Batya Swift Yasgur, MA, LSW
November 11, 2019
Traditionally, it's been widely assumed that patients with mental illness start smoking to self-medicate, but a novel study suggests the reverse may be true — that cigarette smoking actually raises the risk of mental illness, including depression and psychosis.
Investigators at the University of Bristol, United Kingdom, found that lifetime smoking and smoking initiation were both associated risk factors for schizophrenia as well as depression — an association that could, at least in part, be considered causal.
"Smoking is much more common among individuals with mental illness," study investigator Robyn Wootton, PhD, told Medscape Medical News.
"Our evidence suggests that this higher prevalence is due to bidirectional effects such that smoking increases risk of developing depression/schizophrenia and also that having depression/schizophrenia increases smoking behavior," she said.
The study was published online November 6 in Psychological Medicine.
Although it's well-known that smoking is more common in patients with mental illness vs those in the general population, whether there is a cause and effect relationship is unclear.
"Does smoking increase risk for mental illness or does having mental illness increase smoking behavior — or both, or neither?" Wootton asked.
"It has been hard to previously look at causal effects because individuals who smoke might be different to nonsmokers in a variety of ways, such as drinking more alcohol or eating less healthy [food], so it can be hard to control for these differences between groups," she added.
The investigators used the Mendelian randomization model to analyze data from the UK Biobank on over 460,000 people of European ancestry to determine the impact of lifetime smoking on risk for depression and schizophrenia.
Mendelian randomization can get around this issue by "using genetic variants that predispose some individuals to smoke more and others to smoke less, thereby overcoming bias from residual confounding and reverse causation," said Wootton.
"These genetic variants are specific to smoking — not alcohol or diet, etc — and therefore this is a kind of natural experiment where we can look at the effect of smoking independent of other behaviors," she noted.
She added that her group is the "first to apply new genetic variants for smoking to understand the link between smoking and mental health using this method."
To conduct the analysis, the researchers drew on data regarding smoking initiation from the genome-wide association study (GWAS) of the Sequencing Consortium of Alcohol and Nicotine use (GSCAN), identifying 378 conditionally independent genome-wide significant single nucleotide polymorphisms (SNPs) from 1,232,091 individuals of European Ancestry.
These SNPs explain 2% of the variance in smoking initiation, the authors note.
In addition, they generated their own lifetime smoking measure in individuals drawn from the United Kingdom Biobank — a national health research resource of individuals, also of European ancestry — who had phenotype data available and passed genotype inclusion criteria.
Smoking measures, including smoking status, age at initiation, age at cessation, and number of cigarettes smoked per day, were combined into a lifetime smoking index, which determined smoking duration, heaviness, and cessation.
Characteristics of the UK Biobank sample (N = 462,690 individuals) were:
- 54% female
- Mean age 56.7 (SD 8.0) years
- 30% never smokers
- 22% former smokers
- 8% current smokers
Characteristics of those specifically included in the Biobank GWAS of lifetime smoking index were:
- 54% never smokers
- 36% former smokers
- 11% current smokers
- Mean lifetime smoking score 0.359 (SD 0.694)
Inverse variance weighted (IVW) methodology was used to generate the odds ratios (ORs) in the main analysis as well as the sensitivity analyses.
Another Reason to Quit
Based on bidirectional MR analyses, higher lifetime smoking was associated with an increased risk of both schizophrenia and depression (odds ratio [OR] 2.27; 95% confidence interval [CI], 1.67 - 3.08, and OR, 1.99; 95% CI, 1.71 - 2.32, both Ps < .001).
There was a "consistent direction of effect across all 5 MR methods," the authors report.
The same findings were obtained when the researchers analyzed smoking initiation for schizophrenia and depression (OR, 1.53; 95% CI, 1.35 - 1.74; and OR, 1.54; 95% CI, 1.44 - 1.64; both Ps < .001).
Conversely, there was consistent evidence of reverse directionality — ie, that higher genetic liability for both schizophrenia and depression increased lifetime smoking, but the effect was smaller (β = .022; 95% CI, 0.005 - 0.038; P = .009 and β = 0.091; 95% CI, 0.027 - 0.155; P = .005, respectively).
Although these effects remained for depression on smoking initiation (β = 0.083; 95% CI, 0.039 - 0.127; P < .001), they became "even weaker and inconsistent" across the different methods for schizophrenia on smoking initiation (β = 0.010; 95% CI, 0.000 - 0.021; P = .04).
The researchers conducted a sensitivity analysis removing variants from the CHRNA5-A3-B4 gene complex, which were identified in both GWAS of lifetime smoking and of schizophrenia, and the effects remained consistent.
"There was evidence to suggest smoking is a risk factor for developing depression and schizophrenia and that having depression and schizophrenia increases smoking behavior; however, there was no clear evidence that having schizophrenia caused individuals to start smoking, only that they smoked more heavily," said Wootton.
"For individuals already suffering with a mental illness, it is a commonly held belief thought that smoking is a form of 'self-medication' and therefore individuals with mental health problems are not helped to quit as much as they should be," she added.
"We already know that smoking results in a much higher proportion of smoking-related deaths and illnesses for these individuals, so we should be trying to help individuals to quit anyway, but this adds further weight to the argument if smoking is exacerbating mental illness," she said.
Commenting on the study for Medscape Medical News, Simon Gilbody, MBChB, PhD, a professor of psychological medicine at the University of York, UK, called it an "important study" that "represents a significant advance in our understanding of the relationship between smoking and mental ill health."
He pointed to the assumption that elevated smoking rates in people with mental illness are because of "self-medication, whereby smoking, or the use of nicotine, helps to ameliorate stress or the side effects of medication."
However, the "novel and powerful genetic method" used by the researchers is "robust" and "able to test if the opposite is true and answer the question: Does smoking actually increase the risk of developing schizophrenia and depression?"
Gilbody, who was not involved with the research, stated that the "clinical and public health implications are pretty clear."
Smoking is always "bad for human health, and 1 in 2 people who take up smoking will die of a smoking-related illness, [so] now we can add major psychiatric illness to the long list of physical smoking-related illnesses," he said.
"Children and adolescents should be told this to discourage them from taking up smoking," he stressed, "and for adult smokers, the potential psychological and physical benefits of quitting can be emphasized."
The authors express similar views.
"Strong effects of smoking as a risk factor emphasizes the detrimental public health consequences of smoking, especially for mental health, and the need to reduce smoking prevalence not only to reduce the burden of physical illness, but also to reduce the burden of mental illness," they conclude.
Wootton and coauthors are members of the MRC Integrative Epidemiology Unit at the University of Bristol funded by the MRC. The study was supported by the NIHR Biomedical Research Centre at the University Hospitals Bristol NHS Foundation Trust and the University of Bristol. Wootton, study coauthors, and Gilbody have disclosed no relevant financial relationships.