[Contents] [Next] [Previous]

The Scottish Health Survey 1998: Volume 1: Chapter 8

8 SMOKING

Richard Boreham

SUMMARY

  • A third of both men (34%) and women (32%) aged 16-74 reported that they smoked cigarettes.
  • Among those aged 16-64 there had been no significant change in the percentage of men who smoked cigarettes, but the percentage of women who smoked had decreased from 36% to 33% between 1995 and 1998.
  • Prevalence of cigarette smoking among men decreases with age between 16 and 64, but remains constant for women. There was a sharp decline in prevalence between those aged 55-64 and those aged 65-74 among both men and women.
  • Analysis of saliva cotinine levels suggests some level of under reporting of smoking behaviour, particularly among those aged 16-24.
  • Although for men, prevalence of cigarette smokers decreased with age, prevalence of heavy smokers increased with age up to those aged 35-44, remained constant up to age 55-64, then declined sharply among men aged 65-74. A similar pattern was found among women.
  • Among smokers, men were more likely than women to be classified as heavy smokers.
  • There was a strong social class gradient in prevalence of cigarette smoking, with those in manual social classes being much more likely to smoke than those in non-manual classes. Manual class smokers also smoked more cigarettes per day than smokers in non-manual classes.
  • There was no narrowing nor widening of the social class inequalities in cigarette smoking between 1995 and 1998.
  • Some children aged 8-12 had smoked a cigarette, but a negligible number had smoked in the previous week, or claimed to smoke cigarettes at least once a week.
  • 16% of boys aged 13-15 and 15% of girls had saliva cotinine levels of 15 ng/ml or higher which indicates regular smoking behaviour, but levels of self-reported smoking were much lower, which suggests substantial under-reporting of smoking behaviour in this survey among this age group.

 

8.1 INTRODUCTION

8.1.1 Government White Papers

The importance attached by the government to reducing levels of smoking is emphasised by the publication of two White Papers: 'Smoking Kills'1 which sets out the actions to be taken throughout the UK to reduce smoking and 'Towards a Healthier Scotland'2 which focuses on specific targets for Scotland. In the latter there are targets set for both adults' and children's smoking behaviour. The children's target is monitored by a biennial survey of smoking and drinking among secondary school children. The headline target set for children aged 12-15 is:

"Reduce smoking among young people from 14% to 12% between 1995 and 2005 and to 11% by 2010."

The targets for adult smoking are monitored through the Scottish Health Survey and are:

"Reduce rate of smoking from an average of 35% to 33% between 1995 and 2005 and to an average of 31% by 2010."

In addition to these targets, the government is committed to reducing inequalities in health, as demonstrated by the setting up of the Independent Inquiry into Inequalities in Health,3 and by references throughout the 'Smoking Kills' and 'Towards a Healthier Scotland' White Papers to the importance of reducing the social class gradients in the prevalence of smoking.

8.1.2 Smoking questions on the Health Survey

In 1998, as in 1995, information about cigarette smoking was collected from children aged 8-15 and adults aged 16 and 17 by means of a self-completion questionnaire, while for adults aged 18-744 it was collected as part of the main interview.

8.1.3 Cotinine

In 1995, cotinine levels in the Scottish Health Survey were measured in serum in adults, but in 1998 were measured in saliva for both adults and children. This change was made primarily because more informants agree to provide a saliva sample than a blood sample.

Cotinine is a metabolite of nicotine. It is one of several biological markers that are indicators of smoking (others include carbon monoxide and thiocyanate), and is generally considered the most useful. It can be measured in, among other things, saliva or serum. A saliva cotinine level of 15 ng/ml and over is taken as defining that the informant currently smokes (those who use other nicotine products are excluded). Cotinine has a half-life in the body of between 16 and 20 hours, which means that the level of 15 ng/ml in saliva will detect regular smoking (even if the person has not smoked that day) but will not detect occasional smoking if the last occasion was more than a day ago. A level of 15 ng/ml is also sufficiently high to avoid misclassifying people who are exposed to others' smoke as current smokers. Non-smokers who are exposed to heavy passive smoking would not be mis-classified as smokers using 15 ng/ml and over as the definition of a smoker.5

8.2 CURRENT SMOKING PREVALENCE AMONG ADULTS

Smoking prevalence is measured in two ways in the Scottish Health Survey. Informants are asked directly whether they smoke cigarettes, pipes or cigars nowadays, and cotinine levels in saliva are measured for those providing a saliva sample at the nurse interview. Throughout this chapter when referring to self-reported smoking status a distinction is made between 'current cigarette smokers' and 'current smokers'. The latter includes those who smoked pipes or cigars nowadays in addition to those who smoked cigarettes.

The measurement of cotinine levels in the Health Survey series provides an objective cross-check on self-reports of smoking behaviour, which are known not always to be accurate. Inaccuracies in reporting arise in part from difficulties informants may experience in providing quantitative summaries of variable behaviour patterns, but in some cases arise from a desire to conceal the truth from other people, such as household members who may be present during the interview.

Analysis of self-reported cigarette smoking prevalence was conducted using cross-tabulations and logistic regression models. Separate logistic regressions were run for men and women with the following independent variables: age, social class of chief income earner and region. The odds ratios shown in the tables are relative to average. An odds ratio of less than one means that the group was less likely than average to smoke cigarettes currently, and an odds ratio greater than one indicates a greater than average likelihood of smoking.

8.2.1 Smoking prevalence by age and sex

Around a third of informants, 34% of men and 32% of women, were current cigarette smokers (self-reported). A further 4% of men smoked pipes or cigars, but not cigarettes. Negligible numbers of women smoked only pipes or cigars. Women were more likely than men to report that they had never smoked (42% of women compared with 34% of men).

Table 8.1

Self-reported cigarette smoking prevalence among men was highest for those aged 16-24 (37%) and those aged 25-34 (39%). Prevalence then gradually decreased with age to 32% among men aged 55-64, then declined sharply to 20% among men aged 65-74. In the logistic regression model the odds of men aged 16-24 being a current cigarette smoker were 21/2 times as high as for men aged 65-74 (odds ratios relative to average 1.26 for men aged 16-24, 0.47 for men aged 65-74).

Tables 8.1, 8.12, Figure 8A

Figure 8A Prevalence and weight of smoking, by sex and age

Fig 8A

There was a less clear pattern of smoking prevalence by age for men when smoking of pipes and cigars are included in addition to cigarettes. This is predominantly because men aged between 35 and 74 were more likely to smoke only pipes or cigars than were men aged between 16 and 34. This has the effect of flattening the age gradient seen for cigarette smoking, such that overall smoking prevalence was between 38% and 42% for each age group up to age 64, but was still lower for men aged 65-74 at 25%.

Table 8.1

Women do not show the same age-gradient for prevalence of self-reported cigarette smoking as men. There was no clear relationship between age and cigarette smoking for women between 16 and 64, with prevalence varying between 31% and 36%, but smoking prevalence among women aged 65-74 was lower than in the younger age groups at 25%. The pattern was similar when social class and region were taken into account in the logistic regression model. Women aged 16-24 had odds of being a current cigarette smoker that were twice as high as for women aged 65-74 (odds ratios relative to average 1.17 for women aged 16-24, 0.61 for women aged 65-74). Pipe and cigar smoking among women was negligible, so prevalence of cigarette smoking and prevalence of smoking were very similar.

Tables 8.1, 8.12, Figure 8A

An adjusted smoking prevalence can be calculated by classifying people as smokers either if they have reported that they smoke cigarettes, pipes or cigars, or if they have saliva cotinine levels of 15 ng/ml or higher. The overall smoking prevalence and the pattern of smoking by age was different for the sub-sample of people who had a valid cotinine measurement, as people who claimed to smoke were less likely than non-smokers to have agreed to provide a saliva sample.

Table 8.3

When analysing the adjusted smoking prevalence, the sub-sample of those with a valid saliva cotinine measurement was weighted back to the smoking status profile of the whole sample by age and sex to correct for this bias in response.

When prevalence of smoking was adjusted for cotinine levels of 15 ng/ml or higher among informants with a valid saliva cotinine measurement, the proportion of men who smoked rose from 38% to 42%, and the proportion of women smokers from 33% to 35%. This suggests that there was some under-reporting of smoking behaviour.

Analysis of smoking prevalence adjusted for saliva cotinine measurements shows that there was differential under-reporting of smoking for different age groups among men. Men aged 16-24 were the group most likely to under-report - cotinine adjusted smoking prevalence was 11% higher than self-reported smoking prevalence for men aged 16-24. Under-reporting by men aged between 25 and 74 ranged between 2% and 4%. If this pattern of under-reporting among all smokers were similar for cigarette smokers (which it is likely to be as cigarette smokers make up the vast majority of smokers), then it would increase the age gradient in cigarette smoking prevalence described earlier in this section, with the 16-24 age group now having the highest prevalence at 50%.

Women aged 16-24 were more likely than other age groups to under report smoking _ the difference between reported and cotinine-adjusted smoking for women aged 16-24 was 7%. Under-reporting among women in other age-groups varied between 1% and 3%. As with the men, on this measure smoking prevalence was highest in the 16-24 age group (41%).

Table 8.5

8.2.2 Number of cigarettes smoked by age and sex

Although cigarette smoking prevalence decreases with age for men, the pattern for prevalence of heavy smoking by age was very different, as illustrated by Figure 8A. The proportion of all men who were heavy smokers increases with age from 6% of men aged 16-24 to 18% of men aged 35-44, remains around this level for men between 45 and 64, then drops to 8% among men aged 65-74. Thus the prevalence of men who were heavy smokers increases or remains constant between the age of 16 and 64, while the prevalence of men who smoke at all gradually decreases across the age groups.

Prevalence of heavy smokers among women shows the same pattern as for men. Prevalence increases from 4% among women aged 16-24 to a maximum of 16% among women aged 45-54, then decreases monotonically with age to 5% among women aged 64-74.

Table 8.1, Figure 8A

The difference between men and women cigarette smokers regarding the extent of heavy smoking is shown in Table 8.2. Among men, 42% of smokers were heavy smokers (defined as 20 or more cigarettes per day), compared with 35% of women smokers. Conversely women smokers were more likely to be light smokers (under 10 cigarettes per day): 22% of women smokers were light smokers compared with 18% of men smokers. These differential smoking patterns were reflected in the mean number of cigarettes smoked by men and women. The average for men smokers was 17.5 cigarettes per day, while for women smokers it was 14.9 cigarettes per day.

The mean number of cigarettes smoked by smokers increases with age from 12.2 cigarettes per day among men smokers aged 16-24 to 20.7 cigarettes per day among men smokers aged 45-54 and 55-64, then drops to 16.5 cigarettes per day for men smokers aged 65-74. Among women the peak occurs at an earlier age: women smokers aged 16-24 smoke 11.5 cigarettes per day, this rises with age to 17.3 cigarettes per day among women smokers aged 45-54, then decreases to 12.8 cigarettes per day for women smokers aged 65-74.

Table 8.2

8.2.3 Tar levels of cigarettes smoked by age and sex

Around half of informants who smoked, smoked cigarette brands rated as low to middle tar (10mg or more, but less than 15mg of tar per cigarette), 55% of men smokers and 56% of women smokers smoked low to middle tar cigarettes. Women smokers were more likely than men smokers to smoke low tar cigarette brands (31% compared with 17%). Low tar is defined as less than 10mg tar per cigarette.

Men smokers aged between 18 and 34 were more likely than men aged between 35 and 74 to smoke low to middle tar cigarettes; 62% of men smokers aged 18-24 and 25-34 smoked low to middle tar cigarettes, compared with between 46% and 52% of men smokers aged between 35 and 74. Men aged between 35 and 74 were more likely to smoke cigarettes with a high or middle tar content (15 mg/cigarette or higher) than men aged between 18 and 34.

Among women smokers, those aged 18-24 were the most likely to smoke low to middle tar cigarettes: 70% of smokers in this age group smoked low to middle tar compared with between 48% and 59% of women smokers aged between 25 and 74. Unlike men smokers, older women smokers were more likely than those aged 18-24 to smoke lower tar cigarettes _ the proportion of women smokers aged 25 to 74 who smoked a low tar cigarette varied between 30% and 38%, compared with 19% of women smokers aged 18-24.

Table 8.7

8.2.4 Smoking prevalence by social class of chief income earner

In order to ensure that comparison by social class was not confounded by the different age profiles of the individual social classes, the data were age standardised. The percentages presented in this section are age standardised, although the differences between observed and age-standardised percentages were small.

There was a strong social class gradient in prevalence of smoking among both men and women, with those in manual social classes being much more likely to smoke cigarettes than those in non-manual social classes. Among men, 12% in Social Class I were current cigarette smokers compared with 51% in Social Class IV and 45% in Social Class V. The pattern was the same for women -cigarette smoking prevalence was 11% in Social Class I and 56% in Social Class V. The same pattern was found in the regression models -the odds of being a current smoker for people in Social Class V were 61/2 times as high for men, and 12 times as high for women when compared with people in Social Class I.

Tables 8.8, 8.12, Figure 8B

Figure 8B Prevalence of cigarette smoking (age standardised), by social class of chief income earner and sex

Fig 8B

Smokers in manual social classes smoked more than those in non-manual social classes. For men there was a clear class gradient, the mean for men smokers in Social Class I being 12.4 cigarettes per day, compared with 19.1 cigarettes per day for men smokers in Social Class V. The class gradient among women smokers was less pronounced _ women smokers in Social Class II smoked the least number of cigarettes per day (mean of 12.6), while women smokers in Social Class IV smoked the most (mean of 16.8).

Table 8.8

8.2.5 Smoking prevalence by region

Analysis was age standardised to account for any regional age profile differences, and age-standardised percentages are presented in this section. Among men, those in the Highland and Islands showed the lowest prevalence (28% compared with 32% - 36% in other regions). Women in Borders, Dumfries and Galloway had a slightly lower prevalence of cigarette smoking (29%) than other regions. These findings were reflected in the odds ratios. However, region was not a significant predictor of cigarette smoking when age and social class were taken into consideration in the logistic regression model.

Tables 8.10, 8.12

8.3 TRENDS OVER TIME IN CIGARETTE SMOKING AMONG ADULTS

8.3.1 Trends in cigarette smoking by sex

As mentioned in Section 8.1, the White Paper 'Towards a Healthier Scotland' target is to reduce cigarette smoking prevalence from 35% in 1995 to 33% in 2005 and to 31% in 2010. Figure 8C shows the prevalence of cigarette smoking along with 95% confidence intervals for the 1995 and 1998 surveys for men, women and the total sample.

There was no significant change in the overall proportion of self-reported current cigarette smokers between 1995 and 1998, 35% of adults aged 16-64 smoked cigarettes in 1995 compared with 34% in 1998.

Cigarette smoking prevalence among men did not show a significant change between 1995 (34%) and 1998 (36%). There was evidence of a decrease in the prevalence of women aged 16-64 who smoked cigarettes from 36% in 1995 to 33% in 1998.

Table 8.13, Figure 8C

Figure 8C Trends in prevalence of cigarette smoking, by sex with 95% confidence intervals

Figure 8C

8.3.2 Trends in cigarette smoking by social class

People in manual social classes were more likely to be cigarette smokers than those in non-manual social classes, and this was true in both 1995 and 1998. 'Towards a Healthier Scotland' stresses the importance of reducing inequalities in health, by reducing the gap between manual and non-manual social classes in prevalence of cigarette smoking. Figure 8D shows how cigarette smoking prevalence has changed over time (with 95% confidence intervals) by sex and social class.

Figure 8D Trends in cigarette smoking by social class and sex, with 95% confidence intervals

Fig 8D

There was no significant change over time for men or women in either non-manual or manual social classes.

Figure 8D

Analysis of the prevalence of cigarette smoking among different sub-groups over time is not sufficient to determine whether there has been a change in the gap between manual and non-manual classes. However, logistic regression can be used to measure any change in inequality over time. Separate logistic regression models were run for men and women to determine whether there was any change in the social class gradient of prevalence of cigarette smoking between 1995 and 1998. Models were run with the following independent variables: age, social class (manual versus non-manual), survey year and an interaction of social class and survey year. Age was included in the model because there is a strong relationship between age and cigarette smoking, so including age would ensure that any change in the level of social class inequality would not be due to different age profiles in either survey year or the age composition of individual social classes.

The odds ratios for being a cigarette smoker for manual social classes in 1998, controlling for social class, survey year and age were not significantly different from 1 for either men or women. This indicates that there was no narrowing or widening of social class inequalities in cigarette smoking between 1995 and 1998.

Table 8.15

8.3.3 Trends in cigarette smoking by region

Prevalence of smoking among women aged 16-64 increased in Highland and Islands (from 31% to 41%), and decreased in Lothian and Fife (from 37% to 32%) but there were no significant changes among women in other regions. Among men there were no significant changes in cigarette smoking prevalence in any of the regions. Figures reported are not age-standardised.

Table 8.16

8.4 ADULT NON-SMOKERS

8.4.1 Passive smoking among adults

Adults who were not current smokers were asked about whether they were regularly exposed to other people's tobacco smoke in different places. The most common category of place where people were regularly exposed was in public spaces, reported by 55% of men and 46% of women. The next most common place was in someone's home, 31% of men and 35% of women reported that they were exposed to other people's smoke in their own or someone else's home. The least common place for people to report being exposed to tobacco smoke was at work (23% of men and 14% of women), and even among those who worked in the previous week the workplace was still the least common place to be exposed to tobacco smoke (31% of men and 21% of women, not shown). Women non-smokers (39%) were more likely than men non-smokers (33%) not to be regularly exposed to other people's smoke.

Older people were less likely to be exposed to other people's smoke than younger people; 87% of men non-smokers aged 16-24 were regularly exposed to smoke, compared with 46% of men non-smokers aged 65-74. A similar pattern was found for women non-smokers.

Table 8.17

8.4.2 Ex-regular cigarette smokers

Among men, those aged 18-34 were the most likely age group to have given up in the last six months (16% compared with 3% of ex-smokers in 35-54 and 55-74 age groups). A similar pattern was found for women. The majority of ex-smokers aged between 35 and 74 (both men and women) had given up over 10 years ago.

Table 8.18

8.5 CHILDREN'S LEVELS OF SMOKING

8.5.1 Children's self-reported experience of smoking

Around one in five children aged 8-15 claimed to have ever smoked a cigarette, 20% of boys and 23% of girls. Older children were much more likely to have smoked than younger children. Among those aged 14-15, 38% of boys and 48% of girls had ever smoked compared with 9% of boys and 7% of girls aged 8-9 (Note that in Table 8.19 percentages are detailed for each single year of age rather than for two-year age groups shown here).

Prevalence of regular or recent smoking of cigarettes was much lower. Four percent of boys aged 8-15 had smoked in the previous week and 3% reported smoking at least once a week, compared with 20% who had ever smoked a cigarette. The equivalent percentages for girls were 3%, 2% and 23%.

Thirteen was a key age in terms of smoking behaviour. Some children aged 8-12 had experimented with cigarettes, but a negligible number had smoked in the previous week, or reported to smoke cigarettes at least once a week. Thirteen was the age at which some children start to smoke more regularly. Among boys, only 1% of those aged 12 had smoked in the previous week, compared with 7%, 11% and 7% respectively for those aged 13, 14 and 15. The pattern for girls was similar, prevalence of smoking in the previous week rose from 1% of those aged 12 to 5% of those aged 13, 8% of those aged 14 and 12% of those aged 15. Some of the variation between ages is likely to be because individual age sample sizes are fairly small (between 114 and 153 unweighted).

Table 8.19

8.5.2 Children's cotinine levels

Saliva cotinine analysis also identified 13 as a pivotal age. At each age up to 12, among boys and girls 0%-3% had saliva cotinine levels of 15 ng/ml or more. This indicates very low levels of smoking among children in this age group. The proportions of those aged 13-15 with saliva cotinine levels of 15 ng/ml or more were higher (16% of boys and 15% of girls). (Note that in Table 8.20 percentages are shown for each single year of age rather than for the 13-15 age group reported here.)

A level of 15 ng/ml of cotinine in saliva would indicate either regular smoking, or that a cigarette had been smoked very recently. Among boys, 16% of those aged 13-15 had a saliva cotinine level of 15 ng/ml or higher, but only 8% said that they had smoked in the previous week _ the equivalent percentages for girls were 15% and 8%. The fact that the proportions of children aged 13-15 with saliva cotinine above the 15 ng/ml level were higher than self-reported cigarette smoking in the last week suggests that there was substantial under-reporting of smoking behaviour among those aged 13-15.

Table 8.19, 8.20

8.5.3 Comparison with survey of Smoking Among Secondary School Children

The target set in 'Towards a Healthier Scotland' for children's levels of smoking was to reduce prevalence from 14% to 12% among children aged 12-15 between 1995 and 2005. Smoking prevalence is defined as smoking at least one cigarette a week. This target figure is monitored using a survey of children in secondary schools in Scotland, with the interviewing being conducted in schools, using a questionnaire which focuses on smoking, drinking and use of drugs only. There are design features of the schools survey and the Scottish Health Survey which could affect reported smoking prevalence. Firstly the Scottish Health Survey asks questions about smoking within the context of a questionnaire about health, whereas the school based survey is purely about use of cigarettes, alcohol and drugs. Secondly the location of the interview could affect responses - the schools survey takes place with a child answering questions when surrounded by their peers, whereas in the Scottish Health Survey the child's parents or legal guardians are likely to be present. Despite the confidential nature of the self-completion booklet that children complete, the presence of parents or peers could affect children's honesty.

According to Scottish Health Survey data for 1998, 33% of children aged 12-15 have ever smoked, 12% had saliva cotinine levels of 15 ng/ml or higher, 6% reported that they had smoked in the last week, and 5% reported that they smoked cigarettes at least once a week. In the 1998 schools based survey, 12% of children aged 12-15 reported that they smoked at least one cigarette a week. This latter figure is likely to be the more accurate estimation of smoking prevalence among children, given that the cotinine results suggest under-reporting in the Scottish Health Survey.

Although for the reasons above, the Scottish Health Survey figures should not be used to determine whether the target in 'Towards a Healthier Scotland' is being met, it will still be possible to use future Scottish Health Surveys to see whether reported prevalence of cigarette smoking is decreasing among this age group.

8.5.4 Children's exposure to other people's smoke

All children aged 8-15, regardless of whether they smoked or not, were asked a similar question to adults (although with not as many potential places to be exposed to smoke), which makes the results not directly comparable.

Nearly two thirds of boys (63%) and girls (66%) aged 8-15 were exposed to tobacco smoke in their own or someone else's home, although the proportions exposed to smoke in their own home were lower at 44% and 40% for boys and girls respectively. There was no consistent variation in exposure to smoke in people's homes by age. Overall, 56% of boys and 57% of girls reported that they were exposed to smoke somewhere else, and this did show an age gradient; 40% of boys aged 8 were exposed to smoke elsewhere and this rose to 57% among boys aged 15. A similar pattern emerged for girls.

Prevalence of any exposure to cigarette smoke was similar among girls and boys aged 8-15, 85% of girls were exposed to smoke compared with 82% of boys. The proportion of boys exposed to smoke showed no clear pattern by age, but for girls, there was an increase in exposure to smoke at age 12, 92% of girls aged 12-15 were exposed to smoke, compared with 79% of girls aged 8-11.

Table 8.21

8.6 COMPARISONS WITH ENGLAND

8.6.1 Adults' comparisons with England

Among men, prevalence of cigarette smoking was higher in Scotland (34%) than in the North of England (30%), and the whole of England (30%), and men smokers smoked more cigarettes per day on average in Scotland (17.5 cigarettes/day) than in the North of England (16.8 cigarettes/day) and England (15.7 cigarettes/day).

Among women, 32% were a current cigarette smoker in both Scotland and the North of England, compared to 29% in England as a whole. Women smokers in Scotland (14.9 cigarettes/day) and the North of England (14.8 cigarettes/day) also smoked more cigarettes than those in England (13.7 cigarettes/day), although in all areas women smoked less cigarettes per day than men.

Table 8.22

8.6.2 Children's comparisons with England

Children's reported smoking behaviour was very similar in Scotland, the North of England and England. Around one in five children aged 8-15 had ever smoked, around one in twenty had smoked in the previous week, and less than one in twenty smoked cigarettes at least once a week. Evidence from cotinine analysis and the schools based surveys suggest that these reported figures under-estimate the true levels of smoking among children.

Tables 8.24, 8.25

[Contents] [Next] [Previous]