Introduction
Smoking is a risk factor for heart disease, stroke and cancer and contributes to an estimated 21% of deaths.¹ Adolescent smoking is a strong predictor of later smoking.² By age 15-17 one third of Irish young people are smokers.³ The 1999 European School Survey Project on Alcohol and Drug Use (ESPAD) found 31% of Irish 15-16 year old boys and 36% of girls had a lifetime cigarette use of 40 or more.⁴ There is a higher smoking prevalence among young females.^{5, 6}

Smoking behaviour is shaped by social influences. Children whose parents, siblings or peers smoke are more likely to have tried a cigarette and more likely to want to smoke in the future.^{7, 8, 9} Smoking prevalence is influenced by school culture while the media exerts pressure on individuals to use tobacco and can affect the age of smoking onset.^{10, 11} Ease of obtaining cigarettes and economic factors are also strong predictors of adolescent tobacco use.^{11, 12} By age 5-8 children have been exposed to cigarettes and develop attitudes, beliefs and perceptions in relation to smoking behaviour.⁸

School-based interventions have traditionally concerned knowledge transmission of health risks of tobacco. It was assumed that knowing the dangers would discourage students from health compromising activities. This approach is now recognised to be widely ineffective: knowledge alone is insufficient to help people make healthy choices and change behaviours.^{13, 14}

School programmes now adopt a more holistic approach that aims to involve community, increase tobacco knowledge, strengthen refusal skills and improve decision-making ability. This study evaluates the impact of one such school-based intervention.

The Smoke Free Leitrim Project
Smoke Free Leitrim [SFL] was a five-year smoking education/prevention project managed by the NWHB Health Promotion Service. It embraced all County Leitrim students (then aged 9-10 years) in fourth class of primary school in 1996-7. They were tracked through 3 years of primary school and transition to post-primary. In year 3 (1998) the project received funding from the Europe Against Cancer Programme.

Primary school phase
Classroom materials were provided that dealt with smoking, healthy lifestyles and growing up. Teachers determined order and timing of lessons. Classroom work was complemented by additional materials, visits from Health Promotion staff and, at the end of each year, a celebratory events day.

Post-primary school phase
SFL aimed to maintain a supportive environment for students as they moved to bigger schools through a post-primary health education programme that operated in 11 post-primary schools. Classroom materials were not provided; activities included quizzes and competitions; visits by Health Promotion staff; newsletter, website and peer education project; Smoke Free School Buses Campaign and an annual events day.

Method
The project commenced in 1996 with a randomised control design. The SFL cohort was matched with a control group located elsewhere in the NWHB area. Each group provided baseline data on smoking behaviour, knowledge and attitudes; and participation in physical activity. Neither this baseline data, nor the results of surveys conducted in 1998, were used in the study due to non-comparability of survey instruments.

At programme end in 2001 the control (n = 268) and SFL (n = 352) groups completed self-report questionnaires on smoking behaviour, knowledge and attitudes; and participation in physical activity. Students also answered questions about family members smoking; reasons for continuing to smoke; and about giving up. The survey was administered face-to-face to the control group and to SFL students who attended 11 post-primary schools in Leitrim, Sligo and Donegal; it was posted to those at school elsewhere. The full SFL cohort (n = 450) was not surveyed due to: absence; non-return of questionnaire; repeat students; moved from Health Board area.

The intervention did not target specific numeric outcomes. Expectations were that SFL students would, when compared with the control, have smoked less; experimented less; know more about the effects of smoking; have more negative attitudes towards smoking; and exhibit higher levels of physical activity.

Currently smoke was defined as smoking everyday, once a week or less than once a week. Data was analysed using SPSS ¹⁰. Pearsons x2 test was used to determine differences between groups on a number of variables; Spearmans Rank correlation test established significant relationships between variables. Percentages have been rounded upwards for results of .5 and above.

Results
Of the SFL group 185 (52%) were males and 165 (47%) were females - 2 did not declare their sex; 12% were aged 13; 72% were 14; and 16% were 15+. The control group comprised 120 (45%) males and 148 (55%) females: 14% aged 13; 77% aged 14; and 9% 15+.

Smoking experimentation
At intervention end 208 (59%) of the SFL group and 147 (55%) of the control group had tried smoking, with higher levels of experimentation in the SFL group: one hundred and eleven (60%) of SFL males and 68 (57%) of control group males had tried smoking; ninety-four (57%) of SFL females and seventy-eight (53 %) of control group females had tried smoking.

Students in both groups commenced smoking at much the same age. For those who had tried smoking, 19 (9%) of the SFL group (n = 206) and 12 (8%) of the control group (n = 147) did so by age 5-8; 122 (59%) of the SFL group and 82 (56%) of the control group by age 9-12; 60 (29%) of the SFL students by age 13-15 and 46 (31%) of the control group by this age. Six (3%) of the SFL group and 7 (5%) of the control group who had tried smoking did not answer. For smoking experimentation there was no statistically significant difference between the groups.

Current smoking behaviour
Sixty-six (19%) of the SFL group and 63 (24%) of the control group currently smoke. The data was further examined to identify proportions of males and females in both groups smoking everyday, once a week and less than once a week. There are strong similarities between SFL and control group males but a significant difference amongst females (x2 (3, n = 171) = 8.763, p<0.05): those in the control group are twice as likely to smoke every day and 50% more likely to currently smoke (Table 1).

Knowledge about smoking
Students were asked about smoking-related health risks: 326 (93%) SFL students agree that smokers have more risk of heart attack; 246 ( 93%) of the control group responded similarly (x2 (3, n = 614) = 1.609, p = 0.657). Most students agree that smokers have a greater risk of cancer: SFL 318 (92%) and control 240 (89%). The difference between the groups is not significant (x2 (3, n = 607) = 2.109, p = 0.55).

Knowledge and smoking behaviour
Examination of experimentation amongst those in each group with similar levels of knowledge reveals differences: 115 (57%) of those in the SFL group (n = 201) exhibiting high knowledge have tried smoking but in the control group (n = 148) 65 (44%) students with high knowledge have experimented.

Attitudes to smoking
While most students in both groups do not feel that smoking makes them feel grown up, 104 (30%) SFL students and 112 (43%) of the control group do agree with this proposition a significant difference (x2 (3, n = 608) = 14.865, p<0.01). Three hundred and eight (87%) of the SFL group and 230 (86%) of the control group agree that smoking is a waste of money.

To the statement I dont mind if my friends smoke, 249 (72%) of the SFL group and 175 (67%) of the control group agree. The difference between the groups is significant (x2 (3, n = 607) = 10.523 p<0.05). 72% of each group disagrees with the statement smoking helps keep you slim.

Family members and smoking
One hundred and two (29%) SFL students reported that their mother smokes; 109 (31%) have fathers who smoke; 74 (21%) brothers who smoke and 77 (22%) sisters who smoke; 28 (8%) stated that another person who lived with them smoked. The situation is similar in the control group: 88 (33%) have fathers who smoke; 75 (28%) mothers who smoke; 67 (25%) brothers who smoke and 62 (23%) sisters who smoke; 16 (6%) say someone else they live with smokes.

Significant correlations exist between student smoking experimentation and parental or sibling smoking: mothers smoking (rs = 0.165, n = 581, p = 0.01); fathers smoking (rs = 0.172, p = 0.05); mothers and fathers smoking (rs = 0.346, n = 581, p< 0.01); brothers smoking (rs = 0.162, n = 528, p<0.01); and sisters smoking (rs = 0.277, n = 526, p<0.01).

Reasons for continuing to smoke
Twenty-one (31%) of the SFL students who smoked (n = 67) continued to do so because they liked it while 15 (22%) smoked only on social occasions; 15 (22%) believe they do not smoke enough to harm themselves. Other reasons included: friends smoked 4, (6%); stopping would mean weight gain 2 (3%); other reasons 9 (14%); 1 (1%) response missing.
In the control group (n = 63) 19 (30%) continued to smoke because they liked it; 15 (24%) felt they did not smoke enough to harm themselves; 8 (13%) smoked only on social occasions. Four (6%) smoke because their friends smoked; 2 (3%) continue to smoke for fear of weight gain; 13 (21%) gave other reasons; 2 (3%) response missing.

Twenty-three (34%) SFL smokers would like to quit smoking; 27 (40%) say they are not sure if they want to give up; and 9 (14%) smoke but do not want to quit. Eight (12%) responses were missing. In the control group 29 (46%) smokers would like to quit; 22 (35%) are not sure; and 8 (13%) do not want to give up; 4 (6%) responses were missing.

Discussion
SFL aimed to delay or prevent smoking, but data on age of first cigarette and level of smoking experimentation reveals little difference between target and control groups. The key age for starting to smoke is 9-12 and a slightly higher percentage of SFL students (59%) than control group (56%) had smoked their first cigarette by this age. Involvement in the programme had little effect on smoking experimentation; conversely a heightened awareness of smoking may have made participants more curious. Research suggests that smoking education programmes for young people risk having the opposite effect to the one intended.⁹

Nineteen percent of the SFL group and 24% of the control group currently smoke. Though not directly comparable, the 2002 national HBSC survey indicates that 19% of 10-17 year olds are current smokers.¹⁸ Data for daily smoking shows no differences between SFL and control group males. The variation between control group males and females (10% and 16% respectively smoke daily) complements studies that show that girls smoke more than boys from 14 onwards.⁵

A key finding is the significant difference between SFL and control group females for daily smoking: 50% fewer SFL females smoke every day compared to the control group.
Young people are knowledgeable about the effects of smoking, notwithstanding involvement in an education programme.^{6, 15} Interestingly, more of those in the SFL group than the control group with high levels of knowledge have tried smoking. The evaluation indicates, supported by research elsewhere, that interventions that aim to inform young people about the damage of smoking are ineffective in changing smoking rates. ^{6, 9}

Both groups know about passive smoking but express no concern about friends smoking. Students generally regard smoking as a matter of individual choice. This is supported by evidence that links young peoples smoking to peer sociability.⁶ It may also reflect cultural norms within Irish society. The project did not stimulate advocacy about non-smoking. Programmes for young people should involve the empowerment of non-smokers to speak out.¹⁶

Data for both groups on family members smoking was similar and, as with evidence elsewhere,¹⁷ confirms that the likelihood of trying smoking is much greater if either parent, brother or sister smokes.

Smokers in both groups continued to smoke because they like it. Large proportions in both groups are unsure about quitting though many do want to. An understanding of the social context for those ambivalent about giving up and a redirection of interventions to those who want to quit may be more effective.⁹

The evaluation suggests that long-term school-based anti-smoking interventions have little impact on smoking behaviour, especially amongst boys. To increase the number of young people who remain smoke-free an approach that goes beyond the school is needed. Interventions need to tackle young peoples smoking in the broader context of smoking within the family and with other adults in general. For those who want to quit, targeted support programmes may have more potential than general smoking education programmes. Anti-tobacco health promotion initiatives should also relate to the cultural context of adolescent smoking and risk-taking behaviour; give meaningful involvement to young people in design and delivery of programmes; and relate to what is going on in the wider field of tobacco control.

Correspondence: Michelle Share, Public Health /Health Promotion Department,
North Western Health Board
Tel: +353 (0) 71 9852000,
Fax: +353 (0) 71 9851287
Email:[email protected]

1. Peto R et al. Mortality from smoking in developed countries 1950 to 2000, Oxford University Press, New York, 1994.
2. Paavola M, Vartianen E, P Puska. Smoking cessation between teenage years and adulthood. Health Education Research 2001; 16, 1:49-57.
3. Department of Health and Children. The National Health and Lifestyle Surveys: surveys of lifestyle, attitudes and nutrition (SLAN) & The Irish Health Behaviour in school-aged children survey. Dublin: Department of Health, 1999.
4. Hibell B, Andersson B, Ahlstrm S et al. The 1999 ESPAD report: alcohol and other drug use among students in 30 European countries. Stockholm: CAN, 2000.
5. Becher H. Smoking prevalence and cigarette consumption in R Boreham and A Shaw (eds) Smoking and Drinking among young people in England. London: The Stationery Office, 2001.
6. Oakley A, Brannen J, Dodd K. Young people, gender and smoking in the United Kingdom. Health Promotion International 1992; 7, 2:75-88.
7. Patton GC, Carlin JB, Coffey C, Wolfe R, Hibbert M, Bowes G. The course of early life smoking: a population-based cohort study over three years. Addiction, 1998; 93, 8: 1251-1261.
8. Porcellato L, Dugdill L, Springett J, Sanderson FH. Primary schoolchildrens perceptions of smoking: implications for health education. Health Education Research, 1999; 14, 1 71-83.
9. Charlton A. Why are school-based youth centred smoking interventions not as effective as we hoped? Some ideas for research. Health Promotion and Education 2000; 38: 124-128.
10. Lloyd B, Lucas K. Smoking in Adolescence: Images and Identities. London: Routledge, 1998.
11. Bobo JK, Husten C. Sociocultural influences on smoking and drinking. Alcohol Research and Health 2000; 24, 4: 225-232.
12. Layte R, Russell H, Mc Coy, S. The Economics and Marketing of Tobacco: An Overview of Existing Published Evidence. Dublin: ESRI, 2002.
13. Morgan M. The Substance Misuse Prevention Programme: a formative evaluation. Dublin: The Department of Education and Science, 1998.
14. Moon AM, Mullee MA, Rogers L et al. Helping schools become health promoting environments an evaluation of the Wessex Healthy Schools Award. Health Promotion International 1999; 14, 2: 111-112.
15. Department of Health. Smoking and Drinking among young people in Ireland. Dublin: Department of Health, 1996.
16. Glantz S, Jamieson, P. Attitudes towards second-hand smoke, smoking and quitting among young people. Pediatrics 2000: 106 1474-1477.
17. Maguire N, Howell F, Moran A. Smoking and quitting among Irish teenage males. Irish Med J 2000; 93, 9: 272-273.
18. Department of Health and Children, The National Health and Lifestyle Surveys: surveys of lifestyle, attitudes and nutrition (SLAN) & The Irish Health Behaviour in school-aged children survey. Dublin: Department of Health and Children, 2003.