The current study examined sociodemographic and health-related predictors of the use of three forms of social media in an effort to better understand who is accessing and being reached through these emerging communication channels. The results showed that these three forms of social media have distinctly different use patterns and user characteristics, hence different health communication implications. Among the three forms of social media considered in this study, social networking sites by far attract the most users, making them an obvious target for maximizing the reach and impact of health communication and eHealth interventions. Furthermore, with increasing prevalence of personal wireless devices, communication scientists unanimously anticipate the popularity of social networking applications to continue to grow worldwide [2
]. Compared to social networking sites, a much smaller percentage of Internet users reported writing in a blog, suggesting a lower prevalence of blogging. However, reading and commenting on a blog may have been a more reliable measure of blogosphere penetration due to its lower intensity than actively keeping a blog. Moreover, the blogosphere presents a tremendous opportunity for health communication. Particularly so, because bloggers have been observed to act as important communication stakeholders—not only are they information disseminators, but they play a crucial role in directing Internet traffic through opinions and hyperlinks [28
Online support group participation was the only survey item included in the present study that was assessed throughout the three iterations of HINTS, and the weighted prevalence estimates suggest a minor increase: in 2003 and 2005, 3.9% of Internet users had participated in online support groups compared to 4.6% in 2007. User characteristics of support groups differed from blogging and social networking site participation, suggesting that online support group participation is driven by health status. This disease-focused medium may be gradually replaced by more interactive, patient-directed social networking sites and blogs, such as CaringBridge and Patientslikeme. These forms of social media have the potential to serve the social support and empowerment functions previously identified for online support groups [29
Apart from the patterns described above, the results of the study underscore the extent to which age determines who among US adult Internet users are engaging with social media. In this nationally representative sample, age emerged as the single strongest predictor of both social networking and blogging. In light of these findings, it seems reasonable to conclude that health communication efforts utilizing social media will have the broadest reach and impact when the target population is the younger generation. The relatively low penetration in the older population of 55 and older suggests that it is not yet an opportune time to utilize social media in communication with this age group. While this is true currently, we predict a continuing increase in utilization across all generations and groups in the next few years, and it remains a key health communication priority to continue tracking the sociodemographic trends of social media use to be sure that health communicators leverage these dissemination channels most effectively. Finally, for cancer communication efforts, this study found a high prevalence of Internet and social media use among individuals with family members who have/had cancer (see and ), suggesting the potential effectiveness of social media cancer communication efforts targeting “secondary audiences,” that is, caregivers, family, and friends of cancer patients.
A key finding of this study offers new and important implications for health communication in this digital age: among Internet users, social media are found to penetrate the population regardless of education, race/ethnicity, or health care access. In particular, the multivariate analyses showed that having access to a regular health care provider did not predict social media use, suggesting that its significance in the bivariate analyses was primarily due to the effect of age. Specifically, younger individuals are less likely to have a regular health care provider. Considering implications of health communication efforts, the results of this study suggest that in the future, social media promise to be a way to reach the target population regardless of socioeconomic and health-related characteristics. If we can enable broader and more equitable Internet access (eg, increasing broadband access or wireless mobile access), thus reducing the digital divide, the potential for impacting the health and health behavior of the general US population through social media is tremendous. Furthermore, the results showed social networking sites are being utilized by African Americans at a higher rate than by non-Hispanic whites. Given the continuing increase in Internet penetration, these findings suggest a potential systematic shift in the communication pattern that transcends the traditional digital divide. Future studies should continue to examine the impact of changing technologies on patterns of health disparities. On the practice side of health communication, social media outlets may represent an excellent opportunity to reach traditionally underserved members of the population.
The nature of self-report and the current low survey response rates present two major challenges to the generalizability of the results. First, the accuracy of self-reports of specific Internet usage may be affected by recall bias and respondents’ comprehension of survey items. In spite of this issue, this study’s prevalence estimates on Internet and social media penetration are in agreement with the published literature and are the first to be drawn from a nationally representative sample. One aspect to note is that compared to market surveys such as the Pew and Manhattan Research reports, the HINTS estimates are generally more conservative. This is in part attributable to the higher sampling precision mandated for federal surveys. Second, low response rate being a challenge facing all current survey research, HINTS 2007 attempted to boost response rates and extend coverage (especially to cell phone–only households) by adapting a dual sampling frame. As a result, the addition of the mail survey helped remedy the low response rate, to increase the generalizability of the data.
An additional limitation concerns the instrumentation and questions related to blogging and social networking site participation: since neither question asked specifically about health-related use of these technologies, we cannot precisely estimate the prevalence of health-related social media use using HINTS data. Given the growing role of social media in health, future iterations of HINTS may specifically capture health-related social media use [10
]. As well, the question on blogging does not capture individuals who view and comment on blogs and thus may underestimate the degree to which the American public is engaged with this activity.
Finally, with new technologies and social media continuing to evolve rapidly, these data, despite being the most updated national survey data available, may not have been able to capture some emerging social media forms (eg, Twitter and Wikipedia) and rapid changes brought on by the increasing use of personal wireless devices [27
]. In order to track the public’s use of new media, future research should track different age groups’ social media adoption while identifying new forms of social media. Given that the younger age groups are likely to continue their use of social media, we would expect to see a persistent increase across the middle-age population in the near future.
With the goal to develop a better understanding of social media use in the current US population, we have reported on the prevalence and user characteristics of three types of social media using the 2007 HINTS survey. While observations and theories about communication changes brought about by new technologies abound, little is supported by empirical evidence based on nationally representative data. The findings of this study contribute to the knowledge base to inform future programs aiming to utilize social media.
As we have seen, forms of social media present different opportunities for health communication efforts. In particular, social networking sites attract the largest portion of Internet users and are likely to continue to grow, making them an obvious target for maximizing the reach and impact of health communication and eHealth interventions. In addition, recent growth of social media is not uniformly distributed across age groups. New health communication programs aiming to utilize social media must first consider the age of the targeted population. The data also prompt a rethinking of the connection between technologies and health disparities since the findings point to the fact that social media are penetrating individuals of different demographics at the same rate. Opportunities for narrowing the health disparities gap exist through effective use of social media as communication and health promotion platforms. These media will not enable targeted communication messages but may have the capacity to reach a wider audience than traditional media have been able to reach.
Finally, while surveillance research such as the present project is useful for determining the reach of social media, it is less useful for assessing the impact of participation in social media use on health. To assess the multiple levels of social media impact on health, future studies need to bring in diverse disciplines and methods, including intervention studies, longitudinal cohort studies, as well as ethnographic/qualitative observations to examine the effect of the social media–driven changing communication patterns on health.