Health Canada
www.hc-sc.gc.ca
Common menu bar links
Institutional links
1994 Youth Smoking Survey - Technical Report
Help on accessing alternative formats, such as Portable Document Format (PDF), Microsoft Word and PowerPoint (PPT) files, can be obtained in the alternate format help section.
(PDF Version - 394 K)Smoking Behaviour
Edward M. Adlaf, PhD
Social Evaluation and Research Department
Addiction Research Foundation, Toronto
Susan J. Bondy, PhD
Social Evaluation and Research Department
Addiction Research Foundation, Toronto
Acknowledgements: We would like to thank Jürgen Rehm and Frances Hobbs (Addiction Research Foundation) for their contributions and Silvana Luciani (Health Canada) for her review of an earlier draft of this chapter.
Highlights
- Fifteen percent of Canadian youth aged 10-19 years are current smokers. This amounts to almost 400,000 individuals. Two percent are former smokers, while the majority, 83%, are classified as non-smokers (those who have smoked fewer than 100 cigarettes in their lifetime). In comparison, one third of Canadian adults are current smokers.
- Declining rates of smoking among Canadian youth aged 15-19 years not only have halted, but appear to be headed upward. Twenty-four percent of 15- to 19-year-olds in 1994 are current smokers, a rate significantly lower than the 29% in 1985, but higher than the rate of 21% in 1990. This upward trend in smoking has also been noted in several recent surveys.
- Seven percent of 10- to 14-year-olds, representing 138,000 Canadian youth, and 6% of 15- to 19-yearolds, another 118,000 youth, were beginner smokers at the time of the survey.
- Ten percent of youth aged 10-19 years, representing 387,000 Canadian adolescents, smoke daily - a level that makes it difficult for them to quit.
- Age differences in current smoking are notable: rates of current smoking increase from 3% among 10- to 12-year-olds to 29% of 18- to 19-year-olds. Among current smokers, the amount consumed is sharply higher among 15- to 19-year-olds (10.8 cigarettes daily) than among 10- to 14-year-olds (6.5 cigarettes daily).
- The period before 15 years of age, especially the period between 13 and 14 years, is a critical time for the adoption of smoking. First, the proportion of current smokers who smoke daily increases dramatically between 10 and 15 years of age and then remains stable. Second, the percentage of youth beginning smoking peaks at 13-14 years and declines thereafter. And, third, the resistance to smoking is lowest among 13- to 14-year-olds.
- Provincial differences in smoking are striking. Provinces with rates of smoking above the national average of 15% are Newfoundland (19%) and Quebec (18%), whereas provinces with lower than average rates are Ontario (13%) and Saskatchewan (12%). Other important provincial differences are that youth from provinces with the highest rates of smoking also demonstrate the highest perceived access to cigarettes and display the highest increases in smoking between the age groups 10-12 and 13-14. As well, the use of chewing tobacco and snuff is highest among youth from Saskatchewan and Alberta.
- Sex differences are generally minimal. Although current smoking does not differ between males and females (15% vs. 16%), male smokers smoke more cigarettes daily than do females (10.4 vs. 9.4), and males are more likely than females to have tried tobacco products other than cigarettes. At ages 13-14 and 15-17, females are more likely than males to be beginner smokers.
- Current smoking is highest among those reporting poor academic performance and those working more than 10 hours per week.
- These findings underline the importance of prevention programs and indicate that these programs should start early and be reinforced continuously throughout the adolescent years.
Methods
The variables of central interest in this chapter relate to smoking status and amount smoked. As described in detail in Chapter 2, two smoking status measures are employed. First, a three-category measure distinguishes current smokers (those who have smoked at least 100 cigarettes in their lifetime and had smoked during the 30 days preceding the survey), former smokers (those who have smoked 100 or more cigarettes but did not smoke during the 30 days preceding the survey), and non-smokers (those who had smoked fewer than 100 cigarettes in their lifetime by the time of the survey in the fall of 1994). Also, when sample size permits, a six-category measure is used in which current smokers are further divided into daily and non-daily smokers and non-smokers are further divided into beginners (those who have smoked fewer than 100 cigarettes in their lifetime and smoked during the month prior to the survey), past experimenters (those who have smoked fewer than 100 cigarettes but did not smoke during the month preceding the survey), and lifetime abstainers (those who have smoked less than one whole cigarette in their lifetime). In addition, among current smokers, two consumption measures are employed: the typical number of cigarettes smoked daily (presented in categories) and the number of cigarettes consumed during the seven days before the survey (presented as an average). Initiation into smoking was measured by the age when respondents reported smoking their first whole cigarette (see Chapter 2 for details on the above variables).
Several variables are employed to examine variation in smoking behaviours. As described in detail in Chapter 2, three different age groupings are used in this chapter: a two-category measure (10-14 and 15-19 years); a four-category measure (10-12, 13-14, 15-17, and 18-19 years); and, for total population estimates only, a 10-category measure (ages 10 through 19 years). The decision regarding which age grouping to use was based primarily on the need to ensure reliable estimates and to reduce data suppression. In addition to age, five other standard variables described in Chapter 2 are also employed: sex, province, language usually spoken at home, academic performance, and hours worked.
Six additional variables are used in this chapter. The first is a measure of self-identified Aboriginal status. Respondents were asked, "Are you an Aboriginal person, that is, a Native American Indian, Métis, or Inuit (Eskimo)?" (SS4, HH80, that is, School Survey Question 4 and Household Survey Question 80; see Appendix A). An additional smoking behaviour question examines the inhalation of cigarette smoke. Respondents who smoked during the 30 days preceding the survey were asked, "Do you inhale the cigarette smoke when you smoke (that is, breathe the smoke into your lungs)?" (SS20, HH12). As well, three questions are examined regarding accessibility of tobacco and contemplation of smoking. Never smoking respondents (defined as those who never tried smoking, never smoked a whole cigarette, or tried cigarettes only a few times but not during the 30 days prior to the survey) were asked, "Do you think it would be difficult or easy for you to get cigarettes if you wanted to try smoking?" (SS11, HH37). Contemplation of smoking was measured by the following two questions: never smoking respondents (defined as in the access question above) were asked, "Have you ever seriously thought about trying smoking?" (SS9b, HH35) and "Do you think you might try smoking within the next month?" (SS10, HH36). Finally, the association between smoking and accessible income was measured by the following question: "About how much money do you get each week to spend on yourself or save (include money from jobs, allowances or any other source)?" (SS67, HH78). All of the above questions were identical for the school and household survey components, and rates of missing data for the above items were below 2%.
As discussed in Chapter 2, all estimates have been examined to ensure a reasonable level of reliability based on the coefficient of variation. In this chapter, we will see that some of the finer distinctions in smoking status are not releasable or exhibit moderate sampling variability due to small numbers. This is especially apparent for beginner smokers (n = 1,625), former smokers (n = 299), and current non-daily smokers (n = 1,051). For this reason, the six-category smoking status classification is used sparingly. Also, as discussed in Chapter 2, the reader should recall that analyses in this chapter are primarily descriptive and that causal interpretations cannot be drawn from any associations described herein.
Results
Tobacco Use Behaviours
Prevalence of Tobacco Use: All Youth
Of all youth surveyed, 15% (580,000 young Canadians) are classified as current smokers, 2% (56,000) are classified as former smokers, and 83% (3.2 million) are classified as non-smokers (Table 3-1, at end of chapter). The largest category of current smokers is daily smokers (10% of all youth), followed by non-daily smokers (5%), while the 83% of non-smokers are composed of lifetime abstainers, those who have never smoked a whole cigarette in their lifetime (65%), past experimenters, those who have smoked fewer than 100 cigarettes in their lifetime but did not smoke during the 30 days preceding the survey (12%), and beginners, those who smoked during the 30 days prior to the survey and have not smoked more than 99 cigarettes in their lifetime (7%).
Cigarettes are by far the most commonly used tobacco product; 36% of youth report at least trying cigarettes at some point. Other than cigarettes, the most common tobacco products ever tried are cigars and pipes, tried by 20% of all youth, chewing tobacco, tried by 8%, and snuff, tried by 3% (Tables 3-A and 3-2). Rates of use of these tobacco products during the week preceding the survey are 3% or lower.
Regarding the number of cigarettes smoked, over half of current smokers (55%) smoke 10 or fewer cigarettes daily, while 31% smoke 11-20 cigarettes and 15% smoke 21 or more cigarettes daily (Table 3-3). On average, current smokers who smoked in the week preceding the survey smoked 9.9 cigarettes daily. Virtually all current smokers (99%) report inhaling cigarettes when smoking, a rate that remains unchanged regardless of background characteristics (data not shown).
Beginning to Smoke
The percentage classified as beginner smokers peaks at ages 13 and 14 (11% and 12%, respectively), but the relative increase in beginning smoking is most pronounced between 10 and 12 years of age. During this period, initiation increases by 300% (from 2% to 8%) (Table 3-1, Fig. 3-A). In contrast, smoking initiation between 12 and 14 years of age increases by only 50% (from 8% to 12%). It is also important to note that even small and moderate percentages represent a large number of youth. As seen in Table 3-B, 41,000 13-year-olds and 46,000 14-year-olds are beginner smokers, and the 10% of youth who smoke daily represent 387,000 Canadian adolescents.
Beginning smoking varies by sex (Fig. 3-B). Although beginning smoking does not differ between males and females aged 10-12 years (4%, respectively), rates among 13- to 14-year-old females are significantly higher than among their male counterparts (15% vs. 9%). This difference decreases with increasing age.
One means of evaluating the progression to regular smoking is by examining the percentage of current smokers who smoke daily. This percentage grows steadily from 11% among 11-year-olds to 75% among 15-year-olds. After 15 years of age, rates of daily smoking remain stable (Fig. 3-C).
The adoption of smoking can also be evaluated by examining the age at which the first whole cigarette was smoked. As seen in Figure 3-D, the age at which the first whole cigarette is smoked decreases with each age cohort. This shift toward starting smoking at younger ages is quite pronounced; whereas 30% of the 13-year-olds had smoked one whole cigarette by age 12, the same percentage was reached among 15-yearolds by age 13.5, and among 19-year-olds by age 15 (see Methodological Considerations for further discussion regarding this finding).
Figure 3 - A
Current and Beginning Smoking, by Age, Canada, 1994
| Population Estimate | ||
|---|---|---|
| Beginner Smokers | Current Daily Smokers | |
| # Data suppressed due to high sampling variability | ||
| Total, 10-19 | 256,000 | 387,000 |
| 10 | 6,000 | # |
| 11 | 14,000 | # |
| 12 | 31,000 | 5,000 |
| 13 | 41,000 | 17,000 |
| 14 | 46,000 | 22,000 |
| 15 | 30,000 | 46,000 |
| 16 | 22,000 | 66,000 |
| 17 | 22,000 | 55,000 |
| 18 | 28,000 | 75,000 |
| 19 | 16,000 | 99,000 |
Figure 3 - B
Beginning Smoking, by Age and Sex, Canada, 1994
Figure 3 - C
Current Smokers Who Smoke Daily, by Age, Canada, 1994
Figure 3 - D
Cumulative Frequency Distribution of Age First Smoked, Canada, 1994
Access and Contemplation
In addition to smoking consumption measures, the YSS also asked respondents about issues regarding access to tobacco and contemplation of its use. From a policy standpoint, the responses of never smokers are of great interest in these matters, since it is this group that is at risk of smoking adoption.
Just over half (56%) of never smokers report that it would be easy for them to get cigarettes if they wanted to try smoking (Table 3-C). Not surprisingly, age differences are sizeable, with 33% of 10- to 14-yearolds and 88% of 15- to 19-year-olds reporting easy access.
Interestingly, sex differences in perceived access are larger for never smokers aged 10-14 years (36% of males vs. 29% of females) than they are for 15- to 19-year-olds (88% of males vs. 87% of females).
Provincial differences in perceived accessibility of cigarettes are striking in several respects. In Figure 3-E, we display provincial differences in rates of current smoking among all 10- to 14-year-olds by the percentage of never smokers who perceive easy access to cigarettes. The horizontal and vertical lines in this display represent the national averages of 33% for easy access and 7% for current smoking. First, this display shows that the percentage of never smokers perceiving easy access to cigarettes ranges from 30% (Alberta and Ontario) to 38% (Quebec) and 39% (Newfoundland). Moreover, this display shows some association between rates of current smoking and access. Quebec shows higher than average rates of both current smoking (12%) and easy access (38%). Although youth from Newfoundland perceive higher than average access, their rate of current smoking is within 1 percentage point of the national average. Youth from Ontario are at the opposite extreme from youth from Quebec, displaying both lower than average rates of current smoking (4%) and lower than average easy access (30%). Although the association between rates of smoking and easy access is interesting and policyrelevant, we cannot interpret these data causally (see Methodological Considerations for a comment on this finding).
The two measures regarding contemplating cigarette smoking - ever having seriously thought about trying cigarettes and the likelihood of trying cigarettes within the month following the survey - also provide some context to consumption behaviours. Negative responses to these contemplation questions could be viewed as an indicator of resistance to smoking adoption. Among all never smokers, a sizeable majority report not thinking seriously about trying cigarettes or trying cigarettes within the month following the survey (both 88%). Figure 3-F shows that resistance to smoking is lowest among 13- to 14-year-olds. Smoking resistance does not vary significantly according to sex (89% of males vs. 87% of females) or province (with estimates ranging between 85% and 90%), but it does differ according to Aboriginal status (79% of Aboriginals vs. 95% of non-Aboriginals) and academic performance (75% with poor performance vs. 81% with average performance and 85% with above-average performance) (data not shown).
Population Subgroups
Age
Age is by far the most dominant factor related to smoking behaviour. Examining smoking according to each year of age shows strong age effects. In Figure 3-A, we see that rates of current smoking increase sharply among young adolescents, from under 5% among those aged 11 and 12 years to 33% among 19-year-olds. It is also evident that smoking status and other consumption measures show large differences according to age (Table 3-1). First, smoking status differs by age group, especially for current smokers (7% of 10- to 14-year-olds vs. 24% of 15- to 19-yearolds) and non-smokers (93% of 10- to 14-year-olds vs. 74% of 15- to 19-year-olds). The percentage reporting current smoking also represents a significant number of Canadian youth: 63,000 10- to 14-year-olds and 227,000 15- to 19-year-olds. Former smoking is rare and does not vary by age (1% of 10- to 14-year-olds vs. 2% of 15- to 19-year-olds).
In addition to being more likely to smoke, older youth smoke more cigarettes than do younger youth. On average, current smokers aged 10-14 years who smoked during the week preceding the survey consumed 6.5 cigarettes daily, compared to 10.8 cigarettes among those aged 15-19 years (Table 3-3).
Age differences in the percentage trying other tobacco products are negligible for cigars and pipes, chewing tobacco, and snuff (Table 3-2).
Figure 3 - E
Current Smoking, by Perceived Ease of Access to Cigarettes and Province, Age 10-14, Canada, 1994
Figure 3 - F
Never Smokers Not Considering Smoking, by Age, Canada, 1994
Figure 3 - G
Current Smoking, by Age and Sex, Canada, 1994
Figure 3 - H
Current Smoking, by Province, Age 10-19, Canada, 1994
Sex
Sex differences in smoking are generally minimal in the youth population; consequently, sex is not presented in all tables in this chapter. As seen in Table 3-4, 15% of males are current smokers, compared to 16% of females. Sex differences in current smoking, however, vary by age (Fig. 3-G). Between the ages of 13 and 17 years, females are slightly more likely than males to report current smoking, whereas among 18- to 19-year-olds, men are more likely than women to smoke (31% vs. 27%).
Although rates of current smoking do not vary by sex, overall, male smokers consume more cigarettes daily than do female smokers. Male smokers aged 10-14 years consume 7.3 cigarettes daily, compared to 5.7 cigarettes among females. The respective difference among 15- to 19-year-olds is 11.2 versus 10.3 cigarettes (Table 3-4).
As seen in Table 3-D, also absent is any sex difference among smokers for the age at which the first whole cigarette was smoked.
Males are much more likely than females to try cigars and pipes (26% vs. 14%), chewing tobacco (13% vs. 3%), and snuff (5% vs. 2%) during their lifetime (Table 3-2). Current use is also higher for males, but it is low for both sexes compared to cigarette use.
Province
Provincial differences in current smoking are evident in several respects (Table 3-5).
As seen in Figure 3-H, provinces with rates significantly above the national average of current smoking (15%) are Newfoundland (19%) and Quebec (18%), whereas those with below-average rates are Ontario (13%) and Saskatchewan (12%). Youth from the remaining six provinces report average rates of current smoking. This ranking in smoking according to province is similar for both 10- to 14-and 15- to 19-year-olds.
In Figure 3-I, we display the percentage of 10- to 19-year-olds who report current smoking and the average number of cigarettes consumed daily among smokers who smoked during the week preceding the survey. The horizontal and vertical lines in Figure 3-I indicate the national averages for the number of cigarettes (9.9) and current smoking (15%). (The two axes are based on different measurement scales, so we should be cautious about interpreting these data too literally, since the degree of sampling error is dissimilar.) Descriptively, however, this display serves as a useful means of visualizing the interplay between prevalence and amount smoked across provinces. The following are evident from Figure 3-I: (1) youth from four provinces, Manitoba, Alberta, Prince Edward Island, and Nova Scotia, show prevalence and consumption near the national average; (2) youth from British Columbia report average prevalence but lower than average consumption; (3) youth from Saskatchewan and Ontario display lower than average prevalence, those from Ontario show above-average consumption, and those from Saskatchewan display average consumption; (4) youth from New Brunswick show slightly above-average prevalence and aboveaverage consumption; and (5) youth from Quebec and Newfoundland show higher than average prevalence, with the former reporting average consumption and the latter reporting slightly below-average consumption.
Rates of smoking also differ by province in another respect. As seen in Figure 3-J, the relationship between current smoking and province differs by age, although smoking increases sharply with age in all cases. Most notably, we see the following: (1) rates of current smoking increase rapidly among youth from Newfoundland, rising from 2% among 10- to 12-yearolds to 38% among 18- to 19-year-olds (nationally, this increase is from 3% to 29%); (2) youth from Quebec show the most rapid increase in current smoking between the ages 10-12 and 13-14; however, after this point, increases in current smoking by age among Quebec youth remain nominal; and (3) increases in current smoking by age occur more slowly among youth in Ontario and Saskatchewan. Despite these different rates, however, smoking is fairly consistent across most provinces by the end of adolescence.
Provincial differences are also evident for the use of other tobacco products (Table 3-E). Youth from Saskatchewan and Alberta are significantly more likely than the average youth to report trying chewing tobacco (22% and 20% vs. 8% nationally) and snuff (7% and 7% vs. 3% nationally).
Language
Language usually spoken at home is another demographic characteristic associated with differences in rates of smoking (Table 3-6). Francophones report the highest rate of smoking (20%), followed by anglophones (15%) and speakers of other languages (6%, but with moderate sampling variability). Language differences, however, vary by age group.
Figure 3 - I
Current Smoking, by Amount Smoked and Province, Age 10-19, Canada, 1994
Figure 3 - J
Current Smoking, by Age and Province, Age 10-19, Canada,1994
Figure 3 - K
Current Smoking, by Age and Language, Canada, 1994
| # Data suppressed due to high sampling variability a Provinces ordered in increasing order of total tobacco use. |
|||
| Province | Cigars or Pipes (%) | ChewingTobacco (%) | Snuff (%) |
|---|---|---|---|
| Canada | 20 | 8 | 3 |
| Prince Edward Island | 17 | 5 | # |
| Newfoundland | 18 | 5 | # |
| Nova Scotia | 18 | 6 | 2 |
| Quebec | 18 | 3 | 5 |
| New Brunswick | 18 | 6 | 3 |
| Ontario | 20 | 7 | # |
| British Columbia | 21 | 10 | 3 |
| Manitoba | 23 | 10 | 3 |
| Alberta | 23 | 20 | 7 |
| Saskatchewan | 24 | 22 | 7 |
Figure 3 - L
Current Smoking, by School Grades and Age, Canada, 1994
As seen in Figure 3-K, current smoking increases dramatically between 10- to 12- and 13- to 14-year-old francophones (from 4% to 25%), with age differences remaining stable at 25% thereafter. In contrast, increases in current smoking among anglophones show a strong linear progression (2% > 10% > 19% > 32%), whereas smoking among speakers of other languages does not exceed 15% regardless of age. The relationship between age and current smoking is similar among francophone youth and youth from Quebec (Fig. 3-J). It is beyond the scope of this chapter, however, to identify the underlying source of this relationship.
We also see in Table 3-6 that, although speakers of other languages are least likely to smoke, such smokers smoke a greater number of cigarettes (12.0 cigarettes compared to 10.2 cigarettes among francophones and 9.7 cigarettes among anglophones). Finally, the age at which smokers consume their first whole cigarette does not differ significantly according to language (13.0 among anglophones, 12.6 among francophones, and 12.4 among other-language speakers) (data not shown).
Aboriginal Status
As discussed in Chapter 1, an important element of the YSS is the large sample size. Consequently, the smoking behaviours of groups whose sample size is usually too small in other surveys to be useful can be examined. Adolescent Aboriginals are one of these groups. Although 858 youth (4% of the total sample) identified themselves as Aboriginal, estimates based on small percentages (e.g., current smoking) can be unstable. We can, however, examine non-smoking status. As seen in Table 3-7, Aboriginal youth are less likely than non-Aboriginal youth to have never smoked more than 100 cigarettes in their lifetime (76% vs. 84%); among smokers, however, there is no dramatic difference in the number of cigarettes smoked daily (9.0 cigarettes among Aboriginals vs. 10.0 cigarettes among non-Aboriginals). Aboriginal smokers are, however, somewhat more likely than non-Aboriginal smokers to have smoked their first whole cigarette at an earlier age (11.6 vs. 12.9 years) (data not shown).
Academic Performance
In Table 3-8 and Figure 3-L, we see that rates of current smoking differ significantly by self-reported academic performance for youth attending school (94% of sample). Only 8% of those with aboveaverage academic performance report current smoking, compared to 15% of those with average grades and 25% of those with below-average grades. Differences in current smoking hold regardless of age group. Among 10- to 14-year-olds, rates of current smoking increase fivefold, from 4% among those with aboveaverage grades (moderate sampling variability) to 20% among those with below-average grades. The respective increase among 15-to 19-year-olds is from 13% to 36% (moderate sampling variability) - almost three times. Despite differences in the prevalence of smoking, among smokers there is no substantial difference in the number of cigarettes consumed daily (9.0 cigarettes among those with above-average grades, 8.7 among both those with average grades and those with below-average grades). Among smokers, however, those with poorer academic performance initiate smoking earlier. The average age of the first whole cigarette smoked is 11.5 among those with below-average grades, 12.7 among those with average grades, and 13.0 among those with above-average grades (data not shown).
Hours Worked and Income
In Table 3-9, we find that current smoking differs by the number of hours worked during the week preceding the survey for those working at a paid job, including babysitting, cutting lawns, and part-time work (51% of the total sample). Those who work 10 or more hours weekly are twice as likely as those working fewer than 10 hours to be current smokers (26% vs. 10%), and this difference occurs equally among those aged 10-14 years (13% vs. 8%) and those aged 15-19 years (30% vs. 16%). The amount of income available weekly is also associated with smoking status. As seen in Table 3-F, in each age group, current smokers report higher weekly incomes than do former smokers and non-smokers. Readers must recall, however, that the associations between smoking and other variables discussed in this chapter are purely descriptive. Without multivariate analyses, for instance, we cannot separate the influence of hours worked from income earned.
Discussion
Prevalence of Cigarette Smoking
Comparison with Other Surveys
Most adult surveys that address smoking report data for the 15-19 year age group, but the YSS is a rare example of national information for any younger ages (see Table 1-A). Figure 3-M shows the long-term changes in the prevalence of smoking among people aged 15-19 from Canadian national surveys (data for 1965 through 1993 taken from the summary by Hobbs et al.8). Prior to the mid-1970s, male youth were more likely to smoke than female youth; by 1975, however, about half of 15- to 19-year-olds of either sex were current smokers. From 1975 to 1988, the prevalence of smoking by youth of both sexes declined markedly, but this downward trend ended around 1988 or 1989.
| Amount of Money Received Per Week ($) | |||
|---|---|---|---|
| Current Smoker | Former Smoker | Non - smoker | |
| # Data suppressed due to high sampling variability | |||
| 10-12 | 33 | # | 23 |
| 13-14 | 46 | 29 | 32 |
| 15-17 | 52 | 36 | 39 |
| 18-19 | 114 | 96 | 85 |
Figure 3 - M
Current Smoking, by Sex, Age 15-19, Canada, 1965-1994
The 24% smoking prevalence estimate for 15- to 19-year-olds in the YSS is equal to that of the General Social Survey of 199113 and slightly higher than the figure of 21% for the 1990 Health Promotion Survey.11 This evidence of an increase in smoking is strengthened by the consideration of other sources. There are several other surveys around 1994 to which the YSS data may be compared. Cycle 1 of the Survey on Smoking in Canada7 reported that 27% of people aged 15-19 were current smokers. Canada's Alcohol and Drug Survey of 19946 also reported a slightly higher prevalence than the YSS for this age group, at 30%. Given that these other estimates put the prevalence of smoking in 1994 in this age group at least as high as the YSS estimate, the overall picture is that of an increase in the prevalence of smoking by youth. At a minimum, it is quite clear that there has been no overall progress in preventing youthful smoking since the late 1980s.
Figure 3 - N
Smoking Prevalence,+ by Grade, Ontario and United States, 1975-1995
This picture would be clearer still if national estimates existed for the intervening years, 1992 and 1993,14 but other types of studies do indicate increases in this period. There are few places in Canada where trends in smoking among school-age youth can be assessed (see Table 1-B). In Ontario, student drug use has been monitored since 1977.1 That study has found the prevalence of smoking (any use in the year preceding the survey) to have increased steadily between 1991 and 1995 for grade 7, 9, and 11 students (Fig. 3-N). Nova Scotia does not have annual data, but the changes in student smoking can be seen by comparing 1990 data with data from a repeat study in 1994.10 Students in grades 6, 8, 10, and 12 were included in the survey. The Nova Scotia study also shows a significant increase: in 1990, 15% of students smoked seven or more cigarettes weekly (and another 4% smoked less frequently), while in 1994 the percentage smoking at this rate was 20% (with 5% smoking less frequently). These results are paralleled in work from the United States. The Monitoring the Future Study conducted by the University of Michigan has also found increases in cigarette smoking (any use in the month preceding the study) among 8th graders, 10th graders, and 12th graders during the period from 1991 to 19959,16 (Fig. 3-N).
Methodological Considerations
A few methodological issues regarding findings in this chapter are worthy of brief mention. The first regards the shift toward starting smoking at younger ages. It should be noted that our analyses corrected for the fact that starting smoking is possible only in ages younger than the age at interview (see Gmel and Rehm5). Thus, the cumulative frequency distributions in Figure 3-D are censored (i.e., cannot be lower) at the age one year younger than the age at interview (e.g., for 10-yearolds at 9, for 19-year-olds at 18, etc.). It is possible that the mode of administration (school-based data collection for 10- to 14-year-olds, household-based for 15- to 19-year-olds) may have contributed to the differences; however, within each of the two components, younger cohorts also tended to start smoking earlier. Moreover, as noted in Chapter 2, there was no evidence that reports of smoking behaviours differed significantly between the household and school components. Another methodological effect may contribute to the differences: 19-year-olds may no longer remember that they tried a cigarette at age 11, whereas for a 12-year-old this event may be more salient and thus recalled. However, it seems unlikely that these methodological effects can fully account for the large differences in age of first cigarette. The most compelling conclusion is that the age of smoking initiation has decreased among Canadian adolescents.
The second methodological issue concerns the influence of age on smoking and, more generally, the limitations of cross-sectional studies. In studies such as the YSS, two age-related processes are at play: first, those related to maturational effects - influences due to adolescent maturational development - and second, those related to generational or cohort effects - influences due to growing up during different time periods. The descriptive analysis provided in this chapter does not seek to separate these two processes. More generally, with cross-sectional data, even when two variables are highly associated with one another, it is still difficult to establish causality. For example, although we found an association between rates of smoking and easy access (Fig. 3-E), we cannot interpret these data causally, since we cannot establish whether perceived ease of access causes smoking or whether smoking causes perceived ease of access. In any event, these data do point to the usefulness and potential of the YSS data for future policy research.
A third point to note is that the prevalence rates reported here are probably the lower bound of the true range, due to the likely direction of response bias (see Chapter 2) and the possibility that some past experimenters will become smokers at a later stage.
Implications for Regulation and Education
Knowledge of the prevalence of smoking among young Canadians, even knowledge that there is an increase in this prevalence, does not necessarily lead to knowledge about how to prevent initiation of smoking. Nor does this necessarily provide any insight into the process of becoming a long-term smoker as opposed to an early quitter or former experimenter. This chapter underscores how widespread is the problem of tobacco use by young Canadians.
Information presented in Chapters 4 through 9 provides many insights into the beliefs, attitudes, and behaviours of young people around smoking (such as knowledge of health effects and perceptions of why young people smoke). In addition, other parts of this report present information necessary to monitor and evaluate the effectiveness of preventive public policy, such as school-based smoking bans and regulations on sales to minors (see Chapter 10 for a summary).
The young people who participated in the 1994 YSS (born between 1975 and 1984) have grown up with the knowledge that smoking causes health problems, that smoking is addictive, and that it is very difficult for a smoker to quit. Major British, American, and Canadian reports that linked smoking to lung cancer were released between 1962 and 1964, more than a decade before any of these young people were born.4 Legislative efforts to protect them from the influence of cigarette advertising appeared in the early 1970s, and progressive measures to control sales of tobacco, increase prices, and reduce the number of locations where smoking is permitted have taken place (and received significant public debate) throughout the lives of these young people.
As later chapters show, most of these young people (75% of 10- to 19-year-olds; Chapter 7) have been exposed to some form of school-based health education about tobacco and are aware of the health consequences and addictiveness of smoking (Chapter 6). However, the current generation is still drawn to experiment with tobacco as previous ones were. Many of those adolescents who experiment with tobacco will become addicted.
The question remains as to how to design legislation that will keep tobacco out of the hands of children, as well as education campaigns that will reach the majority of young Canadians, convince them that the consequences of smoking are relevant for them, and teach them how to avoid smoking or quit.
Information like the YSS (and many other sources) needs to be examined to explain why access to tobacco is still high for young people (see Chapter 8) and to understand how to mobilize parents, teachers, retailers, and other adults to abide by tobacco restrictions for minors. Firmer legislative initiatives (including, notably, an increase in tobacco prices as well as effective control on distribution and use of tobacco15) need to be in place to help reverse this disturbing increase in smoking prevalence. As well, all of the legislative initiatives of the past 30 years need to be examined critically. While increasing restrictions have been placed on retailers and on smokers themselves, initiatives that fundamentally change the way tobacco is regulated and sold may be far more effective (i.e., restricting sale of tobacco to outlets where there is greater control, and fewer economic advantages to selling to minors). As well, one needs to examine carefully when legislative changes do not achieve their desired goals or are difficult to enforce. Healthy public policy should not focus only on more restrictive regulations, but also on ease of enforcement and public acceptance.
One striking finding is the differences seen in terms of smoking behaviour by province, as well as by language. In this regard, the YSS survey data have the potential to address several compelling questions. Study of the YSS data beyond this descriptive report should use more complex statistical approaches to address the relative importance of cultural and environmental factors and to determine if the observed provincial differences are modified by taking other differences into account. It is important to know to what extent variations in provincial tobacco policies influence provincial differences in the rates of smoking and reported access to cigarettes. Recent advances in hierarchical linear models3 and other methods that can link provincial-level data to individual-level smoking15 can set the stage for policy analyses that may have far-reaching public health implications. Such analysis may reveal that the policy mix that is most readily accepted, and most effective, may not be the same across the entire nation.
Another important observation in this chapter is the profound influence of age-related processes. The age period covered by the YSS includes the years of acquisition of smoking behaviour. We saw that the period under 15 years, particularly the ages 13 and 14, are especially critical to the adoption of smoking. The age of beginning smoking peaks among 13- and 14-year-olds and declines thereafter, and resistance to smoking is lowest among 13- to 14-year-olds. The percentage smoking daily is stable after 15 years of age. In all, there were about 250,000 young Canadians in the process of beginning to smoke at the time of the survey.
Why are these changes prominent among 13- to 14-year-olds? Is this age reflective of a proneness to smoking, or is it an indicator of adolescent emancipation into the adult world? A better understanding of the relationship between age and smoking would require longitudinal studies in order to separate the interactive effects of adolescent maturation from period effects and generational effects and to identify a temporal ordering of important influences. The YSS can, however, further research on other age-related issues discussed in this chapter. One striking finding of some interest is the prominent increase in smoking that begins early among francophone youth. Future multivariate analyses might be able to shed light on this matter.
More study of the stages of progression of smoking (first experimentation, repeated use, purchasing and carrying cigarettes, and the transition to regular smoking) along with normal age-specific developmental psychology is important for designing education campaigns that reach the majority of young Canadians, that address behaviours and beliefs that are important at that particular age, and that come together to complete a comprehensive program of prevention.
Implications for Future Monitoring and Further Research
Given the increases in rates of smoking and decreases in the age of onset, the monitoring of smoking among adolescents will provide critical information for prevention programs. Moreover, these findings illustrate the value of a sample that provides reliable data for single years of age and for each province. Indeed, the YSS findings add to existing epidemiological evidence of increases in smoking among youth. Although national trend data for the population aged 10-14 years are not systematically available (see Table 1-A), student surveys have documented increases in smoking among this age group. The reasons for this, however, are not clear. One partial explanation is the marked reduction in the price of cigarettes in early 1994, which appears to be related to an increase in smoking and tobacco sales.2,15 Still, it is also noteworthy that, in addition to smoking, use of other psychoactive substances has also increased in North America,1,9 which, in turn, poses the question whether increases in adolescent smoking are unique to this behaviour or whether increases in smoking are a manifestation of a larger adolescent social phenomenon. Further research is necessary to explore such a hypothesis.
Another prominent research issue is the finding that adolescents seem to initiate experimental smoking earlier than prior cohorts. This shift in age of onset is quite striking in the YSS. Only further cross-sectional surveys in combination with longitudinal work can assess the stability of this finding as well as the long-term implications of experimental smoking on future addiction. With the exception of women who reached adulthood in the years before it was socially acceptable for women to smoke, few people begin smoking after reaching adulthood. In absolute terms, the highest prevalence of smoking is typically observed among the young adult years, particularly the 20s, and cessation rates become more significant in the 30s. It is not clear what the effect of increased levels of experimentation and smoking among adolescents today will be on the rate of future generations of adult smokers. Clearly, there is an enormous potential for increases in the prevalence of smoking-related illness and deaths (see Figs. 1-C and 1-D).
Perhaps most striking in the YSS is the early development of smoking behaviour. Rates of beginning smoking increase dramatically between 11 and 13 years of age, yet few national surveys since 1978 have monitored smoking among this very group. Indeed, it is hard to imagine how the spectrum of smoking behaviours can be understood by ignoring this population. Also evident from the YSS is the importance of having data on smoking by single years of age rather than the more common and cruder age categories. Indeed, the study of the process of smoking adoption and the effective targeting of prevention programs require fine age distinctions.
Certainly, today's youth are different from those who smoked in large numbers during the 1970s. As a whole, these young people were born after the period of most common drug use and are the first generation of young people who have been educated about the health effects of tobacco (see Chapter 7) and other drugs since birth. It is important to understand what makes tobacco and drug use popular for this generation. The chapters that follow this one provide an insight into this group of young people in terms of their knowledge (Chapter 7) and their beliefs and attitudes (Chapter 6), as well as a number of methods of preventing and reducing smoking among young people (Chapters 8 and 9).
References
- Adlaf EM, Ivis FJ, Smart RG, Walsh GW. Ontario Student Drug Use Survey 1977-1995. Toronto: Addiction Research Foundation, 1995.
- Brown KS, Taylor TE, Madill CL, Cameron R. The Relationship Between the Tobacco Tax Decrease and Smoking Among Youth: Results of a Survey in South-western Ontario. Toronto: Ontario Tobacco Research Unit Working Paper Series No. 14, 1996.
- Bryk AS, Raudenbush SW. Hierarchical Linear Models. Newbury Park, California: Sage, 1992.
- Canadian Council on Smoking and Health. Chronology of Events in Tobacco Control, Canada. Ottawa: Canadian Council on Smoking and Health, September 28, 1995. [Available: http://www.ccsh.ca/ncth/docs/chroncae.html]
- Gmel G, Rehm J. Zum Problem der Schätzung des Alters beim Drogeneinstieg in Querschnittsbefragungen am Beispiel der Schweizerischen Gesundheitsbefragung. (The problem of estimating age of drug onset in cross-sectional survey research - The example of the Swiss Health Survey.) Sozial und Präventivmedizin 41 (in press).
- Health Canada. Canada's Alcohol and Drug Survey 1994. Analysis by authors of the Public Use Data Tape, April 1996.
- Health Canada. Survey on Smoking in Canada, Cycle 1, Fact Sheets, August 1994.
- Hobbs F, Ferrence R, Pope M, Poland B, Ashley MJ, Pederson L. Women and Cessation. Prepared for Health Canada, January 1996.
- Johnston LD, O'Malley PM, Bachman JG. National Survey Results in Drug Use from The Monitoring the Future Study, 1975-1994. Volume I. Secondary School Students. Washington, D.C.: National Institutes of Health publication 95-4026, 1995.
- Nova Scotia Department of Health. The 1994 Survey of Student Tobacco Use in Nova Scotia. Halifax: Smoke Free Nova Scotia, n.d.
- Pederson LL. Smoking. In Stephens T, Fowler Graham D (eds.), Health and Welfare Canada. Canada's Health Promotion Survey 1990: Technical Report. Ottawa: Minister of Supply and Services Canada, 1993.
- Siggner A. Special Study on Youth. Canada's Health Promotion Survey 1985, Technical Report Series. Ottawa: Minister of Supply and Services Canada, 1988.
- Statistics Canada, Housing, Family and Social Statistics Division. General Social Survey Analysis Series. Health Status of Canadians: Report of the 1991 General Social Survey. Ottawa: Minister of Industry, Science and Technology, 1994. Cat. No. 11-612-E, No. 8.
- Stephens T. Trends in the prevalence of smoking, 1991-1994. Chronic Diseases in Canada 1995; 16: 27-32.
- Stephens T, Pederson LL, Koval JJ, Kim C. The relationship of cigarette prices and smoke-free bylaws to the prevalence of smoking in Canada. American Journal of Public Health (in press).
- University of Michigan. Cigarette smoking among American teens rises again in 1995. Press release, December 15, 1995.
- Table 3-1
Type of Smoker, by Age, Canada, 1994 - Table 3-2
Lifetime and Past-week Prevalence of Other Tobacco Use, by Sex and Age, Canada, 1994 - Table 3-3
Number of Cigarettes Smoked, by Sex and Age, Current Smokers, Canada, 1994 - Table 3-4
Type of Smoker and Average Number of Cigarettes Smoked Daily, by Sex and Age, Past-week Smokers, Canada, 1994 - Table 3-5
Type of Smoker and Average Number of Cigarettes Smoked Daily, by Province and Age, Past-week Smokers, Canada, 1994 - Table 3-6
Type of Smoker and Average Number of Cigarettes Smoked Daily, by Language Usually Spoken at Home and Age, Past-week Smokers, Canada, 1994 - Table 3-7
Type of Smoker and Average Number of Cigarettes Smoked Daily, by Aboriginal Status and Age, Past-week Smokers, Canada, 1994 - Table 3-8
Type of Smoker and Average Number of Cigarettes Smoked Daily, by Academic Performance Compared with Peers and Age, 10- to 19-year- olds Attending School, Canada, 1994 - Table 3-9
Type of Smoker and Average Number of Cigarettes Smoked Daily, by Hours Worked Per Week and Age, 10- to 19-year-old Past-week Smokers Working at a Paid Job, Canada, 1994
