If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Neuroepidemiology Unit, Centre for Epidemiology and Biostatistics, Melbourne School of Population & Global Health, The University of Melbourne, Level 3, 207 Bouverie St, Carlton, Melbourne, VIC 3053, Australia
Neuroepidemiology Unit, Centre for Epidemiology and Biostatistics, Melbourne School of Population & Global Health, The University of Melbourne, Level 3, 207 Bouverie St, Carlton, Melbourne, VIC 3053, AustraliaMenzies Institute for Medical Research, University of Tasmania, Hobart, Australia
Neuroepidemiology Unit, Centre for Epidemiology and Biostatistics, Melbourne School of Population & Global Health, The University of Melbourne, Level 3, 207 Bouverie St, Carlton, Melbourne, VIC 3053, Australia
Neuroepidemiology Unit, Centre for Epidemiology and Biostatistics, Melbourne School of Population & Global Health, The University of Melbourne, Level 3, 207 Bouverie St, Carlton, Melbourne, VIC 3053, Australia
Neuroepidemiology Unit, Centre for Epidemiology and Biostatistics, Melbourne School of Population & Global Health, The University of Melbourne, Level 3, 207 Bouverie St, Carlton, Melbourne, VIC 3053, Australia
Systematic review of studies examining associations between smoking and depression and anxiety in people living with MS found;
•
Strong evidence of a smoking-depression relationship, with 1.3–2.3-fold higher depression prevalence in current-smokers.
•
Depression levels were significantly higher in people who quit smoking than in non-smokers.
•
Strong evidence of a smoking-anxiety relationship, with 20% higher anxiety prevalence in current-smokers than non-smokers.
•
Anxiety levels were similar between people who had quit smoking and non-smokers.
Abstract
Background
The link between tobacco smoking and Multiple Sclerosis (MS) onset and progression is well-established. While clinical levels of depression and anxiety are highly prevalent in people living with MS (plwMS), and both are recognized as common MS-related symptoms, the relationships between smoking behavior and depression and anxiety are unclear. This systematic review aimed to synthesize evidence on the relationships between current-smoking and former-smoking and depression and anxiety in plwMS.
Methods
Systematic review of all studies investigating associations between tobacco smoking and depression and anxiety in plwMS was conducted. Relevant studies published before 26 April 2022 were identified by searching seven databases; MEDLINE® (Ovid and PubMed), Embase, CINAHL®, Cochrane Library and PsycInfo), and citation and reference list checking. Joanna Briggs Institute Critical Appraisal Checklists for respective study designs assessed the risk of bias.
Results
Thirteen publications reporting on 12 studies met study inclusion criteria. Nine of 12 studies examining current-smoking and depression in plwMS identified a positive association. Four prospective studies provided evidence supporting a causal smoking-depression relationship, with 1.3–2.3-fold higher depression prevalence found in current-smokers than non-smokers. Three cross-sectional studies found no smoking-depression association. Four of five included studies found current-smoking was associated with anxiety, with three prospective studies indicating anxiety prevalence was around 20% higher in current-smokers. Former-smoking was associated with increased prevalence of depression, but not anxiety.
Conclusion
We provide strong evidence for increased depression prevalence in plwMS who are either current-smokers or former-smokers. However, only current-smoking was associated with increased prevalence of anxiety.
), 25–54% and 36% are predicted to develop a clinical depressive and anxiety disorder in their lifetime, respectively, relative to 16 and 29% in the general population (
). Collectively, depression and anxiety symptoms represent one of the most common comorbidities seen in plwMS, and research in MS suggests this form of psychological comorbidity results in greater consequences compared with single and uncomplicated presentations of other symptoms (
Epidemiology, C.T.I.T., Impact of Comorbidity on Multiple, S. Health-related quality of life in multiple sclerosis: direct and indirect effects of comorbidity.
Smoking is associated with progressive disease course and increased progression in clinical disability in a prospective cohort of people with multiple sclerosis.
Longitudinal associations of modifiable lifestyle factors with positive depression-screen over 2.5-years in an international cohort of people living with multiple sclerosis.
). However, the clinical picture remains unclear, complicated by limited research examining the relationships between smoking and depression and anxiety and mixed findings (
). Therefore, understanding comorbid depression and anxiety in relation to smoking may help clarify these relations and inform existing psychological and pharmacological interventions.
While previous systematic review findings in the general population highlighted associations between smoking and depression and anxiety (
), to date no systematic review on smoking and depression and anxiety in plwMS has been performed. Given the adverse effects of depression, anxiety and smoking on health outcomes in plwMS (
), increased understanding of the association between smoking and these symptoms is highly relevant. In the present study, we conducted a systematic review to clarify relationships between smoking and depression and anxiety. Specifically, we aimed to examine whether current-smokers or former-smokers are at increased risk of depression and/or anxiety compared to non-smokers.
2. Methods
2.1 Study design
A systematic review of relevant quantitative studies following the Preferred Reporting Items for Systematic Reviews and Meta-Analyzes (PRISMA) guidelines was performed (
). The review was pre-registered in PROSPERO (ID: CRD42022329427).
2.2 Search strategy
Seven databases were searched for literature relevant to psychological conditions in plwMS from database conception to 26 April 2022: MEDLINE via OVID and PubMed, Web of Science, EMBASE, APA PsycInfo, Cochrane Central Register of Controlled Trials and Cumulative Index of Nursing and Allied Health Literature (CINAHL). The search strategy was a connexion of general terms for “multiple sclerosis”, “tobacco smoking”, AND combinations of “depression” OR “anxiety” across all fields (i.e., title, abstract, keywords) to capture relevant articles to the exposure (tobacco smoking) and outcomes (depression or anxiety) in plwMS. Exploded terms and Medical Subject Headings (MeSH) were used to maximize results. Search strategies are presented in Supplementary Table 1. Citation and reference checking of included studies was performed to identify articles not identified by electronic searching.
Two independent reviewers (V.V. and S.P.) identified full-text articles complying with study inclusion criteria:
•
study participants ≥18 years with a clinical diagnosis of MS
•
included a sample of current-smokers and/or former-smokers
•
included a quantitative self-report psychometric tool for the assessment of depression and/or anxiety symptoms, validated for the MS population
•
Exclusion criteria:
•
Studies not in English
•
Reviews, protocols, case studies, editorials or conference abstracts
2.3 Data extraction and analysis
Data [study type, country, smoking status (current-, former- and non-smoking), MS type, participant characteristics, outcome measurement tools and results] was extracted from eligible studies by three independent reviewers (V.V., M.I.P. and S.P.) using a standardized data extraction form (
). Checklists contain 8–12 questions with answer options “yes”, “no”, “unclear”, or “not applicable”. The level of bias was determined by the proportion of “yes” answers; >80% positive responses indicate low risk of bias, 60–80% indicate moderate risk and <60% indicate high risk (
). Data from each study was extracted and individual study findings were evaluated and summarized prior to combining results for the narrative synthesis. A meta-analysis could not be performed due to heterogeneity in study design and assessment tools used to measure depression and anxiety across studies. That is, no two included studies had the same study design and used the same assessment tool to measure depression or anxiety.
3. Results
In all, 1666 studies were identified via electronic searching and two studies were identified from citation and reference checking (Fig. 1). The title and abstract of 1361 studies were screened for eligibility, resulting in full-text examination of 22 studies. Of these, nine studies were unsuitable for study inclusion due to: evaluating a different outcome (n = 2) or research question (n = 1); unsuitable (qualitative) outcome measurement tools (n = 5); or were a conference abstract (n = 1). In total, 13 studies reporting on 12 individual studies were eligible for study inclusion, with two studies analysing data from the same study both cross-sectionally (
Longitudinal associations of modifiable lifestyle factors with positive depression-screen over 2.5-years in an international cohort of people living with multiple sclerosis.
Participant characteristics of the 13 included studies are presented according to study design and chronologically in Table 1. Eight studies were cross-sectional (
Aging with, M.S.C. Under-treated depression negatively impacts lifestyle behaviors, participation and health-related quality of life among older people with multiple sclerosis.
Longitudinal associations of modifiable lifestyle factors with positive depression-screen over 2.5-years in an international cohort of people living with multiple sclerosis.
Aging with, M.S.C. Under-treated depression negatively impacts lifestyle behaviors, participation and health-related quality of life among older people with multiple sclerosis.
Population-based; participants of Canadian Survey of Health, Lifestyle and aging with Multiple Sclerosis (study inclusion criteria included >55 years with MS symptoms > 20 years)
Current-smoker vs non-smoking
Depression
HADS-depression HADS-anxiety (a score >11 suggested abnormal symptoms of depression and a report of a confirmed diagnosis was provided)
742 total participants (412 with no depression symptoms; 103 with low depression symptoms; 87 with high depression symptoms; 140 with high symptoms (undiagnosed)
Longitudinal associations of modifiable lifestyle factors with positive depression-screen over 2.5-years in an international cohort of people living with multiple sclerosis.
Current-smoker vs never-smoker Former-smoker vs never-smoker
Depression
PHQ-2 was used to measure changes in depression from baseline to 2.5-year review (cut-off scores > 2 indicated a positive depression screen) PHQ-9 was used to assess for symptoms of depression at 2.5 years. (cut-off scores > 9 indicated positive depression screen)
1309 total participants (701 never-smokers; 527 former-smokers; 102 current-smokers)
Tertiary MS referral center at the Cleveland Clinic, Ohio, US
Current-smoker, former-smoker and never-smokers
Depression
PHQ-9 (no cut-off provided)
1271 total participants (328 current-smokers; 518 former-smokers; 848 never-smokers)
43.0 (10.4)
74.4
RRMS – 100
Multiple sclerosis (MS); Hospital Anxiety and Depression Scale (HADS); Patient Health Questionnaire (PHQ); BECK depression inventory (BDI); Not available (N/A)
Aging with, M.S.C. Under-treated depression negatively impacts lifestyle behaviors, participation and health-related quality of life among older people with multiple sclerosis.
Longitudinal associations of modifiable lifestyle factors with positive depression-screen over 2.5-years in an international cohort of people living with multiple sclerosis.
), where 50% of participants were female compared with 62.3–82.4% in other studies. Mean age ranged from 35 to 65 years. RRMS was the most common MS type (52–95% of cases).
3.1 Risk of bias
Ten studies were judged to have low risk of bias (
Aging with, M.S.C. Under-treated depression negatively impacts lifestyle behaviors, participation and health-related quality of life among older people with multiple sclerosis.
Longitudinal associations of modifiable lifestyle factors with positive depression-screen over 2.5-years in an international cohort of people living with multiple sclerosis.
BDI scores between heavy-smokers and non-smokers (mean ± SE) were not significantly different; 17.05 ± 11.57 for heavy-smokers and 16.04 ± 13.70 for non-smokers; p = 0.516) Backward stepwise logistic regression was used to identify factors statistically predictive of cognitive impairment.
Current-smokers are 2.3 times more likely to screen positive for depression. Adjusted odds ratio (aOR) for examining smoking and the risk of depression: aOR = 2.29 [95% CI 1.60, 3.27], p < 0.001 Former-smokers were at higher odds of screening positive for depression than non-smokers: aOR = 1.54 [95%CI 1.6, 3.27]) p < 0.01 (Adjusted odds ratios of screening positive for depression were adjusted for years since diagnosis, number of comorbidities, level of disability, clinically significant fatigue, age, gender, marital status and level of education.)
Current-smoking vs non-smoking with associated with anxiety in plwMS: aOR = 2.43 [95% CI, 1.18, 5.00] p = 0.015 (Statistical model adjusted for age and sex) After adjusting for further confounders (age, sex, education, smoking, relapse-remitting MS and depression), the association between smoking and anxiety was no longer significant (aOR=2.01; 95%CI 0.90–4.50; p = 0.087)
Non-smoking status was associated with lower depression severity Prevalence ratio of depression in non-smokers = 0.75 (95%CI: 0.59, 0.94) p < 0.05; Symptom severity adjusted -1.12 (95%CI: -1.76, -0.48) p < 0.05 (All models were adjusted for age, sex, Patient‐Determined Disease Steps score, depression treatment, fatigue level, education level and socio‐economic status)
Non-smoking status was associated with lower anxiety severity Prevalence ratio of depression in non-smokers = 0.85 (95%CI: 0.68, 1.00) p ≥ 0.05; Symptom severity adjusted -1.33 (95%CI: -2.10, -0.55) p < 0.05 (All models were adjusted for age, sex, Patient‐Determined Disease Steps score, anxiety treatment, fatigue level, education level and socio‐economic status)
Aging with, M.S.C. Under-treated depression negatively impacts lifestyle behaviors, participation and health-related quality of life among older people with multiple sclerosis.
There was no difference in the % of non-smokers vs non-smokers in plwMS that had symptoms of depression compared with those that had no symptoms of depression (χ2 (3) = 4.54, p = 0.21) (Hierarchical logistic regression was used to examine smoking-depression associations controlling for age, years diagnosed with MS, MS type at diagnosis and level of disability, as measured using the Barthel Index in the first step in the model).
Current-smoking was not associated with the rate of depression (24.9% of people with depression were smokers compared with 71.3% non-smokers; p = 0.113)
BDI scores were significantly higher in current-smokers than non-smokers: BDI score in smokers was 21.32±2.12 compared with the BDI score in non-smokers (BDI=15.95±1.53); p = 0.04 (Adjusted analyzes were not performed as age, gender, marital status and duration of illness, neurological fatigue index-MS, nicotine addiction, and EDSS scores were not different between groups. However, BDI scores were lower in both groups in those with higher education)
BDI scores were significantly higher in smokers than non-smokers (pη2 = 0.015, p = 0.044) (One-way analysis of covariance was used to measure the difference between current-smoker and non-smoker groups, adjusted for age, gender, EDSS, and disease duration.)
Prospective analysis of smoking impact over 4 years in 923 smokers: Current-smokers had worsening HADS-depression scores [0.71 (0.47, 0.96)] over 4 years using linear mixed models. Average HADS-depression scores increased steadily with each year from baseline. Prospective time to event analyzes of 4511 smokers (streak length ranged from 180 days to 8 years): Current-smoking was associated with a quicker time to a worsening in HADS-depression scores (Adjusted Hazard Ratio (aHR)=1.25 (95%CI 1, 1.57; p < 0.0001) Retrospective analyzes At baseline, current-smokers had higher depression [HADS-depression scores = 8 (5–11)] compared with never-smokers [6 (3–9)] Former-smokers had no clinically significant improvement in HADS-depression scores since quitting smoking (−0.08, 95%CI [−0.11, −0.04], n = 2779) For every additional pack year of smoking, there was a significant increase in HADS-depression scores of 0.03. Current-smokers had higher mean increases in HADS-depression scores than never-smokers [HADS increase =0.74 (0.28, 1.20) β (95% CI) p = 0.0015]
(Multivariable analyzes adjusted for age at baseline, time since onset, MS type at diagnosis, ethnicity and DMT usage)
Prospective analysis of smoking impact over 4 years in 923 smokers: Current-smokers had stable HADS-anxiety scores over 4 years using linear mixed models. Current-smokers had higher HADS-anxiety scores compared to never smokers [1.14 (0.52, 1.76)] Prospective time to event analyzes of 4511 smokers (streak length ranged from 180 days to 8 years): Current-smoking was associated with a quicker time to a worsening in HADS-anxiety scores (Adjusted hazard ratio (aHR)=1.25 (95%CI 1, 1.56; p = 0.0017) Retrospective analyzes At baseline, current-smokers had higher anxiety [HADS-anxiety scores = 9 (6–12)] compared with never-smokers [7 (4–10] Former-smokers had improvement in HADS-anxiety scores since quitting smoking (−0.19, 95%CI [−0.22, −0.15], n = 2779) For every additional pack year of smoking, there was a significant increase in HADS-anxiety scores of 0.03 Current-smokers had higher mean increases in HADS-anxiety scores than never smokers [HADS increase =0.79 (0.31, 1.26) β (95% CI) p = 0.0012]) (Multivariable analysis adjusted for age at baseline, time since onset, MS type at diagnosis, ethnicity and DMT usage)
Longitudinal model predicted high depression levels in smokers, with lower BDI score changes found in non-smokers vs smokers [BDI score -2.31 (SE, 0.47)] in adjusted analyzes, p < 0.0001 (For final longitudinal model, a backwards elimination approach was used to find the most suitable predictors. Confounders: Anxiety, fatigue, exercise, recreational drugs, planful coping gender)
Longitudinal model predicted high anxiety levels in smokers, with increased BDI score changes of 2.50 (SE, 0.84) found in smokers vs non-smokers in adjusted analyzes, p = 0.003 (For final longitudinal model, a backwards elimination approach was used to find the most suitable predictors. Confounders: depression, immunotherapy, relaxation, optimism, age, gender, caffeine)
In prevalence analyzes, current-smoking compared with non-smoking was associated with an increased odds of depression at all time points [baseline, 1-year and 2-years] (aOR = 1.37 [95% CI 1.04, 1.78], p unreported) In a prospective follow-up of incident depression cases at the 1- and 2-year follow-up from baseline, current-smoking was NOT associated with an altered risk of depression (aOR = 0.94 [95% CI 0.54, 1.64], p unreported) (multivariable analyzes were adjusted for: age, sex, EDSS, and alcohol dependence in multivariable analyzes)
In prevalence analyzes, current-smoking compared with non-smoking was associated with an increased odds of anxiety (Adjusted Odds ratio = 1.29 [95%CI 1.02, 1.63], p unreported) In a prospective follow-up of incident anxiety cases at the 1- and 2-year follow-up from baseline, current-smoking was NOT associated with an altered risk of anxiety (aOR = 1.20 [95% CI 0.71, 2.04], p unreported) (multivariable analyzes were adjusted for: age, sex, EDSS, and alcohol dependence in multivariable analyzes)
Longitudinal associations of modifiable lifestyle factors with positive depression-screen over 2.5-years in an international cohort of people living with multiple sclerosis.
Longitudinal associations of modifiable lifestyle factors with positive depression-screen over 2.5-years in an international cohort of people living with multiple sclerosis.
replicated their previous cross-sectional results at the 2.5-year follow-up, and found higher depression prevalence in current-smokers than never-smokers (adjusted Prevalence Ratio (aPR)=1.96, 95%CI, 1.51–5.55; p < 0.001). At the 2.5-year follow-up, current-smoking was a predictor of depression risk vs no depression using PHQ-9 (i.e. was a predictor of depression; Minimal depressive symptoms vs no symptoms: Adjusted prevalence ratio (aPR) = 1.34; 95%CI 1.02, 1.75; p = 0.071 for current-smokers compared with never-smokers Major depressive symptoms vs no symptoms: Adjusted prevalence ratio (aPR) = 1.99; 95%CI 1.45, 2.74; p < 0.001 for current-smokers compared with never-smokers Current-smoking was NOT associated with a change in the risk of depression from baseline to 2.5 years. (Multivariable analyzes were adjusted for P-MSSS, age, fatigue, and anti-depressant medication use)
All included/eligible studies used a quantitative self-reported psychometric tool to assess depression and/or anxiety symptoms, validated for the MS population. These included the 14-item Hospital Anxiety and Depression Scale (HADS) with seven questions each for depression and anxiety (
Aging with, M.S.C. Under-treated depression negatively impacts lifestyle behaviors, participation and health-related quality of life among older people with multiple sclerosis.
Longitudinal associations of modifiable lifestyle factors with positive depression-screen over 2.5-years in an international cohort of people living with multiple sclerosis.
). Of note, no studies eligible for inclusion used other psychometric tools to assess for depression or/anxiety such as subscores of the Neuro-QoL (Quality of life in neurological disorders) (
) (Table 2). In a large international study, Taylor et al. found depression prevalence was 2.3 times higher in current-smokers than non-smokers (OR=2.29; p < 0.001) (
In three cross-sectional studies where current-smoking was not associated with depression, a small Turkish study found depressive BDI scores were not significantly different between heavy-smokers (n = 20; 17.05 ± 11.57) and non-smokers (n = 24; 16.04±13.70; p = 0.516) (
). However, this study had a higher proportion of males to females (1:1) than other studies (around 1:4), with lower depression prevalence reported in males (
). Ozcan et al. also examined ‘heavy-smokers’ (≥10 pack-years), limiting the comparability of this study to other studies not considering pack-years. Two further studies from Saudi Arabia and Canada found no difference in the proportion of current-smokers vs non-smokers with depression (
Aging with, M.S.C. Under-treated depression negatively impacts lifestyle behaviors, participation and health-related quality of life among older people with multiple sclerosis.
). In the large UK MS Register (UKMSR) study, HADS-depression scores in current-smokers increased over a 4-year period (HADS = 0.71; 0.47–0.96) and at a quicker rate than never-smokers (p < 0.0001) (
). Similarly, an international study found current-smoking vs never-smoking was a predictor of major depression at the 2.5-year follow-up (aPR=1.99; p < 0.001) (
Longitudinal associations of modifiable lifestyle factors with positive depression-screen over 2.5-years in an international cohort of people living with multiple sclerosis.
), consistent with an earlier Australian study that found current-smoking vs non-smoking predicted higher BDI depression scores at the 2-year follow-up (p < 0.0001) (
also found the prevalence of depression was higher in current-smokers vs non-smokers at baseline, but depression scores did not increase from baseline to the 2-year follow-up. Overall, depression risk was found to be 1.3–2.3 higher in current-smokers vs non-smokers.
In studies examining associations between former-smokers and depression, retrospective analyzes by
found HADS-depression scores in former-smokers at baseline did not significantly improve after smoking cessation. Similarly, a retrospective analysis by
prospective study, the odds of depression were 1.5 higher in former-smokers than never-smokers.
3.4 Smoking and anxiety
3.4.1 Cross-sectional studies
Two cross-sectional studies examining current-smoking and anxiety had conflicting findings (Table 2). Gascoyne et al. found smokers had a higher prevalence of anxiety symptom severity than non-smokers after adjusting for disability, anti-anxiety treatment, fatigue, education and socio-economic status (
found current-smoking was not associated with anxiety, after adjusting for age, sex, education, relapse-remitting MS and depression (aOR=2.01; p = 0.087).
3.4.2 Cohort studies
Three prospective studies found associations between current-smokers and anxiety (
found anxiety prevalence was higher in current-smokers than never-smokers at baseline (OR=1.29, 95%CI 1.02–1.63), and also remained consistent over 2 years.
In retrospective analyzes, Rodgers et al. found current-smokers had higher HADS-anxiety scores at baseline than non-smokers, and smoking cessation improved scores [-0.19, 95%CI, (-0.22)- (-0.15)].
4. Discussion
To the authors’ knowledge, the present study represents the first systematic review examining the association between tobacco smoking and depression and anxiety in plwMS. This review identified convincing evidence of positive associations for both outcomes. Among included studies, nine identified positive associations between current-smoking and depression with 1.3–2.3-fold higher depression prevalence compared with non-smokers. Four studies reported positive associations between current-smoking and anxiety, with around 1.2-fold higher anxiety prevalence compared with non-smoking. Former-smoking was associated with increased depression risk, but not increased anxiety risk. A meta-analysis could not be performed due to heterogeneity across studies.
5. Smoking and depression
All four prospective studies included in the review found positive associations between current-smokers and depression (
Longitudinal associations of modifiable lifestyle factors with positive depression-screen over 2.5-years in an international cohort of people living with multiple sclerosis.
). These pathways may partially explain the role of smoking in increased MS activity via chronic lung inflammation and increased pro-inflammatory cytokines and/or neurotoxic components facilitating neurological damage (
). Further, links between smoking and depression could be partly attributable to the immunomodulatory effects of smoking, with immune system alterations theorized to increase depression risks in plwMS (
found depression in current-smokers worsened over a 4-year period, and at a faster rate than never-smokers analysing UKMSR data after accounting for several confounders including DMT use and time since diagnosis. In particular, DMT use may be relevant to the smoking-depression relationship as smoking can negate immune-based therapies such as interferon-beta (
Longitudinal associations of modifiable lifestyle factors with positive depression-screen over 2.5-years in an international cohort of people living with multiple sclerosis.
). However, these studies did not adjust for DMT usage.
Five cross-sectional studies found current-smoking was associated with depression; however, this study design precludes temporality inferences between smoking and depression. Further, the potential for reverse causality exists, consistent with previous research reporting a bi-directional relationship between smoking and depression in the general population (
A longitudinal study of the bi-directional relationship between tobacco smoking and psychological distress in a community sample of young Australian women.
Notably, three additional cross-sectional studies found no significant association between current-smoking and depression. However, null findings may be attributable to methodological limitations;
Aging with, M.S.C. Under-treated depression negatively impacts lifestyle behaviors, participation and health-related quality of life among older people with multiple sclerosis.
examined data from a select cohort (aged >55 years with MS symptoms >20 years) and used stricter criteria to measure depressive symptoms than other studies (HADS-depression cut-off >11, and a report of a depression diagnosis). In contrast, studies identifying positive smoking-depression associations, including other cross-sectional studies (
used HADS-depression cut-off scores ≥8 inferring potential bias for identifying a significant smoking-depression relationship. Notably, the HADS-depression cut-off >8 also has a sensitivity and specificity of 82% and 68%, respectively, (
). Of note, the cross-sectional study not identifying a significant smoking-anxiety association did not control for several predictive factors, including anti-anxiety medication use (
7. Limitations of studies identifying positive associations
While several studies provided evidence of an association between smoking and depression and anxiety, these studies had limitations that may affect the validity of these results. Many studies were susceptible to selection bias and may represent cohorts of plwMS more susceptible to depression and/or anxiety. In particular, the UKMSR population-based cohort was over-represented by older males with progressive MS and lower smoking rates than the total population (
Longitudinal associations of modifiable lifestyle factors with positive depression-screen over 2.5-years in an international cohort of people living with multiple sclerosis.
Prevalence of depression and anxiety in the different clinical forms of multiple sclerosis and associations with disability: a systematic review and meta-analysis.
), higher proportions of people with progressive MS in studies could over-represent plwMS with depression. Studies were also limited by the psychometric tool employed; the 2-item PHQ-2 tool for assessing depression (
Longitudinal associations of modifiable lifestyle factors with positive depression-screen over 2.5-years in an international cohort of people living with multiple sclerosis.
), has less precision than the 9-item PHQ-9, and the STAI instrument has been found to identify a higher prevalence of anxiety than the HADS-anxiety in certain studies (
Former-smokers had clinically significant improvements in HADS-anxiety scores after quitting smoking, whereas HADS-depression scores did not improve after smoking cessation (
found while former-smokers had lower depressive scores than current-smokers, depressive scores were higher than non-smokers. Collectively, these studies suggest biopsychosocial mechanisms underlying the smoking-depression and smoking-anxiety relationships may be different. Subsequently, detrimental effects of smoking on depression could persist for some time following smoking cessation. Further, depression is often referred to as a symptom of MS (
), which may help explain the persistence of depression in former-smokers and high depression rates in non-smokers. This finding could also suggest the presence of more severe comorbid, or possibly pre-morbid, depressive disorders among plwMS, which is likely maintained by a complex interplay of factors, warranting more intensive, longer-term intervention (
The present study represents the first systematic review to evaluate the relationships between smoking and depression and anxiety. Participant characteristics of all included studies were comparable to other MS studies (
), maximizing potential generalizability of findings to the wider MS community.
Potential study limitations include differences in smoking behaviors and types of tobacco across countries and smoking habits across genders in included studies (
Longitudinal associations of modifiable lifestyle factors with positive depression-screen over 2.5-years in an international cohort of people living with multiple sclerosis.
), were confined to single countries with high human development index (HDI) levels, therefore findings may not be relevant to countries with lower HDI levels.
Confounders adjusted for in multivariable analyzes across studies were not consistent, and future studies should consider as many potential confounders as possible. For instance,
Further, the associations identified between smoking and depression and anxiety may be partially confounded by the extraneous effects of social, educational and financial variables, as well as healthcare access. For instance, smoking has been found to vary according to socio-economic status (SES) (
). However, as this systematic review did not include a meta-analytic synthesis, these factors could not be controlled for and future observational and meta-analytic research should aim to additionally control for social variables (i.e., education, SES) to identify any direct effects of smoking on depression and anxiety.
Included studies were limited by potential recall bias as data across all studies was collected by self-report, with only
validating smoking status using healthcare records. Unstandardized self-reported smoking measures were used, with self-reported smoking status previously found to be an unreliable measure of smoking behavior (
). Additional smoking factors to consider include nicotine levels, tobacco quality, brand, or inhalation method. Similarly, pack-years [packet number (20 cigarettes) smoked per day multiplied by the number of years] was only considered in two studies (
), but increasing pack-years could affect depression and anxiety severity, similar to increased exacerbation of MS symptoms associated with increasing pack-years (
Consequently, future studies implementing uniform measurements for smoking and outcomes (depression and anxiety) with consistent cut-offs and accounting for as many confounders in multivariable analyzes as possible are advised to allow comparisons across studies and a future meta-analysis to be performed.
9.1 Clinical implications
This review contributes to a growing body of literature linking smoking with adverse health outcomes in plwMS (
). Prospective study findings suggest current-smoking predicts increases in depressive and anxiety symptoms. Former-smoking predicted increases in depressive symptoms, but not anxiety, suggesting former-smoking may have specific potency for depression. These findings suggests that current-smokers should be monitored for depression and anxiety, with monitoring of depression in former-smokers also advised.
As seen in other populations, bidirectional relationships between smoking and affective symptoms are plausible (
), whereby depression and anxiety increase the severity of nicotine cravings and withdrawals, and smoking acutely alleviates negative affective states, thereby reinforcing positive meta-beliefs about the value of smoking. The importance of increasing awareness and support for smoking cessation has been emphasised (
), and may have important therapeutic effects on singular or comorbid symptoms of depression or anxiety. Contrariwise, targeting depression and anxiety may help promote smoking cessation, thereby reducing reliance on smoking as a means of regulative negative affective states. Cognitive-behavioral therapies which integrate psychoeducation on the effects of smoking, motivation strategies for reducing smoking ambivalence and restructuring techniques for altering positive meta-beliefs about smoking have been effective in reducing smoking in other populations (
Interventions for tobacco cessation in adults, including pregnant persons: updated evidence report and systematic review for the US preventive services task force.
), and may have benefit for plwMS. Future interventional research is required to determine the efficacy of psychological therapies for smoking cessation, with or without embedded smoking cessation programs. Moreover, it is uncertain whether the treatment of depression and/or anxiety should be prioritized over smoking cessation therapies and beyond the scope of the review of examining positive associations between smoking and depression and anxiety.
Smoking is thought is to have an anxiogenic effect, increasing panic symptoms (
), and so anxiety symptoms may not be as severe when an individual is no longer smoking regularly. To better understand the relationship between current-smoking / former-smoking and depressive symptoms, future research may consider the differential effects of depressive subtypes (i.e., persistent depression vs. acute, time-limited episodes). Some existing research suggests that individuals with dysthymia may be prone to heavier smoking in other populations (
Systematic review of relevant studies found a strong association between current-smoking and former-smoking and depression, and current-smoking and anxiety in plwMS. Findings may add to the evidence for smoking cessation programs in the treatment of depression and anxiety, and the ongoing use of cognitive-behavioral approaches in promoting smoking cessation. Future prospective studies with uniform measurements of smoking and validated tools and consistent cut-offs for depression and anxiety to confirm suggested relationships between current-smoking and former-smoking and depression and anxiety are advised, which in turn would enable a future meta-analysis to be performed.
Author contributions
JCR and SSY were involved in study design and conceptualization. VV and SP identified the articles and screened the titles and abstracts; VV and SP screened the full-text articles; VV, SP, JCR and MIP extracted the data; VV, SSY, RD, MIP, SLN and JCR analyzed and interpreted the data; VV, JCR, RD and SSY wrote the manuscript. All authors reviewed and approved the manuscript.
Data and material availability
All data is available for this study.
Funding
We gratefully acknowledge the financial support of Mr Wal Pisciotta and other anonymous philanthropic donors to the Neuroepidemiology Unit. Funders played no role in the design, review or approval of the study.
Declaration of Competing Interest
SN is a co-editor of Overcoming Multiple Sclerosis Handbook: Roadmap to Good Health. SN is a facilitator of Overcoming MS educational workshops for people with MS.
Acknowledgments
We thank University of Melbourne library staff for their valuable assistance in the systematic review literature search.
A longitudinal study of the bi-directional relationship between tobacco smoking and psychological distress in a community sample of young Australian women.
Interventions for tobacco cessation in adults, including pregnant persons: updated evidence report and systematic review for the US preventive services task force.
Prevalence of depression and anxiety in the different clinical forms of multiple sclerosis and associations with disability: a systematic review and meta-analysis.
Smoking is associated with progressive disease course and increased progression in clinical disability in a prospective cohort of people with multiple sclerosis.
Under-treated depression negatively impacts lifestyle behaviors, participation and health-related quality of life among older people with multiple sclerosis.
Longitudinal associations of modifiable lifestyle factors with positive depression-screen over 2.5-years in an international cohort of people living with multiple sclerosis.
00005-6&id=gr2.jpg){kind=link}
00005-6&id=gr1.jpg){kind=link}