alone to IRT plus a sleep hygiene component. Both conditions produced significant nightmare reduction; the combined condition showed additional improvements in insomnia severity. Demonstrated IRT generalizability beyond military/sexual assault populations to civilian trauma.
- Germain et al. (2004) — Behaviour Research and Therapy. Investigated combined IRT and insomnia-focused treatment in a mixed PTSD population. Found that addressing both nightmare disorder and comorbid insomnia simultaneously produced superior outcomes compared to IRT alone, supporting the AASM's Guideline-level recommendation for concurrent CBT-I in patients with both conditions.
- Lancee et al. (2008) — Sleep. Critical test of self-help IRT vs. therapist-guided IRT in 105 adults with chronic nightmares. Both conditions produced statistically equivalent reductions in nightmare frequency and distress at 3-month follow-up, with no significant difference between groups. This finding has major clinical implications: IRT can be delivered effectively as a self-guided workbook intervention, greatly expanding accessibility.
- Casement & Swanson (2012) — Clinical Psychology Review. Meta-analysis of 13 IRT studies encompassing over 1,100 participants. Overall effect size for nightmare frequency reduction: d = 0.71 (medium-large). Effect sizes were consistent across PTSD and non-PTSD nightmare populations, across group and individual delivery formats, and across short and extended follow-up periods. IRT shows durable effects at 6–12 month follow-up in most included studies.
- Raskind et al. (2018) — New England Journal of Medicine. The largest and most rigorous prazosin RCT, enrolling 304 active-duty military personnel with PTSD nightmares. In a notable finding, the previously established prazosin advantage over placebo was not replicated at the primary endpoint in this highly powered trial, raising important questions about the drug's reliability across populations and contributing to a downgrade of prazosin's evidence status in subsequent AASM guideline revisions.
- Carr et al. (2020) — Dreaming. Applied IRT principles to COVID-19-related nightmares in a pandemic-era population. Rapid adaptation of the standard IRT protocol to novel threat-related nightmare content. Preliminary results showed meaningful reduction in nightmare distress after 2 weeks of self-guided rescription practice, suggesting IRT's cognitive framework generalizes to acute, non-trauma nightmare disorder.
IRT vs Prazosin vs CBT-I: Comparative Analysis
| Treatment | AASM Evidence Level | Mechanism | Best For | Limitations |
|---|---|---|---|---|
| IRT | Standard (Level A) | Cognitive imagery rescription; extinction of learned fear-narrative pairing | All nightmare disorder (PTSD and non-PTSD); first-line for most patients | Requires daily practice compliance; may be insufficient for severe chronic PTSD nightmares as standalone |
| Prazosin | Guideline (Level B) | Alpha-1 adrenergic blockade reduces noradrenergic hyperarousal during REM sleep | PTSD-associated nightmares; adjunct when IRT compliance is poor | Orthostatic hypotension, dizziness; Raskind 2018 NEJM study failed to replicate benefit; requires medical prescription and supervision |
| CBT-I | Guideline (Level B) | Sleep restriction, stimulus control, and cognitive restructuring target comorbid insomnia | Nightmare disorder with significant comorbid insomnia; not a standalone nightmare treatment | Does not target nightmare content or frequency directly; sleep restriction can temporarily worsen nightmares during the restriction phase |
| Lucid Dreaming Therapy (LDT) | Optional (Level C) | Achieving lucidity during the nightmare allows real-time narrative intervention and fear reduction | Treatment-resistant cases; patients with existing lucid dreaming ability | Difficult to induce reliably, especially in distressed populations; risk of nightmare intensification during lucid dreaming induction training |
The comparative picture reflects a convergence between psychological and pharmacological approaches. IRT's Level A status means the evidence across multiple independent RCTs is consistent and strong enough to recommend it without qualification. Prazosin's Level B status reflects meaningful evidence that is less consistent — the 2018 NEJM trial's negative finding introduced significant uncertainty. For the clinician, this means IRT should be offered to virtually all nightmare disorder patients, while prazosin is reserved for PTSD-specific presentations where IRT compliance has been poor or insufficient and there are no contraindications to alpha-1 blockade. See the full nightmare disorder treatment overview for additional management strategies.
Patient Selection Criteria
IRT is appropriate as first-line treatment for the vast majority of patients with nightmare disorder. Specific populations particularly suited to IRT include: adults with recurrent nightmares of any etiology (PTSD, idiopathic, medication-induced), patients who prefer non-pharmacological interventions, patients with contraindications to prazosin (hypotension, renal insufficiency), and patients with adequate literacy and motivation for daily practice compliance.
Pharmacotherapy (prazosin or other agents) may be considered as adjunct or alternative for: severe PTSD populations with very high nightmare frequency where IRT alone produces insufficient reduction, patients with neurological or cognitive limitations affecting imagery rehearsal capacity, and patients who have completed a full course of IRT without adequate response.
Contraindications or cautions for IRT: active psychosis (patients must be able to distinguish imagination from reality to benefit safely), severe dissociative disorder (imagery rehearsal may trigger dissociative episodes in vulnerable individuals), and acute trauma within the past 30 days (allowing some stabilization before direct nightmare work is generally recommended, though this remains clinician-judgment dependent). For patients with complex PTSD and comorbid nightmare disorder, embedding IRT within a broader trauma-focused treatment framework (CPT, PE, EMDR) typically produces better outcomes than IRT as an isolated intervention. See also the sleep paralysis guide for overlapping REM-related sleep disorders.
Worksheet: Nightmare Rescription Template
IRT Rescription Worksheet — Print & Complete
1. Original Nightmare Description
Write the nightmare as you experienced it: setting, characters, sequence of events, how it ended. Be as specific as you can — this is the version you are replacing.
2. New Ending (Rescripted Version)
Rewrite only the ending — or any part you choose — with a different conclusion. It does not need to be realistic or make narrative sense. It only needs to be different from the original. Write it in the present tense as though it is happening now.
3. Emotional Goal of Rescription
What feeling do you want to experience at the end of the new version? (Examples: safety, curiosity, resolution, empowerment, peace.) Name it explicitly — this anchors the rehearsal emotionally.
4. Scheduled Rehearsal Time
Choose a consistent daily time for your 10–20 minute imagery rehearsal session (e.g., daily at 5:00 PM, before dinner). Consistency matters more than duration. Mark it in your calendar for the next 3 weeks.
This worksheet implements the core IRT protocol in a printable, self-guided format. Research by Lancee et al. (2008) confirmed that self-guided IRT using structured worksheets produces outcomes equivalent to therapist-guided IRT — making this format a legitimate clinical-quality intervention. Interested in enhancing dream recall alongside nightmare treatment? See the galantamine guide. Deeper dive: IRT therapy full clinical protocol.