Understanding Food Choices (Aim 2)

The important role that diet plays in health and disease is well established. Excessive intake of energy, saturated fat and sodium increase the risk of heart disease, diabetes and certain cancers. As a result, interventions aimed at encouraging healthier food consumption have been pursued by many countries. These can be broadly grouped into the following categories: price manipulations, food labelling, and behavioral nudges.

No study has previously assessed the potentially interactive effects of a multicomponent intervention that incorporates the strongest features of each intervention component while discarding those that do not meaningfully contribute to healthier consumption. That is the goal of this effort.

At part of the first stage of the MOST framework (Aim 1), the investigators evaluated the effects of evidence-based public health interventions on consumers' diet quality on a web-based grocery store "NUSMart". The investigators focused on the following four interventions: 1. Explicit Tax, 2. Food labels (with the summary of healthiness of shopping baskets \& targets), 3. Ordering and 4. Within Group Healthier Substitution. Based on the study results, the investigators assembled a multicomponent intervention that consists of a subset of the above four interventions that may optimize online grocery platforms in terms of diet improvement.

The multicomponent intervention consists of the following components:

1. Food labels (with the summary of healthiness of shopping baskets \& targets)- Food labels, called Nutri-Traffic-Lights (NTL), indicating whether food and beverage products were healthy, unhealthy or somewhere in between were designed for all products on NUSMart based on the Nutri-Score Nutrition Scoring System. These were supplemented with a video explaining the labels and a dynamic pie chart indicating the overall nutritional quality of the shopper's basket (updated in real time) known as MyCart summary.
2. Ordering- Healthier products, in terms of Nutri-Score Points, were displayed first within each category and subcategory.
3. Within Group Healthier Substitution- At checkout, up to 4 healthier substitutes were suggested for products in the shopper's cart based on the Nutri-Score Points of these products. Substitutes were from the same category as the product that was added to cart and were close to the original product in terms of price as well.

In this study (Aim 2), using a 2-arm randomized controlled trial, the investigators will test whether the assembled multicomponent intervention has a sustained positive effect on diet quality over 3 purchases over a 3-5-week period, wherein the purchased foods are delivered to the participants' homes.

Over the course of the study, participants will log into the NUSMart website once a week and will be asked to purchase their weekly groceries with a minimum spend of $59. Each participant will therefore shop a total of 3 times using the same version of NUSMart randomly assigned to them, during the study.

The investigators' hypotheses about the effects of the multicomponent intervention on diet quality, measured by the weighted (by the number of servings) average Nutri-Score Points (primary) of finalized shopping baskets, are as follows:

1. The multicomponent intervention will significantly improve diet quality as measured by the weighted Nutri-Score Points (primary), calories, sodium, sugar, and saturated fat per serving, which will be calculated based on all purchased products' total nutritional value.
2. The improvements in diet quality will be sustained through repeated shops where food is actually delivered, thus making it more likely that results are externally valid.