August 2018, VOLUME142 /ISSUE 2

Adults’ Perceptions of Nicotine Harm to Children

  1. Catherine B. Kemp, BSN, MHAa,b,
  2. Claire Adams Spears, PhDa,b,
  3. Terry F. Pechacek, PhDa,c, and
  4. Michael P. Eriksen, ScDa,c
  1. aTobacco Center of Regulatory Science,
  2. bDivisions of Health Promotion and Behavior and
  3. cHealth Management and Policy, School of Public Health, Georgia State University, Atlanta, Georgia
  1. Ms Kemp and Dr Spears conceptualized and designed the study, conducted analyses, and wrote the manuscript; Drs Eriksen and Dr Pechacek developed the data collection instrument, oversaw data collection, collaborated on the data analysis plan, and revised the manuscript; and all authors approved the final manuscript as submitted and agree to be accountable for all aspects of the work.


BACKGROUND: Given the changing landscape of tobacco products in recent years, the array of products through which children could be exposed to nicotine has grown substantially. Thus, it is particularly important to understand adults’ perceptions of the harms of nicotine to children and to identify any sociodemographic factors related to inaccurate risk perceptions.

METHODS: Data were drawn from 2015 to 2016 US nationally representative surveys (n = 11 959). Using multinomial logistic regression analyses, we examined whether race, sex, education, tobacco product use, and having a minor child in the home are associated with the level of perceived harmfulness of nicotine to children.

RESULTS: Although the majority of respondents characterized nicotine as “definitely harmful” to children, there were notable subgroup differences. Compared with women, men had significantly lower odds of characterizing nicotine as “definitely harmful” to children. Tobacco product users had significantly lower odds of endorsing “definitely harmful” or “don’t know” than nonusers. African American non-Hispanic individuals, Hispanic individuals, and “other” non-Hispanic individuals had significantly lower odds of endorsing “definitely harmful” or “maybe harmful” than white individuals.

CONCLUSIONS: Although most adults perceive nicotine exposure as harmful for children, there are important differences based on sex, racial and/or ethnic background, and tobacco use status. The results reveal the need for public health efforts to better understand and target inaccurate risk perceptions among specific subgroups.

  • Abbreviations:
    adjusted odds ratio
    confidence interval
    electronic cigarette liquid
    electronic nicotine delivery system
    little cigar, cigarillo, or filtered cigar
    traditional cigar
  • What’s Known on This Subject:

    Nicotine can cause significant harm to children. The array of marketed tobacco products and reported cases of child-nicotine exposure have increased substantially in recent years. Comprehension of adults’ knowledge and perceptions of nicotine’s harmfulness to children is lacking.

    What This Study Adds:

    This study is among the first in which adults’ perceptions of nicotine’s harms to children are examined and subgroups that may be less likely to understand these risks are identified. These findings can guide future research, inform policy, and support public health educational interventions.

    The landscape of tobacco use in the United States has changed dramatically in recent years.1 Although the total consumption of combustible cigarettes declined by ∼39% from 2000 to 2015, the consumption of noncigarette combustible tobacco (eg, cigars and pipe tobacco) increased by ∼117%, and the use of smokeless tobacco products (snuff and chewing tobacco) increased by ∼23% during the same time period.1 In addition, the use of electronic nicotine delivery systems (ENDS; eg, electronic cigarettes) has increased since these products entered the US market in 2007.2,3 More than a fifth of American adults have used an ENDS to consume vaporized nicotine,4 and most electronic cigarette users also smoke combustible cigarettes (ie, dual use).47 Among Americans who have ever smoked cigarettes, ∼38% of current and 49% of recently former smokers have used an ENDS.8

    Alternative sources of nicotine consumption have come into focus in recent years as countries such as the United States and the United Kingdom have been revising public health policies targeting tobacco harm reduction.9,10 Public Health England has published guidance for employers and organizations to facilitate the use of ENDS to consume nicotine (also known as vaping) instead of smoking cigarettes as part of the United Kingdom’s national tobacco harm–reduction strategy.11 The Food and Drug Administration in the United States recently announced that it would be placing greater emphasis on reducing nicotine standards for cigarettes to nonaddictive levels in an effort to pursue a regulatory framework “that recognizes that the core problem of nicotine lies not in the drug itself but in the risk associated with the delivery mechanism.”9 The common theme is that alternative sources of nicotine are being encouraged to reduce exposure to smoked tobacco products. Indeed, the authors of recent studies indicate that many smokers are using noncigarette tobacco products such as ENDS to either aid cessation or provide a less harmful alternative to smoking,47 although the efficacy of these products for smoking cessation is unclear.1215

    Although certain modes of nicotine delivery may reduce smoking-caused harm for adults, both acute and chronic exposures to nicotine are known to cause significant harm to children from fetal development through birth, childhood, and adolescence.1620 Unintended direct nicotine exposure among children caused by handling or ingesting tobacco products such as cigarettes, cigars, nicotine electronic cigarette liquid (e-liquid) used in ENDS, nicotine gums, and other noncigarette nicotine products has been increasing steadily and significantly over the past decade in the United States,21,22 resulting in clinical sequelae ranging from mild illness to death.2123 Although it would be ideal if all adults possessed a basic understanding of nicotine’s harmfulness to children, there are likely more acute needs and information deficits among specific subgroups of the population that contribute to this problem. Research is needed to identify at-risk subgroups (eg, adults who underestimate the risks of nicotine to children) to develop targeted public health interventions to protect children from nicotine exposure.

    Research documenting adults’ risk perceptions regarding children being exposed to nicotine is limited. Although authors of other studies have examined adults’ perceptions of the risks of secondhand smoke to infants and children,24,25 there is a dearth of research on perceptions of the specific harm of nicotine (not necessarily in the context of combustible cigarettes or secondhand smoke) to children. The authors of 1 published study26 examined risk perceptions regarding nicotine exposure from ENDS e-liquids among parents of child patients in an urban pediatric practice. Although most parents believed that child exposure to nicotine e-liquids would be dangerous, many of the parents who used an ENDS did not have adequate knowledge about nicotine-specific harms to children, and e-liquid safe-storage practices were inconsistent.26

    Given the increasing array of sources of nicotine exposure (including not only combustible and smokeless tobacco products but also nicotine e-liquids, which have recently become potentially dangerous sources of child exposure), research is needed to examine adults’ perceptions of the specific harms of nicotine to children. Thus, with the current study, we sought to examine whether demographic and/or behavioral factors (eg, current tobacco use status, having a minor child in the home, level of education) are related to perceptions of nicotine harm to children. Perceptions of nicotine harm to children were assessed generally, regardless of mode of exposure. If there are specific subgroups of adults (eg, tobacco users or those with lower education) who perceive less danger of nicotine to children, this would be important for informing targeted educational interventions to protect children from nicotine exposure.



    Data were obtained from 2 waves of the Tobacco Products and Risk Perceptions Survey administered annually by the Georgia State University Tobacco Center of Regulatory Science. Data collection occurred from August to September 2015 and from August to September 2016; samples were drawn from GfK’s KnowledgePanel, a probability-based Web panel designed to be representative of noninstitutionalized US adults. Details regarding sampling and weighting procedures are documented elsewhere.27 This study included participants providing responses to all study questions (n = 11 948; 99.0% of total sample) and was approved by the Georgia State University Institutional Review Board.



    Age (18–29 years, 30–44 years, 45–59 years, and >60 years), sex, education (less than high school, high school graduate, some college, and bachelor’s degree or higher), and whether at least 1 minor (<18 years) child resided in the home were assessed.

    Tobacco Product Use

    Participants who indicated that they had not smoked at least 100 cigarettes in their life were categorized as nonsmokers. Among respondents who had smoked 100 cigarettes in their life, those who currently smoked some days or every day were characterized as current smokers, and those not currently smoking were categorized as nonsmokers. Participants that responded “no” to the following question were classified as nonusers of the particular product: “Have you ever used electronic vapor products, smoked little cigars, cigarillos, or filtered cigars (LCCs), or smoked traditional cigars (TCs), even 1 or 2 times?” Those responding yes were asked, “Do you now use electronic vapor products, smoke LCCs, or smoke TCs some days, every day, or not at all?” Those responding “some days” or “every day” were categorized as current product users, and those responding “not at all” were categorized as nonusers. A dichotomous combusted tobacco product variable in which the responses for cigarettes, LCCs, and TCs were combined was calculated to represent the current use of smoked tobacco products. For current use of other noncombustible tobacco products, participants were asked if in the last 30 days they had used any of the following: chewing tobacco, dip or snuff, snus, or dissolvable tobacco. Finally, a variable was computed to represent the total number of tobacco product types currently used (0, 1, or 2 or more, regardless of combustible or noncombustible) by each respondent.

    Perceptions of Nicotine Harm to Children

    The following question elicited perceptions of nicotine harm to children regarding children <13 years old: “Tobacco products, including electronic vapor products, contain nicotine. When used by the following groups, how harmful is nicotine in amounts usually found in tobacco products?” Response options were not harmful, unlikely harmful, maybe harmful, definitely harmful, or don’t know.

    Data Analysis

    Using SAS software, version 9.4 (SAS Institute, Inc, Cary, NC), we compared the 2 waves of data for variations in the patterns of bivariate associations between each of the explanatory variables and the outcome of interest; because no substantial differences were found, the 2 samples were combined for this analysis. First, descriptive statistics and χ2 tests were used to examine bivariate associations between demographic characteristics (age, sex, education, race and/or ethnicity, and having ≥1 minor child in the home), tobacco product use (combusted tobacco use, noncombusted tobacco use, ENDS use, and the number of different tobacco products currently used), and perceptions of nicotine harm to children. Next, we used multinomial logistic regression models to examine adjusted relationships between demographic characteristics, the number of different tobacco products currently used, and perceptions of nicotine harm to children. The decision to examine the number of tobacco products used in multivariate analyses, rather than variables representing each individual tobacco product, was based in part on multicollinearity; nearly half of cigarette smokers (44.7%; 95% confidence interval [CI]: 41.9%–47.5%) and noncombustible tobacco product users (48.5%; 95% CI: 41.7%–55.2%), and more than half of ENDS users (71.7%; 95% CI: 67.4%–76.0%), LCC smokers (84.0%; 95% CI: 80.6%–87.5%), and TC smokers (56.1%; 95% CI: 52.1%–60.0%) were current multiple tobacco product users. Furthermore, the research question related to children’s exposure to nicotine rather than particular product types.


    The majority (83.2%; 95% CI: 82.3%–84.1%) of adults characterized nicotine as definitely harmful to children, 6.2% (95% CI: 5.7%–6.8%) as maybe harmful, 1.7% (95% CI: 1.3%–2.0%) as unlikely harmful, 0.6% (95% CI: 0.4%–0.8%) as not harmful, and 8.3% (95% CI: 7.6%–9.0%) responded that they don’t know how harmful nicotine is when used by children. Approximately one-third of current tobacco product users (32.8%; 95% CI: 30.7%–34.9%) reported having at least 1 minor child living in their home.

    Bivariate associations of demographic and tobacco use variables with perceptions of nicotine harm to children are summarized in Table 1. There were significant associations between sex, race and/or ethnicity, age, education, combusted tobacco use status, noncombusted tobacco product use status, ENDS use, and number of nicotine products currently used with perceptions of nicotine harm to children. For example, African American non-Hispanic adults were significantly less likely to characterize nicotine as being definitely harmful to children, and significantly more likely to respond “don’t know,” compared with those of other racial and/or ethnic backgrounds. In addition, respondents with either less than a high school diploma, a high school diploma, or some college were significantly less likely to perceive nicotine as being definitely harmful to children compared with those with a bachelor’s degree or higher.

    TABLE 1

    US Adults’ Perceptions of Nicotine Harmfulness to Children by Demographic and Number and Type of Tobacco Products Used Characteristics

    The multinomial logistic regression model was statistically significant (F56, 11 497 = 8.85, P < .0001). The reference category for the dependent variable was set as “unlikely harmful.” With the exception of having at least 1 minor child in the home, all explanatory variables were significantly associated with nicotine risk perceptions in the overall model (sex: F4, 11 944 = 5.12, P = .0004; race and/or ethnicity: F12, 11 936 = 5.38, P < .0001; education: F12, 11 936 = 5.47, P < .0001; age category: F12, 11 936 = 1.82, P = .04; having a minor child in the home: F4, 11 944 = 1.02, P = .40; and number of nicotine products currently used: F8, 11 940 = 18.04, P < .0001). Adjusted odds ratios (aORs) for all covariates in the model are presented in Table 2. Significant intercategory differences were observed for sex, race and/or ethnicity, and number of tobacco products in the full model. Tobacco product users had significantly lower odds of endorsing “definitely harmful” or “don’t know” than nonusers, and African American non-Hispanic individuals, Hispanic individuals, and “other” non-Hispanic individuals had significantly lower odds of endorsing “definitely harmful” or “maybe harmful” than white individuals. Compared with women, men had significantly lower odds of characterizing nicotine as definitely harmful to children.

    TABLE 2

    Multinomial Logistic Regression Model in Which Associations Between Demographic Factors and Number of Tobacco Products Currently Used With US Adults’ Perceptions of Nicotine’s Harmfulness to Children <13 Years Old Are Examined (N = 11 948)


    We analyzed US adults’ perceptions of nicotine harm to children in a nationally representative sample and found that although the majority of the population perceived nicotine to be harmful when used by children, important subgroup variations exist. For example, the finding that adults who use tobacco are significantly more likely to underestimate nicotine’s harmfulness to children regardless of the number of different products used is particularly concerning. These individuals have the greatest likelihood of possessing nicotine-containing products, providing potential sources of exposure to children. Results also suggest that some men and individuals from racial and/or ethnic minority backgrounds might benefit from interventions to correct inaccurate risk perceptions.

    Authors of previously published research have demonstrated that after cigarettes, ENDS products21,22 and chewing tobacco21 have been the principal sources of unintended nicotine exposure in children since 2012. The increasing use of ENDS and other noncombusted tobacco products in the US adult population15 presents new sources of nicotine consumption and exposure not only to the users but also to the children who live with them. The food-grade flavorings used in these products include many fruit and candy flavors that are appealing to children.28 The number of cases of nicotine poisonings resulting from children ingesting e-liquids has risen sharply in recent years.21,22,29 Although there was a 21% reduction in the number of reported e-liquid exposures in the months after a federal statute for childproof packaging became effective in July 2016, several hundred such exposures continue to be reported monthly in the United States.29 These trends reveal the need for interventions targeting tobacco product users and adults who live with them to educate them on the dangers of child nicotine exposure (from e-liquids as well as other sources) and how to prevent it.

    On the basis of the current findings, such interventions might target specific subgroups. For example, members of racial and/or ethnic minority groups (compared with white adults) and adult men had significantly lower odds of characterizing nicotine as being harmful to children. In addition, adults who use tobacco products were significantly less likely to state that nicotine is harmful to children compared with adults who did not use tobacco products.

    This study and the study by Garbutt et al26 both found that the majority of adults perceived nicotine exposure among children as harmful. Other published studies in which nicotine-specific risk perceptions (ie, without a more general focus on tobacco use) were examined have often been focused on its addictiveness and/or harm to developing fetuses3032; authors of such studies generally consider inhaled rather than ingested nicotine or cutaneous exposures. Researchers have examined agricultural workers’ perceptions related to green tobacco sickness (nicotine toxicity resulting from dermal nicotine absorption that occurs during tobacco harvesting on farms33) resulting from handling tobacco plant constituents. These researchers have documented knowledge deficits related to the potential for nicotine toxicity among children working in tobacco fields and a tendency to attribute symptoms to factors other than nicotine exposure, such as heat and fatigue.3437 Thus, the current research is an important addition to the literature on perceptions about specific harms of nicotine to children. This topic is timely with the recent increase in the availability of different sources of nicotine, such as nicotine e-liquids and other noncigarette tobacco products. Awareness of these findings is particularly important for pediatricians and others providing health care to children and their families.

    This study has limitations. The primary outcome was a response to a general question of nicotine being harmful when used by children <13 years old, and although our analysis controlled for the number of tobacco products currently used, we did not assess harm perceptions specific to nicotine product type. In a similar vein, participants’ responses may or may not be applicable across different modes of nicotine exposure such as ingestion, second-hand inhalation, or absorption through the skin and mucous membranes. Garbutt et al26 found that although 73.7% of adults thought e-liquid ingestion was dangerous to children, less than half that many (31.2%) thought exposure by skin contact would be dangerous. These findings reveal that knowledge deficits vary according to routes of nicotine e-liquid exposure. Future research in which risk perceptions relative to specific sources and routes of nicotine exposure are examined would be helpful in clarifying these differences.

    These findings are limited in that the dependent variable assessed respondents’ perceptions of nicotine’s harmfulness to children <13 years old. We were unable to determine how risk perceptions might vary relative to specific child age ranges <13 years (eg, age 6 and younger, given that more than half the reported nicotine e-liquid exposures occur in this age group38). In the future, researchers should attempt to identify how adults’ perceptions of nicotine harm to children may vary depending on specific child age ranges.


    Although most adults perceive nicotine as harmful to children, there are important differences based on racial and ethnic background and tobacco use status. Given the rapid expansion of potential sources of nicotine exposure (ie, increases in use of noncigarette tobacco products as well as ENDS), public health efforts are needed to better understand and target inaccurate risk perceptions among men, tobacco users, and certain racial and ethnic minorities. Adults’ knowledge of risks associated with unintended nicotine exposures (including ingestion and absorption through skin and mucous membranes), particularly regarding children, has received sporadic attention in the scientific literature. Adding to this, ENDS are widely viewed in the context of cigarette smoking47 and (in this context) are perceived as being safe. Research focused on perceptions of the harm of children’s exposure to sources of nicotine (including cigarettes, cigarette butts, all types of cigars, hookahs, chewing tobacco, snuff, and ENDS e-liquids) and how these materials are handled and stored in the presence of children is needed to identify risk factors associated with such exposures. Such inquiries should include nicotine product users and adults who live with them to assess knowledge of safety issues related to having these products in their homes.


      • Accepted May 1, 2018.
    • Address correspondence to Catherine B. Kemp, BSN, MHA, Tobacco Center of Regulatory Science, Georgia State University School of Public Health, Urban Life Building Room 848, 140 Decatur St, Atlanta, GA 30303. E-mail: ckemp9{at}
    • FINANCIAL DISCLOSURE: Dr Eriksen receives research funding support from Pfizer, Inc (“Diffusion of Tobacco Control Fundamentals to Other Large Chinese Cities,” M.P.E., principal investigator); the other authors have indicated they have no financial relationships relevant to this article to disclose.

    • FUNDING: Supported by grant P50DA036128 from the National Institutes of Health (NIH), National Institute on Drug Abuse, and the Food and Drug Administration Center for Tobacco Products. This work is also supported by grant K23AT008442 from the NIH and National Center for Complementary and Integrative Health. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH or the Food and Drug Administration. Funded by the National Institutes of Health (NIH).

    • POTENTIAL CONFLICT OF INTEREST: The authors have indicated they have no potential conflicts of interest to disclose.