Preliminary list of data items to be extracted from included studies
| Data items | Characteristics |
| Bibliographic data | First author, year of publication, author country, health domain (nutrition/PA/SB),30 55 digital nudge, recommender system or a combination |
| Intervention characteristics | Study design, population, sample size, dose, digital nudge combination |
| Recommender system delivery channel | Mode of delivery (delivery channels: for example, visual, audio, haptic) and delivery devices (eg, desktop, mobile, wearable, ambient)6 |
| Recommender system methods | Hybrid methods, content-based filtering, collaborative filtering, graph-based methods |
| Classification of digital nudges | Social nudges (guide the user’s behaviour by providing references to how other users behave), reinforcement nudges (reinforce behaviours and choices by increasing their salience in the mind of the user), disclosure nudges (adding information that is accessible, clear and relevant to the choice that the user is about to make), friction nudges (encouraging or discouraging behaviour by removing or adding friction), feedback nudges (information about a past or current behaviour of a user), default nudges (assumed desired behaviour), warning nudges (different kinds of warnings and graphics to grab attention), scarcity nudges (information that something is difficult to acquire), deception nudges (affect how users perceive choice alternatives), commitment nudges (to motivate the user to behave)6 55 56 |
| Digital nudge mechanisms | Decision information (translate information, increase the salience of information, make information visible or change the phrasing of information) Decision structure (change range of composition, change choice defaults, change option consequences, change option-related effort) Decision assistance (provide reminders or facilitate commitment) Social decision appeal (increase the reputation of the messenger, provide a social reference point, instigate empathy)1 |
| Personalisation | None, partial (study gathers user data (eg, location, user demographics, user actions) to infer the potential influence of the nudge on user behaviour), full (such information is used to personalise the choice architecture of individual users dynamically)6 |
| Interconnection | None, partial (study investigates how information from other users affects user behaviour), full (study investigates how actions of one user, in turn, dynamically modify the choice architecture of other users)6 |
| Implementation information | Any information about implementation and user engagement |
| Effects | Primary outcome: outcomes related to weight, weight management, PA/SB Secondary outcome: outcomes related to mental health, user engagement, user satisfaction |
PA, physical activity; SB, sedentary behaviour.