JMIR Publications

ABSTRACT

Background:

Nutrition-related apps have great potential to influence eating behaviours and improve adherence to healthy eating guidelines. In addition, when personalised advice is delivered automatically, these apps enable the provision of individual support at-scale with relatively low-cost.

Objective:

To investigate the effectiveness of a mobile web application (eNutri), capable of delivering automated personalised nutrition (PN) advice, in increasing diet quality, relative to general population food-based dietary guidelines.

Methods:

UK adults (aged > 18 years) were randomised to (i) PN advice or (ii) control advice (standard healthy eating guidelines) in a 12-week controlled, parallel, single-blinded, dietary intervention, which was delivered online. Uniquely, PN advice was derived according to an 11-item modified US Alternative Healthy Eating Index (m-AHEI) to align with UK dietary and nutritional recommendations. The primary outcome was change in diet quality (m-AHEI) at 12 weeks. Dietary intake was determined using the eNutri Food Frequency Questionnaire (FFQ). Participant surveys to evaluate the PN report, and longer-term impact of the PN advice were deployed immediately following, and 4.6 months after completion of the study respectively.

Results:

In total, 210 participants completed at least one additional FFQ following baseline. Following exclusion of outliers with unfeasible energy or food group intakes (n=23), 96 participants were included in the PN group (43.5 years, SD 15.9; BMI 24.2 kg/m2, SD 4.4) and 91 in the control (42.8 years, SD 14.0; BMI 24.8 kg/m2, SD 4.4). The average interval between FFQs was 10.8 weeks. A significant increase in the overall m-AHEI score of 3.5 out of 100 (CI 95%: 1.19-5.78), equivalent to an increase of 6.1%, was observed in the PN group, compared with the control group (P = .003). The m-AHEI components ‘nuts and legumes’ and ‘red and processed meat’ also improved significantly more in the PN group (P’s = 0.04). At follow-up (mean 5.9 months after intervention), 64.3% of PN participants agreed that the advice from the app had prompted them to change their eating/drinking habits, and that they were still following some (‘any’) of the advice.

Conclusions:

This result supports the hypothesis that the eNutri app is an effective online tool for the automated delivery of PN advice in UK data. Our data also demonstrate the potential of our app to improve short-term diet quality and engagement in healthy eating behaviours, when compared with standard healthy eating guidelines. Clinical Trial: The EatWellUK study was registered at ClinicalTrials.gov (NCT03250858)