16.3 Tobacco Treatment in Special Populations & Co-Occurring Disorders
Key Takeaways
- The EAGLES trial enrolled 8,144 smokers including a psychiatric cohort of 4,116 and found no significant increase in serious neuropsychiatric events with varenicline or bupropion relative to patch or placebo.
- Tobacco smoke induces CYP1A2, so quitting raises clozapine and olanzapine levels and can require dose reductions of roughly 30 to 40 percent with monitoring.
- Behavioral counseling is first-line in pregnancy, and financial incentives have the strongest evidence among behavioral strategies for pregnant smokers.
- Varenicline is the most effective single agent for most smokers and remains effective in people with serious mental illness and in people in addiction treatment.
- Combination nicotine replacement using a patch plus an as-needed short-acting product outperforms any single nicotine replacement product.
Last updated: September 2026
16.3 Tobacco Treatment in Special Populations & Co-Occurring Disorders
Quick Answer: EAGLES (N = 8,144, including a psychiatric cohort of 4,116) found no significant increase in moderate-to-severe neuropsychiatric events with varenicline or bupropion versus nicotine patch or placebo, and led the FDA to remove the boxed warning. Smoking induces CYP1A2, so cessation raises clozapine and olanzapine concentrations — plan a dose reduction with level monitoring. In pregnancy, behavioral counseling is first line and financial incentives have the strongest behavioral evidence. Combination NRT (patch plus a short-acting product) beats any single NRT product.
1. Serious Mental Illness
| Issue | Clinical response |
|---|---|
| Very high prevalence and heavy consumption | Screen and treat at every contact; do not defer to "when they are more stable" |
| Neuropsychiatric safety concerns | EAGLES resolved these for varenicline and bupropion; monitor mood as you would with any change, but do not withhold effective medication |
| CYP1A2 induction by polycyclic aromatic hydrocarbons in smoke | On cessation, clozapine and olanzapine levels rise; a reduction of roughly 30 to 40 percent with level monitoring is commonly needed. This is caused by smoke, not nicotine — switching to NRT or vaping does not prevent the interaction |
| Other affected agents | Also relevant for fluvoxamine, duloxetine, theophylline, caffeine and, to a lesser degree, haloperidol |
| Bupropion caution | Lowers seizure threshold; avoid with eating disorder history or abrupt sedative withdrawal; has mild antidepressant effect that can be useful |
| Negative symptoms and cognition | Patients often smoke to self-manage; substitute explicit skills and structure rather than leaving a void |
Exam trap: the most frequently missed item in this area is that the clozapine interaction is caused by combustion products, not nicotine. A patient who switches entirely from cigarettes to the patch still loses the CYP1A2 induction and still needs the dose review.
2. Concurrent Alcohol and Other Drug Use Disorders
- Smoking prevalence in addiction treatment populations remains dramatically elevated, and tobacco is a leading killer of people who successfully stop drinking or using opioids.
- Alcohol is the most common single trigger for first-week smoking relapse. Plan for it explicitly.
- Concurrent treatment improves substance use outcomes; the widespread belief to the contrary is unsupported.
- Varenicline has an additional signal for reduced heavy drinking in some trials, which can be raised as a secondary benefit with patients who have alcohol use disorder.
- Patients in opioid treatment programs smoke at extremely high rates and are a captive, under-served population for whom on-site tobacco treatment is high yield.
3. Pregnancy and Postpartum
| Element | Recommendation |
|---|---|
| First line | Intensive behavioral counseling, ideally more than the minimal contact used in general populations |
| Strongest behavioral evidence | Financial incentives, which have the largest effect of any behavioral strategy in pregnancy |
| Pharmacotherapy | Evidence is limited; NRT may be considered when behavioral treatment fails and the patient continues to smoke, after an explicit risk-benefit discussion. Intermittent forms allow lower total daily exposure; some clinicians prefer them for that reason |
| Varenicline and bupropion | Data in pregnancy are limited; generally not first line |
| Postpartum | Relapse rates are very high in the first six months; schedule proactive contact rather than waiting |
| Secondhand exposure | Counsel on complete home and vehicle smoking bans; "smoking on the porch" leaves thirdhand residue on clothing and surfaces |
4. Adolescents
- Behavioral interventions are first line; pharmacotherapy evidence in adolescents is limited and NRT is used selectively in heavily dependent youth.
- Nicotine-salt pod systems deliver high nicotine efficiently, and dependence can develop quickly; ask about vaping explicitly, since many adolescents do not consider it "smoking."
- Flavors and menthol drive initiation; social media marketing is a major exposure route.
- Confidentiality and family engagement must be balanced according to state minor-consent law.
5. Cardiovascular and Perioperative Patients
- NRT is safe in stable cardiovascular disease, including after myocardial infarction, and is far safer than continued smoking. The older blanket contraindication has not survived the evidence.
- Hospitalization for an acute cardiac event is a teachable moment with unusually high quit success; start treatment in hospital and arrange follow-up within one week of discharge, because inpatient-only intervention loses most of its effect.
- Preoperative cessation reduces wound infection and pulmonary complications; benefit increases with longer preoperative abstinence but is measurable even at shorter intervals.
6. Choosing an Agent: A Compact Decision Aid
| Situation | First choice | Rationale |
|---|---|---|
| Most smokers, no contraindication | Varenicline | Highest single-agent efficacy; EAGLES supports safety in psychiatric populations |
| Patient declines varenicline or it is unavailable | Combination NRT (patch plus gum, lozenge, inhaler or spray) | Outperforms any single NRT product |
| Co-occurring depression or weight concern | Bupropion, alone or with NRT | Antidepressant effect and attenuation of post-cessation weight gain |
| Seizure disorder or eating-disorder history | Avoid bupropion; use varenicline or combination NRT | Seizure risk |
| Severe renal impairment | Varenicline at reduced dose (0.5 mg daily, maximum 0.5 mg twice daily) | Predominantly renal elimination |
| Pregnancy | Behavioral counseling plus incentives; NRT only after risk-benefit discussion | Limited pharmacotherapy safety data |
Test Your Knowledge
A patient maintained on clozapine successfully stops smoking and switches to the nicotine patch. Two weeks later he is sedated, hypersalivating and unsteady. What is the most likely explanation?
A
B
C
D
Test Your Knowledge
A patient with stable schizophrenia asks whether varenicline is safe for someone with a psychiatric diagnosis. What does the EAGLES trial support?
A
B
C
D
Test Your Knowledge
A pregnant patient continues to smoke 10 cigarettes daily despite four counseling sessions. Which approach has the strongest behavioral evidence base in pregnancy?
A
B
C
D