• We surveyed hookah & ENDS users exploring Internet use, attitudes and social norms.
  • Internet & social media are used to obtain and share information on these products.
  • Online-behavior related to tobacco was associated with low perceived vulnerability.
  • Photos on social media enforce social acceptability and peer pressure to use hookah.
  • Social media may be a good medium for interventions on alternative tobacco products.



The purpose of this study was to explore the relationship between social norms and attitudes towards ENDS and hookah and use of these products.


We conducted surveys with hookah and ENDS users who regularly used the Internet and social media and analyzed the primary social media account (e.g. Facebook) of each participant, coding all references to nicotine or tobacco products. The survey included domains on perceived favorability, perceived vulnerability and subjective norms.


We surveyed 21 ENDS users and 20 hookah users. Both groups used the Internet to look up information about their respective tobacco product (95% for hookah vs. 90% for ENDS). Seventy percent of hookah users had references to hookah on their social media profiles while 43% of ENDS users had references to ENDS on their page. The majority of both groups were exposed to content posted by friends in their social media network about their respective products online. Those who posted on social media about hookah and those who read about ENDS online had lower perceived vulnerability to the health risks associated with tobacco products.


Hookah and ENDS users actively use the Internet and social media to obtain and share information about nicotine/tobacco products. Study participants who use hookah were more likely to share photos and discuss hookah related activities via social media than those who use ENDS. Social networks also represent valuable and untapped potential resources for communicating with this group about risks and harm reduction related to emerging nicotine/tobacco products.


ENDS, Electronic nicotine delivery system


Alternative tobacco products;Social media;Online behavior;Prevention;Behavioral;Public health

1. Introduction

Emerging tobacco products such as hookah and electronic nicotine delivery systems (ENDS) are increasingly popular in the United States. Estimates of hookah use range from 15–41% for lifetime use, 12–30% for past-year use, and 7–21% for past-month use in the U.S. and Europe (Grekin & Ayna, 2012). From 2010 to 2013, awareness of ENDS among US adults increased from 40% to 86% with self-reported use rising from 3.4% to 15% (Emery et al., 2014 July 1 ;  Pearson et al., 2012/09/01).

Internet searches on hookah and ENDS are on the rise (Ayers et al., 2011 Apr; Salloum et al., 2014 ;  Yamin et al., 2010). Yet, little is known about how users seek and apply information on alternative tobacco products, or the degree of influence that online interactions have on use of these products. Emery et al. found that time spent online and use of social media were associated with awareness of ENDS and searching for ENDS information (Emery et al., 2014).

Social media represent an important forum for the exchange of information as it provides users the capacity to virtually interact with others by sharing and discussing text, photo, video or other multimedia-based content. Social media websites are extremely popular; 75% of those aged 18–29, 50% of those aged 30–45 and 30% of those aged 46–64 report creating a social media profile (Pew Research, 2010). In 2013, users in the US spent 16 min of every hour online on social networking websites (Experian Marketing Services, 2013).

Social media websites provide a daily bulletin of attitudes and behaviors of people in ones social network. Since having at least one friend perceived to be a current smoker is a predictor of initiation (Kandel, Kiros, Schaffran, & Hu, 2004), peer use in online networks may be a powerful influence on experimentation (Freeman & Chapman, 2010). The purpose of this study was to explore the relationship between social norms and attitudes towards ENDS and hookah and use of these products.

2. Methods

We conducted surveys with hookah and ENDS users who regularly used the Internet and social media and analyzed the primary social media account of each participant, coding all references to nicotine or tobacco products.

2.1. Recruitment and participants

We distributed print flyers on The City University of New York and New York University campuses, at/near hookah bars in Manhattan and an ENDS store in Queens, and posted flyers on Craigslist.org and e-cigaretteforum.com. The eligible participants: 1) were ≥ 18 years of age; 2) currently used hookah or ENDS (≥ 2 × in the past 30 days); 3) actively used social media websites (Facebook, Instagram, Twitter or Google +) in the past 2 days; 4) used the Internet > 1 h/day; and 5) spoke English. The eligible participants were invited for a 60–90 minute interview for which they were compensated $50. At the beginning of the interview, all participants provided written informed consent. This study was approved by the Institutional Review Board at the NYU School of Medicine.

2.2. Data collection and measures

Participants also completed a brief survey guided by the Prototype/Willingness Model (PWM). The PWM is a dual-processing model that recognizes two paths to risk behavior: an analytically-driven reasoned action pathway, and a social reaction pathway that relies more on heuristic processing (Gerrard et al., 2008; Hukkelberg and Dykstra, 2009 Mar; Litt and Stock, 2011 Dec; Rivis et al., 2011 ;  Stock et al., 2013). To assess the constructs of the PWM, we included several items used by Gerrard et al. (2006) including: 1) perceived favorability of the typical person their age and gender who uses the given tobacco product; 2) attitudes/perceived vulnerability to the negative consequences of using each type of product (e.g. ‘if you were to smoke hookah, what are the chances that you would get lung cancer’); and 3) subjective normative perceptions of use of each product (e.g. ‘I feel under pressure from friends to smoke hookah’). Each of these were measured on a Likert-scale with the ‘neutral’ option receiving a score of 0: 1) perceived favorability (7 items with a 5-point scale, max score = + 14, indicating the highest favorability); 2) perceived vulnerability (3 items with a 5-point scale, max score = + 6, indicating answers of ‘very likely’ to get lung cancer/other cancer/heart disease); 3) social norms (4 items with a 7-point scale, max score = + 12, indicating the highest perceived social pressure to use the respective products). For categorical analyses, we dichotomized the measures for perceived vulnerability and social norms into high (> 0) vs. low (≤ 0). We adapted questions from the California Tobacco Survey (University of San Diego, 1999) on tobacco use history, peer/parental use, advertising exposure, and plans to quit.

2.3. Social media profile coding

We had participants log in to their most frequently used social media accounts and a trained research assistant used a standardized worksheet to review each participants' homepage and profile, specifically looking for content related to nicotine/tobacco products. Moreno et al. used this strategy to characterize exposure to alcohol content and peer influence online (Moreno, Grant, Kacvinsky, Egan, & Fleming, 2012). We reviewed text, photographs, and groups/pages that the participant followed. We created a codebook informed by the PWM that was used to define key terms and images related to or referencing tobacco use (e.g., photos showing consumption or display of tobacco). For each item containing N/T content, we documented a description of the content, the number of likes (perceived favorability), and the number of comments (social norms).

2.4. Analysis

We calculated descriptive statistics, including means and standard deviations where applicable. We used t-tests to evaluate continuous variables and Fishers Exact Test to evaluate categorical data. For the purpose of this analysis, the results of the social media coding were simplified into dichotomous outcomes (yes/no), based on whether the coder found references to any nicotine/tobacco and references to hookah and ENDS, specifically.

3. Results

The mean age among hookah smokers (N = 20) was 26 and 65% were male, while among the ENDS users (N = 21) the mean age was 36 and 62% were male (Table 1). Among all participants, 41% used both hookah or ENDS and cigarettes, 34% used both hookah and ENDS and 15% used all three. However, among the participants who had used the other alternative tobacco product, all reported infrequent use (less than monthly for hookah and less than weekly for ENDS). Thus, for the analyses we based the two groups on the alternative product that participants used more frequently and on a regular basis. Both groups used the Internet to look up information about their respective tobacco product (95% for hookah vs. 90% for ENDS). Seventy percent of hookah users had references to hookah on their social media profiles while 43% of ENDS users had references to ENDS on their social media page. While the majority of both groups were exposed to content posted by friends in their social media network about their respective products online — hookah users were more likely to see this content compared with ENDS users (90% vs. 52%, p = 0.02) (Table 2). Both groups communicated with others about N/T products online (45% of hookah users, 57% for ENDS users). Sixty-five percent of ENDS users purchased ENDS products over the Internet. Participants who reported having seen N/T related content on a social media site were significantly more likely to have tried a N/T product because of something that they read or saw on the Internet (54% vs. 17%, p = 0.03). Hookah users had more friends who used hookah (50% vs. 4%) while ENDS users had more friends who use ENDS products (19% vs. 3%) (Appendix A — Table A.1).

Table 1. Demographics and characteristics.
ENDS users (n = 21) Hookah users (n = 20)
Age (years, mean ± SD) 36 ± 12 26 ± 6
    18–22 5% 40%
    23–29 38% 30%
    30–39 19% 30%
    40 + 38% 0%
    White 48% 35%
    Black 43% 40%
    Other 10% 25%
Hispanic ethnicity 10% 45%
Male 62% 65%
    High school 24% 5%
    At least some college 52% 65%
    Graduate degree 24% 15%
Current student
    No 76% 55%
    Undergrad 5% 35%
    Graduate 19% 10%
Married or living w. partner 33% 30%
Foreign-born 14% 20%
Cigarette use
    Never cigarette smoker 5% 30%
    Former cigarette smoker 52% 30%
    Current cigarette smoker 43% 40%
Plans to quit cigarettes?
    Yes, within the next 30 days 22% 13%
    Yes, within the next 6 months 56% 38%
    No, not thinking of quitting 22% 50%
Plans to quit alt. tobacco product
    Yes, within the next 30 days 5% 0%
    Yes, within the next 6 months 52% 11%
    No, not thinking of quitting 43% 89%
Dual and poly-use
    Uses the other alternative tobacco product (e.g. hookah or ENDS) 38% 30%
    Currently uses cigarettes, hookah & ENDS 10% 25%
    Former cigarette smoker who uses hookah & ENDS 20% 0%
    Never cigarette smoker who uses hookah & ENDS 5% 5%

Table 2. Internet and social media behaviors relating to N/T products.
ENDS users (n = 21) Hookah users (n = 20) p-Value
Used the Internet to look up info about N/T products 90% 95% 1.0
Seen references to ENDS or hookah on social media site/app 52% 90% 0.02
References to any N/T on social media profile 52% 75% 0.20
References to ENDS or hookah on social media profile 43% 70% 0.12
Conversed about N/T products online 57% 45% 0.75
Tried a N/T product because of something read on the Internet 43% 40% 1.0
Read about anothers N/T quit attempt on social media 62% 70% 0.74
Supported anothers N/T quit attempt through social media 43% 45% 1.0
Interested in a social media based quit intervention 76% 75% 1.0

Bold values indicate significance at *p < 0.05.

3.1. Perceived favorability

Participants generally perceived users of their own product more favorably. While ENDS users had a neutral favorability score towards hookah users (− 0.4 ± 5.5) and a very positive favorability towards other ENDS users (4.6 ± 4.6), hookah users had positive perceived favorability towards both groups (3.9 ± 4.3 for hookah; 1.5 ± 4.2 for ENDS). ENDS users had slightly positive favorability (1.33 ± 5.2) while hookah users had neutral favorability (− 0.1 ± 3.5) towards cigarette users (Appendix A — Table A.1).

3.2. Perceived vulnerability

Both groups perceived high vulnerability to tobacco-related diseases if they were to smoke cigarettes (overall mean = + 2.6). Mean scores for perceived vulnerability to the risks of hookah were concentrated around 0 (unsure) for both hookah and ENDS users. ENDS users perceived themselves to be at low risk for tobacco-related diseases based on their use of ENDS (− 3.5 ± 2.7), while hookah users were unsure about their perceived vulnerability were they to use an ENDS (Appendix A — Table A.1).

Those who self-reported posting hookah-related content on social media were significantly more likely to perceive low vulnerability to tobacco related diseases associated with hookah use than those who have not posted photos (p = 0.04). This finding was consistent with the social media profile coding; 67% of those who had references to hookah on their profile perceived low vulnerability to health risks. Ninety percent of ENDS users with references to ENDS on their profile had low perceived vulnerability for diseases caused by ENDS use. All participants who tried ENDS because of something that they read on the Internet (n = 9) perceived low vulnerability.

3.3. Social norms

Overall, the mean subjective norm scores were negative for both groups across all three forms of N/T, indicating that participants generally did not feel pressure from others to use N/T products. On average, hookah users reported higher social pressure to use hookah than did ENDS users (− 4.3 ± 4.1 vs. − 8.1 ± 5.8, p = 0.04). The same pattern was observed for pressure to use ENDS with ENDS users reporting higher social pressure than hookah users (− 2.6 ± 5.2 vs − 7.5 ± 5.0, p = 0.02). When dichotomized into high vs. low subjective norm scores, all hookah participants who reported high social pressure to smoke hookah had posted photos of themselves smoking hookah on social media.

3.4. Social media & quitting tobacco

Two-thirds of participants had read about and 44% had supported anothers tobacco quit attempt (e.g. liking the post or writing a comment) on a social media site (Table 2). Three-quarters of participants stated that they were interested in a social-media based quit intervention (Table 2); however only 57% of ENDS users and 11% of hookah users were planning to quit their respective products within the next 6 months (Table 1).

4. Discussion

Hookah and ENDS users actively use the Internet and social media to obtain and share information about hookah and ENDS. Based on the social media analysis and the interviews, the study participants who use hookah were more likely to share photos and discuss hookah related activities via social media than those who use ENDS.

Hookah-related social media behavior was associated with low perceived vulnerability to tobacco related diseases, as well as high influence of social pressures. ENDS users strongly believed themselves to be at low risk for tobacco related diseases, a belief even stronger among those who tried an ENDS-product because of something that they had read on the Internet. Since those who posted hookah photos on social media believed themselves to have significantly lower vulnerability to disease than those who didn't, this suggests that hookah users may be influenced by the social norms enforced by photos on social media. Our data support the assertion that both hookah and ENDS users may be easily persuaded by information found on the Internet.

Users of hookah and ENDS perceived other users of these products favorably and reported social pressure to use these products. Indeed a majority of participants' 10 closest friends used tobacco products. Given the social context of hookah use, in particular, this may promote continued use. Furthermore, both groups, and especially hookah users, are continually exposed to risk images depicting product use via social media. Relating our findings back to the Model, there appears to be a cyclic pattern through which peer or prototype behavior exemplified by pro-tobacco content on the Internet contributes to willingness to try and continue using alternative tobacco products.

Our results build on the findings of Emery et al. that 76% of ENDS users had searched for ENDS information online, 23% had searched specifically on Facebook, and 49% had shared ENDS information on Facebook (Emery et al., 2014). In a longitudinal cohort study, positive attitudes and normative beliefs around the acceptability of hookah smoking were associated with increased odds of initiation (Sidani, Shensa, Barnett, Cook, & Primack, 2014). This suggests that exposure to photographs of friends or acquaintances smoking hookah may contribute to initiation as well. Thus, it appears that social media can play an important role in exposure to alternative tobacco products.

This study had several limitations. First, given that this was a pilot study, we had a small sample size. Second, we recruited a convenience sample through free online advertisements and flyers in New York City, and the findings may not be representative of all hookah and ENDS users. Third, the data were cross-sectional and results were based on self-report, aside from the social media profile coding, which may limit the conclusions.

Since the Family Smoking Prevention and Tobacco Control Act reduces the tobacco industrys ability to use traditional marketing to reach potential smokers (U.S. Department of Health and Human Services, 2012), it is important to understand how people share, perceive and discuss tobacco-related content online in order to identify effective anti-tobacco messages and channels. Importantly, the study participants noted high interest in a social media-based quit intervention — this is an avenue that should be pursued further. The Internet may be a double-edged sword for hookah and ENDS users. Our data suggests that those who actively engage in nicotine/tobacco-related social media behavior are less likely to believe that their N/T product of choice makes them vulnerable to disease. However, social networks also represent valuable and untapped potential resources for communicating with this group about risks and harm reduction related to emerging N/T products.

The following is the supplementary data related to this article.

Draft Content 841880011-mmc doc.gif

Table A.1.

Percentage of friends who use tobacco and the Prototype/Willingness Model constructs (mean ± SD).

Role of funding source

This study was funded by the Department of Population Health at the NYU School of Medicine. While Ms. Link and Drs. Shelley and Sherman work for the Department of Population Health, the Department itself was not responsible for the study design, analysis, interpretation of results, writing the manuscript or the decision to submit the manuscript.

Dr. Sherman is supported in part by a grant from the National Institute on Drug Abuse (#1K24DA038345-01).


AL, DS and SS designed the study and wrote the protocol. AL oversaw the data collection and conducted the data analysis. PC conducted literature searches, provided summaries of previous research studies, and assisted with data analysis. All authors contributed to writing and editing the manuscript, and all have approved the final manuscript.

Conflict of interest

The authors have no conflicts of interests to declare.


We would like to thank the research assistants who helped conduct the surveys and code the social media profiles: Jenny Chen, Karishma Kurowski, Jun Lin, Rajkishen Narayanan, Daniele Ngantou, Palak Patel, R. Ivelisse Rozon, and Howard Wong. We would also like to thank Nicholas Lanzieri who helped conduct the surveys and plan the study.


  1. Ayers et al., 2011 Apr J.W. Ayers, K.M. Ribisl, J.S. Brownstein; Tracking the rise in popularity of electronic nicotine delivery systems (electronic cigarettes) using search query surveillance; American Journal of Preventive Medicine, 40 (4) (2011, Apr), pp. 448–453 http://dx.doi.org/10.1016/j.amepre.2010.12.007
  2. Emery et al., 2014 July 1 S.L. Emery, L. Vera, J. Huang, G. Szczypka; Wanna know about vaping? Patterns of message exposure, seeking and sharing information about e-cigarettes across media platforms; Tobacco Control, 23 (suppl 3) (2014, July 1), pp. iii17–iii25 http://dx.doi.org/10.1136/tobaccocontrol-2014-051648
  3. Experian Marketing Services, 2013 Experian Marketing Services. Experian marketing services reveals 27 percent of time spent online is on social networking 2013; http://press.experian.com/united-states/press-release/experian-marketing-services-reveals-27-percent-of-time-spent-online-is-on-social-networking.aspx. (Accessed 8/25/2013).
  4. Freeman and Chapman, 2010 B. Freeman, S. Chapman; British American Tobacco on Facebook: undermining article 13 of the global World Health Organization Framework Convention on Tobacco Control; Tobacco Control, 19 (3) (2010), pp. E1–E9 http://dx.doi.org/10.1136/tc.2009.032847
  5. Gerrard et al., 2006 Jun M. Gerrard, F.X. Gibbons, G.H. Brody, V.M. Murry, M.J. Cleveland, T.A. Wills; A theory-based dual-focus alcohol intervention for preadolescents: the Strong African American Families Program; Psychology of Addictive Behaviors, 20 (2) (2006, Jun), pp. 185–195 http://dx.doi.org/10.1037/0893-164X.20.2.185 (2006-07521-015)
  6. Gerrard et al., 2008 M. Gerrard, F.X. Gibbons, A.E. Houlihan, M.L. Stock, E.A. Pomery; A dual-process approach to health risk decision making: The prototype willingness model; Developmental Review, 28 (1) (2008), pp. 29–61 http://dx.doi.org/10.1016/j.dr.2007.10.001
  7. Grekin and Ayna, 2012/04/01 E.R. Grekin, D. Ayna; Waterpipe smoking among college students in the United States: A review of the literature; Journal of American College Health, 60 (3) (2012/04/01), pp. 244–249 http://dx.doi.org/10.1080/07448481.2011.589419
  8. Hukkelberg and Dykstra, 2009 Mar S.S. Hukkelberg, J.L. Dykstra; Using the Prototype/Willingness model to predict smoking behaviour among Norwegian adolescents; Addictive Behaviors, 34 (3) (2009, Mar), pp. 270–276 http://dx.doi.org/10.1016/j.addbeh.2008.10.024
  9. Kandel et al., 2004 D.B. Kandel, G.E. Kiros, C. Schaffran, M.C. Hu; Racial/ethnic differences in cigarette smoking initiation and progression to daily smoking: A multilevel analysis; American Journal of Public Health, 94 (1) (2004), pp. 128–135 http://dx.doi.org/10.2105/AJPH.94.1.128
  10. Litt and Stock, 2011 Dec D.M. Litt, M.L. Stock; Adolescent alcohol-related risk cognitions: The roles of social norms and social networking sites; Psychology of Addictive Behaviors, 25 (4) (2011, Dec), pp. 708–713 http://dx.doi.org/10.1037/a0024226
  11. Moreno et al., 2012/07/01 M.A. Moreno, A. Grant, L. Kacvinsky, K.G. Egan, M.F. Fleming; College students' alcohol displays on Facebook: Intervention considerations; Journal of American College Health, 60 (5) (2012/07/01), pp. 388–394 http://dx.doi.org/10.1080/07448481.2012.663841
  12. Pearson et al., 2012/09/01 J.L. Pearson, A. Richardson, R.S. Niaura, D.M. Vallone, D.B. Abrams; e-Cigarette awareness, use, and harm perceptions in US adults; American Journal of Public Health, 102 (9) (2012/09/01), pp. 1758–1766 http://dx.doi.org/10.2105/ajph.2011.300526
  13. Pew Research, 2010 Pew Research. Millenials: A portrait of generation next. 2010; http://pewsocialtrends.org/assets/pdf/millenials-confident-connected-open-to-change.pdf. (Accessed 1/11/2013).
  14. Rivis et al., 2011 A. Rivis, C. Abraham, S. Snook; Understanding young and older male drivers' willingness to drive while intoxicated: The predictive utility of constructs specified by the theory of planned behaviour and the prototype willingness model; British Journal of Health Psychology, 16 (2) (2011), pp. 445–456 http://dx.doi.org/10.1348/135910710X522662
  15. Salloum et al., 2014 R.G. Salloum, A. Osman, W. Maziak, J.F. Thrasher; How popular is waterpipe tobacco smoking? Findings from internet search queries; Tobacco Control July 22, 0 (2014), pp. 1–5 http://dx.doi.org/10.1136/tobaccocontrol-2014-051675
  16. Sidani et al., 2014 June 1 J.E. Sidani, A. Shensa, T.E. Barnett, R.L. Cook, B.A. Primack; Knowledge, attitudes, and normative beliefs as predictors of hookah smoking initiation: A longitudinal study of university students; Nicotine & Tobacco Research, 16 (6) (2014, June 1), pp. 647–654 http://dx.doi.org/10.1093/ntr/ntt201
  17. Stock et al., 2013 M.L. Stock, D.M. Litt, V. Arlt, L.M. Peterson, J. Sommerville; The Prototype/Willingness model, academic versus health-risk information, and risk cognitions associated with nonmedical prescription stimulant use among college students; British Journal of Health Psychology, 18 (3) (2013), pp. 490–507 http://dx.doi.org/10.1111/j.2044-8287.2012.02087.x
  18. U.S. Department of Health and Human Services, 2012 August U.S. Department of Health and Human Services; Ending the tobacco epidemic: Progress toward a healthier nation; U.S. Department of Health and Human Services, Office of the Assistant Secretary for Health (2012, August) (Available at: http://www.hhs.gov/ash/initiatives/tobacco/tobaccoprogress2012.pdf. Accessed 8/10/2014)
  19. University of San Diego, 1999 University of San Diego. California tobacco surveys. 1999; http//ssdc.ucsd.edu/tobacco. (Accessed Feb, 2012).
  20. Yamin et al., 2010 C.K. Yamin, A. Bitton, D.W. Bates; E-cigarettes: A rapidly growing internet phenomenon; Annals of Internal Medicine, 153 (9) (2010), pp. 607–609 http://dx.doi.org/10.1059/0003-4819-153-9-201011020-00011
Back to Top

Document information

Published on 26/05/17
Submitted on 26/05/17

Licence: Other

Document Score


Views 7
Recommendations 0

Share this document

claim authorship

Are you one of the authors of this document?