Helping youth to quit smoking what does research suggest
Sponsored Links
This presentation is the property of its rightful owner.
1 / 72

Helping youth to quit smoking: What does research suggest? PowerPoint PPT Presentation

  • Uploaded on
  • Presentation posted in: General

Helping youth to quit smoking: What does research suggest?. Paul McDonald, PhD University of Waterloo Ontario Tobacco Research Unit. Why focus on youth?. 15 to 19 year olds make up 8% of the NL population (age 15+) and nearly 6% of tobacco users

Download Presentation

Helping youth to quit smoking: What does research suggest?

An Image/Link below is provided (as is) to download presentation

Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author.While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server.

- - - - - - - - - - - - - - - - - - - - - - - - - - E N D - - - - - - - - - - - - - - - - - - - - - - - - - -

Presentation Transcript

Helping youth to quit smoking: What does research suggest?

Paul McDonald, PhD

University of Waterloo

Ontario Tobacco Research Unit

Why focus on youth?

  • 15 to 19 year olds make up 8% of the NL population (age 15+) and nearly 6% of tobacco users

  • 20 to 24 year olds make up 8% of the NL population (age 15+) but 12.2% of all tobacco users.

  • Persons 45+ make up 51% of the NL population but less than 40% of tobacco users

Source: CTUMS, 2006

Reasons for optimism

  • Unassisted quit rates for youth are higher than with adults

    • 12 month abstinence rates of up to 40% for adolescents compared to adults rate of 3% to 8% (Driezen et al., submitted)

    • May be because more youth are occasional smokers; daily smokers consume fewer cigarettes; youth frequent more supportive environments; youth smoking patterns are less established.

    • Different puff topography than adults (lower puff volume) (Franken et al., 2006)

Reasons for optimism

Youth have tried to quit in past year

  • 63% of 15-17 yr olds

  • 69% of 18-19 yr olds

  • 62% of 20-24 yrs olds

  • 45 % of 25+

    Trying to quit at least once in the previous year doubles the likelihood of interest in school-based programs (Leatherdale and McDoinald, 2007)

CTUMS, 2006

How should we assess interventions?

  • Should we only be concerned with effectiveness and “reach”?

A revised approach to tobacco control: from prevalence to risk reduction

Never smokers

exposed to ETS

Current smokers and

recent former smokers

% of


at risk

Never smokers

Not exposed to ETS

Low High

Risk of dying or becoming ill

“Rosian” population approach to prevention:Shift the risk of the majority.Impact =“Effect x exposure”

Mean 1 (ū1)

Mean 2 (ū2)




of risk

Probability density

Low High

Level of risk

Not all sub-populations have the same risk distribution; Not all people will yield the same benefit from intervention






Probability density

Low High

Level of risk

A new population paradigm:Effect of sustained program of intervention

3. Reduce variance





1. Shift risk continuum,

2. especially for high risk

(ie high burden)

Probability density

Low High

Level of risk

Implications of the new paradigm

Select intervention mix based on stakeholder

acceptability, feasibility, stage of problem, and potential population impact where:

Effectx Exposure x Benefit


Population impact =



Equity variance

Exposure = % of population who receive or are exposed to intervention program;

Effect = % of people exposed who benefit from an intervention or odds of success;

Benefit = Average expected future benefit from “success” (ie reversible risk)

Cost = Investment (dollars, opportunity etc.) required to achieve given impact

Equity = Variance in impact across designated sub-populations of interest

Not all people benefit equally from quitting: the case for youth cessation

Number of additional years

of life by age of quitter

Life years gained

(relative to continuing smokers)

Average age of quitters

Doll et al., 2004


  • Evaluate programs in terms of:

    • Effectiveness (odds of quitting)

    • Reach or exposure (how many people)

    • Average expected benefit (would addressing the needs of another group yield greater impact?)

    • Impact on sub populations of smokers (not just the majority)

    • Cost to benefit ratio (is it the best way to spend your money relative to alternatives)

How effective are individual treatments?

  • Recent reviews:

    • Mermelstein, 2003

      • a narrative review

    • Garrison et al., 2003

      • limited to 6 published, controlled trials

    • Sussman, 2002; Sussman, Sun and Dent, 2006

      • Systematic review of 66 published and unpublished trials of various designs and outcomes

    • Grimshaw and Stanton, 2006 – Cochrane Review

      • Cochrane protocol applied to 15 eligible studies of published RCTs

    • McDonald et al., 2003

      • Refinement of Sussman, 2002

      • Determined to be the most comprehensive and rigorous

      • Used a panel of experts instead of single reviewer

      • Results based on 20 studies with high or moderate validity (emphasis put on high validity studies)

Expert panel findings(McDonald et al., 2003)

  • Treatments for youth are promising

    • 9 of 20 were effective (incl. 2 of 5 high validity)

  • Treatments based on social cognitive theory/cognitive-behavioural approaches are promising

    • 8 of 14 were effective (incl. 2 of 5 high validity)

    • SLT includes: goal setting, self monitoring, development of coping skills and self efficacy, cognitive reframing, problem solving, positive reinforcement, counter-conditioning, stimulus control, contingency management, anger management, assertiveness training, motivational enhancement

Expert panel findings, cont…

  • Insufficient evidence to support other theoretical approaches

    • NO evidence to support stages of change

      • Lawrance, McDonald et al. (unpublished) – 200+ high schools students

      • Aveyard et al (1999) – 8,000 UK 13/14 yr olds

      • Quinlan & McCall – daily smokers in college

      • Efficacy with adults has also been questioned

        • Riemssma et al, 2003 – BMJ review of matched treatments

        • McDonald, in progress

        • Whitelaw et al., 2000 - review

Expert Panel Findings, cont…

  • Insufficient evidence to draw conclusions about:

    • Pharmacotherapy

      • May not be effective for “light” adult smokers either (Pierce & Gilpin, 2002; Niaura et al., 1994)

    • Best delivery setting

      • Most treatments delivered in school settings (class or outside) and health clinics

      • Some school based and clinic programs were effective (just not enough to draw conclusions)

      • Note: youth smokers less likely to be in school or use clinics

Expert panel findings, cont…

  • Insufficient evidence to draw conclusions about:

  • Best type of provider

    • 4 of 6 teacher/school staff delivered programs were effective (all mod. Validity)

    • 2 of 4 using psychologists, health educators or counselors were effective

    • 1 of 3 using trained peers was effective

  • Voluntary vs. mandatory treatment

    • None of the 3 mandatory treatments were effective


  • Recent completed studies not available to McDonald et al.

    • Zhu et al. – randomized trial of 1200 adolescents re proactive telephone counseling based on SCT. Modest short term effect for 17 to 19 year olds; no effect for 14 to 16 year olds

    • Lipkus et al. Randomized trial of adolescents re proactive telephone counseling based on SCT. Effective at 6 month follow-up.

    • Adelman et al, 2001 – randomized trial of school-based group program. Effective at 4 week follow-up.

    • Yiming et al. - randomized control study of laser acupuncture with 200+ 12 – 18 yr olds in clinic. No effect.

    • Roddy et al, 2006 – randomized trial of 98 teen smokers from low SES backgrounds. Six weeks of NRT was not effective. Adherence was very low (median duration of one week of use).


  • Chen and Yeh, 2006 – 6 wk cessation program + internet assisted instruction with 77 senior high school students. Relative to no treatment control, the treatment group had more positive attitudes toward quitting. Effect on quitting is unknown.

  • Myers and Brown, 2005 – effect of intense counseling with 54 adolescents being treated for substance abuse. Counseling increased # of quit attempts but not likelihood of quitting at 3 mo.

  • Moolchan et al., 2005 – RCT with 120 moderate to heavy teen smokers. Short term, but no long term benefit for NRT.

  • Pbert et al (2006). Randomized trial with 577 smokers receiving brief intervention from school nurses (5A model using social cognitive theory). Relative to usual care control, intervention increased odds of quitting at 3 month follow-up.

A few more words about pharmacotherapy and physician based intervention

  • Moolchan et al (2006) found that NRT was safe for adolescents

  • Recommended by most clinical guidelines for both youth and adults, despite lack of efficacy

  • Youth have access to NRT

    • Johnson et al (2004) found that 81% of those <18 who tried to purchase NRT from retailers were successful

  • Wyman et (2006) found only 8% of parents supported schools providing NRT

(Non-)impact of health professionals* on adolescent smokers in Canada(Stevenson et al, 2007 – from CTUMS)

Per cent





advice to

reduce or quit

Visited a health

care provider

in previous

12 months

Provided with

info on cessation


*includes physician,

dentist, or pharmacist

  • US study done in 2000 with students in grades 6 to 12 found that only 16% received assistance to quit from a physician and 12% received assistance from a dentist

Source: Shelly et al., 2005

Additional observations regarding behavioural treatments

  • Virtually all effective treatments involved 8 to 24 hours of contact with facilitator

  • Few studies examine non-face to face interactions (web-based, telephone)

    • Emerging evidence suggests these may be promising

  • Treatments more likely to be effective with older youth (16+)

  • Few studies with special populations (despite high prevalence and over-representation).

    • One study with psychiatric co-morbidity was not successful

Additional observations

  • No studies looked at interaction between intrapersonal and environmental factors

  • Most studies had high loss to follow-up which suggests:

    • Its difficult, even under ideal conditions, to keep youth engaged

    • It significantly undermines statistical power (increases likelihood of missing a true effect)

Why have individual programs so ineffective?

  • Nearly 30% of adolescent and young adult smokers do NOT smoke on a daily basis

  • Some young occasional smokers do not think of themselves as “smokers”

    • Up to half of Ontario high school smokers do not consider themselves to be “smokers” (Leatherdale and McDonald, 2006)

Proportion of current Canadian smokers who smoke less than daily, 1990 - 2005

Per cent

Source: Various surveys

Why have individual programs so ineffective?

  • Most treatments don’t consider adolescent’s lack of experience

    • Little experience in changing any aspect of behaviour. Not familiar with basic behavioural principals such as re-enforcement, stimulus control, etc.

    • Little experience coping with stress

Why have individual programs so ineffective?

  • Many treatments don’t consider adolescent’s special circumstances – they simply offer adult oriented programs (aimed at moderate to heavy smokers)

    • Adolescents are more likely to smoke with friends, than alone or with family members (Oksuz et al., 2007)

    • More likely have to cope with friends who put them down for not smoking (Falkin et al., 2007)

    • More likely to feel they have to give up things important to them in order to quit (e.g., friends) (Falkin et al., 2007)

Tailored interventions may improve outcomes

  • Does the adolescent think of themselves as a smoker?

  • Does an adolescent live with or social with other smokers (who are influential in their lives)?

  • What is the adolescent’s natural history of smoking?

Smoking predictors among early vs. late initiators(Robinson, et al, 2004)

  • 432 adolescents in a smoking cessation trial

  • Based on self report of age of first whole cigarette

  • Compared to late initiators, early initiators (<13) were more nicotine dependence, took longer to move from initiation to daily smoking (1.3 vs. .45 yrs.) and took longer to seek cessation treatment after becoming a daily smoker (2.9 vs. 1.6 yrs)

The Montreal Natural History of Nicotine Dependence study (NDIT)

  • A six year longitudinal study of 1,300 students recruited in grade 7 from 10 schools in and around Montreal

  • Self report survey and physiological measures collected every 3 months

  • Trajectory study included 369 students who initiated smoking during the study

  • Average age of smoking initiation was 13 years

Four trajectories of smoking among

adolescents who become regular smokers

Karp, O’Loughlin, Paradis et al., 2006

Class IV (6%)

Class III (11%)

Class II (11%)

Class I (72%)

  • 6 9 12 15 18 21 24 27 30 33 36 39 42

  • Months after smoking initiation

Predictors of trajectory

  • Boys more likely to be rapid escalators (class IV)

  • Students with poor academic performance more likely to become regular smokers (ie be in class II, III and IV)

  • Students with non-smoking parents and more than half of friends as non-smokers were less likely to become regular smokers (ie become a class II, III and IV smoker)

  • Attending a school with clear non smoking rules reduces likelihood of being in class II, III and IV


  • Early escalators may require more intensive, longer treatment

  • Early interventions may be more effective with late escalators (before regular patte5rns are established)

How can we improve utilization of individual treatments?

  • Review of literature (mostly based on surveys or interviews with youth in school settings)

  • Factors that increase utilization

    • Program is free or incentive provided

    • Friends are supportive

    • Friends are using it

    • Easy to use/low burden

    • Program is effective

    • Program is lead by adults from outside of school

  • Factors that inhibit utilization

    • Potential breach of confidentiality/privacy (esp. to parents)

    • Program or material cost money

    • Program offered after school

Promoting cessation aids

  • McDonald et al completed meta-analysis of 48 promotion campaigns to youth smokers (12 to 24)

  • Median recruitment rate was 7.8% from a median audience size of 310

Maximizing participation in cessation aids(McDonald et al meta-analysis)

  • Message characteristics

    • Campaign should last one to three months

    • Use credible adult spokesperson (not youth or a combo)

  • Channel characteristics

    • Use media in community rather than rely on school or clinic-based promotion

Maximizing participation in cessation aids(McDonald et al meta-analysis)

  • Source characteristics

    • Health department or research organizations are better than provincial or federal government sponsors

  • Destination characteristics

    • Programs offered thru youth centres and workplaces had higher recruitment than schools or clinics

    • Programs offered during winter are best; spring/summer the worst

    • Programs offered before school, during lunch or during work are better than after school

    • Programs that use a variety of cessation tools and formats are more attractive than programs that use one or two strategies

If individual oriented quitting treatments are only minimally effective, what else should we do?

Quitting: Its not just an individual enterprise

  • Quit rates and quit ratios differ substantially (up to 60%) across provinces, regions, cities, schools, neighborhoods (McDonald, et al., unpublished)

  • Youth prevalence has gone down faster than older adults, despite the lack of effective or available individual cessation treatments

  • This suggests the social and economic environment, and policy measure profoundly influence youth quitting

  • Interventions must not be limited to individuals

Proportion of current adults smokers in

Canada who are 15 to 24 yrs old,


Acase for environmental interventions: Gene-environment interactions

  • Novak, Breslau & Kessler

    • 891 pairs of MZ and DZ twins aged 25-74

    • Heritability index (for daily smoking) = .50, but…

      • H (low SES childhood) = .25

      • H (for high SES childhood) =.79

The effect of population interventions on adult smoking prevalence

  • Levy et al (2007) used data from California to simulate the impact of various population interventions on adult smoking rates

  • Overall, tobacco control policies introduced between 1988 and 2004 reduced smoking rates by 25% more than if policies had been kept at 1988 levels. 59% of the excess reduction was due to price increases

    • 28% of the excess reduction was due to media campaigns

    • 11% of the reduction was due to smoke free places

    • 2% of the excess reduction was due to restrictions in youth access

  • Studies from the UK suggest comprehensive individual cessation treatments only reduced prevalence by 1%

  • Results suggest population interventions are profoundly important

The effect of price on cessation among youth

  • Tauras (2004) used data from longitudinal surveys with young adults.

  • Found that restrictions on smoking in worksites and public places marginally increases the odds of quitting.

  • Largest impact was taxation. Price has an elasticity of 0.35 on smoking cessation – each 10% increase in the price will increase the likelihood young adults will quit smoking by 3.5%

The tobacco industry and consumers are adapting: Discount cigarette use

Speaking of taxes…

  • Find out how much your province collects in taxes from the sale or consumption of tobacco by your more than 5,000 underage tobacco users. How does this compare to expenditures on tobacco control?

What are the effects of media on youth cessation?

  • Beiner et al (2006) survey 787 people who had quit smoking in the previous two years

  • Media advertisements were the most frequently mentioned source of help

  • Older smokers were more likely to find traditional quit aids to be helpful

  • Younger smokers were more likely to cite TV ads as being helpful including those which depict smoking related illness and inspirational quit tips.

Effect of no-smoking policies on youth smoking

  • Wakefield et al (2000) surveyed 17,287 students in 202 US high schools

  • Odds of 30 day smoking based on locations with restrictions compared to no restrictions

    Total home ban.79 (.67 - .91)

    Partial home restrictions.85 (.74 - .95)

    Public places.91 (.83 - .99)

    Enforced school ban.86 (.77 - .94)

    School ban (no enforce).99 (.85, 1.13)

Interaction of pharmacotherapy and smoke free homes

Effect of pharmacotherapy when

the user’s home is not smoke

free and there is another smoker

in the house

Effect of pharmacotherapy when

the user’s home is smoke free

and there is another smoker

in the house

Gilpin, Messer & Pierce, 2006

Effect of number of close friends who smoke on the odds of moving from regular smoking to quitting



Number of close friends who smoke

Source: Leatherdale, 2004

Modeling matters

  • If having friends who smoke can inhibit quitting, can having friends who quit with you enhance success?

Influence of Canadian smoking bans on cessation

  • Shields (2007) used longitudinal data from the National Population Health Survey

  • Smoke free homes significantly increases the odds a person will quit smoking in the next two years

  • Young smokers were more likely to quit due to smoke free homes than any other age group.

  • The effect of smoke free homes is particularly powerful for men (OR = 1.3), those with high school education or higher (OR>1.9), and those with children under age 15 in the home (OR=2.7) .

  • Smoke free homes also significantly reduces the average number of cigarettes people smoke on a daily basis.

Influence of Canadian smoking bans on cessation

  • Smoke free workplaces had the largest effect on 35 to 44 year olds and the least impact on 15 to 24 year olds (likely because they spend less time at work).

Do smoking bans have adverse effects?

  • No known reports of accidents caused by school or institutional smoking bans

  • Callahan et al (2007) studied effect of institutional bans on the likelihood BC youth would attend and stay for inpatient substance abuse treatment

    • Concluded that total smoking bans had no effect on recruitment or retention of youth

A few odds and ends…

Checking our assumptions against the evidence

  • Young women do not continue to smoke to avoid weight gain

    • Weight is seldom mentioned in qualitative interviews

    • Weight was listed as primary barrier to quitting by only .2% of current adolescent smokers (CTUMS, 2001)

    • Coping with negative affect (depression, anger, stress, boredom) is a significant challenge

Checking our assumptions

  • Youth do not dismiss health warnings because of feelings of immortality

    • Henly and Donovan (2003) Australian youth responded as well or better to threat appeals

    • Threat campaigns should include efficacy information (re self and response)

All the good ideas haven’t been tested

  • Putting warning on cigarettes increases salience. Messages could be used to enhance the immediacy of health effects

Checking our assumptions

  • Any quit attempt is a good attempt

    • Zhu et al. (1999), Lawrance and McDonald (1996) and others have found that to be predictive of future abstinence a quit attempt needs to last at least 14 days; short term quit attempts may actually reduce future success (although studies need to adjust for self efficacy and other factors)

    • This suggests we need to do more to support youth during at least the first two weeks of abstinence.

Checking our assumptions

  • Planning for a quit attempt may not be as important as many providers believe.

Checking assumptions: Is planning for a quit attempt necessary?

  • Current counseling models emphasize the importance of planning and preparing for a quit attempt

  • Nearly half of successful quitters did not plan their attempt (including 48% of 16 to 34 year olds))

  • The odds of quitting for 6+ months were higher for unplanned quit attempts relative to those who planned (OR=2.5)


Planned a few

days ahead

Planned for same

day or next day

Source: West & Sohal, 2006

Thinking outside the box

  • Helping parents helps their children.

    • Children with no smoking or former smoking parents are less likely to smoke or more likely to quit (Bricker et al., 2003)

    • Children gain a stronger appreciation of the challenges of quitting when they were involved in their parent’s quit attempts (Tilson et al, 2005)

    • 88% of young smokers reported their parents were not in very good health – compared to 54% of smokers whose parents did not smoke. Youth cessation programs could benefit from targeting smokers who parents smoke and are likely showing the long term effects of smoking related illness (Jamieson and Romer, 2006)

Ontario smoking rates by level of belonging to the community, 2004

*p< .0001 after adjusting for sex, age, and income
















Source: CCHS, 2004

Smoking rates by sex and level of belonging to the community, 2004

Per cent

P =<.001

Source: CCHS, 2004

The effect is strongest for those born most recently

Interaction is significant (p< .0001)












Source: CCHS, 2004

Intentions to quit in next 6 months among Canadian smokers, by level of belonging to the community, 2004




to quit

P < .0001 after controlling for sex, age, education

Source: CCHS, 2004


  • Join with other groups and agencies concerned with broader determinants of health

    • Reduce marginalization and invest in social capital

    • Reduce poverty

    • Invest in education

Consider the wisdom

of dolphin feeding


Future Considerations for individual treatment

  • Study the interaction between individuals and situations. Determine if individual treatments require certain predisposing situational conditions

  • Identify and develop interventions for important sub-populations

    • Separate 12 – 15, 16 to 18 and 19 – 24 year olds

    • Consider aboriginal status, psychiatric co-morbidity and low SES, those with and without social support, those with and without parents who smoke

Future considerations for individual treatment

  • Study new methods of treatment delivery

    • Ubiquitous web-based

    • Telephone-based

    • NRT (for “heavy” youth smokers only; in combo with other treatments)

  • Enhance the rigor of evaluation and research

    • More bad or quick and dirty studies won’t inform us.

    • Major investments are required (>$.5 million per study)

  • Link research, practice and policy

Future considerations for population interventions

  • Test and potential advocate for changes in cigarette design and testing

    • Change the taste (quinine?); the appearance (fecal brown?); warnings on the cigarette?

  • Reduce discounting by setting a minimum price; outlaw the sale of used tobacco manufacturing devices

  • Provide greater enforcement to reduce counterfeiting and smuggling

Future considerations for population interventions

  • Introduce smoking restrictions in youth workplaces, multi-unit dwellings, cars with children, etc.

  • Encourage community action on smoke free homes

  • Restrict youth access to cigarettes (special tobacco control outlets?)

  • Eliminate marketing and sales through the internet

  • Develop media campaigns to encourage youth to quit


  • Most individual approaches to cessation, including pharmatherapy, have not been successful with general populations of youth tobacco users. Modest allocation of resources for behavioural counseling through websites, telephone quitlines, schools, health professionals, youth workplaces, community settings is warranted. However, much more effort should be put into more population oriented strategies.

The end(and not a moment too soon)

  • Login