Table 7: Effects of antitobacco mass media campaigns on smoking initiation, cessation, and prevalence.

Author, yearCountry
(Data source)
Study designDates of data collectionPopulationIntervention, n Smoking measureEffect on smoking initiation, cessation or prevalence

                                                        Smoking initiation

Bauman et al., 1991 [145]US (original data)Cluster RCT1985–1987Youths, ages 12–14(I) 8 30-second radio messages focused on 7 expected consequences of smoking broadcasted over 3 1-month periods;
(C) no mass media campaign*, 951 total nonsmokers at baseline
Ever puffed a cigaretteAmong nonsmokers at baseline, differences relative to comparison group at 11–17 months after broadcasts ended
(i) Smoking experimentation: 1% (P = NS)
(ii) Regular smoking: 2% (P = NS)
(iii) Recent smoking: 1% (P = NS)

Farrelly et al., 2009 [75]US (NLSY97)Longitudinal study1997–2004Youths, ages 12–17(I) TV campaign with cumulative exposure between 2000–2004 of 3096–32137 GRPs across 210 media markets, 8904Ever smoked a cigaretteHR = 0.8 (95% CI: 0.71–0.91; 𝑃 = 0 . 0 0 1 ) (per 10,000 GRP cumulative exposure)

Linkenbach and Perkins, 2003 [72]US (original data)Longitudinal study2000–2001Youths, junior and senior high school students;
mean age = 14.6;
50% male
(I) 1500 GRPs (broadcast TV); 78,000 print and promotional items distributed in schools; 4 theater slides were run over 1 month at 2 movie theaters; 1 billboard design appeared in 4 locations for 1 month, 299;
(C) control, 314
Having tried cigarette smoking12-month follow-up smoking prevalence:
(I) 10%
(C) 17%
Relative measure: 41% lower rate of initiation in intervention group ( 𝑃 < 0 . 0 5 )

Flynn et al., 1997 [73]US (original data)Longitudinal study1985–1991Youths, grades 4–6(I) 540 TV and 350 radio broadcasts per year for 4 years plus school intervention;
(C) school intervention
Smoked >0 cigarettes in past week4-year follow-up smoking prevalence:
(I) 7.5%
(C) 13.0%
6-year followup smoking prevalence:
(I) 15.9%
(C) 20.2%
OR = 0.73

Hafstad et al., 1997 [74]Norway (original data)Longitudinal study1992–1995Youths, ages 14-15(I) 3 annual campaigns of 1 TV and cinema ad 167 times, 3 full-page ads in 5 newspapers, 1 poster in each location run for 3 weeks;
(C) control county
Weekly smokingMales
1-year initiation rate
(I) 10.2%
(C) 14.5%
O R = 0 . 6 7 (95% CI: 0.53; 0.85)
Females
1-year initiation rate
(I) 14.6%
(C) 25.6%
O R = 0 . 7 7 (95% CI: 0.63; 0.95)

                                                        Smoking cessation

Solomon et al., 2009 [80]US (original data)Cluster RCT2001–2004Youths, grades 7–10;
45% male
(I) radio and TV campaign with 380 GRPs/week over 9 months each year for 3 years, 531;
(C) no intervention, 601
Not smoking one cigarette in past 30 days12-month quit rate
(I) 18.1%
(C) 14.8%
24-month quit rate
(I) 14.5%
(C) 12.6%
36-month quit rate
(I) 16.0%
(C) 12.8%
Relative measure: no significant time trend or interaction between condition and time

Terry-McElrath et al., 2011 [81]US (MTF)Longitudinal2001–2008Adults, age 20–3024-month sum of antitobacco TV advertising measured in GRPs, 7997Smoked 0 cigarettes/day in past 30 days<52 GRPS (ref)
52103 GRPs
aOR: 1.15 (95% CI: 0.91; 1.45)
104155 GRPs
aOR: 1.40 (95% CI: 1.07; 1.83)
156207 GRPs
aOR: 1.21 (95% CI: 0.90; 1.63)
208+ GRPs
aOR: 1.22 (95% CI: 0.90; 1.66)

Burns and Levinson, 2010 [82]US (original data collection)Longitudinal2007Adults, age 18+
41–50% male
(I) Spanish-language TV campaign with 1387.4 GRPs for 1 month, radio ads, and 1900 30-second spots on movie screens, 117
(C) non-Spanish speaking population, 193
6-month abstinenceQuit rate prior to campaign
(I) 9.6%
(C) 16.5%
Quit rate post campaign
(I) 18.8%; 𝑃 < 0 . 0 5
(C) 8.8%; 𝑃 = 0 . 0 1

Durkin et al., 2009 [146]US (UMass Tobacco Study)Longitudinal2001–2004Adults
mean age = 41
45% male
24-month GRPs1-month abstinenceQuit rate, 16%

Hyland et al., 2006 [147]US (COMMIT)Longitudinal study1988–2001Adults, ages 24–64(I) TV campaign above 1218 GRPs in 1999-2000
(C) TV campaign below 1218 GRPs in 1999-2000
NR24-month quit rate
(I) 12.9%
(C) 11.0%
RR = 1.1 (95% CI: 0.98–1.24) (per increase in 5000 GRPs of exposure)

Ronda et al., 2004 [76]Netherlands (original data)Longitudinal study1998–2001Adults, ages 18+
39–47% male;
Mean age: 46–50 years
(I) Billboard, print, radio, TV, posters and postcards in waiting rooms and public buildings; 4 months spread over 2 yearsNot having smoking any tobacco in last 7 days24-month quit rate
(I) 12.3%
(C) 14.3%
36-month quit rate
(I) 18.7%
(C) 18.6%
relative measure: no association between intervention and smoking outcome in regression models (not reported)

McVey and Stapleton, 2000 [148]England (original data)Longitudinal study1992–1994Adults, ages 16+
41-42% male;
Mean age: 46 years
(I) 18-month TV campaign, 1744;
(C) no intervention, 719
No smoking at all nowadays18-month quit rate
(I) 9.7%
(C) 8.7%
OR = 1.27 (95% CI: 0.77–2.08; 𝑃 = 0 . 3 5 )

Hafstad et al., 1997 [74]Norway (original data)Longitudinal study1992–1995Youths, ages 14-15(I) 3 annual campaigns of 1 TV and cinema ad 167 times, 3 full-page ads in 5 newspapers, 1 poster in each location run for 3 weeks, 1061;
(C) control county, 1288
Weekly smokingMales
1-year quit rate
(I) 12.7%
(C) 19.1%
O R = 0 . 6 3
Females
1-year quit rate
(I) 25.6%
(C) 17.6%
O R = 1 . 9

                                                        Smoking prevalence

Flynn et al., 2010 [83]US (original data collection)Cluster RCT2001–2005Youths, grades 7–12,
46% male
(I) 380 GRPs from TV ads per week, 215 GRPs from radio ads, 10,412;
(C) no intervention, 9544
Smoking in past 30 daysBaseline smoking prevalence
(I) 18.9%
(C) 17.8%
Smoking intervention at 4 year followup
(I) 16.9%
(C) 15.5%; 𝑃 = 0 . 9 5

Wakefield et al., 2008 [27]Australia (Roy Morgan Single Source)Time series1995–2006Adults, age 18+138-month TV campaign, 288.5 mean monthly GRPs, 343,835Smoke factory-made cigarettesPrevalence percentage point change two months later (i.e., 2 month lag) per 1 GRP per month increase:−0.00077 (95% CI: −0.00144, −0.0001; 𝑃 = 0 . 0 2 5 )

Hafstad et al., 1997 [74]Norway (original data)Longitudinal study1992–1995Youths, ages 14-15(I) 3 annual campaigns of 1 TV and cinema ad 167 times, 3 full-page ads in 5 newspapers, 1 poster in each location run for 3 weeks, 2742;
(C) control county, 3438
Weekly smokingOR = 0.74 (95% CI: 0.64; 0.86)
Males
Baseline prevalence
(I) 6.9%
(C) 9.9%
1-year prevalence
(I) 13.7%
(C) 20.4%
Females
Baseline prevalence
(I) 10.1%
(C) 11.4%
1-year prevalence
(I) 18.7%
(C) 23.8%

Flynn et al., 1995 [149]
Worden et al., 1996 [150]
Flynn et al., 1992 [151]
Flynn et al., 1994 [152]
Worden and Flynn, 2002 [153]
Flynn et al., 1997 [73]
US (original data)Longitudinal study1985–1991Youths, grades 4–6;
mean age: 10.6 years,
48–54% male
(I) 540 TV and 350 radio broadcasts per year for 4 years plus school intervention;
(C) school intervention
Smoked >0 cigarettes in past weekBaseline prevalence
(I) 1%
(C) 1.6%
6-year prevalence
(I) 16.5%
(C) 24%
OR = 0.62 (95% CI: 0.49; 0.78)
Females
4-year prevalence
(I) 12.7%; 𝑃 < 0 . 0 1
(C) 21.1%
6-year prevalence
(I) 16.5%
(C) 29.4%; 𝑃 < 0 . 0 1
Males
4-year prevalence
(I) 9.8%;, 𝑃 = 0 . 1 6
(C) 14.4%
6-year prevalence
(I) 13.0%
(C) 17.1%; 𝑃 = 0 . 2 3

Steenkamp et al., 1991 [77]South Africa (original data)Longitudinal study1979–1983Adults, ages 15–64
46% male
(I) 48-month billboard, print, poster, and mailing campaign, 1531;
(C) control, 1308
Smoking an average of at least 1 cigarette or 1 gram of tobacco per dayBaseline prevalence
(I) 28.1%
(C) 29.5%
4-year prevalence
(I) 18.8%
(C) 19.9%
percentage Reduction
(I) −32.6%
(C) −33.3%
Net percentage change in smoking prevalence relative to control
Males: 2.0%
Females: −19.2%

Meshack et al., 2004 [154]US (original data)Before/after with comparisonSpring 2000–December 2000Youths, grade 6
52% male
(I) 3 × 3 media and community program; media programs involved TV, radio, billboard, and print; $0.50 per capita in low-intensity group; $1.00 per capita in high-intensity group, 3618Tobacco use in past 30 daysPercent change in prevalence at 8.5 months (among groups with no community program):
High intensity: −20.8%
Low intensity: −45.3%
Comparison: −28.3%

Sly et al., 2001 [79]US (original data)Before/after with comparison1998-1999Youths, ages 12–17(I) 12-month campaign with TV, radio, billboard, display ads, promotional items (stickers, lanyards, hats, t-shirts, etc.), 1600 GRPs per quarter, 1800;
(C) control, 1000
At least a puff or two in the past 30 daysBaseline prevalence
(I) 13.8%
(C) 12.6%
12-month prevalence
(I) 12.6%
(C) 14.1%
Percentage change
(I) −8.9%
(C) 11.9%
𝑃 < 0 . 0 5 %

Farrelly et al., 2005 [78]US (MTF)Before/after w/o comparison1997–2002Youths, grades 8, 10, and 12(I) 24-month TV campaign with 3867–20367 GRPs (cumulative exposure over 2-year period for the lowest and highest quintiles of exposure)Any smoking in past 30 daysPercentage annual change in prevalence at 0–2 years prior to intervention:
Total: −3.2% (−3.8, −2.6)
8th: −3.4% (−4.6, −2.1)
10th: −4.6% (−5.6, −3.6)
12th: −1.8% (−2.7, −1.0)
Percentage annual change in prevalence at 0–2 years after intervention:
Total: −6.8% (−7.5, −6.1)
8th: −9.0% (−10.4, −7.6)
10th: −8.7% (−9.8, −7.5)
12th: −5.1% (−6.1, −3.9)

*Additionally, there were 2 other intervention groups that included sweepstakes. Since sweepstakes are not a focus of this paper, they are not included here.
This was part of a cardiovascular disease prevention campaign.
This was part of a coronary risk factor campaign.
C: control group; CI: confidence interval; COMMIT: Community Intervention Trial for Smoking Cessation; GRPs: gross rating points; HR: hazard ratio; I: intervention group; MTF: Monitoring the Future: a Continuing Study of American Youth; NLSY97: National Longitudinal Survey of Youth 1997; NR: not reported; NS: not significant; OR: odds ratio; RCT: randomized controlled trial; RR: relative risk; TV: television; US: United States.