Youth and adult understanding of public displays of information about harmful constituents in cigarette smoke

Abstract

The U.S. FDA is required by law to publicly display a list of harmful and potentially harmful constituents (HPHCs) “by brand and by quantity in each brand and subbrand” in a format that is “understandable and not misleading to a lay person.” An online experiment examined youth and adult understanding of which HPHCs are present in cigarette smoke, understanding of health effects of smoking cigarettes, and endorsement of misleading information after viewing HPHC information displayed in one of six formats. We recruited youth (N = 1324) and adults (N = 2904) from an online panel and randomized them to one of six formats of presenting HPHC information. Participants responded to survey items before and after exposure to an HPHC format. Understanding of HPHCs in cigarette smoke and understanding of health effects of cigarette smoking significantly increased pre- to post-exposure for all formats. Respondents (20.6% to 73.5%) endorsed misleading beliefs after exposure to information about HPHCs. Endorsement of the one misleading belief that was measured pre- and post-exposure significantly increased for viewers of four formats. All formats increased understanding of HPHCs in cigarette smoke and the health effects of smoking cigarettes, but some participants endorsed each misleading belief after exposure to HPHC information.

Tobacco products and tobacco smoke contain thousands of harmful or potentially harmful constituents (HPHCs), many of which are linked to morbidity and mortality (Talhout et al., 2011; U.S. Department of Health and Human Services, 2010). For example, tobacco products and tobacco smoke contain HPHCs that cause addiction (e.g., nicotine), cancer (e.g., formaldehyde), fetal development and sexual problems (e. g., toluene), heart attack and stroke (e.g., benzene), and permanent breathing problems (e.g., ammonia) (Food and Drug Administration, n. d.). The Family Smoking Prevention and Tobacco Control Act (Tobacco Control Act) requires the U.S. Food and Drug Administration (FDA) to publicly display a list of HPHCs in tobacco products and tobacco smoke by brand, and by quantity in each brand and subbrand, in a format that is understandable and not misleading to a layperson (Food and Drug Administration, 2015). The statute does not specify the format or location for display of the HPHC information. Awareness of HPHCs is associated with potentially beneficial outcomes such as discouragement from wanting to try or use tobacco products (Brewer et al., 2016; Wiseman et al., 2016; Jeong et al., 2020), quit intentions (Lazard et al., 2020), and quit attempts (Morgan et al., 2017). However, communicating about HPHCs is challenging due to the public’s lack of understanding about HPHCs (Brewer et al., 2016; Wiseman et al., 2016; Jeong et al., 2020; Morgan et al., 2017; Hall et al., 2014; Moracco et al., 2016; Boynton et al., 2016) and misbeliefs about HPHCs (Brewer et al., 2016; Jeong et al., 2020; Morgan et al., 2017; Hall et al., 2014; Moracco et al., 2016; Pepper et al., 2017a). For example, in general, survey respondents frequently misbelieve that HPHCs mostly originate in additives introduced during cigarette manufacturing rather than naturally occurring in tobacco or created by combustion (Brewer et al., 2016; Jeong et al., 2020; Morgan et al., 2017; Hall et al., 2014; Moracco et al., 2016). Most people tested also incorrectly believe that the amount of an HPHC always correlates with its level of harm (Morgan et al., 2017) (i.e., that tobacco brands with fewer HPHCs, lower levels of an HPHC, or labeled “additive-free” are less harmful) (Wiseman et al., 2016). These misbeliefs are problematic because some adolescent and young adult focus group participants have also indicated that they may switch to brands that have lower concentrations of certain HPHCs (Wiseman et al., 2016), which could lead to a lower likelihood of quitting (Tindle et al., 2009).

Recognizing these communication challenges, researchers have examined whether variations in HPHC information content and format influence participants’ understanding and endorsement of misleading beliefs. For example, Lazard et al. examined adult and adolescents’ responses to webpages displaying cigarette HPHC information (Lazard et al., 2020). Participants who viewed a display with health effects listed first and HPHCs listed underneath had greater understanding of HPHCs than participants who viewed a display with HPHCs grouped by their associated health effects, but layout did not otherwise affect understanding and misleadingness (Lazard et al., 2020). Pepper et al. examined adult cigarette users’ responses to stimuli that varied in content and format displaying HPHCs in a smoked cigarette (Pepper et al., 2017b). Including common uses of each HPHC, describing health harms of each HPHC in clinical language, describing health harms of each HPHC in plain language, and graphic vs. non-graphic format affected understanding and misleadingness. In some cases, the formats that were most understandable were also the most misleading (i.e., viewers were more likely to believe that the amount of an HPHC indicates the chance of experiencing harm) (Pepper et al., 2017b). Byron et al. examined cigarette users’ and nonusers’ responses to webpages that presented chemical and health effect information about HPHCs in cigarette smoke (Byron et al., 2018). Participants’ understanding of HPHCs was greatest when webpages presented both familiar and unfamiliar HPHCs and their understanding of health effects was greatest when health effects were presented as text only (Byron et al., 2018). However, variations in content and format did not affect misleadingness (Byron et al., 2018). The results of these studies suggest there is no consensus about the best format for displaying HPHC information or what information to include to maximize understanding and minimize misleadingness.

Given the need to effectively communicate about HPHCs and tobacco-related harms, FDA conducted an experiment to gain further insight into adult and youth comprehension of information about HPHCs in cigarette smoke. We tested six formats that provided HPHC information and examined whether participants’ understanding of HPHCs in cigarette smoke and understanding of the health effects of cigarette smoking changed after exposure to information about HPHCs. We also examined whether, after viewing HPHC information, participants endorsed beliefs indicating that they were misled by the formats. Unlike previous research, this study tests multiple HPHC information formats for a hypothetical cigarette brand to examine which format, if any, has the potential to meet the Tobacco Control Act requirement that HPHC information be presented in a way that is understandable and not misleading to a lay person.

1. Methods

1.1. Participants and procedure

From October 7, 2020 to January 18, 2021, we conducted an online experiment with youth aged 13–17 (N = 1324) who currently used smokeless tobacco (i.e., used smokeless tobacco on at least one of the past 30 days), currently smoked cigarettes (i.e., smoked cigarettes on at least one of the past 30 days), or were not a cigarette user but were susceptible to smoking cigarettes, and adults (N = 2904) who currently used smokeless tobacco, currently smoked cigarettes, or did not use smokeless tobacco or cigarettes. We recruited smokeless tobacco users and non-cigarette users because of concerns that users of potentially less risky products or tobacco nonusers may have decreased perceptions about the harms of smoking after viewing cigarette HPHC information. We recruited cigarette users because they are the target audience for cigarette HPHC information.

Adult participants were recruited via email and completed an online consent form and study screener. Eligible adult participants who completed the study survey received an incentive via the online panel. Youth were recruited via emails to adult panelists whose profiles indicated they were the parent or guardian of a child aged 13–17. Parents provided affirmative permission and youth provided affirmative assent to participate in the study. The online market research vendor compensated parents of youth regardless of youth survey eligibility or survey completion. Quotas were used to ensure participants included various age and tobacco use groups and inclusion of hard-to-reach populations.

After responding to pre-exposure questions on understanding of HPHCs in cigarette smoke and understanding of the health effects of smoking, we randomized participants using simple random assignment to view HPHC information about cigarette smoke displayed in one of six formats. After viewing the format for a minimum of 30 s, participants immediately responded to post-exposure questions. The format stayed visible during question completion, and participants could refer to it when answering the questions. At the end of the study, participants were debriefed. All materials and procedures for this study were approved by the Institutional Review Boards at FDA (Study #18–038CTP) and RTI (IRB ID: STUDY00021430).

1.2. Formats

In prior research conducted by FDA, we found significant gaps in consumer understanding of HPHCs and their health effects. We then conducted 50 interviews to gain a better understanding of how to convey HPHC information to a layperson. We designed six formats for displaying HPHC information based on previous research, including formats developed as part of Center for Tobacco Product, FDA funded administrative supplements to the National Institutes of Health Tobacco Centers of Regulatory Science, published scientific literature, graphic design principles, and health communication principles. All formats listed information about 18 HPHCs present in a fictitious cigarette brand, Durham Cigarettes. The 18 HPHCs are from the Abbbreviated List of Harmful and Potentially Harmful Constituents provided in draft guidance for industry. All formats included information about health effects associated with HPHCs as previous research has found this led to greater knowledge of those health effects (Byron et al., 2018), higher message effectiveness ratings (Noar et al., 2018), and increased potential influence of HPHC information (Lazard et al., 2019). Several of the formats employed graphical elements, such as icons, to increase understanding (Houts et al., 2006; Lazard et al., 2017). However, the formats varied in the type and amount of information included and in the graphical display of that information (See supplemental material for images of all six formats). The study was not designed to compare formats but rather to examine how well each format conveyed HPHC information in a manner that was understandable and not misleading. Supplemental Table 1 provides an overview of the content and visual display of the formats.

1.3. Measures

1.3.1. Demographic information

We collected information on age, sex, race/ethnicity, and, among adults 18 and older, household income, and educational attainment.

1.3.2. Tobacco use

For youth aged 13–17, current cigarette use was defined as having smoked on one or more of the past 30 days. Current smokeless tobacco use was defined as having used smokeless tobacco on one or more of the past 30 days. Youth who had never smoked a cigarette or had not smoked in the past 30 days were only eligible to participate if they were susceptible to smoking (Nicksic and Barnes, 2019). Susceptibility to cigarette smoking was defined as not having smoked cigarettes in the past 30 days and reporting either (1) being “a little,” “somewhat,” or “very” curious about cigarettes; (2) responding “definitely yes,” probably yes,” or “probably not” to the question “Do you think you will smoke a cigarette at any time in the next year?”; or (3) responding “definitely yes,” probably yes,” or “probably not” to the question “If one of your best friends offered you a cigarette, would you smoke it?”

For adults aged 18 and older, current cigarette use was defined as having smoked on one or more of the past 30 days and having smoked 100 or more cigarettes in lifetime. Current smokeless tobacco use was defined as having used smokeless tobacco on one or more of the past 30 days. Non-users of cigarettes and smokeless tobacco were defined as having not used smokeless tobacco in the past 30 days and is not a current cigarette user (i.e., never smoked, has not smoked in the past 30 days, or has not smoked a minimum of 100 cigarettes in lifetime)

We also assessed participants’ past 30-day use of other tobacco products, including electronic nicotine products; cigars, cigarillos, or filtered cigars; tobacco pipe; or waterpipe and hookah.

1.3.3. Understanding of HPHCs in cigarette smoke

Participants responded to five items before and after format exposure to assess understanding of HPHCs in cigarette smoke (e.g., “Does cigarette smoke contain ammonia?”) (Byron et al., 2018). All HPHCs listed in these items were included in all six of the HPHC formats. Responses were coded as correct (“Yes”) or incorrect (“No” or “Don’t Know”) and correct responses were summed, resulting in a 0–5 scale.

1.3.4. Understanding of the health effects of smoking

Participants responded to seven items before and after format exposure to assess understanding of the health effects of smoking (e.g., “Does smoking cause heart attacks?”) (Byron et al., 2018). Responses were coded as correct (“Yes”) or incorrect (“No” or “Don’t Know”) and correct responses were summed, resulting in a 0–7 scale.

1.3.5. Misleadingness

Participants responded to one misleadingness item pre- and post-exposure: “If you learned that your cigarettes have a lot more of a dangerous chemical than other cigarettes, how likely, if at all, would you be to switch brands?” Participants who responded “Extremely likely” or “Likely” were considered misled.

Participants responded to 14 misleadingness items post-exposure only (Byron et al., 2018). Participants were considered misled on the belief if they responded “Strongly Agree” or “Agree” to the belief item. Three of the 14 items were asked only of current cigarette users.

Five items assessed beliefs that HPHC quantity substantially affects risk (e.g., “A cigarette is much safer to smoke if it has less arsenic than other cigarettes”).

Two items addressed beliefs that brands vary in risk (e.g., “Some cigarette brands are much more harmful to smoke than others”).

Like Byron et al., we asked two items only of current cigarette users to assess whether they made comparisons between cigarette brands with incomplete data (e.g., “My brand of cigarettes is probably safer than Durham”).

Five items assessed whether participants perceived that smoking is less dangerous after viewing the HPHC information. One item was asked of current cigarette users only (i.e., “Reading this information about cigarettes makes me less likely to quit smoking”). The other four items were asked of all participants (e.g., “It’s safer to smoke Durham cigarettes than most other cigarettes”). Participants were considered misled if they responded, “Strongly Agree” or “Agree.” Participants were also considered misled if they responded, “Not at all harmful” or “A little harmful” to the item “How harmful are Durham cigarettes (the cigarettes described)?”

1.4. Analysis

Based on a pre-specified statistical analysis plan, we examined sociodemographic and tobacco use characteristics using descriptive statistics. We initially did not compare outcomes based on tobacco use status or age because an HPHC list must be publicly displayed to meet the regulatory requirement and may therefore be viewed by all members of the public, regardless of their tobacco use status or age. All members of the public must therefore understand and not be misled by HPHC information. However, after reviewer feedback we conducted ad-hoc analyses presenting separate results for youth and adults. In supplemental tables 3, 5, and 6, we provide results further separated by tobacco use status. To assess whether understanding increased pre- to post- exposure, we conducted paired t-tests within each study condition. To assess whether participants endorsed misleading beliefs after exposure to HPHC information, we calculated the proportion of participants misled for each of the misleadingness items across the six formats. For the one misleadingness item assessed pre- and post-exposure, we conducted McNemar tests of paired proportions to determine whether responses differed pre- and post-exposure. One observation was excluded from analysis of understanding of HPHCs in cigarette smoke due to missingness. No other observations were excluded from the results reported here due to missingness.

2. Results

2.1. Sample characteristics

Table 1 reports sociodemographic and tobacco use characteristics among participants. Only sex and household income differed significantly across the six study conditions.

Table 1.

Participant Sociodemographic and Tobacco Use Characteristics by Age Group (N = 4228).

Characteristic	Adult (n = 2904; 68.7%)	%	Youth (n = 1324; 31.3%)	%
Sex *
Male	1356	32.1	821	19.4
Female	1537	36.4	496	11.7
Other	8	0.2	3	0.1
Do not wish to answer	3	0.1	4	0.1
Race/ethnicity
White, non-Hispanic	2155	51.0	1085	25.7
Black or African American, non-Hispanic	554	13.1	148	3.5
American Indian or Alaska native	119	2.8	22	0.5
Asian, non-Hispanic	96	2.3	85	2.0
Native Hawaiian/other Pacific islander	24	0.6	0.2	0.7
Hispanic	408	9.9	210	5.1
Prefer not to answer	25	0.6	19	0.4
Educational attainment
Less than high school graduate or GED	139	4.9	–	–
High school graduate or GED	527	18.1	–	–
Some college or technical training school	918	31.6	–	–
College graduate	961	33.1	–	–
Postgraduate education	359	12.4	–	–
Household income *
Less than $25,000	504	17.4	–	–
$25,000 to $49,999	746	25.7	–	–
$50,000 to $74,999	787	27.1	–	–
$75,000 or more	867	29.9	–	–
Tobacco and nicotine use
Current cigarette user	2385	56.4	545	12.9
Susceptible to cigarette smoking	–	–	778	58.8
Current user of smokeless tobacco	1677	39.7	492	11.6
Current non-user of cigarettes and smokeless tobacco	500	11.8	–	–
Current user of electronic nicotine products	1191	28.2	547	12.9
Current user of traditional cigars, cigarillos, or filtered cigars	1347	31.9	423	10.0
Current user of tobacco pipe	840	19.9	239	5.7
Current user of waterpipe/hookah	838	19.8	274	6.5
Current non-user of tobacco products	444	10.5	–	–

^*Statistically significant difference in proportion among six study conditions, based on chi-square tests (p < 0.05).

2.2. Understanding of HPHCs in cigarette smoke

Understanding of HPHCs in cigarette smoke increased significantly pre- to post-exposure for all six formats (See Table 2). Mean pre-exposure scores on the 0–5 scale ranged from 1.73 (Youth, Format 3) to 1.95 (Adult, Format 5). Mean post-exposure scores ranged from 4.31 (Adult, Format 1) to 4.62 (Youth, Formats 4 and 5). Difference scores ranged from 2.43 (Adult, Format 6) to 2.79 (Youth, Formats 4 and 5). Supplemental Table 2 compares adult and youth difference scores.

Table 2.

Mean Understanding of HPHCs in Cigarette Smoke Pre- and Post-Format Exposure, by Age Group (N = 4227 total participants; N = 2903 adults, N = 1324 youth).

	Age group	Pre-exposure		Post-exposure		Difference (post – Pre)
Format		Mean	95% CI	Mean	95% CI	Mean	95% CI	p-value
1 Bar	Overall	1.81	1.68, 1.95	4.39	4.30, 4.49	2.58	2.42, 2.74	<0.001
	Adult	1.79	1.63, 1.96	4.31	4.20, 4.42	2.52	2.33, 2.43	<0.001
	Youth	1.85	1.61, 2.10	4.57	4.41, 4.73	2.71	2.43, 2.99	<0.001
2 Spaghetti	Overall	1.87	1.74, 2.01	4.50	4.42, 4.59	2.63	2.47, 2.79	<0.001
	Adult	1.91	1.75, 2.08	4.49	4.39, 4.59	2.58	2.39, 2.77	<0.001
	Youth	1.79	1.54, 2.04	4.53	4.37, 4.68	2.74	2.46, 3.02	<0.001
3 Smoke	Overall	1.87	1.74, 2.01	4.46	4.38, 4.54	2.59	2.44, 2.75	<0.001
	Adult	1.92	1.76, 2.08	4.43	4.34, 4.53	2.51	2.33, 2.69	<0.001
	Youth	1.73	1.51, 2.00	4.53	4.39, 4.67	2.78	2.50, 3.05	<0.001
4 Table	Overall	1.87	1.73, 2.01	4.46	4.38, 4.55	2.60	2.44, 2.75	<0.001
	Adult	1.91	1.74, 2.07	4.38	4.28, 4.49	2.47	2.28, 2.67	<0.001
	Youth	1.82	1.58, 2.07	4.62	4.47, 4.76	2.79	2.52, 3.06	<0.001
5 Infographic	Overall	1.91	1.77, 2.04	4.50	4.42, 4.58	2.59	2.44, 2.75	<0.001
	Adult	1.95	1.78, 2.11	4.45	4.35, 4.55	2.50	2.32, 2.69	<0.001
	Youth	1.82	1.58, 2.07	4.62	4.47, 4.76	2.79	2.52, 3.06	<0.001
6 Wildcard	Overall	1.87	1.73, 2.00	4.39	4.30, 4.48	2.52	2.36, 2.67	<0.001
	Adult	1.91	1.75, 2.08	4.34	4.24, 4.45	2.43	2.24, 2.62	<0.001
	Youth	1.77	1.52, 2.01	4.48	4.32, 4.64	2.72	2.44, 2.99	<0.001

Notes: Assessed using a summative scale from 0 to 5; CI = confidence interval.

2.3. Understanding of the health effects of cigarette smoking

Understanding of health effects of cigarette smoking increased significantly pre- to post- exposure for all six formats (see Table 3). Mean pre-exposure scores ranged on the 0–7 scale from 4.96 (Adult, Format 1) to 5.55 (Youth, Format 3). Mean post-exposure scores ranged from 5.95 (Adult, Format 4) to 6.65 (Youth, Format 3). Difference scores ranged from 0.79 (Adult, Format 4) to 1.28 (Youth, Format 6). Supplemental Table 4 compares youth and adult difference scores.

Table 3.

Mean Understanding of Health Effects of Cigarette Smoking Pre- and Post-Format Exposure, by Age Group (N = 4228 total participants; N = 2904 adults, N = 1324 youth).

Format		Pre-exposure		Post-exposure		Difference (post – Pre)		p-value
	Age group	Mean	95% CI	Mean	95% CI	Mean	95% CI
1 Bar	Overall	5.11	4.96, 5.25	6.23	6.12, 6.35	1.12	0.99, 1.25	<0.001
	Adult	4.96	4.78, 5.13	6.07	5.93, 6.20	1.11	0.95, 1.27	<0.001
	Youth	5.44	5.18, 5.69	6.59	6.39, 6.79	1.15	0.92, 1.39	<0.001
2 Spaghetti	Overall	5.20	5.06, 5.34	6.21	6.10, 6.33	1.01	0.89, 1.14	<0.001
	Adult	5.04	4.88, 5.21	6.09	5.95, 6.23	1.05	0.90, 1.20	<0.001
	Youth	5.55	5.30, 5.80	6.49	6.28, 6.70	0.94	0.71, 1.16	<0.001
3 Smoke	Overall	5.24	5.11, 5.37	6.37	6.27, 6.47	1.13	1.01, 1.24	<0.001
	Adult	5.16	5.00, 5.32	6.24	6.12, 6.36	1.09	0.94, 1.23	<0.001
	Youth	5.43	5.19, 5.67	6.65	6.45, 6.83	1.22	1.00, 1.43	<0.001
4 Table	Overall	5.28	5.14, 5.41	6.10	5.98, 6.22	0.82	0.71, 0.93	<0.001
	Adult	5.15	4.99, 5.31	5.95	5.80, 6.09	0.79	0.65, 0.93	<0.001
	Youth	5.53	5.30, 5.77	6.41	6.20, 6.62	0.87	0.68, 1.07	<0.001
5 Infographic	Overall	5.32	5.18, 5.45	6.20	6.08, 6.32	0.88	0.76, 1.00	<0.001
	Adult	5.24	5.08, 5.40	6.11	5.97, 6.25	0.87	0.72, 1.01	<0.001
	Youth	5.49	5.25, 5.73	6.40	6.19, 6.61	0.91	0.69, 1.13	<0.001
6 Wildcard	Overall	5.20	5.06, 5.33	6.39	6.29, 6.49	1.19	1.07, 1.32	<0.001
	Adult	5.16	5.00, 5.32	6.32	6.19, 6.44	1.16	1.01, 1.31	<0.001
	Youth	5.28	5.04, 5.52	6.56	6.37, 6.74	1.28	1.06, 1.50	<0.001

Notes: Assessed using a summative scale with a range of 0–7; CI = confidence interval.

2.4. Misleadingness

2.4.1. Belief that HPHC quantity substantially affects risk

Across formats and age groups, >50% of participants endorsed most of the five misleading beliefs that HPHC quantity in cigarettes substantially affects risk (Table 4).

Table 4.

Endorsement of Beliefs Indicative of Being Misled by Format, by Age Group (N = 4228 for items asked of all participants; N = 2904 adults, N = 1324 youth. N = 2930 for items asked of current cigarette users only; N = 2385 adults, N = 545 youth).

Item	Age group	% endorsement of misleading beliefs by format
		1 bar	2 spaghetti	3 smoke	4 table	5 infographic	6 wildcard
Beliefs that HPHC quantity substantially affects risk
If a person who smokes cigarettes can’t quit, they should switch to a brand with fewer chemicals.	Overall	73.5	69.6	69.0	71.6	71.1	70.6
	Adult	71.9	68.0	71.2	73.4	71.6	68.9
	Youth	76.9	73.4	64.2	68.1	69.9	74.4
It’s much safer to smoke cigarettes with fewer chemicals.	Overall	55.3	53.9	54.3	54.2	56.7	56.1
	Adult	55.0	53.6	56.9	55.0	58.2	54.6
	Youth	56.1	54.6	48.4	52.6	53.4	59.4
A cigarette is much safer to smoke if it has less arsenic than other cigarettes.	Overall	51.6	51.3	52.0	50.1	52.3	52.3
	Adult	52.9	53.4	54.0	52.4	53.9	52.4
	Youth	48.9	46.8	47.4	45.3	48.9	52.1
A cigarette is much safer to smoke if it has less crotonaldehyde than other cigarettes.	Overall	50.8	51.1	49.0	48.1	51.6	51.0
	Adult	49.4	51.1	50.5	48.2	53.5	49.9
	Youth	53.9	50.9	45.6	47.8	47.5	53.4
A cigarette is much safer to smoke if it has less 4-aminobiphenyl than other cigarettes.	Overall	52.3	47.8	46.2	48.1	48.8	48.0
	Adult	52.1	48.1	47.9	47.4	49.4	48.3
	Youth	52.9	47.3	42.3	49.6	47.5	47.5
Beliefs that brands vary in risk
Some cigarette brands are much more harmful to smoke than others.	Overall	70.9	67.8	67.8	68.1	71.6	68.3
	Adult	69.6	65.7	66.9	66.6	70.0	66.8
	Youth	73.8	72.5	69.8	71.1	75.3	71.7
If a website had information like this for all cigarette brands, I would use it to see which cigarettes are safer than others.	Overall	66.2	65.5	67.6	66.0	66.0	66.9
	Adult	64.9	64.7	67.7	64.9	67.9	65.2
	Youth	69.2	67.4	67.4	68.1	61.6	70.8
Evidence of making comparisons with incomplete data
My brand of cigarettes is probably safer than Durham.a	Overall	47.6	45.9	48.7	50.5	49.0	47.7
	Adult	45.5	43.5	47.6	50.9	46.5	46.6
	Youth	55.8	56.7	53.6	48.9	60.2	52.8
My brand of cigarettes has fewer chemicals than Durham.a	Overall	44.8	46.3	50.3	49.5	46.3	48.8
	Adult	42.5	41.7	48.4	49.1	44.3	48.1
	Youth	53.9	66.7	59.5	51.1	55.7	51.7
Perceptions that smoking is less dangerous
Reading this information about cigarettes makes me less likely to quit smoking.a	Overall	40.7	43.6	44.1	44.5	45.5	41.7
	Adult	39.3	41.7	43.6	43.8	43.8	40.7
	Youth	46.2	52.2	46.4	47.7	53.4	46.2
After viewing this information about cigarettes, I now feel that smoking is less dangerous.	Overall	29.2	28.9	30.8	31.5	32.3	33.2
	Adult	32.0	30.2	33.1	34.0	34.2	35.1
	Youth	23.1	26.2	25.6	26.3	28.3	29.2
How harmful are Durham cigarettes (the cigarettes described)?	Overall	22.6	20.6	23.3	22.4	23.0	22.7
	Adult	26.2	22.2	26.6	24.7	24.5	27.0
	Youth	14.5	17.0	15.8	17.7	19.6	13.2
It’s safer to smoke Durham cigarettes than most other cigarettes.	Overall	31.6	28.4	33.0	32.6	33.5	32.5
	Adult	32.9	28.3	34.2	33.2	33.5	33.6
	Youth	29.0	28.4	30.2	31.5	33.3	30.1
Durham cigarettes have fewer chemicals than other cigarettes.	Overall	40.1	33.5	37.5	36.6	38.7	35.7
	Adult	32.9	28.3	34.2	33.2	33.5	33.6
	Youth	29.0	28.4	30.2	31.5	33.3	30.1

Note: Endorsement indicated by responding “Strongly Agree” or “Agree” on a 5-point scale for all belief items except “How harmful are Durham cigarettes (the cigarettes described)?”, for which endorsement was indicated by responding “Not at all harmful” or “A little harmful” on a 4-point scale.

^aAsked of current cigarette users only.

2.4.2. Belief that brands vary in risk

Most participants endorsed misleading beliefs that brands vary in risk after viewing HPHC information (Table 4).

2.4.3. Making comparisons with incomplete data

Approximately half of current cigarette users endorsed misleading beliefs indicative of making comparisons with incomplete HPHC information (Table 4).

2.4.4. Perception that smoking is less dangerous

For most items, fewer than half of participants perceived that smoking cigarettes is less dangerous after viewing HPHC information (Table 4).

2.4.5. Likelihood of switching

Among current cigarette users who viewed formats 1, 3, 5, or 6, there were significant increases from pre- to post- exposure in the proportion of those who reported that they would switch brands if they learned that their cigarettes have a lot more of a dangerous HPHC than other cigarettes (Table 5). There were no significant differences from pre-exposure to post-exposure for formats 2 and 4. The proportion of youth misled only significantly increased for those who viewed format 1 whereas the proportion of adults misled significantly increased for those who viewed formats 1, 3, 5, or 6.

Table 5.

Endorsement of Belief Indicating Switching due to HPHC Information Pre- and Post-Format Exposure (N = 2930 total participants; N = 2385 adults, N = 545 youth).

	Age group	Pre-exposure		Post-exposure		Difference (post – Pre)
Format		%	95% CI	%	95% CI	%	95% CI	p-value
1 Bar	Overall	71.0	67.1, 75.0	77.7	74.1, 81.3	6.6	3.1, 10.2	<0.001
	Youth	76.9	68.8, 85.0	87.5	81.1, 93.9	10.6	2.8, 18.3	0.008
	Adult	69.5	65.1, 74.0	75.2	71.0, 79.4	5.7	1.7, 9.6	0.006
2 Spaghetti	Overall	74.1	70.3, 78.0	75.8	72.0, 79.5	1.6	−1.7, 5.0	0.35
	Youth	86.7	79.6, 93.7	90.0	83.8, 96.2	3.3	−1.5, 8.2	0.18
	Adult	71.4	67.0, 75.8	72.6	68.2, 76.9	1.2	−2.7, 5.2	0.54
3 Smoke	Overall	72.2	68.2, 76.1	76.1	72.3, 79.9	3.9	0.1, 7.7	0.04
	Youth	81.0	72.6, 89.3	83.3	75.4, 91.3	2.4	−7.5, 12.3	0.64
	Adult	70.3	65.8, 74.8	74.6	70.3, 78.8	4.2	0.1, 8.4	0.04
4 Table	Overall	76.0	72.1, 79.9	79.2	75.6, 82.9	3.2	−0.3, 6.7	0.07
	Youth	85.2	77.8, 92.6	89.8	83.4, 96.1	4.5	−2.5, 11.6	0.21
	Adult	73.9	69.5, 78.3	76.8	72.5, 81.0	2.9	−1.1, 6.9	0.15
5 Infographic	Overall	70.0	65.9, 74.1	74.9	71.1, 78.7	4.9	1.1, 8.7	0.01
	Youth	83.0	75.1, 90.8	89.8	83.4, 96.1	6.8	−0.2, 13.9	0.06
	Adult	67.2	62.6, 71.8	71.6	67.2, 76.0	4.5	0.2, 8.8	0.04
6 Wildcard	Overall	71.0	66.9, 75.0	76.6	72.8, 80.3	5.6	2.1, 9.1	0.002
	Youth	82.4	74.6, 90.2	86.8	79.9, 93.8	4.4	−0.9, 9.7	0.10
	Adult	68.3	63.7, 72.9	74.2	69.8, 78.5	5.9	1.8, 10.0	0.005

Notes: CI = confidence interval.

3. Discussion

Public understanding of HPHCs in tobacco products and cigarette smoke is low (Brewer et al., 2016; Jeong et al., 2020; Morgan et al., 2017; Hall et al., 2014; Moracco et al., 2016; Boynton et al., 2016), but, in general, people are interested in learning about HPHCs (Wiseman et al., 2016; Morgan et al., 2017; Moracco et al., 2016; Boynton et al., 2016). Furthermore, the Tobacco Control Act requires that FDA publish a list of HPHCs in a way that is “understandable and not misleading to a layperson.” To address challenges in communicating HPHC information, we used previous research results to design six formats to communicate HPHC information, displayed them to participants, and tested understanding of HPHCs and health effects of smoking after viewing. We also tested whether the information presented in the formats was misleading.

Consistent with previous research (Lazard et al., 2020; Byron et al., 2018; Loken et al., 2021), participants’ understanding of HPHCs in cigarette smoke and the health effects of smoking increased after viewing HPHC information across all formats. All formats provided HPHC information that increased participant understanding and may, therefore, be useful models for providing people with information about HPHCs in tobacco products.

While all formats increased understanding, participants endorsed misleading beliefs after viewing the formats, which is also consistent with previous research. Beliefs that cigarette brands vary in risk and that variations in levels of individual chemicals in cigarettes substantially affects risk were among the most endorsed misleading beliefs, with 67.8–71.6% endorsing the statement that “some cigarette brands are much more harmful to smoke than others” and 69.0–73.5% endorsing the statement that “if a person who smokes cigarettes can’t quit, they should switch to a brand with fewer chemicals.” This finding is consistent with research demonstrating that people mistakenly believe they are at lower risk of harm from a product that contains fewer overall constituents or lower amounts of specific constituents (Lazard et al., 2020; Morgan et al., 2017; Byron et al., 2018) and are susceptible to brand switching in response to HPHC information (U.S. Department of Health and Human Services, 2010; Food and Drug Administration, n.d.; Food and Drug Administration, 2015). People expect that information provided to them is intended to be useful. Participants may, therefore, have assumed that brand and subbrand information about HPHC quantities was provided because it is important and relevant to their behavior (Berman et al., 2017). This may in part explain why participants indicated that they would switch brands if they learned that their cigarettes have higher amounts of a dangerous HPHC than other cigarettes. Displaying HPHC information by product category (e.g., cigarettes, ENDS, smokeless tobacco) rather than by brand may minimize potential misleadingness because individuals would not be able to make comparisons across various brands within a product category. Such displays could also educate consumers that many of the harms of smoking come from the combustion of tobacco and its inhalation, not from chemical additives (King et al., 2021).

Message features (e.g., providing familiar or unfamiliar HPHC names; providing health effect information as text only or text with icons) may influence whether HPHC information is understandable and not misleading to a lay person. Recognizing the importance of message features, we took care in designing the formats; however, our results still indicate that participants were misled. Presenting exact HPHC amounts may encourage participants to make incorrect risk comparisons. This may be due in part to challenges with health numeracy, defined as the ability to understand medical data (Centers for Disease Control and Prevention, 2021). Researchers have examined providing HPHC information with fewer numeric quantities, instead communicating overall harm or communicating information on only a few HPHCs. For example, Borgida et al. showed cigarette users and nonusers a graphic ordering smokeless tobacco brands on a gradient by toxicity and nicotine content (Borgida et al., 2015). Goldstein et al. showed adult cigarette users messages that provided information about one HPHC in cigarette smoke, its health effect, and a photo image of an individual displaying the health effect (Goldstein et al., 2021). The use of graphic elements in communicating HPHC information is also challenging. Including visual risk indicators elicits higher perceptions of usable information among viewers (Lazard et al., 2020) but supplementing quantitative information with visual features such as icons and graphic elements results in less understanding than providing text-only information (Pepper et al., 2017b; Byron et al., 2018).

Study results should be interpreted with caution considering several limitations. The sample was a non-representative convenience sample and included only individuals with access to laptop or desktop computers; thus, results cannot be generalized to the entire U.S. population. It is unclear whether participants held misleading beliefs prior to participating in the study or whether the HPHC information increased misunderstanding. Only the item related to brand switching measured pre- and post- HPHC information exposure. Future research may test misleading beliefs before and after exposure to HPHC information to assess changes in response to study stimuli. Exposure to HPHC information did not occur under real-world circumstances. In the real world, individuals would likely have to seek out HPHC information. They might pay less attention to the information because they are not required to view it for a minimum of 30 s as they were in this study, or they might pay more attention because they are seeking out the information by choice. Additionally, because the Tobacco Control Act requires HPHC information to be presented by brand and subbrand, individuals in the real world could compare across products or only look at their preferred brand, rather than seeing a single hypothetical brand as in this study. Future research may examine these possibilities.

In summary, this study examined whether viewing a format presenting HPHC information increased understanding of HPHCs in cigarette smoke and the health effects of smoking among youth and adults and mitigated misleading beliefs to meet Tobacco Control Act requirements. While understanding increased across all formats, participants still endorsed misleading beliefs post-exposure and the endorsement of one misleading belief (i.e., switching to a brand with fewer HPHCs) significantly increased pre- to post-exposure for four of the six formats. Communicating HPHC information in a way that is not misleading remains a challenge.

Data availability

Data will be made available on request.