Montelukast Side Effects: What the Trial Data and the Boxed Warning Actually Say

Montelukast produces a strange pattern in its own label. In placebo-controlled trials the adverse reaction rates sit almost on top of placebo, yet the drug carries one of the strongest warnings the FDA issues. Both facts are accurate. The distance between them is why most summaries of montelukast side effects end up muddled.

The two are answering different questions. The trial tables record what happened to a few thousand monitored patients over weeks to months. The boxed warning came from what was reported after approval, across tens of millions of prescriptions, over more than two decades. Reading one without the other gives a false picture in either direction.

The boxed warning: serious neuropsychiatric events

The FDA added a boxed warning to montelukast in March 2020. Current generic labelling carries a revision date of 05/2020. The warning is direct about what was reported and equally direct about what is not known.

Serious neuropsychiatric events have been reported with montelukast sodium. The label describes the reported event types as highly variable, including agitation, aggression, depression, sleep disturbances, and suicidal thoughts and behaviour including suicide. It then states that the mechanisms underlying these events are not currently well understood.

The instructions attached to the warning matter more than the warning text itself:

• Discuss benefits and risks with patients and caregivers before prescribing.
• Monitor for neuropsychiatric symptoms during treatment.
• Discontinue montelukast immediately if neuropsychiatric symptoms occur, and contact a healthcare provider.
• For allergic rhinitis, reserve montelukast for patients with an inadequate response or intolerance to alternative therapies.

That last line is a restriction on the indication, not a general caution. The label says the benefits may not outweigh the risk of neuropsychiatric symptoms in patients with allergic rhinitis, so montelukast moved from a first-line option in hay fever to a second-line one. For asthma and exercise-induced bronchoconstriction the wording is different and softer: consider the benefits and risks before prescribing. The indication itself was not narrowed there.

What the adult trials actually reported

Montelukast was evaluated for safety in roughly 2,950 adult and adolescent patients aged 15 and over. The table below lists every adverse experience that occurred in at least 1 percent of patients on montelukast 10 mg daily and at a rate higher than placebo. The placebo column is the part most articles drop, and it is the only part that makes the numbers interpretable.

Adverse experience	Montelukast 10 mg/day (n=1955)	Placebo (n=1180)
Headache	18.4%	18.1%
Influenza	4.2%	3.9%
Abdominal pain	2.9%	2.5%
Cough	2.7%	2.4%
Dyspepsia	2.1%	1.1%
Raised ALT	2.1%	2.0%
Dizziness	1.9%	1.4%
Asthenia or fatigue	1.8%	1.2%
Dental pain	1.7%	1.0%
Nasal congestion	1.6%	1.3%
Raised AST	1.6%	1.2%
Rash	1.6%	1.2%
Fever	1.5%	0.9%
Infectious gastroenteritis	1.5%	0.5%
Trauma	1.0%	0.8%
Pyuria	1.0%	0.9%

How to read that table without misleading yourself

Subtract the placebo column and almost nothing survives. Headache is the loudest number on the page at 18.4 percent, and its placebo rate is 18.1 percent. The difference is 0.3 percentage points, which is noise. Anyone quoting "headache affects nearly one in five people on montelukast" is quoting a true number that carries no information about the drug.

The differences that are not noise are small and unglamorous: dyspepsia at 2.1 versus 1.1 percent, infectious gastroenteritis at 1.5 versus 0.5 percent, fever at 1.5 versus 0.9 percent. Each is roughly a one percentage point excess.

The label also states that the frequency of less common adverse events was comparable between montelukast and placebo, and that with prolonged treatment the profile did not significantly change. Cumulatively, 569 patients took montelukast for at least six months, 480 for one year and 49 for two years within the trial programme.

This is the honest reason montelukast was so widely prescribed for so long. On the measured endpoints in controlled trials, it looks like a very quiet drug.

Why the trial tables missed the neuropsychiatric signal

Three structural reasons, and they apply to most drug safety databases rather than to montelukast specifically.

Size. A few thousand patients cannot detect an event occurring in one in ten thousand. Post-approval exposure was several orders of magnitude larger.

Duration. Most of the registration trials ran for weeks. Fewer than fifty patients in the programme reached two years. Behavioural changes reported in practice often surfaced after months of use.

Ascertainment. Trial adverse event forms capture what investigators code. A parent noticing that a child has become irritable, is sleeping badly or has started having nightmares is not reliably captured by a checklist designed around respiratory outcomes.

The post-marketing psychiatric list in full

Post-approval reports are voluntary, come from a population of uncertain size, and cannot be converted into an incidence rate. The label says this explicitly. What it does provide is the list of psychiatric disorders reported, and the breadth of it is the point:

Agitation, aggressive behaviour or hostility, anxiousness, depression, disorientation, disturbance in attention, dream abnormalities, dysphemia (stuttering), hallucinations, insomnia, irritability, memory impairment, obsessive-compulsive symptoms, restlessness, somnambulism (sleepwalking), suicidal thinking and behaviour including suicide, tic and tremor.

Separately under nervous system disorders: drowsiness, paraesthesia or hypoesthesia, and seizures.

Two patterns are worth knowing before starting montelukast or giving it to a child. The reported events cluster heavily around sleep and mood, which are exactly the domains a parent or partner notices before a prescriber does. And several of them, particularly sleep disturbance, nightmares and irritability, are easy to attribute to something else entirely, especially in a child who also has poorly controlled asthma or allergic rhinitis.

Non-psychiatric post-marketing reports

The rest of the post-marketing section is short but not trivial. Reported events include increased bleeding tendency and thrombocytopenia; hypersensitivity reactions including anaphylaxis and hepatic eosinophilic infiltration; palpitations; epistaxis and pulmonary eosinophilia; and gastrointestinal effects including diarrhoea and dyspepsia.

Pulmonary eosinophilia deserves a note. Rare cases of eosinophilic conditions, sometimes presenting as vasculitis, have been reported in patients on leukotriene antagonists, frequently while oral corticosteroids were being reduced. Whether the drug causes it or unmasks an underlying eosinophilic disorder as steroids come down has never been settled. New rash, worsening lung symptoms, cardiac complications or neuropathy during a steroid taper is a reason to contact the prescriber rather than to push on.

What the paediatric trials reported

Montelukast was assessed in 476 children aged 6 to 14 and 573 children aged 2 to 5. In the 6 to 14 group, events occurring at a frequency of at least 2 percent and more often than placebo were pharyngitis, influenza, fever, sinusitis, nausea, diarrhoea, dyspepsia, otitis, viral infection and laryngitis. In a 56-week growth-rate study in children aged 6 to 8, the additional events meeting that threshold were headache, infective rhinitis, varicella, gastroenteritis, atopic dermatitis, acute bronchitis, tooth infection, skin infection and myopia.

In the 2 to 5 group the list was longer: fever, cough, abdominal pain, diarrhoea, headache, rhinorrhoea, sinusitis, otitis, influenza, rash, ear pain, gastroenteritis, eczema, urticaria, varicella, pneumonia, dermatitis and conjunctivitis.

Read those lists carefully and a pattern shows up: they are dominated by childhood infections, which is what you expect from any group of small children followed for a year. The label notes the frequency of less common events was comparable with placebo. Once again, the controlled data are quiet, and the boxed warning does not come from them.

Dosing, and why timing is not the safety lever people think

Montelukast is taken once daily in the evening for asthma, at 10 mg for ages 15 and over, 5 mg chewable for ages 6 to 14, and 4 mg chewable for ages 2 to 5. For exercise-induced bronchoconstriction the dose is taken at least two hours before exercise, in patients aged 6 and over. Patients with both asthma and allergic rhinitis take a single evening dose covering both.

Age group	Asthma dose	Form	Product
15 years and over	10 mg once daily, evening	Film-coated tablet	Montair 10 mg
6 to 14 years	5 mg once daily, evening	Chewable tablet	Montair 5 mg
2 to 5 years	4 mg once daily, evening	Chewable tablet	Montair 4 mg

A common assumption is that evening dosing exists to reduce side effects or to avoid daytime drowsiness. It does not. The label states there have been no asthma trials comparing morning with evening dosing, and that montelukast pharmacokinetics are similar either way. Evening dosing is how the trials were run and efficacy was demonstrated there. Moving the dose to the morning to avoid nightmares is therefore an off-label improvisation with no trial support behind it, which is worth knowing before treating it as a fix.

A missed dose is taken at the next regular time. Doses are never doubled up.

Montelukast with levocetirizine

Fixed-dose montelukast plus levocetirizine is widely used where allergic rhinitis and asthma coexist, and we stock it as Montair LC and Montair LC Kid. A 2024 systematic review and meta-analysis of the combination in combined allergic rhinitis and asthma syndrome found the combination performed favourably against comparators on efficacy and safety endpoints (PMID: 39394937).

The safety caveat is structural rather than pharmacological. A combination tablet carries the montelukast boxed warning in full, and it delivers montelukast to a population whose main complaint is often rhinitis, which is exactly the group the FDA told prescribers to reserve montelukast for. Adding an antihistamine does not change the reserve-for-second-line instruction on the montelukast half.

Montelukast is not a reliever, and that changes the risk calculation

Montelukast is indicated for prophylaxis and chronic treatment of asthma. It does not relieve an attack. Anyone relying on it as rescue is using the wrong drug, and the correct comparison for weighing its risks is against an inhaled corticosteroid, not against a reliever inhaler.

That comparison is unforgiving. Inhaled corticosteroids are more effective for persistent asthma than leukotriene antagonists in most patients, and they act locally with modest systemic exposure. A patient on montelukast alone who could be well controlled on a low-dose inhaled steroid is carrying the boxed warning without a clinical reason. Our guide to budesonide with formoterol inhalers covers what the inhaled option looks like in practice, and the asthma treatment overview sets out where each drug class sits.

Montelukast still earns its place in specific situations: patients who genuinely cannot or will not use an inhaler correctly, exercise-induced bronchoconstriction as an alternative to pre-exercise dosing, aspirin-exacerbated respiratory disease, and asthma with prominent allergic rhinitis where an inhaled steroid alone leaves the nose untreated.

Stopping montelukast

Montelukast has no withdrawal syndrome and no taper requirement. It can be stopped, and the label instructs immediate discontinuation if neuropsychiatric symptoms appear.

The real risk in stopping is not the drug leaving the system, it is what covers the asthma afterwards. If montelukast was the only controller, stopping it without replacing it removes the anti-inflammatory treatment and leaves the patient on reliever alone, which is the pattern most strongly associated with attacks. Stopping should be a conversation with the prescriber about what takes its place, not a unilateral decision, unless neuropsychiatric symptoms have appeared, in which case the label says stop and contact a provider.

Where behavioural symptoms were the reason for stopping, reported cases generally describe resolution after discontinuation, though the label does not quantify this. Anyone in this position should tell the prescriber explicitly, because it is a fact that belongs in the record before montelukast is prescribed again.

Medical disclaimer

This article is educational and does not replace advice from a qualified healthcare professional. Montelukast requires a prescription in the United States, United Kingdom, Canada and Australia. Trial incidence rates come from studies run under specific conditions and cannot be compared directly with rates from trials of other drugs, or applied to any individual. Post-marketing reports do not establish causation and no frequency can be calculated from them. If you or your child develop changes in mood, behaviour or sleep while taking montelukast, contact your prescriber. Thoughts of self-harm need urgent medical help.

References

• FDA-approved prescribing information, montelukast sodium, Boxed Warning: Serious Neuropsychiatric Events, revised 05/2020.
• FDA-approved prescribing information, montelukast sodium, section 6.1 Clinical Trials Experience, Table 1.
• FDA-approved prescribing information, montelukast sodium, section 6.2 Postmarketing Experience.
• FDA-approved prescribing information, montelukast sodium, sections 1.1 to 1.3 and 2.1 to 2.4.
• DailyMed, montelukast sodium tablets, Contraindications, Interactions and Overdosage sections.
• Efficacy and safety of montelukast-levocetirizine combination therapy in combined allergic rhinitis and asthma syndrome: a systematic review and meta-analysis. PubMed PMID: 39394937.

Related reading on asthma treatment

Montelukast is one option among several. For the inhaled controller and reliever picture, see our guides to budesonide with formoterol dosing and maintenance-and-reliever therapy and salbutamol compared with levosalbutamol. The full range is on our asthma and respiratory medicines page.