Abstract
BACKGROUND: A digital-first strategy is increasingly implemented to reduce participant burden, accelerate recruitment, collect real-world data, and increase the diversity of the study population. However, fully remote studies lack face-to-face interaction, which may affect motivation, particularly in the delivery of personalized nutritional advice. Additionally, self-reported data may vary in terms of standardization and completeness. OBJECTIVE: The study's primary objective is to evaluate the feasibility of conducting a fully remote, fully digital randomized controlled nutritional intervention, including participant experience and the capability to perform do-it-yourself anthropometric measurements at home. Secondary objectives are to determine whether self-collected data could detect changes in body weight and other anthropometric outcomes, and to compare the effectiveness of generic versus personalized nutrition advice, with and without personalized food boxes. METHODS: We conducted a fully online, 3-arm randomized controlled trial including adults with overweight or obesity who were motivated to lose weight. Participants were assigned to a control group that received generic advice (n=43), a personalized intervention group that received personalized advice only (n=40), or a personalized intervention plus group that received personalized advice plus personalized food boxes (n=39). The 6-week intervention was delivered entirely digitally, and all anthropometric measurements, questionnaires, and dietary data were self-collected at home. Feasibility was assessed using adherence metrics, completion of self-measurements, and a user-experience questionnaire. Secondary analyses evaluated weight loss, changes in anthropometry, and exploratory associations, including sex differences. RESULTS: Feasibility was high-102 out of 122 (83.6%) participants found the self-measured anthropometric assessments easy to perform, and 112 (91.8%) participants reported that completing questionnaires from home was easy. For secondary outcomes, participants receiving personalized, but not generic, nutritional advice significantly lost body weight (-1.0 kg; P=.002). Participants receiving personalized food boxes in addition to personalized nutritional advice lost significantly more body weight than the other 2 groups (-2.5 kg; P=.001) and also showed a decrease in hip circumference (-2.9 cm; P=.01). Personalized advice was not easier or more enjoyable to implement than generic nutritional advice, whereas the addition of personalized food boxes improved the ease of implementing personalized nutritional advice (P<.001). All participants, irrespective of the intervention arm, reduced intake of unhealthy food groups, including ready-made meals (113.6 g vs 78.5 g, -30.9%); sauces and gravy (18.8 g vs 10.0 g, -46.8%); sweet snacks (84.8 g vs 64.1 g, -24.4%); savory snacks (50.5 g vs 40.0 g, -20.1%); bread, pasta, rice, and wraps (nutritional quality score of 1.9 vs 1.7, -10.5%); and vegetables (129.0 g vs 118.7 g, -8.0%); and replaced coffee with tea. CONCLUSIONS: This study demonstrates that fully remote, participant-led nutritional intervention studies are feasible, with participants able to independently perform anthropometric measurements and self-report data of sufficient quality to detect meaningful effects. Personalized nutritional advice resulted in greater weight loss than generic advice, and the addition of personalized food boxes further enhanced the beneficial anthropometric effects of the intervention. We conclude that such nutritional intervention studies can be conducted fully online, resulting in measurable anthropometric effects after 6 weeks. TRIAL REGISTRATION: ClinicalTrials.gov NCT06547983; https://clinicaltrials.gov/ct2/show/NCT06547983.