Abstract
Objectives:
In this paper, we explore the adherence patterns to US Centers for Disease Control and Prevention (CDC) COVID-19 mitigation guidelines among current, former, and never smokers.
Methods:
We used an online cross-sectional survey of adults 18 years or older in 5 northeastern states of the US (N=1084).
Results:
Unadjusted analyses revealed that current smokers reported lower adherence to the CDC guidelines than former smokers (27.5 vs 29.4, p<.05). After accounting for sociodemographic covariates, this finding was no longer statistically significant. However, compared to former smokers, never smokers reported wearing their mask less often (OR=0.65; 95% CI=0.45–0.94) and current smokers were less likely to report always practicing illness-related hygiene skills (OR=0.60; 95% CI=0.39–0.93).
Conclusions:
Never smokers had poorer adherence to CDC guidelines than former smokers, namely wearing their masks, and current smokers were less likely to always follow the hygiene recommendations. Results should inform future public health efforts in targeting current smokers with lower adherence to CDC guidelines and learning from the ability of former smokers to demonstrate high adherence.
Keywords: COVID-19, smoking, health behavior, survey, pandemic
Just over 2 years ago, the first cases of the coronavirus disease 2019 (COVID-19) were detected.1 The development and widespread distribution of vaccines to prevent infection with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) have been effective at lowering the risk of getting and spreading the virus, as well as reducing the chances of serious illness and death in those who do contract SARS-CoV-2 or one of its variants.2 Nevertheless, COVID-19 has had a catastrophic impact on global health and is estimated to have killed over 5.5 million people worldwide, including 800,000 Americans, as of January 2022.3 As the world continues to combat COVID-19 and adapt to the ever-evolving nature of this pandemic, one element has remained consistent throughout – the importance of societal and individual behavioral measures as a primary approach to disease containment.
In March 2020, the US Centers for Disease Control and Prevention (CDC) began to publish mitigation guidelines to serve as frontline defense against SARS-CoV-2. Handwashing, face masks, and social distancing were 3 behavioral strategies that demonstrated effectiveness in slowing the spread of the virus.4,5 Implementation of these anti-contagion measures likely prevented 4.8 million additional confirmed COVID-19 cases and 60 million total infections within the first month of the pandemic.6 In addition to heightened individual risk, nonadherence to basic COVID-19 guidelines is met with serious public health implications – failing to social distance or wear a mask increases the likelihood of spreading the disease to others, and puts those most vulnerable at especially high risk.7 These measures play a vital role in the containment of infectious outbreaks and pandemics including, but not limited to, COVID-19, and gaining insight on adherence to these behaviors is critical to inform future public health messaging efforts.
Current tobacco users may be a particularly important population in which to explore COVID-19 guideline adherence due to users’ tendency to engage in other risky health behaviors. Previous epidemiological findings demonstrated that, compared to current non-smokers, current smokers use their seatbelt less frequently,8 participate in riskier sexual behaviors,9 and wear sunscreen less regularly.10 When further analyzing smoking status, important behavioral differences arise. Data from the 2006 Behavioral Risk Factor Surveillance System revealed that current smokers were less likely than either never smokers or former smokers to receive flu or pneumonia vaccines.11 Moreover, former smokers were more likely to receive the influenza vaccine and use routine clinical preventive services such as cervical or prostate cancer screenings than both current and never smokers.11,12 Reasons as to why former smokers have a tendency to demonstrate increased adherence to certain health behaviors might be explained in part by theories of health behavior, such as the Health Belief Model13 and Theory of Planned Behavior.14 Those who quit smoking may become more health conscious in general – a psychological state that might translate into other positive behaviors such as seeking vaccines and using preventive services. These unique differences among current, former, and never smokers make it important to study these 3 groups separately in the context of COVID-19 guideline adherence.
Understanding the relationships between smoking status and CDC virus mitigation guideline adherence is particularly important in light of the association between tobacco use and respiratory and bacterial infection, and poor respiratory infection outcomes.15,16 The influence of smoking on COVID-19 is not fully understood. A living systematic review of the available data up to August 2020 showed that whereas being a current smoker had no association with COVID-19 hospitalization or mortality, former smokers were at an increased risk of hospitalization, disease severity, and mortality compared to never smokers.17 More recent reviews that have employed genetic prediction techniques demonstrated that there is a strong relationship between lifetime tobacco use and COVID-19 infection risk and illness severity.18,19 Research investigating changes in smoking behaviors throughout the pandemic has noted increases, decreases, and no notable changes in tobacco use.20,21 Although the variability in data suggests that there are considerable differences in smoking behaviors, tobacco use remains a primary public health issue, and exploring adherence patterns to CDC virus mitigation guidelines among its users warrants significant attention.
In China, limited evidence found that non-smokers were 1.39 times more likely to practice social-distancing; however, smoking status was only determined by whether the individual had smoked in the past month.22 Jackson et al.23 conducted a study on the associations between COVID-19 and smoking in a population of 53,000 adults in the United Kingdom and reported that current smokers had lower odds of practicing general adherence to COVID-19 protective behaviors compared to never smokers.23 The objective of our study was to investigate whether there is a difference in adherence to a defined and detailed set of COVID-19 mitigation guidelines among current, former, and never smokers. Our study focused on 5 northeastern US states (CT, MA, NJ, NY, RI) that reported the highest rates of COVID-19 cases and deaths in the US per capita at the time.3 We hypothesized that current smokers’ adherence to these guidelines would be lower than that of former smokers. We also hypothesized that former smokers would report adherence to the guidelines that was greater than or equal to never smokers, and never smokers would report adherence levels that were higher than current smokers (Former smokers≥Never smokers>Current smokers).
METHODS
Study Design
This study was an online, cross-sectional panel survey. We recruited participants from 5 northeastern states (CT, MA, NY, NJ, RI) using Amazon’s crowdsourcing platform, Mechanical Turk (MTurk). MTurk is a survey management system that consistently has been shown to be as efficient and reliable at generating samples as other comparable panel data providers in both the broader academic setting,24,25 and within the health and medical research field.26 The final version of the survey was released from June 18-July 19, 2020, a period in which these states were following reopening plans that shared similar restrictions on business capacity and social gatherings.
Study Participants
Eligibility criteria included being≥18 years of age; residing in an eligible state (CT, MA, NY, NJ, or RI); and holding an active MTurk account. In addition, we employed enrollment quotas based on age, gender, race, and ethnicity to ensure a diverse sample. Participants meeting enrollment criteria were filtered into Qualtrics to complete a brief screening to determine if they matched a quota. If the participants filled a quota, they would be directed to review the informed consent document. Participants were compensated $10 for completing the one-hour survey. More details on the study design and enrollment quotas are available elsewhere.27
Measures
CDC guideline adherence.
We assessed adherence to the CDC virus mitigation guidelines using a 13-item self-report measure. We asked participants to rate how often they engaged in 13 recommended behaviors over the past 4 weeks on a scale from 0 (“rarely or never”) to 3 (“always”). Adherence was defined as the sum of item scores (range: 0–39), with higher scores indicative of higher adherence (Table A1).
We further investigated 3 individual guidelines as a secondary outcome. These ‘target’ CDC guidelines, including mask wearing, social distancing, and handwashing, were chosen as the most important and effective behaviors to mitigate COVID-19 transmission.28 Each target CDC guideline had a 0–3 range.
We explored a measure of full adherence as an additional secondary outcome. As Reinders Folmer et al recommends,29 we operationalized full adherence as responding with “always” to all 13 items in the CDC guideline (0=not fully adherent; 1=fully adherent).
Smoking status.
We assessed participants’ tobacco use history and exposure using an adapted measure from the National Health Interview Survey.30 We categorized smoking status into current, former, or never smoker. Never smokers answered “no” to having smoked at least 100 cigarettes in their lifetime. A current smoker was someone who reported smoking at least 100 cigarettes in their lifetime and was currently smoking at the time of the survey. Former smokers were defined as individuals who had smoked at least 100 cigarettes in their lifetime but had not smoked in the last 6 months; a former smoker was also any participant who, although having smoked in the past 6 months, had not smoked in the past 30 days.
Sociodemographic and COVID-19 testing history.
We selected covariates based on their demonstrated relevance and association to COVID-19 mitigation and other health-related behaviors in previous research.31–34 Demographic characteristics included age, sex, gender identity, race, ethnicity, highest level of education, number of people living in the household, annual household income, and home ownership (an indicator of wealth).35 We also asked participants about their essential worker status, operationalized through a question asking if they identified as “someone whose work is critical to business operations and/or meeting basic human needs and is required to attend work during the COVID pandemic” (yes/no/not sure). We measured COVID-19 testing history through a series of questions that asked whether participants had been tested for COVID-19, either through a nose swab test or serology. Answering “yes” to either of these COVID-19 tests prompted a question asking for the result (positive/negative/not sure).
Data Analysis
Data cleaning and analyses were performed using SPSS Version 25 (IBM Corp., Armonk, NY) and Stata 16.0 (StataCorp LLC, College Station, TX). Descriptive analyses assessed demographic and health characteristics overall and by smoking status using one-way ANOVA tests for continuous data – reported as mean (SD) – and chi-square tests for categorical variables – reported as count (percentage). We used one-way ANOVAs to examine unadjusted differences in total CDC adherence scores and conducted post hoc comparisons with Holm-Bonferroni corrections. For the 3 target CDC guidelines, we performed an unadjusted ordinal logistic regression for mask wearing across smoking status (met the proportional odds assumption) and an unadjusted multinomial logistic regression for handwashing and social distancing (did not meet the proportional odds assumption).
We then fit a linear regression model to evaluate differences in total CDC adherence score by smoking status, adjusted for covariates. Table 1 shows the covariates tested for inclusion. We conducted bivariate analyses to test the associations of these covariates with the total CDC adherence score. Significant covariates included in the regression were as follows: gender (reference [ref]=cisgender female), ethnicity (ref=not Hispanic or Latino), essential worker status (ref=not essential worker or unsure), income (eight categories, treated as continuous), and COVID-19 test (ref=no test or unsure). In all regression analyses, we tested for multicollinearity and ensured that the variance inflation factor value for each independent variable was less than 2.5.
TABLE 1.
Demographic and Descriptive Statistics of Study Participants, n=1084
| Characteristic | All (n=1084) | Current (n=220) | Former (n=195) | Never (n=669) | p-value |
|---|---|---|---|---|---|
| Mean Age (SD) | 40.9 (13.4) | 41.8 (12.2) | 48.3 (14.1) | 38.4 (12.9) | < 0.001 |
| Gender a , n (%) | 0.8 | ||||
| Cisgender Female | 521 (48.1) | 103 (46.8) | 96 (49.2) | 322 (48.1) | |
| Cisgender Male | 526 (48.5) | 107 (48.6) | 94 (48.2) | 325 (48.6) | |
| Other Gender Identity, or Prefer not to answer | 37 (3.4) | 10 (4.6) | 5 (2.6) | 22 (3.3) | |
| Ethnicity, n (%) | 0.004 | ||||
| Hispanic or Latino | 276 (25.5) | 75 (34.1) | 44 (22.6) | 157 (23.5) | |
| Not Hispanic or Latino | 808 (74.5) | 145 (65.9) | 151 (77.4) | 512 (76.5) | |
| Race, n (%) | < 0.001 | ||||
| White | 710 (65.5) | 148 (67.3) | 148 (75.9) | 414 (61.9) | |
| Black | 239 (22.1) | 51 (23.2) | 36 (18.5) | 152 (22.7) | |
| Asian | 77 (7.1) | 8 (3.6) | 5 (2.6) | 64 (9.6) | |
| Other, or more than one racial identity | 58 (5.4) | 13 (5.9) | 6 (3.1) | 39 (5.8) | |
| Educational Level, n (%) | 0.005 | ||||
| High school or less | 107 (9.9) | 22 (10.0) | 31 (15.9) | 54 (8.1) | |
| Some college/1-year degree | 215 (19.8) | 52 (23.6) | 40 (20.5) | 123 (18.4) | |
| College graduate/4-year degree | 547 (50.5) | 115 (52.3) | 87 (44.6) | 345 (51.6) | |
| Graduate or Professional school | 215 (19.8) | 31 (14.1) | 37 (18.9) | 147 (21.9) | |
| Mean Household Income in 2020 US$, (SD) | < 0.001 | ||||
| Less than US $49,999 | 397 (36.6) | 100 (45.5) | 68 (34.9) | 229 (34.2) | |
| US $50,000–$99,999 | 475 (43.8) | 101 (45.9) | 88 (45.1) | 286 (42.8) | |
| US $100,000+ | 212 (19.6) | 19 (8.6) | 39 (20.0) | 154 (23.0) | |
| Essential Worker, n (%) | < 0.001 | ||||
| Yes | 388 (35.8) | 115 (52.3) | 64 (32.8) | 209 (31.2) | |
| No/Not Sure | 696 (64.2) | 105 (47.7) | 131 (67.2) | 460 (68.8) | |
| Dwelling Ownership, n (%) | 0.05 | ||||
| No ownership of home or condo | 577 (53.2) | 128 (58.1) | 90 (46.2) | 359 (53.7) | |
| Own a home or condo | 507 (46.8) | 92 (41.8) | 105 (53.9) | 310 (46.3) | |
| Mean Number of People Living in Residence, (SD) | 3.1 (1.3) | 3.4 (1.3) | 2.9 (1.4) | 3.0 (1.3) | 0.0002 |
| Tested for COVID-19 and subsequent result n(%) | < 0.001 | ||||
| Not Tested or Unsure | 805 (74.3) | 132 (60.0) | 149 (76.4) | 524 (78.3) | |
| Yes, Positive Result | 44 (4.1) | 28 (12.7) | 2 (1.0) | 14 (2.1) | |
| Yes, Negative Result | 235 (21.7) | 60 (27.3) | 44 (22.6) | 131 (19.6) | |
Note.
The variable was created by combining and recoding the sex and gender variables.
We also fit an ordinal logistic regression model to assess adjusted differences in adherence to the target CDC guideline of mask wearing by smoking status. We fit 2 multinomial logistic regression models to assess differences in adherence to the target CDC guideline of social distancing and handwashing by smoking status. The significant covariates included in this model were the same as in the previous model, ie, gender, ethnicity, essential worker status, income, and COVID-19 test.
We conducted an additional exploratory analysis to evaluate the extent to which individuals fully adhered to guidelines. The pattern of this dichotomized variable (0=not completely adherent; 1=fully adherent) as it related to smoking status was first analyzed with chi-square tests to determine which of the 13 guidelines had significant differences in proportions of full adherence across these 3 groups. The p-values in the chi-square bivariate analysis were adjusted for multiple comparisons with the Holm-Bonferroni correction. We then fit a binary logistic regression model for each of the guidelines showing significant group differences controlling for the same covariates that were included in the other models.
RESULTS
Sample Characteristics
We assessed 3489 individuals for eligibility. Of these, 1185 participants completed the survey (Figure 1). Prior to analysis, 30 (2.5%) respondents were removed because they incorrectly answered at least 2 of the validity questions or they demonstrated an invalid response pattern to the categorical demographic covariates. Seven participants (0.6%) had missing data on focal study variables and were excluded from further analyses. Duplicate data for 29 MTurk IDs (2.4%) were identified and removed from the dataset. An additional 35 (2.9%) participants had missing covariate data. Thus, the analytic sample for this study with complete information about smoking status, adherence score, and covariate data consisted of 1084 participants (Figure 1). Table 1 presents the characteristics of the 1084 participants.
Figure 1.
Details of the Evaluable Sample Selection Process
There were 220 current smokers (20.3%), 195 former smokers (18.0%), and 669 never smokers (61.7%). The mean age of the sample was 40.9 and half of the sample identified as male. The sample was 22.3% black, and 25.6% Hispanic or Latino. The majority of the sample had at least a college degree. Over half of current smokers reported being an essential worker during the COVID-19 pandemic, compared to only one-third of former and never smokers. Approximately one-fourth of participants (25.9%) responded that they had a history of receiving a COVID-19 test. Of those who had been tested, 15.8% received a positive result (44/279). About 40% of current smokers had been tested – compared to former smokers (23.6%) and never smokers (21.7%) – and they had a higher positivity rate (31.5%).
Total CDC Guideline Adherence
The unadjusted results of CDC guideline adherence compared across smoking groups produced from the one-way ANOVA showed that current smokers had the lowest adherence score, followed by never smokers and former smokers with the highest score (mean=27.5 vs 28.6 vs 29.4, p=.048) (not shown in tables). Current smokers had lower overall adherence to CDC recommendations compared to former smokers (p=.037). There was no statistical difference between current and never smokers, or between former and never smokers.
Results from the regression model showed that the difference in CDC adherence score by smoking status was not significant after accounting for covariates. Covariates accounted for 7.1% of the variance in CDC guideline adherence (R2=0.071; Adjusted [Adj] R2=0.064). The final model with covariates and smoking status (Table 2) accounted for 7.2% of the variance in CDC guideline adherence (R2=0.072; Adjusted [Adj] R2=0.064). Inclusion of the tobacco use variable did not significantly change the variance accounted for in the model. Identifying as Hispanic/Latino, an essential worker, of lower income, and having a positive COVID-19 test were all significantly associated with a lower total CDC adherence score.
TABLE 2.
Multiple Linear Regression Between Smoking Status and Total CDC Guideline Adherence, Accounting for Covariates, n = 1084
| Variable | β | 95% CI |
|---|---|---|
| Intercept | 26.00 | (23.07, 28.95)*** |
| Smoking Status | ||
| Former Smokers (Reference) | ||
| Current Smokers | −0.85 | (−2.40, 0.71) |
| Never Smokers | −0.85 | (−2.14, 0.44) |
| Age | −0.002 | (−0.04, 0.03) |
| Gender | ||
| Cisgender Female (Reference) | ||
| Cisgender Male | 2.15 | (1.20, 3.10)*** |
| Other Gender Identity, or Prefer not to answer | −2.71 | (−5.32, −0.09)* |
| Hispanic or Latino (Reference: No) | −1.30 | (−2.42, −0.17)* |
| Essential Worker (Reference: No) | −1.79 | (−2.82, −0.76)*** |
| Income | 0.49 | (0.22, 0.75)*** |
| COVID Test | ||
| No Test or Unsure (Reference) | ||
| Negative Test | 0.43 | (−0.75, 1.61) |
| Positive Test | −2.96 | (−5.46, −0.46)* |
Note.
p < .05
p < .01
p< .001
Targeted CDC Guidelines Adherence
The unadjusted results of adherence to target CDC guidelines across smoking groups produced from the ordinal logistic regression showed that there were statistically significant differences in adherence to the mask wearing guideline: current smokers were less likely to report wearing their mask compared to former smokers (OR=0.53, 95% CI=0.35–0.79, p=.002) (not shown in tables). Moreover, never smokers also were less likely to adhere to the mask wearing guideline compared to former smokers (OR=0.66, 95% CI=0.47–0.93, p=.016) (not shown in tables). The unadjusted results of adherence to the target CDC guidelines across smoking groups from the multinomial logistic regression showed that current smokers were 2.16 times more likely to social distance “sometimes” than former smokers (95% CI=1.17–3.99) and 1.91 times more likely to adhering to handwashing “sometimes” than former smokers (95% CI=1.07–3.40) (not shown in tables).
The adjusted results from the ordinal logistic regression model showed that the group differences among the mask wearing guideline remained significant for former and never smokers (OR=0.65, 95% CI=0.45–0.94), p=.021) (Table 3). After accounting for covariates, there were no statistically significant differences in mask wearing among former and current smokers, nor were there statistically significant differences in adherence to social distancing or handwashing across any of the smoking groups.
TABLE 3.
Ordinal Logistic Regression Between Smoking Status and Mask-Wearing, Accounting for Covariates, n = 1084
| Variable | OR | 95% CI |
|---|---|---|
| Smoking Status | ||
| Former Smokers (Reference) | ||
| Current Smokers | 0.69 | (0.45, 1.05) |
| Never Smokers | 0.65 | (0.45, 0.94)* |
| Age | 1.00 | (.99, 1.01) |
| Gender | ||
| Cisgender Female (Reference) | ||
| Cisgender Male | 2.13 | (1.65, 2.76)*** |
| Other Gender Identity, or Prefer not to answer | 0.81 | (0.43, 1.54) |
| Hispanic or Latino (Ref: No) | 0.57 | (0.43, 0.75)*** |
| Essential Worker (Ref: No) | 0.64 | (0.49, 0.83)** |
| Income | 1.1 | (1.03, 1.19)** |
| COVID Test | ||
| No Test or Unsure (Reference) | ||
| Negative Test | 0.86 | (0.63, 1.17) |
| Positive Test | 0.63 | (0.34, 1.15) |
Note.
p < .05
p < .01
p< .001
CDC Guideline Full Adherence
Across the 13 individual guidelines, the percentage of participants who reported full adherence ranged from 30.9% (guideline 12: using detergent prior to disinfecting) to 78.9% (guideline 9: disposing used tissues). There were statistically significant differences in full adherence to guidelines by smoking status for 3 of the 13 guidelines (guidelines 4, 8, and 9), where current smokers were less likely to fully adhere compared to former smokers (Table A2). The results from the logistic regression models for the 3 significant guidelines in the bivariate analysis showed that there were statistically significant differences in full guideline adherence between current and former smokers and never and former smokers after accounting for covariates (Table 4). Compared to former smokers, current smokers were less likely to always cover their mouth and nose when coughing/sneezing (OR=0.60, 95% CI=0.39–0.93) and less likely to always throw out their used tissues (OR=0.61, 95% CI=0.37–0.98). Additionally, never smokers were 0.65 times as likely to report always avoiding close contact with sick individuals and to report always covering their face when coughing/sneezing compared to former smokers (OR=0.65, 95% CI=0.43–0.97 and OR=0.65, 95% CI=0.45–0.94).
TABLE 4.
Binary Logistic Regression Model of Association Between Smoking Status and Full Adherence, Accounting for Covariates, n=1084a
| Smoking (reference = Former Smoker): | Current Smoker | Never Smoker | ||
|---|---|---|---|---|
| Full Adherence Item | OR | 95% CI | OR | 95% CI |
| 4. Avoid close contact with people who are sick? | 0.65 | 0.41 – 1.04 | 0.65 | 0.43– 0.97* |
| 8. Cover your mouth and nose with a tissue or use the inside of your elbow when you coughed or sneezed? | 0.60 | 0.39 – 0.93* | 0.65 | 0.45 – 0.94* |
| 9. Throw used tissues in the trash? | 0.61 | 0.37 – 0.98* | 0.83 | 0.54 – 1.26 |
Note.
Results adjusted by covariates: gender, ethnicity, essential worker status, COVID test, and income
p<.05
DISCUSSION
Our study is the first to investigate differences in adherence patterns to the CDC COVID-19 mitigation guidelines across current, former, and never smokers. All participants were residents of 5 northeastern states that had experienced the highest rates of COVID-19 morbidity and mortality at the time of the study. We tested the hypothesis that current smokers would report the lowest adherence to the CDC guidelines, and that former smokers would report adherence that was commensurate with, if not greater than, that of never smokers. Our unadjusted results partially supported this hypothesis, as current smokers had lower adherence to total CDC guidelines scores compared to former smokers. However, this difference was no longer significant after accounting for sociodemographic covariates. Additionally, there was no significant difference in total guideline adherence among former and never smokers, nor between never and current smokers. An exploration of the target CDC guidelines – mask wearing, handwashing, and social distancing – revealed that former smokers had significantly higher adherence to the mask wearing recommendation than never smokers. When the 13 guidelines were assessed on the basis of full adherence, current smokers were significantly less likely to adhere completely to 2 individual guidelines compared to former smokers: practicing proper hygiene and disposing of used tissues. Moreover, never smokers were also significantly less likely to always report avoiding close contact with sick individuals and always practice proper hygiene compared to former smokers.
Our findings showed that there was no difference in total adherence scores across smoking categories after adjusting for key sociodemographic covariates, suggesting that the antecedents of these health behaviors may be different. It is important to contextualize these findings in light of situational and sociodemographic factors. For certain guidelines included in this survey, high adherence would not be possible for those with physical limitations, such as living in a populated household or continuing to go to work in-person. Separately, Pedersen et al.36 found that essential workers reported lower adherence to guidelines about social distancing and isolating in a large survey of US adults. Given over half of current smokers in our sample identified as an essential worker, compared to one-third of former and never smokers, it is plausible that the nature of current smokers’ jobs limits their ability to adhere to the CDC’s COVID-19 mitigation guidelines.
Our results also showed that current smokers were more likely to be Hispanic or Latinx, live in a more populated household, and have a lower total household income compared to former and never smokers. Residing with a greater number of people would make it difficult to report ‘always’ avoiding close contact with people who are sick. Income effects may also drive some of our findings, as those from lower socioeconomic backgrounds are less able to work remotely, refrain from public transportation, or live in environments that facilitate social distancing.37,38 Social distancing and avoiding infected individuals are privileges. Importantly, the lack of significant difference in adherence scores between smoking groups after accounting for sociodemographic covariates highlights the possibility that there are key common socioeconomic drivers of both smoking and adherence. Moreover, the COVID-19 pandemic has exposed long-standing structural inequities of health and access to care that exist within the US. These health disparities likely contribute to the observed differences in health behaviors among smokers and nonsmokers and – in the face of a pandemic – likely lead to detrimental effects that are multiplicative rather than additive.39
In addition to total adherence to all 13 guidelines, we examined adherence patterns to the 3 arguably most important and effective COVID-19 mitigation behaviors prior to the development of the vaccines: handwashing, social distancing, and mask wearing.40 Whereas there were no marked differences in adherence to the handwashing and social distancing measures among smoking status groups, we did find that never smokers were less likely to adhere to mask wearing than former smokers. This result is supported by theories of health behavior13,14 and consistent with other literature that has showed an increase in health-promoting behaviors among former smokers.11,12
A growing number of studies have focused on understanding the patterns of adherence to COVID-19 mitigation guidelines and identifying factors related to strong and weak adherence. However, there is variability in how researchers have defined this outcome, with some who chose to dichotomize the variable into “adherent” and “not-adherent,”29,41 and others who parsed out a more nuanced scale.42,43 It is difficult to capture the complexities that underlie behavioral adherence and selecting one operationalization of this variable is likely to overshadow certain pieces of the overall picture. In this paper, we examined adherence from both a continuous and categorical standpoint. Over the early course of this pandemic, in which vaccines or treatment were not yet available, it was crucial for individuals to completely adhere to the guidelines, as they are most effective when practiced consistently. When participants selected a response other than “always,” they were admitting to being non-adherent at some point in time. With this stricter measurement, we found that current smokers were less likely than former smokers to always practice proper hygiene, such as covering the face when coughing or sneezing. Current smokers have greater cognitive dissonance regarding the risks of smoking compared to former smokers,44,45 and the latter report a more accurate understanding of this risk.46 Covering coughs and sneezes is a hygiene practice that helps reduce the spread of illness.47 Unrealistic optimism regarding health may drive current smokers (relative to former smokers) to minimize COVID-19 risk perceptions and thereby reduce their practice of such health behaviors. Alternatively, although current smokers tend to have unrealistic optimism about their health risk compared to non-smokers,48 perhaps they are more honest with their responses to other health-related behaviors and thus less likely to report “always” adhering to certain guidelines. Our finding that never smokers were less likely to always avoid contact with sick individuals compared to former smokers further underscores the latter group’s potential heightened health-consciousness and motivation to maintain distance from infected people.
Limitations
These findings should be interpreted in the context of key limitations. Generalizability is restricted by the use of a convenience sample (ie, Amazon’s Mturk). In 2019, the CDC calculated that 14.0% of US adults currently smoked cigarettes,49 an observed rate that is 6 percentage points lower than the prevalence calculated in our sample. This divergence from national data should be kept in mind when extrapolating from our results. Analysis for this paper did not account for quantity of cigarettes smoked and did not further differentiate current smokers into daily and non-daily users. We created an assessment that was not yet validated to evaluate adherence to CDC guidelines. It is possible that with the use of self-report data, participants did not accurately disclose their adherence patterns, either because they were unable to recall them or they did not feel able to share their true behaviors. We did not explore potential reasons for weak and strong adherence specifically among the smoking population, and future studies are needed to clarify this relationship by examining health beliefs, risk perception, and other possible explanatory factors.
Moreover, although the results for full adherence draw new distinctions according to smoking status, they must be interpreted with caution. The nature of this dichotomized variable likely taps into a response bias, in which some participants answered “always” to all of the questions due to demand characteristics and social desirability, and/or because they were not attentive to their responses or wanted to maintain internal consistency. Our analyses were not able to differentiate influence from these potential sources of bias.
Implications and Future Directions
This study sheds light on the gap in adherence levels to certain guidelines among current and former smokers and never and former smokers. Our results inform future public health interventions and underscore the importance of a multifaceted approach to raising the adherence of the current smokers. Targeting the smoking population and improving the content and outreach of health-related behavior messaging might be beneficial techniques for achieving control of COVID-19. Furthermore, it is critical that we also consider the large-scale upstream changes that need to occur to support those who face practical constraints, which make practicing preventive behaviors such as social distancing and avoiding infected individuals challenging.
It is important to consider the significance of our data regarding former smokers and their comparatively elevated adherence levels. Questions about what differentiates a former smoker from a current smoker and a never smoker, and how these motivational and situational factors might apply to adherence to COVID-19 guidelines, remain open to answering. A certain mentality, as well as other structural support systems, must be present for someone to quit an addictive substance and resist relapse. We need to improve understanding of former smokers and their seemingly unique ability both to quit smoking and demonstrate higher adherence to certain CDC guidelines. Future studies should investigate whether the qualities that predict higher adherence to these guidelines are similar, overlap, or complement the characteristics that allow formers smokers to successfully quit cigarettes. Findings might then be applied to design interventions that raise the adherence of other subpopulations.
Conclusion
We found that current smokers in 5 northeastern states reported poorer overall adherence to CDC guidelines compared to former smokers. However, these effects were non-significant when accounting for sociodemographic covariates, suggesting that there might be common sociodemographic factors that drive both smoking and poor adherence. When exploring adherence to the 3 most crucial guidelines, we found that never smokers reported wearing face masks significantly less than former smokers. We also found that current smokers were less likely to report always adhering to guidelines that recommend practicing illness-related hygiene skills compared to former smokers. There may be practical and psychosocial barriers at play that make practicing these behaviors challenging. Future interventions to improve adherence must combine efforts that both target the current smoking population by modifying public health messaging and also implement effective structural measures that address these situational constraints, such as the promotion of workplace safety among employers of essential workers.
Supplementary Material
Acknowledgements
This project was supported by an Institutional Development Award from the National Institute of General Medical Sciences of the National Institutes of Health under grant P20GM130414 (Monti). Preparation of the manuscript also was supported by National Institutes of Health grants K23AA024704 (Monnig) and K08DA048137 (Sokolovsky).
Footnotes
Human Subjects Approval Statement
The study was reviewed by Brown University Institutional Review Board (IRB) and determined to be exempt from requiring IRB approval as minimal risk study per federal regulations.
Conflict of Interest Disclosure Statement
None of the authors reports a conflict of interest.
Contributor Information
Claire L. Szapary, Center for Addiction and Disease Risk Exacerbation, Brown University, Providence, RI, United States..
Jaqueline Contrera Avila, Center for Addiction and Disease Risk Exacerbation, Brown University, Providence, RI, United States..
Mollie A. Monnig, Center for Addiction and Disease Risk Exacerbation, Brown University, Providence, RI, and Department of Behavioral and Social Sciences, Brown University, Providence, RI, United States..
Alexander W. Sokolovsky, Center for Addiction and Disease Risk Exacerbation, Brown University, Providence, RI, and Department of Behavioral and Social Sciences, Brown University, Providence, RI, United States..
Grace DeCost, Center for Addiction and Disease Risk Exacerbation, Brown University, Providence, RI, United States..
Jasjit S. Ahluwalia, Center for Addiction and Disease Risk Exacerbation, Brown University, Providence, RI, and Department of Behavioral and Social Sciences, Brown University, Providence, RI, United States..
References
- 1.World Health Organization. Listings of WHO’s response to COVID-19. World Health Organization. https://www.who.int/news/item/29-06-2020-covidtimeline. Published June 29, 2020. Accessed August 6, 2022. [Google Scholar]
- 2.Centers for Disease Control and Prevention. COVID-19 Vaccines are Effective. National Center for Immunization and Respiratory Diseases (NCIRD), Division of Viral Diseases. https://www.cdc.gov/coronavirus/2019-ncov/vaccines/effectiveness/index.html. Published June 29, 2022. Accessed August 6, 2022. [Google Scholar]
- 3.Dong E, Du H, Gardner L. An interactive web-based dashboard to track COVID-19 in real time. Lancet Infect Dis. 2020;20(5):533–34. 10.1016/s1473-3099(20)30120-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Alzyood M, Jackson D, Aveyard H, et al. COVID-19 reinforces the importance of handwashing. J Clin Nurs. 2020;29(15–16):2760–61. 10.1111/jocn.15313 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Howard J, Huang A, Li Z, et al. An evidence review of face masks against COVID-19. Proc Natl Acad Sci U S A. 2021;118(4) 10.1073/pnas.2014564118 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Hsiang S, Allen D, Annan-Phan S, et al. The effect of large-scale anti-contagion policies on the COVID-19 pandemic. Nature. 2020;584(7820):262–67. 10.1038/s41586-020-2404-8 [DOI] [PubMed] [Google Scholar]
- 7.Courtemanche C, Garuccio J, Le A, et al. Strong Social Distancing Measures In The United States Reduced The COVID-19 Growth Rate. Health Aff (Millwood). 2020;39(7):1237–46. 10.1377/hlthaff.2020.00608 [DOI] [PubMed] [Google Scholar]
- 8.Grout P, Cliff KS, Harman ML, et al. Cigarette smoking, road traffic accidents and seat belt usage. Public Health. 1983;97(2):95–101. 10.1016/s0033-3506(83)80005-5 [DOI] [PubMed] [Google Scholar]
- 9.Zabin LS. The association between smoking and sexual behavior among teens in US contraceptive clinics. Am J Public Health. 1984;74(3):261–3. 10.2105/ajph.74.3.261 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Holman DM, Berkowitz Z, Guy GP Jr., et al. Patterns of sunscreen use on the face and other exposed skin among US adults. J Am Acad Dermatol. 2015;73(1):83–92.e1. 10.1016/j.jaad.2015.02.1112 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Pearson WS, Dube SR, Ford ES, et al. Influenza and pneumococcal vaccination rates among smokers: data from the 2006 Behavioral Risk Factor Surveillance System. Prev Med. 2009;48(2):180–3. 10.1016/j.ypmed.2008.11.001 [DOI] [PubMed] [Google Scholar]
- 12.Vander Weg MW, Howren MB, Cai X. Use of routine clinical preventive services among daily smokers, non-daily smokers, former smokers, and never-smokers. Nicotine Tob Res. 2012;14(2):123–30. 10.1093/ntr/ntr141 [DOI] [PubMed] [Google Scholar]
- 13.Janz NK, Becker MH. The Health Belief Model: a decade later. Health Educ Q. 1984;11(1):1–47. 10.1177/109019818401100101 [DOI] [PubMed] [Google Scholar]
- 14.Norman P, Conner M, Bell R. The theory of planned behavior and smoking cessation. Health Psychol. 1999;18(1):89–94. 10.1037//0278-6133.18.1.89 [DOI] [PubMed] [Google Scholar]
- 15.Abadom TR, Smith AD, Tempia S, et al. Risk factors associated with hospitalisation for influenza-associated severe acute respiratory illness in South Africa: A case-population study. Vaccine. 2016;34(46):5649–55. 10.1016/j.vaccine.2016.09.011 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Feldman C, Anderson R. Cigarette smoking and mechanisms of susceptibility to infections of the respiratory tract and other organ systems. J Infect. 2013;67(3):169–84. 10.1016/j.jinf.2013.05.004 [DOI] [PubMed] [Google Scholar]
- 17.Simons D, Shahab L, Brown J, et al. The association of smoking status with SARS-CoV-2 infection, hospitalization and mortality from COVID-19: a living rapid evidence review with Bayesian meta-analyses (version 7). Addiction. 2021;116(6):1319–68. 10.1111/add.15276 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Rosoff DB, Yoo J, Lohoff FW. Smoking is significantly associated with increased risk of COVID-19 and other respiratory infections. Commun Biol. 2021;4(1):1230. 10.1038/s42003-021-02685-y [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Clift AK, von Ende A, Tan PS, et al. Smoking and COVID-19 outcomes: an observational and Mendelian randomisation study using the UK Biobank cohort. Thorax. 2022;77(1):65–73. 10.1136/thoraxjnl-2021-217080 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Sarich P, Cabasag CJ, Liebermann E, et al. Tobacco smoking changes during the first pre-vaccination phases of the COVID-19 pandemic: A systematic review and meta-analysis. EClinicalMedicine. 2022;47:101375. 10.1016/j.eclinm.2022.101375 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Yang H, Ma J. How the COVID-19 pandemic impacts tobacco addiction: Changes in smoking behavior and associations with well-being. Addict Behav. 2021;119:106917. 10.1016/j.addbeh.2021.106917 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Xu H, Gan Y, Zheng D, et al. Relationship Between COVID-19 Infection and Risk Perception, Knowledge, Attitude, and Four Nonpharmaceutical Interventions During the Late Period of the COVID-19 Epidemic in China: Online Cross-Sectional Survey of 8158 Adults. J Med Internet Res. 2020;22(11):e21372. 10.2196/21372 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Jackson SE, Brown J, Shahab L, et al. COVID-19, smoking and inequalities: a study of 53 002 adults in the UK. Tob Control. 2021;30(e2):e111–e21. 10.1136/tobaccocontrol-2020-055933 [DOI] [PubMed] [Google Scholar]
- 24.Buhrmester M, Kwang T, Gosling SD. Amazon’s Mechanical Turk: A New Source of Inexpensive, Yet High-Quality, Data? Perspect Psychol Sci. 2011;6(1):3–5. 10.1177/1745691610393980 [DOI] [PubMed] [Google Scholar]
- 25.Chandler J, Shapiro D. Conducting Clinical Research Using Crowdsourced Convenience Samples. Annu Rev Clin Psychol. 2016;12:53–81. 10.1146/annurev-clinpsy-021815-093623 [DOI] [PubMed] [Google Scholar]
- 26.Mortensen K, Hughes TL. Comparing Amazon’s Mechanical Turk Platform to Conventional Data Collection Methods in the Health and Medical Research Literature. J Gen Intern Med. 2018;33(4):533–38. 10.1007/s11606-017-4246-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Monnig MA, Treloar Padovano H, Sokolovsky AW, et al. Association of Substance Use With Behavioral Adherence to Centers for Disease Control and Prevention Guidelines for COVID-19 Mitigation: Cross-sectional Web-Based Survey. JMIR Public Health Surveill. 2021;7(11):e29319. 10.2196/29319 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Manikandan N Are social distancing, hand washing and wearing masks appropriate measures to mitigate transmission of COVID-19? Vacunas. 2020;21(2):136–37. 10.1016/j.vacun.2020.09.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Reinders Folmer CP, Brownlee MA, Fine AD, et al. Social distancing in America: Understanding long-term adherence to COVID-19 mitigation recommendations. PLoS One. 2021;16(9):e0257945. 10.1371/journal.pone.0257945 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.The National Center for Health Statistics. National Health Interview Survey (NHIS). U.S. Department of Health & Human Services. https://www.cdc.gov/nchs/nhis/index.htm. Published August 2, 2022. Accessed August 6, 2022. [Google Scholar]
- 31.Bekalu MA, McCloud RF, Viswanath K. Association of Social Media Use With Social Well-Being, Positive Mental Health, and Self-Rated Health: Disentangling Routine Use From Emotional Connection to Use. Health Educ Behav. 2019;46(2_suppl):69–80. 10.1177/1090198119863768 [DOI] [PubMed] [Google Scholar]
- 32.Czeisler M, Lane RI, Petrosky E, et al. Mental Health, Substance Use, and Suicidal Ideation During the COVID-19 Pandemic-United States, June 24–30, 2020. MMWR Morb Mortal Wkly Rep. 2020;69(32):1049–57. 10.15585/mmwr.mm6932a1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Fridman I, Lucas N, Henke D, et al. Association Between Public Knowledge About COVID-19, Trust in Information Sources, and Adherence to Social Distancing: Cross-Sectional Survey. JMIR Public Health Surveill. 2020;6(3):e22060. 10.2196/22060 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Park CL, Russell BS, Fendrich M, et al. Americans’ COVID-19 Stress, Coping, and Adherence to CDC Guidelines. J Gen Intern Med. 2020;35(8):2296–303. 10.1007/s11606-020-05898-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Turner TM, Luea H. Homeownership, wealth accumulation and income status. Journal of Housing Economics. 2009;18(2):104–14. 10.1016/j.jhe.2009.04.005 [DOI] [Google Scholar]
- 36.Pedersen MJ, Favero N. Social Distancing during the COVID-19 Pandemic: Who Are the Present and Future Noncompliers? Public Adm Rev. 2020;80(5):805–14. 10.1111/puar.13240 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Bonacini L, Gallo G, Scicchitano S. Working from home and income inequality: risks of a ‘new normal’ with COVID-19. J Popul Econ. 2021;34(1):303–60. 10.1007/s00148-020-00800-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Garnier R, Benetka JR, Kraemer J, et al. Socioeconomic Disparities in Social Distancing During the COVID-19 Pandemic in the United States: Observational Study. J Med Internet Res. 2021;23(1):e24591. 10.2196/24591 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Bhaskar S, Rastogi A, Menon KV, et al. Call for Action to Address Equity and Justice Divide During COVID-19. Front Psychiatry. 2020;11:559905. 10.3389/fpsyt.2020.559905 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Doung-Ngern P, Suphanchaimat R, Panjangampatthana A, et al. Case-Control Study of Use of Personal Protective Measures and Risk for SARS-CoV 2 Infection, Thailand. Emerg Infect Dis. 2020;26(11):2607–16. 10.3201/eid2611.203003 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Coroiu A, Moran C, Campbell T, et al. Barriers and facilitators of adherence to social distancing recommendations during COVID-19 among a large international sample of adults. PLoS One. 2020;15(10):e0239795. 10.1371/journal.pone.0239795 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Fancourt D, Bu F, Mak HW, et al. Covid-19 Social Study: Results Release 2. UCL COVID-19 Social Study 2020. https://www.covidsocialstudy.org/results [Google Scholar]
- 43.Williams SN, Armitage CJ, Tampe T, et al. Public perceptions of non-adherence to pandemic protection measures by self and others: A study of COVID-19 in the United Kingdom. PLoS One. 2021;16(10):e0258781. 10.1371/journal.pone.0258781 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Fotuhi O, Fong GT, Zanna MP, et al. Patterns of cognitive dissonance-reducing beliefs among smokers: a longitudinal analysis from the International Tobacco Control (ITC) Four Country Survey. Tob Control. 2013;22(1):52–8. 10.1136/tobaccocontrol-2011-050139 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Halpern MT. Effect of smoking characteristics on cognitive dissonance in current and former smokers. Addict Behav. 1994;19(2):209–17. 10.1016/0306-4603(94)90044-2 [DOI] [PubMed] [Google Scholar]
- 46.Finney Rutten LJ, Augustson EM, Moser RP, et al. Smoking knowledge and behavior in the United States: sociodemographic, smoking status, and geographic patterns. Nicotine Tob Res. 2008;10(10):1559–70. 10.1080/14622200802325873 [DOI] [PubMed] [Google Scholar]
- 47.Centers for Disease Control and Prevention. Water, Sanitation, and Environmentally Related Hygiene (WASH): Hygiene Etiquette and Practice. Centers for Disease Control and Prevention. https://www.cdc.gov/healthywater/hygiene/etiquette/coughing_sneezing.html. Published June 15, 2022. Accessed August 6, 2022. [Google Scholar]
- 48.Weinstein ND, Marcus SE, Moser RP. Smokers’ unrealistic optimism about their risk. Tob Control. 2005;14(1):55–9. 10.1136/tc.2004.008375 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49.Centers for Disease Control and Prevention. Current Cigarette Smoking Among Adults in the United States. Office on Smoking and Health, National Center for Chronic Disease Prevention and Health Promotion. https://www.cdc.gov/tobacco/data_statistics/fact_sheets/adult_data/cig_smoking/index.htm. Published March 17, 2022. Accessed August 6, 2022. [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.