p,a,vn ting Tobacco Use Among Young People USDHHS 1989b). Since the recalled age at initiation is often 10 or more years younger than the age of the respondent at the time of the survey, recall bias may affect the reliability of these estimates. In the 1991 NHSDA, 69 percent of respondents aged 30 through 39 years reported trying a cigarette by age 18. Of all persons who had ever tried a cigarette, 88 percent had tried their first cigarette by age 18. The mean age of first trying a cigarette was 14.5 years. Thirty-five percent of the respondents had become daily smokers by age 18. Of those who had ever sinoked daily, 71 percent had smoked daily by age 18. The mean age of becoming a daily smoker was 17.7 years. Surveys conducted in 1991 among school-aged stu- dents, while lacking information on postadolescent ini- tiation, provide information of more recent initiation patterns (i.e., during the 1980s and early 1990s). Among 12th~grade students surveyed in 1991,22 percent of TAPS respondents, 40 percent of NHSDA respondents, 40 per- cent of MTFI' respondents, and 37 percent of YRBS re- spondents first tried a cigarette by age 14 (Table 8). About 60 percent of the respondents in the NHSDA, the MTFP, and the YRBS and about 50 percent of the TAPS respondents had smoked by their senior year. Daily cigarette use began by age 16 (or the 10th grade) for 18 to 23 percent of respondents to the NHSDA, the MTFP, and the YRBS (Table 9). By their senior year, 22 to 29 percent of these respondents had become daily smokers. Other Patterns of Smoking Two of the surveys gathered further information about smoking patterns-the number of days per month an adolescent smoked and the number of cigarettes the adolescent smoked per day. In the 1991 YRBS, responses indicated that in general, the greater number of days students reported smoking during the 30 days preceding the survey, the greater the number of cigarettes they smoked per day (Table 10). For example, 49 percent of students who smoked cigarettes on only one or two days during the preceding 30 days smoked fewer than one cigarette per day; among students who smoked ciga- rettes on all 30 days, 47 percent smoked 11 or more per day. Smoking patterns were also reported recently by Moss et al. (1992), using 1989 TAPS data (Table 11). About 41 percent of teenage smokers-whether male or female-smoked every day, and about one in four smoked on fewer than five of the preceding 30 days. The percentage of smokers who smoked every day increased with increasing age; 48 percent of 16- through l&year- old smokers smoked every day. About twice as many white as black teenagers smoked every day (42 vs. 22 percent), and blacks were more likely than whites to have smoked on fewer than five days. Non-Hispanics were more likely than Hispanics to smoke every day. Sixteen percent of 12- through l&year-old TAPS respondents who smoked during the week preceding the survey smoked 20 or more cigarettes daily. Males smoked more cigarettes daily than females. Older stu- dents smoked more cigarettes daily than younger stu- dents; 47 percent of 16- through l&year-old smokers and 11 percent of 12- and 13-year-old smokers reported smoking 10 or more cigarettes daily. Whites smoked more cigarettes daily than blacks, and non-Hispanics Table 10. Percent distribution of the number of cigarettes smoked per day, by the number of days on which cigarettes were smoked during the 30 days preceding the survey, Youth Risk Behavior Survey, United States, 1991 Number of days cigarettes were smoked 20 Total N 1-2 49.2 29.2 18.0 1.7 1.0 0.2 100 756 3-5 25.3 29.2 41.5 3.6 0.4 0.0 100 452 6-9 7.0 32.5 54.4 5.8 0.4 0.0 100 273 lo-19 7.4 13.0 66.5 10.8 1.8 0.4 100 326 20-29 0.7 4.6 61.4 27.9 5.4 0.0 100 294 30 0.1 0.3 26.5 26.0 36.6 10.8 100 803 Average 14.8 15.0 37.2 14.8 14.1 4.0 100 2,904 Source: Centers for Disease Control and Prevention, Division of Adolescent and School Health (unpublished data). Epidemiology 51 Surgeon General's Report Table 11. Percentage of current smokers by the number of days smoked during the past month and the average number of cigarettes smoked daily, by gender, age, and race/Hispanic origin, Teenage Attitudes and Practices Survey, United States, 1989 Number of days Number of cigarettes smoked during past month* smoked daily+ Every Category <5 5-9 10-29 day <5 5-9 l&19 220 Overall 24.1 8.7 26.4 40.8 37.9 20.4 25.7 16.0 Gender Male 23.9 8.5 26.6 41.0 33.9 19.3 27.6 19.2 Female 24.3 8.9 26.2 40.6 42.7 21.6 ' 23.5 12.1 Age (years) 12-13 51.9 8.3* 23.3 16.5t 64.3 24.6* 11.0* 0.0 14-15 28.4 9.8 34.5 27.3 55.5 17.2 23.0 4.3* 16-18 20.0 8.4 24.1 47.5 31.6 21.1 27.2 20.1 Race White 23.4 8.4 26.2 42.0 36.6 20.1 26.5 16.8 Black 37.0 15.0* 26.5 21.6 60.3 20.5t 16.3t 2.9* Hispanic origin Hispanic 30.7 11.2* 31.9 26.3 59.2 22.5 11.6s 6.6% Non-Hispanic 23.5 8.5 26.0 42.0 36.3 20.2 26.9 16.7 Source: Moss et al. (1992). *Excludes unknown number of days smoked. `Excludes unknown number of cigarettes smoked daily and none smoked during the past week. SEstimate does not meet standards of reliability or precision (< 30 percent relative standard error). were heavier smokers than Hispanics. Thus, not only were black and Hispanic adolescents less likely to smoke than whites, but those who did smoke, smoked fewer cigarettes each day than their white adolescent counter- parts. On average, persons 12 through 18 years old who smoked the week before the survey (N = 1,099) smoked 9 cigarettes each day. Males smoked IO cigarettes daily and females smoked 8. Whites averaged 9 cigarettes per day and blacks averaged 6 (1989 TAPS, CDC, Office on Smoking and Health [OSHI, unpublished data). The overall average for adult smokers is 19 cigarettes a day (CDC 1992aI. Initiation Continuum of Smoking The 1989 Surgeon General's report on smoking and health described the continuum of smoking be- havior as one that occurs in four stages: initiation, experimentation, regular smoking, and dependence or addiction (USDHHS 1989b). The report also ac- knowledged a preparatory stage that occurred before any initial smoking (Flay et al. 1983). These five stages are examined in detail in Chapter 4 (see "Develop- mental Stages of Smoking"). Data from the 1989 TAPS were used to create an initiation continuum similar to the smoking continuum for adults that was described in the 1989 Surgeon General's report (Pierceand Hatziandreu 199O;USDHHS 1989b). This initiation continuum incorporates mea- sures of smoking behavior and measures of the possibil- ity that a respondent will smoke in the future. In 1989, 54.5 percent of persons 12 through 18 years old reported that they had never smoked a cigarette, not even a few puffs (TabIe 12). These respondents were asked to report (1) whether they thought they would try a cigarette soon ("yes, " "no," and "don't know"), (2) whether they would 52 Epidemiology Prewnting Tobacco Use Among Young People Table 12. Percent distribution of an initiation continuum for cigarette smoking among persons aged 12-18 yeas, by age, gender, and race/Hispanic origin, Teenage Attitudes and Practices Survey, United States, 1989 Uptake continuum category Age (years) Gender Overall lZ-14 15-16 17-18 Male Female Race/Hispanic origin White/ Black/ non- non- His- His- panic panic Hispanic 1. Never tried smoking, not susceptible 44.3 2. Never tried smoking, susceptible 10.2 3. Tried smoking, not a whole cigarette, not susceptible 7.9 4. Tried smoking, not a whole cigarette, susceptible 3.3 5. Smoked l-99 cigarettes, 13.5 but none in the f&t 30 days, and not intending to smoke in a year 6. Smoked l-99 cigarettes, but none in the last 30 days, and might smoke in a year 7. Smoked 2 100 cigarettes, but none in the last 30 days, and not intending to smoke in a year 8. Smoked 2 100 cigarettes, but none in the last 30 days, and might smoke in a year 9. 10. 11. Smoked l-99 cigarettes, at least some in the past 30 days Smoked 2 100 cigarettes and smoked on l-19 days during the past 30 days Smoked at least 100 cigarettes and smoked on at least 20 days during the past 30 days 4.1 0.9 0.4 5.9 2.2 7.3 55.5 40.1 32.9 42.0 46.8 42.3 54.0 40.3 15.8 8.4 4.3 10.1 10.3 9.4 10.5 15.9 6:6 4.3 7.5 4.2 0.2 0.2 3.7 0.7 1.3 8.3 9.5 8.6 7.2 7.1 12.7 8.0 3.2 2.1 3.8 2.7 2.6 5.2 5.4 16.6 18.8 13.6 13.4 14.6 9.6 12.6 4.8 3.1 4.2 3.9 4.4 1.9 5.4 1.0 1.9 1.2 0.7 1.2 0.0 0.8 0.4 0.7 0.4 0.4 0.5 0.3 0.5 7.3 7.4 5.8 5.9 6.3 4.1 5.6 2.6 3.8 2.3 2.0 2.6 0.6 1.7 7.5 15.5 7.8 6.7 9.1 1.2 4.0 Source: Centers for Disease Control and Prevention, Office on Smoking and Health (unpublished data). Epidemiology 53 Surgeon General's Report smoke a cigarette if one of their best friends were to offer them one ("definitely yes," "probably yes," "probably not," "definitelynot," and "don't know"), and (3) whether they thought they would be smoking cigarettes in one year ("definitely yes," "probably yes," "probably not," "definitely not," and "don't know"). Never smokers who answered "no" to the first question, "definitely not" to the second question, and "definitelynot" to the third question were categorized as "not susceptible" to smoking. Those who answered these three questions in any other way were considered susceptible to smoking in the future (Pierce et al. 1993). According to these criteria, 44 percent of all TAPS respondents had never tried a cigarette and were not considered susceptible to smoking, and 10 percent had never tried smoking but were con- sidered susceptible. Adolescents who had tried smoking but had not smoked a whole cigarette accounted for 11 percent of TAPS respondents; 8 percent were judged to be not susceptible to smoking in the future, and 3 percent were judged susceptible. Those who had smoked at least one cigarette were only asked question 3, above, concerning whether or not they thought they would be smoking in a year. A large category (14 percent of all respondents) was composed of those who had smoked at least 1 but fewer than 100 cigarettes, who had not smoked in the preceding 30 days, and who definitely did not intend to smoke in a year. Another 4 percent had smoked from 1 to 99 cigarettes, had not smoked in the preceding 30 days, and were not definite in their resolve to not be smoking in a year. Slightly more than 1 percent of TAPS respondents had smoked at least 100 cigarettes but had not smoked in the preceding 30 days; these respondents are considered to be former smokers (USDHHS 1989b, 199Ob). Finally, among the 15 percent of respondents who smoked in the preceding 30 days, about 45 percent (6 percent of all respondents) had smoked fewer than 100 cigarettes in their lifetime. Although current smok- ers, these persons were still at a relatively early stage in the process of smoking initiation. Among those who had smoked at least 100 cigarettes and had smoked in the preceding month, more than three-fourths (7 percent of all respondents) had smoked on 20 or more of those 30 days. The distribution of this continuum was similar for males and females. White adolescents were more likely to be further along the continuum than were Hispanic and black adolescents. Cigarette Brand Preference Knowing what brands of cigarettes are preferred by young smokers may aid the development of smoking prevention programs and may provide insight into the influence that cigarette advertising may have on YOU% people. ln 19781980, the NHE assessed the brands of ciga- rettes most often used by current smokers (CDC, OSH, unpublished data). Among 707 respondents who were 18 or 19 years old, the most commonly used brands were Marlboro (37 percent), Kool (14 percent), Salem (10 per- cent), Winston (9 percent), Newport (8 percent), Virginia Slims (5 percent), Merit (4 percent), Benson & Hedges (3 percent), and Camel (2 percent). Ten percent of females and no males used Virginia Slims. Among whites, Marlboro (42 percent), Kool (10 percent), Winston (10 percent), Salem (8 percent), Virginia Slims (6 percent), and Newport (6 percent) were the most commonly used brands. Among blacks, Kool (46 percent), Newport (25 percent), Salem (20 percent), and Benson & Hedges (6 percent) were the most commonly smoked brands. In the 1989 TAPS, adolescent respondents who generally bought their own cigarettes were asked what brand they usually purchased. More than two-thirds of these smokers usually purchased Marlboro (Table 13). Preference for Marlboro did not differ appreciably by gender, Hispanic origin, age, or region of the country. White adolescent smokers were much more likely to smoke Marlboro cigarettes than were black adolescent smokers (71 vs. 9 percent). The next most popular brands, Newport and Camel, each accounted for only 8 percent of the overall population's preference. Black smokers, however, were much more likely to smoke Newport cigarettes than were white smokers (61 vs. 6 percent), although sample sizes of blacks were small. Smokers who resided in the Northeast and the Midwest were more likely to smoke Newport cigarettes than were smokers in the South and the West. Among white adolescents, Newport was more popular in the Northeast (14 percent) and the Midwest (7 percent) than in the South (1 percent) and the West (1 percent) (CDC 1992b). The Camel brand was more popular among male (11 percent) than female smokers (5 percent), among white (8 percent) than black smokers (3 percent), and among smokers residing in the West (18 percent) than among those residing in the other three regions (from 4 to 7 percent). Several nonnational studies conducted since the 1989 TAPS suggest that Camel cigarettes may be gaining in popularity among young smokers In a 1990 survey of ninth-grade students in 10 U.S. communities included in the Community Intervention Trial for Smoking Cessa- tion (COMMIT) evaluation, 43 percent of smokers who usually bought their own cigarettes bought Marlboro, 30 percent bought Camel, and 20 percent bought New- port (CDC 1992b). As TAPS data also indicated, adoles- cent smokers residing in communities in the western 54 Epidemiology prL~wting Tobacco Use Among Young People Table 13. Percent distribution of cigarette brands that 12-18-year-old current smokers* reported usually buying, by gender, race/Hispanic origin,' age, and region, Teenage Attitudes and Practices Survey, United States, 1989 Category Benson & Number Marlboro Newport Camel Winston Hedges Salem Kool Merit Vantage Other OverallS 865 Gender Male 477 FemaIe 388 Race White 807 Black 41 Hispanic origin Hispanic 46 Non-Hispanic 817 Age (years) 12-15 195 16-18 670 Region Northeast 184 Midwest 247 south 281 West 153 Overall market share, 1989 68.7 8.2 8.1 3.2 1.5 1.4 1.0 0.5 0.1 7.3 68.9 7.3 10.9 3.6 0.5 0.2 1.9 0.7 0.2 6.0 68.4 9.4 4.6 2.6 2.9 2.9 0.0 0.3 0.0 8.9 71.4 5.6 8.4 3.4 1.0 1.3 0.6 0.5 0.1 7.6 8.7 61.3 3.1 0.0 9.7 3.3 10.9 0.0 0.0 2.9 60.9 12.8 7.6 0.0 2.8 3.7 5.8 0.0 0.0 6.5 69.1 8.0 8.1 3.3 1.5 1.3 0.8 0.5 0.1 7.3 74.8 6.1 8.7 2.5 0.9 0.4 67.0 8.8 7.9 3.3 1.7 1.6 1.1 1.0 0.0 1.1 2.1 0.0 5.9 0.0 0.0 6.5 0.6 0.1 7.8 68.4 16.2 4:l 0.0 2.3 0.0 70.2 10.0 7.3 3.4 2.2 0.0 67.2 5.b 6.1 6.2 1.1 2.9 69.6 2.0 18.1 0.7 0.6 2.3 0.6 0.5 7.9 0.5 0.0 5.3 0.4 0.0 9.1 0.6 0.0 6.2 26.3 4.7 3.9 9.1 6.2 3.9 3.8 2.5 33.7 Sources: Centers for Disease Control (1992b); Maxwell (1992). *Persons who reported smoking on one or more of the 30 days preceding the survey. `Excludes the racial category "other" (N = 17). Ethnicity for two persons was unknown. `Data were weighted to provide national estimates. United States showed more preference for Camel ciga- rettes than did smokers from other regions of the nation. Other studies conducted after TAPS report rates of Camel preference among adolescent smokers that are consis- tent with the COMMIT survey results (DiFranza et al. 1991; Pierce, Gilpin, et al. 1991). In June and July 1992, the George H. Gallup Inter- national Institute 0992) conducted a telephone survey of a nationwide sample of 1,125 youths 12 through 17 years old. Smokers (those who reported having smoked at least one cigarette during the 30 days preceding the interview) were disproportionately oversampled, and the data were weighted to represent the adolescent population. Smokers were asked, `Thinking now about the last time you bought cigarettes for yourself, what brand did you happen to buy on that occasion?" Marlboro was the brand bought by 53 percent of these teenage smokers, Camel by 16 percent, and Newport by 8 percent. The most popular brand among blacks in this survey was Newport (54 percent preference). Epidemiology 55 Surgeon Generul's Reporf Trends in Cigarette Smoking Current Smoking Ever Smoking Data from the NTIS, the NHSDA, and the MTFP suggest that the prevalence of ever smoking among ado- lescents has declined since the 1970s (Table 14). In the NHSDA, the prevalence of smoking among youths 17 through 19 years old declined from 78 percent in 1979 to 64 percent in 1991, an average decline of 1.2 percentage points per year. In the MTFP, the prevalence among 17- and 18-year-olds decreased from 76 percent in 1977 to 62 percent in 1992, an average decline of 0.9 percentage points per year. In the NHIS, the percentage of 18- and 19-year-olds who had smoked at least 100 cigarettes dropped from 41 percent in 1974 to 25 percent in 1991, an average decline of 1 .O percentage points each year. NHIS data have been used to examine historical trends in smoking by reconstructing the prevalence of cigarette smoking for the decades in this century before systematic surveillance of cigarette smoking was con- ducted WSDHHS 1980,1985,1991b; Harris 1983). Us- ing information on a respondent's date of birth, age at initiation of fairly regular smoking, and duration of abstinence (for former smokers), the smoking status of the respondent can be assessed for any given year. For this report, the reconstructed prevalence of smoking among those aged 10 through 19 years is reported for the years 1920 through 1980. ,. Except for 1980, smoking during this 60-year pe- riod was more common among white and black ado- Iescent males than among white and black adolescent females (Figure I). The prevalence of cigarette smoking Figure 1. Trends in the reconstructed prevalence* of cigarette smoking among IO-19-year-olds, by gender and race, United States, 1920 -1980 25 1 m White males IIII~II Black males - White females - - Black females 0 I I I I I 1 1920 1930 1940 1950 1960 1970 1980 Year Source: U.S. Department of Health and Human Services (1991b). Data sources are the 1970,1978,1979,1980, and 1987 National Health Interview Surveys. The smoking prevalence for each of the years indicated was calculated for people who would have been 10-19 years old in each of those years by using the survey respondents' date of birth, age when they first began smoking regularly, and age when they quit smoking (see Appendix 2). 56 Epidemiology Pmwrlting TO~UCCO Use Among Young People Table 14. Trends in the prevalence (%I of ever smoking among young people, National Teenage Tobacco Surveys (NTIS), National Household Surveys on Drug Abuse (NHSDA),Monitoring the Future Projeci (MTFP), National Health Interview Surveys (NHIS), United States, 1968-1992 Year NITS' NHSDA' MTFPt NHIS_ 1968 1970 1972 1974 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 36.1 40.8 39.2 41.3 69.5 64.1 67.8 34.0 78.1 72.6 63.2 66.2 61.4 63.6 41.1 75.4 75.8 75.3 36.7 74.0 39.3 71.0 34.1 71 .o 70.1 70.6 34.5 69.7 68.8 29.8 67.6 67.2 26.2 66.4 27.7 65.7 64.4 27.6 63.1 25.3 61.8 Sources: NTTS: U.S. Department of Health, Education, and Welfare (USDHEW) (1972,1976,1979b); NHSDA: Centers for Disease Control and Prevention (CDC), Office on Smoking and Health (OSH) (unpublished data on 1974-1991 surveys); MTFP: Johnston, O'Malley, Bachman (in press); NHIS: CDC, OSH (unpublished data on 1974-1991 surveys). `NTTS, aged 17-18 years. Published reports (USDHEW 1972,1976,1979b) merge never smokers and experimenters (those who tried or experimented with smoking, but who had not yet smoked 100 cigarettes) into one category. By definition, therefore, the NTIS will underestimate the percentage of ever smokers. The trends, however, use the same definition. `NHSDA, aged 17-19 years. Those who reported in 1974,1976, and 1977 that they were current smokers and those who were not current smokers but who responded "yes" to the question, "Have you ever smoked cigarettes?" were classified as ever smokers for those years. For the years 1979 through 1991, ever smoking status was determined by response to the question, "About how old were you when you first tried a cigarette?" The prevalence of ever smoking is the complement of the response "Never tried a cigarette." tMTFP high school seniors, aged 17-18 years. Based on response to the question, "Have you ever smoked cigarettes?" `NHIS, aged 18-19 years. Based on response to the question, "Have you smoked at least 100 cigarettes in your entire life?" Those who had smoked at least 100 cigarettes by the time of the survey were classified as ever smokers. `Available information from published sources (USDHEW 1972,1976,1979b) do not permit exact comparisons with the 1989 TAPS data. Epidemiology 57 Surgeon General's Report remained higher among white adolescent males than among black adolescent males. Smoking prevalence gradually increased among white males during the six decades covered by the data. Among black males, preva- lence declined between 1950 and 1980. Among female adolescents, the reconstructed prevalence of current smoking increased steadily from 1920 through 1980; in 1980, the prevalence among fe- males surpassed that among males for the first time during the six-decade study period. Prevalence among white females has been higher than among black females since 1950. The data indicate a sharp increase in female smoking prevalence between 1970 and 1980. Trends in current smoking prevalence over the past two decades indicate that for both males and females, past-month smoking declined sharply in the late 1970s or, early 1980s (Table 15). Progress then slowed consider- ably, especially for males. In the MTFP surveys, the past- month smoking prevalence among males actually increased from 27 percent in 1981 to 29 percent in 1992; in the NHSDA and the THIS, male smoking prevalence was about the same in 1985 and in 1991. The prevalence among adolescent females in the MTFP and NHIS surveys was only slightly lower in 1991 and 1992 than in 1985; in the 1991 NHSDA, female smoking prevalence was about the same as in 1985. By the early 198Os, smoking was gener- ally more common among females than among males. By 1991, however, adolescent females and males had almost equivalent smoking prevalence. In all three surveys with information on race, the prevalence of current smoking declined during the late 1970s or early 1980s for both black and white older adolescents (Table 16). In the middle 197Os, current smoking was almost equally common among blacks and whites. At the end of that decade, black adolescents were less likely to be current smokers than white adoles- cents; this trend continued during the 1980s. Among white high school seniors in the MTF'P, current smoking was more prevalent in 1992 (32 percent) than in 1981(30 percent). In all three surveys, prevalence among older white adolescents was slightly higher in 1991 and 1992 than it was in 1985. Wallace and Bachman (1991) reported that white high school seniors were more than twice as likely as black high school seniors to report smoking in the past month, even after statistical control was made for factors such as parental education, number of parents living at home, urban or rural location, educational plans, academic per- formance, and religious attitudes and practices. MTJ!P trend data are available for daily smoking among racial and ethnic subgroups (Bachmanet al. 1991). In general, for Asian, black, white, Hispanic, and Ameri- can Indian male and female high school seniors, the prevalence of daily smoking declined from 1976-1984. The decline continued at a reduced rate during the late 1980s for most groups and ceased altogether among white males. Overall, the prevalence of daily smoking among high school seniors was 29 percent in 1976,21 percent in 1980, and 17 percent in 1992. Among males, the preva- lence was 28 percent in 1976,19 percent in 1980, and 17 percent in 1992; among females, 29 percent smoked daily in 1976, 24 percent in 1980, and 17 percent in 1992. Among whites, the prevalence of daily smoking de- clined from 29 percent in 1976 to 22 percent in 1980; the prevalence was 20 percent in 1992. Among blacks, the prevalence of daily smoking declined from 27 percent in 1976 to 16 percent in 1980 and continued to decline to 4 percent in 1992 (Bachman, Johnston, O'Malley 198Oa, 1981; ISR, University of Michigan, unpublished data). Data on smoking among the nation's high school seniors have also been reported as a function of parental education (NCHS 1993). Interestingly, the prevalence of past-month smoking decreased slightly from 1980 through 1991 among those seniors whose parents had completed fewer years of formal education and increased slightly during that period among those seniors whose parents had relatively more years of formal education. For example, among those seniors whose parents, on average, did not graduate from high school, the preva- lence of past-month smoking decreased from 33 percent in 1980 to 31 percent in 1991; among seniors whose parents graduated from high school, prevalence of smok- ing was 34 percent in 1980 and 29 percent in 1991. Among seniors whose parents had some postgraduate educa- tion, the prevalence of smoking was 24 percent in 1980 and 27 percent in 1991. Age or Grade When Smoking Begins The age at which people become regular cigarette smokers has been measured in national surveys con- ducted in 1955,1966, 1970,1978,1979,1980,1987, and 1988 (Haenszel, Shimkin, Miller 1955; NCHS 1970; USDHHS 1980, 1989b, 1991b; CDC 1991b). Data from the 1955 Current Population Survey (Haenszel, Shimkin, Miller 1955) suggest that during the first half of the century, people became regular smokers at progressively younger ages. The data for males are limited, however, because before 1974 many of the reports for men were provided by proxy respondents. To reduce proxy responses, Ahmed and Gleeson (NCHS 1970) limited their analysis of data from the 1966 Current Population Survey to females. These investiga- tors concluded that between 1955 and 1966, U. S. women began smoking at an earlier age. For the present report, the likelihood of having become a regular cigarette smoker by age 18 was deter- mined for females surveyed in the 1970,1978-1980, and 58 Epidemiology preventing TO~UCCO Use Among Young People Table 15. Trends in the prevalence (o/o) of current smoking* among young people, by gender, National Teenage Tobacco Surveys WITSl, National Household Surveys on Drug Abuse (NHSDA), Monitoring the Future Project (MTFP), National Health Interview Surveys (NHIS), United States, 1968-1992 Year N-l-l-S NHSDA MTFP NHIS Males Females Males Females Males Females Males Females (aged 17-18 years) (aged 17-19 years) (aged 17-18 years) (aged 18 -19 years) 1968 34.0 1970 37.8 1972 31.2 1974 32.6 1976 1977 1978 1979 19.6 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 21.0 24.1 26.0 26.4 47.8 35.1 39.0 38.7 52.0 47.2. 27.0 41.7t 41.7t 35.6 37.3 27.8 26.7 28.3. 32.9 28.9 20.2 27.0 27.0 37.7 36.7 34.5 31.2 26.8 26.5 26.8 28.0 25.9 28.2 27.9 27.0 28.0 27.7 29.1 29.0 29.2 39.1 39.7 38.1 37.1 33.4 31.6 32.6 31.6 31.9 31.4 30.6 31.4 28.9 29.0 29.2 27.5 26.1 36.9 30.8 30.6 33.5 29.5 34.2 24.9 27.8 23.3 31.4 20.1 24.5 21.6 20.9 19.6 23.1 21.7 18.0 22.0 20.6 Sources: N'ITS: U.S. Department of Health, Education, and Welfare (USDHEW) (1972,1976,1979b); NHSDA: Centers for Disease Control and Prevention (CDC), Office on Smoking and Health (OSH) (unpublished data on 1974-1991 surveys); MTFP: Bachman, Johnston, O'Malley (1980a, b, 1981,1984,1985,1987,1991); Johnston, Bachman, O'Malley (1980a, b, 1982, 1984,1986,1991,1992); Johnston, O'Malley, Bachman (1991a, in press); Institute for Social Research, University of Michigan (unpublished data); NHIS: CDC, OSH (unpublished data in 1974-1991 surveys). *For the NTTS, current smokers are those who state that they smoke less than one cigarette per week, one or more cigarettes per week, or one or more cigarettes a day (USDHEW 1979b). For the NHSDA and the MTFP, current smoking is defined as any cigarette smoking during the 30 days preceding the survey. For the NHIS, current smokers are those who report that they have smoked at least 100 cigarettes and who respond "yes" to the question, "DO you smoke now?" `The 1979 NHSDA determined current smoking status only for those respondents who had smoked at least 100 cigarettes (lifetime). The National Institute on Drug Abuse later published adjusted 1979 estimates using data from the 1982 NHSDA (Miller et al. 1983). The adjusted 1979 estimates used the ratio of the 1982 prevalence estimate, based on the 1979 definition, to the prevalence estimate based on the definition used in other years (i.e., any smoking in the last 30 days, regardless of whether the respondent had ever smoked 100 lifetime cigarettes). This table reports estimates based on the same adjust- ment procedure. rAvailable information from published sources (USDHEW 1972,1976,1979b) does not permit exact comparisons with the 1989 TAPS data. Epidemiology 59 Surgeon General's Report Table 16. Trends in the prevalence (X) of current smoking* among white and black young people, National Household Surveys on Drug Abuse (NHSDAI, Monitoring the Future Project (MTFP), National Health Interview Surveys (NHISI, United States, 1974-1992 NHSDA+ MTPP NHIS Year White Black (aged 17-19 years) White Black (aged 17-18 years) White Black (aged 18-19 years) 1974 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 41.9 47.4 43.0 47.2 42.9 44.3 44.4$ 37.7$ 39.2 20.9 28.6 20.8 33.0 17.6 28.3 7.2 30.5 11.4 38.3 39.7 38.4 34.4 37.0 31.5 34.9 28.7 31.0 25.2 30.1 22.3 31.3 21.2 31.3 21.2 31.0 17.6 31.7 18.7 32.0 14.6 32.2 13.9 32.3 12.8 32.1 12.4 32.5 12.0 31.8 9.4 31.8 8.2 33.6 33.7 33.3 26.3 32.6 30.8 26.1 29.0 28.6 18.5 23.4 18.4 23.4 15.3 23.7 9.4 22.2 10.3 24.9 7.6 Sources: NTTS: U.S. Department of Health, Education, and.Welfare (1972,1976,1979b); NHSDA: Centers for Disease Control and Prevention (CDC), Office on Smoking and Health (OSHI (unpublished data on 1974-1991 surveys); MTFP: Bachman, Johnston, O'Malley (1980a, b, 1981, 1984,.1985,1987,1991); Johnston, Bachman, O'Malley (1980a, b, 1982,19&k 1986,1991,1992); Johnston, O'MaIIey, Bachman (1992a); Institute for Social Research, University of Michigan (unpublished data); NHIS: CDC, DSH (unpublished data on 1974-1991 surveys). *For the NHSDA and the MTFP, current smoking is defined as any cigarette smoking during the 30 days preceding the survey. For the NHIS, current smokers are those who report that they have smoked at least 100 cigarettes and who respond "yes" to the question, "Do you smoke now?" `In the NHSDA, "white" and "black" include respondents of Hispanic origin, except for 1985. tithe 1979 NHSDA determined current smoking status only for those respondents who had smoked at least 100 cigarettes (lifetime). The National Institute on Drug Abuse later published adjusted 1979 estimates using data from the 1982 NHSDA (Miller et al. 1983). The adjusted 1979 estimates used the ratio of the 1982 prevalence estimate, based on the 1979 definition, to the prevalence estimate based on the definition used in other years (i.e., any smoking in the last 30 days, regardless of whether the respondent had ever smoked 100 Iifetime cigarettes). This table reports estimates based on the same adjust- ment procedure. 1987-1988 NHE (Figure 2). The data confirm that women in the United States have started to smoke at increasingly younger ages. The largest differences exist for women who were at least 45 years old at the time of the survey. The initiation curve for 18 through 24year-old females surveyed in 1987 and 1988 is, by age 18, lower than that for 18 through 24year-old females surveyed in 1978 through 1980, which is consistent with the notion that the prevalence of cigarette smoking has declined recently among young females (Table 15). Johnston, (YMalley, and Bachman (1992a) used retrospective reports from h4TFP high school seniors to describe trends in the initiation of daily smoking among seniors. Their data show that the likelihood of becoming a daily smoker at an earlier grade level increased sharply during the early to middle 1970s for the 1976 through 1978 senior classes. From 1975 through 1977, this likeli- hood decreased, and the grade of initiation declined or leveled for the 1979-1986 and 1988 classes. The lifetime 60 Epidemiology praenting Tobacco Use Among Young People Figure 2. Cumulative percentage of females becoming regular cigarette smokers by age 18, by age at time of survey, United States, 1970,1978-1980, and 1987-1988 25 z g 20 L 15 Age at time of survey: - 18-24 years 3 5 7 9 11. 13 15 17 Age when respondent began smoking regularly (years) - 30- Age at time of survey: 25- 35-44years 20- 2 2 & 15- c lo- 3 5 7 9 11 13 15 17 Age when respondent began smoking regularly (years) 25 1 Age at time of survey: 20- 55-64 years 2 15- 8 5 f lo- 3 5 7 9 11 13 15 17 Age when respondent began smoking regularly (years) 40- Age at time of survey: Age at time of survey: 25-34 years 30- 30- 25- 25- 2 2 8 20- 8 % % 20- a 15- a 15- 10- 10- 5- 5- 3 3 5 7 5 7 9 9 11 11 13 13 15 15 17 17 Age when respondent began smoking regularly (years) Age when respondent began smoking regularly (years) 30 30- 25 1 Age at time Age at time of survey: of survey: 25 - 45-54 years 45-54 years 20- 20- F F 8 15- 8 ti ti 15- f, f, lo- lo- 5- 5- 0 0 3 5 7 9 11 13 15 17 Age when respondent began smoking regularly (years) 12- 12- Age at time of survey: Age at time of survey: lo- lo- 2 65 years 2 65 years 8- 8- 5 5 2 2 6- 6- k k 4- 4- 2- 2- 0 0 3 3 5 7 9 11 13 15 17 5 7 9 11 13 15 17 Age when respondent began smoking regularly (years) 3 5 7 9 11 13 15 17 Age when respondent began smoking regularly (years) Age when respondent began smoking regularly (years) Source: National Health Interview Surveys 1970,1978,1979,1980,1987,1988, Centers for Disease Control and Prevention, Office on Smoking and Health (unpublished data). - 1987-1988 Surveys - 1978-1980 Surveys - - 1970 Survey Epidemiology 61 Surgeon Germal's Report prevalence of daily cigarette smoking at aII grade levels increased among the classes of 1989,1990, and 1991. Number of Cigarettes Smoked Each Day Trends in the intensity of smoking among MTFI' high school seniors indicate that since 1976, the propor- tion of heavy smokers (2 one-haIf pack per day) has decreased and the proportion of never smokers has in- creased (Figure 3). For example, in 1976,25 percent of high school seniors had never smoked, and 19 percent were heavy smokers; by 1992, 38 percent had never smoked, and 10 percent were heavy smokers (Bachman, Johnston, 0'MaIIey 1980a; ISR, University of Michigan, unpublished data). Attempts to Quit Smoking Cessation attempts are common among young smokers. In the 1989 TAPS, 74 percent of 12- through l&year-old smokers reported that they had seriously thought about quitting, 64 percent reported that they had tried to quit smoking, and 49 percent reported that they had tried to quit during the previous six months (AIIen et al. 1993). Nearly half of aII smokers among high school se- niors surveyed by the MTFP between 1976 and 1984 reported that they wanted to stop smoking (Table 17). Interest in quitting declined slightly thereafter. About 30 percent of current smokers reported that at one time in their lives they had tried but failed to stop smoking. About 40 percent of daily smokers reported that they had tried at least once to stop smoking but had faiied. The percentage of seniors who at some time had smoked regularly but had not smoked during the 30 days pre- ceding the survey (former smokers) increased sharply for males from 1977 through 1980 and for females from 1977 through 1981 (Figure 4). This measure declined sharply after 1980 for males and after 1981 for females. Table 17. Trends in high school senior smokers' interest in quitting smoking and attempts to quit smoking, by frequency of smoking during the past 30 days, Monitoring the Future Project, United States, 1976-1989 Respondents answering "Yes" Survey Question Do you want to stop smoking now? Among those who smoked at all during the last 30 days 1976 -1979 1980-1984 1985-1989 N (weighted) % N (weighted) % N (weighted) % 3,872 46.1 3,805 47.1 3,418 42.5 Among those who smoked 2 1 cigarette/day during the last 30 days 3,396 46.1 3,262 47.6 2,761 43.9 Have you ever tried to stop smoking and found that you could not? Among those who smoked at all during the last 30 days 4,740 31.5 4,942 31.4 4,534 27.8 Among those who smoked 3,604 38.5 3,464 41.6 2 1 cigarette/day during the last 30 days Source: Centers for Disease Control and Prevention, Office on Smoking and Health (unpublished data). 2,953 39.4 62 Epidemiology pmwnting Tobacco Use Among Young People Figure 3. Trends in the intensity of smoking among high school seniors, Monitoring the Future Project, United States, 1976-1992 45 1 40- 35- 30- 2 25- 8 Never smoked Have smoked but not in the past 30 days . . . . . . . . . . . . . . . . < 1 cigarette/day in the past 30 days ---- l-5 cigarettes/day in the past 30 days II- . . 2 one-half pack/day in the past 30 days 1984 Year 1986 1988 1990 1992 Sources: Bachman, Johnston, O'Malley (1980a, b, 1981,1984,1985,1987,1991); Johnston, Bachman, O'Malley (1980a, b, 1982, 1984,1986,1991,1992); Institute for Social Research, University of Michigan (unpublished data). Figure 4. Trends in the percentage of former smdkers among ever smokers,* by gender, high school seniors, Monitoring the Future Project, United States, 1976-1989 . . . . . . . Male Z?- - Female 16- 14 I 1 I I I I I I I I I I I 1976 1978 1980 1982 1984 1986 1988 Year Source: Centers for Disease Control and Prevention, Office on Smoking and Health (unpublished data). *Percentage of those who had ever smoked regularly who had not smoked during the previous 30 days. Epidemiology 63 161-235 95-3 The trend of cessation is similar to the trend for current smoking prevaIence. Substantial progress occurred in the late 197Os, but this progress slowed considerably in the 1980s. Trends in Knowledge and Attitudes About Smoking Trends in Perceived Health Risks of Smoking Data from the M'TFP allow comparisons of trends in beliefs about the risks associated with cigarette smok- ing and in actual smoking behavior. The decline in the prevalence of ever smoking has been associated with an increase in the percentage of high school seniors who believe that smoking one or more packs of cigarettes each day is a serious health risk (Figure 5). This associa- tion has been observed for both genders and for whites and blacks (Bachman, Johnston, 0'MalIey 198Oa, b, 1981, 1984, 1985, 1987, 1991; Johnston, Bachman, O'Malley 198Oa, b, 1982,1984,1986,1991; ISR, University of Michi- gan, unpublished data). For example, during the early 1980s the percentage of black high school seniors who felt that there is great risk associated with smoking a pack or more per day increased substantially. At the same time, the percentage of black youth who had smoked Surgeon General's Report at all and who had smoked daily declined rapidly. In 1989, over 50 percent of smokers and 74 percent of non- smokers reported that they believed that smoking a pack or more per day is a serious health risk 0989 MTFP, CDC, OSH, unpublished data). The percentage of seniors who believed that smok- ing entails a great risk to health increased from 56 per- cent in 1976 to 69 percent in 1991, and the percentage who believed that the health effects of smoking had been exaggerated decreased from 16 percent in 1981 to 14 percent in 1991 (Table 18). Nonetheless, 3 out of 10 seniors in 1991 still did not believe that heavy smoking poses a serious threat to health. Among 12- through l&year-olds in the 1989 TAPS, 32 percent believed that there is no harm in having an occasional cigarette; 57 percent of smokers in the survey endorsed that statement (Allen et al. 1993). Twenty-one percent of smokers and 3 percent of never smokers be- lieved that it is safe to smoke for only a year or two. Trends in Perceptions About Smoking The percentage of high school seniors surveyed by the MTFP who considered smoking a "dirty habit" in- creased between 198lf66 percent) and 1991 (72 percent) (Table 18). About 73 percent of white and 74 percent of black adolescents now feel this way, compared with only Figure 5. Trends in the percentage of high school seniors who believe that smoking is a serious health risk and in the percentage who have ever smoked, Monitoring the Future Project, United States, 1976-1991 - Ever smoked --/ 1 / , , , , , , t , ( ( / , 1 1976 1978 1980 1982 1984 1986 1988 1990 Year Sources: Bachman, Johnston, O'MaIley (1980a. b, 1981,1984,1985,1987,1991); Johnston, Bachman, O'MalIey (1980a, b, 1982, 1984,1986,1991,1992); Institute for Social Research, University of Michigan (unpublished data). 64 Epidemiology Preoenting Tobacco Use Among Young People 69 percent of whites and 54 percent of blacks surveyed in 1981 (Johnston, Bachman, 0'Malley 1982; ISR, Univer- sity of Michigan, unpublished data). The perception that smoking is a dirty habit has increased among males, females, smokers, and nonsmokers. Fii percent of smokers and 81 percent of nonsmokers classified smok- ing as a dirty habit in 1989 (Johnston, Bachman, O'Malley 1982,1984,1986,1991,1992;Bachman, Johnston,CYMalley 1984,1985,1991; 1981-1989 MTFI', CDC, OSH, unpub- lished data). Between 1977 and 1981, the percentage of seniors who felt that their close friends would not, or did not, approve of their smoking increased substantially (Table 18). The percentages reported for 1981 and 1991, however, were essentially identical. The percentage of senior who believed that adults should be prohibited by law from smoking in certain public places increased from 42 percent in 1977 to 45 percent in 1986 and remained about the same in 1991. TARS data on 12- through Wyear-olds provide further information on beliefs about smoking. In 1989, smokers were from two to five times more likely than never smokers to report that they believed that cigarette smoking helps people relax, reduce stress, feel more comfortable in social situations, reduce boredom, and keep their weight down (Allen et al. 1993). Smokers may also deny the addictive properties of cigarettes (LJSDI-IHS 1988b). TAPS data indicated that 39 percent of smok- ers-but only 11 percent of never smokers-believed that they would be able to quit smoking anytime they wanted. Trends in Perceptions About Smokers The overwhelming majority of high school seniors surveyed by the MTFP did not believe that cigarette smoking makes smokers their age look mature, in con- trol, or independent (Table 18). About half believed that smoking makes smokers look insecure, and more than 60 percent perceived cigarette smoking as something smokers use to try to look mature. between 1981 and 1991, smoking among seniors became less of the behav- ioral norm; fewer than 20 percent of seniors in 1991 reported feeling that smoking is an attempt to conform to such a norm. Responses to the MTFP indicate that the majority of high school seniors prefer to date nonsmokers and that this is becoming a trend. Since 1981, the propor- tion of respondents who prefer to date nonsmokers has increased by over 10 percent, to about 74 percent. The most substantial change occurred among black high school seniors (Figure 6). The percentage of white seniors who preferred to date nonsmokers increased only slightly. Over 85 percent of nonsmokers and Figure 6. Trends in the percentage of high school seniors who prefer to date nonsmokers, by race, Monitoring the Future Project, United States, 1981-1991 55 1 I I I I I I 1 I 1 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 Year Sources: Bachman, Johnston, O'Malley (1981,1984,1985,1987,19911; Johnston, Bachman, O'Malley (1982,1984,1986,1991, 1992); Institute for Social Research, University of Michigan (unpublished data). Epidemiology 65 surgeon General's Report Table 18. Trends in high school seniors' beliefs and attitudes about smoking and smokers, Monitoring the Future Project, United States, 1976,1981,1986,1991 Beliefs and attitudes 1976 1981 1986 1991 About smoking How much do you think people risk harming themselves if they smoke one or tiore packs of cigarettes per day?* (percentage who say great risk) The harmful effects of cigarettes have been exaggerated.+ (percentage who agree) Smoking is a dirty habit. (percentage who agree) How do you think your close friends feel (or would feel) about your smoking one or more packs of cigarettes per day?* (percentage who disapprove) Do you think that people (who are 18 or older) should be prohibited by law from smoking tobacco in certain specified public places? (percentage who say yes) About smokers 56.4 63.3 66.0 69.4 15.5 16.2 13.8 65.5 68.6 71.6 6&Y 73.9 76.2 74.3 42.P 43.0 45.1 44.9 In my opinion, when a guy my age is smoking a cigarette, it makes him look (percentage who agree) . . . like he's trying to appear mature and sophisticated 61.4 62.7 60.8 . . . insecure 42.0 43.6 47.9 . . . conforming 25.4 21.3 16.5 . . . rugged, tough, independent 8.6 9.9 9.8 . . . mature, sophisticated 5.3 4.6 5.0 . . . cool, calm, in control 6.2 5.5 5.3 Sources: Bachman, Johnston, O'MaUey (198Oa, 1987); Johnston, Bachman, O'Malley (198Oa, 1982); Institute for Social Research, University of Michigan (unpublished data). *Possible responses included "no risk," "slight risk," "moderate risk," "great risk," "can't say-drug unfamiliar." Percentages include those who say "great risk." `Possible responses included "disagree," "mostly disagree," "neither," "mostly agree," "agree." Percentages include those who "agree" or "mostly agree." Qossibie responses included "not disapprove," "disapprove," "strongly disapprove." Percentages include those who "disapprove" or "strongly disapprove." 51977 data. &i Epidemiology prtwzting Tobacco Use Among Young People Table 18. Continued Beliefs and attitudes 1976 1981 1986 1991 About smokers In my opinion, when a girl my age is smoking a cigarette, it makes her look (percentage who agree) . . like she's trying to appear mature and sophisticated . . . insecure . . conforming . . . independent and liberated . . mature, sophisticated . , . cool, calm, in control I prefer to date people who don't smoke. (percentage who agree) Smokers know how to enjoy life more than nonsmokers. (percentage who agree) I think that becoming a smoker reflects poor judgment. (percentage who agree) I strongly dislike being near people who are smoking. (percentage who agree) I personally don't mind being around people who are smoking. (percentage who agree) Do you disapprove of people (2 age 18) who smoke one or more packs of cigarettes per day? (percentage who disapprove) 64.6 65.0 64.1 47.4 49.5 52.0 26.5 21.7 19.5 11.2 9.5 9.6 6.9 5.4 4.5 5.5 4.5 4.1 66.5 71.0 74.0 2.8 2.4 3.6 57.0 59.3 61.0 38.2 36.9 33.1 65.9 70.0 75.4 71.4 45.4 48.9 about one-third of smokers preferred to date nonsmok- ers in 1989 (1989 MTF'P, CDC, OSH, unpublished data). Findings from the 1989 TAPS also suggest that few adolescents consider smoking a norm for their age group. Twothirds of 12- through Wyear-old respondents agreed with the statement, "Seeing someone smoking turns me off," and 86 percent (94 percent of never smokers and 51 percent of current smokers> preferred to date nonsmok- ers (Allen et al. 1993). Adolescents seem to be more concerned about people smoking around them. In the MTFP, the percent- age of high school seniors who strongly disliked being near smokers increased between 1986 (45 percent) and 1991(49 percent), and the percentage who reported that they did not mind being around smokers declined (from 38 percent in 198i to 33 percent in 1991) (Table 18). Males were consistently more likely than females to mind being around smokers (Johnston, Bachman, O'Malley 1982, Epidemiology 67 1984,1986,1991,1992; Bachman, Johnston, O'MalIey 1984, 1985, 1991; ISR, University of Michigan, unpublished data). The percentage of female seniors who.did not mind being around smokers changed little over time. From 1981 through 1991, the proportion of high school seniors who did not mind being around people who were smoking decreased by about 50 percent among blacks and by only 5 percent among whites (Figure 7). Smokers' acceptance of being around other smokers re- mained constant, at approximately 70 percent, from 1981 through 1989, whereas the percentage of nonsmokers who did not mind being around smokers decreased from 25 to 21 percent (1981-1989 MTFF surveys, CDC, OSH, unpublished data). Adult Implications of Adolescent Smoking Some notable findings regarding young peopIe's expectations to smoke, or to abstain from smoking, have emerged from the MTFP (see Johnston, O'Malley, Bachman 1992b). In their senior year, respondents who answered one of five questionnaire forms were asked, "Do you think you wiII be smoking cigamttes five years from now?" Overall, about 1 percent said they "definitely' would be smoking in five years, 14 percent said they "prob- ably" would, 27 percent said they probably would not, and 58 percent said they definitely would not flable 19). About 55 percent of past-month smokers and about 45 percent of daily smokers stated that they probably would not or defi- nitely would not be smoking in five years. Of the seniors in the fuII panel, 68 percent indicated that they had not smoked in the 30 days preceding the senior-year survey; 9 percent had smoked less than one Surgeon General's Report cigarette per day; 8 percent had smoked one to five ciga- rettes per day; 7 percent had smoked about one-half pack per day; and 8 percent had smoked a pack or more per day (Table 20). Five years after graduation, the same total proportion (32 percent) were past-month smokers. Some- what more (26 vs. 23 percent), however, were daily smokers. Further, for each smoking group defined by senior-year smoking level, those who continued to smoke increased their frequency of smoking (Tables 20-21). Of the respondents who were nonsmokers at the end of their senior year, 86 percent remained nonsmok- ers five to six years later, whereas only 13 percent of those who smoked one pack each day in their senior year became nonsmokers (Table 20). Those students who smoked one-half pack per day in their senior year were nearly as likely to continue use as were those students who smoked one pack daily; 81 percent of half- pack-a-day smokers &II smoked, and the majority of them increased their rate of smoking (Table 21). Seventy percent of respondents who in their senior year smoked one to five cigarettes per day continued to smoke five years later; most of these continuing smokers increased their rate of use. Even among the seniors who smoked the least &ass than one cigarette per day), 42 percent continued to smoke five to six years later, and two-thirds of these had increased their rate of smoking. When earlier smoking behavior is controlled, se- niors' expectations to smoke had very limited power to predict subsequent smoking behavior (Table 22). Many seniors who smoked one pack per day had expectations of discontinuing use. These expectations showed no relationship to the actual rate of smoking five to six years later. The same is true for those seniors who smoked Table 19. High school seniors predicting whether they will be smoking in five years, by smoking status in senior year, Monitoring the Future Project, United States, 1976-1986 senior classes Predicted likelihood of smoking in five years (o/o)* Senior year smoking status Definitely Probably Probably (use in past 30 days) will will will not None 0.4 1.3 21.0 < 1 cigarette/day 0.5 14.7 56.5 l-5 cigarettes/day 1.8 37.6 44.1 About % pack/day 0.6 57.7 30.3 2 1 pack/day 5.1 62.9 26.7 Total 0.9 14.2 27.0 Source: Institute for Social Research, University of Michigan (unpublished data). o Entnes are row percentages. Definitely Number will not (weighted) 77.3 1,926 28.3 248 16.5 211 11.3 197 5.2 228 58.0 2,810 68 Epidemiology prevating Tobaca, Use Among Young People Figure 7. Trends in the percentage of high school seniors who do not mind being around people who are smoking, by race, Monitoring the Future Project, United States, 1981-1991 5 5 5 5 5 5 .' ,.`"S `, 5 ?? o ?? ,.** `0.. ??? ??? ?? ,.*h ? ???? ???? ???? ??? ???? ??? ? ??? ??? ?? 0 I I I I I I I I I I 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 Year Sources: Bachman, Johnston, O'Malley (1981,19&Q, 1985,1987,1991); Johnston, Bachman, O'Malley (1982,1984,1986,1988, 1991,1992); Institute for sodal Research, University of Michigan (unpublished data). Table 20. Intensity of smoking (%) in senior year of high school, by intensity of smoking 5- 6 years later, Monitoring the Future Project, United States, 1976-1986 Smoking intensity (past 30 days) 5-6 years later (%)* Senior-year < 1 ciga- l-5 ciga- smoking intensity r&e rettes Number Column (use in past 30 days) None i&Y l&Y M pack r 1 pack (weighted) percentage None 85.6 4.9 2.6 2.7 4.1 9,238 67.6 < 1 cigarette/day 57.8 14.4 9.6 7.8 10.4 1,268 9.3 l- 5 cigaretes day per 29.6 8.8 17.2 20.5 23.9 1,058 7.7 About ?4 pack/ day 18.8 4.9 8.7 21.7 46.0 Lo(Jo 7.3 2 1 pack/day 13.4 2.7 4.1 10.1 69.7 1,100 8.1 Total 68.0 5.9 5.0 6.6 Source: Institute for Social Research, University of Michigan (unpublished data). *Entries are row percentages. 14.6 13,665 100.0 Epidemiology 69 Surgeon General's Report Table 21. Direction of change in smoking behavior (%) between senior year of high school and 5 - 6 years later, Monitoring the Future Project, United States, 1976-1986 senior classes Senior-year smoking status (use in past 30 days) Quit Smoking status 5-6 years later" Number Less use Same level More use (weighted) None 85.6 14.4 9,238 < 1 cigarette/day 57.8 14.4 27.8 1,268 l-5 cigarettes/day 29.6 8.8 17.2 44.4 1,058 About `/z pack/day 18.8 13.6 21.7 46.0 1,000 2 1 pack/day 13.2 17.7 40.2 29.0 869 Source: Institute for Social Research, University of Michigan (unpublished data). *Entries are row percentages. Table 22. Smoking intensity 5-6 years after high school, by senior-year smoking status and expectation to smoke in 5 years, Monitoring the Future Project, United States, 1976-1986 senior classes Senior-year smoking intensity (use in past 30 days) and predicted likelihood of smoking in 5 years None Smoking intensity (past 30 days) 5-6 years later* < 1 cigarette l-5 cigarettes 2 1 pack Number /day /day 112 pa&day /day (weighted) None Will smoke Will not smoke Total < 1 cigarette/day Will smoke Will not smoke Total l-5 cigarettes/day Will smoke Will not smoke Total About l/z pack/day Will smoke Will not smoke Total 2 1 pack/day Will smoke Will not smoke Total 55.3 10.6 19.8 8.3 5.9 30 84.7 5.6 2.9 2.5 4.3 1,829 84.2 5.7 3.2 2.6 4.3 1,859 41.7 18.4 19.5 14.0 6.4 36 58.4 14.7 9.7 9.7 7.5 208 55.9 15.2 11.1 10.4 7.3 244 32.3 3.0 15.5 23.0 26.2 83 31.8 5.8 15.9 23.0 23.5 125 32.0 4.7 15.7 23.0 24.6 208 15.5 4.9 6.5 21.0 52.1 115 17.6 2.5 6.5 21.1 52.3 81 16.4 3.9 6.5 21.1 52.2 196 13.3 2.2 3.2 9.6 71.8 153 13.2 1.6 5.3 6.3 73.6 72 13.3 2.0 3.8 8.5 72.4 225 Grand Total 67.0 6.0 5.2 6.6 15.2 2,731 Source: Institute for Social Research, University of Michigan (unpublished data). *Entries are row percentages. 70 Epidemiology Preventing Tobacco Use Among Young People one-half pack-or even as little as one to five cigarettes- per day in high school. Expectations were predictive only for those smokers who smoked less than one ciga- rette per day; 58 percent of those who thought they probably or definitely would be smoking in the future did, in fact, continue to smoke, whereas only 42 percent of those who did not expect to smoke in the future did smoke. Among seniors who had never smoked, less than 2 percent thought they would be smoking in five years (Table 19). This small group did, in fact, have a higher rate of subsequent smoking (45 percent) than never smokers who did not expect to be smoking in five years (15 percent) (Table 22). Thus, the expectation to avoid smoking seemed to make some difference among nonsmokers and very light smokers in high school, although very few seniors in these groups reportecl an expectation to smoke. On the other hand, among light, moderate, and heavy daily smokers, the expectation to abstain from smoking in the future seemed overwhelmed by the strong forces that tend to maintain or advance smoking behavior once it is established. One implication of these results is that young people should be made aware of the strongly addictive nature of nicotine and its ability to overwhelm future good expectations. Clearly, prevention is the major goal, but immediate cessation is of aitical importance for adolescents, even for those who smoke very little in high schooL Smoking and Other Drug Use In Chapter 2, tobacco use is discussed as a possible predictor of other drug use (see "Smoking as a Risk Factor for Other Drug Use" and "Smokeless Tobacco Use as a Risk Factor for Other Drug Use"). The present chapter presents detailed information on high school seniors' usage patterns for cigarettes, alcohol, marijuana, cocaine, inhalants, and smokeless tobacco. Both preva- lence of past-month use and comparisons of the self- reported age at first use of each will be presented. Prevalence of Smoking and Other Drug Use Among high school seniors in the MTFP studies, the majority of alcohol users (60 percent) and smokeless tobacco users (57 percent) did not smoke (Table 23). The majority of marijuana (62 percent), cocaine (68 percent), and inhalant (56 percent) users smoked cigarettes. Ciga- rette smoking prevalence was from 1.9 to 3.9 times higher among users of these drugs than among nonusers. Although most drinkers (60 percent) did not smoke, almost all smokers (88 percent) were drink- ers. Almost one-half (45 percent) of cigarette smok- ers were also marijuana smokers, 11 percent were cocaine users, 5 percent used inhalants, and 33 per- cent used smokeless tobacco (which will be discussed separately later in this chapter). The prevalence of Table 23. Prevalence (%) of cigarette smoking among users of other drugs and prevalence of other drug use among smokers, + high school seniors, Monitoring the Future Project, United States, 1985-1989 Prevalence of Prevalence of smoking among smoking among Prevalence of Prevalence of users of other nonusers of drug use among drug use among Other substances drugs other drugs smokers nonsmokers Alcohol 40.0 10.3 87.6 54.8 Marijuana 62.1 20.3 44.9 11.2 Cocaine+ 68.1 27.2 10.9 2.1 Inhalantst 56.1 28.5 4.8 1.5 Smokeless tobacco5 43.0 22.4 32.5 15.6 Source: Centers for Disease Control and Prevention, Office on Smoking and Health (unpublished data). *Any use of cigarettes or other drugs during the past month. `Includes "coke," "crack," and "rock." %lue, aerosols, laughing gas, etc. SMales only, 1986-1989 senior classes only. Epidemiology 71 Surgeon General's Report other drug use was from 1.6 to 5.2 times more preva- lent among cigarette smokers than nonsmokers. Grade When Smoking and Other Drug Use Begins MTFP data from 1986 through 1989 were merged to observe the grade at which seniors reported trying cigarettes, smokeless tobacco, alcohol, marijuana, and cocaine (Figure 8). Among ever smokers, 31 percent hieci their first cigarette by the sixth grade, and 61 per- cent first smoked by the eighth grade. Among those who had used smokeless tobacco, 23 percent had first done so by the sixth grade, and 53 percent by the eighth grade. Proportionately fewer users of alcohol, marijuana, and cocaine initiated use as early as respondents initiated use of cigarettes and smokeless tobacco. Thirty-four percent of alcohol users, 26 percent of marijuana users, and 6 percent of cocaine users first tried these drugs by the eighth grade. By the 12th grade, only 8 percent of MTFP respon- dents had not tried cigarettes or alcohol; 68 percent had tried both, and 24 percent had tried alcohol but not cigarettes (Table 24). Of those students who had tried both cigarettes and alcohol by 12th grade, almost half (49 percent) had tried cigarettes before trying alcohol; 33 percent had tried both at about the same time. About 30 percent of all students had not tried ciga- rettes or marijuana by the 12th grade (Table 25); 44 percent had tried both, and 22 percent had tried cigarettes but not marijuana. Of those-who had tried both by 12th grade, most students (65 percent) had tried cigarettes before mari- juana; 23 percent had tried both at about the same time. About one-third of seniors (34 percent) had not tried cigarettes or cocaine; 12 percent had tried both, and over half (53 percent) had tried cigarettes but not cocaine (Table 26). Of those who had tried both by 12th grade, 90 percent had tried cigarettes before trying cocaine, and 9 percent had tried both at about the same time. These data support the contention that tobacco use falls early in the sequence of drug use for young adoles- cents and therefore may be considered a "gateway' drug. Figure 8. Grade when respondents (high school seniors) first tried cigarettes, smokeless tobacco, alcohol, marijuana, and cocaine, among respondents who had ever used these substances by grade 12, Monitoring the Future Project, United States, 1986-1989 100 I- 60 2 8 k L 40 Grade 12 Cigarettes Smokeless tobacco Alcohol Marijuana Cocaine Source: Centers for Disease Control and Prevention, Office on Smoking and Health (unpublished data). 72 Epidemiology Grade 11 Grade 10 Grade 9 Grades 7-8 Grade 16 pr~wnting Tobacco Use Among Young People Table 24. Percent distribution of high school seniors (N [weighted] = 19,831), by grade in which they first (if ever) used cigarettes and alcohol, Monitoring the Future Project, United States, 1986-1989 Grade when respondent first tried alcohol Grade when respondent first tried Never Row cigarettes S6 7-a 9 10 11 12 used total $6 4.2 7.2 4.9 2.5 1.5 0.6 0.3 21.2 7-8 1.3 8.0 6.4 3.1 1.3 0.5 0.2 20.8 9 0.4 2.0 4.9 2.4 1.0 0.4 0.1 11.1 10 0.3 1.1 1.9 2.8 1.0 0.3 * 7.4 11 0.2 0.5 1.2 1.6 1.6 0.3 0.1 5.5 12 0.1 0.3 0.5 0.6 0.6 0.6 0.1 2.7 Never used 2.0 3.8 5.3 5.3 4.7 2.8 7.5 31.4 Source: Centers for Disease Control and Prevention, Office on Smoking and Health (unpublished data). * < 0.05. Note: Totals may not equal the sum of individual percentages because of rounding. Table 25. Percent distribution of high school seniors (N [weighted] = 20,657), by grade in which they first (if ever) tried cigarettes and marijuana, Monitoring the Future Project, United States, 1986-1989 Grade when respondent first tried cigarettes 5 6 Grade when respondent first tried marijuana 7-8 9 10 11 12 Never Row used total 56 2.0 4.5 3.3 2.2 1.4 0.8 6.2 20.3 7-8 0.3 4.1 4.4 2.9 1.5 0.8 5.8 19.8 9 0.1 0.5 2.5 2.3 1.2 0.6 3.5 10.7 10 0.1 0.2 0.5 1.7 1.4 0.5 2.6 6.9 11 * 0.1 0.3 0.4 1.3 0.6 2.5 5.2 12 * * 0.1 0.2 0.2 0.5 1.5 2.6 Never used 0.2 0.5 0.9 0.9 0.9 0.6 30.5 34.5 Source: Centers for Disease Control and Prevention, Office on Smoking and Health (unpublished data). *< 0.05. Note: Totals may not equal the sum of individual percentages because of rounding. Epidemiology 73 Surgeon General's Report Table 26. Percent distribution of high school seniors (N [weighted] = 21,007), by grade in which they first (if ever) used cigarettes and cocaine, Monitoring the Future Project, United States, 1986-1989 Grade when respondent first tried cocaine Grade when respondent first tried Never Row cigarettes 56 7-8 9 10 11 12 used total 16 0.1 0.4 0.9 1.2 1.4 0.9 15.4 20.3 7-8 * 0.2 0.6 1.1 1.3 0.9 15.6 19.7 9 * * 0.2 0.5 0.6 0.3 9.0 10.7 10 * * * 0.2 0.4 0.2 6.1 7.0 11 o ? o ? 0.2 0.2 4.8 5.2 12 o ? ? ? ? 0.1 .2.5 2.6 Never used * o 0.1 0.2 0.2 0.3 33.8 34.5 Source: Centers for Disease Control and Prevention, Office on Smoking and Health (unpublished data). o 0.05. < Note: Totals may not equal the sum of individual percentages because of rounding. Table 27. Percentage of high school students who used tobacco, by behaviors that contribute Lo unintentional and intentional injuries, Youth Risk Behavior Survey, United States, 1991 Current Current bY Current frequent smokeless Risk behavior Number cigarette use* cigarette use' cigarette use' tobacco us& Seat belt used Always Most the time/sometimes Rarely/ never Physical fighting' 0 times l-5 times 26times Weapon carrying** 0 days 2 1 day Attempted suicide' 0 times 11 time 2,908 60.2 17.8 6.8 13.5 5,651 70.1 26.3 11.4 17.6 3,548 80.6 40.3 21.8 26.5 6,864 63.9 20.3 8.1 13.9 4358 77.8 35.4 17.3 23.2 789 82.6 49.3 30.5 32.1 8,703 65.5 22.6 9.4 13.3 3,171 82.8 41.1 22.2 27.5 10,060 68.2 24.8 10.6 17.8 824 85.0 52.5 33.8 33.6 Sources: Centers for Disease Control and Prevention (CDC), Division of Adolescent and School Health (unpublished data); CDC, Office on Smoking and Health (unpublished data). *During the respondent's lifetime. `Cigarette use on 2 1 day during the 30 days preceding the survey. *Cigarette use on 2 20 days during the 30 days preceding the survey. sDuring the 30 days preceding the survey; includes chewing tobacco or snuff; males only. A When riding in a car driven by someone else. `During the 12 months preceding the survey. **During the 30 days preceding the survey; includes any weapon such as a gun, knife, or club. 74 Epidemiology preuen ting Tobacco Use Among Young People Cigarette use is most likely to precede use of other sub- stances and to be prevalent among users of other drugs. Cigarette Smoking and Other Health-Related Behaviors Available data on the relationships between ciga- rette smoking and other health-related behaviors are derived from cross-sectional studies and thus suggest that other behaviors may covary with adolescent smok- kg. Even if the direction of influence is not established, information on the extent of these relationships is useful for intervention, since such data may suggest a syn- drome of health-compromising behaviors that need to be considered together. Data from the 1991 YRBS indicate that high school students who reported practicing other selected health-. risk behaviors were more likely to be past-month or frequent smokers than were those who reported fewer selected health-risk behaviors. For example, students in the survey were more likely to be past-month or fre quent smokers if they rarely or never wore seat belts, had participated in a physical fight six or more times during the preceding year, had carried weapons one or more days during the preceding month, or had made one or more suicide attempts during the preceding year (Table 27). Students were also more likely to be past-month or frequent smokers if they had ever had sexual intercourse, had had sexual intercourse with four or more partners during their lifetime, or had not used a condom during their most recent sexual intercourse (Table 28). These relationships for sexual risk behaviors held for males and females, regardless of age (CDC, OSH, unpublished data). Lastly, students were more likely to be past- month or frequent smokers if they had not participated on any sponsored sports teams during the preceding year or if they had used steroids without a doctor's prescription (Table 29). Cigarette Smoking and Health Status Pregnancy and Smoking Data on maternal smoking status during pregnancy are recorded on birth certificates in 43 states and the District of Columbia (NCHS 1992b). ln these states, the overall maternal smoking prevalence was 20 percent in 1989. Maternal smoking among adolescent women Table 28. Percentage of high school students who used fobacco, by sexual risk behaviors, Youth Risk Behavior Survey, United States, 1991 , hY Current Current Current cigarette cigarette frequent smokeless Risk behavior Number use* use' cigarette usez tobacco uses Sexual intercoursed No 5,011 55.1 13.8 3.1 12.9 Yes 6,508 82.6 38.8 20.7 23.9 Number of sexual partners* l-3 4,048 81.0 33.8 15.4 23.2 24 2,443 85.4 47.9 30.3 24.9 Condom use1 No 2,494 86.4 46.2 27.5 23.8 Yes 2,091 79.3 36.0 18.5 26.6 Sources: Centers for Disease ControI and Prevention (CDC), Division of Adolescent and School Health (unpublished data); CDC, Office on Smoking and Health (unpublished data). * During the respondent's lifetime. ' Cigarette use on 11 day during the 30 days preceding the survey. $ Cigarette use on 2 20 days during the 30 days preceding the survey. 5 Any smokeless tobacco use, including chewing tobacco or snuff, during the 30 days preceding the survey; males only. * During the respondent's lifetime. 7 During last sexual intercourse, among students who had sexual intercourse during the 3 months preceding the survey. Epidemiology 75