Non-Restorative Sleep: Drugs That Cause It, Drugs That Treat It, and What Else to Do

What Non-Restorative Sleep Actually Means

Non-restorative sleep is the experience of waking up tired, mentally foggy, or physically unrefreshed after a full night's sleep. The complaint is distinct from insomnia, where the core problem is falling or staying asleep. A person can sleep eight hours and still report non-restorative sleep, which makes it diagnostically tricky.

The 2014 International Classification of Sleep Disorders, Third Edition (ICSD-3), lists unrefreshing sleep as a criterion for several diagnoses, including insomnia disorder, fibromyalgia-related sleep dysfunction, and idiopathic hypersomnia. The American Academy of Sleep Medicine (AASM) notes that unrefreshing sleep is among the most common sleep-related complaints presenting to primary care, reported by roughly 1 in 10 adults in large population surveys. (AASM clinical practice guideline, 2017)

Why Sleep Quality Differs from Sleep Quantity

Total sleep time does not capture what happens inside each sleep cycle. Restorative sleep depends on adequate slow-wave sleep (SWS, also called N3) and REM sleep. SWS drives growth hormone secretion and physical recovery. REM supports emotional memory consolidation. Drugs, hormonal shifts, and primary sleep disorders can shorten or fragment these stages even when total sleep time looks normal on a diary.

The Role of Sleep Architecture

A standard overnight polysomnogram divides sleep into N1, N2, N3, and REM. Healthy adults spend roughly 13-23% of total sleep time in N3 and 20-25% in REM. (Ohayon et al., 2004, Sleep) When N3 drops below 10% of total sleep time, morning fatigue scores rise significantly, independent of total sleep duration.

Drugs That Cause Non-Restorative Sleep

Several widely prescribed and over-the-counter agents directly suppress slow-wave or REM sleep, producing the subjective experience of waking unrefreshed even after a full night.

Beta-Blockers

Propranolol and other lipophilic beta-blockers cross the blood-brain barrier and block melatonin synthesis by inhibiting beta-1 adrenoceptors in the pineal gland. A crossover trial published in Sleep (Brismar et al., N=15) found propranolol reduced melatonin by approximately 48% compared to placebo and shortened REM latency while increasing REM fragmentation. (Brismar et al., 1988, Sleep) Atenolol, a hydrophilic alternative, produces less central nervous system penetration and is preferred when a beta-blocker is clinically required in a patient with sleep complaints.

Benzodiazepines and Z-Drugs

Triazolam, temazepam, zolpidem, and zaleplon all enhance GABA-A receptor activity. They reduce sleep onset latency but simultaneously suppress N3 slow-wave sleep. (Gottesmann, 2002, Neuroscience) Patients who take zolpidem nightly for 30 or more days often report feeling "groggy, not rested" in the morning, a phenomenon consistent with SWS suppression and next-day residual sedation. The FDA added a black-box warning to zolpidem extended-release in 2013 specifically citing next-morning impairment.

SSRIs and SNRIs

Fluoxetine, sertraline, venlafaxine, and most other antidepressants in this class suppress REM sleep and increase the number of brief awakenings per night. (Wilson and Argyropoulos, 2005, Drugs) This is not a reason to discontinue antidepressants in patients who need them, but clinicians should recognize the mechanism when a patient on stable antidepressant therapy complains of unrefreshing sleep. Switching to bupropion, which does not suppress REM, is an option worth discussing with a prescribing psychiatrist.

Antihistamines (Diphenhydramine, Hydroxyzine)

Diphenhydramine, found in Benadryl and most OTC sleep aids (ZzzQuil, Unisom SleepTabs), blocks H1 histamine receptors and induces sedation. That sedation does not translate to restorative sleep. Tolerance to the hypnotic effect develops within four consecutive nights, yet anticholinergic side effects including morning cognitive fog, dry mouth, and urinary retention persist. (Rickels et al., 1983, J Clin Pharmacol) For older adults (age 65+), the American Geriatrics Society Beers Criteria explicitly lists diphenhydramine as potentially inappropriate due to anticholinergic burden and fall risk.

Alcohol

A single alcoholic drink within two hours of bedtime increases N3 in the first half of the night and then produces a rebound in light sleep and REM in the second half. (Ebrahim et al., 2013, Alcohol Clin Exp Res) The net result is fragmented early-morning sleep and non-restorative awakening. The effect is dose-dependent: high-dose alcohol (defined in that meta-analysis as >1 g/kg) suppresses REM across the entire night.

Corticosteroids

Prednisone and dexamethasone activate the hypothalamic-pituitary-adrenal axis and raise cortisol, which is normally near its nadir at midnight. A case series of 23 patients on chronic prednisone (mean dose 20 mg/day) documented polysomnographic N3 suppression and increased stage 1 sleep in all subjects. When corticosteroids cannot be reduced, melatonin 0.5-3 mg at bedtime may partially offset the circadian disruption.

Stimulants and Decongestants

Pseudoephedrine, methylphenidate, and amphetamine salts all delay sleep onset and reduce total sleep time. Even when a patient on a stimulant achieves adequate total hours, the arousal threshold remains elevated throughout the night, fragmenting deeper sleep stages. Morning dosing of stimulant medications reduces but does not eliminate the sleep architecture impact.

Hormonal Causes of Non-Restorative Sleep

The endocrine system drives sleep architecture at multiple points. Non-restorative sleep is common in the context of hormonal imbalance, and treating the underlying hormonal disorder often resolves the sleep complaint without any additional sleep-specific medication.

Low Testosterone in Men

Testosterone levels follow a circadian rhythm, peaking during REM sleep. When testosterone is low (typically <300 ng/dL by most laboratory reference ranges), REM sleep duration decreases. A randomized trial of 67 men with late-onset hypogonadism found that 6 months of testosterone undecanoate therapy improved subjective sleep quality scores by 22% compared to placebo. (Shigehara et al., 2011, Aging Male) Testosterone replacement does not treat obstructive sleep apnea and can worsen OSA at supraphysiologic doses.

Low Progesterone and Perimenopause

Progesterone has GABA-A agonist properties through its metabolite allopregnanolone. Declining progesterone during perimenopause directly reduces sleep continuity and N3 amplitude. The SWAN Sleep Study (N=3,045) found women in late perimenopause had a 40% higher likelihood of reporting frequent waking compared to premenopausal controls, independent of vasomotor symptoms. (Kravitz et al., 2008, Sleep) Micronized progesterone (Prometrium) 100-200 mg at bedtime may improve sleep quality independent of its estrogenic effects.

Thyroid Dysfunction

Both hypothyroidism and hyperthyroidism disrupt sleep. Hypothyroid patients show reduced N3 and increased sleep fragmentation; hyperthyroid patients typically report difficulty falling asleep and reduced total sleep time. A TSH outside the 0.5-4.5 mIU/L range warrants treatment before attributing non-restorative sleep to a primary sleep disorder.

How Non-Restorative Sleep Is Diagnosed

Diagnosis starts with a structured history. The clinician needs total time in bed, estimated total sleep time, sleep efficiency (time asleep/time in bed x 100), number of awakenings, and the subjective refreshment rating on waking.

Validated Questionnaires

The Pittsburgh Sleep Quality Index (PSQI) score >5 indicates poor sleep quality. The Epworth Sleepiness Scale (ESS) score >10 suggests daytime hypersomnia that may accompany non-restorative sleep. The Insomnia Severity Index (ISI) score >14 indicates moderate-to-severe insomnia. All three are freely available and have published psychometric validation.

Wrist Actigraphy

A 2-week actigraphy study provides objective sleep-wake timing data at home. The AASM considers actigraphy appropriate for evaluating circadian rhythm disorders and for quantifying treatment response, though it cannot stage sleep. (AASM, 2015 Actigraphy Practice Parameters)

When to Order Polysomnography

Polysomnography (PSG) is indicated when the history suggests obstructive sleep apnea (snoring, witnessed apneas, neck circumference >17 inches in men), periodic limb movement disorder, or when non-restorative sleep fails to respond to 8 weeks of CBT-I plus pharmacotherapy. An apnea-hypopnea index (AHI) >5 on PSG with associated symptoms meets criteria for OSA, which is a common missed cause of non-restorative sleep. (AASM diagnostic criteria for OSA, 2014)

First-Line Treatment: CBT-I

Cognitive behavioral therapy for insomnia is the AASM's first-line treatment for chronic insomnia disorder, placed above pharmacotherapy in the 2021 clinical practice guideline. CBT-I improves sleep efficiency, reduces wake after sleep onset, and increases subjective sleep quality across multiple randomized controlled trials.

What CBT-I Includes

A standard CBT-I course runs 6-8 weekly sessions (in-person or digital). Core components are:

• Sleep restriction therapy. Temporarily limits time in bed to actual sleep time (minimum 5.5 hours) to build sleep pressure.
• Stimulus control. Restricts the bed to sleep and sex only; patients get out of bed when awake for more than 20 minutes.
• Cognitive restructuring. Identifies and challenges unhelpful beliefs (for example, "I must get 8 hours or tomorrow is ruined").
• Relaxation training. Progressive muscle relaxation or diaphragmatic breathing reduces pre-sleep arousal.
• Sleep hygiene. Addresses caffeine timing, light exposure, and bedroom environment, though hygiene alone is insufficient as a standalone intervention.

A Cochrane meta-analysis of 20 RCTs (N=1,162 participants) found CBT-I reduced wake after sleep onset by a mean of 62 minutes and improved sleep efficiency from 81.4% to 88.0% at post-treatment. (van Straten et al., 2018, Sleep Medicine Reviews) Digital CBT-I platforms (Sleepio, Somryst) show comparable effects to therapist-delivered CBT-I in RCT data.

FDA-Approved Drugs That Treat Non-Restorative Sleep or Insomnia

When CBT-I alone is insufficient, or while waiting for it to take effect, the following agents have FDA approval for sleep maintenance insomnia or sleep onset insomnia and have evidence bearing on sleep quality specifically.

Suvorexant (Belsomra)

Suvorexant is a dual orexin receptor antagonist (DORA) approved by the FDA in 2014. It blocks orexin-1 and orexin-2 receptors, reducing the wake-promoting signal rather than sedating via GABA pathways, which means it does not suppress SWS or REM. A phase 3 trial (N=1,021) demonstrated suvorexant 20 mg reduced subjective wake after sleep onset by 28 minutes vs. 14 minutes for placebo at week 3, with sustained effect at month 3. (Herring et al., 2012, Neurology) Standard doses are 10 mg (starting) to 20 mg. Next-day driving impairment was reported at 40 mg in a controlled driving study, so the 20 mg ceiling for most patients is important.

Low-Dose Doxepin (Silenor)

Doxepin 3 mg and 6 mg (brand name Silenor) received FDA approval in 2010 specifically for sleep maintenance insomnia. At these doses, doxepin acts selectively as an H1 histamine antagonist. Unlike tricyclic antidepressant doses of doxepin (75-150 mg), the sleep formulation does not produce next-morning sedation or anticholinergic side effects at a clinically meaningful level in most patients. A 35-night randomized trial (N=240) showed doxepin 6 mg improved total sleep time by 25.8 minutes and wake after sleep onset by 32 minutes vs. Placebo. (Roth et al., 2007, Sleep)

Ramelteon (Rozerem)

Ramelteon is a melatonin MT1/MT2 receptor agonist approved for sleep onset insomnia. It does not suppress any sleep stage and carries no Schedule IV classification, making it suitable for patients with a history of substance use disorder. The standard dose is 8 mg taken 30 minutes before bed. It is particularly useful when a circadian phase delay contributes to non-restorative sleep.

Eszopiclone (Lunesta)

Eszopiclone is a non-benzodiazepine GABA-A modulator. Unlike zolpidem, eszopiclone has FDA approval for use without a duration restriction, and 6-month data show it maintains efficacy without tolerance development. The ESTEEM 1 trial (N=788) showed eszopiclone 3 mg improved total sleep time by 37 minutes and next-day functioning scores vs. Placebo at 6 months. (Krystal et al., 2003, Archives of General Psychiatry) Dose should start at 1 mg in adults over 65.

Trazodone (Off-Label)

Trazodone 50-100 mg at bedtime is the most commonly prescribed off-label sleep agent in the United States, used by an estimated 5 million patients per year. It is a serotonin antagonist and reuptake inhibitor (SARI) that increases SWS at low doses. A small crossover study (N=15) found trazodone 100 mg increased N3 by approximately 15% compared to placebo without reducing REM. (Parrino et al., 1994, Psychopharmacology) Priapism, orthostatic hypotension, and next-day sedation are the main risks; orthostatic hypotension is most relevant in older men.

Mirtazapine (Off-Label)

Mirtazapine 7.5-15 mg at bedtime blocks H1, 5-HT2A, and 5-HT2C receptors. The 5-HT2A blockade specifically increases SWS. A weight gain of 2-4 kg over 6 weeks is a common side effect that limits use in patients already managing metabolic concerns.

A Drug Review Framework for Non-Restorative Sleep

Before prescribing a sleep-specific agent, apply this four-step medication review:

1. List all current medications including OTC agents and supplements. Flag any beta-blocker, benzodiazepine, SSRI, antihistamine, stimulant, corticosteroid, or alcohol use pattern.
2. Assess timing. Many offenders are harmless at the right time. Methylphenidate taken before noon rarely disrupts sleep architecture by bedtime. Prednisone taken at 8 a.m. Produces less circadian disruption than evening dosing.
3. Check hormone panel. Order TSH, free T4, morning cortisol, testosterone (total and free in men), and estradiol/progesterone (in perimenopausal women) before attributing the complaint to a primary sleep disorder.
4. Apply the substitution test. If a beta-blocker is needed, trial hydrophilic atenolol. If an antihistamine is being used for sleep, switch to melatonin 0.5-1 mg 90 minutes before bed (evidence from a Cochrane review of 19 RCTs: melatonin reduced sleep onset latency by 7 minutes and improved sleep quality ratings vs. Placebo). (Ferracioli-Oda et al., 2013, PLOS ONE)

Supplements: What the Evidence Actually Shows

Melatonin

Melatonin is most effective for circadian-related non-restorative sleep (jet lag, shift work, delayed sleep phase). It does not increase total sleep time in primary insomnia but does shift the circadian clock when taken 90-120 minutes before the desired sleep time at 0.5-1 mg. Higher doses (5-10 mg) are common but produce supraphysiologic peaks that may desensitize MT1/MT2 receptors over time.

Magnesium

Magnesium glycinate 200-400 mg taken at bedtime receives wide attention online. A small RCT (N=46) of older adults with insomnia found magnesium supplementation improved ISI scores and increased serum melatonin vs. Placebo. (Abbasi et al., 2012, J Res Med Sci) The trial is underpowered, and results have not been replicated in larger populations. Magnesium supplementation is low-risk at this dose for most adults.

L-Theanine

L-Theanine 200 mg at bedtime has shown modest effects on subjective sleep quality in two small RCTs (N=30 and N=22). The mechanism is thought to involve alpha-wave induction and reduced glutamatergic arousal. Evidence is insufficient to recommend it as primary treatment.

When to Refer and When to Worry

Non-restorative sleep lasting fewer than 3 months often resolves with CBT-I, medication review, and attention to sleep hygiene. Refer to a sleep specialist or arrange PSG when any of these features are present:

• Loud snoring, witnessed apneas, or morning headaches (suggesting OSA)
• Restless legs, leg kicking reported by a partner, or periodic leg jerks waking the patient (suggesting PLMD or RLS)
• Excessive daytime sleepiness with sleep attacks despite adequate nighttime sleep (suggesting narcolepsy or idiopathic hypersomnia)
• Symptoms persisting after 8 weeks of CBT-I plus an appropriately selected pharmacologic agent
• A new diagnosis of fibromyalgia, ME/CFS, or Parkinson's disease, all of which carry high rates of non-restorative sleep and require specialist co-management

The AASM states: "Chronic insomnia disorder is defined as difficulty initiating or maintaining sleep, or early morning awakening, occurring at least 3 nights per week, for at least 3 months, with associated daytime impairment." (AASM ICSD-3, 2014) Non-restorative sleep meeting that temporal threshold warrants formal evaluation, not watchful waiting.