Nutri-score Labelling in a UK Restaurant Setting: a Randomised Control Trial

An adapted version of the University of North Carolina PME scale will be used to measure health concern, product attitude, and discouragement of item consumption in response to the food menu. Participants will be given the menu they ordered from to look at. All participants will answer 3 questions using a Likert scale ranging from 1 - 5 anchored by "not at all" and "a great deal". The mean response to the three items will be calculated. PME is used as an early indicator of a health message's potential to change behaviour (e.g., reduce selection of less healthy items). The scale has been used previously to identify the potential impact of food labels and is predictive of long-term behaviour.

Nutritional quality for meal choices will be calculated using the UK NPM scoring system. This calculates a score (continuous variable) for a product by deducting a total score for positive nutrients from a total score for negative nutrients. A lower score represents greater nutritional quality and a higher score represents poorer nutritional quality.

We will calculate the mean NPM score weighted by energy content of items for each participant. This will ensure that the NPM score of main dishes will be weighted higher than that of starters or sides (assuming energy content of selected main is higher than any starters or sides), leading to a more representative overall NPM score for food orders. This will be done using the weighted.mean function in R. This function multiplies each value (i.e. NPM scores for individual items) by its corresponding weight (kcal content of individual items), sums all items and then divides this by the sum of the weights.

A NS category value will be recorded for each item selected by participants (A,B,C,D,E)

The sum of energy (kcal) across all foods ordered for the meal will be calculated

Researchers will take pictures of plates before and after each item is given to participants. Researchers will estimate the percentage of each dish consumed.

Participants will be asked whether they shared any of their meal with someone else and this will be taken into account when making percentage estimates. A random 10% of percentage estimates will be performed by a second researcher to measure reliability.

Energy and nutrient consumption will be calculated by multiplying the total energy/nutritional content of food items selected/consumed by the estimated proportion of the item consumed. Nutrients explored will be salt, sugar, fat, saturated fat, protein and fibre.

The morning after participants take part in the study, they will be emailed with a link to complete a dietary recall survey (via intake 24, https://intake24.co.uk/). Participants will report everything they ate after the study session up until they went to bed. They will be asked to provide as much detail as possible on what they had for dinner, snacks, and drinks after the study session. We will estimate based on data from Intake 24, later consumption (kcal, salt, fat, saturated fat, sugar, protein, fibre).