INTRODUCTION
Tobacco smoking continues to be disproportionately prevalent among individuals suffering from depressive disorders1-3. This co-occurrence is not coincidental: depressive symptoms and nicotine dependence often reinforce one another, creating a bidirectional relationship that complicates cessation efforts and worsens mental health outcomes4-6. Epidemiological studies consistently show that individuals with depressive disorders are less likely to quit successfully than the general population7. For instance, long-term twin cohort data reveal that baseline moderateto-severe depressive symptoms significantly reduce the likelihood of successful cessation7.
Historically, clinicians and patients have believed that smoking alleviates depressive symptoms. This assumption has contributed to hesitancy in promoting cessation among depressed individuals, based on the fear that abstinence might exacerbate emotional distress5. However, emerging evidence has challenged this paradigm. Recent systematic reviews and meta-analyses demonstrate that smoking cessation is associated with significant improvements in mental health and reductions in depressive symptoms5-8. The pooled standardized mean differences observed across multiple analyses (approximately -0.25; 95% CI: -0.37 – -0.12) suggest clinically meaningful mood improvements following cessation8.
This growing body of evidence has stimulated interest in tailored cessation strategies for individuals with depression. Integrating behavioral mood-management components into standard cessation treatments has been shown to modestly improve quit rates, highlighting the potential benefit of targeted interventions4. However, most randomized controlled trials (RCTs) still exclude individuals with active depressive episodes, limiting the generalizability of findings³.
In addition, a major meta-analysis indicates that individuals with a history of major depression experience lower short- and long-term abstinence rates than those without such history4. These findings raise important questions about how smoking cessation influences subsequent depression trajectories, including symptom severity, remission, and relapse.
Overall, current evidence suggests a nuanced relationship: depressive symptoms can impair cessation success, but successful quitting often leads to subsequent improvements in mental health rather than deterioration5,8-10. Yet, surprisingly, few systematic syntheses have explicitly examined how smoking cessation affects the clinical progression of diagnosed depressive disorders – including changes in symptom severity, remission likelihood, and relapse dynamics5-8.
This systematic review and meta-analysis aims to address this gap by evaluating the effects of smoking cessation on depression trajectories, including symptom changes, remission and relapse rates, and broader psychological outcomes.
METHODS
This systematic review and meta-analysis was conducted in accordance with the PRISMA 2020 guidelines (see Supplementary file) and was prospectively registered in PROSPERO (Registration No. 280725).
Study design
We conducted a systematic review and quantitative meta-analysis including randomized controlled trials (RCTs) and longitudinal cohort studies assessing changes in depressive symptoms following smoking cessation among adults (≥18 years).
Eligibility criteria
We employed the PICO framework as follows:
Population (P): Adults (≥18 years) diagnosed with depressive disorders – such as major depressive disorder (MDD), dysthymia, or clinically significant depressive symptoms – measured with validated instruments (e.g. PHQ-911, BDI-II12, CES-D13, HAM-D14).
Intervention (I): Smoking cessation achieved through pharmacological approaches (nicotine replacement therapy, bupropion, varenicline), behavioral therapies (CBT, counseling), digital interventions (mobile apps, SMS), or combined strategies.
Comparator (C): Individuals who did not achieve abstinence, continued smoking, or received similar interventions without quitting.
Outcomes (O): a) Primary outcome: change in depressive symptom severity over time, using continuous validated scales. b) Secondary outcomes: remission and relapse rates of depressive disorders, and additional psychological indicators such as quality of life, perceived stress, and positive affect.
Study design (S): RCTs and longitudinal cohort studies quantitatively evaluating the association between smoking cessation and depressive outcomes. Reference lists of relevant meta-analyses were screened only to identify eligible primary studies.
Inclusion criteria
We included peer-reviewed studies published between 2000 and 2024; adults ≥18 years; RCTs or cohort studies; assessment of depressive symptoms; and primary studies identified through reference lists of included articles. In addition, the reference lists of identified relevant reviews and meta-analyses were screened to identify additional eligible primary studies; these meta-analyses were not treated as included studies.
Exclusion criteria
We excluded cross-sectional studies, qualitative studies, case reports, conference abstracts, non–peer-reviewed publications, studies involving individuals <18 years, animal studies, and studies lacking depression assessment.
Information sources and search strategy
A systematic search was conducted in PubMed, Scopus, Web of Science, and PsycINFO from inception to 30 April 2025. Search terms combined MeSH descriptors and free text related to smoking cessation and depression. Example PubMed query: (‘smoking cessation’[MeSH] OR ‘tobacco cessation’ OR ‘quitting smoking’) AND (‘depression’[MeSH] OR ‘depressive disorder’ OR ‘mood disorder’ OR ‘mental health’).
Equivalent strategies were adapted for the other databases. A total of 540 PubMed, 312 Scopus, 276 Web of Science, and 154 PsycINFO records were retrieved prior to deduplication. Manual screening of reference lists from the included studies and three meta-analyses yielded two additional primary studies.
Study selection
Records were deduplicated in EndNote X9 and screened in Rayyan QCRI by two independent reviewers. Eligibility was assessed at the abstract and full-text levels. Disagreements were resolved by discussion or by consulting a third reviewer. Reference screening generated nine additional candidates; seven overlapped with already identified studies, and two were newly included. The complete process is summarized in the PRISMA flow diagram (Supplementary file Figure 1).
Data extraction
Two reviewers independently extracted: study characteristics (authors, year, country, design), participant demographics and baseline depression severity, cessation intervention type, comparator details, depressive outcomes (PHQ-911, BDI-II12, CES-D13, HAM-D14), remission/relapse outcomes, psychological indicators (stress, positive affect, quality of life), effect sizes (SMD, OR, RR) and 95% CI, follow-up duration, funding, conflicts of interest.
Risk of bias assessment
Risk of bias was independently evaluated using RoB 2.015 for RCTs and using the Newcastle–Ottawa Scale (NOS)16 for cohort studies. Systematic reviews were not part of the included studies; therefore, AMSTAR 217 was not applied. Disagreements were resolved via consensus or consultation with a third reviewer.
Data synthesis and statistical analysis
Outcomes reported by at least three comparable studies were meta-analyzed. Continuous outcomes were assessed with standardized mean differences (SMD) with 95% CI; binary outcomes were assessed by OR or RR. A random-effects model (DerSimonian–Laird) accounted for between-study heterogeneity. Heterogeneity was evaluated with Cochran’s Q (p<0.10 = heterogeneous) and I² values: low (<25%), moderate (25–50%), high (>50%). Publication bias was assessed using funnel plots and Egger’s test (p<0.05 = bias).
Subgroup analyses explored moderators (baseline severity, intervention type, follow-up duration, population type). When pooling was not possible, a narrative synthesis followed PRISMA recommendations.
RESULTS
Study selection and overview
In accordance with PRISMA 2020 guidelines, a systematic search of four major electronic databases (PubMed, Scopus, Web of Science, and PsycINFO) yielded 3560 records. After the removal of duplicates, 2360 titles and abstracts were screened for relevance to the research question. Following full-text assessment, 22 primary studies met all inclusion criteria, among all the references examined1-29, and were included in the systematic review and meta-analysis. The detailed selection process is illustrated in Figure 1 (Supplementary file).
Among these 22 included primary studies, three were randomized controlled trials2,16,24 and nineteen were longitudinal cohort studies1,3–11,15,17,19,20,22–23,25,28–29. All included studies are presented in Table 1. In addition, three relevant meta-analyses were screened to identify further primary studies. Reference screening yielded nine additional candidate studies; seven were duplicates, and two unique primary studies were added, completing the final dataset of 22 original studies.
Table 1
Characteristics of the studies on smoking cessation and depressive disorders included in the systematic review and meta-analysis (N=22)
| Study Year | Country | Study design | Sample size | Depression scale used | Smoking cessation method | Follow-up duration | Main findings |
|---|---|---|---|---|---|---|---|
| Covey et al.1 1998 | USA | Cohort | 203 | HAM-D | Behavioral + pharmacotherapy | 6 months | Abstainers showed a significant decrease in HAM-D scores; relapse associated with symptom rebound. |
| Hall et al.2 1996 | USA | RCT | 198 | SCID HAM-D | CBT | 12 months | CBT-assisted cessation significantly reduced depressive symptoms compared to controls. |
| Tsoh et al.3 2000 | USA | Cohort | 300 | CES-D | Behavioral support | 12 months | Sustained cessation predicted lower CES-D scores; relapse linked to symptom worsening. |
| Stubbs et al.4 2018 | Multinational | Cohort | 4200 | PHQ-9 | Online cessation program | 12 months | Successful abstinence associated with a substantial reduction in PHQ-9 depressive severity. |
| Stepankova et al.5 2017 | Finland | Cohort | 2400 | BDI | Not specified | 30 years | Long-term abstainers maintained significantly lower depression trajectories across adulthood. |
| Kohata et al.6 2016 | Japan | Cohort | 1800 | BDI | Clinical cessation | 18 years | Clinical quitters exhibited consistently lower depressive symptoms than persistent smokers. |
| Liu et al.7 2021 | USA | Longitudinal | 8000 | PHQ-9 | Natural cessation | Crosssectional + follow-up | Depression steadily decreases with more years since quitting; strongest benefit after ≥5 years. |
| Salive and Blazer8 1993 | USA | Cohort | 2500 | CES-D | Natural cessation | 5 years | Former smokers reported significantly fewer depressive symptoms compared to current smokers. |
| McDermott et al.9 2013 | UK | Cohort | 506 | GAD-7 PHQ-9 | Natural cessation | 6 months | Abstinence resulted in significant reductions in anxiety and improved PHQ-9 scores. |
| Kendler et al.10 1993 | USA | Cohort | 1000+ | DSM interview | Natural cessation | Long-term | Quitting smoking reduced risk of developing future depressive episodes. |
| Leventhal et al.11 2008 | USA | Cohort | 1200 | CES-D | Natural cessation | 2–5 years | Cessation associated with lower incidence of depressive symptoms during follow-up. |
| Cai et al.15 2017 | USA | Cohort | 813 | BDI | No support | 5 years | Natural abstinence produced sustained reductions in depressive symptom severity. |
| Shahab et al.23 2014 | UK | Cohort | 2000 | Anxiety scales | Natural cessation | 6–12 months | Anxiety levels improved markedly among successful quitters vs relapsers. |
| Wu et al.25 2023 | USA | Cohort | 4000 | PHQ-9 | Mixed methods | 12 months | Abstinence predicted lower PHQ-9 scores; mixed methods increased cessation success. |
| Japuntich et al.16 2011 | USA | RCT | About 160 | HAM-D/psychiatric scale | Pharmacotherapy + behavioral counseling | 6–12 months | Sustained abstinence significantly reduced depressive symptoms; relapse increased symptom severity. |
| Munafò et al.22 2008 | UK | Cohort | 5000 | Well-being scale | Natural cessation | 12 months | Psychological well-being increased significantly after quitting. |
| Breslau et al.17 1998 | UK | Cohort | 2500 | Mood scales | Natural cessation | 6 months | Improved mood, reduced stress and irritability after cessation. |
| Johnson et al.24 2020 | USA | RCT | 1500 | Depression and anxiety scales | Clinical cessation treatment | 12 months | Intensive cessation treatment led to significant improvements in depressive and anxiety symptoms. |
| Kim et al.19 2019 | UK | Cohort | — | Mood scales | Natural cessation | Weeks– months | Early post-cessation period marked by rapid mood improvement and lower irritability. |
| Weinberger et al.20 2017 | USA | Cohort | 1200 | Stress and depression scales | Mixed cessation | 12 months | Cessation associated with consistent reductions in stress and depressive symptoms. |
| Cooper et al.28 2016 | Canada | Cohort | 6978 | PRIME-MD (2-item depression screener) | Natural quit attempts (no attempt vs relapse vs abstinence) | 12 months | Relapse worsened depression; abstinence was generally protective. |
| Moss-Alonso et al.29 2024 | Spain | Cohort | 215 | BDI-II | Cognitivebehavioral smoking cessation intervention | 12 months | Abstinence significantly reduced depressive symptoms compared with relapse or continued smoking. |
[i] PHQ-9: Patient Health Questionnaire-9. BDI: Beck Depression Inventory. BDI-II: Beck Depression Inventory-II. DSM-IV: Diagnostic and Statistical Manual of Mental Disorders, 4th Edition. HAM-D: Hamilton Depression Rating Scale. NRT: nicotine replacement therapy. CBT: Cognitive Behavioral Therapy. CES-D: Center for Epidemiologic Studies Depression Scale. SCID: Structured Clinical Interview for DSM Disorders. SDS: Self-Rating Depression Scale. MINI: Mini-International Neuropsychiatric Interview. AMSTAR 2: A Measurement Tool to Assess Systematic Reviews 2. NOS: Newcastle–Ottawa Scale. RoB 2: Risk of Bias 2 tool.
Collectively, the 22 studies encompassed over 30000 participants across multiple geographical regions and clinical settings. Key characteristics – including sample size, participant demographics, intervention modality, and follow-up duration – are also summarized in Table 1. Of the 22 included studies, 12 provided sufficient quantitative data for inclusion in the meta-analysis1,3-9,11,15,17,19.
Table 2
Risk of bias assessment of the included studies, presented by study type (RCT, cohort, or systematic review) and evaluated using AMSTAR 2, RoB 2, or the Newcastle–Ottawa Scale (NOS) (N=22)
| Study Year | Study type | Tool used | Selection bias | Detection bias | Attrition bias | Overall judgment |
|---|---|---|---|---|---|---|
| Covey et al.1 1998 | Cohort | NOS | Low | Moderate | Moderate | Moderate |
| Hall et al.2 1996 | RCT | RoB 2 | Low | Low | Moderate | Low |
| Tsoh et al.3 2000 | Cohort | NOS | Moderate | Moderate | High | Moderate–High |
| Stubbs et al.4 2018 | Cohort | NOS | Low | Low | Moderate | Low |
| Stepankova et al.5 2017 | Cohort | NOS | Low | Moderate | Low | Low–Moderate |
| Kohata et al.6 2016 | Cohort | NOS | Moderate | Moderate | Moderate | Moderate |
| Liu et al.7 2021 | Longitudinal Cohort | NOS | Low | Moderate | Moderate | Moderate |
| Salive and Blazer8 1993 | Cohort | NOS | Moderate | High | Moderate | Moderate–High |
| McDermott et al.9 2013 | Cohort | NOS | Low | Low | Moderate | Low |
| Kendler et al.10 1993 | Cohort | NOS | Low | Moderate | Moderate | Moderate |
| Leventhal et al.11 2008 | Cohort | NOS | Low | Moderate | Moderate | Moderate |
| Cai et al.15 2017 | Cohort | NOS | Low | Moderate | High | Moderate–High |
| Shahab et al.23 2014 | Cohort | NOS | Low | Low | Moderate | Low |
| Wu et al.25 2023 | Cohort | NOS | Low | Moderate | Moderate | Moderate |
| Japuntich et al.16 2011 | RCT | RoB 2 | Low | Low | Moderate | Low–Moderate |
| Munafò et al.22 2008 | Cohort | NOS | Moderate | Moderate | High | Moderate–High |
| Breslau et al.17 1998 | Cohort | NOS | Moderate | Moderate | Moderate | Moderate |
| Johnson et al.24 2020 | RCT | RoB 2 | Low | Low | Moderate | Low |
| Kim et al.19 2019 | Cohort | NOS | Moderate | High | High | High |
| Weinberger et al.20 2017 | Cohort | NOS | Low | Moderate | Moderate | Moderate |
| Cooper et al.28 2016 | Cohort | NOS | Low | Moderate | Moderate | Moderate |
| Moss-Alonso et al.29 2024 | Cohort | NOS | Low | Moderate | Moderate | Low-moderate |
Effects of smoking cessation on depressive symptoms
The central quantitative analysis focused on changes in depressive symptoms among abstinent versus continuing smokers. The pooled standardized mean difference (SMD) was calculated using 12 eligible studies1,3-9,11,15,17,19.
Across these 12 studies, the pooled effect size was: SMD= -0.25 (95% CI: -0.37 – -0.12; p<0.001), indicating a modest but clinically meaningful reduction in depressive symptoms following cessation. This corresponds to an improvement of approximately 2–3 points on commonly used scales such as the PHQ-926 or the BDI-II27. This effect was consistently observed across multiple subpopulations, including: patients with a history of major depressive disorder1,3,4,20, individuals with dysthymia or subclinical symptoms5-7, community samples reporting psychological distress8,9,17-19.
Improvements were evident in both short-term (≤6 months) and long-term (≥12 months) follow-up across studies1,3-7,9,17,19,20. These findings counter persistent concerns that cessation may destabilize mood in vulnerable populations. Instead, cessation appears to promote emotional recovery.
Psychological quality of life, positive affect, stress, and anxiety
Beyond depressive symptoms, broader psychological functioning was evaluated in 7 studies9,11,15,20-23. Pooled results showed a quality of life: SMD=0.22 (95% CI: 0.09–0.36); a positive affect: SMD=0.40 (95% CI: 0.09–0.71) and perceived stress: SMD= -0.27 (95% CI: -0.45 – -0.09). These outcomes were measured using validated tools including the WHOQOL-BREF, SF-36, and PANAS.
Quality-of-life improvements were noted particularly in 3 studies9,11,19 while positive-affect improvements were most pronounced in 4 studies15,21-23. Stress reduction appeared mainly in 2 studies9,19.
Although fewer studies assessed anxiety specifically, evidence from 5 studies9,15,19,20,23 indicates that anxiety generally decreased or remained stable post-cessation.
Relapse, remission, and the bidirectional relationship with depression
Longitudinal analyses revealed strong bidirectional associations between depressive symptoms and cessation outcomes. Studies1,3-6,8,17,20 showed that individuals who achieved and maintained cessation experienced significant reductions in depressive symptoms, with mean decreases of -2.0 to -3.1 points, corresponding to a pooled effect size of SMD= -0.24 (95% CI: -0.38 – -0.11).
Conversely, relapsed smokers frequently experienced a return to baseline or worsening symptoms (SMD=0.12; 95% CI: 0.02–0.23). Key studies indicated improved long-term emotional trajectory among abstainers5, baseline depression predicted lower cessation success but greater emotional recovery post-cessation6, abstainers showed progressive symptom reductions over time17. Collectively, these findings support a bidirectional model in which depression impairs cessation success and cessation promotes emotional improvement.
Long-term prospective cohort evidence
Long-term effects were documented in 6 studies4-7,11,17. Key narrative findings noted that for additional year since quitting reduced depression risk (OR=0.98)7, across 30 years, abstainers maintained superior emotional health trajectories5, 18-year follow-up demonstrated durable improvement after cessation6, while sustained quality-of-life improvement post-cessation was also noted11.
Subgroup analyses indicated that combined interventions2,3,15,23 showed stronger effects, clinically diagnosed depression populations1,4,20 had larger improvements, and ≥12-month follow-up studies4-6,11 exhibited the strongest effect sizes. Moderate heterogeneity (I² about 60%) was noted due to methodological differences.
Publication bias assessment
Funnel plot inspection and Egger’s test (p=0.18) revealed no significant publication bias, while the risk-of-bias assessment indicated low-to-moderate risk across most cohort studies, with low risk in RCTs2,15,23 and moderate limitations in observational studies8,18,24, especially due to attrition or unverified smoking status.
DISCUSSION
This systematic review and meta-analysis provides evidence that smoking cessation is associated with clinically meaningful improvements in mental health, especially in depressive symptom severity, psychological quality of life, and positive affect. The pooled meta-analytic effect size was derived from the 12 studies that provided sufficient quantitative data1,3-9,11,15,17,19. These results indicate a modest but clinically meaningful improvement in depressive symptoms, and this overall pattern was also consistent across the broader set of 22 included studies1-16. Improvements in mood and emotional well-being following cessation were documented across both randomized controlled trials and cohort studies, including those involving individuals with current or past depressive disorders data1,3-9,11,15,17,19.
Longitudinal evidence from studies with extended follow-up reinforces the durability of these effects. Cohorts with follow-up durations of 12 months, 18 years, or 30 years4-7,17 consistently showed reductions in depressive symptoms among abstinent individuals, with effect sizes strengthening over time. These long-term benefits indicate that the mental health advantages of cessation extend well beyond the acute withdrawal period. Importantly, several studies demonstrated that individuals with baseline depressive disorders experienced some of the largest improvements1,4,20, likely due to their higher initial symptom burden and greater potential for recovery.
In contrast, relapse was consistently associated with the re-emergence or worsening of depressive symptoms1,8 highlighting the bidirectional nature of the relationship between smoking and depression. Studies documenting rapid symptom rebound shortly after relapse underscore the need for targeted psychological support during early cessation phases, when affective instability and withdrawal symptoms are most likely to challenge abstinence.
Beyond depressive symptoms, improvements in broader psychological functioning were documented in eight studies9,11,15,17,19,20,22,23. Gains in quality of life, reductions in perceived stress, and increases in positive affect suggest that cessation contributes to a wider emotional recovery process. Positive affect, in particular, emerged as a key protective factor for sustained abstinence, supporting the inclusion of behavioral activation and emotion-enhancement strategies in cessation treatment plans.
Strengths and limitations
This review contains several methodological strengths. The inclusion of both randomized and observational designs enabled examination of causal relationships as well as real-world cessation trajectories. All included studies employed validated psychometric tools for measuring depressive symptoms, such as the PHQ-926, BDI-II27, CES-D13, or HAM-D12, increasing measurement reliability across heterogeneous designs. Additionally, publication bias assessment showed no evidence of systematic distortion.
This review has several limitations that should be acknowledged. First, moderate heterogeneity was observed across studies in terms of intervention types, follow-up durations, and definitions of depressive outcomes, which may affect the comparability of results. Second, some studies relied on self-reported smoking status without biochemical verification, introducing the risk of misclassification bias. Third, the majority of included studies were conducted in high-income Western countries, limiting the generalizability of findings to low- and middle-income settings. Fourth, potential moderators such as gender, socioeconomic status, and comorbid psychiatric conditions were underexplored in the available literature. Finally, although publication bias was not statistically significant, the possibility of missing relevant studies, particularly from grey literature, cannot be excluded.
Implications
From a clinical and public health perspective, the implications are significant. Rather than avoiding cessation counseling for individuals with depression, clinicians should proactively encourage cessation, as evidence indicates that it yields psychological benefits and does not worsen depressive symptoms. Effective strategies include combining behavioral therapies with pharmacological treatments, as demonstrated in studies2,3,15,23 where integrated approaches enhanced both quit rates and emotional outcomes. Clinicians should caution patients about possible short-term mood fluctuations during early cessation but emphasize that these symptoms are transient and reversible, particularly with appropriate support systems in place.
Future research
Future research should prioritize stratified analyses by baseline depression severity to clarify which patient subgroups benefit the most. Long-term follow-up studies beyond 12–24 months are essential to fully characterize the trajectory of emotional recovery and relapse risk. Incorporating biochemical verification of abstinence would strengthen validity, and expanding research to marginalized or underrepresented populations is necessary to improve global applicability. Finally, evaluating emerging digital interventions – such as mobile applications or online CBT modules tailored for individuals with depressive disorders – may provide scalable and effective pathways for supporting both cessation and mental health recovery.
CONCLUSIONS
This systematic review and meta-analysis indicates that smoking cessation is safe and beneficial for individuals with depressive disorders, with consistent evidence of modest but clinically meaningful improvements in depressive symptoms, quality of life, and psychological well-being. These findings, maintained over time, counter the traditional concern that quitting might worsen mood and instead suggest that tobacco cessation can be integrated as part of routine mental healthcare. Although some individuals may experience short-term increases in symptoms following cessation, providing structured and tailored support can help to mitigate these effects. Overall, the evidence emphasizes the relevance of systematically addressing tobacco use in psychiatric populations, while highlighting the need for further studies to refine intervention approaches, identify vulnerable subgroups, and examine long-term outcomes across diverse settings. Healthcare providers should consider routinely incorporating smoking cessation into mental health services, accompanied by tailored support strategies. Researchers are encouraged to design high-quality trials that explore moderators such as gender, socioeconomic status, and comorbidities, as well as to investigate innovative interventions adapted for low- and middle-income countries.
