research article on tobacco

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings

Preview improvements coming to the PMC website in October 2024. Learn More or Try it out now .

• Advanced Search
• Journal List
• Taylor & Francis Open Select

Tobacco smoking: Health impact, prevalence, correlates and interventions

Robert west.

a Department of Behavioural Science and Health , University College London , London, UK

Background and objectives : Despite reductions in prevalence in recent years, tobacco smoking remains one of the main preventable causes of ill-health and premature death worldwide. This paper reviews the extent and nature of harms caused by smoking, the benefits of stopping, patterns of smoking, psychological, pharmacological and social factors that contribute to uptake and maintenance of smoking, the effectiveness of population and individual level interventions aimed at combatting tobacco smoking, and the effectiveness of methods used to reduce the harm caused by continued use of tobacco or nicotine in some form.

Results and conclusions : Smoking behaviour is maintained primarily by the positive and negative reinforcing properties of nicotine delivered rapidly in a way that is affordable and palatable, with the negative health consequences mostly being sufficiently uncertain and distant in time not to create sufficient immediate concern to deter the behaviour. Raising immediate concerns about smoking by tax increases, social marketing and brief advice from health professionals can increase the rate at which smokers try to stop. Providing behavioural and pharmacological support can improve the rate at which those quit attempts succeed. Implementing national programmes containing these components are effective in reducing tobacco smoking prevalence and reducing smoking-related death and disease.

Introduction

The continued popularity of tobacco smoking appears to defy rational explanation. Smokers mostly acknowledge the harm they are doing to themselves and many report that they do not enjoy it – yet they continue to smoke (Fidler & West, 2011 ; Ussher, Brown, Rajamanoharan, & West, 2014 ). The reason is that nicotine from cigarettes generates strong urges to smoke that undermine and overwhelm concerns about the negative consequences of smoking, and the resolve not to smoke in those trying to stop (West & Shiffman, 2016 ). Progress is being made in many countries in reducing smoking prevalence but it remains one of the main causes of ill health and premature death worldwide (Gowing et al., 2015 ).

This paper provides a broad overview of smoking in terms of: the health effects, benefits of stopping, prevalence and patterns of use, psychological, pharmacological and social factors leading to uptake and maintenance of the behaviour, effectiveness of population level and individual level interventions to combat it, and methods used to reduce the harm despite continued use of tobacco or nicotine.

Definitions of smoking and smoking cessation

Tobacco smoking consists of drawing into the mouth, and usually the lungs, smoke from burning tobacco (West & Shiffman, 2016 ). The type of product smoked is most commonly cigarettes, but can also include cigarillos, cigars, pipes or water pipes. ‘Smokeless’ tobacco is also popular in some parts of the world. This typically involves using tobacco preparations for chewing, sniffing into the nose or placing as a wad in the mouth between the cheeks and gums (Critchley & Unal, 2003 ). Smokeless tobacco use has features that are similar to smoking and can carry significant health risks (Critchley & Unal, 2003 ); however, this article focuses on smoked tobacco only as this has been the subject of by far the largest volume of research and is the most harmful form of tobacco use.

Stopping smoking usually involves an intention not to smoke any more cigarettes from a given point in time (a ‘quit attempt’), followed by self-conscious resistance of urges to smoke resulting in a period of abstinence. If someone making a quit attempt smokes one or more cigarettes on an occasion but then resumes abstinence, this is usually termed a ‘lapse’. If this person resumes smoking on a regular basis s/he is said to have ‘relapsed’. ‘Short-term abstinence’ is commonly defined in terms of achieving up to 4 weeks of abstinence. ‘Long-term abstinence’ often refers to abstinence for at least 6 months but more typically involves abstinence for at least 12 months. There is no agreed criterion for deciding when someone has ‘stopped smoking’ so it is essential when using the term to be clear about how long the abstinence period has been.

Health impact of smoking and the benefits of stopping

Tobacco smoking increases the risk of contracting a wide range of diseases, many of which are fatal. Stopping smoking at any age is beneficial compared with continuing to smoke. For some diseases, the risk can be reversed while for others the risk is approximately frozen at the point when smoking stopped.

Health impact of smoking

Table ​ Table1 1 lists the main causes of death from smoking. Tobacco smoking is estimated to lead to the premature death of approximately 6 million people worldwide and 96,000 in the UK each year (Action on Smoking and Health, 2016b ; World Health Organization, 2013 ). A ‘premature death from smoking’ is defined as a death from a smoking-related disease in an individual who would otherwise have died later from another cause. On average, these premature deaths involve 10 years of life years lost (US Department of Health and Human Services, 2004 ). Many of these deaths occur in people who have stopped smoking but whose health has already been harmed by smoking. It also happens to be the case that smokers who do not stop smoking lose an average of 10 years of life expectancy compared with never-smokers and they start to suffer diseases of old age around 10 years earlier than non-smokers (Jha & Peto, 2014 ).

Most smoking-related deaths arise from cancers (mainly lung cancer), respiratory disease (mainly chronic obstructive pulmonary disease – COPD), and cardiovascular disease (mainly coronary heart disease) (Action on Smoking and Health, 2016b ). Smoking is an important risk factor for stroke, blindness, deafness, back pain, osteoporosis, and peripheral vascular disease (leading to amputation) (US Department of Health and Human Services, 2004 ). After the age of 40, smokers on average have higher levels of pain and disability than non-smokers (US Department of Health and Human Services, 2004 ).

Smoking in both women and men reduces fertility (Action on Smoking and Health, 2013 ). Smoking in pregnancy causes underdevelopment of the foetus and increases the risk of miscarriage, neonatal death, respiratory disease in the offspring, and is probably a cause of mental health problems in the offspring (Action on Smoking and Health, 2013 ).

People used to think that smoking was protective against Alzheimer’s disease but we now know that the opposite is the case: it is a major risk factor for both Alzheimer’s and vascular dementia (Ferri et al., 2011 ; US Department of Health and Human Services, 2004 ).

There is a positive association between average daily cigarette consumption and risk of smoking-related disease, but in the case of cardiovascular disease the association is non-linear, so that low levels of cigarette consumption carry a higher risk than would be expected from a simple linear relationship (US Department of Health and Human Services, 2004 ).

Tobacco smoke contains biologically significant concentrations of known carcinogens as well as many other toxic chemicals. Some of these, including a number of tobacco-specific nitrosamines (particularly NNK and NNN) are constituents of tobacco, largely as a result of the way it is processed, while others such as benzopyrine result from combustion of tobacco (Action on Smoking and Health, 2014b ). These chemicals form part of the particulate matter in smoke. Tobacco smoke also contains the gas, carbon monoxide (CO). CO is a potent toxin, displacing oxygen from haemoglobin molecules. However, acutely the amount of CO in tobacco smoke is too small to lead to hypoxia and the body produces increased numbers of red blood cells to compensate.

The nicotine in tobacco smoke may cause a small part of the increase in cardiovascular disease but none or almost none of the increase in risk of respiratory disease or cancer (Benowitz, 1997 , 1998 ). It is the other components of cigarette smoke that do almost all the damage. It has been proposed on the basis of studies with other species that nicotine damages the adolescent brain but there is no evidence for clinically significant deficits in cognition or emotion in adults who smoked during adolescence and then stopped (US Department of Health and Human Services, 2004 ).

Exposure to second-hand smoke carries a significant risk for both children and adults. Thus, non-smokers who are exposed to a smoky environment have an increased risk of cancer, heart disease and respiratory disease (Action on Smoking and Health, 2014a ).

Benefits of stopping smoking

Table ​ Table1 1 lists the main benefits of stopping smoking. Smokers who stop before their mid-30s have approximately the same life expectancy as never smokers (Doll, Peto, Boreham, & Sutherland, 2004 ; Pirie, Peto, Reeves, Green, & Beral, 2013 ). After the age of 35 years or so, stopping smoking recovers 2–3 months of healthy life expectancy for every year of smoking avoided, or 4–6 h for every day (Jha & Peto, 2014 ).

Stopping smoking has different effects on different smoking-related diseases. Excess risk of heart attack caused by smoking reduces by 50% within 12 months of stopping smoking. Stopping smoking returns the rate of decline in lung function to the normal age-related decline, but does not reverse this; it reduces the frequency of ‘exacerbations’ (acute attacks of breathing difficulty resulting in death or hospitalisation) in COPD patients (US Surgeon General, 1990 ). Stopping smoking ‘freezes’ the risk of smoking-related cancers at the level experienced when stopping occurs but does not decrease it in absolute terms (US Surgeon General, 1990 ).

Smokers who stop show reduced levels of stress and mood disorder than those who continue (Royal College of Physicians and Royal College of Psychiatrists, 2013 ). They also report higher levels of happiness and life satisfaction than those who continue (Shahab & West, 2009 , 2012 ). This suggests that smoking may harm mental health, though other explanations cannot be ruled out on the current evidence.

Prevalence and patterns of smoking

Smoking prevalence.

There are estimated to be approximately 1 billion tobacco smokers worldwide (Eriksen, Mackay, & Ross, 2013 ), amounting to approximately 30% of men and 7% of women (Gowing et al., 2015 ).

Cigarette smoking prevalence in Great Britain was estimated to be 16.9% in 2015, the most recent year for which figures are available at the time of writing: slightly lower in women than men (Office of National Satistics, 2016 ). Smoking in Great Britain has declined by 0.7 percentage points per year since 2001 (from 26.9% of adults in 2001). In Australia, daily cigarette smoking has declined by 0.6 percentage points per year over a similar time period (from 22.4% of adults aged 18 + years in 2001 to 14.5% in 2015) (Australian Bureau of Statistics, 2015 ). However, international comparisons are confused by different countries using a different definition of what counts as being a smoker, and different methods for assessing prevalence. Australia only counts daily smokers in their headline figures. The situation in the US is even more misleading. The headline prevalence figure for the US is below 16%, but this does not include occasional smokers and people who smoke cigarillos which are essentially cigarettes in all but name and which have become increasingly popular in recent years. So the figure for prevalence that is most comparable to the figure for Great Britain is 20% (Jamal, 2016 ).

With the above caveats in mind, the figures in Table ​ Table2 2 for smoking prevalence in world regions in men and women provide very broad estimates (Gowing et al., 2015 ). Most noteworthy is that smoking prevalence in men is more than four times that in women globally but that the difference is much less in most parts of Europe, and that Eastern Europe as a whole has the highest smoking prevalence of any region in the world.

Note: Current smoking of any tobacco product, adults aged 15 years and older, age-standardised rate, by gender. ‘Tobacco smoking’ includes cigarettes, cigars, pipes or any other smoked tobacco products. ‘Current smoking’ includes both daily and non-daily or occasional smoking. From Gowing et al. ( 2015 ).

Smoking patterns

The most common age of first trying a cigarette in countries that have been studied is 10–15 years (Action on Smoking and Health, 2015b ; Talip, Murang, Kifli, & Naing, 2016 ); take up of regular smoking usually continues up to early 20s (Dierker et al., 2008 ).

Average daily cigarette consumption among smokers in the US and UK has declined steadily since the 1970s. In the UK, it is currently 11 cigarettes per day, and non-daily smoking is very rare (Action on Smoking and Health, 2016c ; Jarvis, Giovino, O’Connor, Kozlowski, & Bernert, 2014 ). Smokers take in an average of 1–1.5 mg of nicotine per cigarette (US Department of Health Human Services, 2014 ). The US figures on patterns of smoking are distorted by not counting ‘cigarillos’ and other smoked tobacco products which are used very much like cigarettes, whose prevalence has increased in recent years (Jamal et al., 2015 ). The reduction in daily cigarette consumption has not been accompanied by a reduction in daily nicotine intake (Jarvis et al., 2014 ). This could be due to the use of other smoked tobacco products (in the case of the US) or smokers smoking their cigarettes more intensively (taking more, deeper or longer puffs).

Smokers in England spend an average of £23 per week on cigarettes and this figure is slowly rising (West & Brown, 2015 ). In the UK, hand-rolled cigarettes have become increasingly popular with 34% of smokers currently reporting use of these products (Action on Smoking and Health, 2016c ). Men and people in more deprived socio-economic groups are more likely to smoke hand-rolled cigarettes (Action on Smoking and Health, 2016c ).

In most countries, there are strong negative associations between smoking prevalence and educational level, affluence and mental health; and positive associations with alcohol use disorder and substance use disorder (Action on Smoking and Health, 2016a , 2016c ; Royal College of Physicians and Royal College of Psychiatrists, 2013 ; Talati, Keyes, & Hasin, 2016 ). In the UK, average daily cigarette consumption is higher for men than women, and higher in smokers in more deprived socio-economic groups and those with mental health problems (Action on Smoking and Health, 2016c ).

Psychological, pharmacological and social factors involved in smoking and smoking cessation

The natural history of smoking can be modelled as states and factors that influence the transition between these. Figure ​ Figure1 1 shows transitions that have been researched – the variables identified in the diagram are listed descriptively without attempting to explain how they may be connected.

Factors associated with transitions in the natural history of smoking (parentheses indicate negative associations).

Smoking initiation

Important factors predicting initiation in western societies are: having friends who smoke, having parents who smoke, low social grade, tendency to mental health problems and impulsivity (Action on Smoking and Health, 2015b ). Transition to daily smoking follows a highly variable pattern sometimes being very rapid and sometimes taking several years (Schepis & Rao, 2005 ). Important factors predicting transition to regular smoking are: having friends who smoke, weak academic orientation, low parental support, pro-smoking attitudes, drinking alcohol and low socio-economic status (Action on Smoking and Health, 2015b ).

Smoking initiation has a ‘heritability’ (the proportion of variance in a characteristic that is attributable to genetic rather than environmental variance) of approximately 30–50% in western societies (Vink, Willemsen, & Boomsma, 2005 ). This means that differences in genetic make-up account for almost half of the difference in likelihood of starting smoking between individuals. This does not mean that environmental factors do not also play a crucial role as is evident from the very large decline in smoking initiation since the 1970s in many western countries.

The heritability of cigarette addiction (as distinct from smoking) is approximately 70–80% in western societies (Vink et al., 2005 ). Cigarette addiction here refers to the extent to which someone experiences a strong need to smoke. It is usually indexed by a combination of number of cigarettes per day and time from waking to smoking the first cigarette of the day (Kozlowski, Porter, Orleans, Pope, & Heatherton, 1994 ). It can also be indexed by the self-reported strength of urges to smoke (Fidler, Shahab, & West, 2011 ). Heritability of cigarette addiction, as indexed by failure of attempts to stop, is higher than the heritability for smoking and for initiation of smoking. This suggests that differences in genetic inheritance play a larger role in being able to stop smoking than in starting to smoke.

Cigarette addiction

Cigarette addiction stems from the fact that smoking provides highly controllable doses of the drug, nicotine, rapidly to the brain in a form that is accessible, affordable and palatable (West, 2009 ; West & Shiffman, 2016 ). Nicotine provided more slowly, for example by the nicotine transdermal patch, is much less addictive. It is possible that one or more mono-amine oxidase inhibitors in cigarette smoke add to, or synergise, the addictive properties of nicotine (Hogg, 2016 ).

The psychopharmacology of cigarette addiction is complex and far from fully understood. The following paragraphs summarise the current narrative.

Nicotine resembles the naturally occurring neurotransmitter, acetylcholine, sufficiently to attach itself to a subset of neuronal receptors for this neurotransmitter in the brain. These are called ‘nicotinic acetylcholine receptors’. When it does this with receptors in the ventral tegmental area in the midbrain, it causes an increased rate of firing of the nerves projecting forward from that area to another part of the brain called the nucleus accumbens. This causes release of another neurotransmitter called dopamine in the nucleus accumbens.

Dopamine release and uptake by neurones in the nucleus accumbens is believed to be central to all addictive behaviours. It acts as a neural ‘teaching signal’ which causes the brain to form an association between the current situation as perceived and the impulse to engage in whatever action immediately preceded this release. In the case of smoking, this creates an urge to smoke in situations in which smoking frequently occurs. These are often referred to as ‘cue-driven smoking urges’ or ‘situational cravings’ (West, 2009 ; West & Shiffman, 2016 ). This explains why even non-daily smokers often find it difficult to stop smoking altogether.

Repeated ingestion of nicotine from cigarettes causes changes to the functioning of the ventral tegmental area and nucleus accumbens such that when brain concentrations of nicotine are lower than usual, there is an abnormally low level of neural activity in these regions. This leads to feelings of need for behaviours that have in the past restored normal functioning, typically smoking. This feeling of need can be thought of as a kind of ‘nicotine hunger’, also called ‘background craving’ (West, 2009 ; West & Shiffman, 2016 ). This is probably why time between waking and first cigarette of the day is a useful predictor of difficulty stopping smoking (Vangeli, Stapleton, Smit, Borland, & West, 2011 ). So ‘cue-driven smoking urges’ and ‘nicotine hunger’ are important factors contributing to smoking behaviour and thought to be the primary mechanisms underpinning cigarette addiction (West, 2009 ; West & Shiffman, 2016 ).

When smokers abstain from cigarettes, within a few hours many of them start to experience nicotine withdrawal symptoms. Withdrawal symptoms from a drug are temporary symptoms that arise when the drug dose is reduced or use is terminated. They arise from neural adaptation to the presence of the drug in the central nervous system. For smoking, the most common early onset symptoms are: irritability, restlessness and difficult concentrating. Depression and anxiety have also been observed in some smokers. These symptoms typically last 1 to 4 weeks (West, 2009 ; West & Shiffman, 2016 ).

After a day or two of stopping smoking, many smokers experience other symptoms: increased appetite, constipation, mouth ulcers, cough, and weight gain. Increased appetite tends to last for at least 3 months; weight gain (averaging around 6 kg) tends to be permanent; other symptoms tend to last a few weeks. The increased appetite, weight gain and constipation arise from termination of nicotine intake but the others are probably related to other effects of stopping smoking (West, 2009 ; West & Shiffman, 2016 ).

Any of the above effects of abstinence may in individual cases promote resumption of smoking following a quit attempt but statistically the association is inconsistent and weak; the main factors driving relapse appear to be cue-driven smoking urges and nicotine hunger (Fidler & West, 2011 ; West, 2009 ; West & Shiffman, 2016 ).

Many smokers report that smoking helps them cope with stress and increases their ability to concentrate. However, this appears to be because when they go for a period without smoking they experience nicotine withdrawal symptoms that are relieved by smoking. Long-term smokers who stop report lower levels of stress than when they were smoking and no reduction in ability to concentrate (West, 2009 ; West & Shiffman, 2016 ).

It is commonly thought that smokers with mental health problems are using cigarettes to ‘self-medicate’ or treat their psychological symptoms. However, the evidence indicates that neither nicotine nor smoking improves psychological symptoms, and people with serious mental health disorders who stop smoking do not experience a worsening of mental health. In fact some studies have found an improvement (Royal College of Physicians and Royal College of Psychiatrists, 2013 ).

Smoking cessation

For most smokers, cessation requires a determined attempt to stop and then sufficient resolve in the following weeks and months to overcome what are often powerful urges to smoke. Factors that predict quit attempts differ from those that predict the success of those attempts (Vangeli et al., 2011 ). Approximately 5% of unaided quit attempts succeed for at least 6 months (Hughes, Keely, & Naud, 2004 ). Relapse after this point is estimated to be around 50% over subsequent years (Stapleton & West, 2012 ).

The most common self-reported reasons for smoking are stress relief and enjoyment, with around half of smokers reporting these smoking motives. Weight control, aiding concentration and socialising are also quite commonly cited (Fidler & West, 2009 ). Smoking for supposed stress relief, improved concentration, weight control or other functions has not been found to be related to attempts to stop or success of attempts to stop (Fidler & West, 2009 ). Smokers who report enjoying smoking are less likely to try to stop but not less likely to succeed if they do try (Fidler & West, 2011 ). In addition, having a positive smoker identity (liking being a smoker) predicts not trying to quit, over and above enjoyment of smoking (Fidler & West, 2009 ).

No clear association has been found between the number of times smokers have tried to stop in the past and their chances of success the next time they try (Vangeli et al., 2011 ). However, having tried to stop in the past few months is predictive of failure of the next quit attempt (Zhou et al., 2009 ). Belief in the harm caused by smoking is predictive of smokers making quit attempts but not the success of those attempts (Vangeli et al., 2011 ).

Some clinical studies have found that women were less likely to succeed in quit attempts than men but large population studies have found no difference in success rates between the genders (Vangeli et al., 2011 ) so it may be the case that women who seek help with stopping have greater difficulty than men who seek help with stopping.

Number of cigarettes smoked per day, time between waking and the first cigarette of the day and rated strength of urges to smoke prior to a quit attempt have been found to predict success of quit attempts (Vangeli et al., 2011 ).

Quit attempts that involve gradual reduction are less likely to succeed than those that involve quitting abruptly, even after controlling statistically for measures of cigarette addiction, confidence in quitting, other methods used to quit (e.g. nicotine replacement therapy) and sociodemographic factors (Lindson-Hawley et al., 2016 ).

Interventions to combat smoking

There is extensive evidence on interventions that can reduce smoking prevalence, either by reducing initiation or promoting cessation. Table ​ Table3 3 lists those that have the strongest evidence.

Population-level interventions

Increasing the financial cost of smoking through tax increases and control of illicit supply on average reduces overall consumption with a typical price elasticity globally of 0.4 (meaning that for every 10% increase in the real cost there is a 4% decrease in the number of cigarettes purchased). Most of the effect is in getting smokers to reduce their daily cigarette consumption so the effect on smoking prevalence has been found to be an average of a 1–2 percentage point prevalence reduction for every 10% increase in the real cost (Levy, Huang, Havumaki, & Meza, 2016 ). It has been claimed that increasing taxes on tobacco increases the amount of smuggling of cheap tobacco, but the evidence does not support this (Action on Smoking and Health, 2015a ; Joossens & Raw, 2003 ).

Social marketing campaigns (e.g. TV advertising) can prevent smoking uptake, increase the rate at which smokers try to quit and improve the chances of success. This can lead to a reduction in smoking prevalence. Their effectiveness varies considerably with intensity, type of campaign and context (Bala, Strzeszynski, Topor-Madry, & Cahill, 2013 ; Hoffman & Tan, 2015 ).

Legislating to ban smoking in all indoor public areas may have a one-off effect on reducing smoking prevalence but findings are inconsistent across different countries (Bala et al., 2013 ). For example, in countries such as France it was not possible to detect an effect while in England, there did appear to be a decline in prevalence following the ban.

Although it is hard to show conclusively, circumstantial evidence suggests that banning tobacco advertising and putting large graphic health warnings on cigarette packets may have reduced smoking prevalence in some countries (Hoffman & Tan, 2015 ; Noar et al., 2016 ).

Individual-level interventions to promote smoking cessation

Brief advice.

Brief advice to stop smoking from a physician and offer of support to all smokers, regardless of motivation to quit, has been found in randomised trials to increase rate of quitting by an average of 2 percentage points of all those receiving it, whether or not they were initially interested in quitting (Stead et al., 2013 ). The offer of support appears to be more effective in getting smokers to try to quit than just advising smokers to stop (Aveyard, Begh, Parsons, & West, 2012 ).

Pharmacotherapy

Using a form of nicotine replacement therapy (NRT: transdermal patch, chewing gum, nasal spray, mouth spray, lozenge, inhalator, dissolvable strip) for at least 6 weeks from the start of a quit attempt increases the chances of long-term success of that quit attempt by about 3–7 percentage points if the user is under the care of a health professional or provided as part of a structured support programme (Stead et al., 2012 ). Some studies have found that NRT when bought from a shop and used without any additional structured support does not improve the chances of success at stopping (Kotz, Brown, & West, 2014a , 2014b ). A small proportion of people who use NRT to stop smoking continue to use it for months or even years after stopping smoking, but NRT appears to carry minimal risk to long-term users (Royal College of Physicians, 2016 ; Stead et al., 2012 ).

Data are sparse but at present, using an electronic cigarette in a quit attempt appears to increase the chances of success at stopping on average by an amount broadly similar to that from NRT; the variety of products available and the greater similarity to smoking appear to make them more attractive to many smokers as a means of stopping than NRT (McNeill et al., 2015 ; Royal College of Physicians, 2016 ). Electronic cigarettes deliver nicotine to users by heating a liquid containing nicotine, propylene glycol or glycerol and usually flavourings to create a vapour that is inhaled. They appear to carry minimal acute risk to users. If they are used long-term, their risk is almost certainly much less than that of smoking (based on concentrations of chemicals in the vapour) (McNeill et al., 2015 ; Royal College of Physicians, 2016 ).

‘Dual-form NRT’ (combining a transdermal NRT patch and one of the other forms) increases the chances of success at stopping more than ‘single-form NRT’ (just using one of the products) (Stead et al., 2012 ). Starting to use a nicotine transdermal patch several weeks before the target quit date may improve the chances of success at quitting compared with starting on the quit date (Stead et al., 2012 ).

Taking the prescription anti-depressant, bupropion (brand name Zyban), improves the chances of success of quit attempts by a similar amount to single-form NRT (Hughes, Stead, Hartmann-Boyce, Cahill, & Lancaster, 2014 ). Bupropion often leads to sleep disturbance and carries a very small risk of seizure. Bupropion probably works by reducing urges to smoke rather than any effect on depressed mood, but how it does this is not known. It is contra-indicated in pregnant smokers and people with an elevated seizure risk or history of eating disorder (Hughes et al, 2014 ). Taking the tricyclic anti-depressant, nortriptyline also improves the chances of success of quit attempts, probably by about the same amount as bupropion and NRT (Hughes et al., 2014 ). Its mechanism of action is not known. Nortriptyline often leads to dry mouth and sleep disorder and can be fatal in overdose (Hughes et al., 2014 ).

Taking the nicotinic-acetylcholine receptor partial agonist, varenicline (brand name Chantix in the US and Champix elsewhere), improves the chances of success by about 50% more than bupropion or single-form NRT (Cahill, Lindson-Hawley, Thomas, Fanshawe, & Lancaster, 2016 ). This is true for smokers with or without a psychiatric disorder (Anthenelli et al., 2016 ). Varenicline appears to work both by reducing urges to smoke and the rewarding effect of nicotine should a lapse occur (West, Baker, Cappelleri, & Bushmakin, 2008 ). Varenicline often leads to sleep disturbance and nausea. Serious neuropsychiatric and cardiovascular adverse reactions have been reported, but in comparative studies these have not been found to be more common than placebo or NRT (Anthenelli et al., 2016 ; Cahill et al., 2016 ; Sterling, Windle, Filion, Touma, & Eisenberg, 2016 ).

Taking the nicotinic-acetylcholine receptor partial agonist, cytisine, appears to improve the chances of success at least as much as single-form NRT and probably more (Cahill et al., 2016 ). Cytisine often causes nausea. No serious adverse reactions have been reported to date (Cahill et al., 2016 ). Where it is licensed for sale, cytisine is less than 1/10th the cost of other smoking cessation medications (Cahill et al., 2016 ).

Behavioural support

There is good evidence that behavioural interventions of many kinds, delivered though several modalities can help smokers to stop. Thus, behavioural support (encouragement, advice and discussion) from a trained stop-smoking specialist, provided at least weekly until at least 4 weeks following the target quit date can increase the chances of long-term success of a quit attempt by about 3–7 percentage points, whether it is given by phone or face-to-face (Lancaster & Stead, 2005 ). Group behavioural support (specialist-led groups of smokers stopping together and engaging in a structured discussion about their experiences), involving at least weekly sessions lasting until at least 4 weeks after the target quit date can increase the chances of success of a quit attempt by a similar amount or possibly more than individual support (Stead & Lancaster, 2005 ). Scheduled, multi-session telephone support can improve rates of success at stopping smoking by a broadly similar amount (Stead, Hartmann-Boyce, Perera, & Lancaster, 2013 ) but some large studies have failed to detect an effect so contextual factors and/or the precise type of support could be crucial to success. The effects of behavioural support and medication/NRT on success at stopping smoking appear to combine roughly additively (Stead, Koilpillai, & Lancaster, 2015 ). Smoking cessation support appears to be effective in primary care, secondary care and worksite settings (Cahill & Lancaster, 2014 ; West et al., 2015 ). Financial incentives, in the form of vouchers, have been found to increase smoking cessation rates for as long as they are in place (Cahill, Hartmann-Boyce, & Perera, 2015 ; Higgins & Solomon, 2016 ). Printed self-help materials can improve the chances of success at stopping long term by around 1–2 percentage points (Hartmann-Boyce, Lancaster, & Stead, 2014 ).

There is still relatively limited evidence on the effectiveness of digital support interventions for smoking cessation. Thus, while there is evidence that tailored, interactive websites can improve the chances of success at stopping smoking compared with no support, brief written materials or static information websites, many of those tested have not been found to be effective and it is not clear what differentiates those that are effective from those that are not (Graham et al., 2016 ). Text messaging programmes have been found to increase the chances of success of quit attempts by about 2–7 percentage points (Whittaker, McRobbie, Bullen, Rodgers, & Gu, 2016 ). There is currently insufficient evidence to know whether smartphone applications can improve success rates of quit attempts, although preliminary data suggest that they might (Whittaker et al., 2016 ). Evidence on alternative and complementary therapies is not sufficient to make confident statements about their effectiveness as aids to smoking cessation (Barnes et al., 2010 ; White, Rampes, Liu, Stead, & Campbell, 2014 ).

Overall, the highest smoking cessation rates appear to be achieved using specialist face-to-face behavioural support together with either varenicline or dual form NRT. With this support, continuous abstinence rates up to 52 weeks, verified by expired-air carbon monoxide tests, of more than 40% have been achieved (Kralikova et al., 2013 ). More commonly, 52-week continuous abstinence rates with this treatment are between 15 and 25% (West et al., 2015 ).

Smoking cessation support for pregnant smokers

In pregnant smokers, there is some evidence that NRT can help promote smoking cessation but evidence for an effect sustained to end of pregnancy is not conclusive (Sterling et al., 2016 ). There is also evidence that written self-help materials and face-to-face behavioural support can aid smoking cessation (Jones, Lewis, Parrott, Wormall, & Coleman, 2016 ), and financial incentives have also been found to improve quitting rates among pregnant smokers (Tappin et al., 2015 ). Almost half of women who stop smoking during pregnancy as a result of a clinical intervention relapse to smoking within 6 months of the birth (Jones et al., 2016 ).

Effectiveness of programmes to reduce smoking uptake

School-based programmes that involve both social competence training and peer-led social influence have been found to reduce smoking uptake (Georgie, Sean, Deborah, Matthew, & Rona, 2016 ) but educational programmes have not (Thomas, McLellan, & Perera, 2013 ). Mass media campaigns and increasing the financial cost of smoking reduce smoking uptake (Brinn, Carson, Esterman, Chang, & Smith, 2012 ; van Hasselt et al., 2015 ).

Reducing the harm from tobacco and nicotine use

Smokers who report that they are reducing their cigarette consumption smoke only 1–2 fewer cigarettes per day on average than when they say they are not (Beard et al., 2013 ). Clinical trials have found that use of NRT while smoking can substantially reduce cigarette consumption compared with placebo (Royal College of Physicians, 2016 ) but national surveys show very little reduction in cigarette consumption when smokers take up use of NRT in real-world settings (Beard et al., 2013 ). The benefit from using NRT while continuing to smoke appears to be in promoting subsequent smoking cessation. Using NRT (or varenicline) to reduce cigarette smoking with no immediate plans to quit leads to increased rates of quitting subsequently (Wu, Sun, He, & Zeng, 2015 ).

‘Snus’, a form of tobacco that is placed between the gums and the cheek and which is prepared in a way that is very low in carcinogens, gives high doses of nicotine but without evidence of an increase in risk of major tobacco-related cancers and either no, or a small, increase in risk of heart disease. It does appear to increase risk of periodontal disease, however. Snus is very popular in Sweden. Sweden has very low rates of smoking and tobacco-related disease indicating that a form of nicotine intake other than smoking can become popular and suggesting that this can contribute to a substantial reduction in tobacco-related harm (Royal College of Physicians, 2016 ).

The introduction of complete bans on smoking in indoor public areas can also be considered as a harm reduction measure. In this case, the main issue is harm to non-tobacco users. The evidence shows that such bans have been rapidly followed in the UK and several other jurisdictions by a reduction in heart attacks in non-smokers (Action on Smoking and Health, 2014a ).

Conclusions

Tobacco smoking causes death and disability on a huge scale and only about half of smokers report enjoying it. Despite this, approximately 1 billion adults engage in this behaviour worldwide and only around 5% of unaided quit attempts succeed for 6 months or more. The main reason appears to be that cigarettes deliver nicotine rapidly to the brain in a form that is convenient, and palatable. Nicotine acts on the brain to create urges to smoke in situations where smoking would normally occur and when brain nicotine levels become depleted. Concern about the harm from, and financial cost of, smoking are mostly not sufficient to counter this.

Governments can reduce smoking prevalence by raising the cost of smoking through taxation, mounting sustained social marketing campaigns, ensuring that health professionals routinely advise smokers to stop and offer support for quitting, and make available pharmacological and behavioural support for stopping.

Statement of competing interests

RW has, within the past 3 years, undertaken research and consultancy for companies that develop and manufacture smoking cessation medications (Pfizer, GSK, and J&J). He is an unpaid advisor to the UK’s National Centre for Smoking cessation and Training. His salary is funded by Cancer Research UK.

Disclosure statement

No potential conflict of interest was reported by the author.

This work was supported by Cancer Research UK [grant number C1417/A22962].

• Action on Smoking and Health (2013). Smoking and reproduction . London: ASH; Retrieved from http://www.ash.org.uk/files/documents/ASH_112.pdf [ Google Scholar ]
• Action on Smoking and Health (2014a). Secondhand smoke . London: ASH; Retrieved from http://www.ash.org.uk/files/documents/ASH_113.pdf [ Google Scholar ]
• Action on Smoking and Health (2014b). What’s in a cigarette? London: ASH; Available from http://www.ash.org.uk/files/documents/ASH_117.pdf [ Google Scholar ]
• Action on Smoking and Health (2015a). Illicit trade in tobacco . London: ASH; Available from http://www.ash.org.uk/files/documents/ASH_122.pdf [ Google Scholar ]
• Action on Smoking and Health (2015b). Young people and smoking . London: ASH; Retrieve from http://www.ash.org.uk/files/documents/ASH_108.pdf [ Google Scholar ]
• Action on Smoking and Health (2016a). Smoking and mental health . London: ASH; Retrieved from http://www.ash.org.uk/files/documents/ASH_120.pdf [ Google Scholar ]
• Action on Smoking and Health (2016b). Smoking statistics: Illness and death . London: ASH; Retrieved from http://www.ash.org.uk/files/documents/ASH_107.pdf [ Google Scholar ]
• Action on Smoking and Health (2016c). Smoking statistics: Who smokes and how much? London: ASH; Retrieved from http://www.ash.org.uk/files/documents/ASH_106.pdf [ Google Scholar ]
• Anthenelli R. M., Benowitz N. L., West R., St Aubin L., McRae T., Lawrence D., & Evins A. E. (2016). Neuropsychiatric safety and efficacy of varenicline, bupropion, and nicotine patch in smokers with and without psychiatric disorders (EAGLES): A double-blind, randomised, placebo-controlled clinical trial . The Lancet , , 2507–2520. doi: 10.1016/s0140-6736(16)30272-0 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Australian Bureau of Statistics (2015). National health survey . Canberra: Author. [ Google Scholar ]
• Aveyard P., Begh R., Parsons A., & West R. (2012). Brief opportunistic smoking cessation interventions: a systematic review and meta-analysis to compare advice to quit and offer of assistance . Addiction , , 1066–1073. doi: 10.1111/j.1360-0443.2011.03770.x [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Bala M. M., Strzeszynski L., Topor-Madry R., & Cahill K (2013). Mass media interventions for smoking cessation in adults . Cochrane Database of Systematic Reviews , , Cd004704. doi: 10.1002/14651858.CD004704.pub3 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Barnes J., Dong C. Y., McRobbie H., Walker N., Mehta M., & Stead L. F. (2010). Hypnotherapy for smoking cessation . Cochrane Database of Systematic Reviews , , Cd001008. doi: 10.1002/14651858.CD001008.pub2 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Beard E., McNeill A., Aveyard P., Fidler J., Michie S., & West R. (2013). Association between use of nicotine replacement therapy for harm reduction and smoking cessation: A prospective study of English smokers . Tobacco Control , , 118–122. 10.1136/tobaccocontrol-2011-050007 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Benowitz N. L. (1997). The role of nicotine in smoking-related cardiovascular disease . Preventive Medicine , , 412–417. 10.1006/pmed.1997.0175 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Benowitz N. L. (1998). Nicotine safety and toxicity . Oxford: Oxford University Press. [ Google Scholar ]
• Brinn M. P., Carson K. V., Esterman A. J., Chang A. B., & Smith B. J. (2012). Cochrane review: Mass media interventions for preventing smoking in young people . Evidence-based Child Health: A Cochrane Review Journal , , 86–144. 10.1002/ebch.v7.1 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Cahill K., Hartmann-Boyce J., & Perera R (2015). Incentives for smoking cessation . Cochrane Database of Systematic Reviews , , Cd004307. doi: 10.1002/14651858.CD004307.pub5 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Cahill K., & Lancaster T (2014). Workplace interventions for smoking cessation . Cochrane Database of Systematic Reviews , , Cd003440. doi: 10.1002/14651858.CD003440.pub4 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Cahill K., Lindson-Hawley N., Thomas K. H., Fanshawe T. R., & Lancaster T (2016). Nicotine receptor partial agonists for smoking cessation . Cochrane Database of Systematic Reviews , , Cd006103. doi: 10.1002/14651858.CD006103.pub7 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Critchley J., & Unal B. (2003). Health effects associated with smokeless tobacco: A systematic review . Thorax , , 435–443. 10.1136/thorax.58.5.435 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Dierker L., He J., Kalaydjian A., Swendsen J., Degenhardt L., Glantz M., & Merikangas K. (2008). The importance of timing of transitions for risk of regular smoking and nicotine dependence . Annals of Behavioral Medicine , , 87–92. doi: 10.1007/s12160-008-9051-x [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Doll R., Peto R., Boreham J., & Sutherland I. (2004). Mortality in relation to smoking: 50 years’ observations on male British doctors . British Medical Journal , , 1519. doi: 10.1136/bmj.38142.554479.AE [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Eriksen M., Mackay J., & Ross H (2013). The tobacco atlas . New York, NY: American Cancer Society. [ Google Scholar ]
• Ferri C. P., West R., Moriyama T. S., Acosta D., Guerra M., Huang Y., … Prince M. J. (2011). Tobacco use and dementia: Evidence from the 1066 dementia population-based surveys in Latin America, China and India . International Journal of Geriatric Psychiatry , , 1177–1185. doi: 10.1002/gps.2661 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Fidler J., Shahab L., & West R. (2011). Strength of urges to smoke as a measure of severity of cigarette dependence: Comparison with the fagerström test for nicotine dependence and its components . Addiction , , 631–638. 10.1111/add.2011.106.issue-3 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Fidler J., & West R. (2009). Self-perceived smoking motives and their correlates in a general population sample . Nicotine & Tobacco Research , , 1182–1188. doi: 10.1093/ntr/ntp120 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Fidler J., & West R. (2011). Enjoyment of smoking and urges to smoke as predictors of attempts and success of attempts to stop smoking: A longitudinal study . Drug and Alcohol Dependence , , 30–34. doi: 10.1016/j.drugalcdep.2010.10.009 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Georgie J. M., Sean H., Deborah M. C., Matthew H., & Rona C. (2016). Peer-led interventions to prevent tobacco, alcohol and/or drug use among young people aged 11–21 years: a systematic review and meta-analysis . Addiction , , 391–407. doi: 10.1111/add.13224 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Gowing L. R., Ali R. L., Allsop S., Marsden J., Turf E. E., West R., & Witton J. (2015). Global statistics on addictive behaviours: 2014 status report . Addiction , , 904–919. doi: 10.1111/add.12899 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Graham A. L., Carpenter K. M., Cha S., Cole S., Jacobs M. A., Raskob M., & Cole-Lewis H. (2016). Systematic review and meta-analysis of internet interventions for smoking cessation among adults . Substance Abuse and Rehabilitation , , 55–69. doi: 10.2147/sar.s101660 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Hartmann-Boyce J., Lancaster T., & Stead L. F (2014). Print-based self-help interventions for smoking cessation . Cochrane Database of Systematic Reviews , , Cd001118. doi: 10.1002/14651858.CD001118.pub3 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Higgins S. T., & Solomon L. J. (2016). Some recent developments on financial incentives for smoking cessation among pregnant and newly postpartum women . Current Addiction Reports , , 9–18. doi: 10.1007/s40429-016-0092-0 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Hoffman S. J., & Tan C. (2015). Overview of systematic reviews on the health-related effects of government tobacco control policies . BMC Public Health , , 744. doi: 10.1186/s12889-015-2041-6 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Hogg R. C. (2016). Contribution of monoamine oxidase inhibition to tobacco dependence: A review of the evidence . Nicotine & Tobacco Research , , 509–523. doi: 10.1093/ntr/ntv245 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Hughes J. R., Keely J., & Naud S. (2004). Shape of the relapse curve and long-term abstinence among untreated smokers . Addiction , , 29–38. 10.1111/add.2004.99.issue-1 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Hughes J. R., Stead L. F., Hartmann-Boyce J., Cahill K., & Lancaster T (2014). Antidepressants for smoking cessation . Cochrane Database of Systematic Reviews , , Cd000031. doi: 10.1002/14651858.CD000031.pub4 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Jamal A. (2016). Current cigarette smoking among adults – United States, 2005–2015 . Morbidity and Mortality Weekly Report , , 1205–1211 [ PubMed ] [ Google Scholar ]
• Jamal A., Homa D., O’Connor E., Babb S., Caraballo R., Singh T., & King B. (2015). Current cigarette smoking among adults – United States, 2005–2014 . Morbidity and Mortality Weekly Report , , 1233–1240. 10.15585/mmwr.mm6444a2 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Jarvis M. J., Giovino G. A., O’Connor R. J., Kozlowski L. T., & Bernert J. T. (2014). Variation in nicotine intake among U.S. cigarette smokers during the past 25 years: Evidence from NHANES surveys . Nicotine & Tobacco Research , , 1620–1628. doi: 10.1093/ntr/ntu120 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Jha P., & Peto R. (2014). Global effects of smoking, of quitting, and of taxing tobacco . New England Journal of Medicine , , 60–68. doi: 10.1056/NEJMra1308383 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Jones M., Lewis S., Parrott S., Wormall S., & Coleman T. (2016). Re-starting smoking in the postpartum period after receiving a smoking cessation intervention: A systematic review . Addiction , , 981–990. doi: 10.1111/add.13309 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Joossens L., & Raw M. (2003). Turning off the tap: The real solution to cigarette smuggling . International Journal of Tuberculosis and Lung Disease , , 214–222. [ PubMed ] [ Google Scholar ]
• Kotz D., Brown J., & West R. (2014a). Prospective cohort study of the effectiveness of smoking cessation treatments used in the “real world” . Mayo Clinic Proceedings , , 1360–1367. doi: 10.1016/j.mayocp.2014.07.004 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Kotz D., Brown J., & West R. (2014b). ‘Real-world’ effectiveness of smoking cessation treatments: A population study . Addiction , , 491–499. doi: 10.1111/add.12429 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Kozlowski L. T., Porter C. Q., Orleans C. T., Pope M. A., & Heatherton T. (1994). Predicting smoking cessation with self-reported measures of nicotine dependence: FTQ, FTND, and HSI . Drug and Alcohol Dependence , , 211–216. 10.1016/0376-8716(94)90158-9 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Kralikova E., Kmetova A., Stepankova L., Zvolska K., Davis R., & West R. (2013). Fifty-two-week continuous abstinence rates of smokers being treated with varenicline versus nicotine replacement therapy . Addiction , , 1497–1502. doi: 10.1111/add.12219 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Lancaster T., & Stead L. F (2005). Individual behavioural counselling for smoking cessation . Cochrane Database of Systematic Reviews , , Cd001292. doi: 10.1002/14651858.CD001292.pub2 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Levy D. T., Huang A. T., Havumaki J. S., & Meza R. (2016). The role of public policies in reducing smoking prevalence: Results from the Michigan SimSmoke tobacco policy simulation model . Cancer Causes and Control , , 615–625. doi: 10.1007/s10552-016-0735-4 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Lindson-Hawley N., Banting M., West R., Michie S., Shinkins B., & Aveyard P. (2016). Gradual versus abrupt smoking cessation: A randomized, controlled noninferiority trial . Annals of Internal Medicine , , 585–592. doi: 10.7326/m14-2805 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• McNeill A., Brose L., Calder R., Hitchman S., Hajek P., & McRobbie H. (2015). E-cigarettes: An evidence update. A report commissioned by Public Health England . London: Public Health England. [ Google Scholar ]
• Noar S. M., Francis D. B., Bridges C., Sontag J. M., Ribisl K. M., & Brewer N. T. (2016). The impact of strengthening cigarette pack warnings: Systematic review of longitudinal observational studies . Social Science & Medicine , , 118–129. doi: 10.1016/j.socscimed.2016.06.011 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Office of National Satistics (2016). Smoking prevalence in Great Britain . London: ONS; Retrieved from https://www.ons.gov.uk/peoplepopulationandcommunity/healthandsocialcare/drugusealcoholandsmoking/datasets/adultsmokinghabitsingreatbritain [ Google Scholar ]
• Royal College of Physicians and Royal College of Psychiatrists (2013). Smoking and mental health . London: RCP. [ Google Scholar ]
• Pirie K., Peto R., Reeves G. K., Green J., & Beral V. (2013). The 21st century hazards of smoking and benefits of stopping: A prospective study of one million women in the UK . Lancet , , 133–141. doi: 10.1016/s0140-6736(12)61720-6 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Royal College of Physicians. (2016). Nicotine without smoke: Tobacco harm reduction . London: RCP. [ Google Scholar ]
• Schepis T. S., & Rao U. (2005). Epidemiology and etiology of adolescent smoking . Current Opinion in Pediatrics , , 607–612. 10.1097/01.mop.0000176442.49743.31 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Shahab L., & West R. (2009). Do ex-smokers report feeling happier following cessation? Evidence from a cross-sectional survey . Nicotine & Tobacco Research , , 553–557. doi: 10.1093/ntr/ntp031 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Shahab L., & West R. (2012). Differences in happiness between smokers, ex-smokers and never smokers: cross-sectional findings from a national household survey . Drug and Alcohol Dependence , , 38–44. doi: 10.1016/j.drugalcdep.2011.08.011 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Stapleton J. A., & West R. (2012). A direct method and ICER tables for the estimation of the cost-effectiveness of smoking cessation interventions in general populations: application to a new cytisine trial and other examples . Nicotine & Tobacco Research , , 463–471. doi: 10.1093/ntr/ntr236 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Stead L. F., Buitrago D., Preciado N., Sanchez G., Hartmann-Boyce J., & Lancaster T (2013). Physician advice for smoking cessation . Cochrane Database of Systematic Reviews , , Cd000165. doi: 10.1002/14651858.CD000165.pub4 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Stead L. F., Hartmann-Boyce J., Perera R., & Lancaster T (2013). Telephone counselling for smoking cessation . Cochrane Database of Systematic Reviews , , Cd002850. doi: 10.1002/14651858.CD002850.pub3 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Stead L. F., Koilpillai P., & Lancaster T (2015). Additional behavioural support as an adjunct to pharmacotherapy for smoking cessation . Cochrane Database of Systematic Reviews , , Cd009670. doi: 10.1002/14651858.CD009670.pub3 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Stead L. F., & Lancaster T (2005). Group behaviour therapy programmes for smoking cessation . Cochrane Database of Systematic Reviews , , Cd001007. doi: 10.1002/14651858.CD001007.pub2 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Stead L. F., Perera R., Bullen C., Mant D., Hartmann-Boyce J., Cahill K., & Lancaster T. (2012). Nicotine replacement therapy for smoking cessation . Cochrane Database of Systematic Reviews , , Cd000146. doi: 10.1002/14651858.CD000146.pub4 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Sterling L. H., Windle S. B., Filion K. B., Touma L., & Eisenberg M. J. (2016). Varenicline and adverse cardiovascular events: A systematic review and meta-analysis of randomized controlled trials . Journal of the American Heart Association , , doi: 10.1161/jaha.115.002849 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Talati A., Keyes K., & Hasin D. (2016). Changing relationships between smoking and psychiatric disorders across twentieth century birth cohorts: Clinical and research implications . Molecular Psychiatry , , 464–471. doi: 10.1038/mp.2015.224 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Talip T., Murang Z., Kifli N., & Naing L. (2016). Systematic review of smoking initiation among Asian Adolescents, 2005–2015: Utilizing the frameworks of triadic influence and planned behavior . Asian Pacific Journal of Cancer Prevention , , 3341–3355. [ PubMed ] [ Google Scholar ]
• Tappin D., Bauld L., Purves D., Boyd K., Sinclair L., MacAskill S., & Coleman T. (2015). Financial incentives for smoking cessation in pregnancy: Randomised controlled trial . British Medical Journal , , h134. doi: 10.1136/bmj.h134 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Thomas R. E., McLellan J., & Perera R. (2013). School-based programmes for preventing smoking . Evidence-based Child Health: A Cochrane Review Journal , , 1616–2040. 10.1002/ebch.v8.5 [ CrossRef ] [ Google Scholar ]
• US Department of Health and Human Services (2004). The health consequences of smoking: a report of the surgeon general (p. 62). Atlanta, GA: US Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health. [ Google Scholar ]
• US Department of Health Human Services (2014). The health consequences of smoking – 50 years of progress: A report of the surgeon general (p. 17). Atlanta, GA: US Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health. [ Google Scholar ]
• US Surgeon General. (1990). The health benefits of smoking cessation . Washington, DC: Department of Health and Human Services. [ Google Scholar ]
• Ussher M., Brown J., Rajamanoharan A., & West R. (2014). How do prompts for attempts to quit smoking relate to method of quitting and quit success? Annals of Behavioral Medicine , , 358–368. doi: 10.1007/s12160-013-9545-z [ PubMed ] [ CrossRef ] [ Google Scholar ]
• van Hasselt M., Kruger J., Han B., Caraballo R. S., Penne M. A., Loomis B., & Gfroerer J. C. (2015). The relation between tobacco taxes and youth and young adult smoking: What happened following the 2009 US federal tax increase on cigarettes? Addictive Behaviors , , 104–109. 10.1016/j.addbeh.2015.01.023 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Vangeli E., Stapleton J., Smit E. S., Borland R., & West R. (2011). Predictors of attempts to stop smoking and their success in adult general population samples: A systematic review . Addiction , , 2110–2121. doi: 10.1111/j.1360-0443.2011.03565.x [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Vink J. M., Willemsen G., & Boomsma D. I. (2005). Heritability of smoking initiation and nicotine dependence . Behavior Genetics , , 397–406. doi: 10.1007/s10519-004-1327-8 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• West R. (2009). The multiple facets of cigarette addiction and what they mean for encouraging and helping smokers to stop . Chronic Obstructive Pulmonary Disease , , 277–283. 10.1080/15412550903049181 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• West R., Baker C. L., Cappelleri J. C., & Bushmakin A. G. (2008). Effect of varenicline and bupropion SR on craving, nicotine withdrawal symptoms, and rewarding effects of smoking during a quit attempt . Psychopharmacology (Berl) , , 371–377. doi: 10.1007/s00213-007-1041-3 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• West R., & Brown J. (2015). Smoking in England 2007–2014 . Retrieved from www.smokinginengland.info/latest-statistics/ [ Google Scholar ]
• West R., Raw M., McNeill A., Stead L., Aveyard P., Bitton J., & Borland R. (2015). Health-care interventions to promote and assist tobacco cessation: A review of efficacy, effectiveness and affordability for use in national guideline development . Addiction , , 1388–1403. doi: 10.1111/add.12998 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• West R., & Shiffman S (2016). Smoking cessation (3rd ed.). Abingdon: Health Press. [ Google Scholar ]
• White A. R., Rampes H., Liu J. P., Stead L.F., & Campbell J (2014). Acupuncture and related interventions for smoking cessation . Cochrane Database of Systematic Reviews , , Cd000009. doi: 10.1002/14651858.CD000009.pub4 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Whittaker R., McRobbie H., Bullen C., Rodgers A., & Gu Y. (2016). Mobile phone-based interventions for smoking cessation . Cochrane Database of Systematic Reviews , , Cd006611. doi: 10.1002/14651858.CD006611.pub4 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• World Health Organization (2013). WHO report on the global tobacco epidemic, 2013: enforcing bans on tobacco advertising, promotion and sponsorship . Geneva: WHO. [ Google Scholar ]
• Wu L., Sun S., He Y., & Zeng J. (2015). Effect of smoking reduction therapy on smoking cessation for smokers without an intention to quit: An updated systematic review and meta-analysis of randomized controlled . International Journal of Environmental Research and Public Health , , 10235–10253. doi: 10.3390/ijerph120910235 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Zhou X., Nonnemaker J., Sherrill B., Gilsenan A. W., Coste F., & West R. (2009). Attempts to quit smoking and relapse: Factors associated with success or failure from the ATTEMPT cohort study . Addictive Behaviors , , 365–373. doi: 10.1016/j.addbeh.2008.11.013 [ PubMed ] [ CrossRef ] [ Google Scholar ]

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

• View all journals
• My Account Login
• Explore content
• About the journal
• Publish with us
• Sign up for alerts
• Open access
• Published: 10 October 2022

Health effects associated with smoking: a Burden of Proof study

• Xiaochen Dai   ORCID: orcid.org/0000-0002-0289-7814 1 , 2 ,
• Gabriela F. Gil 1 ,
• Marissa B. Reitsma 1 ,
• Noah S. Ahmad 1 ,
• Jason A. Anderson 1 ,
• Catherine Bisignano 1 ,
• Sinclair Carr 1 ,
• Rachel Feldman 1 ,
• Simon I. Hay   ORCID: orcid.org/0000-0002-0611-7272 1 , 2 ,
• Jiawei He 1 , 2 ,
• Vincent Iannucci 1 ,
• Hilary R. Lawlor 1 ,
• Matthew J. Malloy 1 ,
• Laurie B. Marczak 1 ,
• Susan A. McLaughlin 1 ,
• Larissa Morikawa   ORCID: orcid.org/0000-0001-9749-8033 1 ,
• Erin C. Mullany 1 ,
• Sneha I. Nicholson 1 ,
• Erin M. O’Connell 1 ,
• Chukwuma Okereke 1 ,
• Reed J. D. Sorensen 1 ,
• Joanna Whisnant 1 ,
• Aleksandr Y. Aravkin 1 , 3 ,
• Peng Zheng 1 , 2 ,
• Christopher J. L. Murray   ORCID: orcid.org/0000-0002-4930-9450 1 , 2 &
• Emmanuela Gakidou   ORCID: orcid.org/0000-0002-8992-591X 1 , 2  

Nature Medicine volume  28 ,  pages 2045–2055 ( 2022 ) Cite this article

25k Accesses

40 Citations

168 Altmetric

Metrics details

• Risk factors

Matters Arising to this article was published on 14 April 2023

As a leading behavioral risk factor for numerous health outcomes, smoking is a major ongoing public health challenge. Although evidence on the health effects of smoking has been widely reported, few attempts have evaluated the dose–response relationship between smoking and a diverse range of health outcomes systematically and comprehensively. In the present study, we re-estimated the dose–response relationships between current smoking and 36 health outcomes by conducting systematic reviews up to 31 May 2022, employing a meta-analytic method that incorporates between-study heterogeneity into estimates of uncertainty. Among the 36 selected outcomes, 8 had strong-to-very-strong evidence of an association with smoking, 21 had weak-to-moderate evidence of association and 7 had no evidence of association. By overcoming many of the limitations of traditional meta-analyses, our approach provides comprehensive, up-to-date and easy-to-use estimates of the evidence on the health effects of smoking. These estimates provide important information for tobacco control advocates, policy makers, researchers, physicians, smokers and the public.

Similar content being viewed by others

The Burden of Proof studies: assessing the evidence of risk

Health effects associated with exposure to secondhand smoke: a Burden of Proof study

Health effects associated with chewing tobacco: a Burden of Proof study

Among both the public and the health experts, smoking is recognized as a major behavioral risk factor with a leading attributable health burden worldwide. The health risks of smoking were clearly outlined in a canonical study of disease rates (including lung cancer) and smoking habits in British doctors in 1950 and have been further elaborated in detail over the following seven decades 1 , 2 . In 2005, evidence of the health consequences of smoking galvanized the adoption of the first World Health Organization (WHO) treaty, the Framework Convention on Tobacco Control, in an attempt to drive reductions in global tobacco use and second-hand smoke exposure 3 . However, as of 2020, an estimated 1.18 billion individuals globally were current smokers and 7 million deaths and 177 million disability-adjusted life-years were attributed to smoking, reflecting a persistent public health challenge 4 . Quantifying the relationship between smoking and various important health outcomes—in particular, highlighting any significant dose–response relationships—is crucial to understanding the attributable health risk experienced by these individuals and informing responsive public policy.

Existing literature on the relationship between smoking and specific health outcomes is prolific, including meta-analyses, cohort studies and case–control studies analyzing the risk of outcomes such as lung cancer 5 , 6 , 7 , chronic obstructive pulmonary disease (COPD) 8 , 9 , 10 and ischemic heart disease 11 , 12 , 13 , 14 due to smoking. There are few if any attempts, however, to systematically and comprehensively evaluate the landscape of evidence on smoking risk across a diverse range of health outcomes, with most current research focusing on risk or attributable burden of smoking for a specific condition 7 , 15 , thereby missing the opportunity to provide a comprehensive picture of the health risk experienced by smokers. Furthermore, although evidence surrounding specific health outcomes, such as lung cancer, has generated widespread consensus, findings about the attributable risk of other outcomes are much more heterogeneous and inconclusive 16 , 17 , 18 . These studies also vary in their risk definitions, with many comparing dichotomous exposure measures of ever smokers versus nonsmokers 19 , 20 . Others examine the distinct risks of current smokers and former smokers compared with never smokers 21 , 22 , 23 . Among the studies that do analyze dose–response relationships, there is large variation in the units and dose categories used in reporting their findings (for example, the use of pack-years or cigarettes per day) 24 , 25 , which complicates the comparability and consolidation of evidence. This, in turn, can obscure data that could inform personal health choices, public health practices and policy measures. Guidance on the health risks of smoking, such as the Surgeon General’s Reports on smoking 26 , 27 , is often based on experts’ evaluation of heterogenous evidence, which, although extremely useful and well suited to carefully consider nuances in the evidence, is fundamentally subjective.

The present study, as part of the Global Burden of Diseases, Risk Factors, and Injuries Study (GBD) 2020, re-estimated the continuous dose–response relationships (the mean risk functions and associated uncertainty estimates) between current smoking and 36 health outcomes (Supplementary Table 1 ) by identifying input studies using a systematic review approach and employing a meta-analytic method 28 . The 36 health outcomes that were selected based on existing evidence of a relationship included 16 cancers (lung cancer, esophageal cancer, stomach cancer, leukemia, liver cancer, laryngeal cancer, breast cancer, cervical cancer, colorectal cancer, lip and oral cavity cancer, nasopharyngeal cancer, other pharynx cancer (excluding nasopharynx cancer), pancreatic cancer, bladder cancer, kidney cancer and prostate cancer), 5 cardiovascular diseases (CVDs: ischemic heart disease, stroke, atrial fibrillation and flutter, aortic aneurysm and peripheral artery disease) and 15 other diseases (COPD, lower respiratory tract infections, tuberculosis, asthma, type 2 diabetes, Alzheimer’s disease and related dementias, Parkinson’s disease, multiple sclerosis, cataracts, gallbladder diseases, low back pain, peptic ulcer disease, rheumatoid arthritis, macular degeneration and fractures). Definitions of the outcomes are described in Supplementary Table 1 . We conducted a separate systematic review for each risk–outcome pair with the exception of cancers, which were done together in a single systematic review. This approach allowed us to systematically identify all relevant studies indexed in PubMed up to 31 May 2022, and we extracted relevant data on risk of smoking, including study characteristics, following a pre-specified template (Supplementary Table 2 ). The meta-analytic tool overcomes many of the limitations of traditional meta-analyses by incorporating between-study heterogeneity into the uncertainty of risk estimates, accounting for small numbers of studies, relaxing the assumption of log(linearity) applied to the risk functions, handling differences in exposure ranges between comparison groups, and systematically testing and adjusting for bias due to study designs and characteristics. We then estimated the burden-of-proof risk function (BPRF) for each risk–outcome pair, as proposed by Zheng et al. 29 ; the BPRF is a conservative risk function defined as the 5th quantile curve (for harmful risks) that reflects the smallest harmful effect at each level of exposure consistent with the available evidence. Given all available data for each outcome, the risk of smoking is at least as harmful as the BPRF indicates.

We used the BPRF for each risk–outcome pair to calculate risk–outcome scores (ROSs) and categorize the strength of evidence for the association between smoking and each health outcome using a star rating from 1 to 5. The interpretation of the star ratings is as follows: 1 star (*) indicates no evidence of association; 2 stars (**) correspond to a 0–15% increase in risk across average range of exposures for harmful risks; 3 stars (***) represent a 15–50% increase in risk; 4 stars (****) refer to >50–85% increase in risk; and 5 stars (*****) equal >85% increase in risk. The thresholds for each star rating were developed in consultation with collaborators and other stakeholders.

The increasing disease burden attributable to current smoking, particularly in low- and middle-income countries 4 , demonstrates the relevance of the present study, which quantifies the strength of the evidence using an objective, quantitative, comprehensive and comparative framework. Findings from the present study can be used to support policy makers in making informed smoking recommendations and regulations focusing on the associations for which the evidence is strongest (that is, the 4- and 5-star associations). However, associations with a lower star rating cannot be ignored, especially when the outcome has high prevalence or severity. A summary of the main findings, limitations and policy implications of the study is presented in Table 1 .

We evaluated the mean risk functions and the BPRFs for 36 health outcomes that are associated with current smoking 30 (Table 2 ). Following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines 31 for each of our systematic reviews, we identified studies reporting relative risk (RR) of incidence or mortality from each of the 36 selected outcomes for smokers compared with nonsmokers. We reviewed 21,108 records, which were identified to have been published between 1 May 2018 and 31 May 2022; this represents the most recent time period since the last systematic review of the available evidence for the GBD at the time of publication. The meta-analyses reported in the present study for each of the 36 health outcomes are based on evidence from a total of 793 studies published between 1970 and 2022 (Extended Data Fig. 1 – 5 and Supplementary Information 1.5 show the PRISMA diagrams for each outcome). Only prospective cohort and case–control studies were included for estimating dose–response risk curves, but cross-sectional studies were also included for estimating the age pattern of smoking risk on cardiovascular and circulatory disease (CVD) outcomes. Details on each, including the study’s design, data sources, number of participants, length of follow-up, confounders adjusted for in the input data and bias covariates included in the dose–response risk model, can be found in Supplementary Information 2 and 3 . The theoretical minimum risk exposure level used for current smoking was never smoking or zero 30 .

Five-star associations

When the most conservative interpretation of the evidence, that is, the BPRF, suggests that the average exposure (15th–85th percentiles of exposure) of smoking increases the risk of a health outcome by >85% (that is, ROS > 0.62), smoking and that outcome are categorized as a 5-star pair. Among the 36 outcomes, there are 5 that have a 5-star association with current smoking: laryngeal cancer (375% increase in risk based on the BPRF, 1.56 ROS), aortic aneurysm (150%, 0.92), peripheral artery disease (137%, 0.86), lung cancer (107%, 0.73) and other pharynx cancer (excluding nasopharynx cancer) (92%, 0.65).

Results for all 5-star risk–outcome pairs are available in Table 2 and Supplementary Information 4.1 . In the present study, we provide detailed results for one example 5-star association: current smoking and lung cancer. We extracted 371 observations from 25 prospective cohort studies and 53 case–control studies across 25 locations (Supplementary Table 3 ) 5 , 6 , 32 , 33 , 34 , 35 , 36 , 37 , 38 , 39 , 40 , 41 , 42 , 43 , 44 , 45 , 46 , 47 , 48 , 49 , 50 , 51 , 52 , 53 , 54 , 55 , 56 , 57 , 58 , 59 , 60 , 61 , 62 , 63 , 64 , 65 , 66 , 67 , 68 , 69 , 70 , 71 , 72 , 73 , 74 , 75 , 76 , 77 , 78 , 79 , 80 , 81 , 82 , 83 , 84 , 85 , 86 , 87 , 88 , 89 , 90 , 91 , 92 , 93 , 94 , 95 , 96 , 97 , 98 , 99 , 100 , 101 , 102 , 103 , 104 , 105 , 106 , 107 . Exposure ranged from 1 pack-year to >112 pack-years, with the 85th percentile of exposure being 50.88 pack-years (Fig. 1a ).

a , The log(RR) function. b , RR function. c , A modified funnel plot showing the residuals (relative to 0) on the x axis and the estimated s.d. that includes reported s.d. and between-study heterogeneity on the y axis.

We found a very strong and significant harmful relationship between pack-years of current smoking and the RR of lung cancer (Fig. 1b ). The mean RR of lung cancer at 20 pack-years of smoking was 5.11 (95% uncertainty interval (UI) inclusive of between-study heterogeneity = 1.84–14.99). At 50.88 pack-years (85th percentile of exposure), the mean RR of lung cancer was 13.42 (2.63–74.59). See Table 2 for mean RRs at other exposure levels. The BPRF, which represents the most conservative interpretation of the evidence (Fig. 1a ), suggests that smoking in the 15th–85th percentiles of exposure increases the risk of lung cancer by an average of 107%, yielding an ROS of 0.73.

The relationship between pack-years of current smoking and RR of lung cancer is nonlinear, with diminishing impact of further pack-years of smoking, particularly for middle-to-high exposure levels (Fig. 1b ). To reduce the effect of bias, we adjusted observations that did not account for more than five confounders, including age and sex, because they were the significant bias covariates identified by the bias covariate selection algorithm 29 (Supplementary Table 7 ). The reported RRs across studies were very heterogeneous. Our meta-analytic method, which accounts for the reported uncertainty in both the data and between-study heterogeneity, fit the data and covered the estimated residuals well (Fig. 1c ). After trimming 10% of outliers, we still detected publication bias in the results for lung cancer. See Supplementary Tables 4 and 7 for study bias characteristics and selected bias covariates, Supplementary Fig. 5 for results without 10% trimming and Supplementary Table 8 for observed RR data and alternative exposures across studies for the remaining 5-star pairs.

Four-star associations

When the BPRF suggests that the average exposure of smoking increases the risk of a health outcome by 50–85% (that is, ROS > 0.41–0.62), smoking is categorized as having a 4-star association with that outcome. We identified three outcomes with a 4-star association with smoking: COPD (72% increase in risk based on the BPRF, 0.54 ROS), lower respiratory tract infection (54%, 0.43) and pancreatic cancer (52%, 0.42).

In the present study, we provide detailed results for one example 4-star association: current smoking and COPD. We extracted 51 observations from 11 prospective cohort studies and 4 case–control studies across 36 locations (Supplementary Table 3 ) 6 , 8 , 9 , 10 , 78 , 108 , 109 , 110 , 111 , 112 , 113 , 114 , 115 , 116 . Exposure ranged from 1 pack-year to 100 pack-years, with the 85th percentile of exposure in the exposed group being 49.75 pack-years.

We found a strong and significant harmful relationship between pack-years of current smoking and RR of COPD (Fig. 2b ). The mean RR of COPD at 20 pack-years was 3.17 (1.60–6.55; Table 2 reports RRs at other exposure levels). At the 85th percentile of exposure, the mean RR of COPD was 6.01 (2.08–18.58). The BPRF suggests that average smoking exposure raises the risk of COPD by an average of 72%, yielding an ROS of 0.54. The results for the other health outcomes that have an association with smoking rated as 4 stars are shown in Table 2 and Supplementary Information 4.2 .

a , The log(RR) function. b , RR function. c , A modified funnel plot showing the residuals (relative to 0) on th e x axis and the estimated s.d. that includes the reported s.d. and between-study heterogeneity on the y axis.

The relationship between smoking and COPD is nonlinear, with diminishing impact of further pack-years of current smoking on risk of COPD, particularly for middle-to-high exposure levels (Fig. 2a ). To reduce the effect of bias, we adjusted observations that did not account for age and sex and/or were generated for individuals aged >65 years 116 , because they were the two significant bias covariates identified by the bias covariate selection algorithm (Supplementary Table 7 ). There was large heterogeneity in the reported RRs across studies, and our meta-analytic method fit the data and covered the estimated residuals well (Fig. 2b ). Although we trimmed 10% of outliers, publication bias was still detected in the results for COPD. See Supplementary Tables 4 and 7 for study bias characteristics and selected bias covariates, Supplementary Fig. 5 for results without 10% trimming and Supplementary Table 8 for reported RR data and alternative exposures across studies for the remaining health outcomes that have a 4-star association with smoking.

Three-star associations

When the BPRF suggests that the average exposure of smoking increases the risk of a health outcome by 15–50% (or, when protective, decreases the risk of an outcome by 13–34%; that is, ROS >0.14–0.41), the association between smoking and that outcome is categorized as having a 3-star rating. We identified 15 outcomes with a 3-star association: bladder cancer (40% increase in risk, 0.34 ROS); tuberculosis (31%, 0.27); esophageal cancer (29%, 0.26); cervical cancer, multiple sclerosis and rheumatoid arthritis (each 23–24%, 0.21); lower back pain (22%, 0.20); ischemic heart disease (20%, 0.19); peptic ulcer and macular degeneration (each 19–20%, 0.18); Parkinson's disease (protective risk, 15% decrease in risk, 0.16); and stomach cancer, stroke, type 2 diabetes and cataracts (each 15–17%, 0.14–0.16).

We present the findings on smoking and type 2 diabetes as an example of a 3-star risk association. We extracted 102 observations from 24 prospective cohort studies and 4 case–control studies across 15 locations (Supplementary Table 3 ) 117 , 118 , 119 , 120 , 121 , 122 , 123 , 124 , 125 , 126 , 127 , 128 , 129 , 130 , 131 , 132 , 133 , 134 , 135 , 136 , 137 , 138 , 139 , 140 , 141 , 142 , 143 , 144 . The exposure ranged from 1 cigarette to 60 cigarettes smoked per day, with the 85th percentile of exposure in the exposed group being 26.25 cigarettes smoked per day.

We found a moderate and significant harmful relationship between cigarettes smoked per day and the RR of type 2 diabetes (Fig. 3b ). The mean RR of type 2 diabetes at 20 cigarettes smoked per day was 1.49 (1.18–1.90; see Table 2 for other exposure levels). At the 85th percentile of exposure, the mean RR of type 2 diabetes was 1.54 (1.20–2.01). The BPRF suggests that average smoking exposure raises the risk of type 2 diabetes by an average of 16%, yielding an ROS of 0.15. See Table 2 and Supplementary Information 4.3 for results for the additional health outcomes with an association with smoking rated as 3 stars.

a , The log(RR) function. b , RR function. c , A modified funnel plot showing the residuals (relative to 0) on the x axis and the estimated s.d. that includes the reported s.d. and between-study heterogeneity on the y axis.

The relationship between smoking and type 2 diabetes is nonlinear, particularly for high exposure levels where the mean risk curve becomes flat (Fig. 3a ). We adjusted observations that were generated in subpopulations, because it was the only significant bias covariate identified by the bias covariate selection algorithm (Supplementary Table 7 ). There was moderate heterogeneity in the observed RR data across studies and our meta-analytic method fit the data and covered the estimated residuals extremely well (Fig. 3b,c ). After trimming 10% of outliers, we still detected publication bias in the results for type 2 diabetes. See Supplementary Tables 4 and 7 for study bias characteristics and selected bias covariates, Supplementary Fig. 5 for results without 10% trimming and Supplementary Table 8 for observed RR data and alternative exposures across studies for the remaining 3-star pairs.

Two-star associations

When the BPRF suggests that the average exposure of smoking increases the risk of an outcome by 0–15% (that is, ROS 0.0–0.14), the association between smoking and that outcome is categorized as a 2-star rating. We identified six 2-star outcomes: nasopharyngeal cancer (14% increase in risk, 0.13 ROS); Alzheimer’s and other dementia (10%, 0.09); gallbladder diseases and atrial fibrillation and flutter (each 6%, 0.06); lip and oral cavity cancer (5%, 0.05); and breast cancer (4%, 0.04).

We present the findings on smoking and breast cancer as an example of a 2-star association. We extracted 93 observations from 14 prospective cohort studies and 9 case–control studies across 14 locations (Supplementary Table 3 ) 84 , 87 , 145 , 146 , 147 , 148 , 149 , 150 , 151 , 152 , 153 , 154 , 155 , 156 , 157 , 158 , 159 , 160 , 161 , 162 , 163 , 164 , 165 . The exposure ranged from 1 cigarette to >76 cigarettes smoked per day, with the 85th percentile of exposure in the exposed group being 34.10 cigarettes smoked per day.

We found a weak but significant relationship between pack-years of current smoking and RR of breast cancer (Extended Data Fig. 6 ). The mean RR of breast cancer at 20 pack-years was 1.17 (1.04–1.31; Table 2 reports other exposure levels). The BPRF suggests that average smoking exposure raises the risk of breast cancer by an average of 4%, yielding an ROS of 0.04. See Table 2 and Supplementary Information 4.4 for results on the additional health outcomes for which the association with smoking has been categorized as 2 stars.

The relationship between smoking and breast cancer is nonlinear, particularly for high exposure levels where the mean risk curve becomes flat (Extended Data Fig. 6a ). To reduce the effect of bias, we adjusted observations that were generated in subpopulations, because it was the only significant bias covariate identified by the bias covariate selection algorithm (Supplementary Table 7 ). There was heterogeneity in the reported RRs across studies, but our meta-analytic method fit the data and covered the estimated residuals (Extended Data Fig. 6b ). After trimming 10% of outliers, we did not detect publication bias in the results for breast cancer. See Supplementary Tables 4 and 7 for study bias characteristics and selected bias covariates, Supplementary Fig. 5 for results without 10% trimming and Supplementary Table 8 for observed RR data and alternative exposures across studies for the remaining 2-star pairs.

One-star associations

When average exposure to smoking does not significantly increase (or decrease) the risk of an outcome, once between-study heterogeneity and other sources of uncertainty are accounted for (that is, ROS < 0), the association between smoking and that outcome is categorized as 1 star, indicating that there is not sufficient evidence for the effect of smoking on the outcome to reject the null (that is, there may be no association). There were seven outcomes with an association with smoking that rated as 1 star: colorectal and kidney cancer (each –0.01 ROS); leukemia (−0.04); fractures (−0.05); prostate cancer (−0.06); liver cancer (−0.32); and asthma (−0.64).

We use smoking and prostate cancer as examples of a 1-star association. We extracted 78 observations from 21 prospective cohort studies and 1 nested case–control study across 15 locations (Supplementary Table 3 ) 157 , 160 , 166 , 167 , 168 , 169 , 170 , 171 , 172 , 173 , 174 , 175 , 176 , 177 , 178 , 179 , 180 , 181 , 182 , 183 , 184 , 185 . The exposure among the exposed group ranged from 1 cigarette to 90 cigarettes smoked per day, with the 85th percentile of exposure in the exposed group being 29.73 cigarettes smoked per day.

Based on our conservative interpretation of the data, we did not find a significant relationship between cigarettes smoked per day and the RR of prostate cancer (Fig. 4B ). The exposure-averaged BPRF for prostate cancer was 0.94, which was opposite null from the full range of mean RRs, such as 1.16 (0.89–1.53) at 20 cigarettes smoked per day. The corresponding ROS was −0.06, which is consistent with no evidence of an association between smoking and increased risk of prostate cancer. See Table 2 and Supplementary Information 4.5 for results for the additional outcomes that have a 1-star association with smoking.

The relationship between smoking and prostate cancer is nonlinear, particularly for middle-to-high exposure levels where the mean risk curve becomes flat (Fig. 4a ). We did not adjust for any bias covariate because no significant bias covariates were selected by the algorithm (Supplementary Table 7 ). The RRs reported across studies were very heterogeneous, but our meta-analytic method fit the data and covered the estimated residuals well (Fig. 4b,c ). The ROS associated with the BPRF is −0.05, suggesting that the most conservative interpretation of all evidence, after accounting for between-study heterogeneity, indicates an inconclusive relationship between smoking exposure and the risk of prostate cancer. After trimming 10% of outliers, we still detected publication bias in the results for prostate cancer, which warrants further studies using sample populations. See Supplementary Tables 4 and 7 for study bias characteristics and selected bias covariates, Supplementary Fig. 5 for results without 10% trimming and Supplementary Table 8 for observed RR data and alternative exposures across studies for the remaining 1-star pairs.

Age-specific dose–response risk for CVD outcomes

We produced age-specific dose–response risk curves for the five selected CVD outcomes ( Methods ). The ROS associated with each smoking–CVD pair was calculated based on the reference risk curve estimated using all risk data regardless of age information. Estimation of the BPRF, calculation of the associated ROS and star rating of the smoking–CVD pairs follow the same rules as the other non-CVD smoking–outcome pairs (Table 1 and Supplementary Figs. 2 – 4 ). Once we had estimated the reference dose–response risk curve for each CVD outcome, we determined the age group of the reference risk curve. The reference age group is 55–59 years for all CVD outcomes, except for peripheral artery disease, the reference age group for which is 60–64 years. We then estimated the age pattern of smoking on all CVD outcomes (Supplementary Fig. 2 ) and calculated age attenuation factors of the risk for each age group by comparing the risk of each age group with that of the reference age group, using the estimated age pattern (Supplementary Fig. 3 ). Last, we applied the draws of age attenuation factors of each age group to the dose–response risk curve for the reference age group to produce the age group-specific dose–response risk curves for each CVD outcome (Supplementary Fig. 4 ).

Using our burden-of-proof meta-analytic methods, we re-estimated the dose–response risk of smoking on 36 health outcomes that had previously been demonstrated to be associated with smoking 30 , 186 . Using these methods, which account for both the reported uncertainty of the data and the between-study heterogeneity, we found that 29 of the 36 smoking–outcome pairs are supported by evidence that suggests a significant dose–response relationship between smoking and the given outcome (28 with a harmful association and 1 with a protective association). Conversely, after accounting for between-study heterogeneity, the available evidence of smoking risk on seven outcomes (that is, colon and rectum cancer, kidney cancer, leukemia, prostate cancer, fractures, liver cancer and asthma) was insufficient to reject the null or draw definitive conclusions on their relationship to smoking. Among the 29 outcomes that have evidence supporting a significant relationship to smoking, 8 had strong-to-very-strong evidence of a relationship, meaning that, given all the available data on smoking risk, we estimate that average exposure to smoking increases the risk of those outcomes by >50% (4- and 5-star outcomes). The currently available evidence for the remaining 21 outcomes with a significant association with current smoking was weak to moderate, indicating that smoking increases the risk of those outcomes by at least >0–50% (2- and 3-star associations).

Even under our conservative interpretation of the data, smoking is irrefutably harmful to human health, with the greatest increases in risk occurring for laryngeal cancer, aortic aneurysm, peripheral artery disease, lung cancer and other pharynx cancer (excluding nasopharynx cancer), which collectively represent large causes of death and ill-health. The magnitude of and evidence for the associations between smoking and its leading health outcomes are among the highest currently analyzed in the burden-of-proof framework 29 . The star ratings assigned to each smoking–outcome pair offer policy makers a way of categorizing and comparing the evidence for a relationship between smoking and its potential health outcomes ( https://vizhub.healthdata.org/burden-of-proof ). We found that, for seven outcomes in our analysis, there was insufficient or inconsistent evidence to demonstrate a significant association with smoking. This is a key finding because it demonstrates the need for more high-quality data for these particular outcomes; availability of more data should improve the strength of evidence for whether or not there is an association between smoking and these health outcomes.

Our systematic review approach and meta-analytic methods have numerous benefits over existing systematic reviews and meta-analyses on the same topic that use traditional random effects models. First, our approach relaxes the log(linear) assumption, using a spline ensemble to estimate the risk 29 . Second, our approach allows variable reference groups and exposure ranges, allowing for more accurate estimates regardless of whether or not the underlying relative risk is log(linear). Furthermore, it can detect outliers in the data automatically. Finally, it quantifies uncertainty due to between-study heterogeneity while accounting for small numbers of studies, minimizing the risk that conclusions will be drawn based on spurious findings.

We believe that the results for the association between smoking and each of the 36 health outcomes generated by the present study, including the mean risk function, BPRF, ROS, average excess risk and star rating, could be useful to a range of stakeholders. Policy makers can formulate their decisions on smoking control priorities and resource allocation based on the magnitude of the effect and the consistency of the evidence relating smoking to each of the 36 outcomes, as represented by the ROS and star rating for each smoking–outcome association 187 . Physicians and public health practitioners can use the estimates of average increased risk and the star rating to educate patients and the general public about the risk of smoking and to promote smoking cessation 188 . Researchers can use the estimated mean risk function or BPRF to obtain the risk of an outcome at a given smoking exposure level, as well as uncertainty surrounding that estimate of risk. The results can also be used in the estimation of risk-attributable burden, that is, the deaths and disability-adjusted life-years due to each outcome that are attributable to smoking 30 , 186 . For the general public, these results could help them to better understand the risk of smoking and manage their health 189 .

Although our meta-analysis was comprehensive and carefully conducted, there are limitations to acknowledge. First, the bias covariates used, although carefully extracted and evaluated, were based on observable study characteristics and thus may not fully capture unobserved characteristics such as study quality or context, which might be major sources of bias. Second, if multiple risk estimates with different adjustment levels were reported in a given study, we included only the fully adjusted risk estimate and modeled the adjustment level according to the number of covariates adjusted for (rather than which covariates were adjusted for) and whether a standard adjustment for age and sex had been applied. This approach limited our ability to make full use of all available risk estimates in the literature. Third, although we evaluated the potential for publication bias in the data, we did not test for other forms of bias such as when studies are more consistent with each other than expected by chance 29 . Fourth, our analysis assumes that the relationships between smoking and health outcomes are similar across geographical regions and over time. We do not have sufficient evidence to quantify how the relationships may have evolved over time because the composition of smoking products has also changed over time. Perhaps some of the heterogeneity of the effect sizes in published studies reflects this; however, this cannot be discerned with the currently available information.

In the future, we plan to include crude and partially adjusted risk estimates in our analyses to fully incorporate all available risk estimates, to model the adjusted covariates in a more comprehensive way by mapping the adjusted covariates across all studies comprehensively and systematically, and to develop methods to evaluate additional forms of potential bias. We plan to update our results on a regular basis to provide timely and up-to-date evidence to stakeholders.

To conclude, we have re-estimated the dose–response risk of smoking on 36 health outcomes while synthesizing all the available evidence up to 31 May 2022. We found that, even after factoring in the heterogeneity between studies and other sources of uncertainty, smoking has a strong-to-very-strong association with a range of health outcomes and confirmed that smoking is irrefutably highly harmful to human health. We found that, due to small numbers of studies, inconsistency in the data, small effect sizes or a combination of these reasons, seven outcomes for which some previous research had found an association with smoking did not—under our meta-analytic framework and conservative approach to interpreting the data—have evidence of an association. Our estimates of the evidence for risk of smoking on 36 selected health outcomes have the potential to inform the many stakeholders of smoking control, including policy makers, researchers, public health professionals, physicians, smokers and the general public.

For the present study, we used a meta-analytic tool, MR-BRT (metaregression—Bayesian, regularized, trimmed), to estimate the dose–response risk curves of the risk of a health outcome across the range of current smoking levels along with uncertainty estimates 28 . Compared with traditional meta-analysis using linear mixed effect models, MR-BRT relaxes the assumption of a log(linear) relationship between exposure and risk, incorporates between-study heterogeneity into the uncertainty of risk estimates, handles estimates reported across different exposure categories, automatically identifies and trims outliers, and systematically tests and adjusts for bias due to study designs and characteristics. The meta-analytic methods employed by the present study followed the six main steps proposed by Zheng et al. 28 , 29 , namely: (1) enacting a systematic review approach and data extraction following a pre-specified and standardized protocol; (2) estimating the shape of the relationship between exposure and RR; (3) evaluating and adjusting for systematic bias as a function of study characteristics and risk estimation; (4) quantifying between-study heterogeneity while adjusting for within-study correlation and the number of studies; (5) evaluating potential publication or reporting biases; and (6) estimating the mean risk function and the BPRF, calculating the ROS and categorizing smoking–outcome pairs using a star-rating scheme from 1 to 5.

The estimates for our primary indicators of this work—mean RRs across a range of exposures, BRPFs, ROSs and star ratings for each risk–outcome pair—are not specific to or disaggregated by specific populations. We did not estimate RRs separately for different locations, sexes (although the RR of prostate cancer was estimated only for males and of cervical and breast cancer only for females) or age groups (although this analysis was applied to disease endpoints in adults aged ≥30 years only and, as detailed below, age-specific estimates were produced for the five CVD outcomes).

The present study complies with the PRISMA guidelines 190 (Supplementary Tables 9 and 10 and Supplementary Information 1.5 ) and Guidelines for Accurate and Transparent Health Estimates Reporting (GATHER) recommendations 191 (Supplementary Table 11 ). The study was approved by the University of Washington Institutional Review Board (study no. 9060). The systematic review approach was not registered.

Selecting health outcomes

In the present study, current smoking is defined as the current use of any smoked tobacco product on a daily or occasional basis. Health outcomes were initially selected using the World Cancer Research Fund criteria for convincing or probable evidence as described in Murray et al. 186 . The 36 health outcomes that were selected based on existing evidence of a relationship included 16 cancers (lung cancer, esophageal cancer, stomach cancer, leukemia, liver cancer, laryngeal cancer, breast cancer, cervical cancer, colorectal cancer, lip and oral cavity cancer, nasopharyngeal cancer, other pharynx cancer (excluding nasopharynx cancer), pancreatic cancer, bladder cancer, kidney cancer and prostate cancer), 5 CVDs (ischemic heart disease, stroke, atrial fibrillation and flutter, aortic aneurysm and peripheral artery disease) and 15 other diseases (COPD, lower respiratory tract infections, tuberculosis, asthma, type 2 diabetes, Alzheimer’s disease and related dementias, Parkinson’s disease, multiple sclerosis, cataracts, gallbladder diseases, low back pain, peptic ulcer disease, rheumatoid arthritis, macular degeneration and fracture). Definitions of the outcomes are described in Supplementary Table 1 .

Step 1: systematic review approach to literature search and data extraction

Informed by the systematic review approach we took for the GBD 2019 (ref. 30 ), for the present study we identified input studies in the literature using a systematic review approach for all 36 smoking–outcome pairs using updated search strings to identify all relevant studies indexed in PubMed up to 31 May 2022 and extracted data on smoking risk estimates. Briefly, the studies that were extracted represented several types of study design (for example, cohort and case–control studies), measured exposure in several different ways and varied in their choice of reference categories (where some compared current smokers with never smokers, whereas others compared current smokers with nonsmokers or former smokers). All these study characteristics were catalogued systematically and taken into consideration during the modeling part of the analysis.

In addition, for CVD outcomes, we also estimated the age pattern of risk associated with smoking. We applied a systematic review of literature approach for smoking risk for the five CVD outcomes. We developed a search string to search for studies reporting any association between binary smoking status (that is, current, former and ever smokers) and the five CVD outcomes from 1 January 1970 to 31 May 2022, and included only studies reporting age-specific risk (RR, odds ratio (OR), hazard ratio (HR)) of smoking status. The inclusion criteria and results of the systematic review approach are reported in accordance with PRISMA guidelines 31 . Details for each outcome on the search string used in the systematic review approach, refined inclusion and exclusion criteria, data extraction template and PRISMA diagram are given in Supplementary Information 1 . Title and/or abstract screening, full text screening and data extraction were conducted by 14 members of the research team and extracted data underwent manual quality assurance by the research team to verify accuracy.

Selecting exposure categories

Cumulative exposure in pack-years was the measure of exposure used for COPD and all cancer outcomes except for prostate cancer, to reflect the risk of both duration and intensity of current smoking on these outcomes. For prostate cancer, CVDs and all the other outcomes except for fractures, we used cigarette-equivalents smoked per day as the exposure for current smoking, because smoking intensity is generally thought to be more important than duration for these outcomes. For fractures, we used binary exposure, because there were few studies examining intensity or duration of smoking on fractures. The smoking–outcome pairs and the corresponding exposures are summarized in Supplementary Table 4 and are congruent with the GBD 2019 (refs. 30 , 186 ).

Steps 2–5: modeling dose–response RR of smoking on the selected health outcomes

Of the six steps proposed by Zheng et al. 29 , steps 2–5 cover the process of modeling dose–response risk curves. In step 2, we estimated the shape (or the ‘signal’) of the dose–response risk curves, integrating over different exposure ranges. To relax the log(linear) assumption usually applied to continuous dose–response risk and make the estimates robust to the placement of spline knots, we used an ensemble spline approach to fit the functional form of the dose–response relationship. The final ensemble model was a weighted combination of 50 models with random knot placement, with the weight of each model proportional to measures of model fit and total variation. To avoid the influence of extreme data and reduce publication bias, we trimmed 10% of data for each outcome as outliers. We also applied a monotonicity constraint to ensure that the mean risk curves were nondecreasing (or nonincreasing in the case of Parkinson’s disease).

In step 3, following the GRADE approach 192 , 193 , we quantified risk of bias across six domains, namely, representativeness of the study population, exposure, outcome, reverse causation, control for confounding and selection bias. Details about the bias covariates are provided in Supplementary Table 4 . We systematically tested for the effect of bias covariates using metaregression, selected significant bias covariates using the Lasso approach 194 , 195 and adjusted for the selected bias covariates in the final risk curve.

In step 4, we quantified between-study heterogeneity accounting for within-study correlation, uncertainty of the heterogeneity, as well as small number of studies. Specifically, we used a random intercept in the mixed-effects model to account for the within-study correlation and used a study-specific random slope with respect to the ‘signal’ to capture between-study heterogeneity. As between-study heterogeneity can be underestimated or even zero when the number of studies is small 196 , 197 , we used Fisher’s information matrix to estimate the uncertainty of the heterogeneity 198 and incorporated that uncertainty into the final results.

In step 5, in addition to generating funnel plots and visually inspecting for asymmetry (Figs. 1c , 2c , 3c and 4c and Extended Data Fig. 6c ) to identify potential publication bias, we also statistically tested for potential publication or reporting bias using Egger’s regression 199 . We flagged potential publication bias in the data but did not correct for it, which is in line with the general literature 10 , 200 , 201 . Full details about the modeling process have been published elsewhere 29 and model specifications for each outcome are in Supplementary Table 6 .

Step 6: estimating the mean risk function and the BPRF

In the final step, step 6, the metaregression model inclusive of the selected bias covariates from step 3 (for example, the highest adjustment level) was used to predict the mean risk function and its 95% UI, which incorporated the uncertainty of the mean effect, between-study heterogeneity and the uncertainty in the heterogeneity estimate accounting for small numbers of studies. Specifically, 1,000 draws were created for each 0.1 level of doses from 0 pack-years to 100 pack-years or cigarette-equivalents smoked per day using the Bayesian metaregression model. The mean of the 1,000 draws was used to estimate the mean risk at each exposure level, and the 25th and 95th draws were used to estimate the 95% UIs for the mean risk at each exposure level.

The BPRF 29 is a conservative estimate of risk function consistent with the available evidence, correcting for both between-study heterogeneity and systemic biases related to study characteristics. The BPRF is defined as either the 5th (if harmful) or 95th (if protective) quantile curve closest to the line of log(RR) of 0, which defines the null (Figs. 1a , 2b , 3a and 4a ). The BPRF represents the smallest harmful (or protective) effect of smoking on the corresponding outcome at each level of exposure that is consistent with the available evidence. A BPRF opposite null from the mean risk function indicates that insufficient evidence is available to reject null, that is, that there may not be an association between risk and outcome. Likewise, the further the BPRF is from null on the same side of null as the mean risk function, the higher the magnitude and evidence for the relationship. The BPRF can be interpreted as indicating that, even accounting for between-study heterogeneity and its uncertainty, the log(RR) across the studied smoking range is at least as high as the BPRF (or at least as low as the BPRF for a protective risk).

To quantify the strength of the evidence, we calculated the ROS for each smoking–outcome association as the signed value of the log(BPRF) averaged between the 15th and 85th percentiles of observed exposure levels for each outcome. The ROS is a single summary of the effect of smoking on the outcome, with higher positive ROSs corresponding to stronger and more consistent evidence and a higher average effect size of smoking and a negative ROS, suggesting that, based on the available evidence, there is no significant effect of smoking on the outcome after accounting for between-study heterogeneity.

For ease of communication, we further classified each smoking–outcome association into a star rating from 1 to 5. Briefly, 1-star associations have an ROS <0, indicating that there is insufficient evidence to find a significant association between smoking and the selected outcome. We divided the positive ROSs into ranges 0.0–0.14 (2-star), >0.14–0.41 (3-star), >0.41–0.62 (4-star) and >0.62 (5-star). These categories correspond to excess risk ranges for harmful risks of 0–15%, >15–50%, >50–85% and >85%. For protective risks, the ranges of exposure-averaged decreases in risk by star rating are 0–13% (2 stars), >13–34% (3 stars), >34–46% (4 stars) and >46% (5 stars).

Among the 36 smoking–outcome pairs analyzed, smoking fracture was the only binary risk–outcome pair, which was due to limited data on the dose–response risk of smoking on fracture 202 . The estimation of binary risk was simplified because the RR was merely a comparison between current smokers and nonsmokers or never smokers. The concept of ROS for continuous risk can naturally extend to binary risk because the BPRF is still defined as the 5th percentile of the effect size accounting for data uncertainty and between-study heterogeneity. However, binary ROSs must be divided by 2 to make them comparable with continuous ROSs, which were calculated by averaging the risk over the range between the 15th and the 85th percentiles of observed exposure levels. Full details about estimating mean risk functions, BPRFs and ROSs for both continuous and binary risk–outcome pairs can be found elsewhere 29 .

Estimating the age-specific risk function for CVD outcomes

For non-CVD outcomes, we assumed that the risk function was the same for all ages and all sexes, except for breast, cervical and prostate cancer, which were assumed to apply only to females or males, respectively. As the risk of smoking on CVD outcomes is known to attenuate with increasing age 203 , 204 , 205 , 206 , we adopted a four-step approach for GBD 2020 to produce age-specific dose–response risk curves for CVD outcomes.

First, we estimated the reference dose–response risk of smoking for each CVD outcome using dose-specific RR data for each outcome regardless of the age group information. This step was identical to that implemented for the other non-CVD outcomes. Once we had generated the reference curve, we determined the age group associated with it by calculating the weighted mean age across all dose-specific RR data (weighted by the reciprocal of the s.e.m. of each datum). For example, if the weighted mean age of all dose-specific RR data was 56.5, we estimated the age group associated with the reference risk curve to be aged 55–59 years. For cohort studies, the age range associated with the RR estimate was calculated as a mean age at baseline plus the mean/median years of follow-up (if only the maximum years of follow-up were reported, we would halve this value and add it to the mean age at baseline). For case–control studies, the age range associated with the OR estimate was simply the reported mean age at baseline (if mean age was not reported, we used the midpoint of the age range instead).

In the third step, we extracted age group-specific RR data and relevant bias covariates from the studies identified in our systematic review approach of age-specific smoking risk on CVD outcomes, and used MR-BRT to model the age pattern of excess risk (that is, RR-1) of smoking on CVD outcomes with age group-specific excess RR data for all CVD outcomes. We modeled the age pattern of smoking risk on CVDs following the same steps we implemented for modeling dose–response risk curves. In the final model, we included a spline on age, random slope on age by study and the bias covariate encoding exposure definition (that is, current, former and ever smokers), which was picked by the variable selection algorithm 28 , 29 . When predicting the age pattern of the excess risk of smoking on CVD outcomes using the fitted model, we did not include between-study heterogeneity to reduce uncertainty in the prediction.

In the fourth step, we calculated the age attenuation factors of excess risk compared with the reference age group for each CVD outcome as the ratio of the estimated excess risk for each age group to the excess risk for the reference age group. We performed the calculation at the draw level to obtain 1,000 draws of the age attenuation factors for each age group. Once we had estimated the age attenuation factors, we carried out the last step, which consisted of adjusting the risk curve for the reference age group from step 1 using equation (1) to produce the age group-specific risk curves for each CVD outcome:

We implemented the age adjustment at the draw level so that the uncertainty of the age attenuation factors could be naturally incorporated into the final adjusted age-specific RR curves. A PRISMA diagram detailing the systematic review approach, a description of the studies included and the full details about the methods are in Supplementary Information 1.5 and 5.2 .

Estimating the theoretical minimum risk exposure level

The theoretical minimum risk exposure level for smoking was 0, that is, no individuals in the population are current or former smokers.

Model validation

The validity of the meta-analytic tool has been extensively evaluated by Zheng and colleagues using simulation experiments 28 , 29 . For the present study, we conducted two additional sensitivity analyses to examine how the shape of the risk curves was impacted by applying a monotonicity constraint and trimming 10% of data. We present the results of these sensitivity analyses in Supplementary Information 6 . In addition to the sensitivity analyses, the dose–response risk estimates were also validated by plotting the mean risk function along with its 95% UI against both the extracted dose-specific RR data from the studies included and our previous dose–response risk estimates from the GBD 2019 (ref. 30 ). The mean risk functions along with the 95% UIs were validated based on data fit and the level, shape and plausibility of the dose–response risk curves. All curves were validated by all authors and reviewed by an external expert panel, comprising professors with relevant experience from universities including Johns Hopkins University, Karolinska Institute and University of Barcelona; senior scientists working in relevant departments at the WHO and the Center for Disease Control and Prevention (CDC) and directors of nongovernmental organizations such as the Campaign for Tobacco-Free Kids.

Statistical analysis

Analyses were carried out using R v.3.6.3, Python v.3.8 and Stata v.16.

Statistics and reproducibility

The study was a secondary analysis of existing data involving systematic reviews and meta-analyses. No statistical method was used to predetermine sample size. As the study did not involve primary data collection, randomization and blinding, data exclusions were not relevant to the present study, and, as such, no data were excluded and we performed no randomization or blinding. We have made our data and code available to foster reproducibility.

Reporting summary

Further information on research design is available in the Nature Research Reporting Summary linked to this article.

Data availability

The findings from the present study are supported by data available in the published literature. Data sources and citations for each risk–outcome pair can be downloaded using the ‘download’ button on each risk curve page currently available at https://vizhub.healthdata.org/burden-of-proof . Study characteristics and citations for all input data used in the analyses are also provided in Supplementary Table 3 , and Supplementary Table 2 provides a template of the data collection form.

Code availability

All code used for these analyses is publicly available online ( https://github.com/ihmeuw-msca/burden-of-proof ).

Doll, R. & Hill, A. B. Smoking and carcinoma of the lung. Br. Med. J. 2 , 739–748 (1950).

Article   CAS   PubMed   PubMed Central   Google Scholar  

Di Cicco, M. E., Ragazzo, V. & Jacinto, T. Mortality in relation to smoking: the British Doctors Study. Breathe 12 , 275–276 (2016).

Article   PubMed   PubMed Central   Google Scholar  

World Health Organization. WHO Framework Convention on Tobacco Control 36 (WHO, 2003).

Dai, X., Gakidou, E. & Lopez, A. D. Evolution of the global smoking epidemic over the past half century: strengthening the evidence base for policy action. Tob. Control 31 , 129–137 (2022).

Article   PubMed   Google Scholar  

Dikshit, R. P. & Kanhere, S. Tobacco habits and risk of lung, oropharyngeal and oral cavity cancer: a population-based case-control study in Bhopal, India. Int. J. Epidemiol. 29 , 609–614 (2000).

Article   CAS   PubMed   Google Scholar  

Liaw, K. M. & Chen, C. J. Mortality attributable to cigarette smoking in Taiwan: a 12-year follow-up study. Tob. Control 7 , 141–148 (1998).

Gandini, S. et al. Tobacco smoking and cancer: a meta-analysis. Int. J. Cancer 122 , 155–164 (2008).

Deng, X., Yuan, C. & Chang, D. Interactions between single nucleotide polymorphism of SERPINA1 gene and smoking in association with COPD: a case–control study. Int. J. Chron. Obstruct. Pulmon. Dis. 12 , 259–265 (2017).

Leem, A. Y., Park, B., Kim, Y. S., Jung, J. Y. & Won, S. Incidence and risk of chronic obstructive pulmonary disease in a Korean community-based cohort. Int. J. Chron. Obstruct. Pulmon. Dis. 13 , 509–517 (2018).

Forey, B. A., Thornton, A. J. & Lee, P. N. Systematic review with meta-analysis of the epidemiological evidence relating smoking to COPD, chronic bronchitis and emphysema. BMC Pulmon. Med. 11 , 36 (2011).

Article   Google Scholar  

Tan, J. et al. Smoking, blood pressure, and cardiovascular disease mortality in a large cohort of Chinese men with 15 years follow-up. Int. J. Environ. Res. Public Health 15 , E1026 (2018).

Doll, R., Peto, R., Boreham, J. & Sutherland, I. Mortality in relation to smoking: 50 years’ observations on male British doctors. Br. Med. J. 328 , 1519 (2004).

Huxley, R. R. & Woodward, M. Cigarette smoking as a risk factor for coronary heart disease in women compared with men: a systematic review and meta-analysis of prospective cohort studies. Lancet 378 , 1297–1305 (2011).

Hbejan, K. Smoking effect on ischemic heart disease in young patients. Heart Views 12 , 1–6 (2011).

Chao, H. et al. A meta-analysis of active smoking and risk of meningioma. Tob. Induc. Dis. 19 , 34 (2021).

Shi, H., Shao, X. & Hong, Y. Association between cigarette smoking and the susceptibility of acute myeloid leukemia: a systematic review and meta-analysis. Eur. Rev. Med Pharm. Sci. 23 , 10049–10057 (2019).

CAS   Google Scholar  

Macacu, A., Autier, P., Boniol, M. & Boyle, P. Active and passive smoking and risk of breast cancer: a meta-analysis. Breast Cancer Res. Treat. 154 , 213–224 (2015).

Pujades-Rodriguez, M. et al. Heterogeneous associations between smoking and a wide range of initial presentations of cardiovascular disease in 1 937 360 people in England: lifetime risks and implications for risk prediction. Int. J. Epidemiol. 44 , 129–141 (2015).

Kanazir, M. et al. Risk factors for hepatocellular carcinoma: a case-control study in Belgrade (Serbia). Tumori 96 , 911–917 (2010).

Pytynia, K. B. et al. Matched-pair analysis of survival of never smokers and ever smokers with squamous cell carcinoma of the head and neck. J. Clin. Oncol. 22 , 3981–3988 (2004).

Barengo, N. C., Antikainen, R., Harald, K. & Jousilahti, P. Smoking and cancer, cardiovascular and total mortality among older adults: the Finrisk Study. Prev. Med. Rep. 14 , 100875 (2019).

Guo, Y. et al. Modifiable risk factors for cognitive impairment in Parkinson’s disease: a systematic review and meta-analysis of prospective cohort studies. Mov. Disord. 34 , 876–883 (2019).

Aune, D., Vatten, L. J. & Boffetta, P. Tobacco smoking and the risk of gallbladder disease. Eur. J. Epidemiol. 31 , 643–653 (2016).

Qin, L., Deng, H.-Y., Chen, S.-J. & Wei, W. Relationship between cigarette smoking and risk of chronic myeloid leukaemia: a meta-analysis of epidemiological studies. Hematology 22 , 193–200 (2017).

Petrick, J. L. et al. Tobacco, alcohol use and risk of hepatocellular carcinoma and intrahepatic cholangiocarcinoma: the Liver Cancer Pooling Project. Br. J. Cancer 118 , 1005–1012 (2018).

United States Department of Health, Education and Welfare. Smoking and Health. Report of the Advisory Committee on Smoking and Health to the Surgeon General of the United States Public Health Service https://www.cdc.gov/tobacco/data_statistics/sgr/index.htm (US DHEW, 1964).

United States Public Health Service Office of the Surgeon General & National Center for Chronic Disease Prevention and Health Promotion (US) Office on Smoking and Health. Smoking Cessation: A Report of the Surgeon General . (US Department of Health and Human Services, 2020).

Zheng, P., Barber, R., Sorensen, R. J. D., Murray, C. J. L. & Aravkin, A. Y. Trimmed constrained mixed effects models: formulations and algorithms. J. Comput. Graph Stat. 30 , 544–556 (2021).

Zheng, P. et al. The Burden of Proof studies: assessing the evidence of risk. Nat. Med. in press (2022).

Reitsma, M. B. et al. Spatial, temporal, and demographic patterns in prevalence of smoking tobacco use and attributable disease burden in 204 countries and territories, 1990–2019: a systematic analysis from the Global Burden of Disease Study 2019. Lancet 397 , 2337–2360 (2021).

Moher, D., Liberati, A., Tetzlaff, J. & Altman, D. G. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. Br. Med. J. 339 , b2535 (2009).

Liu, Z. Y., He, X. Z. & Chapman, R. S. Smoking and other risk factors for lung cancer in Xuanwei, China. Int. J. Epidemiol. 20 , 26–31 (1991).

Brownson, R. C., Reif, J. S., Keefe, T. J., Ferguson, S. W. & Pritzl, J. A. Risk factors for adenocarcinoma of the lung. Am. J. Epidemiol. 125 , 25–34 (1987).

Marugame, T. et al. Lung cancer death rates by smoking status: comparison of the Three-Prefecture Cohort study in Japan to the Cancer Prevention Study II in the USA. Cancer Sci. 96 , 120–126 (2005).

Dosemeci, M., Gokmen, I., Unsal, M., Hayes, R. B. & Blair, A. Tobacco, alcohol use, and risks of laryngeal and lung cancer by subsite and histologic type in Turkey. Cancer Causes Control 8 , 729–737 (1997).

Freedman, N. D. et al. Impact of changing US cigarette smoking patterns on incident cancer: risks of 20 smoking-related cancers among the women and men of the NIH-AARP cohort. Int. J. Epidemiol. 45 , 846–856 (2016).

Bae, J.-M. et al. Lung cancer incidence by smoking status in Korean men: 16 years of observations in the Seoul Male Cancer Cohort study. J. Korean Med. Sci. 28 , 636–637 (2013).

Everatt, R., Kuzmickienė, I., Virvičiūtė, D. & Tamošiūnas, A. Cigarette smoking, educational level and total and site-specific cancer: a cohort study in men in Lithuania. Eur. J. Cancer Prev. 23 , 579–586 (2014).

Nordlund, L. A., Carstensen, J. M. & Pershagen, G. Are male and female smokers at equal risk of smoking-related cancer: evidence from a Swedish prospective study. Scand. J. Public Health 27 , 56–62 (1999).

Siemiatycki, J., Krewski, D., Franco, E. & Kaiserman, M. Associations between cigarette smoking and each of 21 types of cancer: a multi-site case–control study. Int. J. Epidemiol. 24 , 504–514 (1995).

Chyou, P. H., Nomura, A. M. & Stemmermann, G. N. A prospective study of the attributable risk of cancer due to cigarette smoking. Am. J. Public Health 82 , 37–40 (1992).

Potter, J. D., Sellers, T. A., Folsom, A. R. & McGovern, P. G. Alcohol, beer, and lung cancer in postmenopausal women. The Iowa Women’s Health Study. Ann. Epidemiol. 2 , 587–595 (1992).

Chyou, P. H., Nomura, A. M., Stemmermann, G. N. & Kato, I. Lung cancer: a prospective study of smoking, occupation, and nutrient intake. Arch. Environ. Health 48 , 69–72 (1993).

Pesch, B. et al. Cigarette smoking and lung cancer–relative risk estimates for the major histological types from a pooled analysis of case–control studies. Int. J. Cancer 131 , 1210–1219 (2012).

Jöckel, K. H. et al. Occupational and environmental hazards associated with lung cancer. Int. J. Epidemiol. 21 , 202–213 (1992).

Jöckel, K. H., Ahrens, W., Jahn, I., Pohlabeln, H. & Bolm-Audorff, U. Occupational risk factors for lung cancer: a case-control study in West Germany. Int. J. Epidemiol. 27 , 549–560 (1998).

Lei, Y. X., Cai, W. C., Chen, Y. Z. & Du, Y. X. Some lifestyle factors in human lung cancer: a case-control study of 792 lung cancer cases. Lung Cancer 14 , S121–S136 (1996).

Pawlega, J., Rachtan, J. & Dyba, T. Evaluation of certain risk factors for lung cancer in Cracow (Poland)—a case–control study. Acta Oncol. 36 , 471–476 (1997).

Mao, Y. et al. Socioeconomic status and lung cancer risk in Canada. Int. J. Epidemiol. 30 , 809–817 (2001).

Barbone, F., Bovenzi, M., Cavallieri, F. & Stanta, G. Cigarette smoking and histologic type of lung cancer in men. Chest 112 , 1474–1479 (1997).

Matos, E., Vilensky, M., Boffetta, P. & Kogevinas, M. Lung cancer and smoking: a case–control study in Buenos Aires, Argentina. Lung Cancer 21 , 155–163 (1998).

Simonato, L. et al. Lung cancer and cigarette smoking in Europe: an update of risk estimates and an assessment of inter-country heterogeneity. Int. J. Cancer 91 , 876–887 (2001).

Risch, H. A. et al. Are female smokers at higher risk for lung cancer than male smokers? A case–control analysis by histologic type. Am. J. Epidemiol. 138 , 281–293 (1993).

Sankaranarayanan, R. et al. A case–control study of diet and lung cancer in Kerala, south India. Int. J. Cancer 58 , 644–649 (1994).

Band, P. R. et al. Identification of occupational cancer risks in British Columbia. Part I: methodology, descriptive results, and analysis of cancer risks, by cigarette smoking categories of 15,463 incident cancer cases. J. Occup. Environ. Med. 41 , 224–232 (1999).

Becher, H., Jöckel, K. H., Timm, J., Wichmann, H. E. & Drescher, K. Smoking cessation and nonsmoking intervals: effect of different smoking patterns on lung cancer risk. Cancer Causes Control 2 , 381–387 (1991).

Brockmöller, J., Kerb, R., Drakoulis, N., Nitz, M. & Roots, I. Genotype and phenotype of glutathione S-transferase class mu isoenzymes mu and psi in lung cancer patients and controls. Cancer Res. 53 , 1004–1011 (1993).

PubMed   Google Scholar  

Vena, J. E., Byers, T. E., Cookfair, D. & Swanson, M. Occupation and lung cancer risk. An analysis by histologic subtypes. Cancer 56 , 910–917 (1985).

Cascorbi, I. et al. Homozygous rapid arylamine N -acetyltransferase (NAT2) genotype as a susceptibility factor for lung cancer. Cancer Res. 56 , 3961–3966 (1996).

CAS   PubMed   Google Scholar  

Chiazze, L., Watkins, D. K. & Fryar, C. A case–control study of malignant and non-malignant respiratory disease among employees of a fiberglass manufacturing facility. Br. J. Ind. Med 49 , 326–331 (1992).

CAS   PubMed   PubMed Central   Google Scholar  

Ando, M. et al. Attributable and absolute risk of lung cancer death by smoking status: findings from the Japan Collaborative Cohort Study. Int. J. Cancer 105 , 249–254 (2003).

De Matteis, S. et al. Are women who smoke at higher risk for lung cancer than men who smoke? Am. J. Epidemiol. 177 , 601–612 (2013).

He, Y. et al. Changes in smoking behavior and subsequent mortality risk during a 35-year follow-up of a cohort in Xi’an, China. Am. J. Epidemiol. 179 , 1060–1070 (2014).

Nishino, Y. et al. Cancer incidence profiles in the Miyagi Cohort Study. J. Epidemiol. 14 , S7–S11 (2004).

Papadopoulos, A. et al. Cigarette smoking and lung cancer in women: results of the French ICARE case–control study. Lung Cancer 74 , 369–377 (2011).

Shimazu, T. et al. Alcohol and risk of lung cancer among Japanese men: data from a large-scale population-based cohort study, the JPHC study. Cancer Causes Control 19 , 1095–1102 (2008).

Tindle, H. A. et al. Lifetime smoking history and risk of lung cancer: results from the Framingham Heart Study. J. Natl Cancer Inst. 110 , 1201–1207 (2018).

PubMed   PubMed Central   Google Scholar  

Yong, L. C. et al. Intake of vitamins E, C, and A and risk of lung cancer. The NHANES I epidemiologic followup study. First National Health and Nutrition Examination Survey. Am. J. Epidemiol. 146 , 231–243 (1997).

Hansen, M. S. et al. Sex differences in risk of smoking-associated lung cancer: results from a cohort of 600,000 Norwegians. Am. J. Epidemiol. 187 , 971–981 (2018).

Boffetta, P. et al. Tobacco smoking as a risk factor of bronchioloalveolar carcinoma of the lung: pooled analysis of seven case-control studies in the International Lung Cancer Consortium (ILCCO). Cancer Causes Control 22 , 73–79 (2011).

Yun, Y. D. et al. Hazard ratio of smoking on lung cancer in Korea according to histological type and gender. Lung 194 , 281–289 (2016).

Suzuki, I. et al. Risk factors for lung cancer in Rio de Janeiro, Brazil: a case–control study. Lung Cancer 11 , 179–190 (1994).

De Stefani, E., Deneo-Pellegrini, H., Carzoglio, J. C., Ronco, A. & Mendilaharsu, M. Dietary nitrosodimethylamine and the risk of lung cancer: a case–control study from Uruguay. Cancer Epidemiol. Biomark. Prev. 5 , 679–682 (1996).

Google Scholar  

Kreuzer, M. et al. Risk factors for lung cancer in young adults. Am. J. Epidemiol. 147 , 1028–1037 (1998).

Armadans-Gil, L., Vaqué-Rafart, J., Rosselló, J., Olona, M. & Alseda, M. Cigarette smoking and male lung cancer risk with special regard to type of tobacco. Int. J. Epidemiol. 28 , 614–619 (1999).

Kubík, A. K., Zatloukal, P., Tomásek, L. & Petruzelka, L. Lung cancer risk among Czech women: a case–control study. Prev. Med. 34 , 436–444 (2002).

Rachtan, J. Smoking, passive smoking and lung cancer cell types among women in Poland. Lung Cancer 35 , 129–136 (2002).

Thun, M. J. et al. 50-year trends in smoking-related mortality in the United States. N. Engl. J. Med. 368 , 351–364 (2013).

Zatloukal, P., Kubík, A., Pauk, N., Tomásek, L. & Petruzelka, L. Adenocarcinoma of the lung among women: risk associated with smoking, prior lung disease, diet and menstrual and pregnancy history. Lung Cancer 41 , 283–293 (2003).

Hansen, M. S., Licaj, I., Braaten, T., Lund, E. & Gram, I. T. The fraction of lung cancer attributable to smoking in the Norwegian Women and Cancer (NOWAC) Study. Br. J. Cancer 124 , 658–662 (2021).

Zhang, P. et al. Association of smoking and polygenic risk with the incidence of lung cancer: a prospective cohort study. Br. J. Cancer 126 , 1637–1646 (2022).

Weber, M. F. et al. Cancer incidence and cancer death in relation to tobacco smoking in a population-based Australian cohort study. Int. J. Cancer 149 , 1076–1088 (2021).

Guo, L.-W. et al. A risk prediction model for selecting high-risk population for computed tomography lung cancer screening in China. Lung Cancer 163 , 27–34 (2022).

Mezzoiuso, A. G., Odone, A., Signorelli, C. & Russo, A. G. Association between smoking and cancers among women: results from the FRiCaM multisite cohort study. J. Cancer 12 , 3136–3144 (2021).

Hawrysz, I., Wadolowska, L., Slowinska, M. A., Czerwinska, A. & Golota, J. J. Adherence to prudent and mediterranean dietary patterns is inversely associated with lung cancer in moderate but not heavy male Polish smokers: a case–control study. Nutrients 12 , E3788 (2020).

Huang, C.-C., Lai, C.-Y., Tsai, C.-H., Wang, J.-Y. & Wong, R.-H. Combined effects of cigarette smoking, DNA methyltransferase 3B genetic polymorphism, and DNA damage on lung cancer. BMC Cancer 21 , 1066 (2021).

Viner, B., Barberio, A. M., Haig, T. R., Friedenreich, C. M. & Brenner, D. R. The individual and combined effects of alcohol consumption and cigarette smoking on site-specific cancer risk in a prospective cohort of 26,607 adults: results from Alberta’s Tomorrow Project. Cancer Causes Control 30 , 1313–1326 (2019).

Park, E. Y., Lim, M. K., Park, E., Oh, J.-K. & Lee, D.-H. Relationship between urinary 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanol and lung cancer risk in the general population: a community-based prospective cohort study. Front. Oncol. 11 , 611674 (2021).

De Stefani, E., Deneo-Pellegrini, H., Mendilaharsu, M., Carzoglio, J. C. & Ronco, A. Dietary fat and lung cancer: a case–control study in Uruguay. Cancer Causes Control 8 , 913–921 (1997).

Wünsch-Filho, V., Moncau, J. E., Mirabelli, D. & Boffetta, P. Occupational risk factors of lung cancer in São Paulo, Brazil. Scand. J. Work Environ. Health 24 , 118–124 (1998).

Hu, J. et al. A case-control study of diet and lung cancer in northeast China. Int. J. Cancer 71 , 924–931 (1997).

Jia, G., Wen, W., Massion, P. P., Shu, X.-O. & Zheng, W. Incorporating both genetic and tobacco smoking data to identify high-risk smokers for lung cancer screening. Carcinogenesis 42 , 874–879 (2021).

Rusmaully, J. et al. Risk of lung cancer among women in relation to lifetime history of tobacco smoking: a population-based case–control study in France (the WELCA study). BMC Cancer 21 , 711 (2021).

Jin, K. et al. Tobacco smoking modifies the association between hormonal factors and lung cancer occurrence among post-menopausal Chinese women. Transl. Oncol. 12 , 819–827 (2019).

Tse, L. A., Wang, F., Wong, M. C.-S., Au, J. S.-K. & Yu, I. T.-S. Risk assessment and prediction for lung cancer among Hong Kong Chinese men. BMC Cancer 22 , 585 (2022).

Huang, C.-C. et al. Joint effects of cigarette smoking and green tea consumption with miR-29b and DNMT3b mRNA expression in the development of lung cancer. Genes 13 , 836 (2022).

Hosseini, M. et al. Environmental risk factors for lung cancer in Iran: a case–control study. Int. J. Epidemiol. 38 , 989–996 (2009).

Naghibzadeh-Tahami, A. et al. Is opium use associated with an increased risk of lung cancer? A case–control study. BMC Cancer 20 , 807 (2020).

Shimatani, K., Ito, H., Matsuo, K., Tajima, K. & Takezaki, T. Cumulative cigarette tar exposure and lung cancer risk among Japanese smokers. Jpn J. Clin. Oncol. 50 , 1009–1017 (2020).

Lai, C.-Y. et al. Genetic polymorphism of catechol- O -methyltransferase modulates the association of green tea consumption and lung cancer. Eur. J. Cancer Prev. 28 , 316–322 (2019).

Schwartz, A. G. et al. Hormone use, reproductive history, and risk of lung cancer: the Women’s Health Initiative studies. J. Thorac. Oncol. 10 , 1004–1013 (2015).

Kreuzer, M., Gerken, M., Heinrich, J., Kreienbrock, L. & Wichmann, H.-E. Hormonal factors and risk of lung cancer among women? Int. J. Epidemiol. 32 , 263–271 (2003).

Sreeja, L. et al. Possible risk modification by CYP1A1, GSTM1 and GSTT1 gene polymorphisms in lung cancer susceptibility in a South Indian population. J. Hum. Genet. 50 , 618–627 (2005).

Siemiatycki, J. et al. Are the apparent effects of cigarette smoking on lung and bladder cancers due to uncontrolled confounding by occupational exposures? Epidemiology 5 , 57–65 (1994).

Chan-Yeung, M. et al. Risk factors associated with lung cancer in Hong Kong. Lung Cancer 40 , 131–140 (2003).

Lawania, S., Singh, N., Behera, D. & Sharma, S. Xeroderma pigmentosum complementation group D polymorphism toward lung cancer susceptibility survival and response in patients treated with platinum chemotherapy. Future Oncol. 13 , 2645–2665 (2017).

De Stefani, E. et al. Mate drinking and risk of lung cancer in males: a case-control study from Uruguay. Cancer Epidemiol. Biomark. Prev. 5 , 515–519 (1996).

Pérez-Padilla, R. et al. Exposure to biomass smoke and chronic airway disease in Mexican women. A case-control study. Am. J. Respir. Crit. Care Med. 154 , 701–706 (1996).

Zhang, X.-R. et al. Glucosamine use, smoking and risk of incident chronic obstructive pulmonary disease: a large prospective cohort study. Br. J. Nutr . https://doi.org/10.1017/S000711452100372X (2021).

Johannessen, A., Omenaas, E., Bakke, P. & Gulsvik, A. Incidence of GOLD-defined chronic obstructive pulmonary disease in a general adult population. Int. J. Tuberc. Lung Dis. 9 , 926–932 (2005).

Fox, J. Life-style and mortality: a large-scale census-based cohort study in Japan. J. Epidemiol. Community Health 45 , 173 (1991).

Article   PubMed Central   Google Scholar  

Thomson, B. et al. Low-intensity daily smoking and cause-specific mortality in Mexico: prospective study of 150 000 adults. Int. J. Epidemiol. 50 , 955–964 (2021).

van Durme, Y. M. T. A. et al. Prevalence, incidence, and lifetime risk for the development of COPD in the elderly: the Rotterdam study. Chest 135 , 368–377 (2009).

Li, L. et al. SERPINE2 rs16865421 polymorphism is associated with a lower risk of chronic obstructive pulmonary disease in the Uygur population: a case–control study. J. Gene Med. 21 , e3106 (2019).

Ganbold, C. et al. The cumulative effect of gene-gene interactions between GSTM1 , CHRNA3 , CHRNA5 and SOD3 gene polymorphisms combined with smoking on COPD risk. Int. J. Chron. Obstruct. Pulmon. Dis. 16 , 2857–2868 (2021).

Omori, H. et al. Twelve-year cumulative incidence of airflow obstruction among Japanese males. Intern. Med. 50 , 1537–1544 (2011).

Manson, J. E., Ajani, U. A., Liu, S., Nathan, D. M. & Hennekens, C. H. A prospective study of cigarette smoking and the incidence of diabetes mellitus among US male physicians. Am. J. Med. 109 , 538–542 (2000).

Lv, J. et al. Adherence to a healthy lifestyle and the risk of type 2 diabetes in Chinese adults. Int. J. Epidemiol. 46 , 1410–1420 (2017).

Waki, K. et al. Alcohol consumption and other risk factors for self-reported diabetes among middle-aged Japanese: a population-based prospective study in the JPHC study cohort I. Diabet. Med. 22 , 323–331 (2005).

Meisinger, C., Döring, A., Thorand, B. & Löwel, H. Association of cigarette smoking and tar and nicotine intake with development of type 2 diabetes mellitus in men and women from the general population: the MONICA/KORA Augsburg Cohort Study. Diabetologia 49 , 1770–1776 (2006).

Huh, Y. et al. Association of smoking status with the risk of type 2 diabetes among young adults: a nationwide cohort study in South Korea. Nicotine Tob. Res. 24 , 1234–1240 (2022).

Sawada, S. S., Lee, I.-M., Muto, T., Matuszaki, K. & Blair, S. N. Cardiorespiratory fitness and the incidence of type 2 diabetes: prospective study of Japanese men. Diabetes Care 26 , 2918–2922 (2003).

Will, J. C., Galuska, D. A., Ford, E. S., Mokdad, A. & Calle, E. E. Cigarette smoking and diabetes mellitus: evidence of a positive association from a large prospective cohort study. Int. J. Epidemiol. 30 , 540–546 (2001).

Nakanishi, N., Nakamura, K., Matsuo, Y., Suzuki, K. & Tatara, K. Cigarette smoking and risk for impaired fasting glucose and type 2 diabetes in middle-aged Japanese men. Ann. Intern. Med. 133 , 183–191 (2000).

Sairenchi, T. et al. Cigarette smoking and risk of type 2 diabetes mellitus among middle-aged and elderly Japanese men and women. Am. J. Epidemiol. 160 , 158–162 (2004).

Hou, X. et al. Cigarette smoking is associated with a lower prevalence of newly diagnosed diabetes screened by OGTT than non-smoking in Chinese men with normal weight. PLoS ONE 11 , e0149234 (2016).

Hu, F. B. et al. Diet, lifestyle, and the risk of type 2 diabetes mellitus in women. N. Engl. J. Med. 345 , 790–797 (2001).

Teratani, T. et al. Dose-response relationship between tobacco or alcohol consumption and the development of diabetes mellitus in Japanese male workers. Drug Alcohol Depend. 125 , 276–282 (2012).

Kawakami, N., Takatsuka, N., Shimizu, H. & Ishibashi, H. Effects of smoking on the incidence of non-insulin-dependent diabetes mellitus. Replication and extension in a Japanese cohort of male employees. Am. J. Epidemiol. 145 , 103–109 (1997).

Patja, K. et al. Effects of smoking, obesity and physical activity on the risk of type 2 diabetes in middle-aged Finnish men and women. J. Intern. Med. 258 , 356–362 (2005).

White, W. B. et al. High-intensity cigarette smoking is associated with incident diabetes mellitus in Black adults: the Jackson Heart Study. J. Am. Heart Assoc. 7 , e007413 (2018).

Uchimoto, S. et al. Impact of cigarette smoking on the incidence of Type 2 diabetes mellitus in middle-aged Japanese men: the Osaka Health Survey. Diabet. Med . 16 , 951–955 (1999).

Rimm, E. B., Chan, J., Stampfer, M. J., Colditz, G. A. & Willett, W. C. Prospective study of cigarette smoking, alcohol use, and the risk of diabetes in men. Br. Med. J. 310 , 555–559 (1995).

Article   CAS   Google Scholar  

Hilawe, E. H. et al. Smoking and diabetes: is the association mediated by adiponectin, leptin, or C-reactive protein? J. Epidemiol. 25 , 99–109 (2015).

InterAct, Consortium et al. Smoking and long-term risk of type 2 diabetes: the EPIC-InterAct study in European populations. Diabetes Care 37 , 3164–3171 (2014).

Jee, S. H., Foong, A. W., Hur, N. W. & Samet, J. M. Smoking and risk for diabetes incidence and mortality in Korean men and women. Diabetes Care 33 , 2567–2572 (2010).

Rasouli, B. et al. Smoking and the risk of LADA: results from a Swedish population-based case-control study. Diabetes Care 39 , 794–800 (2016).

Wannamethee, S. G., Shaper, A. G. & Perry, I. J., British Regional Heart Study. Smoking as a modifiable risk factor for type 2 diabetes in middle-aged men. Diabetes Care 24 , 1590–1595 (2001).

Radzeviciene, L. & Ostrauskas, R. Smoking habits and type 2 diabetes mellitus in women. Women Health 58 , 884–897 (2018).

Carlsson, S., Midthjell, K. & Grill, V., Nord-Trøndelag Study. Smoking is associated with an increased risk of type 2 diabetes but a decreased risk of autoimmune diabetes in adults: an 11-year follow-up of incidence of diabetes in the Nord-Trøndelag study. Diabetologia 47 , 1953–1956 (2004).

Akter, S. et al. Smoking, smoking cessation, and the risk of type 2 diabetes among Japanese adults: Japan Epidemiology Collaboration on Occupational Health Study. PLoS ONE 10 , e0132166 (2015).

Pirie, K. et al. The 21st century hazards of smoking and benefits of stopping: a prospective study of one million women in the UK. Lancet 381 , 133–141 (2013).

Park, C.-H. et al. [The effect of smoking status upon occurrence of impaired fasting glucose or type 2 diabetes in Korean men]. J. Prev. Med. Public Health 41 , 249–254 (2008).

Doi, Y. et al. Two risk score models for predicting incident Type 2 diabetes in Japan. Diabet. Med. 29 , 107–114 (2012).

van den Brandt, P. A. A possible dual effect of cigarette smoking on the risk of postmenopausal breast cancer. Eur. J. Epidemiol. 32 , 683–690 (2017).

Dossus, L. et al. Active and passive cigarette smoking and breast cancer risk: results from the EPIC cohort. Int. J. Cancer 134 , 1871–1888 (2014).

Kawai, M., Malone, K. E., Tang, M.-T. C. & Li, C. I. Active smoking and the risk of estrogen receptor-positive and triple-negative breast cancer among women ages 20 to 44 years. Cancer 120 , 1026–1034 (2014).

Reynolds, P. et al. Active smoking, household passive smoking, and breast cancer: evidence from the California Teachers Study. J. Natl Cancer Inst. 96 , 29–37 (2004).

Ellingjord-Dale, M. et al. Alcohol, physical activity, smoking, and breast cancer subtypes in a large, nested case-control study from the Norwegian Breast Cancer Screening Program. Cancer Epidemiol. Biomark. Prev. 26 , 1736–1744 (2017).

Arthur, R. et al. Association between lifestyle, menstrual/reproductive history, and histological factors and risk of breast cancer in women biopsied for benign breast disease. Breast Cancer Res. Treat. 165 , 623–631 (2017).

Luo, J. et al. Association of active and passive smoking with risk of breast cancer among postmenopausal women: a prospective cohort study. Br. Med. J. 342 , d1016 (2011).

White, A. J., D’Aloisio, A. A., Nichols, H. B., DeRoo, L. A. & Sandler, D. P. Breast cancer and exposure to tobacco smoke during potential windows of susceptibility. Cancer Causes Control 28 , 667–675 (2017).

Gram, I. T. et al. Breast cancer risk among women who start smoking as teenagers. Cancer Epidemiol. Biomark. Prev. 14 , 61–66 (2005).

Gammon, M. D. et al. Cigarette smoking and breast cancer risk among young women (United States). Cancer Causes Control 9 , 583–590 (1998).

Magnusson, C., Wedrén, S. & Rosenberg, L. U. Cigarette smoking and breast cancer risk: a population-based study in Sweden. Br. J. Cancer 97 , 1287–1290 (2007).

Chu, S. Y. et al. Cigarette smoking and the risk of breast cancer. Am. J. Epidemiol. 131 , 244–253 (1990).

Lemogne, C. et al. Depression and the risk of cancer: a 15-year follow-up study of the GAZEL cohort. Am. J. Epidemiol. 178 , 1712–1720 (2013).

Morabia, A., Bernstein, M., Héritier, S. & Khatchatrian, N. Relation of breast cancer with passive and active exposure to tobacco smoke. Am. J. Epidemiol. 143 , 918–928 (1996).

Conlon, M. S. C., Johnson, K. C., Bewick, M. A., Lafrenie, R. M. & Donner, A. Smoking (active and passive), N -acetyltransferase 2, and risk of breast cancer. Cancer Epidemiol. 34 , 142–149 (2010).

Ozasa, K., Japan Collaborative Cohort Study for Evaluation of Cancer. Smoking and mortality in the Japan Collaborative Cohort Study for Evaluation of Cancer (JACC). Asian Pac. J. Cancer Prev. 8 , 89–96 (2007).

Jones, M. E., Schoemaker, M. J., Wright, L. B., Ashworth, A. & Swerdlow, A. J. Smoking and risk of breast cancer in the Generations Study cohort. Breast Cancer Res. 19 , 118 (2017).

Bjerkaas, E. et al. Smoking duration before first childbirth: an emerging risk factor for breast cancer? Results from 302,865 Norwegian women. Cancer Causes Control 24 , 1347–1356 (2013).

Gram, I. T., Little, M. A., Lund, E. & Braaten, T. The fraction of breast cancer attributable to smoking: the Norwegian women and cancer study 1991–2012. Br. J. Cancer 115 , 616–623 (2016).

Li, C. I., Malone, K. E. & Daling, J. R. The relationship between various measures of cigarette smoking and risk of breast cancer among older women 65–79 years of age (United States). Cancer Causes Control 16 , 975–985 (2005).

Xue, F., Willett, W. C., Rosner, B. A., Hankinson, S. E. & Michels, K. B. Cigarette smoking and the incidence of breast cancer. Arch. Intern. Med. 171 , 125–133 (2011).

Parker, A. S., Cerhan, J. R., Putnam, S. D., Cantor, K. P. & Lynch, C. F. A cohort study of farming and risk of prostate cancer in Iowa. Epidemiology 10 , 452–455 (1999).

Sawada, N. et al. Alcohol and smoking and subsequent risk of prostate cancer in Japanese men: the Japan Public Health Center-based prospective study. Int. J. Cancer 134 , 971–978 (2014).

Hiatt, R. A., Armstrong, M. A., Klatsky, A. L. & Sidney, S. Alcohol consumption, smoking, and other risk factors and prostate cancer in a large health plan cohort in California (United States). Cancer Causes Control 5 , 66–72 (1994).

Cerhan, J. R. et al. Association of smoking, body mass, and physical activity with risk of prostate cancer in the Iowa 65+ Rural Health Study (United States). Cancer Causes Control 8 , 229–238 (1997).

Watters, J. L., Park, Y., Hollenbeck, A., Schatzkin, A. & Albanes, D. Cigarette smoking and prostate cancer in a prospective US cohort study. Cancer Epidemiol. Biomark. Prev. 18 , 2427–2435 (2009).

Butler, L. M., Wang, R., Wong, A. S., Koh, W.-P. & Yu, M. C. Cigarette smoking and risk of prostate cancer among Singapore Chinese. Cancer Causes Control 20 , 1967–1974 (2009).

Lotufo, P. A., Lee, I. M., Ajani, U. A., Hennekens, C. H. & Manson, J. E. Cigarette smoking and risk of prostate cancer in the physicians’ health study (United States). Int. J. Cancer 87 , 141–144 (2000).

Hsing, A. W. et al. Diet, tobacco use, and fatal prostate cancer: results from the Lutheran Brotherhood Cohort Study. Cancer Res. 50 , 6836–6840 (1990).

Veierød, M. B., Laake, P. & Thelle, D. S. Dietary fat intake and risk of prostate cancer: a prospective study of 25,708 Norwegian men. Int. J. Cancer 73 , 634–638 (1997).

Meyer, J., Rohrmann, S., Bopp, M. & Faeh, D. & Swiss National Cohort Study Group. Impact of smoking and excess body weight on overall and site-specific cancer mortality risk. Cancer Epidemiol. Biomark. Prev . 24 , 1516–1522 (2015).

Putnam, S. D. et al. Lifestyle and anthropometric risk factors for prostate cancer in a cohort of Iowa men. Ann. Epidemiol. 10 , 361–369 (2000).

Taghizadeh, N., Vonk, J. M. & Boezen, H. M. Lifetime smoking history and cause-specific mortality in a cohort study with 43 years of follow-up. PLoS ONE 11 , e0153310 (2016).

Park, S.-Y. et al. Racial/ethnic differences in lifestyle-related factors and prostate cancer risk: the Multiethnic Cohort Study. Cancer Causes Control 26 , 1507–1515 (2015).

Nomura, A. M., Lee, J., Stemmermann, G. N. & Combs, G. F. Serum selenium and subsequent risk of prostate cancer. Cancer Epidemiol. Biomark. Prev. 9 , 883–887 (2000).

Rodriguez, C., Tatham, L. M., Thun, M. J., Calle, E. E. & Heath, C. W. Smoking and fatal prostate cancer in a large cohort of adult men. Am. J. Epidemiol. 145 , 466–475 (1997).

Rohrmann, S. et al. Smoking and risk of fatal prostate cancer in a prospective U.S. study. Urology 69 , 721–725 (2007).

Giovannucci, E. et al. Smoking and risk of total and fatal prostate cancer in United States health professionals. Cancer Epidemiol. Biomark. Prev. 8 , 277–282 (1999).

Rohrmann, S. et al. Smoking and the risk of prostate cancer in the European Prospective Investigation into Cancer and Nutrition. Br. J. Cancer 108 , 708–714 (2013).

Lund Nilsen, T. I., Johnsen, R. & Vatten, L. J. Socio-economic and lifestyle factors associated with the risk of prostate cancer. Br. J. Cancer 82 , 1358–1363 (2000).

Hsing, A. W., McLaughlin, J. K., Hrubec, Z., Blot, W. J. & Fraumeni, J. F. Tobacco use and prostate cancer: 26-year follow-up of US veterans. Am. J. Epidemiol. 133 , 437–441 (1991).

Murray, C. J. L. et al. Global burden of 87 risk factors in 204 countries and territories, 1990–2019: a systematic analysis for the Global Burden of Disease Study 2019. Lancet 396 , 1223–1249 (2020).

Bero, L. A. & Jadad, A. R. How consumers and policymakers can use systematic reviews for decision making. Ann. Intern. Med. 127 , 37–42 (1997).

Centers for Disease Control and Prevention (CDC). Cigarette smoking among adults and trends in smoking cessation—United States, 2008. MMWR Morb. Mortal. Wkly Rep. 58 , 1227–1232 (2009).

Prochaska, J. O. & Goldstein, M. G. Process of smoking cessation: implications for clinicians. Clin. Chest Med. 12 , 727–735 (1991).

Page, M. J. et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. Br. Med. J. 372 , n71 (2021).

Stevens, G. A. et al. Guidelines for Accurate and Transparent Health Estimates Reporting: the GATHER statement. Lancet 388 , e19–e23 (2016).

BMJ Best Practice. What is GRADE? https://bestpractice.bmj.com/info/us/toolkit/learn-ebm/what-is-grade (BMJ, 2021).

The GRADE Working Group. GRADE handbook . https://gdt.gradepro.org/app/handbook/handbook.html (The GRADE Working Group, 2013).

Efron, B., Hastie, T., Johnstone, I. & Tibshirani, R. Least angle regression. Ann. Stat. 32 , 407–499 (2004).

Tibshirani, R. Regression shrinkage and selection via the lasso. J. R. Stat. Soc. Ser. B Stat. Methodol. 58 , 267–288 (1996).

von Hippel, P. T. The heterogeneity statistic I2 can be biased in small meta-analyses. BMC Med. Res. Methodol. 15 , 35 (2015).

Kontopantelis, E., Springate, D. A. & Reeves, D. A re-analysis of the Cochrane Library data: the dangers of unobserved heterogeneity in meta-analyses. PLoS ONE 8 , e69930 (2013).

Biggerstaff, B. J. & Tweedie, R. L. Incorporating variability in estimates of heterogeneity in the random effects model in meta-analysis. Stat. Med. 16 , 753–768 (1997).

Egger, M., Smith, G. D., Schneider, M. & Minder, C. Bias in meta-analysis detected by a simple, graphical test. Br. Med. J. 315 , 629–634 (1997).

Lee, P. N., Forey, B. A. & Coombs, K. J. Systematic review with meta-analysis of the epidemiological evidence in the 1900s relating smoking to lung cancer. BMC Cancer 12 , 385 (2012).

Rücker, G., Carpenter, J. R. & Schwarzer, G. Detecting and adjusting for small-study effects in meta-analysis. Biometr. J. 53 , 351–368 (2011).

Wu, Z.-J., Zhao, P., Liu, B. & Yuan, Z.-C. Effect of cigarette smoking on risk of hip fracture in men: a meta-analysis of 14 prospective cohort studies. PLoS ONE 11 , e0168990 (2016).

Thun, M. J. et al. in Cigarette Smoking Behaviour in the United States: changes in cigarette-related disease risks and their implication for prevention and control (eds Burns, D.M. et al.) Tobacco Control Monograph No. 8 Ch. 4 (National Cancer Institute, 1997).

Tolstrup, J. S. et al. Smoking and risk of coronary heart disease in younger, middle-aged, and older adults. Am. J. Public Health 104 , 96–102 (2014).

Jonas, M. A., Oates, J. A., Ockene, J. K. & Hennekens, C. H. Statement on smoking and cardiovascular disease for health care professionals. American Heart Association. Circulation 86 , 1664–1669 (1992).

Khan, S. S. et al. Cigarette smoking and competing risks for fatal and nonfatal cardiovascular disease subtypes across the life course. J. Am. Heart Assoc. 10 , e021751 (2021).

Download references

Acknowledgements

Research reported in this publication was supported by the Bill & Melinda Gates Foundation and Bloomberg Philanthropies. The content is solely the responsibility of the authors and does not necessarily represent the official views of the funders. The study funders had no role in study design, data collection, data analysis, data interpretation, writing of the final report or the decision to publish.

We thank the Tobacco Metrics Team Advisory Group for their valuable input and review of the work. The members of the Advisory Group are: P. Allebeck, R. Chandora, J. Drope, M. Eriksen, E. Fernández, H. Gouda, R. Kennedy, D. McGoldrick, L. Pan, K. Schotte, E. Sebrie, J. Soriano, M. Tynan and K. Welding.

Author information

Authors and affiliations.

Institute for Health Metrics and Evaluation, University of Washington, Seattle, WA, USA

Xiaochen Dai, Gabriela F. Gil, Marissa B. Reitsma, Noah S. Ahmad, Jason A. Anderson, Catherine Bisignano, Sinclair Carr, Rachel Feldman, Simon I. Hay, Jiawei He, Vincent Iannucci, Hilary R. Lawlor, Matthew J. Malloy, Laurie B. Marczak, Susan A. McLaughlin, Larissa Morikawa, Erin C. Mullany, Sneha I. Nicholson, Erin M. O’Connell, Chukwuma Okereke, Reed J. D. Sorensen, Joanna Whisnant, Aleksandr Y. Aravkin, Peng Zheng, Christopher J. L. Murray & Emmanuela Gakidou

Department of Health Metrics Sciences, School of Medicine, University of Washington, Seattle, WA, USA

Xiaochen Dai, Simon I. Hay, Jiawei He, Peng Zheng, Christopher J. L. Murray & Emmanuela Gakidou

Department of Applied Mathematics, University of Washington, Seattle, WA, USA

Aleksandr Y. Aravkin

You can also search for this author in PubMed   Google Scholar

Contributions

X.D., S.I.H., S.A.M., E.C.M., E.M.O., C.J.L.M. and E.G. managed the estimation or publications process. X.D. and G.F.G. wrote the first draft of the manuscript. X.D. and P.Z. had primary responsibility for applying analytical methods to produce estimates. X.D., G.F.G., N.S.A., J.A.A., S.C., R.F., V.I., M.J.M., L.M., S.I.N., C.O., M.B.R. and J.W. had primary responsibility for seeking, cataloguing, extracting or cleaning data, and for designing or coding figures and tables. X.D., G.F.G., M.B.R., N.S.A., H.R.L., C.O. and J.W. provided data or critical feedback on data sources. X.D., J.H., R.J.D.S., A.Y.A., P.Z., C.J.L.M. and E.G. developed methods or computational machinery. X.D., G.F.G., M.B.R., S.I.H., J.H., R.J.D.S., A.Y.A., P.Z., C.J.L.M. and E.G. provided critical feedback on methods or results. X.D., G.F.G., M.B.R., C.B., S.I.H., L.B.M., S.A.M., A.Y.A. and E.G. drafted the work or revised it critically for important intellectual content. X.D., S.I.H., L.B.M., E.C.M., E.M.O. and E.G. managed the overall research enterprise.

Corresponding author

Correspondence to Xiaochen Dai .

Ethics declarations

Competing interests.

The authors declare no competing interests.

Peer review

Peer review information.

Nature Medicine thanks Frederic Sitas and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. Primary Handling Editors: Jennifer Sargent and Ming Yang, in collaboration with the Nature Medicine team.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Extended data

Extended data fig. 1 prisma 2020 flow diagram for an updated systematic review of the smoking and tracheal, bronchus, and lung cancer risk-outcome pair..

The PRISMA flow diagram of an updated systematic review on the relationship between smoking and lung cancer conducted on PubMed to update historical review from previous cycles of the Global Burden of Disease Study. Template is from: Page MJ, McKenzie JE, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ 2021;372:n71. doi: 10.1136/bmj.n71. For more information, visit: http://www.prisma-statement.org/ .

Extended Data Fig. 2 PRISMA 2020 flow diagram for an updated systematic review of the Smoking and Chronic obstructive pulmonary disease risk-outcome pair.

The PRISMA flow diagram of an updated systematic review on the relationship between smoking and chronic obstructive pulmonary disease conducted on PubMed to update historical review from previous cycles of the Global Burden of Disease Study. Template is from: Page MJ, McKenzie JE, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ 2021;372:n71. doi: 10.1136/bmj.n71. For more information, visit: http://www.prisma-statement.org/ .

Extended Data Fig. 3 PRISMA 2020 flow diagram for an updated systematic review of the Smoking and Diabetes mellitus type 2 risk- outcome pair.

The PRISMA flow diagram of an updated systematic review on the relationship between smoking and type 2 diabetes conducted on PubMed to update historical review from previous cycles of the Global Burden of Disease Study. Template is from: Page MJ, McKenzie JE, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ 2021;372:n71. doi: 10.1136/bmj.n71. For more information, visit: http://www.prisma-statement.org/ .

Extended Data Fig. 4 PRISMA 2020 flow diagram for an updated systematic review of the Smoking and Breast cancer risk-outcome pair.

The PRISMA flow diagram of an updated systematic review on the relationship between smoking and breast cancer conducted on PubMed to update historical review from previous cycles of the Global Burden of Disease Study. Template is from: Page MJ, McKenzie JE, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ 2021;372:n71. doi: 10.1136/bmj.n71. For more information, visit: http://www.prisma-statement.org/ .

Extended Data Fig. 5 PRISMA 2020 flow diagram for an updated systematic review of the Smoking and Prostate cancer risk-outcome pair.

The PRISMA flow diagram of an updated systematic review on the relationship between smoking and prostate cancer conducted on PubMed to update historical review from previous cycles of the Global Burden of Disease Study. Template is from: Page MJ, McKenzie JE, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ 2021;372:n71. doi: 10.1136/bmj.n71. For more information, visit: http://www.prisma-statement.org/ .

Extended Data Fig. 6 Smoking and Breast Cancer.

a , log-relative risk function. b , relative risk function. c , A modified funnel plot showing the residuals (relative to 0) on the x-axis and the estimated standard deviation (SD) that includes reported SD and between-study heterogeneity on the y-axis.

Supplementary information

Supplementary information.

Supplementary Information 1: Data source identification and assessment. Supplementary Information 2: Data inputs. Supplementary Information 3: Study quality and bias assessment. Supplementary Information 4: The dose–response RR curves and their 95% UIs for all smoking–outcome pairs. Supplementary Information 5: Supplementary methods. Supplementary Information 6: Sensitivity analysis. Supplementary Information 7: Binary smoking–outcome pair. Supplementary Information 8: Risk curve details. Supplementary Information 9: GATHER and PRISMA checklists.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

Cite this article.

Dai, X., Gil, G.F., Reitsma, M.B. et al. Health effects associated with smoking: a Burden of Proof study. Nat Med 28 , 2045–2055 (2022). https://doi.org/10.1038/s41591-022-01978-x

Download citation

Received : 11 April 2022

Accepted : 28 July 2022

Published : 10 October 2022

Issue Date : October 2022

DOI : https://doi.org/10.1038/s41591-022-01978-x

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

This article is cited by

Estimating effects of whole grain consumption on type 2 diabetes, colorectal cancer and cardiovascular disease: a burden of proof study.

Nutrition Journal (2024)

• Gabriela F. Gil
• Jason A. Anderson
• Emmanuela Gakidou

Nature Communications (2024)

• Luisa S. Flor

Nature Medicine (2024)

Metabolic profiling of smoking, associations with type 2 diabetes and interaction with genetic susceptibility

• Sofia Carlsson

European Journal of Epidemiology (2024)

Global burden of prostate cancer attributable to smoking among males in 204 countries and territories, 1990–2019

• Hanfei Zhang
• Dingping Huang
• Daqing Hong

BMC Cancer (2023)

Quick links

• Explore articles by subject
• Guide to authors
• Editorial policies

Sign up for the Nature Briefing: Cancer newsletter — what matters in cancer research, free to your inbox weekly.

Tobacco, Nicotine, and E-Cigarettes Research Report What are the physical health consequences of tobacco use?

Cigarette smoking harms nearly every organ in the body, 1,44 and smoking is the leading preventable cause of premature death in the United States. Although rates of smoking have declined, it is estimated that it leads to about 480,000 deaths yearly. 1 Smokers aged 60 and older have a twofold increase in mortality compared with those who have never smoked, dying an estimated 6 years earlier. 45 Quitting smoking results in immediate health benefits, and some or all of the reduced life expectancy can be recovered depending on the age a person quits. 46

Although nicotine itself does not cause cancer, at least 69 chemicals in tobacco smoke are carcinogenic, 1 and cigarette smoking accounts for at least 30 percent of all cancer deaths. 22 The overall rates of death from cancer are twice as high among smokers as nonsmokers, with heavy smokers having a four times greater risk of death from cancer than nonsmokers. 1

Foremost among the cancers caused by tobacco use is lung cancer. Cigarette smoking has been linked to about 80 to 90 percent of all cases of lung cancer, the leading cause of cancer death for both men and women, and it is responsible for roughly 80 percent of deaths from this disease. 22,47 Smoking increases lung cancer risk five to tenfold, with greater risk among heavy smokers. 48 Smoking is also associated with cancers of the mouth, pharynx, larynx, esophagus, stomach, pancreas, cervix, kidney, and bladder, as well as acute myeloid. 1 Cigarette smoking is not the only form of tobacco use associated with cancers. Smokeless tobacco (see " Other Tobacco Products ") has been linked to cancer of the pharynx, esophagus, stomach, and lung, as well as to colorectal cancer. 49

In addition to cancer, smoking causes lung diseases such as chronic bronchitis and emphysema, and it has been found to exacerbate asthma symptoms in adults and children. Cigarette smoking is the most significant risk factor for chronic obstructive pulmonary disease (COPD). 50 Survival statistics indicate that quitting smoking results in repair to much of the smoking-induced lung damage over time. However, once COPD develops, it is irreversible; COPD-related lung damage is not repaired with time.

Smoking also substantially increases the risk of heart disease, including stroke, heart attack, vascular disease, and aneurysm. 51,52 Cardiovascular disease is responsible for 40 percent of all smoking-related deaths. 53 Smoking causes coronary heart disease, the leading cause of death in the United States. Smoking is also linked to many other major health conditions—including rheumatoid arthritis, inflammation, and impaired immune function. 1 Even young smokers aged 26 to 41 report reduced health-related quality of life compared with nonsmoking peers, according to a cross-sectional population study. 54 Recent animal research also identified a pathway between the pancreas and a part of the brain active in nicotine intake, potentially linking cigarette smoking to the risk of developing Type 2 Diabetes.

• Skip to main content
• Skip to FDA Search
• Skip to in this section menu
• Skip to footer links

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you're on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

U.S. Food and Drug Administration

•   Search
•   Menu
• Tobacco Products

Tobacco Science & Research

FDA supports science and research to help us better understand tobacco use and associated risks so that we can reduce the public health burden of tobacco in the United States. 

Research programs and projects include, but are not limited to, the scientific fields of epidemiology, behavior, biology, medicine, economics, chemistry, engineering, toxicology, pharmacology, addiction, public health, communications, marketing, and statistics. 

Research News

• PATH Study Data User Workshop: Apply Now for the Summer 2024 Events
• PATH Study: Wave 7 Restricted-Use Files (RUFs) now available
• CTP and NIDA Announce the Continuation of the Population Assessment of Tobacco and Health (PATH) Study
• FDA & NIH Announce Funding Awards for TCORS 3.0 
• PATH Study: Wave 6 Public-Use Files now available
• PATH Study: Wave 6 Restricted-Use Files now available
• PATH Study: Wave 4 and Wave 5 Biomarker Restricted-Use Files (BRUFs) now available, including data from youth participants
• PATH Study: New Special Collection Public-Use Files (SCPUFs): Wave 5.5 and PATH Study Adult Telephone Survey (PATH-ATS)
• PATH Study: New Data Tables and Figures Available: Waves 1 - 5 (2013 – 2019)
• Find statistics about youth tobacco use  from the most recent National Youth Tobacco Survey

Stay current on FDA's tobacco regulatory science and research efforts, tobacco scientific publications and study findings, and research grants by subscribing to CTP's quarterly Spotlight on Science newsletter.

Subscribe to Spotlight on Science

Studies involving administration of unauthorized tobacco products.

If you plan to study tobacco products that do not have marketing authorization or that do not comply with an applicable tobacco product standard, you may submit your proposed protocol to FDA for review. FDA intends to evaluate specific uses of investigational tobacco products on a case-by-case basis according to potential human subject protection concerns or other impacts on public health.

Using Investigational Tobacco Products

Investigators who are designing a protocol involving administration of a tobacco product to humans should review the information below regarding the need for submitting their protocol to Food and Drug Administration (FDA) Center for Tobacco Products (CTP) for review.  

Investigators are encouraged to work with tobacco product manufacturers to ensure availability of products to complete planned studies. FDA evaluates the specific uses of investigational tobacco products (ITPs) on a case-by-case basis according to potential human subject protection concerns or other impacts on public health. Generally, submission of protocols by industry and academic researchers for FDA review is a voluntary process; however, FDA will review all protocols submitted. FDA recommends submission of proposed use of ITPs to FDA for review only if the study design is more likely to raise concerns about human subject protection, public health, or both. As discussed by FDA in its February 2019 guidance, Use of Investigational Tobacco Products , factors to consider would be studies that plan to enroll vulnerable populations, particularly those < 21 years old, studies that involve significant increases over the participants’ usual exposure to nicotine, studies that modify the tobacco product in a manner different from that described by the manufacturer or study of a novel product for which there is limited experience and knowledge.  

For all clinical studies involving use of ITPs, we recommend that you notify FDA, all participating clinical investigators, and any committee or group formally designated to oversee the study of any serious or unexpected adverse experience associated with the tobacco product you are investigating within a few weeks after initial notification, and that you supply FDA with a completed case report form for the adverse experience. We encourage the reporting of adverse experiences associated with a clinical investigation of an investigational tobacco product to FDA through the FDACTP Safety Reporting Portal for Researchers.

FDA is committed to furthering scientific research on tobacco products and has a major investment in regulatory science. If you plan to study tobacco products that do not have marketing authorization or that do not comply with an applicable tobacco product standard, you may submit your proposed protocol to FDA for review based on the criteria described above. FDA will review any protocols submitted and intends to evaluate specific uses of investigational tobacco products on a case-by-case basis according to potential human subject protection concerns or other impacts on public health. Generally, FDA does not recommend that investigators correspond with us about the use of investigational tobacco products in nonclinical studies as these are not ordinarily reviewed. You may refer to the draft guidance, Use of Investigational Tobacco Products , for more information regarding how to submit your proposed use of an investigational tobacco product and how FDA intends to make enforcement decisions regarding the use of investigational tobacco products.  

FDA understands that investigators may choose to obtain tobacco products directly from a tobacco product manufacturer with the sole intent to use the products for research investigations without commercializing the products. In such cases, FDA recommends that investigators add language to all product labels to indicate that these products are limited to investigational use, that study participants be instructed that the products may not be further distributed, and that study protocols include a plan to collect and account for all investigational tobacco products after the study has concluded. 

If there are additional questions, investigators should reach out to the FDA CTP at: [email protected]. 

The email should:

• Clearly and uniquely identify the product(s) you wish to study by brand and sub-brand—including the type or category of tobacco product (e.g., cigarette, smokeless tobacco, cigar, electronic nicotine delivery systems [ENDS], waterpipe tobacco) and subcategory (e.g., closed or open e-cigarette, closed or open e-liquid).
• Provide additional available information such as packaging type, package quantity, and/or characterizing flavor that may help answer the specific question(s)

Once the FDA CTP receives the email, they will make every effort to respond via email within 2weeks. 

Note that the FDA CTP intends to respond to investigators within 60 days of receipt of protocols for review. Investigators should receive acknowledgement of the submission with the name and contact information for the assigned Regulatory Health Project Manager (RHPM). If investigators do not receive a response within 60 days, they should contact the RHPM. Investigators may also contact their NIH Program Officer to discuss additional steps/actions.

If the marketed products will be used with investigator-manipulated modification(s), then the investigator should submit an ITP request. In addition to the protocol and other information described in the FDA Draft Guidance, the ITP request should also include: 

• A description of the planned modification(s).
• A rationale for how these modification(s) support the study design and do not increase risk to human participants.

Tobacco Researcher Interviews: Meet some of the people who lead tobacco research

Ongoing Research

• Read about the research goals of the Population Assessment of Tobacco and Health (PATH) Study , a collaboration between FDA and NIH, as well as availability of Restricted Use Files (RUF) and Public Use Files (PUF) for Waves 1 & 2.
• Learn about CTP's research priorities that build the science base behind FDA's authority to regulate tobacco products.
• Find out more about the Tobacco Regulatory Science Program (TRSP), FDA's partnership with NIH to foster tobacco regulatory research, including the  Tobacco Centers of Regulatory Science (TCORS). 
• Learn about FDA's collaboration with CDC on the National Youth Tobacco Survey .
• Review information on Harmful and Potentially Harmful Constituents .
• Understand more about Modified Risk Tobacco Products and the rigorous standards in place to protect the public's health.

Additional Resources

• Products, Ingredients and Components
• FDA's New Regulations for E-Cigarettes, Cigars, and All Other Tobacco Products
• FDA Safety Reporting Portal for Tobacco Products
• Connect with Us
• Tobacco Control Act

Stay current on tobacco regulatory science and research, tobacco scientific publications and study findings, CTP grants, and more in this quarterly newsletter.

A .gov website belongs to an official government organization in the United States.

A lock ( ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites.

• About Health Effects of Cigarette Smoking
• Secondhand Smoke
• E-cigarettes (Vapes)
• Menthol Tobacco Products
• Morbidity and Mortality Weekly Reports (MMWR)
• About Surveys
• Other Tobacco Products
• Smoking and Tobacco Use Features
• Patient Care Settings and Smoking Cessation
• Patient Care
• Funding Opportunity Announcements
• State and Community Work
• National and State Tobacco Control Program
• Multimedia and Tools
• Tobacco - Health Equity
• Tobacco - Surgeon General's Reports
• State Tobacco Activities Tracking and Evaluation (STATE) System
• Global Tobacco Control

E-Cigarette Use Among Youth

What to know.

E-cigarettes are the most commonly used tobacco product among U.S. youth. No tobacco products, including e-cigarettes, are safe, especially for children, teens, and young adults. Learn more about e-cigarette use among youth.

• In the United States, youth use e-cigarettes, or vapes, more than any other tobacco product. 1
• No tobacco products, including e-cigarettes, are safe, especially for children, teens, and young adults. 2
• Most e-cigarettes contain nicotine, which is highly addictive. Nicotine can harm the parts of an adolescent's brain that control attention, learning, mood, and impulse control. 2
• E-cigarette marketing, the availability of flavored products, social influences, and the effects of nicotine can influence youth to start or continue vaping. 3 4
• Most middle and high school students who vape want to quit. 5
• Many people have an important role in protecting youth from vaping including parents and caregivers, educators and school administrators, health care providers, and community partners.
• States and local communities can implement evidence-based policies, programs, and services to reduce youth vaping.

E-cigarette use among U.S. youth

In 2023, e-cigarettes were the most commonly used tobacco product among middle and high school students in the United States. In 2023: 6

• 550,000 (4.6%) middle school students.
• 1.56 million (10.0%) high school students.
• Among students who had ever used e-cigarettes, 46.7% reported current e-cigarette use.
• 1 in 4 (25.2%) used an e-cigarette every day.
• 1 in 3 (34.7%) used an e-cigarette on at least 20 of the last 30 days.
• 9 in 10 (89.4%) used flavored e-cigarettes.
• Most often used disposable e-cigarettes (60.7%) followed by e-cigarettes with prefilled or refillable pods or cartridges (16.1%).
• Most commonly reported using the following brands: Elf Bar, Esco Bars, Vuse, JUUL, and Mr. Fog.

Most middle and high school students who vape want to quit and have tried to quit. 5 In 2020:

• 63.9% of students who currently used e-cigarettes reported wanting to quit.
• 67.4% of students who currently used e-cigarettes reported trying to quit in the last year.

Most tobacco use, including vaping, starts and is established during adolescence. There are many factors associated with youth tobacco product use . These include:

• Tobacco advertising that targets youth.
• Product accessibility.
• Availability of flavored products.
• Social influences.
• Adolescent brain sensitivity to nicotine.

Some groups of middle and high school students use e-cigarettes at a higher percentage than others. For example, in 2023: 6

• More females than males reported current e-cigarette use.
• Non-Hispanic multiracial students: 20.8%.
• Non-Hispanic White students: 18.4%.
• Hispanic or Latino students: 18.2%.
• Non-Hispanic American Indian and Alaska Native students: 15.4%.
• Non-Hispanic Black or African American students: 12.9%.

Many young people who vape also use other tobacco products, including cigarettes and cigars. 7 This is called dual use. In 2020: 8

• About one in three high school students (36.8%) who vaped also used other tobacco products.
• One in two middle school students (49.0%) who vaped also used other tobacco products.

E-cigarettes can also be used to deliver other substances, including cannabis. In 2016, nearly one in three (30.6%) of U.S. middle and high school students who had ever used an e-cigarette reported using marijuana in the device. 9

• Park-Lee E, Ren C, Cooper M, Cornelius M, Jamal A, Cullen KA. Tobacco product use among middle and high school students—United States, 2022 . MMWR Morb Mortal Wkly Rep. 2022;71:1429–1435.
• U.S. Department of Health and Human Services. E-cigarette Use Among Youth and Young Adults: A Report of the Surgeon General . Centers for Disease Control and Prevention; 2016. Accessed Feb 14, 2024.
• Apelberg BJ, Corey CG, Hoffman AC, et al. Symptoms of tobacco dependence among middle and high school tobacco users: results from the 2012 National Youth Tobacco Survey . Am J Prev Med. 2014;47(Suppl 1):S4–14.
• Gentzke AS, Wang TW, Cornelius M, et al. Tobacco product use and associated factors among middle and high school students—National Youth Tobacco Survey, United States, 2021 . MMWR Surveill Summ. 2022;71(No. SS-5):1–29.
• Zhang L, Gentzke A, Trivers KF, VanFrank B. Tobacco cessation behaviors among U.S. middle and high school students, 2020 . J Adolesc Health. 2022;70(1):147–154.
• Birdsey J, Cornelius M, Jamal A, et al. Tobacco product use among U.S. middle and high school students—National Youth Tobacco Survey, 2023 . MMWR Morb Mortal Wkly Rep. 2023;72:1173–1182.
• Wang TW, Gentzke AS, Creamer MR, et al. Tobacco product use and associated factors among middle and high school students—United States, 2019 . MMWR Surveill Summ. 2019;68(No. SS-12):1–22.
• Wang TW, Gentzke AS, Neff LJ, et al. Characteristics of e-cigarette use behaviors among US youth, 2020 . JAMA Netw Open. 2021;4(6):e2111336.
• Trivers KF, Phillips E, Gentzke AS, Tynan MA, Neff LJ. Prevalence of cannabis use in electronic cigarettes among U.S. youth . JAMA Pediatr. 2018;172(11):1097–1099.

Smoking and Tobacco Use

Commercial tobacco use is the leading cause of preventable disease, disability, and death in the United States.

For Everyone

Health care providers, public health.

• Search Menu
• Sign in through your institution
• Supplements
• Advance articles
• Editor's Choice
• Special Issues
• Author Guidelines
• Submission Site
• Why Publish With Us?
• Open Access
• About Nicotine & Tobacco Research
• About Society for Nicotine & Tobacco Research
• Editorial Board
• Advertising and Corporate Services
• Journals Career Network
• Self-Archiving Policy
• Dispatch Dates
• Journals on Oxford Academic
• Books on Oxford Academic

Article Contents

Association of covid-19 and the prevalence of in-person vs telehealth primary care visits and subsequent impacts on tobacco use assessment and referral for cessation assistance.

• Article contents
• Figures & tables
• Supplementary Data

Susan A Flocke, Elizabeth Albert, Steven Lewis, Eileen Seeholzer, Steffani R Bailey, Association of COVID-19 and the prevalence of in-person vs telehealth primary care visits and subsequent impacts on tobacco use assessment and referral for cessation assistance, Nicotine & Tobacco Research , 2024;, ntae126, https://doi.org/10.1093/ntr/ntae126

• Permissions Icon Permissions

The COVID-19 pandemic dramatically altered patterns of health care delivery. Smoking remains an important risk factor for multiple chronic conditions and may exacerbate more severe symptoms of COVID-19. Thus, it is important to understand how pandemic-induced changes in primary care practice patterns affected smoking assessment and cessation assistance.

Electronic health record (EHR) data from 8 community health centers were examined from March 1, 2019 to February 28, 2022. Data include both telehealth (phone and video) and in-person office visits and represent 310,388 visits by adult patients. Rates of smoking assessment, provision of referral to counseling and orders for smoking cessation medications were calculated. Comparisons by visit mode and time period were examined using generalized estimating equations and logistic regression.

The proportion of telehealth visits was <0.1% one year prior to COVID-19 onset and, 54.5% and 34.1% 1 and 2 years after. The odds of asking about smoking status and offering a referral to smoking cessation counseling were significantly higher during in-person vs. telehealth visits; AOR (95% CI) = 15.0 (14.7 –15.4) and AOR (95% CI)= 6.5 (3.0 – 13.9), respectively. The interaction effect of visit type * time period was significant for ordering smoking cessation medications.

Telehealth visits were significantly less likely to include smoking status assessment and referral to smoking cessation counseling compared to in-person visits. Given that smoking assessment and cessation assistance do not require face-to-face interactions with health care providers, continued efforts are needed to ensure provision at all visits, regardless of modality.

The COVID-19 pandemic dramatically altered patterns of health care seeking and delivery with a considerable rise in telehealth visits. This study examined one year prior to the onset of COVID-19 and two years after to evaluate the assessment of tobacco use and assistance with tobacco cessation and differences during in-person vs telehealth visits. Tobacco assessment was 15 times more likely during in-person vs. telehealth visits in the two years post onset of COVID-19. Given that telehealth visits are likely to continue, ensuring that patients are regularly assessed for tobacco regardless of visit modality is an important concern for health systems.

Email alerts

Citing articles via.

• About Nicotine & Tobacco Research
• Recommend to your Library

Affiliations

• Online ISSN 1469-994X
• Copyright © 2024 Society for Research on Nicotine and Tobacco
• About Oxford Academic
• Publish journals with us
• University press partners
• What we publish
• New features  
• Open access
• Institutional account management
• Rights and permissions
• Get help with access
• Accessibility
• Advertising
• Media enquiries
• Oxford University Press
• Oxford Languages
• University of Oxford

Oxford University Press is a department of the University of Oxford. It furthers the University's objective of excellence in research, scholarship, and education by publishing worldwide

• Copyright © 2024 Oxford University Press
• Cookie settings
• Cookie policy
• Privacy policy
• Legal notice

This Feature Is Available To Subscribers Only

Sign In or Create an Account

This PDF is available to Subscribers Only

For full access to this pdf, sign in to an existing account, or purchase an annual subscription.

• My View My View
• Following Following
• Saved Saved

Tobacco industry aims to hook new generation on vapes, WHO says

• Medium Text

Sign up here.

Reporting by Emma Rumney; editing by Hugh Lawson, Mark Heinrich and Cynthia Osterman

Our Standards: The Thomson Reuters Trust Principles. New Tab , opens new tab

Business Chevron

Japan, at G7 meet, renews push to keep yen bears in check

Japan renewed its push to counter excessive yen falls during a weekend gathering of Group of Seven (G7) finance leaders, after a recent rise in bond yields to a 12-year high failed to slow the currency's stubborn decline.

Take the Quiz: Find the Best State for You »

What's the best state for you ».

Tobacco Industry Aims to Hook New Generation on Vapes, WHO Says

FILE PHOTO: A woman holds an e-cigarette as she vapes on a street in Manchester, Britain March 6, 2024. REUTERS/Temilade Adelaja

LONDON (Reuters) -Tobacco companies still actively target young people via social media, sports and music festivals and new, flavoured products, the World Health Organization (WHO) said on Thursday, accusing companies of trying to hook a new generation on nicotine.

Amid ever-stricter regulation targeting cigarettes, big tobacco companies and new entrants have begun offering smoking alternatives such as vapes, which they say are aimed at adult smokers.

But the WHO and industry watchdog STOP said in a joint report these products are often marketed to youth, their design and fruity flavours appeal to children and young people are more likely to use them than adults in many countries.

Tedros Adhanom Ghebreyesus, the WHO's director-general, rejected the industry's claim that it is working to reduce the harm from smoking. "It's dishonest to talk about harm reduction when they are marketing to children," he said in the report.

The WHO's increasingly tough stance on newer nicotine products follows a sharp rise in youth vaping across several countries.

The WHO pointed to flavours like bubblegum as one driver of this rise. The industry says flavours are an important tool in encouraging adults to switch away from smoking.

Large tobacco companies have mostly steered away from such flavours. But firms including Philip Morris International and British American Tobacco target youth via the sponsorship of music and sports festivals and the use of social media, the WHO said.

These provide platforms to promote their brands to younger audiences and hand out free samples, it continued.

Both companies said they aimed to move smokers away from cigarettes. BAT added that it follows principles for responsible marketing and ensuring its products were for adults only.

PMI said scientifically substantiated smoking alternatives must be part of sound tobacco policy and it was ready to engage with any government and the WHO on the issue.

The WHO said there is insufficient evidence vapes help people quit smoking. The body said there is evidence that vaping increases traditional cigarette use, especially among youth.

However, Sarah Jackson, principal research fellow at University College London's Tobacco and Alcohol Research Group, said these statements "do not accurately reflect current evidence on e-cigarettes".

(Reporting by Emma Rumney; editing by Hugh Lawson, Mark Heinrich and Cynthia Osterman)

Copyright 2024 Thomson Reuters .

Join the Conversation

America 2024

Health News Bulletin

Stay informed on the latest news on health and COVID-19 from the editors at U.S. News & World Report.

Sign in to manage your newsletters »

Sign up to receive the latest updates from U.S News & World Report and our trusted partners and sponsors. By clicking submit, you are agreeing to our Terms and Conditions & Privacy Policy .

You May Also Like

The 10 worst presidents.

U.S. News Staff Feb. 23, 2024

Cartoons on President Donald Trump

Feb. 1, 2017, at 1:24 p.m.

Photos: Obama Behind the Scenes

April 8, 2022

Photos: Who Supports Joe Biden?

March 11, 2020

U.S. Military Deaths, by War

Elliott Davis Jr. May 25, 2024

Takeaways From the NCAA’s Settlement

Laura Mannweiler May 24, 2024

Noncitizen Voting: the Fiction and Facts

Aneeta Mathur-Ashton May 24, 2024

Quiz: Who Said What in Trump’s Trial?

U.S. News Staff May 24, 2024

CDC: COVID-19 Strains Are on the Rise

Cecelia Smith-Schoenwalder May 24, 2024

Consumers See Worsening Job Market

Tim Smart May 24, 2024

• Share full article

Advertisement

Supported by

Cannabis Tops Alcohol as Americans’ Daily Drug of Choice

A new study shows a growing number of people are regularly using cannabis, while frequent alcohol consumption has remained stable.

By Christina Caron

For the first time on record, cannabis has outpaced alcohol as the daily drug of choice for Americans.

In 2022 there were 17.7 million people who reported using cannabis either every day or nearly every day, compared with 14.7 million who reported using alcohol with the same frequency, according to a study, published on Wednesday in the journal Addiction that analyzed data from the U.S. National Survey on Drug Use and Health.

While far more people drink than use cannabis, drinking frequently has become slightly less common than it was around 15 years ago, the study found. But the proportion of people in the U.S. who use cannabis frequently has increased 15-fold in the three decades since 1992, when daily cannabis use hit a low point.

Cannabis legalization has also rapidly accelerated since the ’90s. The drug is now legal for recreational use in 24 states and Washington, D.C. , and for medical use in 38 states and D.C.

The sharp increase in the prevalence of high-frequency cannabis use over the last three decades might partly be attributed to a growing acceptance of the drug, said Jonathan P. Caulkins, a professor of public policy at Heinz College at Carnegie Mellon University. And because the survey data was self-reported, people may now feel more comfortable disclosing how often they use it.

Even so, “I don’t think that for most daily or near-daily users it is a health-promoting activity,” he added. “For some, it’s truly harmful.”

Several experts who were not involved in the research said the study’s findings were concerning. Those in favor of legalizing cannabis have argued that making the drug widely available would draw people away from the harms of alcohol, said Beatriz Carlini, a research associate professor in the psychiatry department of the University of Washington in Seattle.

But the study’s data, which shows only a slight decline in frequent alcohol use, suggests this has not been the case.

“It is disheartening,” she said.

Dr. Carlini and others noted that the concentrations of THC, the psychoactive component in marijuana, have increased dramatically over the years.

In 1995, the concentration of THC in cannabis samples seized by the Drug Enforcement Administration was about 4 percent. By 2021, it was about 15 percent . And now cannabis manufacturers are extracting THC to make oils, edibles, wax, sugar-size crystals and glass-like products called shatter with THC levels that can exceed 95 percent.

In the last decade , research has shown that frequent cannabis use — and particularly the use of high-potency products with levels of THC greater than 10 percent — is a risk factor for the onset of schizophrenia and other psychotic disorders.

“But that isn’t to say that use less frequent — monthly or yearly — is necessarily safe,” said Dr. Michael Murphy, an assistant professor of psychiatry at Harvard Medical School and a psychiatrist at McLean Hospital in Belmont, Mass.

“As we see higher rates of cannabis use in young people, I expect to see higher rates of psychotic disorders,” he said.

The risks of developing psychotic symptoms are higher for those who use cannabis before age 25, people who use it frequently, those with a genetic predisposition (for example, a parent or sibling with a psychotic disorder) or individuals who experienced stressful events like abuse, poverty or neglect during childhood.

In states that have legalized cannabis for recreational use, anyone 21 and over can purchase it.

Those who use cannabis frequently are also at risk of developing cannabis addiction as well as cannabinoid hyperemesis syndrome, a condition that causes recurrent vomiting, the experts said.

This latest study arrives on the heels of the Biden administration’s move last week to downgrade marijuana from the most restrictive category of drugs, known as Schedule I, to Schedule III, which includes drugs thought to have a low-to-moderate risk of abuse.

The survey did not collect information about the concentrations of THC in the products purchased by frequent users or note how often the respondents used cannabis each day.

“A lot of people go home and have a vape after work or take a gummy to go to sleep at night,” said Aaron Smith, the co-founder and chief executive of the National Cannabis Industry Association. He didn’t see that kind of casual daily use as a problem, he added.

At the same time, there may be young people who are using throughout the day “and are exposing themselves to a lot more THC than those people who are just taking a puff a day,” said Ziva D. Cooper, the director of the Center for Cannabis and Cannabinoids at the University of California, Los Angeles. “The mental health and the physical health outcomes are probably going to vary drastically when you look at those different groups of people.”

Christina Caron is a Times reporter covering mental health. More about Christina Caron