Table 1 Description of all included studies about social media and nutrition-related outcomes in young adults

Gow, 2010 [43], United States

RCT: 4 arms

18.1

24.4 (5.1)

74%

A private discussion board with scheduled weekly group discussions and asynchronous discussion groups facilitated by researcher.

BMI, weight, Fruit and vegetable (combined) intake (serves/day)

4

Poor

Napolitano, 2013 [39], United States

RCT: 3 arms

20.5 (2.2)

31.4 (5.3)

87%

A private Facebook group containing posts with content such as handouts and podcasts, suggested caloric intake, access to polls and healthy activity or eating event invitations. Used group postings and messages.

Weight

2

High

Ashton, 2017 [6], Australia

RCT

22.1 (2.0)

25.5 (4.6)

0%

A private Facebook discussion group to facilitate social support, send reminders for upcoming face-to- face sessions and send notifications for new material added to the website

BMI, weight, Fruit & Vegetable intake (serves/day), Diet Quality, Energy Intake (kJ/day), body fat mass (g)

3

High

Beetham, 2015 [40], United States

RCT

18.6

100%

100%

A private Facebook group containing only members of group (n = 7–9); expected to post once weekly as well as the lifestyle coaches from their groups. The facilitators shared tips about nutrition, physical activity, inspirational quotes, reminders to read weekly material and attend sessions, and self-compassion and prompted feedback about the weekly lesson.

BMI, weight

6

Poor

Dadkhah [51], 2013 [49], United States

RCT

18 (NR for overall)

22 (NR for overall)

65%

A Facebook group, similar to Phase 2, with weekly posts of relevant tools, news articles, educational Web sites, pages, images, or short texts that supported daily tips.

BMI, weight, energy intake (kcal/day)

7

Poor

Godino, 2016 [42], United States

RCT

22.7 (3.8)

100%

70%

A public Facebook group allowing for social support, accountability, and healthy social norms from existing social networks. The research team delivered 17 challenges and campaigns that were often culturally themed and promoted changes to weight-related behaviours.

BMI, weight

24

High

Hebden, 2014 [49], Australia

RCT

23 (NR for overall)

mean BMI 27 (NR for overall)

80%

A private Internet forum where participants and investigator could to contribute comments, questions and information. New healthy eating information was posted by the investigator biweekly.

BMI, weight, Vegetable intake, fruit intake, sugar-sweetened beverages, energy-dense takeaway meals

3

High

Laska 2016 [47], United States

RCT

22.7 (5.0)

25.4 (3.8)

68%

see Laska 2016

Fast food, sugary beverages, eat breakfast 5–7 days/week; weekly meals prepared at home

24

High

Lytle, 2017 [63], United States

RCT

22.7 (5.0)

25.4 (3.8)

68%

see Laska 2016

BMI, weight, waist circumference, % body fat

24

High

Partridge, 2015 [64], Australia

RCT

27.7 (4.9)

27.1 (2.5)

61%

A community blog (no more information given).

BMI, weight

3

Moderate

Dadkhah, 2013 [51], United States

Qualitative: focus groups

18.0 (0.5)

NR

40%

Formative research: to determine the need for a weight gain prevention program using social media for first-year college students

n/a

n/a

Moderate

Dennison, 2013 [52], United Kingdom

Qualitative: focus groups

23.8 (7.9)

NR

68%

Exploring young adult perspectives on apps relating to health behaviour change.

n/a

n/a

High

Vaterlaus, 2015 [65], United States

Qualitative: focus groups and interviews

20.4

26%

79%

Exploring young adult perceptions of social media on health behaviours.

n/a

n/a

High

Laska, 2016 [47], United States

Process evaluation (of Lytle 2015). Quantitative methods; survey data and engagement metrics.

22.9 (5.0)

25.4 (3.8)

67%

A private social networking and support website with a discussion forum available to participants and a limited number of their invited guests was designed to reinforce, inform and encourage exchange and support between participants. Students were encouraged to track their weight and up to 10 weight-related behaviours on the website. Trained interventionists interacted with participants through the website and there was an “Ask the Expert” section where students could ask confidential questions about a personal challenge or health issue. The website included articles, recipes, quizzes, videos and ways to accumulate points for prizes.

n/a

n/a

Poor

Merchant, 2014 [46], United States

Process evaluation (of Godino 2016). Mixed methods: engagement analytics and semi-structured interviews.

22.0 (3.8)

28.7 (3.5)

70%

Participants were invited to ‘like’ an open Facebook page. Non-study participants could also like the page, view content and interact with it. Behaviour change campaigns were posted, as well as healthy tips and tailored messages from the health coach. Participants were encouraged to self-monitor their diet and physical activity on Facebook.

n/a

n/a

Moderate

Partridge, 2016 [50], Australia

Process evaluation (or Partridge 2015). Mixed methods: survey data, engagement metrics and semi-structured interviews

NR

NR

NR

A community blog (no more information given).

n/a

n/a

Moderate

Pappa, 2017 [48], Brazil

Mixed methods; engagement analytics and content analysis.

26 (6)

NR

57%

Reddit is a public social media site. /r “LoseIt” is a subreddit community where people interact about weight loss issues.

n/a

51^a

Poor

Chung, 2016 [41], United States

Experimental; single arm

20.3 (overweight group), 19.0 (healthy weight group)

58%

67%

A private Twitter group where participants received both text and photo-based Tweets from the study team that focused on health messages; encouragements and reminders to wear their Fitbits and log their dietary intake.Participants were encouraged to post questions to the study team or to their Twitter group. The study team posted questions to the group on topics such as “What small change are you going to make this week?” to encourage interactivity and to tailor message content to participant needs. Fitbit accounts were set up to autotweet daily steps and distance travelled to the assigned private Twitter group so that individuals could see how others were doing, which was the basis of some of the competitions.

n/a

2

Poor

Dadkhah 2013 [51], United States

Experimental; single arm

18.0 (0.4)

23

77%

A Facebook group (not identified as public or private) where daily health tips were shared. Participants chose the timing of the postings through Facebook.

BMI, weight, energy intake (kcal/day)

7

Moderate

Harvey-Berino, 2012 [44], United States

Experimental; single arm

NR

53%

87%

A bulletin board and discussion forum where 1-h weekly “group meetings” were led by an interventionist trained in behaviour modification and online facilitation. The bulletin board could be used for group communication.

BMI, weight

4

Poor

Meng, 2017 [45], United States

Experimental: 5 arms

19.9 (1.7)

22.7 (3.0)

67%

A private community group consisting of 3 modules: 1) group goal; 2) self-track message wall to post their F&V consumptions 3) bar graph illustrating weekly summaries. Groups consisted of 1 participant and “confederates” i.e. fake people.

Fruit and vegetable intake

1

Moderate