Clinical Psychopharmacology · Paper A
Smoking and drug metabolism
Tobacco smoke changes drug levels through liver enzyme inductionA drug increasing the amount or activity of a liver enzyme, which speeds up the breakdown of other drugs it shares that enzyme with.Read more →, not through nicotine itself, a distinction that matters enormously when a patient's smoking status changes. Take a patient on clozapine admitted to a smoke-free ward who cannot smoke for the first time in years. Their clozapine level can rise into the toxic range within days, purely from losing that enzyme induction, e.g. a level nearly doubling within a week of admission. Clinically, monitor clozapine levels proactively whenever a smoking patient is admitted somewhere smoke-free.
Polycyclic aromatic hydrocarbons in combusted tobacco smoke induce hepatic CYP1A2, which speeds up the breakdown of CYP1A2 substrates, especially clozapine and olanzapine. Stopping smoking removes this induction over days, so concentrations of these drugs rise; starting or resuming smoking has the opposite effect and can lower levels, risking reduced efficacy. A patient who resumes heavy smoking after discharge from a smoke-free ward, e.g. going back to twenty cigarettes a day within a week, will see levels fall again as induction returns. Clinically, nicotine replacement therapy or vaping does NOT reproduce this effect, since nicotine itself is not the inducing agent, only combusted smoke is.
A huge number of psychiatric inpatients smoke, and wards are now smoke-free. That single policy change means every patient who smokes and gets admitted is, pharmacologically, starting a mini drug-interaction experiment the moment they walk through the door. The reason this matters clinically, not just as trivia, is clozapine: a small number of patients on clozapine who suddenly stop smoking have had seizures or severe toxicity within days of admission, because their blood level climbed sharply while nobody adjusted the dose. Getting this concept right is a genuine patient-safety skill, not an exam curiosity.
Most drugs are cleared from the body by liver enzymes, chiefly a family called cytochrome P450 (CYP), each named with a number-letter-number code such as CYP1A2 or CYP3A4. INDUCTION means a substance makes the liver build MORE of a particular enzyme, so drugs broken down by that enzyme (its SUBSTRATES) get cleared faster and blood levels fall. INHIBITION is the opposite: a substance blocks an enzyme, so substrates build up and levels rise. Smoking is a classic INDUCER of CYP1A2 specifically. Two drugs used a great deal in psychiatry, clozapine and olanzapine, are CYP1A2 substrates, so their levels are directly hostage to a patient's smoking status.
The single most commonly missed point is this: the inducing agent is polycyclic aromatic hydrocarbons (PAHs), chemicals produced by burning tobacco leaf, not nicotine itself. This means nicotine replacement therapy (NRT, patches or gum) and vaping (e-cigarettes) do NOT induce CYP1A2, because neither involves combustion. A patient who switches from cigarettes to vaping, or who starts nicotine patches instead of smoking, is pharmacokinetically in exactly the same position as a patient who has stopped smoking altogether: their CYP1A2 induction fades, and drugs like clozapine will rise in the blood over the following one to two weeks, even though the patient still feels like they are 'still smoking, just differently'.
is not a light switch. When PAHs from smoke are present, the liver cell gradually makes more CYP1A2 enzyme protein over roughly one to two weeks of continued exposure; the induction plateaus once enzyme levels reach a new . When smoking stops, existing CYP1A2 enzyme does not vanish instantly, it has to be degraded and not replaced, so DE-induction also takes about one to two weeks. This matters practically: a patient who stops smoking on admission to a smoke-free ward will not see their clozapine level spike on day one, but rather creep upward over the following one to two weeks, which is exactly the window where a busy ward can miss it if no one is watching proactively.
Consider a patient established on clozapine 400mg who smokes twenty cigarettes a day and is admitted informally to a smoke-free psychiatric ward. For the first few days they may feel fine. By around day seven to ten, as CYP1A2 induction fades, their clozapine level can climb sharply, e.g. nearly doubling from its baseline, producing sedation, hypersalivation, tachycardia (a fast heart rate) or, in severe cases, seizures. Now flip the scenario: a patient who has been smoke-free in hospital for three weeks, stabilised on a now-higher effective dose, is discharged home and resumes heavy smoking. Within one to two weeks their clozapine level falls back down, and if the dose is not raised again they risk a relapse of psychotic symptoms from a now sub-therapeutic level. The drug dose never changed in either story; only the smoking status did.
A rising white cell count with clozapine suggests infection or, rarely, agranulocytosis (a dangerous drop in white cells, ironically opposite in direction) rather than a smoking-related level change, and should prompt a full blood count, not just a level check. Genuine non-adherence (a patient simply not taking their tablets) can also cause apparent 'level changes' but moves levels down, not up, and is distinguished by directly asking about missed doses. Caffeine intoxication can mimic some clozapine-toxicity symptoms like tachycardia and agitation; caffeine is also a CYP1A2 substrate, so heavy caffeine intake alongside a smoking change is a genuine confound worth asking about directly.
The key investigation is a clozapine plasma level (a blood test measuring the drug concentration), which should be checked at baseline and repeated whenever smoking status changes materially, in addition to routine monitoring. Full blood count monitoring for clozapine (weekly, then fortnightly, then monthly per its mandatory monitoring schedule) continues regardless, since agranulocytosis screening is a separate safety requirement. Clinically, any admission to or discharge from a smoke-free environment should itself trigger a documented plan for level monitoring, not be left to chance.
The Maudsley Prescribing Guidelines in Psychiatry, the standard UK psychiatric prescribing reference, recommend that clozapine and olanzapine doses be reviewed proactively whenever a patient's smoking status changes, ideally with a plasma level check rather than waiting for toxicity symptoms to appear. NICE guidance on smoking cessation in secondary care settings also flags smoke-free wards as a moment requiring medicine reviews for CYP1A2 substrates. In practice, many services offer nicotine replacement therapy on admission both to support the patient and, crucially, because NRT itself does NOT reproduce the induction, so it is a genuinely honest substitute rather than a hidden dose-adjustment trick.
If a vignette describes a patient on clozapine or olanzapine whose ward is smoke-free, or who has recently stopped or resumed smoking, think CYP1A2 induction change before you think 'new side effect' or 'non-adherence': the exam is testing the enzyme-induction mechanism, and specifically that nicotine replacement and vaping do not reproduce it.
Combusted tobacco smoke, not nicotine, induces CYP1A2 and raises of clozapine and olanzapine, so any change in smoking status is itself a trigger for level monitoring.