Does social media affect our perceptions of what most people eat, what we choose to eat and our body weight?

A diet high in fruit and vegetables can help prevent obesity and decrease the risk of associated chronic comorbidities and mortality, however many people do not consume the recommended amounts. Thus, innovative ways of encouraging healthier food choices are needed, for example, by using social influences. Social norms, which describe how the majority of people behave, have consistently been shown to be associated with, and to affect, eating behaviour. Nowadays, one way such norms may be communicated is via social media, however, the relationship between social norms conveyed via social media and eating behaviour has received little attention. Therefore, this thesis aimed to quantify the extent to which social norms communicated via social media influenced food consumption, and whether social media could be used as the basis for an online social norms intervention, aimed at nudging healthier consumption. Associations with body weight were also considered. Through cross-sectional and longitudinal survey studies, it was established that our perceptions of what social media users eat predicts our own choices, both acutely and over time. A laboratory-based experimental study also demonstrated that exposing participants to socially endorsed social media images significantly nudged the proportion of food consumed towards low energy-dense vs. high energy-dense snacks, allowing causal conclusions to be established. Additionally, social norms may enhance attention towards LED foods in those with overweight and obesity, however, these findings were not statistically significant. Finally, in a two-week, online social media intervention using social norms, participants reported consuming significantly more fruit and vegetables than those in the control condition. These results therefore demonstrate the potential utility of a social norms-based social media intervention in nudging consumption towards healthier food choices. Wider implications for policy, as well as advertising industries are discussed. Future research should aim to replicate these results with larger and more diverse samples