摘要：Objectives. We examined whether state tobacco control programs are effective in reducing the prevalence of adult smoking. Methods. We used state survey data on smoking from 1985 to 2003 in a quasi-experimental design to examine the association between cumulative state antitobacco program expenditures and changes in adult smoking prevalence, after we controlled for confounding. Results. From 1985 to 2003, national adult smoking prevalence declined from 29.5% to 18.6% ( P <.001). Increases in state per capita tobacco control program expenditures were independently associated with declines in prevalence. Program expenditures were more effective in reducing smoking prevalence among adults aged 25 or older than for adults aged 18 to 24 years, whereas cigarette prices had a stronger effect on adults aged 18 to 24 years. If, starting in 1995, all states had funded their tobacco control programs at the minimum or optimal levels recommended by the Centers for Disease Control and Prevention, there would have been 2.2 million to 7.1 million fewer smokers by 2003. Conclusions. State tobacco control program expenditures are independently associated with overall reductions in adult smoking prevalence. Recent data from the Centers for Disease Control and Prevention (CDC) showed that adult smoking remained constant at 20.8% from 2004 to 2005 after years of steady decline. 1 The CDC study cited a 27% decline in funding for tobacco control programs from 2002 through 2006 and smaller annual increases in cigarette prices in recent years as 2 possible explanations for stalled smoking rates. Our study is a systematic assessment of the association between adult smoking, funding for state tobacco control programs, and state cigarette excise taxes. In 1989, California began the first comprehensive statewide tobacco control program in the United States after passage of a state ballot measure that raised cigarette excise taxes by $0.25. 2 Comprehensive programs include interventions such as mass media campaigns, increased cigarette excise taxes, telephone quit lines, reduced out-of-pocket costs for smoking cessation treatment, health care provider assistance for cessation, and restrictions on secondhand smoke in public places. 3 – 6 Subsequently, other states, including Massachusetts in 1992, Arizona in 1995, and Florida in 1998, began similar large-scale state tobacco control programs. 3 Multistate tobacco control interventions with substantial financial support began in the 1990s, with assistance from US government programs (e.g., the CDC’s Initiatives to Mobilize for the Prevention and Control of Tobacco Use [IMPACT] and the National Cancer Institute’s Americans Stop Smoking Intervention Study [ASSIST]) and other national programs. 3 Some states also committed resources from other sources, such as revenue from the 1998 Master Settlement Agreement (MSA) between the 4 largest tobacco companies in the United States and 46 US states. 7 The MSA imposes restrictions on the advertising, promotion, and marketing or packaging of cigarettes, including a ban on tobacco advertising that targets people younger than 18, and requires the tobacco companies to pay $246 billion over 25 years to the states. The MSA also established a foundation that became the American Legacy Foundation. Extensive research has shown that state tobacco control programs, combined with other efforts, such as the American Legacy Foundation’s national truth campaign, have been effective in reducing adolescent tobacco use. 3 , 8 , 9 Following a large increase in adolescent smoking during the mid-1990s, there has been an unprecedented decline, with the national prevalence among high school students dropping from 36.4% in 1997 to 21.9% in 2003. 10 In marked contrast, there has been little research into the effects of state programs on the prevalence of adult smoking, which is unfortunate given that smoking cessation confers substantial health benefits to adults. 3 , 11 , 12 To date, findings from California, Massachusetts, and Arizona suggest that state tobacco control programs have had some effect on adults. 13 – 16 From 1988 through 1999, the prevalence of adult smoking in California declined from 22.8% to 17.1%, compared with an overall national decline from 28.1% to 23.5% (a relative percentage decline of 25% in California and 16% elsewhere). 13 , 14 From 1992 through 1999, the relative percentage decline in adult smoking was 8% in Massachusetts compared with 6% nationwide. 14 , 15 Findings from Arizona from 1996 to 1999 suggest a greater effect: the relative percentage decline was 21% compared with 8% nationwide. 16 In addition, per capita cigarette sales—a proxy for cigarette consumption—have declined faster in Arizona, California, Massachusetts, and Oregon (where another large-scale program began in 1997) than in the rest of the United States since the programs’ implementation. 17 The ASSIST evaluation showed that smoking prevalence decreased more in ASSIST states than in non-ASSIST states by the end of an 8-year intervention; by contrast, the evaluation found no difference in per capita cigarette consumption. 6 , 18 These few state-specific studies on the prevalence of adult smoking had important limitations. First, state-specific findings may not be generalizable. Second, none of the studies considered the key role of cigarette price increases on prevalence (i.e., through higher cigarette excise taxes, which have consistently been shown to reduce cigarette consumption and prevalence) 3 or controlled for other state characteristics, such as demographic changes or secular trends. Third, the studies did not assess the potential effects of programs on adults of different ages. Although the ASSIST evaluation provides a more comprehensive view of state tobacco control programs, it failed to control for baseline differences in state-level demographics and policy variables between ASSIST and non-ASSIST states. Finally, none of the studies considered the possible long-term effects of tobacco control programs on adult smoking. In 1999, the CDC published Best Practices for Comprehensive Tobacco Control Programs , 19 which provided states with guidelines and recommendations for 9 tobacco control program activities (e.g., community programs, counter-marketing, cessation), along with minimum and optimum funding levels for each specific activity. On the basis of this document, in fiscal year 2006, states should have allocated $6.47 per capita minimum and $17.14 optimally to tobacco control programs (i.e., the $5.98 and $15.85, respectively, recommended in the 1999 CDC document, adjusted for inflation). We used data on state tobacco control program expenditures and periodic surveys of adult smoking prevalence conducted by the US Census Bureau from 1985 to 2003 to answer the following questions: (1) After control for potentially confounding factors (e.g., cigarette excise taxes), were increases in state tobacco control program expenditures independently associated with declines in adult smoking prevalence, and did effects differ by age group? (2) What would have been the predicted effect of state tobacco control program expenditures on adult smoking prevalence if all states had met CDC-recommended minimum or optimum per capita funding levels from 1995 to 2003?
