This article appeared in a journal published by Elsevier. The attached
copy is furnished to the author for internal non-commercial research
and education use, including for instruction at the authors institution
and sharing with colleagues.
Other uses, including reproduction and distribution, or selling or
licensing copies, or posting to personal, institutional or third party
websites are prohibited.
In most cases authors are permitted to post their version of the
article (e.g. in Word or Tex form) to their personal website or
institutional repository. Authors requiring further information
regarding Elsevier’s archiving and manuscript policies are
encouraged to visit:
Author's personal copy
Smoke-free laws and adult smoking prevalence
Ellen J. Hahna,b,⁎, Mary Kay Rayensa,b, Karen M. Butlera, Mei Zhanga, Emily Durbina, Doug Steinkec
aUniversity of Kentucky, College of Nursing, 760 Rose Street, Lexington, KY 40536-0232, USA
bUniversity of Kentucky, College of Public Health, Lexington, KY, USA
cUniversity of Kentucky, College of Pharmacy, Lexington, KY, USA
A B S T R A C T A R T I C L EI N F O
Available online 29 April 2008
Adult smoking prevalence
Objective. To evaluate whether the adult smoking rate changed in Lexington-Fayette County, Kentucky,
following the enactment of a smoke-free public places ordinance.
Methods. Behavioral Risk Factor Surveillance System (BRFSS) data from 2001–2005 were used to test
whether smoking rates changed in Fayette County from the pre- to post-law period, relative to the change in
30 Kentucky counties with similar demographics. The sample consisted of 10,413 BRFSS respondents: 7139
pre-law (40 months) and 3274 post-law (20 months).
Results. There was a 31.9% decline in adult smoking in Fayette County (25.7% pre-law to 17.5% post-law). In
the group of 30 Control counties, the rate was 28.4% pre-law and 27.6% post-law. Controlling for seasonality,
time trend, age, gender, ethnicity, education, marital status, and income, there was a significant Time (pre- vs.
post-law) by Group (Fayette vs. Controls) interaction. There were an estimated 16,500 fewer smokers in
Fayette County during post-law compared to pre-law.
Conclusion. There was a significant effect of smoke-free legislation on adult smoking rates.
© 2008 Elsevier Inc. All rights reserved.
Smokers in communities with comprehensive smoke-free work-
place ordinances are more likely to quit than those who live in
communities with no smoke-free workplace laws (Moskowitz et al.,
2000). However, few studies have examined population smoking
prevalence and smoking cessation rates (Levy et al., 2004b) as
outcomesof smoke-free legislation.Voluntaryrestrictionson smoking
in public places and private workplaces reduce both smoking
prevalence and average daily cigarette consumption (Heloma and
Jaakkola, 2003, Farrelly et al., 1999, Brownson et al., 2002, Levy et al.,
2004a, Longo et al., 2001, Evans et al., 1999, Chaloupka and Saffer,
1992, Chaloupka and Wechsler, 1997, Townsend, 1998) and increase
cessation attempts (Hopkins et al., 2001, Farkas et al.,1999, Hammond
et al., 2004). Smoke-free workplaces are associated with a 29% drop in
cigarette consumption (Glasgow et al.,1997). Restrictions on smoking
may alter the perceived norms related to smoking by changing
attitudes concerning the social acceptability of smoking (U.S. Depart-
ment of Health and Human Services, 1994) and increase public
awareness about the dangers of cigarette smoking (Evans et al.,1999).
On April 27, 2004, Lexington-Fayette County, Kentucky, imple-
mented a smoke-free ordinance prohibiting smoking in all public
buildings including restaurants, bars, bingo parlors, pool halls, public
areas of hotels/motels, and all other buildings open to the public.
Kentucky leads the U.S. in smoking prevalence, with 28.6% of adults
who smoke cigarettes, compared to 20.2% nationally (Centers for
Disease Control and Prevention, 2007). In the United States, from
1997–2001, smoking cigarettes and exposure to secondhand smoke
were estimated to result in 438,000 premature deaths and 5.5 million
years of potential life lost (Centers for Disease Control and Prevention,
2005). In 1998, smoking-attributable health care costs were estimated
at $75.5 billion, accounting for 6 to 14% of personal health expen-
ditures (Max, 2001, Warner et al., 1999).
This study determined whether there was a change in the rate of
adult smoking in Lexington-Fayette County following implementation
of a smoke-free public places ordinance and evaluated how the
smoking rates pre- and post-law in Fayette compared to smoking
prevalence in a group of Control counties with similar demographics,
but without smoke-free laws.
The study was a quasi-experimental, two-group design. Behavioral Risk Factor
Surveillance Survey (BRFSS) data from 2001–05 were used to test whether adult
smoking rates changed significantly in Fayette County from pre- to post-law, relative to
the degree of change during the same timeframe in the Kentucky counties that were
most similar in education, income and smoking prevalence but that did not have
smoke-free laws. To form the Control group, Fayette County and its contiguous counties
were omitted from consideration, as were two counties with partial smoke-free laws
enacted during the study period; all retained counties did not have a smoke-free law.
The remaining 112 counties and Fayette County were ordered on three demographic
characteristics: education (percent of adults aged 25 or above with a high school
diploma from the 2000 U.S. Census); income (median household income from the 2000
Preventive Medicine 47 (2008) 206–209
⁎ Corresponding author. University of Kentucky, College of Nursing, 760 Rose Street,
Lexington, KY 40536-0232, USA.
E-mail address: email@example.com (E.J. Hahn).
0091-7435/$ – see front matter © 2008 Elsevier Inc. All rights reserved.
Contents lists available at ScienceDirect
journal homepage: www.elsevier.com/locate/ypmed
Author's personal copy
U.S. Census); and smoking rate (percent of current smokers in the population aged 18
and above from the BRFSS, aggregated from 2000–02). The counties were ranked for
each demographic factor (from high to low for education and income and low to high
for smoking rate). An index score was formed for each county by summing the three
corresponding ranks. A cohort of the top 30 counties from among this group of 112 was
chosen, representing one-fourth of the counties in Kentucky. Fayette County had the
third highest index score when combined with the accessible population of Control
The 30 counties in the Control group were similar to Fayette County on education,
income and smoking prevalence at baseline (see Table 1). For example, the percent of
adults in Fayette County with at least a high school diplomawas 85.8%, comparedwith a
population-weighted percentage of 79.3% in the group of 30 Control counties. Both
estimates exceed the percentage of high school graduates among the entire group of
112 potential Control counties. There were similar relationships for household income
and smoking status, with the weighted values for the 30 Control counties more similar
toFayetteCounty than the weighted values for the entirecohortof 112 potential Control
The national BRFSS survey of adults aged 18 and older, administered by the
Behavioral Surveillance Branch (BSB) of the Centers for Disease Control and Prevention
and by the states, uses a disproportionate stratified random sampling design. States
conduct the telephone interviews and forward data to the BSB, which weights the data
to account for probability of selection and differential participation by age, sex, and
The BRFSS items included current smoking status, date of interview, county of
residence, age, gender, ethnicity, marital status, education, and household income.
Current smokers reported smoking cigarettes ‘some days’ or ‘every day’ and they
smoked at least 100 cigarettes in their lifetime. Nonsmokers included both former and
The sample for this study included 10,413 BRFSS respondents: 7139 pre-law and
3274 post-law. Of the pre-law sample, 579 (8.1%) were from Fayette County, while 281
(8.6%) of the post-law sample lived in Fayette; the remaining respondents were from
the Control counties. The pre-law period was January 2001 to April 2004 (40 months);
post-law was May 2004 to December 2005 (20 months). All BRFSS data were analyzed
using methods appropriate for survey data in SAS, taking into account the sampling
scheme and data weights (SAS Institute, Cary, NC); data weights were adjusted prior to
analysis to account for multiple years of survey data. The SAS SURVEYFREQ procedure
was used to obtain the weighted smoking prevalence estimates (and corresponding
confidence intervals). SURVEYLOGISTIC was used to test for differences in the smoking
rate over time and between Fayette and the Control counties, controlling for time trend,
seasonality, and respondent demographic factors. Contrasts were used to assess post-
hoc pairwise comparisons of the interaction effect.
Smoking rates were compared between the two groups and between the two time
periods, controlling for month of the year (as an adjustment for seasonality), time
(a continuous variable to control for secular trends), and the respondent demographic
characteristics of age in years, gender, ethnicity (White versus minority), education
(categorical, with four categories ranging from less than high school to a college
degree), marital status (married versus other), and household income (ordinal, with
eight categories). The two categorical variables with more than two levels, month of
year and education, were included in the model using dummy coding (with 11 and
3 indicator variables, respectively).
The weighted smoking rate and 95% confidence interval (CI) for
Fayette County pre-law, on average, was 25.7% (CI: 21.2–30.1), and this
declinedto17.5%(CI: 11.8–23.1) post-law, a decrease of 31.9%.In the 30
counties without a law (Control group), the rates in the pre-law and
post-law periods were 28.4% (CI: 26.8–30.0) and 27.6% (CI: 25.2–30.0),
respectively. As shown in Fig. 1, Fayette County had smoking rates
similar to the group of Control counties for the 40 months before the
smoke-free law, while Fayette versus Control group differences are
Descriptive summary of the three demographic indicators used to rank counties for
inclusion in the Control group
Percent of adults aged 25 and above
with a high school diplomab
Median annual household incomeb
Percent of adults aged 18 and above
who were current smokersc
aEstimates for the control and potential control groups were obtained by weighting
the county-level values by the corresponding population size.
bEstimated from the 2000 U.S. Census for Fayette County, Kentucky and subsets of
potential Control counties in Kentucky.
cEstimated from the 2000–02 Behavioral Risk Factor Surveillance Survey summaries
for Fayette County, Kentucky and subsets of potential Control counties in Kentucky.
Fig. 1. Weighted percent of smokers for each four-month period by group (N=10,413);
estimates obtained from Behavioral Risk Factor Surveillance Survey data (2001–06), for
Fayette County, Kentucky and 30 Control counties in Kentucky.
Logistic regression estimates from the model to determine differences between Fayette
County (with a smoke-free law) and Control counties (all without laws) and differences
between the pre- and post-law periods in the smoking among adults, using Behavioral
Risk Factor Surveillance Survey data (2001–06)
interval for OR
Month of year
(December is the reference)a
Gender (female=1; male=0)
Ethnicity (white=1; minority=0)
Education (college degree is the reference)a
Marital status (married=1; other=0)
Post-law indicator (post=1; pre=0)
Fayette County indicator
(Fayette=1; other county=0)
aFor the nominal variables with more than 2 categories (month of year, education),
the odds ratios are for the comparison category relative to the reference category, while
the regression coefficients are estimates for the comparison category indicator in the
model, with corresponding p-value levels shown.
E.J. Hahn et al. / Preventive Medicine 47 (2008) 206–209
Author's personal copy
Regression estimates from the logistic model are displayed in
differences in personal characteristics, the interaction between Time
(pre- vs. post-law) and Group (Fayette vs. Controls) was significant
(Wald χ2=5.5, p=.02). The magnitude of change from pre-law to post-
law was significantly different for Fayette County compared to the
Control counties, controlling for seasonality, secular trends, and
demographic characteristics. Post-hoc analysis determined that
while Fayette and the Control counties did not differ in smoking rate
during the pre-law period (Wald χ2=0.2, p=.7), Fayette was sig-
nificantlylower than Controls during post-law(Wald χ2=7.5, p=.006).
The degree of decrease over time in Fayette was significant (Wald
χ2=4.4, p=.04), while the change from pre- to post-law in the Control
counties was not significant (Wald χ2=0.1, p=.7).
Based on the weighted smoking rates (25.7% pre-law and 17.5%
post-law) and the Fayette County population, it is estimated that
during the 40-month pre-law there were, on average, 53,437 adult
cigarette smokers living in Fayette County compared with an average
of 36,982 during the first 20 months after the law took effect, a
decrease of approximately 16,500 smokers.
Adult smoking prevalence declined by nearly one-third during
the 20 months after implementation of a smoke-free public places
ordinance in Lexington-Fayette County, Kentucky. The fact that adult
smoking did not decline in Kentucky counties without smoke-free
laws and with similar educational attainment, income and smoking
rates pre-law indicates that the smoke-free law was associated with
a significant decline in adult smoking rates. During the 20-month
post-law period, there was no change in access to tobacco treatment
activities or tobacco control programs in Fayette County; there
were no community-wide discounts for tobacco treatment medica-
tions and no media campaigns promoting quitting. There were
reports that tobacco company marketing had declined in Fayette
County bars and nightclubs frequented by college students after the
law went into effect (Ridner et al., submitted for publication), but
the decrease in adult smoking prevalence in Lexington-Fayette
County was significant even with age included as a control variable
in the logistic model.
The magnitude of the decline in adult smoking prevalence in this
study (32%) is higher than reported in other related studies; one
reason may be that the outcome measures and/or types of restrictions
were different. In Italy, a smoke-free policy covering all indoor public
places accounted for an 8% decrease in cigarette consumption in the
short run (Gallus et al., 2006). Evans et al. (1999) report that voluntary
workplace bans lead to a 5.7 percentage point decline in smoking
prevalence and a decrease in average cigarette consumption of 2.3
cigarettes per day. Although the Evans et al. study did not examine the
effect of community-wide smoke-free ordinances, our results are
similar in that the decline in smoking prevalence after Lexington's
smoke-free law was 8.2 percentage points, from 25.7% to 17.5%. Given
the small number of smoke-free laws in Kentucky, one limitation of
this study is that the data are from one county with a smoke-free law
and thirty without; this precluded controlling for differences in
county-level risk factors or secular trends and did not allow for
determining testing of differences in the smoking rate by strength of
law, as has been done in other studies (Juster et al., 2007). Future
research to assess the link between smoke-free laws and smoking rate
will benefit from the inclusion of more geographic units with these
laws, and by determining if there are differences in the smoking rate
by strength of law.
Implementation of comprehensive tobacco control programs,
including policy change, can reduce the prevalence of smoking
along with its associated mortality and health care costs (Centers for
Disease Control and Prevention, 2004). Based on the estimate of
16,500 fewer smokers post-law in Fayette County, we projected an
annual cost savings of approximately $21 million 1998 USD in
smoking-attributable expenditures (95% CI $19.9–$22.2 million 1998
USD) using an ecological approach (Miller et al., 1999). This method
relies on the smoking-attributable fraction (SAF) and may be an
oversimplified model because it does not control for individual-level
lifetime costs of smoking, including those due to early death and
disability (Sloan et al., 2004). Further research is needed to document
the effect of smoke-free policies on the all-inclusive cost of smoking
using a life cycle approach (Sloan et al., 2004).
Globally, 29% of people 15 years or older were regular smokers
in 1995 (Jha et al., 2002). In 2000, it is estimated that 4.84 million
premature deaths attributable to smoking occurred worldwide (Ezzati
and Lopez, 2004). Although there are no global estimates of smoking-
attributable medical expenditures, reported annual costs range from
$88.09 million (Korea) (Kang et al., 2003) to $2.68 billion (Canada)
(Single et al., 1998). The enactment of smoke-free laws may be one
cost-effective way to reduce premature death and healthcare costs
due to smoking worldwide.
Previous studies have focused primarily on the effect of voluntary
smoke-free workplace policies on smoking prevalence (Levy et al.,
2004b). This study focuses on the impact of a municipal smoke-free
ordinance on a population measure of adult smoking prevalence.
While smoke-free legislation is typically enacted to protect non-
smokers from secondhand smoke, this study provides evidence that
smoke-free laws may also positively affect the health of both current
smokers and those at risk of initiation.
Brownson, R., Hopkins, D., Wakefield, M., 2002. Effects of smoking restrictions in the
workplace. Annu. Rev. Public Health 23, 333–348.
Centers For Disease Control And Prevention, 2004. The Health Consequences of
Smoking: a Report of the Surgeon General. US Department of Health and Human
Services, Atlanta, GA.
Centers For Disease Control And Prevention, 2005. Annual smoking-attributable
mortality, years of potential life lost, and productivity losses—United States,
1997–2001. MMWR 54, 625–628.
Centers For Disease Control And Prevention, 2007. State-specific prevalence of cigarette
smoking among adults and quitting among persons aged 18–35 years — United
States, 2006. MMWR 56, 993–996.
Chaloupka, F., Saffer, H., 1992. Clean indoor air laws and the demand for cigarettes.
Contemp. Policy Issues 64, 72–83.
Chaloupka, F.J., Wechsler, H., 1997. Price, tobacco control policies and smoking among
young adults. J. Health Econ. 16, 359–373.
Evans, W.N., Farrelly, M.C., Montgomery, E., 1999. Do workplace smoking bans reduce
smoking? Am. Econ. Rev. 89, 728–747.
Ezzati, M., Lopez, A., 2004. Regional, disease specific patterns of smoking-attributable
mortality in 2000. Tob. Control 13, 388–389.
Farkas, A.J., Gilpin, E.A., Distefan, J.M., Pierce, J.P., 1999. The effects of household and
workplace smoking restrictions on quitting behaviours. Tob. Control 8, 261–265.
Farrelly, M.C., Evans, W.N., Sfekas, A.E.S., 1999. The impact of workplace smoking bans:
results from a national survey. Tob. Control 8, 272–277.
Gallus, S., Zuccaro, P., Colombo, P., Apolone, G., Pacifici, R., Garattini, S., La Vecchi, C.,
2006. Effects of new smoking regulations in Italy. Ann. Oncol. 17, 346–347.
Glasgow, R.E., Cummings, K.M., Hyland, A., 1997. Relationship of worksite smoking
policy to changes in employee tobacco use: findings from COMMIT. Community
Intervention Trial for Smoking Cessation. Tob. Control 6 (Suppl 2), 44–48.
Hammond, D., Mcdonald, P., Fong, G., Brown, K., Cameron, R., 2004. The impact of
cigarette warning labels and smoke-free bylaws on smoking cessation. Revue
Canadienne De Sante Publique. 95, 201–204.
Heloma, A., Jaakkola, M., 2003. Four-year follow-up of exposure, attitudes, and smoking
behaviour following enactment of Finland's national smoke-free workplace law.
Addiction 98, 1111–1117.
Hopkins, D.P., Briss, P.A., Ricard, C.J., Husten, C.G., Carande-kulis, V.G., Fielding, J.E., Alao,
M.O., Mckenna, J.W., Sharp, D.J., Harris, J.R., Woollery, T.A., Harris, K.W., 2001.
Reviews of evidence regarding interventions to reduce tobacco use and exposure to
environmental tobacco smoke. Am. J. Prev. Med. 20, 16–66.
Jha, P., Ranson, M.K., Nguyen, S.N., Yach, D., 2002. Estimates of global and regional
smoking prevalence in 1995, by age and sex. Am. J. Public Health 92, 1002–1006.
Juster, H.R., Loomis, B.R., Hinman, T.M., 2007. Declines in hospital admissions for acute
myocardial infarction in New York State after implementation of a comprehensive
smoking ban. Am. J. Public Health 97, 2035–2039.
E.J. Hahn et al. / Preventive Medicine 47 (2008) 206–209
Author's personal copy Download full-text
Kang, H., Kim, H., Park, T., Jee, S., Nam, C., Park, H., 2003. Economic burden of smoking in
Korea. Tob. Control 12, 37–44.
Levy, D., Chaloupka, F., Gitchell, J., 2004a. The effects of tobacco control policies on
Levy, D., Romano, E., Mumford, E., 2004b. Recent trends in home and work smoking
bans. Tob. Control 13, 258–263.
Max, W., 2001. The financial impact of smoking on health-related costs: a review of the
literature. Am. J. Health Promot. 15, 321–331.
Miller, V., Ernst, C., Collin, F.,1999. Smoking-attributable medical care costs in the U.S.A.
Soc. Sci. Med. 48, 375–391.
Moskowitz, J.M., Lin, Z., Hudes, E.S., 2000. The impact of workplace smoking ordinances
in California on smoking cessation. Am. J. Public Health 90, 757–761.
Ridner, S.L., Walker, K.L. & Hahn, E.J., submitted for publication. Tobacco marketing
among college students and smoke-free laws.
Single, E., Robson, L., Xie, X., Rehm, J., 1998. The economic costs of alcohol, tobacco and
illicit drugs in Canada, 1992. Addiction 93, 991–1006.
The MIT Press, Cambridge, MA.
Townsend, J.L., 1998. UK smoking targets: policies to attain them and effects on
premature mortality. In: Abedian, I., Vander Merwe, R., Wilins, N., Jha, P. (Eds.), The
Economics of Tobacco Control: Toward an Optimal Policy Mix. Applied Fiscal
Research Centre, University of Cape Town, Cape Town, SA.
U.S. Department Of Health And Human Services (1994) Preventing tobacco use among
young people: a report of the Surgeon General. Atlanta, GA, U.S. Department of
Health and Human Services, Public Health Service, Centers for Disease Control and
Prevention, National Center for Chronic Disease Prevention and Health Promotion,
Office of Smoking and Health.
Warner, K., Hodgson, T., Carroll, C., 1999. Medical costs of smoking in the United
States: estimates, their validity, and their implications. Tob. Control 8,
E.J. Hahn et al. / Preventive Medicine 47 (2008) 206–209