Abstract: Is Online Health Information Seeking Behavior Associated with More Use of E-Cigarettes? A Moderated Mediation Model

◆ Qinghua Yang, Texas Christian University

Electronic devices have become one of the primary avenues for health information seeking. However, due to the scientific uncertainty surrounding electronic cigarettes (e-cigarettes) and the lack of gatekeeping on the Internet, the online communication environment may not objectively portray the risks and benefits of e-cigarettes. For instance, content analyses documented that e-cigarette-related texts were mostly positive/pro-e-cigarette on websites and social media (Grana & Ling; 2014; Luo et al., 2014). In the same vein, the e-cigarette information seeking of youth and young adults, who are heavy users of Internet and social media, significantly predicted their e-cigarette use six months later, and the information they sought were predominantly pro-e-cigarette (Yang, Liu, Lochbuehler, & Hornik, 2019).

Despite the available empirical evidence, the underlying mechanism and moderator of the relationship between individuals’ online health information seeking behavior (OHISB) and their e-cigarette use remain unclear. Given the pro-e-cigarette online environment, a moderated mediation model was hypothesized that the relationship between individuals’ OHISB and e-cigarette use is mediated by their lower perceived harmfulness of e-cigarettes, with perceived trustworthiness of online health information moderating the association between OHISB and perceived harmfulness.

The Health Information National Trends Survey 5 Cycle 2 (2018, NCI) was analyzed to test the moderated mediation model, with potential confounders (i.e., age, gender, race/ethnicity, income, smoking status, and health status) being adjusted and sampling weights incorporated in the analyses. The proposed model, tested using structural equation modeling in Mplus 8.2, fits the data very well (χ^2(8) = 10.20, p = .25; RMSEA = .013 (90%CIs [.000, .034], CFI = .98, TLI = .96, SRMR= .05). Individuals’ e-cigarettes use was negatively predicted by perceived harmfulness of e-cigarettes compared to combustible cigarettes (= -.30, p< .001), which was further predicted by OHISB for themselves (= -.36, p< .01), but not OHISB for others (p= .25). Consistent with hypothesis, the relationship between OHISB and perceived harmfulness was moderated by perceived trustworthiness of online health information (p< .01).

To examine the moderated mediation, bootstrapping with 1,000 iterations was implemented to obtain bias-corrected 95%CIs of the indirect effects (Preacher & Hayes, 2008). The indirect effect between OHISB for themselves and e-cigarette use through perceived harmfulness of e-cigarettes was only significant when individuals show high- (Mean+1SD; .14, 95%CI [.05, .25]) or medium-level (mean; .11, 95%CI [.03, .21]) perceived trustworthiness of online health information. However, when individuals do not trust online health information (Mean-1SD), the indirect effect was not significant (.08, 95%CI [-.05, .22]).

The results show consistency with existing empirical evidence of pro-e-cigarette online public communication environment, which could influence individuals’ attitudes towards and risk perceptions around e-cigarette use and validate their vaping behavior. However, critical thinking of online health information could be a protective factor of potential negative effects of OHISB. The divergence of OHISB for oneself and for others in influencing health behaviors was also highlighted. The findings, which inform FDA’s regulations regarding e-cigarettes online marketing and misinformation, respond to the conference theme by contributing to the interdisciplinary scholarship of both health communication and tobacco regulatory science.