Drugs That Distort the Salivary Melatonin Profile Test

What a Salivary Melatonin Profile Actually Measures

The salivary melatonin profile captures the rise, peak, and decline of melatonin secretion across an evening and night by sampling saliva at regular intervals under dim-light conditions (typically <30 lux). The most clinically useful data point extracted from this curve is the dim-light melatonin onset, or DLMO, defined as the time when salivary melatonin crosses a fixed threshold, usually 3 pg/mL or 4 pg/mL depending on the assay [1]. DLMO is considered the most reliable peripheral marker of the central circadian pacemaker in the suprachiasmatic nucleus [2].

Clinicians order this test to diagnose delayed sleep-wake phase disorder (DSWPD), advanced sleep-wake phase disorder, non-24-hour sleep-wake rhythm disorder, and to time chronotherapy interventions such as exogenous melatonin or bright-light therapy. The Endocrine Society's 2010 Clinical Practice Guideline on the evaluation and treatment of circadian rhythm sleep disorders identifies DLMO as "the most accurate marker of circadian phase position available in clinical practice" [3]. Because the test relies on detecting subtle shifts in melatonin timing, even modest pharmacological interference can push a diagnosis in the wrong direction. A drug that suppresses melatonin amplitude by 50% can delay the apparent DLMO by 1 to 2 hours, converting a normal result into one that looks like delayed-phase disorder.

Beta-Blockers: The Strongest Single-Drug Effect

Beta-adrenergic blockers are the most potent suppressors of endogenous melatonin. Melatonin synthesis in the pineal gland depends on norepinephrine binding to beta-1 adrenergic receptors, which activates arylalkylamine N-acetyltransferase (AANAT), the rate-limiting enzyme in melatonin production. When a non-selective beta-blocker such as propranolol occupies those receptors, melatonin output drops sharply.

A study by Stoschitzky et al. (2006) demonstrated that atenolol 100 mg daily reduced nocturnal melatonin levels by approximately 80%, while nebivolol (a beta-1 selective agent with nitric oxide modulation) reduced levels by only about 38% [4]. Propranolol produced similar 60% to 80% reductions in earlier pharmacokinetic work [5]. This suppression is dose-dependent. Patients on low-dose metoprolol (25 mg) may see a 20% to 30% reduction, while those on high-dose propranolol (160 mg/day) can experience near-complete suppression of the nocturnal melatonin rise.

The clinical implication is direct. A patient on atenolol 50 mg who undergoes a salivary melatonin profile will likely show a flattened curve with no identifiable DLMO, leading to a false impression of absent circadian rhythmicity. The American Academy of Sleep Medicine (AASM) recommends documenting all beta-blocker use when interpreting DLMO results [6].

Washout guidance by beta-blocker half-life:

• Propranolol (t½ 3 to 6 hours): hold for 24 to 30 hours pre-test
• Atenolol (t½ 6 to 7 hours): hold for 35 hours
• Metoprolol succinate (t½ 3 to 7 hours): hold for 35 hours
• Nadolol (t½ 20 to 24 hours): hold for 5 days

Any washout decision must be discussed with the prescribing physician, particularly in patients taking beta-blockers for rate control in atrial fibrillation or post-myocardial infarction prophylaxis.

NSAIDs and Prostaglandin-Pathway Interference

Nonsteroidal anti-inflammatory drugs suppress melatonin through a different mechanism: inhibition of prostaglandin synthesis. Prostaglandin D2 and E2 participate in the signaling cascade that drives pineal melatonin release. When cyclooxygenase (COX) is blocked, the downstream signal to AANAT weakens.

Murphy et al. (1996) published one of the foundational studies on this interaction, showing that ibuprofen 400 mg taken at 6 PM reduced overnight urinary 6-sulphatoxymelatonin (the primary melatonin metabolite) by 75% compared to placebo in healthy volunteers [7]. Indomethacin and aspirin produced similar, though slightly less pronounced, reductions in salivary melatonin amplitude in a crossover trial by Surrall et al. [8].

Even a single dose matters. A patient who takes two ibuprofen tablets for a headache on the afternoon of a DLMO collection can lose most of the melatonin signal that evening. Selective COX-2 inhibitors (celecoxib, meloxicab) have shown a smaller effect in limited data, but the evidence is insufficient to call them safe for pre-test use. The safest approach is acetaminophen for pain control in the 48 hours before collection, as it does not significantly affect melatonin secretion [7].

Caffeine and CYP1A2 Metabolic Interference

Caffeine affects salivary melatonin through two pathways. First, caffeine is a competitive inhibitor of adenosine receptors, and adenosine contributes to sleep-pressure signaling that interacts with circadian timing. Second, both caffeine and melatonin are metabolized primarily by hepatic CYP1A2. High caffeine intake can alter CYP1A2 activity and shift melatonin clearance rates.

A controlled study by Burke et al. (2015) published in Science Translational Medicine showed that 200 mg of caffeine (roughly one 12-oz drip coffee) consumed 3 hours before habitual bedtime delayed DLMO by approximately 40 minutes, an effect size nearly half that of 3 hours of bright-light exposure (3,000 lux) [9]. The combination of caffeine plus bright light delayed DLMO by 105 minutes. That 40-minute caffeine-induced delay is clinically significant when diagnosing circadian disorders where the DLMO shift of interest may itself be only 1 to 2 hours.

Dr. Kenneth Wright, the study's senior author from the University of Colorado Boulder, noted: "Caffeine's effect on the circadian clock has been underestimated in clinical testing protocols. A single cup of evening coffee can shift melatonin timing by nearly an hour" [9]. Patients should abstain from all caffeine sources (coffee, tea, energy drinks, chocolate, pre-workout supplements) for at least 12 hours before the start of a salivary melatonin collection.

Benzodiazepines, Z-Drugs, and GABAergic Agents

Benzodiazepines reduce salivary melatonin amplitude by approximately 25% to 35%, likely through GABA-A receptor modulation of the suprachiasmatic nucleus output. Alprazolam and diazepam have the best-documented effects. A study by McIntyre et al. (1993) found that alprazolam 0.5 mg reduced peak nocturnal melatonin by 32% in healthy subjects [10].

Z-drugs (zolpidem, zopiclone, eszopiclone) are sometimes assumed to be "safer" for melatonin testing because they are more selective for the alpha-1 subunit of the GABA-A receptor. Limited evidence suggests zolpidem has a smaller effect on melatonin amplitude (roughly 10% to 15% suppression), but data are scarce, and a conservative approach still calls for washout [10]. Gabapentin and pregabalin, which act on voltage-gated calcium channels rather than GABA receptors, do not appear to affect salivary melatonin concentrations meaningfully based on available pharmacokinetic data.

Antidepressants, Antipsychotics, and Serotonergic Agents

The relationship between serotonergic drugs and melatonin is complex because serotonin (5-HT) is the biochemical precursor to melatonin. The pineal gland converts serotonin to N-acetylserotonin and then to melatonin via AANAT and hydroxyindole-O-methyltransferase (HIOMT).

SSRIs and SNRIs. Fluvoxamine is a potent CYP1A2 inhibitor and dramatically increases serum and salivary melatonin levels, often by 200% to 300%, by blocking melatonin's primary clearance pathway [11]. Other SSRIs (sertraline, fluoxetine, citalopram) have weaker CYP1A2 effects and produce smaller, less predictable changes. The practical concern is that fluvoxamine can create a falsely high melatonin curve, making the DLMO appear earlier than it truly is.

Tricyclic antidepressants. Desipramine, a norepinephrine-reuptake inhibitor, has been shown to increase nocturnal melatonin by approximately 50% in some studies, likely by increasing noradrenergic tone at the pineal gland [12].

Atypical antipsychotics. Quetiapine, olanzapine, and risperidone have variable effects on melatonin. Quetiapine's antihistaminic and anti-alpha-1 properties may slightly suppress melatonin, but published data are inconsistent [12]. Given the risk of discontinuation in patients with serious psychiatric illness, the clinician ordering the salivary melatonin profile should coordinate closely with the prescribing psychiatrist rather than reflexively requesting a washout.

Exogenous Melatonin and Melatonin-Receptor Agonists

This may seem obvious, but it warrants explicit mention. Exogenous melatonin supplements (even low doses of 0.5 mg) and melatonin-receptor agonists such as ramelteon (Rozerem) and tasimelteon (Hetlioz) will contaminate the salivary melatonin profile. Melatonin supplements sold over the counter in the United States have highly variable actual content. A 2017 study in the Journal of Clinical Sleep Medicine analyzed 31 melatonin supplements and found that actual melatonin content ranged from 83% less to 478% more than the labeled dose [13].

Ramelteon, a selective MT1/MT2 receptor agonist, does not raise salivary melatonin per se (it is not melatonin), but its metabolite M-II has weak melatonin-like activity and cross-reacts with some radioimmunoassay (RIA) melatonin assays, producing a falsely elevated reading [14]. Patients should discontinue exogenous melatonin for at least 1 week (given its short half-life but variable supplement purity) and ramelteon for at least 5 days before testing.

Light Exposure, Smoking, and Alcohol

Not all test distortions come from prescription bottles.

Light. The defining condition of DLMO measurement is dim light, typically <30 lux, ideally <10 lux. Exposure to 200 lux (a normally lit room) during the collection window suppresses salivary melatonin by more than 50% [15]. Checking a smartphone screen (80 to 100 lux at the cornea, with a high blue-light fraction) during sampling can acutely suppress the melatonin signal. The 2015 Endocrine Society Scientific Statement on endocrine-disrupting chemicals noted that "short-wavelength light exposure in the 446 to 477 nm range is the most potent acute suppressor of pineal melatonin release" [16].

Smoking. Tobacco smoke induces CYP1A2 activity. Chronic smokers clear melatonin faster, resulting in lower salivary melatonin concentrations and a compressed nocturnal curve. Studies show smokers have approximately 30% lower peak melatonin compared to non-smokers of the same age [17]. This effect reverses within 2 to 4 weeks of smoking cessation, but asking a patient to quit smoking for a test is not realistic. Document smoking status and interpret accordingly.

Alcohol. Acute alcohol intake (2 to 3 standard drinks) in the evening suppresses nocturnal melatonin by roughly 15% to 20% and may delay DLMO by 30 to 45 minutes [18]. Patients should avoid alcohol for 24 hours before and during the collection period.

How to Prepare a Patient for an Accurate Salivary Melatonin Profile

A reliable DLMO result requires more pre-test coordination than most lab tests. The ordering clinician should send patients a written preparation checklist at least 2 weeks in advance so that washout timelines can be discussed with other prescribers.

Step 1: Medication audit (14 days before test). Review every prescription, OTC, and supplement the patient takes. Flag beta-blockers, NSAIDs, benzodiazepines, SSRIs (especially fluvoxamine), exogenous melatonin, ramelteon, and tasimelteon. Determine which drugs can be safely held for 5 half-lives.

Step 2: Caffeine and alcohol restriction (48 hours before test). No caffeine from any source for at least 12 hours before the first sample; ideally, abstain for 24 to 48 hours. No alcohol for 24 hours.

Step 3: Light control (collection evening). Dim home lighting to <30 lux starting 5 to 6 hours before habitual bedtime. No screens without blue-light filters. Use dim red or amber lighting only.

Step 4: Sampling protocol. Collect saliva every 30 to 60 minutes from approximately 5 PM (or 6 hours before habitual bedtime) through 1 hour past habitual bedtime. Do not eat, drink, or brush teeth for 15 minutes before each sample. Label tubes with exact clock time.

Step 5: Interpret with context. The lab report gives raw melatonin concentrations; the clinician calculates DLMO by identifying the sample time at which melatonin first exceeds the assay threshold. If the patient could not wash out a known interfering drug, note this on the requisition and consider repeating the test after successful washout.

Normal Salivary Melatonin Range and What Abnormal Results Mean

In healthy adults, evening pre-DLMO salivary melatonin is typically <3 pg/mL, rising to a peak of 10 to 50 pg/mL between 2 AM and 4 AM, then declining to daytime baseline levels (<3 pg/mL) by 7 to 9 AM [1]. DLMO in adults with normal sleep timing generally occurs between 8 PM and 10 PM.

A high salivary melatonin level (peak above 50 pg/mL) may reflect exogenous melatonin contamination, fluvoxamine co-administration, or (rarely) a pineal cyst or tumor with autonomous melatonin secretion [14]. A low or absent melatonin rise may indicate beta-blocker suppression, NSAID use, excessive light exposure during collection, advanced age (melatonin declines by approximately 10% to 15% per decade after age 20 [19]), or true circadian disruption as seen in non-24-hour sleep-wake disorder or Smith-Magenis syndrome.