Trends in Cigarette Smoking Prevalence and Status among U.S. Adults with Disabilities, 2015–2019

Abstract

Introduction:

The prevalence of current cigarette smoking is higher in people with disabilities compared to those without. However, little is known about smoking status and trends in smoking by disability functioning domain.

Methods:

Data from the 2015–2019 National Surveys on Drug Use and Health were used to estimate the prevalence of past-month and daily cigarette smoking, former smoking, and nicotine dependence for people with any disability and six disability functioning domains. Logistic regression models estimated the odds of each outcome by disability domain compared to no disability, adjusting for sociodemographic factors.

Results:

From 2015–2019, the overall prevalence of current cigarette smoking (23.3% vs. 16.7%) and the proportion of those with nicotine dependence (14.6% vs 8.0%) was higher in people with any disability compared to those without (ps < 0.001). The prevalence of current cigarette smoking decreased while the prevalence of former cigarette smoking increased from 2015–2019 (ps < 0.05). People with any disability had higher odds of current smoking (AOR=1.20; 95% CI 1.16, 1.25) and similar odds of former smoking (AOR=1.00; 95% CI 0.95, 1.06) compared to people without disabilities. Odds of current and former smoking varied by domain.

Conclusion:

The prevalence of cigarette smoking among people with any disability is decreasing over time but remains higher than those without. People with any disability have similar odds of former smoking and differences exist by disability domain. Future research should explore additional smoking and quit behaviors by disability functioning domain.

1. Introduction

People with disabilities (PWDs)—approximately 25% of the population—¹have a higher prevalence of tobacco use compared to people without disabilities;^2,3 however, little is known about trends in cigarette smoking and cigarette quit behavior of PWDs. Population cigarette smoking has decreased in recent years,^2,4 but the reduction in current cigarette smoking has been slower for PWDs compared to those without a disability.⁵ According to the social model of disability, disability relates to the interaction between a person’s health condition and personal and environment factors,⁶ and PWDs encounter more environmental barriers related to social determinates of health (e.g., less access to health resources, discrimination),^1,7 factors that may influence smoking and cessation. There remains a paucity of information on cessation behavior in PWDs and available research reports mixed findings. For example, one study suggests that PWDs make fewer quit attempts than those without a disability⁸ while other studies indicate either that PWDs do not differ significantly on planning and making a quit attempt^9,10 or are more likely to make a quit attempt compared to those without a disability.¹¹

Additionally, little is known about factors that may affect quitting like nicotine dependence in PWDs.¹² Although some research has indicated PWDs are more nicotine dependent than people without disabilities as they are more likely to smoke within 5 minutes of waking and smoke more cigarettes per day,⁹ there is little information on nicotine dependence by disability functioning domain, which could provide insight into both smoking prevalence and cessation behaviors in PWDs.

Quitting smoking may be particularly important for PWDs, as this population has higher background rates of chronic disease such as diabetes, asthma, cancer, and stroke, which are also causally associated with smoking,¹³ and lower access to health resources that may provide cessation treatment.¹ Limited research suggests that PWDs report improvements in mental health, general health, and energy and vitality after quitting cigarettes,¹⁴ underscoring the benefits of cessation in this population. Various functioning domains may also experience different barriers or have distinct needs, requiring adaptation to existing programs.

The purpose of this study was to provide information on recent trends in current cigarette smoking prevalence and status in adults with disabilities by functioning domain.

2. Methods

Data were from the National Survey on Drug Use and Health (NSDUH), which is an annual survey that provides health information on people 12 years or older in the United States. Datasets were limited to the years 2015–2019 as 2015 was the first year NSDUH included the American Community Survey six-item set of questions to measure disability¹⁵ and NSDUH cautions comparing 2020 to previous years due to methodological changes caused by the COVID-19 pandemic.¹⁶ The survey questions the civilian, noninstitutionalized population, using in-person data collection and computer-assisted interviewing techniques in English or Spanish language. Analyses included respondents who were 18 or older (n = 214,505). The University of Vermont Institutional Review Board determined the study not to be human subjects research.

2.1. Measures

2.1.1. Disability

Participants were categorized by disability and functioning domain based on their response to the American Community Survey six-item set of questions to measure disability:¹⁵ 1) “Are you deaf or do you have serious difficulty hearing?” (hearing); 2) “Are you blind or do you have serious difficulty seeing, even when wearing glasses?” (vision); 3) “Because of a physical, mental or emotional condition, do you have serious difficulty concentrating, remember, or making decisions?” (cognition); 4) “Do you have serious difficulty walking or climbing stairs?” (mobility); 5) “Do you have difficulty dressing or bathing?” (self-care); and 6) “Because of a physical, mental or emotional condition, do you have difficulty doing errands alone such as visiting a doctors’ office or shopping?” (independent living). Respondents were considered to have any disability if they responded “yes” to any of the questions.

2.1.2. Smoking status

Lifetime smoking was defined as those who had smoked at least 100 cigarettes in their life. Past-month cigarette smoking (current) were respondents who reported lifetime smoking and also indicated smoking at least one cigarette in the previous 30 days. Past-month daily smoking was defined as those who reported lifetime smoking and indicated smoking cigarettes daily in the last month. Nicotine dependence was assessed from a derived variable using the Fagerström Test of Nicotine Dependence.¹⁷ Former smoking was defined for respondents who reported lifetime smoking but had not smoked cigarettes in the previous year. For current smoking, the reference group was noncurrent smoking (never and former); for daily smoking, the reference group was nondaily smoking (never, former, people who did not smoke daily in the past month); for former smoking, the reference group was those who smoked at least 100 cigarettes in their lifetime and had smoked in the past year.

2.1.3. Covariates

Sociodemographic variables included as covariates were the following: sex (male, female), age (18–25 years, 26–34 years, 35–49 years, 50+ years), sexual orientation (bisexual, heterosexual, lesbian or gay), race/ethnicity (Hispanic, Non-Hispanic (NH) White, NH Black, Other), education (less than high school, high school diploma/GED, some college/associate degree, Bachelor’s degree or more), annual household income (<$20,000, $20,000–$49,000, $50,000–$74,999, ≥$75,000), health insurance (yes/no), and past month any illicit drug use (yes/no). Past year serious psychological distress (yes/no) was included in sensitivity analyses.

2.2. Data Analysis

The prevalence of current cigarette smoking, daily cigarette smoking, former smoking, and nicotine dependence by disability type were estimated for combined 2015–2019 data. The prevalence of past-month cigarette smoking and former smoking were estimated for each year by disability type. Linear time trends in current cigarette smoking and former smoking were evaluated using logistic regression models with disability status as the exposure of interest; an interaction between disability status and year was explored to identify differences in prevalence and cessation trends by disability status over time. For any disability and the six disability functioning domains, adjusted logistic regression models for current smoking, daily smoking, former smoking, and nicotine dependence included disability status, year, and all covariates with the reference group being people without any or the respective disability type for all analyses. Sensitivity analyses added serious psychological distress to the adjusted logistic regression models consistent with prior studies² and also compared 2015–2019 data with 2015–2020 data.

Analyses were completed using Stata Version 16 and R, which accounted for NSDUH sampling weights and complex clustered sampling to produce accurate standard errors using Taylor series estimation. Missing data were removed using list-wise deletion (range = 0%–1.9%). Data presented followed NSDUH suppression guidelines.¹⁶

3. Results

The prevalence of adults aged 18+ with any disability was 20.2% and PWD were more likely to be female, complete less education, have lower income, have health insurance, have past-year serious psychological distress, and have any past-month illicit drug use than people without disability (Table S1).

Pooled data from 2015–2019 by disability and functioning domain indicated people with any disability had higher prevalence of past-month cigarette smoking (23.3% vs. 16.7%), daily smoking (15.8% vs. 10.3%), and nicotine dependence (14.6% vs. 8.0%) than those without a disability (all ps < 0.001; Table S2). This pattern was consistent across all disability functioning domains, except those with a hearing disability who had similar past-month (17.6%) smoking prevalence as those with no disability, but a higher prevalence of daily smoking (12.0%) and nicotine dependence (11.1%). The prevalence of current cigarette smoking decreased overall from 2015–2019 (Figure 1; p < 0.001); while there was no interaction between disability status and year, smoking prevalence remained higher in people with any disability and for each disability type, except hearing, than those without a disability over this time period.

Figure 1.

Weighted prevalence of adults indicating current cigarette smoking by disability and disability functioning domain by year. Data are from the 2015–2019 National Survey on Drug Use and Health.

The odds of current, former, and daily smoking and nicotine dependence varied across any disability and disability type (Figure 2). After adjusting for covariates, people with any disability had higher odds of current smoking (AOR=1.20; 95% CI 1.16, 1.25), daily smoking (AOR=1.20; 95% CI 1.15, 1.26), and nicotine dependence (AOR=1.37; 95% CI 1.30, 1.34) compared to people with no disability. Former smoking (AOR=1.00; 95% CI 0.95, 1.06) was similar in people with and without disabilities. People with a hearing disability had lower odds of currently smoking (AOR=0.85; 95% CI 0.78, 0.93) and daily smoking (AOR=0.85; 95% CI 0.77, 0.93) than those without a hearing disability; all other functioning domains had a higher odds of current and daily smoking compared to those who did not have that respective disability (all ps < 0.001). People with a cognitive (AOR=0.79; 95% CI 0.71, 0.87), independent living (AOR=0.84; 95% CI 0.77, 0.93), self-care (AOR=0.81; 95% CI 0.72, 0.92), and visual (AOR=0.78; 95% CI 0.70, 0.87) disability had lower odds of former smoking than those who reported not having the respective disability. However, people who reported a hearing disability had higher odds of former smoking (AOR=1.39; 95% CI 1.27, 1.51) than those without a hearing disability. The prevalence of former cigarette smoking increased overall from 2015–2019 (Figure S1; p < 0.05). Although there was no interaction between disability status and year, the prevalence of former smoking was lower in people with cognitive, independent living, self-care, and vision; higher for hearing and mobility disabilities; and similar for people with any disability than those without a disability over this time period. Sensitivity analyses including 2020 data (Tables S3–S4 and Figures S2–S4) and past-year psychological distress (Figures S5–S6) indicated overall similar results. People with any disability were more likely to formerly smoke after adjusting for past-year psychological distress (Figure S5–S6).

Figure 2.

Adjusted odds ratios of current, former, and daily smoking and nicotine dependence by disability and functioning domain. Data are combined from the 2015–2019 National Survey on Drug Use and Health.

Note. Odds ratio with 95% confidence intervals adjusting for sex, age group, sexual orientation, race/ethnicity, education, income, health insurance, past-month any illicit drug use, and year. Along the y-axis is disability and type. The x-axis uses a log scale and represents the adjusted odds ratio. Estimates are weighted. Boldface indicates statistical significance (p<0.05) with people without each respective disability serving as the reference group.

4. Discussion

Using a large national study with repeated cross-sectional data, our findings support that despite population-level decreases in the prevalence of past-month cigarette smoking, smoking prevalence remained higher for people with any disability compared with no disability from 2015–2019. Higher prevalence of daily smoking and nicotine dependence were consistent across people with any disability and all disability types compared to people without a disability. The prevalence of former smoking has increased overall across study years; however, odds of former smoking varied by disability functioning domain.

These results extend a limited evidence base on cigarette smoking behavior among PWDs. Similar to previous findings, this research suggests people with any disability have a higher smoking prevalence than those without a disability.^2,3 Results of this study highlight that higher past-month smoking prevalence is seen across all disability types except for hearing. Previous findings indicate people with hearing difficulties have a higher prevalence of smoking cigarettes³ or are either less likely to smoke or no more likely to smoke, depending on if they are prelingually or postlingually deaf, respectively.¹⁸ This study supports earlier research that found people with any and various types of disabilities are more nicotine dependent.⁹ Finally, findings related to former smoking extend the current literature that generally focuses on either disability or no disability and indicates mixed results by providing information on disability type.^8–11 The current study highlights differences in disability type that other studies may have obscured by grouping all disabilities together, indicating that people with any, cognitive, independent living, self-care, and vision disabilities have lower odds of formerly smoking. It is important to note the lack of a standard way to measure disability may impact comparisons, as some previous studies use different definitions of disability or phrase survey questions on disability differently.

The results of this study indicate higher cigarette smoking prevalence but varying odds of former smoking in PWD. This finding, integrated with previous research that suggests PWDs may be more likely to be given the advice to stop smoking¹⁰ and to use cessation treatment^10,11 but do not differ significantly on sustained long-term cessation than those without a disability,¹¹ may indicate PWDs are either quitting and then returning to smoke or are initiating smoking at high rates. Some disability domains may also be less likely to quit altogether. Further research should continue to examine cigarette smoking over time, incidence, return to use, and duration of quitting to improve the long-term success of cessation interventions and better tailor prevention methods to reduce initiation of smoking in people with disabilities. Data should also continue to be disaggregated by disability type to identify for whom prevention and cessation programs should be tailored.

This study has a number of strengths. First, this study investigates smoking status among multiple disability types using one national dataset. Next, this study provides information on trends in cigarette smoking by disability type, information which to this point has remained unknown. However, this study has several limitations. First, data are self-reported and were cross-sectional, therefore respondents were not followed across time. Second, only cigarette smoking was evaluated in this study. Given higher rates of various tobacco products in PWDs (e.g., e-cigarettes),^3,19 future research should explore quit behavior of other tobacco products. Additionally, due to survey data collection limitations, people who had formerly smoked were defined as people who did not smoke in the last year and physical health could not be included as a covariate. Next, although the survey is administered with in-person and computer-assisted methods, it is unknown whether this is sufficient to make the survey accessible to a diverse sample of people with disabilities. Finally, 2020 data were not included due to methodological changes.¹⁶ However, analyses including 2020 show similar results. Future research should estimate additional factors that may affect quitting smoking (e.g., menthol use, quantity of cigarettes smoked) and quit related behaviors (e.g., quit methods, duration of quit) by disability type. Understanding quit behaviors of PWDs can help appropriately tailor cessation interventions and build on existing cessation interventions adapted for PWDs²⁰ to ultimately reduce tobacco related disparities.

Highlights.

• Past-month cigarette smoking has decreased for people with any disability.

• However, smoking prevalence remained higher for people with any disability.

• Daily smoking and nicotine dependence was higher for all types of disabilities.

• Odds of former smoking varied by disability functioning domain.