Abstract
The United States (US) has identified income-based disparities in smoking as a critical public health issue, but the extent to which these disparities are changing over time within states is not well documented. This study examined recent trends in current cigarette smoking in each state and the District of Columbia by self-reported annual household income. Data came from the Behavioral Risk Factor Surveillance System, a state-representative survey of US adults. Sample sizes for each state and year ranged from 2,914 to 36,955 participants. We fit logistic regression models to examine linear time trends in cigarette smoking status in each state between 2011 and 2017. In every state, the odds of smoking were 1.4 to 3.0 times greater in the lower-income group as compared to the higher-income group in 2017. Among 47 states, linear time trends in smoking did not significantly differ by income group, suggesting no change in income-based disparities. In three states (Florida, Maine, West Virginia) disparities widened, primarily because smoking prevalence only dropped among higher-income groups. Disparities declined in only one state. In New York, smoking prevalence declined more for lower-income groups compared to higher-income groups. Findings from this study suggest that little progress has been made toward reducing income-based differences in smoking and additional policy and tobacco control efforts may be required to meet national disparity reduction goals.
Keywords: tobacco use, health disparities, income
Despite marked reductions in smoking prevalence in the United States (US), disparities in smoking by income persist. In 2017 14% of US adults were current cigarette smokers.1 Smoking prevalence was higher among individuals reporting an annual household income less than $35,000 (21%) as compared to those reporting an income of $35,000 - $74,999 (15%).1 The disparity in smoking prevalence was even greater when compared to higher-income groups.1 Nationally, although smoking prevalence has decreased among all income groups, the greatest declines have been among higher-income groups.2,3
The Centers for Disease Control and Prevention (CDC) states that reducing tobacco-related disparities is a priority for tobacco control.4 Numerous reports, including those by the Surgeon General, have documented income-based disparities in smoking and the consequences of these disparities on health outcomes.4,5,6 One of the major conclusions of the 2014 Surgeon General’s Report, The Health Consequences of Smoking-50 Years of Progress, was that large disparities in smoking according to socioeconomic status remain to be addressed.6 CDC has launched initiatives, such as the National Networks for Tobacco Control and Prevention, that focus on reducing income-based disparities in tobacco use.4
Federal initiatives and policies, however, are only one component of tobacco control efforts. States have significant authority to implement tobacco control programs and policies. Studies suggest that certain tobacco control policies have differential effects dependent on income.7 For example, raising the price of cigarettes reduces smoking more among lower-income smokers compared to higher-income smokers.8,9 State programs support other activities that may reduce income-based disparities in smoking, such as increasing the number of people covered by smoke-free air laws and reducing exposure to tobacco industry advertising.4,10 Given variation in state tobacco control efforts, as well as other policies affecting health and income, income-based differences in smoking likely vary across states. Tracking state-specific trends in these differences could identify those places making progress toward national equity.
To examine whether progress has been made in reducing income-based disparities in smoking in the different US states, the present study used data from the annual, state-representative Behavioral Risk Factor Surveillance System (BRFSS) to examine disparities in cigarette use by self-reported annual household income, as well as changes in those disparities in each state and the District of Columbia (DC) from 2011 to 2017. To our knowledge, this is the first state-based analysis to examine recent trends in income-based disparities in the US.
Methods
Data were drawn from the BRFSS, a state-based, random-digit-dialed telephone (landline and cellular telephone) survey that collects data on the non-institutionalized adult population (≥18 years) in the US. Detailed information about the BRFSS survey design and methods are available at www.cdc.gov/brfss. The survey is conducted annually in 50 states, Washington, D.C. (DC), and participating US territories (Guam, Puerto Rico, US Virgin Islands). In the present study, BRFSS data from the 50 states and DC were used. The 50 states and DC will be referred to as states. The BRFSS core survey includes questions about current cigarette smoking. For the present study, BRFSS data from 2011 to 2017 were used. Over 350,000 respondents completed a survey each year between 2011 and 2017. Sample sizes for each state and year ranged from 2,914 to 36,955 participants (Appendix 1), and the BRFSS weighting procedures make the survey state representative. The median response rate across states ranged from 45.2% to 49.7% over the time period, comparable to other federal surveys.11 The BRFSS makes up to 15 attempts to reach a respondent before designating the respondent as non-responsive.11
Measures
Smoking.
Cigarette smoking was assessed in the core survey using two questions. Respondents were asked, “Have you smoked at least 100 cigarettes in your entire life?” and “Do you now smoke cigarettes every day, some days, or not at all?” Current cigarette smokers were defined as respondents who reported smoking at least 100 cigarettes in their entire lifetime and currently smoke every day or some days.
Demographic Characteristics.
Demographic characteristics were collected in the core questionnaire. The following variables and categories were employed for analysis: age group (18–24, 25–34, 35–44, 45–54, 55–64, 65 or older); sex (male, female); race/ethnicity (non-Hispanic White only, non-Hispanic Black only, non-Hispanic Other race only/Multiracial, Hispanic); educational attainment (did not graduate high school, graduated high school, attended college or technical school, graduated from college or technical school); annual household income (less than $15,000, $15,000 to less than $25,000, $25,000 to less than $35,000, $35,000 to less than $50,000, $50,000 to less than $75,000, and $75,000 or more); and survey year (2011, 2012, 2013, 2014, 2015, 2016, 2017).
Data Analysis
Descriptive statistics were examined for the study sample. Due to a large percentage of missing data on income (15.5%), multiple imputation was used to generate 25 imputed datasets following procedures recommended for use with BRFSS data.12 Specifically, the fully conditional specification method was used to impute missing income data.13 First, variables that were correlated (i.e., Cramer’s V greater than 0.1) with income or missingness of income were identified as potential covariates for the imputation model.12 Next, logistic regression with stepwise model selection was used to eliminate redundant covariates from the final imputation model. The final imputation model included all non-redundant covariates and the sampling weight.
Within each state, respondents who were in the lower 33% of all reported annual household incomes were categorized as having lower-income. Respondents who reported an annual household income within the upper-two thirds of all reported incomes in their state were categorized as having higher-income. This cut-off was selected to reflect the percentage of the population that is poor in the US; approximately one-third of the US population is considered poor or nearly poor.14 Across states, respondents in the lower-income groups reported a maximum annual household income of less than $35,000. The majority of respondents in the higher-income groups reported an annual household income of $35,000 or higher. For descriptive purposes, smoking prevalence with 95% confidence intervals were obtained in 2011 and 2017 for the lower- and higher-income groups. A ratio that compared smoking by income groups was calculated by dividing the smoking prevalence of the lower-income group by that of the higher-income group.
Next, state-specific linear time trends of smoking status were assessed using logistic regression models that included BRFSS data from 2011 through 2017. For each state, logistic regression models were fit to examine the relationship between year (2011–2017), income (lower-income group=1; higher-income group=0), and cigarette smoking status (current smoker=1; non-current smoker=0), adjusting for age, sex, race/ethnicity, and education level. To determine if there were differential time trends in smoking by income group, a subsequent logistic regression model included an interaction term between year and income. For statistically significant interaction terms, simple effects tests were conducted to examine the linear time trend separately in each income group. Analyses were conducted using the BRFSS sampling weights and complex survey procedures in SAS 9.4 (SAS V.9.4, SAS Institute Inc).
Results
Smoking Prevalence among Lower- and Higher-Income Groups
Table 1 provides sociodemographic characteristics of the study sample. In 2011, smoking prevalence among the lower-income groups ranged from 18.6% to 40.3% (Table 2). Among the higher-income groups, smoking prevalence ranged from 7.9% to 22.8%. The ratio of smoking rates between the lower- and higher- income groups ranged from 1.5 to 2.4. In 2017, smoking prevalence among the lower-income groups ranged from 15.1% to 38.5%. Among the higher-income groups, smoking prevalence ranged from 7.0% to 19.4%. The ratio of smoking prevalence between the lower- and higher- income groups ranged from 1.4 to 3.0. As shown in Figure 1, although smoking prevalence typically declined in both lower- and higher-income groups over the study period, income-based disparities in smoking prevalence were typically maintained.
Table 1.
Group | Sample size | Percentage of the Total Sample | Smoking Prevalence |
---|---|---|---|
Age | |||
18–24 | 220, 158,839 | 12.8% | 16.3% |
25–34 | 297,604,815 | 17.3% | 22.0% |
35–44 | 285,708,821 | 16.6% | 19.1% |
45–54 | 303513191 | 17.7% | 19.9% |
55–64 | 281,321,111 | 16.4% | 17.2% |
65 or older | 327,917,743 | 19.1% | 8.5% |
Sex | |||
Female | 880,865,485 | 51.3% | 19.3% |
Male | 835,180,909 | 48.7% | 14.9% |
Missing | 178,126 | <1% | 22.9% |
Income | |||
Less than $10,000 | 92,626,842 | 5.4% | 29.4% |
$10,000 to less than $15,000 | 82,701,729 | 4.8% | 26.9% |
$15,000 to less than $20,000 | 118,285,358 | 6.9% | 25.5% |
$20,000 to less than $25,000 | 137, 721,331 | 8.0% | 23.1% |
$25,000 to less than $35,000 | 157,355,655 | 9.2% | 20.3% |
$35,000 to less than $50,000 | 199700002 | 11.6% | 18.1% |
$50,000 to less than $75,000 | 217,493,157 | 12.7% | 15.1% |
$75,000 or more | 445,191,716 | 25.9% | 9.6% |
Missing | 265,148,729 | 15.4% | 14.0% |
Education | |||
Did not graduate high school | 246,667,136 | 14.4% | 26.5% |
Graduated high school | 486,573,230 | 28.4% | 21.6% |
Attended college or technical school | 526,324,450 | 30.7% | 17.0% |
Graduated from college or technical school | 447,269,335 | 26.1% | 7.0% |
Missing | 9,390,368 | <1% | 7.2% |
Race/ethnicity | |||
White only, non-Hispanic | 1,093,518, 887 | 63.7% | 18.0% |
Black only, non-Hispanic | 198,950,812 | 11.6% | 19.1% |
Other race only, non-Hispanic or Multiracial, non-Hispanic | 133,637,273 | 7.8% | 14.4% |
Hispanic | 261,325,378 | 15.2% | 13.1% |
Missing | 28,792,170 | 1.7% | 15.1% |
Year | |||
2011 | 235,054,070 | 13.7% | 19.8% |
2012 | 240,130,580 | 14.0% | 18.4% |
2013 | 243,095,138 | 14.2% | 17.4% |
2014 | 245,561,099 | 14.3% | 16.4% |
2015 | 248,437,417 | 14.5% | 15.8% |
2016 | 251,162,036 | 14.6% | 15.5% |
2017 | 252,784,180 | 14.7% | 15.5% |
Notes. BRFSS = Behavioral Risk Factor Surveillance System. The table presents descriptive statistics for combined BRFSS data from 2011 to 2017. The Other race only, non-Hispanic group includes persons who are American Indian or Alaskan Native only, non-Hispanic, Asian only, non-Hispanic, Native Hawaiian or other Pacific Islander only, non-Hispanic, and Other race only, non-Hispanic.
Table 2.
State | 2011 | 2017 | Change in ratio | ||||
---|---|---|---|---|---|---|---|
Lower income | Higher income | Ratioa | Lower income | Higher income | Ratioa | ||
Alabama | 33.4 (30.7, 36.0) | 18.4 (16.6, 20.2) | 1.8 | 32.2 (29.4, 35.0) | 15.7 (14.0, 17.3) | 2.1 | −0.3 |
Alaska | 35.1 (31.1, 39.2) | 16.3 (14.1, 18.6) | 2.2 | 33.0 (27.1, 38.9) | 16.1 (13.3, 19.0) | 2.0 | 0.2 |
Arizona | 27.0 (22.9, 31.1) | 15.1 (12.9, 17.3) | 1.8 | 24.0 (22.2, 25.8) | 13.0 (12.0, 13.9) | 1.8 | 0.0 |
Arkansas | 38.2 (34.3, 42.0) | 18.3 (15.9, 20.6) | 2.1 | 35.3 (31.1, 39.6) | 15.6 (13.0, 18.2) | 2.3 | −0.2 |
California | 18.6 (17.1, 20.1) | 11.1 (10.3, 11.9) | 1.7 | 17.2 (15.1, 19.2) | 10.5 (9.3, 11.7) | 1.6 | 0.1 |
Colorado | 26.4 (24.3, 28.6) | 13.1 (12.0, 14.3) | 2.0 | 24.2 (22.0, 26.3) | 11.5 (10.4, 12.5) | 2.1 | −0.1 |
Connecticut | 23.4 (20.7, 26.1) | 13.5 (12.0, 15.2) | 1.7 | 21.0 (18.7, 23.3) | 10.3 (9.1, 11.6) | 2.0 | −0.3 |
Delaware | 32.0 (27.8, 36.2) | 17.7 (15.7, 19.8) | 1.8 | 26.0 (22.1, 29.8) | 14.4 (12.5, 16.4) | 1.8 | 0.0 |
District of Columbia | 31.6 (27.8, 35.5) | 13.4 (11.2, 15.6) | 2.4 | 26.8 (23.4, 30.3) | 9.0 (7.4, 10.7) | 3.0 | −0.6 |
Florida | 24.7 (22.6, 26.7) | 16.1 (14.7, 17.5) | 1.5 | 24.3 (21.9, 26.7) | 12.7 (11.4, 14.0) | 1.9 | −0.4 |
Georgia | 30.7 (28.0, 33.4) | 15.4 (13.9, 16.9) | 2.0 | 25.8 (23.1, 28.5) | 14.7 (13.1, 16.3) | 1.8 | 0.2 |
Hawaii | 24.5 (21.8, 27.2) | 12.4 (10.9, 14.0) | 2.0 | 18.3 (16.1, 20.6) | 11.3 (9.9, 12.6) | 1.6 | 0.4 |
Idaho | 26.2 (22.6, 29.7) | 13.0 (11.3, 14.8) | 2.0 | 24.9 (21.2, 28.6) | 11.2 (9.6, 12.8) | 2.2 | −0.2 |
Illinois | 27.0 (23.8, 30.1) | 17.0 (14.9, 19.1) | 1.6 | 22.7 (20.1, 25.4) | 12.3 (10.8, 13.7) | 1.8 | −0.2 |
Indiana | 34.4 (31.8, 37.0) | 21.0 (19.4, 22.6) | 1.6 | 33.3 (31.2, 35.5) | 17.8 (16.7, 18.9) | 1.9 | −0.3 |
Iowa | 28.9 (26.6, 31.2) | 14.7 (13.3, 16.1) | 2.0 | 26.1 (23.8, 28.3) | 13.0 (11.9, 14.2) | 2.0 | 0.0 |
Kansas | 31.2 (29.8, 32.7) | 15.1 (14.2, 16.0) | 2.1 | 27.0 (25.5, 28.5) | 13.2 (12.4, 14.0) | 2.0 | 0.1 |
Kentucky | 40.3 (37.4, 43.3) | 22.7 (21.0, 24.5) | 1.8 | 37.1 (33.6, 40.6) | 19.4 (17.6, 21.2) | 1.9 | −0.1 |
Louisiana | 32.6 (30.1, 35.1) | 21.2 (19.5, 22.9) | 1.5 | 32.8 (29.6, 36.0) | 18.3 (16.4, 20.2) | 1.8 | −0.3 |
Maine | 34.0 (31.8, 36.3) | 17.0 (15.8, 18.1) | 2.0 | 30.4 (27.2, 33.6) | 13.8 (12.5, 15.2) | 2.2 | −0.2 |
Maryland | 28.2 (25.4, 31.0) | 14.4 (13.0, 15.9) | 2.0 | 23.3 (20.7, 25.9) | 10.9 (9.8, 12.0) | 2.1 | −0.1 |
Massachusetts | 25.6 (23.9, 27.4) | 13.9 (12.8, 15.0) | 1.8 | 22.9 (19.7, 26.0) | 10.4 (8.9, 11.9) | 2.2 | −0.4 |
Michigan | 33.6 (31.0, 36.3) | 18.4 (17.0, 19.9) | 1.8 | 31.4 (29.0, 33.8) | 15.4 (14.3, 16.6) | 2.0 | −0.2 |
Minnesota | 27.6 (25.7, 29.5) | 14.1 (13.0, 15.2) | 2.0 | 23.2 (21.5, 24.8) | 11.3 (10.5, 12.1) | 2.1 | −0.1 |
Mississippi | 34.6 (32.0, 37.2) | 21.6 (20.0, 23.3) | 1.6 | 33.1 (29.0, 37.1) | 18.9 (16.8, 20.9) | 1.8 | −0.2 |
Missouri | 37.1 (34.1, 40.1) | 18.8 (17.0, 20.6) | 2.0 | 34.3 (31.1, 37.5) | 16.1 (14.6, 17.6) | 2.1 | −0.1 |
Montana | 33.2 (30.7, 35.7) | 15.6 (14.1, 17.0) | 2.1 | 31.3 (27.9, 34.6) | 11.8 (10.3, 13.3) | 2.7 | −0.6 |
Nebraska | 28.5 (26.9, 30.1) | 16.5 (15.6, 17.3) | 1.7 | 25.7 (23.3, 28.2) | 12.6 (11.6, 13.6) | 2.0 | −0.3 |
Nevada | 32.0 (28.0, 36.0) | 17.9 (15.7, 20.1) | 1.8 | 24.1 (20.2, 28.1) | 15.2 (12.9, 17.4) | 1.6 | 0.2 |
New Hampshire | 29.7 (26.9, 32.6) | 13.9 (12.2, 15.5) | 2.1 | 26.2 (22.6, 29.8) | 12.5 (10.6, 14.4) | 2.1 | 0.0 |
New Jersey | 21.6 (19.9, 23.4) | 14.0 (12.9, 15.1) | 1.5 | 18.4 (16.2, 20.6) | 12.3 (10.9, 13.7) | 1.5 | 0.0 |
New Mexico | 28.9 (26.8, 31.1) | 16.2 (14.8, 17.6) | 1.8 | 26.4 (23.6, 29.1) | 13.0 (11.4, 14.6) | 2.0 | −0.2 |
New York | 26.7 (24.1, 29.4) | 13.8 (12.4, 15.1) | 1.9 | 20.6 (18.5, 22.6) | 12.7 (11.6, 13.8) | 1.6 | 0.3 |
North Carolina | 30.9 (28.4, 33.3) | 16.5 (15.0, 18.0) | 1.9 | 25.9 (22.8, 29.0) | 14.0 (12.3, 15.7) | 1.9 | 0.0 |
North Dakota | 27.7 (24.8, 30.6) | 17.7 (15.8, 19.6) | 1.6 | 26.9 (24.2, 29.7) | 15.0 (13.5, 16.4) | 1.8 | −0.2 |
Ohio | 38.0 (35.3, 40.7) | 18.6 (17.1, 20.0) | 2.0 | 34.2 (31.6, 36.9) | 16.7 (15.4, 17.9) | 2.0 | 0.0 |
Oklahoma | 38.1 (35.3, 40.8) | 19.6 (18.0, 21.1) | 1.9 | 29.7 (26.8, 32.6) | 16.3 (14.8, 17.9) | 1.8 | 0.1 |
Oregon | 32.0 (29.0, 35.1) | 13.4 (11.9, 15.0) | 2.4 | 26.3 (23.7, 29.2) | 13.0 (11.5, 14.4) | 2.0 | 0.4 |
Pennsylvania | 32.5 (30.2, 34.8) | 17.8 (16.5, 19.1) | 1.8 | 28.9 (25.9, 32.0) | 15.8 (14.3, 17.2) | 1.8 | 0.0 |
Rhode Island | 28.7 (25.7, 31.7) | 15.7 (14.1, 17.4) | 1.8 | 23.1 (19.6, 26.6) | 12.8 (11.1, 14.6) | 1.8 | 0.0 |
South Carolina | 31.8 (29.6, 34.0) | 17.9 (16.4, 19.5) | 1.8 | 29.0 (26.7, 31.3) | 14.7 (13.5, 16.0) | 2.0 | −0.2 |
South Dakota | 31.9 (28.1, 35.8) | 18.7 (16.4, 21.0) | 1.7 | 28.8 (23.9, 33.6) | 16.4 (14.3, 18.6) | 1.8 | −0.1 |
Tennessee | 31.8 (27.4, 36.3) | 17.2 (14.4, 20.0) | 1.8 | 35.7 (32.5, 39.0) | 16.9 (15.1, 18.8) | 2.1 | −0.3 |
Texas | 25.8 (23.5, 28.1) | 15.4 (14.0, 16.8) | 1.7 | 20.0 (17.3, 22.7) | 14.5 (12.7, 16.3) | 1.4 | 0.3 |
Utah | 18.6 (16.9, 20.3) | 7.9 (7.0, 8.8) | 2.4 | 15.1 (13.4, 16.8) | 7.0 (6.2, 7.9) | 2.2 | 0.2 |
Vermont | 28.7 (26.0, 31.3) | 12.5 (11.0, 14.0) | 2.3 | 25.3 (22.5, 28.1) | 11.6 (10.1, 13.2) | 2.2 | 0.1 |
Virginia | 30.6 (27.7, 33.6) | 14.9 (13.1, 16.7) | 2.1 | 25.4 (23.2, 27.7) | 12.5 (11.3, 13.7) | 2.0 | 0.1 |
Washington | 26.6 (24.4, 28.8) | 11.4 (10.2, 12.6) | 2.3 | 21.5 (19.6, 23.3) | 10.7 (9.7, 11.6) | 2.0 | 0.3 |
West Virginia | 38.0 (35.2, 40.8) | 22.8 (20.7, 24.8) | 1.7 | 38.5 (35.5, 41.4) | 19.0 (17.2, 20.9) | 2.0 | −0.3 |
Wisconsin | 34.2 (29.9, 38.5) | 15.5 (13.8, 17.2) | 2.2 | 27.6 (23.9, 31.3) | 13.3 (11.7, 14.8) | 2.1 | 0.1 |
Wyoming | 32.3 (29.4, 35.1) | 16.4 (14.7, 18.1) | 2.0 | 27.8 (24.7, 30.9) | 14.0 (12.3, 15.7) | 2.0 | 0.0 |
Note. The value is the ratio of smoking prevalence in the lower-income group to that in the higher-income group.
Logistic Regression
In the majority of states, the odds of smoking decreased from 2011 to 2017 (Odds ratios [ORs]: 0.95 – 0.98), after controlling for age, sex, race/ethnicity, and education level. However, in 17 states (Alabama, Alaska, Arkansas, California, Florida, Georgia, Idaho, Iowa, Louisiana, Maine, Mississippi, Missouri, New Hampshire, Ohio, Oregon, Tennessee, Vermont), there were no significant linear trends in the odds of smoking. In each state, the odds of smoking were greater in the lower-income group as compared to the higher-income group. Odds ratios ranged from 1.37 in North Dakota to 2.11 in Arkansas (see Appendix Table 2). In each state, older age, being female, and greater education were associated with lower odds of smoking. The relationship between race/ethnicity and smoking, however, was inconsistent across states (results not shown).
There were significant interactions between year and income in four states, indicating time trends in smoking status varied by income group. In only one state (New York), smoking prevalence declined more for lower-income groups compared to higher-income groups. Simple slopes analyses revealed that in New York there was a significant, negative effect for year in the lower-income group (OR, 95% confidence interval [CI]: 0.96 [0.93, 0.98]), but not in the higher-income group (0.99 [0.97, 1.0]). However, in three states, smoking prevalence only dropped among higher-income groups, increasing income-based disparities. In West Virginia (0.96 [0.94, 0.99]), Florida (0.97 [0.95, 0.99]), and Maine (0.97 [0.95, 0.99]), there was a significant, negative effect for year in the higher-income groups, but no significant effect for year in the lower-income groups (West Virginia: 1.00 [0.98, 1.03], Florida: 1.01 [0.99, 1.03], and Maine: 1.0 [0.98, 1.03]).
Discussion
Although eliminating disparities in tobacco use is a priority for tobacco control, all states continue to exhibit income-based disparities in smoking prevalence. Between 2011 and 2017 only one state reduced disparities in smoking between lower- and higher-income groups. In three states income-based disparities in smoking actually widened. In all other states there were no differences in linear time trends between the lower- and higher-income groups, suggesting no change in income-based disparities in smoking. Smoking prevalence among the lower-income groups was often twice as high as smoking prevalence in the higher-income groups. Eliminating income-based disparities in smoking is critical to making progress in tobacco control and reducing the disproportionate burden of smoking-related disease experienced among those with lower-incomes.15
Findings from this study underscore the importance of identifying and implementing policies that reduce income-based disparities in smoking. Research suggests that policies that raise the prices of tobacco through tax and non-tax means have potential to reduce income-based disparities in smoking.9,16 Although the present study is descriptive and did not evaluate the relationships between state-level tobacco control policies and smoking, a discussion of the tobacco policy environment in New York, where disparities reduced, may provide insights into effective interventions for reducing income-based disparities. New York raised its cigarette tax by $1.60 in the middle of 2010, just before the observed trends.17 In addition, New York City, where nearly two thirds of the state’s population live, implemented a $10.50 minimum price for cigarettes in 2014.18,19
Tobacco control programs, if they seek to reduce disparities, should consider reviewing their policies and programs to ensure they reduce, and do not maintain or exacerbate, income-based disparities in smoking. Studies suggest that population-level smoking cessation programs may increase disparities in smoking because of higher cessation rates among higher-income smokers.9 In addition, mass media campaigns are typically less effective among lower-income smokers.9 Hill et al.9 state that smoking cessation programs that target support and increase recruitment among lower-income smokers can help make up for the different cessation rates. Mass media campaigns may need to be tailored to lower-income smokers to exhibit pro-equity effects. For example, studies suggest that campaigns that use personal testimony are more effective among lower-income smokers.9
Ideally, tobacco control programs and policies should either have a pro-equity impact (i.e., have a greater health promoting impact among more disadvantaged populations) or be adapted specifically for lower-income smokers. A tax increase is an example of a pro-equity tobacco control policy.9 Though a tax increase is the same for all groups, the impact is greater among those with lower-incomes. In other cases, tobacco control programs and interventions may need to deliver a greater “dose” to the priority population or be tailored to that priority population. Policies that are not specific to tobacco control may also play an important role in reducing income-based disparities in smoking. For example, Medicaid expansion may have increased access to cessation resources in certain states and have an impact on income-based disparities in smoking. More research examining the impact of tobacco control and other social policies on income-based disparities is needed.
The findings should be interpreted considering study limitations. This study was designed to document trends in income-based disparities within states; further research is required to examine the role of tobacco control, health and income-related policies in contributing to those trends. The present study is not able to make definitive conclusions about the impact of tobacco control policies on income-based disparities in smoking. In addition, this study focuses on cigarette use and does not examine trends in other tobacco product, e-cigarette, or poly-tobacco use. Trends in income-based disparities may be affected by use of other products. Also, the present study categorized respondents into two groups based on their self-reported annual household income but did not take into account the number of individuals in one’s household. The BRFSS only asks participants who completed the landline survey about the number of individuals in their household. A household of two with an annual household income of $50,000, however, may be in a different economic position than a household of five with the same income. A measure of income that incorporates household size may better represent the resources available to support the household. Smoking prevalence was also not examined according to more finely-grained income levels because of sample size limitations in smaller states.20 Classifications of participants into more income groups may provide more detailed information about income-based disparities in smoking. In addition, the present study focused only on income-based disparities in smoking. Prior research has found that educational attainment is a stronger risk factor for smoking than poverty status, an income-based measure.21 Nevertheless, the present study provides important epidemiological findings in response to the CDC’s priority of reducing income-based disparities in smoking. Also, guidelines for social policies (e.g., Medicaid) in the US are typically based on income as opposed to education, so an assessment of income-based disparities in smoking may be more useful for guiding policy.
Furthermore, the present study focuses on reducing income-based disparities in smoking prevalence considering relative, as opposed to absolute, differences in smoking. For example, we would consider a scenario in which smoking prevalence dropped from 10% to 7% among higher-income groups and from 25% to 22% among lower-income groups to be an increase in disparities because the ratio of lower-to-higher smoking prevalence increases from 2.5 to 3.1. Although the absolute change for each group is the same (a drop of three percentage points), a higher proportion of the initial higher-income smoking population benefited than the initial lower-income smoking population. There is no consensus on which measure (absolute or relative) should be used to assess disparities in smoking.22 Therefore it is critical to recognize what type of disparity is being examined and the accurate corresponding interpretation for the measure used.
A system to consistently monitor income-based disparities in smoking and disparities among other demographic groups, at both state and national levels, is needed to continue to examine progress toward reducing disparities. The present study focused on income-based disparities in smoking, but disparities exist according to groups defined by race/ethnicity, education, US census region, sexual orientation, disability, gender, and report of serious psychological distress.1 Although not a focus of the present study, consistent with prior research at the national level,2,23 there were significant relationships between age, sex, education and smoking in each state. Tracking smoking among these other sociodemographic groups at the state-level is a critical step in advocating and planning for effective tobacco control programs.
Conclusion
Lower-income populations have higher rates of smoking as compared to higher-income populations.15 In addition, in most states there were no differences in linear time trends between the lower- and higher-income groups, suggesting no change in income-based disparities in smoking from 2011 to 2017. Findings from this study suggest that little progress has been made toward reducing income-based differences in smoking and additional policy and tobacco control efforts may be required to meet national disparity reduction goals.
Highlights.
In each state, the odds of smoking were 1.4 to 3.0 times greater in the lower-income group as compared to the higher-income group
Between 2011 and 2017 only one state reduced disparities in smoking between lower-and higher-income groups
Pro-equity strategies are needed because current tobacco control efforts are maintaining or widening income-based disparities
Acknowledgments
This projected was funded in part by the Advancing Science and Practice in the Retail Environment (ASPiRE), National Cancer Institute award P01 CA225597. Dr. Mills also received funding from the National Cancer Institute award T32 CA057726. Dr. Ribisl serves as an expert consultant in litigation against tobacco companies. All other authors declare no conflicts of interest. We would like to acknowledge Dr. Chris Wiesen at the Odum Institute at the University of North Carolina, Chapel Hill for support with statistical modeling.
Appendix
Appendix Table 1.
State | 2011 | 2012 | 2013 | 2014 | 2015 | 2016 | 2017 |
---|---|---|---|---|---|---|---|
Alabama | 7689 | 9026 | 6503 | 8652 | 7950 | 7031 | 6754 |
Alaska | 3543 | 4345 | 4578 | 4388 | 3657 | 2914 | 3203 |
Arizona | 6489 | 7306 | 4252 | 14867 | 7946 | 10952 | 15499 |
Arkansas | 4739 | 5187 | 5268 | 5258 | 5256 | 5298 | 5322 |
California | 18004 | 14574 | 11518 | 8832 | 12601 | 11393 | 9358 |
Colorado | 13612 | 12255 | 13649 | 13399 | 13537 | 14958 | 9802 |
Connecticut | 6829 | 8781 | 7710 | 7950 | 11899 | 11041 | 10588 |
Delaware | 4777 | 5174 | 5206 | 4300 | 4070 | 4057 | 4139 |
District of Columbia | 4560 | 3827 | 4931 | 4074 | 3994 | 3852 | 3868 |
Florida | 12399 | 7624 | 34186 | 9821 | 9739 | 36955 | 22059 |
Georgia | 9960 | 6100 | 8138 | 6351 | 4678 | 5381 | 6056 |
Hawaii | 7606 | 7582 | 7858 | 7247 | 7163 | 8087 | 7754 |
Idaho | 6077 | 5896 | 5630 | 5487 | 5802 | 5258 | 4894 |
Illinois | 5475 | 5579 | 5608 | 5052 | 5289 | 4764 | 5545 |
Indiana | 8495 | 8645 | 10338 | 11530 | 6067 | 11066 | 13829 |
Iowa | 7354 | 7166 | 8157 | 8130 | 6227 | 7257 | 7699 |
Kansas | 20712 | 11801 | 23282 | 13743 | 23236 | 12188 | 21843 |
Kentucky | 10894 | 11223 | 11013 | 11197 | 8806 | 10265 | 8642 |
Louisiana | 10926 | 9068 | 5251 | 6781 | 4716 | 5256 | 4809 |
Maine | 13243 | 9921 | 8097 | 9137 | 9063 | 10019 | 9692 |
Maryland | 10117 | 12812 | 13011 | 12569 | 12598 | 18473 | 13588 |
Massachusetts | 22328 | 21723 | 15071 | 15654 | 9294 | 8415 | 6912 |
Michigan | 11049 | 10499 | 12759 | 8466 | 8935 | 12024 | 10889 |
Minnesota | 15401 | 12246 | 14340 | 16419 | 16761 | 16831 | 17095 |
Mississippi | 8907 | 7788 | 7453 | 4205 | 6035 | 5135 | 5076 |
Missouri | 6405 | 6754 | 7118 | 7081 | 7307 | 7126 | 7601 |
Montana | 10265 | 8679 | 9693 | 7502 | 6051 | 5971 | 5915 |
Nebraska | 25416 | 19173 | 17139 | 22420 | 17561 | 15183 | 15350 |
Nevada | 5493 | 4846 | 5101 | 3763 | 2926 | 4344 | 3764 |
New Hampshire | 6362 | 7530 | 6463 | 6192 | 7022 | 6420 | 5751 |
New Jersey | 15383 | 15761 | 13386 | 13045 | 11465 | 7652 | 11719 |
New Mexico | 9417 | 8776 | 9316 | 8937 | 6734 | 6024 | 6538 |
New York | 7735 | 6060 | 8979 | 6865 | 12357 | 34190 | 12249 |
North Carolina | 11550 | 11898 | 8860 | 7289 | 6698 | 6536 | 4916 |
North Dakota | 5306 | 4879 | 7806 | 7786 | 4972 | 5742 | 6992 |
Ohio | 9948 | 13026 | 11971 | 10933 | 11929 | 12389 | 12289 |
Oklahoma | 8523 | 8015 | 8244 | 8448 | 6943 | 6925 | 6638 |
Oregon | 6247 | 5302 | 5949 | 5227 | 5359 | 5439 | 5370 |
Pennsylvania | 11509 | 19958 | 11429 | 11000 | 5740 | 6810 | 6542 |
Rhode Island | 6533 | 5480 | 6531 | 6450 | 6206 | 5457 | 5632 |
South Carolina | 12948 | 12795 | 10717 | 11027 | 11607 | 11236 | 11311 |
South Dakota | 8259 | 7878 | 6895 | 7401 | 7221 | 5767 | 7012 |
Tennessee | 5914 | 7056 | 5815 | 5142 | 5979 | 6167 | 5843 |
Texas | 14973 | 9129 | 10917 | 15436 | 14697 | 11709 | 12255 |
Utah | 12669 | 12436 | 12769 | 15006 | 11401 | 10988 | 10251 |
Vermont | 7096 | 6056 | 6392 | 6475 | 6489 | 6540 | 6516 |
Virginia | 6605 | 7398 | 8464 | 9472 | 8646 | 9002 | 9630 |
Washington | 14772 | 15319 | 11162 | 10092 | 16116 | 14259 | 13279 |
West Virginia | 5282 | 5409 | 5899 | 6199 | 5957 | 7151 | 5472 |
Wisconsin | 5302 | 5299 | 6589 | 7045 | 6188 | 5271 | 5810 |
Wyoming | 6870 | 6273 | 6454 | 6416 | 5492 | 4497 | 4463 |
Note. Sample size is unweighted.
Appendix Table 2.
State | Odds Ratio (95% Confidence Interval) | |
---|---|---|
Income | Year | |
Alabama | 1.8 (1.6, 1.9) | 0.99 (0.97, 1.00) |
Alaska | 1.8 (1.6, 2.0) | 0.99 (0.96, 1.02) |
Arizona | 1.7 (1.6, 1.9) | 0.97 (0.95, 0.99) |
Arkansas | 2.1 (1.9, 2.4) | 0.98 (0.96, 1.01) |
California | 1.6 (1.5, 1.7) | 1.00 (0.98, 1.01) |
Colorado | 1.8 (1.7, 1.9) | 0.98 (0.97, 1.00) |
Connecticut | 1.7 (1.6, 1.9) | 0.96 (0.95, 0.98) |
Delaware | 1.6 (1.4, 1.7) | 0.97 (0.95, 0.99) |
District of Columbia | 1.8 (1.6, 2.1) | 0.95 (0.93, 0.97) |
Florida | 1.7 (1.6, 1.8) | 0.99 (0.97, 1.00) |
Georgia | 1.7 (1.6, 1.9) | 0.98 (0.97, 1.00) |
Hawaii | 1.5 (1.4, 1.7) | 0.98 (0.96, 1.00) |
Idaho | 2.0 (1.8, 2.2) | 0.98 (0.96, 1.01) |
Illinois | 1.7 (1.5, 1.8) | 0.96 (0.94, 0.98) |
Indiana | 1.8 (1.7, 2.0) | 0.98 (0.97, 1.00) |
Iowa | 1.9 (1.8, 2.0) | 0.99 (0.97, 1.00) |
Kansas | 1.8 (1.7, 1.9) | 0.97 (0.96, 0.98) |
Kentucky | 1.8 (1.7, 2.0) | 0.98 (0.96, 0.99) |
Louisiana | 1.6 (1.5, 1.8) | 0.99 (0.97, 1.01) |
Maine | 2.0 (1.8, 2.1) | 0.99 (0.97, 1.00) |
Maryland | 1.6 (1.5, 1.7) | 0.96 (0.95, 0.98) |
Massachusetts | 1.8 (1.6, 1.9) | 0.97 (0.95, 0.98) |
Michigan | 1.8 (1.7, 1.9) | 0.98 (0.96, 0.99) |
Minnesota | 1.8 (1.7, 1.9) | 0.96 (0.95, 0.97) |
Mississippi | 1.8 (1.6, 1.9) | 0.99 (0.97, 1.01) |
Missouri | 1.9 (1.7, 2.0) | 0.99 (0.97, 1.00) |
Montana | 2.0 (1.8, 2.1) | 0.98 (0.96, 1.00) |
Nebraska | 1.7 (1.6, 1.8) | 0.97 (0.95, 0.98) |
Nevada | 1.8 (1.6, 2.0) | 0.97 (0.94, 0.99) |
New Hampshire | 2.0 (1.8, 2.2) | 1.00 (0.98, 1.02) |
New Jersey | 1.4 (1.3, 1.6) | 0.97 (0.95, 0.99) |
New Mexico | 1.7 (1.6, 1.9) | 0.97 (0.95, 0.99) |
New York | 1.6 (1.5, 1.7) | 0.98 (0.96, 0.99) |
North Carolina | 1.8 (1.7, 1.9) | 0.97 (0.96, 0.99) |
North Dakota | 1.4 (1.3, 1.5) | 0.98 (0.96, 0.99) |
Ohio | 2.0 (1.8, 2.1) | 0.99 (0.97, 1.00) |
Oklahoma | 1.8 (1.7, 2.0) | 0.96 (0.95, 0.98) |
Oregon | 2.0 (1.9, 2.2) | 0.99 (0.97, 1.01) |
Pennsylvania | 1.8 (1.7, 1.9) | 0.97 (0.96, 0.99) |
Rhode Island | 1.8 (1.7, 2.0) | 0.96 (0.94, 0.98) |
South Carolina | 1.8 (1.7, 1.9) | 0.98 (0.97, 0.99) |
South Dakota | 1.6 (1.5, 1.8) | 0.97 (0.95, 0.99) |
Tennessee | 2.0 (1.8, 2.2) | 1.00 (0.99, 1.03) |
Texas | 1.5 (1.4, 1.7) | 0.97 (0.95, 0.99) |
Utah | 1.8 (1.6, 1.9) | 0.97 (0.96, 0.99) |
Vermont | 2.0 (1.8, 2.2) | 1.00 (0.98, 1.02) |
Virginia | 1.7 (1.6, 1.8) | 0.96 (0.95, 0.98) |
Washington | 1.9 (1.8, 2.1) | 0.97 (0.96, 0.99) |
West Virginia | 1.9 (1.8, 2.0) | 0.98 (0.97, 1.00) |
Wisconsin | 1.9 (1.7, 2.1) | 0.97 (0.95, 0.99) |
Wyoming | 1.8 (1.6, 2.0) | 0.97 (0.95, 0.99) |
Notes. Logistic regression models were adjusted for age, sex, education, and race/ethnicity. The income term was coded where 1 = lower-income group and 0 = higher-income group. Models were run separately for each state.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- 1.Wang TW, Asman K, Gentzke AS, et al. Tobacco Product Use Among Adults - United States, 2017. MMWR Morbidity and mortality weekly report. 2018;67(44):1225–1232. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Drope J, Liber AC, Cahn Z, et al. Who’s still smoking? Disparities in adult cigarette smoking prevalence in the United States. CA: A Cancer Journal for Clinicians. 2018;68(2):106–115. [DOI] [PubMed] [Google Scholar]
- 3.Agaku IT, Odani S, Okuyemi KS, Armour B. Disparities in current cigarette smoking among US adults, 2002–2016. Tobacco Control. 2019. [DOI] [PubMed]
- 4.Centers for Disease Control and Prevention. Best Practices User Guide: Health Equity in Tobacco Prevention and Control. Atlanta: 2015. [Google Scholar]
- 5.U.S. Department of Health and Human Services. Reducing Tobacco Use: A Report of The Surgeon General. Atlanta, Georgia: 2000. [Google Scholar]
- 6.U.S. Department of Health and Human Services. The Health Consequences of Smoking-50 Years of Progress: A Report of the Surgeon General. Atlanta: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health; 2014. [Google Scholar]
- 7.Thomas S, Fayter D, Misso K, et al. Population tobacco control interventions and their effects on social inequalities in smoking: systematic review. Tobacco Control. 2008;17(4):230. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Farrelly MC, Bray JW, Pechacek T, Woollery T. Response by Adults to Increases in Cigarette Prices by Sociodemographic Characteristics. Southern Economic Journal. 2001;68(1):156–165. [Google Scholar]
- 9.Hill S, Amos A, Clifford D, Platt S. Impact of tobacco control interventions on socioeconomic inequalities in smoking: review of the evidence. Tobacco Control. 2014;23(e2):e89. [DOI] [PubMed] [Google Scholar]
- 10.American Lung Association. State of Tobacco Control 2017.
- 11.Centers for Disease Control and Prevention. 2017 Summary Data Quality Report. 2018.
- 12.Li J, Sussell AL. Multiple Imputation of Family Income Data in the 2015 Behavioral Risk Factor Surveillance System. 2017; https://www.mwsug.org/proceedings/2017/AA/MWSUG-2017-AA06.pdf Accessed November 7, 2019.
- 13.van Buuren S. Multiple imputation of discrete and continuous data by fully conditional specification. Statistical Methods in Medical Research. 2007;16(3):219–242. [DOI] [PubMed] [Google Scholar]
- 14.Haymes S, de Haymes MV, Miller R. The Routledge Handbook of Poverty in the United States. London, United Kingdom: Routledge; 2015. [Google Scholar]
- 15.Henley SJ, Thomas CC, Sharapova SR, et al. Vital Signs: Disparities in Tobacco-related cancer incidence and mortality. Morbidity and Mortality Weekly Report. 2016;65(44):1212–1218. [DOI] [PubMed] [Google Scholar]
- 16.Golden SD, Farrelly MC, Luke DA, Ribisl KM. Comparing projected impacts of cigarette floor price and excise tax policies on socioeconomic disparities in smoking. Tobacco Control. 2016;25(Suppl 1):i60. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Campaign for Tobacco-free Kids. Cigarette Tax Increases by State Per Year 2000–2019. 2019; https://www.tobaccofreekids.org/assets/factsheets/0275.pdf Accessed November 6, 2019.
- 18.DeLong H, Chiriqui JF, Leider J, Chaloupka FJ. Tobacco Product Pricing Laws: A State-by-State Analysis, 2015. Chicago: 2016. [Google Scholar]
- 19.NYC Health: New Law: Minimum price for cigarettes is now $13. https://www1.nyc.gov/assets/doh/downloads/pdf/smoke/cig-min-price.pdf Accessed May 27, 2019. [Google Scholar]
- 20.Centers for Disease Control and Prevention. The BRFSS Data User Guide. 2013; https://www.cdc.gov/brfss/data_documentation/pdf/UserguideJune2013.pdf Accessed November 7, 2019.
- 21.Higgins ST, Kurti AN, Redner R, et al. Co-occurring risk factors for current cigarette smoking in a U.S. nationally representative sample. Prev Med. 2016;92:110–117. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Houweling TAJ, Kunst AE, Huisman M, Mackenbach JP. Using relative and absolute measures for monitoring health inequalities: experiences from cross-national analyses on maternal and child health. International Journal for Equity in Health. 2007;6(1):15. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Jamal A, Phillips E, Gentzke AS, et al. Current Cigarette Smoking Among Adults - United States, 2016. MMWR Morbidity and mortality weekly report. 2018;67(2):53–59. [DOI] [PMC free article] [PubMed] [Google Scholar]