Abstract
Objectives:
We examined the prevalence of and factors associated with usually purchasing tobacco online.
Methods:
We analyzed Waves 1 (2013-14) and 4 (2016-17) of the US Population Assessment of Tobacco and Health adult (18+) and youth (12-17) studies: 15,450 adults and 495 youth in 2013-14 and 15,037 adults and 465 youth in 2016-17. Z-tests compared the prevalence of usually purchasing tobacco online between waves and weighted multivariable regressions identified associations between purchasing online and sociodemographics.
Results:
The prevalence of usually purchasing tobacco online increased from 2.5% to 3.3% among adults (p < .05) and from 2.5% to 4.4% among youth (p < .05), generalizing to a US population of 2,000,000 adults and 35,000 youth. E-cigarettes and cigars and e-cigarettes and waterpipe tobacco were the most common products among adults and youth, respectively. Men, adults with greater education, adults with higher income, and non-Hispanic black youth had greater odds of purchasing tobacco online (p < .05).
Conclusions:
Usually purchasing tobacco online remains low, although ever purchasing was not assessed. Efforts should be made to expand Internet tobacco purchasing surveillance and extend and enforce restrictions broadly across tobacco products to reduce youth access.
Keywords: tobacco access, tobacco control, tobacco prevention
Purchasing tobacco products online is easy and convenient. Consumers report purchasing tobacco online primarily because these products are less expensive.1,2 Other reasons include product variety, ease in ordering, and delivery convenience.1 The Internet market is expected to increase significantly as more consumers become aware of the broader product availability and potential to acquire products not available in their own states or at lower prices.3 Whereas these features are beneficial for adult consumers, online markets have the potential to increase youth access and use, which would be detrimental to public health. Additionally, there may be concern if adults are increasingly purchasing tobacco products online as these purchases are historically associated with increased tobacco use, higher levels of dependence, and fewer quit attempts.4,5
Online purchases of cigarettes, smokeless tobacco, and now e-cigarettes, are restricted at the federal level in the United States (US). In 2005, after several studies showed youth could easily access cigarettes online,6–8 UPS, FedEx, Visa, Mastercard, and PayPal entered into voluntary federal agreements to prohibit sale/distribution of Internet cigarettes. To reduce Internet cigarette sales further, largely due to the billions lost in tax revenue through Internet sales, the Prevent All Cigarette Trafficking (PACT) Act was passed in 2009.9 Specific provisions include prohibiting online sale of cigarettes and smokeless tobacco to anyone under the legal age of sale, requiring vendors to verify age and identity at purchase and delivery, and collecting all applicable local, state, and federal taxes. In December 2020, The PACT Act was expanded to include e-cigarettes.10 The PACT Act does have limitations, including lacking regulations for all products, including cigars and waterpipe tobacco. These gaps could be covered through state or other federal policy. For example, the Family Smoking Prevention and Tobacco Control Act instructed the US Food and Drug Administration (FDA) to adopt a regulation to control Internet sales to youth.11 However, the FDA has yet to propose such regulation.
Despite these regulations, the number of tobacco vendors selling online has proliferated.12 There are several reasons for this. For example, flavored e-cigarette products have been removed from retail stores in several states yet are available online across the US. Similarly, the age of legal tobacco sale increased to 21 at the federal level in 2019.13 These and other changes might lead more users, especially youth, to purchase tobacco online. Internet retailers commonly lack age verification strategies, allowing youth to access tobacco despite age restrictions.14,15 Previous research has found that the vast majority of Internet tobacco retailers do not meet age restriction requirements.14,16
National estimates suggest Internet sales represent 25%-30% of the market, but the exact amount is unknown.17 A handful of studies have examined the prevalence of Internet purchasing among adults and youth. A 2002 study examining any tobacco Internet purchasing found that 6.7% of people who smoked cigarettes usually purchased tobacco online.4 More recent data from a 2016 probability sample of US adults found 17.5% of e-cigarette users usually purchased products online.5 According to 2016 data from the International Tobacco Control study, 26.8% of US adults who currently used e-cigarettes made their last tobacco purchase online.18 Among youth, we see slightly lower rates; a 2015 convenience sample reported that 10% of youth had ever purchased e-cigarettes using the Internet.19 National data from the 2019 Youth Risk Behavior Survey indicate the Internet was the usual source of e-cigarette purchase for 3.6% of youth under 18.20
We extend this nascent area of research by conducting secondary data analyses of the Population Assessment of Tobacco and Health (PATH) Study, a nationally representative cohort study. To our knowledge, this is the first analysis to examine online purchase behavior for multiple products among youth and adults. This study had 2 primary aims, each conducted among the adult and youth samples separately: (1) Identify the prevalence of Internet tobacco purchases over time, and (2) Identify factors associated with purchasing tobacco on the Internet. Identifying the prevalence of adult Internet tobacco purchasing can inform regulatory actions, such as whether enforcement is needed for existing restrictions on cigarette and smokeless tobacco sales or whether new regulations are warranted for other tobacco products. Identifying the prevalence and characteristics of youth accessing tobacco online is critical to addressing youth tobacco use.
METHODS
Data Source
Data are from Waves 1 (2013-14) and 4 (2016-17) of the PATH Study Public-Use Files.21 The PATH Study is sponsored by the National Institutes of Health and the FDA’s Center for Tobacco Products and assesses the impact of tobacco use on health. The target population is US citizens 12 years and older across all 50 states and the District of Columbia. The PATH Study team used address-based sampling to identify participants in Wave 1. Wave 1 respondents were eligible for participation in subsequent waves if they were living in the US and not incarcerated. Youth were identified through their parents, and parental consent was obtained for all participating youth. Following the same methodology in Wave 1, new adults and youth were added at Wave 4 in a replenishment sample which allowed for 2 distinct cross-sectional cohort samples. Wave 1 data collection occurred between September 12, 2013 and December 14, 2014. Wave 4 data collection occurred between December 1, 2016 and January 3, 2018.
The weighted response rate for the 2013-14 adult and youth surveys were 74.0% and 78.4%, respectively. The weighted response rates for the replenishment sample in 2016-17 for adults and youth were 68.0% and 70.6%, respectively. The weighted retention rates in 2016-17 for adults and youth were 73.5% and 79.5%, respectively.21 Further details on the PATH Study are published elsewhere.22 The analytic sample for this study includes all adult and youth who reported now using any tobacco product and responded to the item about usual purchase behavior.
Measures
Dependent variable.
Those who reported current or former past 30-day tobacco product use were asked: “How do/did you usually buy your [product]?” to measure access for each of the assessed tobacco products. Response options included in person, from the Internet, by telephone, and I do not buy my own. These responses were dichotomized into one variable of usually purchasing tobacco online, where online was coded as “yes” and all other responses as “no.” Individual tobacco products assessed by PATH in Waves 1 and 4 included cigarettes, e-cigarettes, traditional cigars, cigarillos, filtered cigars, pipe, hookah (waterpipe tobacco), snus, and smokeless tobacco. In the adult survey, access to roll-your-own cigarettes was assessed separately from manufactured cigarettes. At Wave 4 for the adult and youth surveys, each of the cigar types was assessed separately for use with joints; these items were combined for analyses to match data collected in Wave 1. Due to small sample sizes, for analyses, we combined all dichotomized responses from each individual tobacco access variable into one “any tobacco” variable (no, yes). Additional analyses for adult traditional cigar and e-cigarette Internet purchasing were conducted due to sufficient samples.
Independent variables.
For the demographic variables age, sex, annual household income, education, and region we used the derived variables provided by the PATH Study Team.21 The specific variables used for all analyses are listed in Supplemental Table 5. We collapsed categories from Wave 1 to match those assessed at Wave 4 for age and education. Education and income were not included in the youth analyses due to limited sample size. We combined the derived variables for race and ethnicity to create a combination race/ethnicity variable. For days used, we summed the responses for each available individual tobacco product using the item “On how many of the past 30 days did you [smoke/use tobacco product]?” For those who used multiple products, we summed days used rather than other approaches, based on the increased risk exposure by using 2 or more products.
Data Analysis
We conducted weighted descriptive analyses to characterize the samples. We used Z-tests to compare the prevalence of purchasing tobacco products online in 2013-14 to the prevalence in 2016-17. We used weighted multivariable logistic regression analyses to examine associations between purchasing any tobacco online and demographic variables (age, sex, race and ethnicity, education, income, region), days used, and polytobacco use status. Survey weights provided by the PATH Study team were applied and compensate for variable probabilities of selection, differential nonresponse rates, possible deficiencies in the sampling frame, sampling design factors such as the stratification and sampling of primary sampling units and area segments, and the use of oversampling and nonresponse adjustment factors.21 We used SAS v 9.4 to complete analyses. Statistical significance was set at p < .05.
RESULTS
Sample Characteristics
In 2013-14, the adult sample (N = 15,450) was 59.3% male and 68.4% non-Hispanic white (Supplemental Table 1); 59.9% reported past 30-day use of one product; 13.3% of 2 products, 4.8% reported use of 3 or more products. In 2016-17, the adult sample (N = 15,037) was 59.9% male and 66.0% non-Hispanic white; 57.3% reported use of one product, 17.4% of 2 products, and 7.2% of 3 or more products.
In 2013-14, the youth sample (N = 495) was 64.3% male, 61.7% non-Hispanic white, and 18.6% Hispanic (Supplemental Table 2); 37.6% reported use of one product, 26.0% of 2 products, 14.4% of 3 products, and 10.8% of 4 or more products. In 2016-17, the youth sample (N = 465) was 58.7% male, 59.5% non-Hispanic white, and 18.4% Hispanic; 46.3% reported past 30-day use of one product, 29.7% of 3 products, 11.8% of 3 products, and 9.1% of 4 or more products.
Prevalence of Usually Purchasing Tobacco Online
Among adults who reported now using tobacco, 2.5% reported usually purchasing tobacco online in 2013-14, and 3.3% reported usually purchasing tobacco online in 2016-17 (p = .001; Table 1). The individual products most commonly purchased online were e-cigarettes (13.0% in 2013-14; 6.2% in 2016-17) and traditional cigars (7.5% in 2013-14; 6.8% in 2016-17). There was a statistically significant decrease in purchasing e-cigarettes online from 2013-14 to 2016-17 (p < .0001), whereas there was a statistically significant increase in purchasing cigarettes online from 2013-14 to 2016-17 (p < .001). Among polytobacco users, most (93.9%) reported purchasing one product type online in 2013-14, 4.4% reported purchasing 2, and 1.7% reported purchasing 3-5 different product types online. In 2016-17, most (88.9%) reported purchasing one product type online, 7.9% reported purchasing 2, and 3.3% reported purchasing 3-6 different product types online.
Table 1.
Representative Population and Weighted Prevalence for Usually Purchasing Tobacco Online, Adults
| Wave 1 -- 2013-2014 | Wave 4 -- 2016-2017 | Z-score | |||
|---|---|---|---|---|---|
| N | % (95% CI) | N | % (95% CI) | ||
| Any Tobacco | 1,515,482 | 2.6 (2.4-2.9) | 2,111,929 | 3.3 (2.9-3.7) | −3.27 (p = .001) |
|
| |||||
| Cigarettes | 29,113 | 0.1 (0.0-0.2) | 124,061 | 0.3 (0.2-0.4) | −3.59(p < .001) |
|
| |||||
| Roll-Your-Own | 49,516 | 1.2 (0.7-2.2) | 103,179 | 1.9 (1.2-2.9) | −1.83 (p = .07) |
|
| |||||
| E-Cigarettes | 778,860 | 12.8 (11.0-14.8) | 845,030 | 7.5 (6.4-8.7) | 6.13 (p < .00001) |
|
| |||||
| Traditional Cigars | 330,302 | 8.0 (6.3-10.1) | 728,457 | 7.9 (6.3-9.9) | 1.84 (p = .07) |
|
| |||||
| Cigarillos | 38,292 | 0.9 (0.5-1.6) | 110,683 | 0.9 (0.6-1.4) | −0.23 (p = .81) |
|
| |||||
| Filtered Cigars | 8492 | 0.4 (0.1-1.4) | 35,715 | 0.9 (0.4-2.1) | −1.38 (p = .17) |
|
| |||||
| Pipe | 45,151 | 3.4 (1.9-6.2) | 120,515 | 7.6 (4.9-11.6) | −1.47 (p = .14) |
|
| |||||
| Waterpipe Tobacco | 223,666 | 2.4 (1.8-3.1) | 107,940 | 2.6 (1.5-4.4) | 0.05 (p = .96) |
|
| |||||
| Snus | 9945 | 0.7 (0.2-2.4) | 25,782 | 1.3 (0.8-1.9) | −0.53 (p = .60) |
|
| |||||
| Smokeless Tobacco | 4519 | 0.1 (0.0-0.5) | 41,312 | 0.6 (0.3-1.3) | −2.48 (p = .01) |
Note.
Dissolvables were not included in this table because they were only assessed at Wave 1.
Among youth who reported current or former tobacco use, 2.5% reported usually purchasing tobacco online in 2013-14, and 4.4% reported usually purchasing tobacco online in 2016-17 (p = .02; Table 2). The products most commonly purchased online by youth were e-cigarettes (6.9% in 2013-14; 7.2% in 2016-17) and waterpipe tobacco (9.5% in 2013-14; 11.0% in 2016-17). Among polytobacco users in 2013-14, 100% reported purchasing only one of their products online; in 2016-17, 88.5% reported purchasing one product type online, and 11.5% reported 2.
Table 2.
Representative Population and Weighted Prevalence for Usually Purchasing Tobacco Online, Youth
| Wave 1 -- 2013-2014 | Wave 4 -- 2016-2017 | Z-score | |||
|---|---|---|---|---|---|
| N | % (95% CI) | N | % (95% CI) | ||
| Any Tobacco | 22,320 | 2.5 (1.3-4.6) | 34,686 | 4.4 (2.9-6.8) | −2.33 (p = .02) |
|
| |||||
| Cigarettes | 0 | -- | 4775 | 1.9 (0.6-5.9) | −2.13 (p = .03) |
|
| |||||
| E-Cigarettes | 12,287 | 6.9 (3.1-15.0) | 19,193 | 7.2 (4.2-12.0) | −0.89 (p = .37) |
|
| |||||
| Traditional Cigars | 0 | -- | 1673 | 5.3 (0.7-30.2) | −1.19 (p = .23) |
|
| |||||
| Cigarillos | 1281 | 0.4 (0.1-3.1) | 1673 | 1.7 (0.2-11.8) | −0.74 (p = .46) |
|
| |||||
| Filtered Cigars | 0 | -- | 1566 | 6.0 (0.8-33.0) | −0.84 (p = .40) |
|
| |||||
| Waterpipe Tobacco | 10,768 | 9.5 (3.9-21.2) | 2974 | 11.0 (2.4-38.7) | −0.68 (p = .50) |
|
| |||||
| Smokeless | 0 | -- | 4506 | 4.3 (1.1-16.0) | −1.77 (p = .77) |
Note.
Pipes were assessed at each wave but there were no users who reported purchasing online. Bidis and kretek were assessed at Wave 1 but there were no users who reported purchasing online.
Factors Associated with Purchasing Online
Among adults, sex, race, education, income, and polytobacco use status were associated with purchasing online in adjusted models (Table 3). Women had decreased odds than men for purchasing tobacco online in 2013-14 (Adjusted Odds Ratios [AOR] = 0.70; 95% CI = 0.54-0.90) and 2016-17 (AOR = 0.47; 95% CI = 0.35-0.63). Non-Hispanic black adults had decreased odds compared to Non-Hispanic white adults for purchasing tobacco online in 2013-14 (AOR = 0.40; 95% CI = 0.22-0.71) and 2016-17 (AOR = 0.42; 95% CI = 0.26-0.67). Those with a baccalaureate degree or higher had increased odds compared to those with less than a high school degree for purchasing tobacco online in 2013-14 (AOR = 4.53; 95% CI = 2.44-8.43) and 2016-17 (AOR = 2.47; 95% CI = 1.52-4.02). Those with a household income greater than $100,000 had increased odds compared to those with an income less than $10,000 to usually purchase tobacco online in 2013-14 (AOR = 2.45; 95% CI = 1.57-3.82) and 2016-17 (AOR = 2.27; 95% CI = 1.53-3.36). Those who reported polytobacco use (using 2 or more tobacco products in the past 30 days) had increased odds of usually purchasing online in 2013-14 (AOR = 1.53; 95% CI = 1.02-2.31) and 2016-17 (AOR = 2.49; 95% CI = 1.80-3.45). Usually purchasing tobacco online did not vary by number of days used, region, or age.
Table 3.
Factors Associated with Adults Usually Purchasing Tobacco Online
| Wave 1 -- 2013-2014 | Wave 4 -- 2016-2017 | |||
|---|---|---|---|---|
|
|
||||
| Weighted Prevalence for Purchasing Online % (95% CI) |
Adjusted Odds for Purchasing Online N = 13,892 AOR (95% CI) |
Weighted Prevalence for Purchasing Online % (95% CI) |
Adjusted Odds for Purchasing Online N = 13,381 AOR (95% CI) |
|
| Age | ||||
|
| ||||
| 18-24 | 2.6 (2.1-3.2) | Ref | 3.7 (3.0-4.5) | Ref |
| 25-34 | 2.9 (2.3-3.7) | 0.99 (0.70-1.40) | 3.5 (2.7-4.4) | 0.85 (0.60-1.22) |
| 35-44 | 2.8 (2.2-3.6) | 0.96 (0.67-1.39) | 2.5 (1.9-3.4) | 0.72 (0.46-1.13) |
| 45-54 | 2.1 (1.6-2.7) | 0.84 (0.57-1.25) | 2.4 (1.8-3.3) | 0.72 (0.47-1.12) |
| 55-64 | 3.0 (2.3-4.0) | 1.36 (0.92-2.00) | 3.3 (2.3-4.9) | 1.14 (0.71-1.84) |
| 65+ | 2.1 (1.2-3.6) | 0.91 (0.48-1.71) | 5.1 (3.6-7.1) | 1.50 (0.89-2.51) |
|
| ||||
| Sex | ||||
|
| ||||
| Male | 3.1 (2.7-3.5) | Ref | 4.1 (3.5-4.7) | Ref |
| Female | 2.0 (1.6-2.4) | 0.70 (0.54-0.90) | 2.1 (1.7-2.5) | 0.47 (0.35-0.63) |
|
| ||||
| Race/Ethnicity | ||||
|
| ||||
| NH White | 3.0 (2.7-3.5) | Ref | 3.9 (3.4-4.4) | Ref |
| NH Black | 1.0 (0.6-1.6) | 0.40 (0.22-0.71) | 1.4 (0.9-2.2) | 0.42 (0.26-0.67) |
| NH Other | 4.0 (2.7-5.8) | 1.16 (0.77-1.77) | 2.8 (1.7-4.6) | 0.64 (0.35-1.19) |
| Hispanic | 1.7 (1.1-2.4) | 0.65 (0.43-1.00) | 2.5 (1.8-3.5) | 0.81 (0.53-1.23) |
|
| ||||
| Education | ||||
|
| ||||
| Less than high school | 0.8 (0.5-1.3) | Ref | 1.8 (1.3-2.4) | Ref |
| GED | 1.5 (1.0-2.2) | 1.67 (0.86-3.25) | 1.9 (1.3-2.8) | 0.92 (0.57-1.47) |
| High school graduate | 1.7 (1.3-2.2) | 1.85 (1.01-3.39) | 2.4 (1.8-3.2) | 1.11 (0.68-1.80) |
| Some college | 3.1 (2.7-3.7) | 2.90 (1.58-5.30) | 3.5 (3.0-4.2) | 1.58 (1.00-2.48) |
| Bachelor’s degree or higher | 6.0 (5.0-7.3) | 4.53 (2.44-8.43) | 6.5 (5.2-8.2) | 2.47 (1.52-4.02) |
|
| ||||
| Income | ||||
|
| ||||
| Less than $10,000 | 1.3 (0.9-1.8) | Ref | 1.9 (1.4-2.5) | Ref |
| $10,000 to $24,999 | 1.8 (1.4-2.3) | 1.18 (0.75-1.87) | 2.5 (1.9-3.3) | 1.05 (0.68-1.61) |
| $25,000 to $49,999 | 2.4 (1.9-3.1) | 1.43 (0.93-2.20) | 2.4 (1.8-3.1) | 0.92 (0.55-1.54) |
| $50,000 to $99,999 | 4.0 (3.3-4.7) | 1.92 (1.28-2.87) | 4.2 (3.3-5.3) | 1.44 (0.93-2.21) |
| $100,000 or more | 6.1 (4.6-8.0) | 2.45 (1.57-3.82) | 7.4 (5.9-9.1) | 2.27 (1.53-3.36) |
|
| ||||
| Region | ||||
|
| ||||
| Northeast | 2.7 (1.9-3.7) | Ref | 3.7 (2.7-5.0) | Ref |
| Midwest | 2.5 (2.1-3.1) | 1.02 (0.68-1.52) | 2.8 (2.1-3.6) | 0.76 (0.46-1.26) |
| South | 2.5 (2.1-3.1) | 1.10 (0.73-1.66) | 3.2 (2.7-3.9) | 0.97 (0.64-1.47) |
| West | 3.0 (2.4-3.6) | 1.13 (0.72-1.75) | 3.5 (2.9-4.3) | 0.98 (0.63-1.54) |
|
| ||||
| Days Used | ||||
|
| ||||
| 0-9 Days | 2.4 (2.1-2.8) | Ref | 2.8 (2.3-3.3) | Ref |
| 10-30 Days | 3.5 (3.0-4.2) | 1.23 (0.92-1.63) | 3.2 (2.6-4.0) | 0.98 (0.69-1.40) |
|
| ||||
| Polytobacco Use | ||||
|
| ||||
| No | 2.4 (2.1-2.8) | Ref | 2.9 (2.5-3.4) | Ref |
| Yes | 4.4 (3.3-5.8) | 1.53 (1.02-2.31) | 5.5 (4.5-6.8) | 2.49 (1.80-3.45) |
Note.
Bold indicates p < .05.
Factors associated with purchasing e-cigarettes online varied by year (Supplemental Table 3). In 2013-14, those with greater education and higher income had increased odds to purchase online. In 2016-17, those 65 and older had increased odds compared to those 18-24 to purchase online (AOR = 2.68; 95% CI = 1.22-5.88), and only those with a baccalaureate degree or higher had increased odds compared to those with less than a high school degree to purchase online (AOR = 2.30; 95% CI = 1.05-5.04). Those living in the West region had decreased odds to purchase e-cigarettes online than those in the Northeast in 2013-14 (AOR = 0.46; 95% CI = 0.25-0.84) and 2016-17 (AOR = 0.52; 95% CI = 0.30-0.91).
Age and income were associated with purchasing traditional cigars online (Supplemental Table 4). Adults in each of the 3 oldest age groups (45-54, 55-64, 65+) had increased odds compared to those 18-24 years to purchase traditional cigars online in 2013-14 and 2016-17. Those in the 3 highest income groups had increased odds compared to those with an annual household income less than $10,000 to purchase online in 2013-14.
Among youth, sex, race, and number of days used were associated with purchasing tobacco online in 2016-17. Women had decreased odds compared to men for purchasing tobacco online in 2016-17 (AOR = 0.23; 95% CI = 0.09-0.63). Non-Hispanic black youth had increased odds compared to Non-Hispanic white youth for purchasing tobacco online in 2016-17 (AOR = 2.85; 95% CI = 1.01-8.05). In 2016-17, youth who reported using tobacco 10-30 days had decreased odds compared to those who reported tobacco use 0-9 days for usually purchasing tobacco online (AOR = 0.20; 95% CI = 0.08-0.51).
Discussion
Less than 5% of persons who report current or former tobacco use usually purchase their tobacco products online. The prevalence of usually purchasing tobacco online appears to be increasing; we identified increases in usually purchasing tobacco online among both adults and youth from 2013-14 to 2016-17. Importantly, the number of people who purchase tobacco online may be even higher overall, as PATH only measures usual purchase behavior.
E-cigarettes followed by traditional cigars were the products most usually purchased by adults, while waterpipe tobacco then e-cigarettes were the products most likely to be purchased online by youth. Waterpipe tobacco is less often available than other products in brick-and-mortar retail locations, often limited to specialty tobacco shops or cafes. This may be driving more youth to attempt purchasing waterpipe tobacco online. Similarly, the increased prevalence of adults purchasing traditional cigars online likely reflects the greater availability of products, as opposed to what is available in traditional tobacco retailers. Purchasing traditional cigars online appears to be more heavily concentrated among older adults, though it is unclear whether income also plays a role. The massive increase in online e-cigarette vendors during this time12 likely accounts for the popularity of purchasing e-cigarettes online relative to other tobacco products. Among adults only, we observed a significant decrease in the percentage of people usually purchasing e-cigarettes online from 2013-14 to 2016-17, which may reflect an additional increase in vape shops and retail access as e-cigarettes increased in popularity. However, purchasing of e-cigarettes among youth remained uniform throughout the 4-year study period. With changes in alternative tobacco product regulations, monitoring online e-cigarette purchases in particular will be important, especially if mostly youth, rather than adults, are accessing these products through that channel. Although purchasing e-cigarettes online in 2013-14 appeared more common among those with higher income and education, shifts in 2016-17 suggest the practice is more broadly across demographic groups. The finding that those in the West are less likely than those in the Northeast to purchase online may reflect more widespread vape shop prevalence in western states. These findings will be important to monitor with continuing policy changes.
During the study period, cigarettes were the most consumed tobacco product among adults. However, purchasing cigarettes and smokeless tobacco online is heavily restricted by federal and state laws. Thus, despite the popularity of cigarettes, usual online purchases remain very low (0.1%-0.3%). We observed a significant increase in the prevalence of usually purchasing cigarettes online among both adults and youth between 2013-14 and 2016-17. Importantly, in the beginning of this study, no youth reported usually purchasing cigarettes online, which changed in later years. Considering the PATH data only capture usual purchasing, the total amount could be higher among adults and youth. As such, it will be important to continue monitoring the trends in online cigarette purchasing to ensure policies are continually enforced.
Female adults and youth were less likely than their male counterparts to report usually purchasing tobacco online. Non-Hispanic black adults were less likely than non-Hispanic white adults to report purchasing products online. We saw the opposite pattern among youth, with black youth more likely than white youth to report purchasing online. We are unsure why these differences exist, but suspect it may be tied to product preferences. For example, non-Hispanic white adults are more likely than non-Hispanic black adults to use traditional cigars and e-cigarettes, 2 products more commonly purchased online.
Among adults, both education level and income were associated with purchasing tobacco online, with those with higher education and income more likely to purchase products online. At first, this finding appears counterintuitive to literature that indicates tobacco products are less expensive online.2,23 However, it is possible the hidden costs of having Internet access, paying for shipping, and buying in bulk upfront may reduce the accessibility of purchasing online. Additionally, prices may be increased for certain consumers if purchasing within other states. Future research examining reasons for purchasing online might offer explanations for demographic differences in purchasing behavior.
Regarding tobacco use behaviors, we found frequency of use and number of products used were associated with purchasing online. Among adults, those who reported using 2 or more tobacco products were more likely to report purchasing online. In analyses among polytobacco users, we found that most only purchased one product online rather than all products used, suggesting there are specific reasons for purchasing some products rather than all products online. Among youth, days used was associated with purchasing online, such that youth who used more days per month were less likely to purchase online. Online purchasing may not be the quickest or best source for underage youth who are frequent tobacco users. This finding, and the lack of association between use and purchasing among adults contrast previous literature that found that those who purchase cigarettes online smoke more often.4
Notable legislation and events occurred after the study period which will be important to consider for future studies on the prevalence of online tobacco purchasing. In December 2019, the federal minimal legal age for tobacco sale was increased to 21. Because online age verification practices are traditionally inadequate to prevent youth access,12 those under 21 may switch to online purchases as brick-and-mortar stores begin enforcing the new minimum age of sale. In March 2020, COVID-19 led to brick-and-mortar retail closures which then led to increases in purchasing consumer goods online. We anticipate increases in online tobacco purchases during this time period. Lastly, in December 2020, e-cigarettes were added to the PACT Act. This regulation is likely to lead to decreased online purchases of e-cigarettes due to restrictions on shipping. The enforcement of this new regulation will be particularly important to monitor given the higher prevalence of e-cigarette purchases online.
There are several limitations to these findings. First, data are self-reported and subject to recall and response biases. Second, the purchase item asks about “usual” purchase behavior, not whether consumers ever purchase through this method. This is difficult to interpret, particularly among less frequent tobacco users. Accessing tobacco through the Internet is likely much higher than what is reported, as this measure captures more frequent online purchases. Additionally, the PATH study includes ‘telephone’ as a response option for usual purchases, which could be misinterpreted as purchases could be made through a smart phone device, and thus, the Internet. These factors likely contribute to the differences in estimates reported by the present study and other national surveys.18,19 Lastly, the days used item was not asked of each individual tobacco product. The total days used may represent a lower total due to not capturing use of some products (eg, waterpipe tobacco).
IMPLICATIONS FOR TOBACCO REGULATION
We identified increases in purchasing tobacco online among both youth and adults. The overall prevalence of usually purchasing tobacco online remains low, yet any purchase of tobacco online may be higher. Unfortunately, access measures do not currently assess any online tobacco purchasing which likely underestimates total prevalence.18,19 Future surveillance efforts should consider assessing both usual and any purchase behavior to provide greater understanding of how youth and adults access tobacco. Regardless, estimates from the present study indicate over 2 million adults and nearly 35,000 youth usually purchased tobacco through the Internet in 2016-17. Existing restrictions limiting cigarette and smokeless tobacco sales online appear effective as rates for purchasing these products were much lower than products such as e-cigarettes, cigars, and waterpipe tobacco. Efforts should be made to extend provisions more broadly across tobacco products to reduce youth access. Additionally, because online cigarette purchases increased among both youth and adults during the study period, renewed enforcement efforts may be needed.
Supplementary Material
Table 4.
Factors Associated with Youth Usually Purchasing Tobacco Online
| Wave 1 -- 2013-2014 | Wave 4 -- 2016-2017 | |||
|---|---|---|---|---|
| Weighted Prevalence for Purchasing Online % (95% CI) |
Adjusted Odds for Purchasing Online N = 487 AOR (95% CI) |
Weighted Prevalence for Purchasing Online % (95% CI) |
Adjusted Odds for Purchasing Online N = 442 AOR (95% CI) |
|
| Age | ||||
|
| ||||
| 12-14 | 6.3 (2.0-18.0)a | Ref | 0 | Ref |
| 15-17 | 2.2 (1.0-4.4) | 0.34 (0.07-1.72) | 4.9 (3.2-7.4) | 7.15x1010 (436x1010−1.18x1011) |
|
| ||||
| Sex | ||||
|
| ||||
| Male | 2.8 (1.3-5.7) | Ref | 6.2 (3.9-10.0) | Ref |
| Female | 2.1 (0.6-6.5)a | 0.68 (0.16-2.84) | 2.0 (0.8-4.9) | 0.23 (0.09-0.63) |
|
| ||||
| Race/Ethnicity | ||||
|
| ||||
| NH White | 2.7 (1.2-6.0) | Ref | 3.1 (1.5-6.2) | Ref |
| NH Black | 4.8 (1.1-18.3)a | 1.22 (0.22-6.89) | 9.6 (4.5-19.2) | 2.85 (1.01-8.05) |
| NH Other | 1.9 (0.2-13.3)a | 0.52 (0.05-5.07) | 7.7 (2.6-20.6)a | 3.19 (0.80-12.75) |
| Hispanic | 0.8 (0.1-5.3)a | 0.14 (0.02-1.10) | 3.4 (1.3-9.1)a | 0.97 (.44-2.14) |
|
| ||||
| Region | ||||
|
| ||||
| Northeast | 2.9 (0.7-10.8)a | Ref | 3.3 (1.1-9.4)a | Ref |
| Midwest | 0.6 (0.1-3.8)a | 0.18 (0.01-2.30) | 4.0 (1.6-9.9 | 0.78 (0.19-3.18) |
| South | 2.6 (0.9-7.4)a | 0.92 (0.14-5.98) | 5.8 (3.2-10.4) | 1.22 (0.40-3.70) |
| West | 4.3 (1.6-11.2)a | 1.84 (0.28-12.19) | 3.9 (1.3-10.9)a | 1.02 (0.31-3.36) |
|
| ||||
| Days Used | ||||
|
| ||||
| 0-9 Days | 4.1 (1.6-9.7) | Ref | 8.2 (5.1-12.7) | Ref |
| 10-30 Days | 1.8 (0.8-4.3) | 0.42 (0.14-1.120) | 1.8 (0.8-4.3) | 0.20 (0.08-0.51) |
|
| ||||
| Polytobacco Use | ||||
|
| ||||
| No | 2.8 (1.2-6.3) | Ref | 4.9 (2.8-8.6) | Ref |
| Yes | 2.3 (0.9-5.6) | 0.97 (0.24-3.85) | 3.9 (2.1-7.3) | 1.00 (0.37-2.70) |
Note.
Bold indicates p < .05
indicates cell size < 5, results should be interpreted with caution.
Acknowledgements
Research reported in this publication was supported by NCI and FDA Center for Tobacco Products (CTP), grant number K01CA253235. 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.
Footnotes
Human Subjects Approval Statement
Westat’s Institutional Review Board approved the study design and protocol and the Office of Management and Budget approved the data collection.
Conflict of Interest Disclosure Statement
The authors have no conflicts of interest to report.
Contributor Information
Jessica L. King, Department of Health & Kinesiology, University of Utah, Salt Lake City, UT, United States.
Julie W. Merten, Brooks College of Health, University of North Florida, Jacksonville, FL, United States.
Nicole E. Nicksic, Department of Health Behavior and Policy, Virginia Commonwealth University, Richmond, VA, United States.
References
- 1.Kim AE, Ribisl KM, Delnevo CD, Hrywna M. Smokers’ beliefs and attitudes about purchasing cigarettes on the Internet. Public Health Rep. 2006;121(5):594–602. doi: 10.1177/003335490612100515 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Opazo Breton M, Britton J, Huang Y, Bogdanovica I. Comparison between online and offline price of tobacco products using novel datasets. Int J Environ Res Public Health. 2018;15(10). doi: 10.3390/ijerph15102282 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Cision PR Newswire. Global Vapor Products Market Out-look and Forecast 2018-2023: Market and is Expected to Cross $43 Billion by 2023. https://www.prnewswire.com/news-releases/global-vapor-products-market-out-look-and-forecast-2018-2023-market-and-is-expected-to-cross-43-billion-by-2023-300589389.html. Published January 29, 2018. Accessed September 29, 2020.
- 4.Hrywna M, Delnevo CD, Staniewska D. Prevalence and correlates of internet cigarette purchasing among adult smokers in New Jersey. Tob Control. 2004;13(3):296–300. doi: 10.1136/tc.2003.006783 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Hsu G, Gamst AC, Zhuang Y-L, et al. A comparison of e-cigarette use patterns and smoking cessation behavior among vapers by primary place of purchase. Int J Environ Res Public Health. 2019;16(5). doi: 10.3390/ijerph16050724 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Ribisl KM, Williams RS, Kim AE. Internet sales of cigarettes to minors. JAMA. 2003;290(10):1356–1359. doi: 10.1001/jama.290.10.1356 [DOI] [PubMed] [Google Scholar]
- 7.Ribisl KM, Kim AE, Williams RS. Are the sales practices of Internet cigarette vendors good enough to prevent sales to minors? Am J Public Health. 2002;92(6):940–941. doi: 10.2105/AJPH.92.6.940 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Ribisl KM, Kim AE, Williams RS. Web sites selling cigarettes: how many are there in the USA and what are their sales practices? Tob Control. 2001;10(4):352–359. doi: 10.1136/tc.10.4.352 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.US Department of Commerce and Trade. United States Code, 2010 Edition. U.S.C. Title 15 - Chapter 10A – Collection of States Cigarette Taxes. https://www.gov-info.gov/content/pkg/USCODE-2010-title15/html/USCODE-2010-title15-chap10A.htm. Published March 31, 2010. Accessed March 7, 2021.
- 10.JD_Supra.com. Congress Passes Legislation Sweeping ENDS into PACT Act. https://www.jdsupra.com/legal-news/congress-passes-legislation-sweeping-83179/. Accessed January 11, 2021.
- 11.US Food and Drug Administration, HHS. Family Smoking Prevention and Tobacco Control Act. 2009. http//www.govinfo.gov/content/pkg/PLAW-111publ31/pdf/PLAW-111publ31.pdf. Accessed October 8, 2019.
- 12.Williams RS, Derrick J, Liebman AK, et al. Content analysis of age verification, purchase and delivery methods of internet e-cigarette vendors, 2013 and 2014. Tob Control. 2018;27(3):287–293. doi: 10.1136/tobaccocontrol-2016-053616 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.US Food and Drug Administration, Center for Tobacco Products. Newly Signed Legislation Raises Federal Minimum Age of Sale of Tobacco Products to 21. https://www.fda.gov/tobacco-products/ctp-newsroom/newly-signed-legislation-raises-federal-minimum-age-sale-tobacco-products-21. Published July 15, 2020. Accessed January 11, 2021.
- 14.Williams RS, Derrick J, Ribisl KM. Electronic cigarette sales to minors via the Internet. JAMA Pediatr. 2015;169(3):e1563–e1563. doi: 10.1001/jamapediatrics.2015.63 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Grana RA, Ling PM. Smoking revolution. Am J Prev Med. 2014;46(4):395–403. doi: 10.1016/j.amepre.2013.12.010 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Williams RS, Derrick J, Phillips KJ. Cigarette sales to minors via the internet: how the story has changed in the wake of federal regulation. Tob Control. 2017;26(4):415–420. doi: 10.1136/tobaccocontrol-2015-052844 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Herzog B Tobacco Talk: Independent Vapor Mfr. Survey. Wells Fargo Securities; 2015. https:www.ecigarette-politics.com/files/4g14-weils-fargo.pdf. Accessed March 31, 2021. [Google Scholar]
- 18.Braak DC, Cummings KM, Nahhas GJ, et al. Where do vapers buy their vaping supplies? Findings from the International Tobacco Control (ITC) 4 Country Smoking and Vaping Survey. Int J Environ Res Public Health. 2019;16(3):338. doi: 10.3390/ijerph16030338 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Kong G, Morean ME, Cavallo DA, et al. Sources of electronic cigarette acquisition among adolescents in Connecticut. Tob Regul Sci. 2017;3(1):10–16. doi:10.18001/TRS.3.1.2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Creamer MR. Tobacco product use among high school students – Youth Risk Behavior Survey, United States, 2019. MMWR Suppl. 2020;69(1):56. doi:10.15585/mmwr.su6901a7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.National Addiction & HIV Archive Program. Population Assessment of Tobacco and Health (PATH) Public-Use Files User Guide. https://www.icpsr.umich.edu/icpsrweb/NAHDAP/studies/36231. Published June 1, 2017. Accessed December 1, 2017.
- 22.Hyland A, Ambrose BK, Conway KP, et al. Design and methods of the Population Assessment of Tobacco and Health (PATH) Study. Tob Control. 2017;26(4):371–378. doi: 10.1136/tobaccocontrol-2016-052934 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Hall MG, Williams RS, Gammon DG, Ribisl KM. Internet cigarette vendors make tax-free claims and sell cigarettes cheaper than retail outlets. Tob Control. 2016;25(6):616–618. doi: 10.1136/tobaccocontrol-2015-052359 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.