A Randomized Trial of Nortriptyline for Smoking Cessation | Lifestyle Behaviors | JAMA Internal Medicine | JAMA Network
[Skip to Navigation]
Figure 1. 
Trial profile.

Trial profile.

Figure 2. 
Mean (SE) daily anxiety/tense withdrawal symptom score during the first 8 days after quit day. Day 8, P=.002 (Mann-Whitney U test with Bonferroni correction).

Mean (SE) daily anxiety/tense withdrawal symptom score during the first 8 days after quit day. Day 8, P=.002 (Mann-Whitney U test with Bonferroni correction).

Figure 3. 
Validated continuous abstinence rate. Day 180, P =.003 (Fisher exact test).

Validated continuous abstinence rate. Day 180, P =.003 (Fisher exact test).

Table 1. 
Baseline Characteristics*
Baseline Characteristics*
Table 2. 
Tobacco Withdrawal Symptoms*
Tobacco Withdrawal Symptoms*
Table 3. 
Adverse Reactions*
Adverse Reactions*
Original Investigation
October 12, 1998

A Randomized Trial of Nortriptyline for Smoking Cessation

Author Affiliations

From the Ambulatory Care, Denver Veterans Affairs Medical Center, Denver, Colo (Dr Prochazka and Mss Licari and Lofaso); the Division of General Internal Medicine (Dr Prochazka and Ms Licari) and the Department of Family Medicine (Dr Fryer), University of Colorado Health Sciences Center, Denver; Fitzsimons Army Medical Center, Aurora, Colo (Dr Weaver and Mr Keller); Inteck Incorporated, Denver (Dr Weaver); and the 528th Medical Detachment, United States Army, Fort Bragg, NC (Mr Keller).

Arch Intern Med. 1998;158(18):2035-2039. doi:10.1001/archinte.158.18.2035

Background  Smoking cessation rates with current therapy are suboptimal. One class of drugs that may improve cessation is the tricyclics.

Objective  To add nortriptyline hydrochloride to a behavioral smoking cessation program to enhance cessation rates and reduce withdrawal symptoms.

Subjects and Methods  We conducted a randomized, double-blind, placebo-controlled trial at an affiliated Department of Veterans Affairs Medical Center and an Army Medical Center. Subjects were aged 18 through 70 years, smoked 10 or more cigarettes per day, and were without current major depression. Nortriptyline hydrochloride or matched placebo was started at 25 mg before bed 10 days prior to quit day and titrated to 75 mg/d or to the maximal tolerated dose. The behavioral intervention consisted of 2 group sessions and 12 individual follow-up visits. Withdrawal symptoms were measured using a daily diary, and smoking cessation was defined as self-reported abstinence, expired carbon monoxide of 9 ppm or less, and a 6-month urine cotinine level of less than 50 ng/mL.

Results  A total of 214 patients were randomized (108 to nortriptyline and 106 to placebo). There was a significant reduction in several withdrawal symptoms including anxious/tense, anger/irritability, difficulty concentrating, restlessness, and impatience by day 8 after quit day in the nortriptyline group. The cessation rate at 6 months was 15 (14%) of 108 and 3 (3%) of 106, respectively (P=.003; absolute difference, 11%; 95% confidence interval, −18% to −4%). Nortriptyline caused frequent adverse effects, including dry mouth (64%) and dysgeusia (20%).

Conclusions  We conclude that nortriptyline led to an increased short-term cessation rate compared with placebo. In addition, there were significant, but relatively small, reductions in withdrawal symptoms. Nortriptyline may represent a new therapeutic approach to smoking cessation.

SMOKING IS an important preventable cause of mortality; however, current approaches to cessation are only partially successful.1,2 The relationship between depressed mood and smoking behavior3,4 suggests that antidepressant drugs might have a role in smoking cessation. Several antidepressants, including bupropion,5 doxepin,6 nortriptyline,7 and moclobemide,8 have shown some effectiveness in smoking cessation.

This study was designed to test our hypothesis that the addition of a tricyclic to a behavioral smoking cessation program would increase cessation rates. We chose nortriptyline as the experimental agent because it is generally well tolerated, even in elderly subjects, and has a known therapeutic blood level range.9-12

Subjects and methods

This was a double-blind, randomized, controlled trial of nortriptyline vs placebo in subjects who were enrolled in a behavioral smoking cessation program.


The sites for the study were Fitzsimons Army Medical Center, an urban military teaching hospital in Aurora, Colo, and the Denver Veterans Affairs Medical Center in Denver, Colo. This project was approved by the institutional review boards of both institutions.


Subjects were recruited from the outpatient clinics and from campus advertisements and provided written informed consent before participation. Inclusion criteria were readiness to set a quit date in the next 3 weeks, age 18 through 70 years, and currently smoking at least 10 cigarettes per day. Potential subjects were excluded for current major depression or other significant psychiatric disorder, alcohol or drug dependence, significant cardiopulmonary or liver disease, active malignancy, pregnancy, contraindication or allergy to nortriptyline, use of thyroid or anticonvulsant medications, other therapy for smoking cessation, or inability to give informed consent.

Baseline measurements

We used a questionnaire to assess smoking history and estimated the degree of nicotine dependence with the Fagerstrom Nicotine Dependence Questionnaire.13 The Computerized Diagnostic Interview Schedule14 was used to identify subjects with current major depression or prior depression. All subjects had a medical history taking and physical examination performed by a physician or nurse practitioner along with a 12-lead electrocardiogram to identify other potential exclusions. Women of childbearing potential had a serum pregnancy test at the qualification visit and were asked to use accepted means of contraception during the study period.

Smoking cessation

Subjects set a quit date within 21 days of study entry and participated in 2 behavioral group sessions based on an American Cancer Society model to prepare for quitting before the start of drug therapy.


Subjects were stratified by study site and randomized after the completion of the group smoking cessation sessions.

Drug therapy

Drug therapy began after completion of the group sessions. Subjects took 1 capsule of the study drug before bedtime, either nortriptyline hydrochloride (25 mg) or placebo, 10 days before quit day, then increased to 2 capsules per day (50 mg of nortriptyline hydrochloride) after 3 days, and finally after 3 more days increased to 3 capsules per day (75 mg of nortriptyline hydrochloride), if tolerated. They continued receiving the maximal tolerated dose for 8 weeks after their quit date.

All visits while subjects were receiving study drugs were individual sessions with the study nurse who reviewed problems quitting and helped develop strategies for abstinence from smoking. The study drug was dispensed in weekly increments and the returned pills were collected at each visit to reduce the possibility of overdose and to monitor compliance. If adverse effects developed, the dosage was reduced as necessary.

Blood samples were drawn 1 week after quit day, when subjects had been taking the maximal dose of study drug for at least 11 days. An unblinded research pharmacist recommended dosage reductions for those above the therapeutic range (≥150 ng/mL) and dosage increases for those who were subtherapeutic (<50 ng/mL). To maintain blinding, dose reductions and increases on an equal number of randomly selected placebo-treated subjects were also recommended. The dosage of study drug was tapered to 50 mg/d beginning 8 weeks after quit day (day 56). All subjects stopped receiving study drug by week 10 (day 70).

Expired carbon monoxide was determined using a Bedfont expired carbon monoxide analyzer (Bedfont Technical Instruments Ltd, Sittingbourne, Kent, England). Urine cotinine levels were measured using a radioimmunoassay with gas chromatographic/mass spectroscopic confirmation by the Toxicology Laboratory, Colorado Department of Health, Denver.

Subjects completed a daily diary of the number of cigarettes smoked for the first week after quit day and selected nicotine withdrawal symptoms rated on a 0- to 5-point scale, with 5 representing severe symptoms.15 We used the Beck Depression Inventory to identify depressive symptoms at baseline, 1 week after cessation, and at the end of nortriptyline therapy.16

Statistical analysis

The primary outcome was sustained smoking abstinence, defined as self-reported cessation within 1 week of the quit day, expired carbon monoxide of 9 ppm or less at each visit, and verified by urine cotinine level of less than 50 ng/mL at the final visit (6 months).17 The sample size for the study (100 per group) was estimated based on assumptions of an 80% power, α of .05, and a 10% absolute difference in the 6-month validated quit rates between the groups. Data were analyzed using SPSSPC software (SPSS Inc, Chicago, Ill) and Stat-Exact 3 (Cytel Software Corporation, Cambridge, Mass). We compared the groups for continuous variables using t tests and Mann-Whitney U tests, as appropriate, and for categorical variables using χ2 tests. We used the Bonferroni adjustment for the analysis of the 8 daily withdrawal symptom scores, leading to critical value for statistical significance of P<.006.18 We compared the 6-month cessation rates in the 2 groups using Fisher exact tests based on an intention-to-treat analysis.


We screened 269 patients, of whom 26 were ineligible and 29 failed to attend the scheduled cessation classes, leaving 214 who were randomized (108 to nortriptyline and 106 to placebo) (Figure 1). Table 1 shows there were no significant differences between the baseline characteristics of the study groups. Only 26 (12%) of 214 patients had a history of depression as assessed using the Computerized Diagnostic Interview Schedule. Of the patients who continued in the study until 1 week after quit day, 77 (85%) of 91 nortriptyline subjects and 80 (85%) of 94 placebo subjects were receiving the 3-capsule (75 mg of nortriptyline hydrochloride) target dose. The median nortriptyline level at the final drug dose was 67 ng/mL (range, 25-146 ng/mL). There was a statistically significant reduction (P<.006, Mann-Whitney U test with Bonferroni adjustment) in being anxious/tense (Figure 2) and several other withdrawal symptoms by 1 week after quit day including irritability/anger, difficulty concentrating, restlessness, impatience, and insomnia, but not for craving, excess hunger, drowsiness, and increased eating (Table 2).

The validated cessation rate over the follow-up period was higher in the nortriptyline group at every time point. The nortriptyline group had a 6-month sustained abstinence rate higher than the placebo group (absolute difference, 11%; 95% confidence interval, 4%-18%) (Figure 3).

There was no significant effect on the quit rate for history of depression, baseline pack-years smoked, sex, age, Fagerstrom score, or Beck Depression Inventory scores.

Adverse effects were significantly higher in the nortriptyline group compared with the placebo group (Table 3). Nortriptyline therapy was discontinued due to adverse events in 10 of 108 subjects (4 with constipation and indigestion, 1 with rash, 1 with trouble urinating, 2 with subjective elevated heart rate, 1 with micturition syncope and fractured humerus, and 1 with hot flashes) and placebo was discontinued in 3 of 106 subjects (1 with headache and sleep disturbance, 1 with dry mouth and abdominal pain, and 1 with constipation).

At the end of drug therapy the study nurse identified the treatment assignment correctly in 78% of cases, while subjects were correct 67% of the time.


This study demonstrates that the addition of nortriptyline to a smoking cessation program results in improved smoking cessation rates. These results are similar to the preliminary data presented by Humfleet et al,7 who found a short-term cessation rate of 50% with nortriptyline at 12 weeks compared with 35% with placebo. We also observed an improvement in withdrawal symptoms and an increase in anticholinergic adverse effects, which were typical of the drug and are greater than those seen with nicotine replacement therapy.

Smoking behavior is determined by a complex interaction of biological, psychological, and behavioral factors.19 The mechanisms of nicotine dependence and withdrawal are incompletely understood.20 Nevertheless, there are several possible mechanisms of action for nortriptyline's effect in enhancing smoking cessation. Nortriptyline may reduce depressive symptoms21 and the need for "negative affect reduction smoking."22,23 Other antidepressant agents (bupropion5 and moclobemide7) are also effective in smoking cessation, suggesting the antidepressant effect may be the common mechanism. However, we showed no significant difference in Beck Depression Inventory scores between subjects in the nortriptyline and placebo groups, indicating that treatment of depressive symptoms is probably not the only mechanism.

Anxiety and stress are important factors in stimulating smoking and preventing cessation, and nortriptyline has antianxiety effects.24 We observed a significant reduction in anxiety/tension and irritability/anger in the nortriptyline group, which may partially explain its efficacy.

It is also possible that antidepressants suppress the symptoms of nicotine withdrawal with central noradrenergic receptor systems. Nortriptyline affects a number of neurotransmitter systems, predominantly acting to block reuptake of norepinephrine with a lesser effect on serotonin.25,26 Other drugs that are effective in smoking cessation, including bupropion and clonidine, also have effects on the central noradrenergic systems.27

The cessation effect of nortriptyline may be due to anticholinergic actions,28 especially the dry mouth and taste changes. Subjects in this study frequently reported that cigarettes "did not taste good" when they were receiving the study drug, reminiscent of the frequent complaint of patients with hepatitis. Rose29 and Westman et al30 have demonstrated the importance of upper airway sensory stimulation in smoking and cessation. Most likely, the beneficial effects of nortriptyline use are due to a combination of these, and perhaps other mechanisms, with different mechanisms being more or less important in different individuals.

There are several limitations that must be kept in mind when interpreting the results of this study. First, we required subjects to be smoking 10 cigarettes per day at the time of study entry, and many were trying to quit and had already reduced the number of cigarettes smoked before their formal quit date. If they were already past the peak of their withdrawal symptoms, the apparent benefits of nortriptyline use would have been minimized. Second, our blinding was only partially effective. Because of the high frequency of dry mouth, the nurse and subjects were often able to identify the active drug. Third, the relatively small sample size limits our analysis of the effect of potential predictors of cessation such as sex, prior depression, level of nicotine dependence, and number of prior quit attempts. Since we enrolled fewer subjects with symptoms of depression than anticipated, we cannot determine whether nortriptyline would be more effective in depressed smokers.

We have demonstrated the efficacy of nortriptyline in smoking cessation; however, the ideal dosage and duration of treatment remain to be determined. The optimal effect of nortriptyline use may also require a longer precessation period of drug therapy. The relapse in the nortriptyline group after the drug was discontinued suggests that a longer duration of treatment may be more effective. The role of nortriptyline compared with or in combination with other agents is not yet known. Nortriptyline may prove to be most useful in those smokers who have failed standard smoking cessation therapy.

Accepted for publication March 12, 1998.

This study was sponsored by a grant from the Department of Veterans Affairs and the US Department of Defense.

The opinions expressed are the private views of the authors and do not represent official statements of the Department of Veterans Affairs or the US Department of Defense.

We acknowledge the assistance of Stephen Bartlett, RPh, MSc, Anita Huttenhower, RPh, Barbara Martin, Patricia Schoch, RPh, Jaime Soria, ANP, and Thomas Miyoshi with this project.

Reprints: Allan V. Prochazka, MD, MSc, Ambulatory Care 11B, 1055 Clermont, Denver, CO 80220 (e-mail: prochaza@essex.uchsc.edu).

Kottke  TEBattista  RNDeFriese  GHBrekke  ML Attributes of successful smoking cessation interventions in medical practice.  JAMA. 1988;2592883- 2889Google ScholarCrossref
Fiore  MCSmith  SSJorenby  DEBaker  TB The effectiveness of nicotine patch for smoking cessation: a meta-analysis.  JAMA. 1994;2711940- 1947Google ScholarCrossref
Glassman  AHStetner  FWalsh  BT  et al.  Heavy smokers, smoking cessation and clonidine: results of a double-blind, randomized trial.  JAMA. 1988;2592863- 2866Google ScholarCrossref
Breslau  NKilbey  MMAndreski  P Nicotine withdrawal symptoms and psychiatric disorders: findings from an epidemiologic study of young adults.  Am J Psychiatry. 1992;149464- 469Google Scholar
Hurt  RDSachs  DPLGlover  ED  et al.  A comparison of sustained-release bupropion and placebo for smoking cessation.  N Engl J Med. 1997;3371195- 1202Google ScholarCrossref
Edwards  NBSimmons  RCRosenthal  TLHoon  PWDowns  JM Doxepin in the treatment of nicotine withdrawal.  Psychosomatics. 1988;29203- 206Google ScholarCrossref
Humfleet  GHall  SReus  V  et al.  The efficacy of nortriptyline as an adjunct to psychological treatment for smokers with and without depressive histories. Harris  LSed. Problems of Drug Dependence Rockville, Md US Dept of Health and Human Services1996;334Google Scholar
Berlin  ISaid  SSpreux-Varoquaux  O  et al.  A reversible monoamine oxidase A inhibitor (moclobemide) facilitates smoking cessation and abstinence in heavy, dependent smokers.  Clin Pharmacol Ther. 1995;58444- 452Google ScholarCrossref
Georgotas  AMcCue  RECooper  BA A placebo-controlled comparison of nortriptyline and phenelzine in maintenance therapy of elderly depressed patients.  Arch Gen Psychiatry. 1989;46783- 786Google ScholarCrossref
Fabre  LFScharf  MBItil  TM Comparative efficacy and safety of nortriptyline and fluoxetine in the treatment of major depression: a clinical study.  J Clin Psychiatry. 1991;52(suppl)62- 67Google Scholar
Reynolds  CFFrank  EPerel  JM Combined pharmacotherapy and psychotherapy in the acute and continuation treatment of elderly patients with recurrent major depression: a preliminary report.  Am J Psychiatry. 1992;1491687- 1692Google Scholar
Borson  SMcDonald  GJGayle  T  et al.  Improvement in mood, physical symptoms and function with nortriptyline for depression in patients with chronic obstructive pulmonary disease.  Psychosomatics. 1992;33190- 201Google ScholarCrossref
Fagerstrom  KO Measuring degree of physical dependence to tobacco smoking with reference to individualization of treatment.  Addict Behav. 1978;3235- 241Google ScholarCrossref
Bucholz  KKRobins  LNShayka  JJ  et al.  Performance of two forms of a computer psychiatric screening interview: version I of the DISSI.  J Psychiatr Res. 1991;25117- 129Google ScholarCrossref
Gritz  ERCarr  CRMarcus  AC The tobacco withdrawal syndrome in unaided quitters.  Br J Addict. 1991;8657- 69Google ScholarCrossref
Beck  ATRial  WYRickels  K Short form of the Depression Inventory: cross validation.  Psychol Rep. 1974;341184- 1186Google Scholar
Perez-Stable  EJMarin  BVMarin  GBrody  DJBenowitz  NL Apparent under-reporting of cigarette consumption among Mexican American smokers.  Am J Public Health. 1990;801057- 1061Google ScholarCrossref
Fleiss  JL The Design and Analysis of Clinical Experiments.  New York, NY John Wiley & Sons Inc1986;103- 107
Carmody  TP Affect regulation, nicotine addiction and smoking cessation.  J Psychoactive Drugs. 1989;21331- 342Google ScholarCrossref
Pomerleau  OFPomerleau  CS Neuroregulators and the reinforcement of smoking: towards a biobehavioral explanation.  Neurosci Biobehav Rev. 1984;8503- 513Google ScholarCrossref
Covey  LSGlassman  AHStetner  F Major depression following smoking cessation.  Am J Psychiatry. 1997;154263- 266Google Scholar
Ikard  FFTomkins  S The experience of affect as a determinant of smoking behavior.  J Abnorm Psychol. 1973;81172- 181Google ScholarCrossref
Shiffman  S Relapse following smoking cessation: a situational analysis.  J Consult Clin Psychol. 1982;5071- 86Google ScholarCrossref
Potter  WZManji  HKRudorfer  MV Tricyclics and tetracyclics. Schatzberg  AFNemeroff  CBeds. Textbook of Psychopharmacology Washington, DC American Psychiatric Press Inc1995;141- 160Google Scholar
Dilsaver  SCPariser  CFChurchill  CMLarson  CN Is there a relationship between failing efforts to stop smoking and depression?  J Clin Psychopharmacol. 1990;10153- 154Google ScholarCrossref
Corrigal  WA Understanding brain mechanisms in nicotine reinforcement.  Br J Addict. 1991;86507- 510Google ScholarCrossref
Gourlay  SGBenowitz  NL Is clonidine an effective smoking cessation therapy?  Drugs. 1995;50197- 207Google ScholarCrossref
Bachynsky  N The use of anticholinergic drugs for smoking cessation: a pilot study.  Int J Addict. 1986;21789- 805Google Scholar
Rose  JE The role of upper airway stimulation in smoking. Pomerleau  OFPomerleau  CSeds. Nicotine Replacement A Critical Evaluation New York, NY Alan R Liss Inc1988;95- 106Google Scholar
Westman  ECBehm  FMRose  JE Airway sensory replacement combined with nicotine replacement for smoking cessation: a randomized, placebo-controlled trial using a citric acid inhaler.  Chest. 1995;1071358- 1364Google ScholarCrossref