Public Support for Cigarette Pack Pictorial Health Warnings Among US Adults: A Cross-sectional Analysis of the 2020 Health Information National Trends Survey

Abstract

Introduction

The US Food and Drug Administration issued a final rule requiring new warnings for cigarette packages and advertisements. This study examines population-level characteristics of support for—versus neutrality or opposition toward—cigarette pack warnings that use text and images to portray the negative health effects of smoking.

Methods

We used nationally representative cross-sectional data of US adults age 18 and older from the 2020 Health Information National Trends Survey (n = 3865). Frequencies and weighted proportions were calculated for neutrality toward, opposition to, and support for pictorial warnings across sociodemographics and other predictors. Weighted, multivariable logistic regression examined predictors of being neutral or opposed versus supportive of pictorial warnings.

Results

In 2020, an estimated 69.9% of US adults supported pictorial warnings, 9.1% opposed, and 20.9% neither supported nor opposed them. In fully adjusted models, current smokers had almost twice the odds of being neutral or opposed to pictorial warnings as never smokers (odds ratio [OR] = 1.99, confidence interval [CI] 1.12, 3.52). Adults 75 years and older (vs. 18–34) (OR = 0.55, CI 0.33, 0.94) and those with children under 18 in their household (vs. no children) (OR = 0.67, CI 0.46, 0.98) were less likely to be neutral or opposed.

Conclusions

In advance of the Food and Drug Administration’s implementation of pictorial warnings on cigarette packages, nearly 70% of American adults support this policy. Disseminating information about the effectiveness of pictorial warnings may further strengthen support among current smokers who are less supportive than never smokers. Furthermore, framing messages around the benefits of pictorial warnings for protecting youth may increase public support.

Implications

While public support for pictorial warnings on cigarette packages is high in the United States, it may increase further after policy implementation and be strengthened by utilizing information campaigns that convey the evidence that pictorial warnings are an effective public health strategy.

Introduction

The 2009 Family Smoking Prevention and Tobacco Control Act (TCA) directed the Food and Drug Administration (FDA) to issue regulations requiring color graphics depicting the negative health consequences of smoking with new textual warning statements on cigarette packages and advertisements in the United States.¹ In March 2020, the FDA finalized the “Required Warnings for Cigarette Packages and Advertisements” rule, which specifies 11 textual warning statements as well as accompanying concordant photo-realistic color images.² The rule also requires the random and equal display and distribution of the warnings on packages and the quarterly rotation of the warnings in alternating sequence in advertisements.² Due to ongoing litigation, the effective date of the final rule has been postponed.³ This rule is significant for tobacco prevention and control efforts, as longitudinal studies around the globe have found that stronger cigarette pack warnings are associated with increases in tobacco-related knowledge, quitline calls, and reductions in smoking behavior.⁴ Globally, 126 countries/jurisdictions have finalized requirements for cigarette pictorial health warnings (“pictorial warnings”).⁵ Pictorial warnings have led to decreased smoking in some countries.^6,7

Assessing public support for pictorial warnings is important because public support can stimulate and reinforce implementation, be used by public policy makers, and serve as an evaluation of a policy’s impact.⁸ Political will is a necessary component of effective public health policy implementation,⁹ and public opinion influences political will.^10,11 However, few studies to date have examined public support for pictorial warnings in the United States.

One study using data from 13 waves of the National Adult Tobacco Survey found that overall support for “warning labels on cigarette packs that show graphic images of damage caused by smoking, such as black lungs” increased from approximately 58% in 2007 to 68% in 2012.¹² This study found that support was highest among Hispanic individuals followed by African Americans and non-Hispanic Caucasian individuals, and that those who were less affluent and less educated were more supportive than their counterparts. Younger smokers were more supportive than older smokers. Findings from other studies have similarly found that tobacco-related policy support changes over time.^13–15 A web-based nationally representative sample of US adults showed that support for graphic warnings on ads and packs was about 45% in 2013.¹⁶ Differences in these findings may be due to variation in methods and specificity of assessment. Both studies reported lower support for pictorial warnings among smokers than nonsmokers and found that support was associated with sociodemographic characteristics of the populations surveyed.

The goal of the current study is to build on this prior work to understand current population-level characteristics associated with support for and neutrality or opposition to cigarette pack pictorial health warnings. Findings may help guide targeted public health communications as changes are implemented. This is the first nationally representative survey to assess public support for pictorial warnings among US adults since publication of the 2020 final rule. It will provide a baseline for examining whether policy implementation increases support for this warning policy over time and may inform ancillary interventions and communication campaigns.

Methods

Data were from the Health Information National Trends Survey (HINTS) 5 Cycle 4, a nationally representative, cross-sectional, self-administered postal survey of civilian, noninstitutionalized US adults age 18 and older. HINTS 5 received an expedited approval from the Westat IRB on March 28, 2016 and was subsequently reviewed by the NIH Office of Human Subjects Research and given a nonhuman subjects determination via exemption #13204 on April 25, 2016. Data were collected February through June 2020 (N = 3865). The weighted response rate was 36.7%. Details about HINTS have been published elsewhere.^17–19

Sociodemographics, political viewpoint, children in the household, smoking, and electronic cigarette use were assessed in association with the item, “To what extent would you support or oppose that cigarette packs should be required to have warning labels that use both images and words to show the negative health effects of smoking?” (strongly oppose, oppose, neither support nor oppose, support, and strongly support). Selection of covariates was based on prior research and their availability in HINTS. Analyses were conducted in SUDAAN using sample weights to produce population-level point estimates and a set of 50 jackknife replicate weights to compute variance estimates. Weighted, unadjusted cross-tabulations provided bivariate estimates. Weighted, multivariable logistic regression examined predictors of being neutral/opposed versus in support (referent) in order to isolate and identify characteristics of those who do not report a supportive stance toward the pictorial warning policy. A sensitivity analysis utilizing multivariable, multinomial logistic regression to examine the odds of being “opposed versus supportive” and “neutral versus supportive” is provided in Supplementary Table 1.

Results

Overall, more US adults strongly supported (37.8%) or supported (32.2%) cigarette pack pictorial health warnings than opposed (4.8%) or strongly opposed (4.3%). Approximately 20.9% of US adults neither supported nor opposed pictorial warnings. Participant characteristics, weighted unadjusted estimates, and weighted adjusted odds of neutrality or opposition are shown in Table 1. More than half (54.7%) of current smokers, 66.4% of former smokers, and 74.6% of never smokers supported pictorial warnings. The data suggest that there was widespread support for pictorial warnings across population subgroups, ranging from the lowest proportion of support among current smokers (54.7%, 95% confidence interval [CI; 45.5, 63.6]) to the highest proportion of support among individuals age 75 or older (78.5%, 95% CI [72.9, 83.2]).

Table 1.

Descriptive Characteristics, Weighted Unadjusted Estimates, and Weighted, Adjusted Odds of Neutrality or Opposition Toward Cigarette Pack Pictorial Health Warnings, by Participant Characteristics. Health Information National Trends Survey (HINTS) 5 Cycle 4, 2020.

Characteristic	Analytic sample (n = 3865)	Unadjusted		Adjusted Neutral or Oppose vs. Support (n = 2997)
	Weighted % (n)	Weighted % (95% CI)		Odds ratio	95% CI
		Neutral or Oppose	Support
Age group
18–34 (ref)	25.5 (484)	29.1 (22.9, 36.3)	70.9 (63.7, 77.1)	—	—
35–49	24.8 (703)	31.9 (26.8, 37.5)	68.1 (62.5, 73.2)	1.17	0.74–1.84
50–64	26.9 (1142)	32.5 (27.8, 37.6)	67.5 (62.4, 72.2)	0.98	0.61–1.57
65–74	11.6 (869)	27.9 (23.4, 32.8)	72.1 (67.2, 76.6)	0.79	0.47–1.32
75+	8.4 (540)	21.5 (16.8, 27.1)	78.5 (72.9, 83.2)	0.55*	0.33–0.94
Sex on birth certificate
Female	50.2 (2204)	27.0 (24.2, 30.0)	73.0 (70.0, 75.8)	0.77	0.56–1.05
Male (ref)	47.6 (1561)	33.4 (29.0, 38.1)	66.6 (61.9, 71.0)	—	—
Race/ethnicity
Non-Hispanic Black or African American	10.3 (481)	28.6 (22.2, 35.9)	71.4 (64.1, 77.8)	0.88	0.55–1.38
Hispanic	15.7 (596)	23.5 (17.4, 30.8)	76.5 (69.2, 82.6)	0.64	0.40–1.05
Non-Hispanic Asian	4.8 (161)	22.6 (12.6, 37.1)	77.4 (62.9, 87.4)	0.72	0.30–1.7
Non-Hispanic Other	3.1 (119)	34.3 (19.6, 52.8)	65.7 (47.2, 80.4)	1.02	0.48–2.16
Non-Hispanic White (ref)	58.7 (2133)	32.6 (29.2, 36.1)	67.4 (63.9, 70.8)	—	—
Income
<$35 000	28.1 (1255)	29.7 (25.4, 34.4)	70.3 (65.6, 74.6)	0.89	0.60–1.32
$35 000–49 999	12.4 (516)	32.3 (25.1, 40.3)	67.7 (59.7, 74.9)	1.04	0.66–1.64
$50 000–74 999	17.8 (649)	28.4 (22.6, 35.1)	71.6 (64.9, 77.4)	0.79	0.53–1.18
$75 000+ (ref)	41.1 (1427)	30.6 (26.8, 34.7)	69.4 (65.3, 73.2)	—	—
Education
Less than high school	7.8 (273)	33.4 (22.4, 46.6)	66.6 (53.4, 77.6)	1.31	0.53–3.27
High school graduate	21.9 (705)	33.8 (27.7, 40.4)	66.2 (59.6, 72.3)	1.46	0.85–2.50
Some college	38.1 (1081)	30.9 (26.9, 35.3)	69.1 (64.7, 73.1)	1.14	0.77–1.69
College graduate or more (ref)	29.4 (1663)	26.2 (22.5, 30.2)	73.8 (69.8, 77.5)	—	—
Sexual orientation
Homosexual, gay, or lesbian	2.5 (81)	31.6 (15.0, 54.6)	68.4 (45.4, 85.0)	0.72	0.18–2.83
Bisexual	2.6 (82)	27.3 (13.7, 47.0)	72.7 (53.0, 86.3)	0.98	0.34–2.81
Other	1.6 (59)	29.9 (15.8, 49.1)	70.1 (50.9, 84.2)	1.19	0.33–4.31
Heterosexual or straight (ref)	88.4 (3402)	30.4 (27.8, 33.1)	69.6 (66.9, 72.2)	—	—
Marital status
Married	49.1 (1806)	29.1 (25.8, 32.7)	70.9 (67.3, 74.2)	0.86	0.64–1.17
Not married (ref)	47.9 (1915)	31.2 (27.5, 35.1)	68.8 (64.9, 72.5)	—	—
Smoking status
Current	13.6 (436)	45.3 (36.4, 54.5)	54.7 (45.5, 63.6)	1.99*	1.12–3.52
Former	22.6 (935)	33.6 (28.8, 38.8)	66.4 (61.2, 71.2)	1.25	0.90–1.73
Never (ref)	62.1 (2422)	25.4 (22.7, 28.3)	74.6 (71.7, 77.3)	—	—
E-cigarette use status
Current	6.3 (114)	40.1 (23.3, 59.5)	59.9 (40.5, 76.7)	1.10	0.46–2.62
Former	12.5 (382)	39.3 (30.9, 48.4)	60.7 (51.6, 69.1)	1.20	0.67–2.16
Never (ref)	79.7 (3314)	27.6 (25.1, 30.3)	72.4 (69.7, 74.9)	—	—
Geography
Rural	12.2 (430)	39.0 (31.8, 46.6)	61.0 (53.4, 68.2)	1.30	0.89–1.89
Urban (ref)	87.8 (3435)	28.8 (26.1, 31.7)	71.2 (68.3, 73.9)	—	—
Any children under age 18 in household
Yes	32.2 (879)	26.2 (21.7, 31.2)	73.8 (68.8, 78.3)	0.67*	0.46–0.98
No (ref)	61.8 (2676)	32.3 (29.3, 35.3)	67.7 (64.7, 70.7)	—	—
Political viewpoint
Moderate	33.9 (1200)	31.3 (27.6, 35.3)	68.7 (64.7, 72.4)	1.33	0.91–1.92
Conservative	30.7 (1223)	31.6 (26.8, 36.9)	68.4 (63.1, 73.2)	1.32	0.89–1.97
Liberal (ref)	27.0 (1058)	25.3 (20.4, 30.8)	74.7 (69.2, 79.6)	—	—

Missing data ranged from 0.5% to 8.4%: age 2.7% (n = 123), sex 2.2% (n = 100), race/ethnicity 7.3% (n = 375), income 0.7% (n = 18), education 2.8% (n = 143), sexual orientation 4.9% (n = 239), marital status 3.0% (n = 144), smoking status 0.5% (n = 17), e-cigarette use status 1.5% (n = 55), any children under 18 in household 6.0% (n = 310), political viewpoint 8.4% (n = 384). CI = confidence interval.

*p < .05.

Results from the multivariable logistic regression revealed that current smokers had almost twice the odds of being neutral/opposed to pictorial warnings as never smokers (odds ratio [OR] = 1.99, 95% CI 1.12–3.52). Those who were age 75 or older (OR = 0.55, 95% CI 0.33–0.94) and those with children in their household (OR = 0.67, 95% CI 0.46–0.98) had lower odds of being neutral/opposed to pictorial warnings, compared to those ages 18–34 and those with no children in their household, respectively.

Discussion

The current analyses estimate levels of public support for cigarette pack pictorial health warnings contemporaneous with FDA’s issuance of the 2020 final rule, but prior to implementation of the rule. Our study finds that a majority of adults (69.9%) either support or strongly support pictorial warnings, with less than 10% opposed. While pictorial warnings on cigarette packs and advertisements are mandated by the TCA, they have faced opposition from tobacco companies.³ Our results suggest that pictorial warnings are broadly supported across numerous demographic parameters, with over 60% of all subgroups reporting support across age, sex, race/ethnicity, income, education, sexual orientation, marital status, geography, and political viewpoint. This is noteworthy because broad public support can be an important lever for policy implementation.²⁰

Two previous national studies similarly examined support for pictorial warnings in the United States.^12,16 These studies showed that smokers were more neutral/opposed than never smokers.^12,16 Consistent with prior studies, we found that current smokers were more neutral or opposed to pictorial warnings than never smokers; nonetheless, our study found that more than half of current smokers (54.7%) were supportive of pictorial warnings in 2020. In 2013, 29% of smokers supported pictorial warnings on packs.¹⁶ Findings from the National Adult Tobacco Survey showed that smoker support for pictorial warnings increased significantly from approximately 35% in 2007 to 62% in 2011 but fell in 2012 to 40%.¹² Our study is the first to demonstrate that adults with children in the household had lower odds of being neutral/opposed than those without children. This is likely a reflection of strong support for policies that help protect children and youth from becoming tobacco users and is consistent with existing literature that suggests parents support tobacco control policies that aim to protect young people.²¹

Research has shown that public support for tobacco prevention and control policies rises once they take effect.¹⁵ An experimental study of smokers found that exposure to pictorial warnings on their cigarette packs increased policy support after four weeks compared to text-only warnings.²² The increase in support for pictorial warnings post-exposure was mediated by greater perceived message effectiveness and having more conversations about the policy. Communicating evidence of the policy’s effectiveness may further increase already high levels of public support, as supported by a meta-analysis on public support for policies in a range of domains.¹³

There are some limitations to this study. Due to the cross-sectional nature of the data, these results are descriptive, and causality cannot be inferred. Although the response rate for HINTS 5 Cycle 4 (36.7%) is lower than response rates for some in-person, interviewer administered surveys, prior research suggests that the response rates for HINTS are comparable to other population-based health surveys that employ a similar methodology.²³ Future research can build on these findings including continued monitoring of how support for pictorial warnings and related behaviors (eg, initiation, escalation, cessation) change post-implementation. A focused examination of current smokers and the association between their smoking-related behaviors, such as heaviness of smoking or intentions to quit, and support for pictorial warnings could elucidate subgroups for targeted messaging.

Our study demonstrates that, prior to implementation of cigarette pack pictorial health warnings, there is broad public support for pictorial warnings (69.9% overall) among a nationally representative sample of US adults. Opposition and neutrality varied by smoking status, age, and presence of children in the household. While the already high level of public support may increase further with policy implementation, it could also be strengthened through information campaigns that convey the evidence that pictorial warnings are an effective public health strategy.¹³ Campaigns to increase perceptions of effectiveness among current smokers could be particularly valuable, as this group consistently reports the lowest levels of support for policies requiring pictorial warnings. Furthermore, those with children in the household had lower odds of being neutral/opposed, suggesting that message framing which focuses on the protection of youth from tobacco may be an effective way to increase public support for pictorial warnings.

Supplementary Material

A Contributorship Form detailing each author’s specific involvement with this content, as well as any supplementary data, are available online at https://academic.oup.com/ntr.

Acknowledgments

The authors gratefully acknowledge Westat for HINTS data collection and Tim McNeel at Information Management Services, Inc. for analytic support. The content of this publication does not necessarily reflect the views or policies of the Department of Health and Human Services, nor does the mention of trade names, commercial products, or organizations imply endorsement by the US government. The findings and conclusions in this report are those of the authors and may not necessarily represent NIH or CDC positions or policies.

Funding

HINTS 5 was funded by National Cancer Institute via contract to Westat (HHSN2612012000028I).

Declaration of Interests

None declared.

Data Availability

Data available at https://hints.cancer.gov/data/default.aspx.