We will use the population, intervention, comparator, and outcome (PICO) framework to identify appropriate Medical Subject Headings (MeSH) terms for the literature search. Meanwhile, the PRISMA-P (Preferred Reporting Items for Systematic Reviews and Meta-Analysis Protocols) 2015 guideline will be used as the standard reporting protocol guideline checklist. The PRISMA-P 2015 guideline will aid in describing the rationale, hypothesis, and planned methods of the review [15]. The checklist is provided in Multimedia Appendix 1.

This systematic review will comprise a literature search, article selection, data extraction, quality appraisal, data analysis, and data synthesis. The protocol of this systematic review was prospectively registered on PROSPERO (CRD42023458336).

The PICO framework presented in Table 1 is derived from the research questions previously mentioned.

We will search the following databases: Science Direct, PubMed, Cochrane, Embase, JMIR, ProQuest, and Google Scholar. Key terms relating to telehealth and telenutrition were extracted from an initial review of the literature. Specific search terms, such as “obesity,” “adolescent,” “dietary behaviors,” and “physical activity,” were identified in a preliminary scan of the literature and chosen in consultation with a medical librarian. Search terms will include MeSH terms and related keywords and are grouped into 3 themes presented in Table 2. All the MeSH terms and Keywords are included in Table 2 with the following structure: Obesity (MeSH OR keywords) AND Adolescent (MeSH OR keywords) AND Telenutrition (MeSH OR keywords). To broaden the scope of the article collection, we will not limit the publication period or language of publication in our search strategy. In addition to exploring these databases, a manual search will be performed to identify the reference lists of included studies. Authors of conference and poster abstracts selected for inclusion will be contacted to see if a full text is available to be included.

Only human intervention studies, either quasi-experimental studies or randomized controlled trials, will be included in this review. This review will evaluate original research articles using digital health promotion interventions delivered via any type of telehealth and involving adolescents aged 13-18 years with overweight or obesity. Overweight and obesity among adolescents aged 5-19 years are defined by their BMI per age z score, as outlined in the World Health Organization growth chart references [16]. We will include any study that categorizes participants with a BMI per age z score >1 SD as overweight and those with a BMI >2 SD as obese.

The review focused mainly on the changes in healthy behaviors and nutritional outcomes, including dietary behavior, physical activity levels, and BMI. Any outcomes other than these will be excluded from our review. In addition, articles for which full texts cannot be obtained despite requests to the corresponding authors will be excluded.

All articles found through database searches will be saved in the citation and article management software Mendeley and Rayyan AI, which will be used to eliminate duplicates and screening. The titles and abstracts of all studies will be reviewed independently by 2 reviewers. Studies that do not meet the eligibility criteria will be excluded, and any disagreements will be discussed until a consensus is reached. The full text of the remaining studies will then be examined by 2 reviewers independently to determine final eligibility, with any disagreements resolved by a third and fourth reviewer. A PRISMA flow diagram will be used to record the details of the screening and selection process so that the study can be replicated.

One reviewer will read the full text of all the papers included in the final selection to extract the predetermined outcomes, which will be validated by a second reviewer. The outcomes will be extracted into a standard extraction form, which will be designed in Covidence (Veritas Health Innovation). The detailed extraction custom-built form is summarized in Table 3. Disagreements will be resolved through discussion, and if consensus cannot be reached, a third and fourth reviewer will be consulted. Missing data will be considered in the risk of bias assessment, but due to time constraints, authors will not be contacted.

Following the final selection of studies, 2 reviewers will independently assess the risk of bias in all papers included in the final selection. If there is a disagreement in judgment, the reviewers will discuss it before consulting the third and fourth reviewers. The Cochrane Collaboration Risk of Bias tool will be used to assess the randomized controlled trials included in the review and assign low, unclear, or high risk to the studies for each of the potential biases [17]. Studies will be classified as low risk if the identified bias is unlikely to significantly alter the results, and they are classified as high risk if the bias clearly has the potential to significantly alter the results. Unclear risk occurs when the bias imposes some doubt on the outcome. The risk of bias assessment will cover 6 domains: selection, performance, detection, attrition, reporting, and other bias [17].

Meanwhile, ROBINS-I (Risk Of Bias In Non-randomized Studies of Interventions) will be used for evaluating the risk of bias in estimates of the comparative effectiveness of interventions from studies that did not use randomization to allocate units to comparison groups [18]. A table will be created that summarizes the quality of all included studies.

A meta-analysis is unlikely to be conducted due to the expected diversity of study aims, methods, and reported outcomes. As a result, we will conduct a descriptive analysis to summarize the extracted data. Each outcome will be coded as having no (0), some (+), or significant evidence (++) of effectiveness in both outcomes. Significant evidence will be coded only when the intervention (ie, nutritional literacy and behavior change communication via telehealth delivery) outperforms a comparator or control. The intervention will be considered to have some evidence of effectiveness if there is a significant difference over time but not between groups or a significant improvement in only a subgroup of the population. Studies will be grouped by target health behavior (eg, dietary pattern only, physical activity only, or both) and analyzed together to describe the effectiveness of telehealth or telenutrition interventions in general for both those target behaviors; in particular, the analysis will focus on either dietary pattern or physical activity. The discussion will synthesize the data to describe specific intervention components that characterize a successful telenutrition intervention, such as the type of digital platform, engagement features, theory, method, and behavior change strategy used. The theory, method, and behavior change strategy will be analyzed and presented based on the intervention mapping approach and its taxonomy, as described by Bartholomew et al [19] and Kok et al [20]. Furthermore, the information regarding the identified barriers and challenges of telehealth interventions will be summarized to give directions for future research and development.

There will be no involvement from patients or the public in this study. This study synthesizes data from studies that were already peer reviewed and published. Thus, no ethical approval will be required.