Systematic review
Effects of using social media in health awareness campaigns to promote healthy lifestyle habits
Systematic review
|Updated
Facilitating and encouraging healthy behaviours is important to prevent illness and promote health. This systematic review investigates whether the use of social media can have an impact on healthy behavours. We included seven studies.
Key message
Facilitating and encouraging healthy behaviours is important to prevent illness and promote health. This systematic review investigates whether the use of social media can have an impact on healthy behavours. We included seven studies. We found many different interventions and most of the comparisons and outcomes were measured only by one small study.
- Studies comparing an online social network to no information or information delivered via other channels showed no or only small effects for change in knowledge, attitudes, subjective norms, perceived control, intentions and behaviour.
- Several studies combined an online social network (OSN) with other interventions. Outcome measures were change in knowledge, attitudes, behaviour, eating habits, weight, energy and fat intake, self efficacy and sharing in social media. One of the studies found small effects in favour of OSN in combination with other interventions for change in knowledge, attitudes and behaviour related to food safety. None of the other comparisons showed significant differences.
- One study which compared watching motivational videos to not watching such videos found small to no effects for the outcome measures self efficacy and physical activity.
- We found no studies which compared the use of social media to the use of traditional media such as newspapers, radio or television.
- We found no relevant studies about tobacco or alcohol.
Because of small studies and weaknesses in how they were performed, we have low to very low confidence in these results. We cannot draw any clear conclusions about whether or not the use of social media in health promotion campaigns have an effect on healthy behaviours.
The fact that we are very unsure about the potential effect does not mean that there is no effect; it means that we are unsure about whether or not there is an effect.
Summary
Background
The Norwegian health authorities seek to increase effort to prevent illness and to promote health. An important part of this effort is to facilitate and encourage healthy behaviours. Different types of health promotion and public awareness campaigns are used to influence knowledge, attitudes and behaviour in this regard. Such campaigns have been presented via mass media, and in recent years, also via social media. When commissioning this report from the Norwegian Knowledge Centre the Health Directorate said that their main purpose for taking part in social media is to listen to and communicate with the population where they are, in addition to setting the agenda and to influence attitudes and thereby promote good health and quality of life. Public awareness campaigns are costly and the Health Directorate requested an updated systematic review of the relevant research about using social media in health promotion campaigns.
Social media may potentially reach vast populations. Even more people can be reached when users of social media share and discuss messages. Individuals, businesses, organisations and others spread health related content in social media. This can be useful, but also confuse the users who may not be able to assess the quality and trustworthiness of the content. Social media have become part of everyday life for a large part of the population. It is important to examine whether or not the use of social media in health promotion campaigns is effective.
Method
We performed systematic literature searches in the following databases: Cochrane Database of Systematic Reviews, Database of Abstracts of Reviews of Effects, HTA Database, Cochrane Central Register of Controlled Trials, MEDLINE, Embase, PsycINFO, CINAHL, ERIC, SveMed+ and ISI Web of Science. We also searched Google Scholar. The search strategies consisted of subject headings and text words for social media which were combined with subject headings and text words to identify health promotion campaigns. The search was completed in October 2014. Two persons assessed all titles and abstracts independently to identify studies that potentially met the inclusion criteria and two persons then assessed full text of the identified studies independently. We used the Cochrane Effective Practice and Organisation of Care Group’s Risk of Bias Tool to assess risk of bias. The quality of evidence was assessed for each outcome using the GRADE methodology. The documentation was judged to be of either of high quality (⨁⨁⨁⨁), moderate quality (⨁⨁⨁◯), low quality (⨁⨁◯◯), or very low quality (⨁◯◯◯).
Results
We included seven studies, four of which were randomised controlled trials, two non-randomised controlled studies and one controlled before and after study. All studies were relatively new, published between 2010 and 2013. One study was performed in Australia, the rest in the United States of America.
From the included studies, we identified three categories of interventions using social media: 1) online social networks such as Facebook, Twitter etc., 2) online social networks combined with other interventions and 3) video sharing sites such as YouTube etc. These were compared to no information or waiting list, to information given via other channels, or to each other.
We found no studies which compared the use of social media to use of traditional media such as newspapers, radio or television.
The outcome measures in the included studies were changes in attitudes, knowledge and behaviour regarding nutrition and/or physical activity. There were many different interventions and most of the comparisons and outcomes were measured by only one small study. We found no relevant studies about tobacco or alcohol.
In the included studies we found 51 different comparisons and there was large variation between studies in interventions, objectives and outcomes. Because of this heterogeneity, we did not combine the results in meta-analyses.
Two out of ten comparisons from three studies which compared online social network to no information or waiting list showed statistically significant results in favour of online social network, (95 % CI 0,1 to 0,5 for change in attitude, 95 % CI 0,02 to 0,4 for change in behavior measured on a scale from 1 to 5) (⨁◯◯◯). There were no significant differences for the outcome measures change in knowledge and attitude regarding nutrition and physical activity or for attitudes, beliefs, subjective norms, intentions and perceived control regarding multivitamins.
Two out of five comparisons from two studies which compared online social network to information given via other channels showed statistically significant results in favour of online social network, (95 % CI 0,03 to 0,4 for change in behaviour, 95 % CI 0 to 0,4 for change in attitude measured on a scale from 1 to 5) (⨁◯◯◯). There were no significant differences for the outcome measures change in knowledge and behaviours regarding nutrition and physical activity.
All of the three comparisons from one study which compared online social network combined with other interventions to no information or waiting list found results in favour of online social network combined with other interventions, (95 % CI 0,1 to 0,5 for change in attitudes, 95 % CI 0,04 to 0,1 for change in knowledge and 95 % CI 0,01 to 0,4 for change in behavior measured on a scale from 1 to 5) (⨁◯◯◯).
Two of ten comparisons from two studies which compared online social network combined with other interventions to information given via other found results in favour of online social network combined with other interventions, (95 % CI 0,04 to 0,4 for change in attitudes and 95 % CI 0,02 to 0,04 for change in behaviour measured on a scale from 1 to 5) (⨁⨁◯◯) (⨁◯◯◯). There were no significant differences for the outcome measures change in knowledge, weight, physical activity, energy and fat intake or self efficacy.
None of the six comparisons from one study which compared watching motivational videos on YouTube to not watching such videos found statistically significant differences (⨁◯◯◯).
Our GRADE assessments revealed that the documentation was of low to very low quality, which means that we have limited confidence in these results.
Discussion
Due to large variations between studies concerning interventions, objectives and outcomes we did not combine the results in meta-analyses. The quality of the documentation was low to very low. This is due to the fact that in most cases there was only one study reporting each outcome, and most studies had very few participants which resulted in wide confidence intervals. In several of the studies, the outcome measures were self-reported and often subjective, which can be a source of bias.
The comprehensive search strategies has increased the probability that we have identified all relevant studies in the bibliographic databases. There is always a risk that we might have missed relevant studies, particularly grey literature. A strength of this systematic review is the systematic methods used to identify, select, appraise and summarise studies.
Conclusion
We identified and included seven studies, including four randomised controlled trials, two non-randomised controlled trials and one controlled before and after study which examined the effect of using social media to influence attitudes, knowledge and behaviour regarding nutrition and/or physical activity. We found no studies that examined the effect of using social media in campaigns to influence the use of tobacco or alcohol.
The studies were small and had several methodological weaknesses. Our grading of the documentation reveals that we have low to very low confidence in the effect estimates. The fact that we are unsure about the potential effect does not mean that there is no effect, it means that we are unsure about whether or not it there is an effect.
On this basis, we cannot draw any clear conclusions about whether or not the use of social media in health promotion campaigns is effective.
Further research about the use of social media in health promotion campaigns is needed before we can conclude about effect. To be able to say something about the effect of social media without co-interventions, studies examining the effect of social media alone compared to other interventions is needed.