• Overview
• Definition
• Incidence, prevalence and trends
• Risk factors
• Health consequences
• Costs
• Contributors

Incidence, Prevalence and Trends

Smoking incidence

The incidence of a condition is the number of new cases arising in a population in a given time. In adolescence, people start and stop smoking frequently. ¹ The incidence would measure the number of starts. What counts as a new commencement of smoking: one cigarette? At least weekly smoking? Or the onset of nicotine dependence? How long does a person have to quit smoking for restarting to constitute a new incident case? In adulthood, a new incident case would be someone who has returned to smoking after attempting to stop. How long would the abstinence have to be to constitute a new incident case?

These problems make the incidence of smoking impossible to measure. Fortunately, knowing the incidence of smoking is of no obvious value for public health.

Smoking prevalence

Smoking prevalence in the UK is 0% until age 10–11 years. It rises through adolescence to a peak during the early twenties, and falls slowly with age thereafter. By the end of adolescence about 80% of people have tried smoking, but few class that as having been a smoker or having smoked regularly. ² The prevalence of regular smoking in adolescents aged 11–15 is monitored annually, and this shows steady rises with each year of age from 1% of 11 year olds to 20% of 15 year olds in England in 2005. ³

However, longitudinal studies in adolescence show that this rise is not due simply to adding new smokers. About 25% of regular smokers in one year will not be regular smokers by the next, so the incidence of regular smoking is higher than implied by year on year prevalence. ^{1 2 3 4} These data have shown that the prevalence of smoking in schoolchildren in England has remained roughly constant between 2000 and 2006 at around 10%, with the prevalence in girls being about 2% higher than in boys. ³

The data on smoking in adolescents are not collected in the same way or across the same age ranges in all countries of the UK. Allowing for this, however, the prevalence in the UK appears similar to that in England and similarly constant since 2000. ⁵

In every generation throughout the twentieth and twenty first centuries, the prevalence of smoking is highest among people in their twenties. ^{6 7} It declines as people age for three reasons:

People stop smoking

Almost nobody takes up smoking for the first time after age 25

Smokers have a higher death rate than non-smokers.

The third factor is an important cause of the declining prevalence only in the older cohorts.

Surveys of schoolchildren show consistently that the prevalence of smoking is higher for girls than boys. Data on peak smoking prevalence suggest that the higher smoking prevalence among adolescent girls probably reflects early maturation rather than a specific problem in young women.

Declining prevalence since 1970s

The dramatic fall in prevalence during the 1970s was mainly because there was a peak prevalence of around 70% in men in cohorts born before and around World War II. ⁶ These cohorts have now largely died and the prevalence is much lower in current cohorts.

Fortunately, this drop in peak prevalence seems to be continuing. The prevalence at age 20–24 in England declined from 40% in 1998 to 37% in men in 2004. In women it dropped from 40% to 30%. ³ Overall prevalence is falling at around 0.4% per year, which is too small to measure reliably from one year to the next. ³

Cigarette consumption

In England, men who smoke consume 16 cigarettes per day on average, while women who smoke consume 14 cigarettes per day. Some 34% of men and 15% of women smoke hand rolled cigarettes. The prevalence of hand rolled smoking has increased dramatically since 1980s, but it is concentrated in people with routine and manual working backgrounds.

Socioeconomic status

The prevalence of smoking varies by social class and did not change from 2001–5 in the UK. However, between 1921 and 1970 smoking went from being a behaviour of the educated upper classes to one concentrated in the lower classes. ⁹

Ethnic group

The prevalence of smoking varies by ethnic group, but much more for women than for men. The prevalence of smoking is strikingly low among women from Asian ethnic backgrounds. However, the HABITS study shows that the prevalence of regular smoking among Asian girls in school year 11 in south London was about 20%, which is similar to black girls (22%) but lower than in white girls (36%). ¹⁰ If these data are followed nationally the prevalence in Asian women will rise substantially over the next 30 years.

Concealment of smoking

The prevalence figures are self reported. While data show that people are usually honest when questioned about smoking, they may be more likely to deny their habit as smoking becomes seen as less desirable.

In the 2003 Health Survey for England the reported prevalence of smoking was 24% in men and 23% in women. However, levels of cotinine (a metabolite of nicotine) were above a threshold that indicates active smoking or chewing tobacco in 30% of men and 25% of women. ¹¹ The true prevalence of smoking may be higher than suggested and the decline in prevalence may, in part, be due to increased stigmatisation of smoking.

Smoking cessation

Wanting and intending to stop

Around 70% of the population want to stop smoking. This figure has not changed since 1998, when it was first included in surveys. ⁸ Most people give health and expense as the main reasons for wanting to stop smoking.

Some 76% intend to stop smoking at some point; only 12% intend to do so imminently (within a month), but more than half intend to in the next year, and nearly half expect not to be smoking in a year’s time. ⁸

Prevalence and incidence of attempts to stop

In surveys run by the Office for National Statistics people are asked whether they have ever tried to stop smoking. Over the years a consistent 80% of the British population report they have done so. ⁸ Of these, in 2005 27% reported they had tried to stop in the past year. However, the recent Smoking Toolkit Study found that annual surveys underestimate the prevalence of attempts to stop smoking. ¹³ Using different methods the study estimated that 46% of smokers try to stop in a year, and on average those that do make 1.7 attempts each. Expressed as an incidence rate, the incidence of attempts at cessation is 78 attempts per 100 smokers per year. ¹⁴

Hardcore smoking

“Hardcore smokers” are people who have not attempted to stop in the past year and do not want or intend to do so in the future. About 16% of smokers are hardcore smokers. ¹⁵ They tend to be older and more dependent, and do not believe smoking is influencing their health. Whether these beliefs are causes or consequences of their continued heavy smoking is not clear and the stability of their status as hardcore over time is uncertain.

Successful cessation

The incidence of successful lifetime quitting is hard to define. We can, however, measure the success of quit attempts using the Russell standard. In unaided quit attempts about 5% of smokers sustain abstinence for six months and 4% do so for 12 months. ¹⁶ But these success rates are higher with treatment (see Evidence sections).

The incidence of successful smoking cessation since the 1960s is estimated to be about 1–2% of smokers per year. ¹³ The current estimate for successful smoking cessation is twice as high, at 2–3%. This may reflect the greater imperative to stop smoking in society and the increased use of aids for cessation compared with the 1960s. ¹³

Smoking in pregnancy

About one third of women smoke before becoming pregnant and one half give up immediately before or during early pregnancy. Data on smoking in pregnancy in England are collected in the Infant Feeding Survey, which is based on interviews with women who have given birth. ³ Data may be unreliable because pregnant women who smoke are particularly likely to deny do so. ¹⁷

In 2000, 35% of mothers reported smoking before pregnancy, with 19% reporting smoking while pregnant. This dropped to 33% and 17%, respectively, in 2005. This implies nearly half of all pregnant women give up smoking for the pregnancy. Nearly all the remainder say that they try to cut down, but data indicate that the reduction in consumption occurs only in the first half of pregnancy and probably does not produce a worthwhile reduction in toxin exposure for the fetus. ¹⁸

Nearly all cessation occurs before pregnancy or in the very early days, with little occurring after the first antenatal contacts with healthcare professionals. ^{19 20 21} More than nine out of 10 women who stop smoking in pregnancy start smoking again afterwards. ²²