Journal of Medical Internet Research

Introduction

College college students are in a important developmental stage characterised by the transition from adolescence to maturity, throughout which they encounter a number of challenges, together with elevated tutorial calls for and evolving social roles. Evidence means that faculty college students usually exhibit inadequate self-management capability associated to wholesome life-style behaviors [], with insufficient bodily exercise, extended sedentary conduct, irregular weight loss program patterns, and sleep disturbances being significantly prevalent. Previous analysis has demonstrated that well being behaviors established throughout this developmental interval are likely to exhibit substantial stability and continuity over time []. The adoption of unhealthy life-style behaviors throughout this stage has been proven to considerably improve the danger of power illnesses, despair, and anxiousness later in maturity [,]. Therefore, the implementation of early and efficient interventions focusing on these 4 key life-style behaviors amongst faculty college students is of considerable public well being significance [].

With the speedy development of digital applied sciences and the widespread adoption of sensible units, digital well being interventions (DHIs) have emerged as an progressive strategy to well being promotion and are more and more acknowledged as an necessary technique of bettering life-style behaviors amongst faculty college students [,]. Particularly within the put up–COVID-19 period, DHIs have demonstrated larger adaptability and broader software potential than conventional face-to-face well being intervention fashions [-]. In current years, a rising physique of empirical proof has proven that DHIs are efficient in selling bodily exercise amongst faculty college students [-], lowering sedentary time [,], bettering weight loss program behaviors [,], and enhancing sleep high quality [,]. These interventions—encompassing cell apps, wearable units, on-line platforms, and social media—provide a number of benefits, together with low value, excessive scalability, and a excessive diploma of personalization [,], and have been proven to boost consumer engagement and facilitate sustained conduct change [,]. Concurrently, developments in rising applied sciences, resembling synthetic intelligence, proceed to drive the refinement of DHI implementation methods and additional improve intervention effectiveness [].

However, the prevailing physique of analysis on DHIs focusing on life-style behaviors amongst faculty college students stays topic to a number of limitations. On the one hand, nearly all of unique intervention research have targeted on single life-style behaviors or particular technological modalities, with a relative lack of complete designs that combine a number of behaviors and intervention approaches. At the identical time, key intervention dimensions—resembling practical traits, intervention period, participant demographics, and adherence—have but to achieve unified requirements or methodological consensus [,]. On the opposite hand, current systematic evaluations and meta-analyses on this discipline additionally reveal limitations when it comes to specificity and methodological rigor. First, systematic syntheses that particularly goal the school scholar inhabitants stay comparatively scarce, with inadequate consideration paid to life-style behaviors resembling sedentary conduct and sleep. Second, current analyses haven’t adequately synthesized the mixed results of a number of life-style behaviors throughout numerous DHI intervention codecs [].

In mild of the present analysis context and recognized limitations, this evaluation is guided by the next analysis questions: (1) What is the present state of the literature on DHIs focusing on 4 key life-style behaviors amongst faculty college students (bodily exercise, sedentary conduct, weight loss program, and sleep)? (2) What are the particular implementation methods and modalities of DHIs addressing these behaviors? (3) To what extent are DHIs efficient in influencing these 4 goal life-style behaviors amongst faculty college students? Through a complete synthesis and evaluation of related major analysis proof, this evaluation will explicitly think about the traits of school college students as “digital natives” []. The evaluation will systematically look at the varieties, capabilities, and key parts of various DHIs, and comprehensively consider their results on the 4 goal behaviors which might be intently associated to varsity scholar well being. This evaluation goals to make clear the applicability and effectiveness of DHIs inside this inhabitants, thereby offering evidence-based suggestions for optimizing DHI instruments, informing well being promotion methods in larger schooling settings, and guiding future analysis.

Methods

Search Strategy

This systematic evaluation was prospectively registered in PROSPERO (International Prospective Register of Systematic Reviews) on August 4, 2025 (registration quantity: CRD420251119078), and the reporting of the evaluation findings adheres to the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) 2020 pointers (see ). A complete literature search was carried out throughout 10 main English-language digital databases, together with Scopus, Web of Science, PubMed, ProQuest Central, and 6 databases accessed by way of the EBSCOhost platform (MEDLINE, PsycINFO, SPORTDiscus, APA PsycArticles, ERIC, and Academic Search Premier), with Google Scholar used as a supplementary search supply. In addition, the reference lists of related articles had been screened to establish doubtlessly missed research. The preliminary search was accomplished on August 5, 2025, masking research printed between January 1, 2010, and June 1, 2025, for major examine identification, and an up to date search was carried out on December 27, 2025, to seize research printed inside the newest 6 months; the identical search technique was utilized constantly throughout each searches. No printed search filters had been used, and the search technique was neither tailored from nor reused, in complete or partly, from earlier evaluations. The search technique was initially developed by the authors and subsequently peer reviewed by an skilled searcher with experience in scientific data retrieval. Beyond these approaches, no examine registries had been searched, no purposeful looking out or shopping (eg, desk of contents screening, print convention proceedings, or web site searches) was carried out, and no further data was sought by contacting authors, specialists, producers, or different related events.

The literature search technique was systematically developed in accordance with the PRISMA-S (Preferred Reporting Items for Systematic Reviews and Meta-Analyses—Search Extension) pointers to make sure transparency and reproducibility of the search course of. The technique mixed Medical Subject Headings phrases with free-text phrases and constructed key phrase mixtures round 3 core ideas: (1) intervention codecs (eg, digital well being, digital intervention, eHealth, and cell well being [mHealth]); (2) goal behaviors (eg, bodily exercise, sedentary conduct, sleep, and weight loss program); and (3) examine populations (eg, college college students, faculty college students, and undergraduate college students). Boolean operators (AND and OR) had been utilized to steadiness search sensitivity and specificity. Using Scopus for instance, the search question was as follows: TITLE-ABS-KEY (“digital health” OR “eHealth” OR “mHealth” OR “mobile health” OR “digital intervention” OR “health app”) AND TITLE-ABS-KEY (“college students” OR “university students” OR “undergraduate students” OR “young adults”) AND TITLE-ABS-KEY (“lifestyle behavior” OR “health behavior” OR “physical activity” OR “exercise” OR “diet” OR “nutrition” OR “sleep” OR “sedentary behavior”). The full English-language search phrases used throughout all databases are offered in . To guarantee complete protection, no geographical restrictions had been utilized throughout the literature search, permitting for the inclusion of related research from numerous world areas.

Inclusion and Exclusion Criteria

The examine inclusion standards had been developed in accordance with the PICOS (Population-Intervention-Comparison-Outcome -Study Design) framework, as outlined in .

The exclusion standards had been as follows: (1) The examine inhabitants was not explicitly recognized as “college students,” “university students,” or people enrolled in larger schooling establishments. (2) The nondigital parts constituted nearly all of the intervention (≥50%), or the examine relied solely on wearable units for passive behavioral monitoring with out incorporating suggestions mechanisms or energetic intervention methods. (3) The examine didn’t implement a behavioral intervention, or the intervention description lacked enough element to find out its content material and implementation procedures. (4) Studies that didn’t report any life-style conduct–associated final result measures. (5) Conference abstracts, theses, unpublished manuscripts, and different types of grey literature. (6) Full-text articles had been unavailable, or the publication was not in English.

Study Selection

All retrieved data had been imported into FinishNote 20 (Clarivate Plc) reference administration software program for duplicate elimination and standardized report numbering. Subsequently, 2 reviewers (QYZ and JJJ) independently screened titles and abstracts for preliminary eligibility. Records that handed the preliminary screening had been subjected to full-text evaluation to find out last eligibility for inclusion. To guarantee standardization and consistency within the screening course of, all reviewers obtained standardized coaching on the predefined inclusion and exclusion standards. Interrater reliability between the two reviewers was assessed utilizing the Cohen κ coefficient, yielding a worth of 0.86, which signifies a excessive stage of screening settlement. In circumstances of disagreement relating to particular person data, a 3rd reviewer (ZHY) was consulted to facilitate dialogue and obtain a last consensus.

Data Extraction and Synthesis

To guarantee standardization and consistency within the knowledge extraction course of, the analysis crew developed a structured knowledge extraction kind prematurely, masking the examine title, first creator, publication 12 months, examine area, examine design, intervention inhabitants traits, intervention protocol traits, final result measures, intervention effectiveness, and examine conclusions. The knowledge extraction kind was pilot-tested utilizing 5 research to evaluate its feasibility. During the formal knowledge extraction course of, 2 reviewers (QYZ and JJJ) independently extracted the info. In circumstances of lacking knowledge or discrepancies in interpretation, a 3rd reviewer (ZHY) was consulted to resolve disagreements. The last extracted knowledge had been consolidated right into a standardized desk and are offered in .

Quality Assessment

All included research had been initially assessed for methodological high quality utilizing the Mixed Methods Appraisal Tool (MMAT, 2018 model) to acquire an total preliminary appraisal of examine high quality. The MMAT is designed to guage 5 classes of examine designs: qualitative analysis (QR), quantitative randomized managed trials (QRCTs), quantitative nonrandomized research (QNRSs), quantitative descriptive research (QDSs), and blended strategies research (MMSs), every comprising 5 appraisal standards []. To improve specificity and methodological rigor, the Risk of Bias 2 (RoB 2) device was additional utilized to evaluate the danger of bias in QRCTs. For all different examine designs, the Joanna Briggs Institute (JBI) important appraisal instruments had been utilized. This 2-stage high quality evaluation strategy was meant to steadiness breadth and depth in methodological analysis. Quality assessments had been carried out independently by 2 reviewers (QYZ and JJJ), with discrepancies resolved by means of dialogue. To guarantee consistency, each reviewers obtained standardized coaching on the MMAT, RoB 2, and JBI important appraisal instruments and accomplished pilot scoring workouts earlier than the formal evaluation.

Results

Screening and Inclusion Results

Search and Screening Results

In this examine, a complete of 2998 data had been retrieved from 10 main English-language databases. After deduplication and preliminary title and summary screening, 273 articles had been chosen for full-text evaluation. Based on the predefined exclusion standards, exclusions had been made for the next causes: nonuniversity samples (n=68); interventions not primarily digital-based (n=20); wearable units solely, with out energetic intervention parts (n=5); absence of behavioral interventions (n=83); lack of related behavioral outcomes (n=35); and protocol or summary solely (n=19). Additionally, 3 extra articles had been recognized by means of guide reference tracing of related evaluation papers. Ultimately, 46 publications met the inclusion standards and had been included within the last evaluation, as depicted in .

‎

Quality Assessment Results

Following 2 rounds of high quality evaluation, the first-round MMAT analysis indicated that the 46 included research demonstrated an total excessive stage of methodological high quality. Specifically, 28 (61%) research had been rated as top quality, 14 (30%) as average high quality, and 4 (9%) as low high quality (see ; additionally see [,,-,,,,-]). Major methodological considerations recognized throughout the evaluation had been primarily concentrated in MMAT gadgets C4 and C5. Item C4 was primarily associated to the implementation of blinding procedures, the adequacy of final result interpretation, and the management of threat of bias, whereas merchandise C5 mirrored points resembling inadequate intervention adherence and the dearth of rigorous statistical analyses. In the second spherical of evaluation, the RoB 2 device was utilized to guage 30 QRCTs, indicating that the first sources of bias had been associated to final result measurement, deviations from meant interventions, and the dealing with of lacking final result knowledge (see ; see additionally [,,,-,-,,,-,-,,,,,-]). Concurrently, the JBI important appraisal of the remaining 16 research indicated that key components influencing examine high quality primarily included pattern representativeness, intervention adherence, and the objectivity of final result measurement (see ; see additionally [,,,,,,, -,,,,,,]).

The emergence of those methodological points may be primarily attributed to 2 components. On the one hand, the behavioral nature of DHIs makes the implementation of blinding inherently difficult, and a number of other key behavioral outcomes depend on participant self-report measures. On the opposite hand, comparatively excessive dropout charges related to DHIs contribute to points resembling low intervention adherence and elevated loss-to-follow-up charges in some research. When mixed with insufficiently rigorous statistical analyses, these challenges could lead to suboptimal dealing with of lacking knowledge or deviations from meant interventions. Although research rated as average to low high quality represent a notable proportion of the included literature, it is very important acknowledge that lots of their methodological limitations are intently associated to the inherent traits of DHIs. Moreover, many of those research primarily aimed to discover the feasibility and applicability of DHIs moderately than to supply definitive proof of intervention efficacy. Therefore, these research retain substantial worth for informing future analysis and intervention improvement. Given these concerns, no research had been excluded from this evaluation solely based mostly on methodological high quality. Instead, all eligible research had been included, and findings from risk-of-bias assessments had been systematically integrated into the narrative synthesis. This strategy permits for a complete presentation of the present proof panorama whereas explicitly figuring out each its strengths and limitations.

‎

‎

‎

Data Extraction Results

This evaluation included a complete of 46 research. The fundamental traits of the included research are summarized in , with detailed knowledge extraction outcomes offered in . Given the substantial heterogeneity among the many included research with respect to review design, goal behaviors, intervention codecs, core capabilities, and first final result measures, in addition to variations in methodological high quality, a meta-analysis was not carried out. Instead, a complete evaluation was carried out utilizing descriptive synthesis and comparative approaches. By systematically organizing and describing key traits of DHIs—together with intervention targets, participant traits, pattern sizes, codecs, capabilities, durations, outcomes, and results—this evaluation delineates the general patterns and heterogeneity throughout the discipline. Specific particulars are elaborated within the subsequent sections and illustrated by means of related tables, charts, and figures.

In phrases of annual distribution (see ), the variety of research throughout the early interval (2014-2015) was low, with only one publication per 12 months. Since 2016, the variety of publications elevated markedly, reaching a primary minor peak in 2016 (n=8), presumably related to the speedy adoption of smartphones and cell apps amongst faculty college students. From 2017 to 2020, the variety of research fluctuated between 1 and 5 yearly, sustaining an total average stage. The quantity elevated once more and stabilized in 2021-2022, declined barely in 2023, reached a second peak in 2024 (n=8), and remained excessive in 2025 (n=4). Publications from the final 5 years accounted for greater than half of all research recognized.

‎

The regional and nation distribution of the included research demonstrates a transparent geographical focus. At the regional stage, most research had been carried out in North America (n=18, 39%), adopted by Oceania (n=10, 22%) and Europe (n=9, 20%). Asia accounted for six (13%) research, whereas Africa contributed the smallest share with 3 (7%) research. At the nation stage, the United States recorded the best variety of publications (n=15, 33%), adopted by Australia (n=9, 20%). The United Kingdom, Germany, Canada, South Korea, and India every contributed 2 research. The remaining nations had been represented by a single examine, indicating a comparatively dispersed distribution past the main contributors.

The distribution of examine design varieties among the many included research exhibited a transparent structural sample. The largest proportion comprised QRCTs (n=30, 65%). This was adopted by QNRSs (n=10) and QDSs (n=4), which had been primarily used for exploratory analyses and descriptive accounts of phenomena. By distinction, QR and MMS had been represented by only one article every, accounting for lower than 2% of the full. Overall, DHI research addressing faculty college students’ life-style behaviors are predominantly quantitative, with a marked choice for QRCTs.

With respect to moral compliance, all included research adhered to related moral pointers, with all 46 (100%) explicitly reporting knowledgeable consent procedures and ethics committee approval or evaluation standing. Regarding privateness safety and knowledge safety, 24 (52%) research explicitly reported the implementation of protecting measures, together with safe server storage compliant with knowledge security requirements, encrypted knowledge transmission, knowledge deidentification, and strict entry management mechanisms. With respect to adversarial occasions and intervention-related dangers, no critical adversarial occasions had been reported throughout the included research. Only a small variety of research reported minor destructive points associated to expertise use, resembling fluctuations in intervention engagement, larger dropout charges, or decreased compliance attributable to members’ competing tutorial or private commitments. No well being dangers had been recognized that had been straight attributable to the DHIs.

Intervention Design and Implementation Results

Intervention Objectives

Among the intervention aims examined within the included research, 30 addressed bodily exercise, 26 addressed weight loss program, 10 focused sedentary conduct, and 6 focused sleep. Single-behavior interventions accounted for a big proportion of the research; nonetheless, multibehavior crossover interventions had been additionally substantial, with mixed bodily exercise and weight loss program interventions being the commonest (n=18). Notably, bodily exercise was each probably the most frequent single-behavior intervention goal and the first entry level for multibehavior mixed interventions, whereas sleep was comparatively underemphasized in intervention design.

Intervention Participants

Based on the PROGRESS-Plus (Place of Residence, Race/Ethnicity, Occupation, Gender/Sex, Religion, Education, Socioeconomic Status, Plus Other Relevant Factors) framework, a synthesis of sociodemographic traits from 46 DHI research recognized 10 major participant classes (see ), together with well being standing (n=46), age (n=45), gender/intercourse (n=45), schooling (n=41), occupation (n=39), place of residence (n=36), race/ethnicity (n=28), socioeconomic standing (n=14), social capital (n=8), and faith (n=1). The evaluation revealed the next: (1) All members had been faculty college students, predominantly aged 18-30 years, which is in keeping with typical faculty scholar demographics and confirmed no substantial deviation throughout research. (2) Most interventions focused college students with typically wholesome standing, whereas 14 out of 46 (30%) targeted on subpopulations with particular well being dangers or particular wants, resembling obese or weight problems, sleep issues, psychological stress, or disordered consuming behaviors. (3) Gender/intercourse distribution was comparatively balanced throughout research, whereas schooling and occupation exhibited restricted variability owing to the homogeneity of the examine inhabitants. (4) By distinction, PROGRESS-Plus dimensions resembling race/ethnicity, socioeconomic standing, social capital, and faith obtained notably restricted consideration, with a scarcity of systematic evaluation from a well being fairness perspective.

Intervention Sample

The pattern sizes of the included research diversified significantly. Histograms indicated that almost all research had pattern sizes concentrated beneath 200 members, with a median of roughly 95, whereas just a few research had small (<50) or extraordinarily giant (>400) samples. As proven in , box-and-whisker plots additional revealed an uneven distribution with long-tailed traits. Variations in pattern dimension had been intently related to examine design. Rigorous QRCTs usually require bigger samples to make sure statistical energy and subsequently are likely to make use of medium- to large-scale pattern sizes. By distinction, QDSs and QR are extra inclined towards small-sample explorations, generally recruiting only some dozen members, and are extra vulnerable to choice bias.

‎

Intervention Modalities

The intervention codecs within the included research fell into 3 fundamental classes. The first class, single, referred to interventions using only one digital well being expertise (n=29), resembling cell apps. The second class, a number of, concerned combining a number of digital well being applied sciences throughout the identical intervention (n=10). For instance, the TXT2BFiT program built-in cellphone calls, web sites, apps, and SMS textual content messaging concurrently to realize intervention objectives. The third class, mixed (n=7), in contrast the effectiveness of various mixtures of digital well being applied sciences, resembling a “web-based nutrition intervention only” versus a “web-based intervention combined with daily SMS text message reminders.” Regarding the forms of intervention applied sciences, these may very well be categorized into 7 teams: (1) cell apps, used 21 occasions; (2) web-based platforms, together with web sites (13 occasions), on-line boards (3 occasions), and digital studying or eHealth instruments (4 occasions); (3) cell communications, together with SMS textual content messages (11 occasions), emails (5 occasions), and cellphone calls (3 occasions); (4) social media (7 occasions); (5) wearable units (4 occasions); (6) gamification and multimedia, together with gamification and exergames (5 occasions), movies (2 occasions), and digital actuality (1 time); and (7) clever applied sciences, represented solely by synthetic intelligence (1 time). Overall, cell apps and web-based platforms had been probably the most regularly used applied sciences.

Intervention Functionalities

The technological capabilities of the DHIs included on this evaluation exhibited distinct patterns of emphasis. Educational and guidance-related capabilities predominated throughout most interventions, adopted by monitoring and prompting capabilities; against this, suggestions and interactive capabilities had been used much less regularly, whereas immersive, screening, and engagement-related capabilities had been not often integrated. Coding these interventions utilizing the Behavior Change Technique Taxonomy model 1 (BCTTv1) indicated that probably the most regularly employed methods had been “4.1 Instruction on how to perform the behavior” and “5.1 Information about health consequences,” suggesting that present DHIs primarily emphasize foundational behavioral help capabilities. Further frequency evaluation of BCT coding amongst efficient intervention research (see ) confirmed that BCTTv1 codes 4.1 (16/87, 18%), 5.1 (14/87, 16%), and a couple of.3 (13/87, 15%) constituted the core set of methods, collectively accounting for practically half of all methods utilized in efficient interventions.

Intervention Duration

The period of interventions diversified significantly throughout the included research (see ; see additionally [,,-,,,,-]), with the bulk concentrated within the short- to medium-term vary (1-16 weeks). Studies involving long-term interventions (>16 weeks) had been comparatively scarce, with solely 4 research recognized. Among these research, most integrated follow-up intervals, and medium- to long-term interventions had been usually related to extra systematic follow-up protocols. With respect to review design, randomized managed trials predominantly employed interventions of medium period (8-16 weeks). Among the QDSs (n=4) and MMS (n=1) analyzed, some research employed longer intervention durations to look at behavioral upkeep; nonetheless, these accounted for a comparatively small proportion of the proof base. Subgroup evaluation demonstrated a progressive improve within the proportion of research categorised as “effective” with growing intervention period (see ): 2 out of 4 (50.0%) for ultra-short-term (<1 week), 10 out of 16 (63%) for short-term (>1 and <8 weeks), 12 out of 18 (67%) for medium-term (8-16 weeks), and three out of 4 (75%) for long-term (>16 weeks). Notably, medium-duration interventions (8-16 weeks) not solely represented the most important proportion of the prevailing proof but additionally demonstrated each a comparatively excessive “effective” price (12/18, 67%) and a low “ineffective” price (1/18, 6%). These findings point out that present DHI analysis stays skewed towards short- and medium-term interventions, with the 8-16-week class standing out when it comes to proof quantity and the obvious stability of intervention results.

‎

Intervention Outcomes

As a results of substantial heterogeneity among the many included research with respect to final result measurement devices, final result definitions, and evaluation time factors, it was not possible to outline a unified major final result or to conduct a statistically legitimate meta-analysis. Accordingly, this evaluation adopted a descriptive synthesis framework to summarize and combine the related outcomes. The final result metrics within the included research had been categorised into 2 fundamental classes. The major outcomes targeted on life-style behaviors, together with bodily exercise (eg, exercise stage, step depend, and exercise depth), sedentary behaviors (eg, complete sedentary time and frequency of breaks from sitting or resting), weight loss program (eg, dietary high quality; consumption of fruits, greens, and sugar-sweetened drinks; vitality consumption; and dietary information), and sleep (eg, sleep high quality, period, effectivity, and severity of insomnia). These indicators straight replicate modifications in core well being behaviors ensuing from the intervention and function a key foundation for evaluating its effectiveness. Secondary outcomes, serving as supplementary indicators, had been extra numerous and encompassed bodily well being standing and psychosocial dimensions, resembling weight and physique composition (eg, weight, BMI, waist circumference, and physique fats proportion), bodily health indicators (eg, flexibility, muscle power, and cardiorespiratory health), cardiometabolic indicators (eg, blood strain, blood glucose, and blood lipid profiles), and psychological and self-perception measures (eg, self-efficacy, physique picture, and life satisfaction). Overall, present research stay primarily targeted on major outcomes, whereas secondary outcomes have expanded however proceed to exhibit restricted protection.

Intervention Effectiveness

Based on the reported impact measure varieties, impact estimates, confidence ranges (%), and CIs throughout the included research, along with a complete evaluation of the authors’ conclusions (see ), the outcomes indicated that 31 (67%) research demonstrated proof of intervention effectiveness, suggesting that DHIs are typically related to optimistic outcomes in bettering life-style behaviors amongst faculty college students. Four research reported no statistically vital results, with limitations primarily attributed to small pattern sizes or brief intervention durations. The remaining 11 research demonstrated restricted effectiveness, with enhancements noticed solely in chosen secondary outcomes or throughout short-term follow-up intervals.

Based on a complete evaluation of every behavioral area utilizing the Grading of Recommendations Assessment, Development and Evaluation (GRADE) framework, the knowledge of proof for the bodily exercise and weight loss program domains was rated as “moderate,” whereas the proof for the sedentary conduct and sleep domains was rated as “low.” With respect to proof credibility, this evaluation signifies a average stage of confidence within the total estimate that DHIs are efficient in bettering life-style behaviors amongst faculty college students. The certainty of proof in some domains was downgraded attributable to methodological limitations within the current major research, together with small pattern sizes, challenges in implementing blinding, and inconsistencies in final result evaluation instruments. Nevertheless, these GRADE assessments present an correct reflection of the present state of the proof and its total power for DHIs amongst faculty college students, thereby providing invaluable steering for interpretation and future analysis.

Discussion

Principal Findings

Discussion on Current Research Status

In phrases of temporal developments, analysis on DHIs focusing on faculty college students’ life-style behaviors has steadily emerged since 2014, expanded quickly after 2016, and reached a peak up to now 5 years []. This pattern has been pushed primarily by 4 classes of things. First, technological advances have laid a strong basis for DHIs, with the proliferation of smartphones, wearable units, and app ecosystems considerably enhancing their accessibility and operability []. Second, conceptual developments have accelerated theoretical and methodological improvements in DHIs, underscoring their distinctive benefits in facilitating behavioral enchancment []. Third, demand has elevated considerably, significantly throughout the COVID-19 pandemic, with DHIs gaining broad recognition as viable alternate options when conventional approaches had been constrained []. Fourth, useful resource funding has continued to broaden, with funding, supportive coverage frameworks, and interdisciplinary collaboration creating a positive surroundings for analysis. Overall, future analysis is predicted to shift from assessing short-term feasibility to evaluating long-term effectiveness, scalability, and the capability to accommodate personalization [,].

In phrases of spatial distribution, analysis on DHIs is predominantly concentrated in high-income nations, significantly within the United States and Australia. This focus is primarily pushed by a mixture of technological infrastructure, analysis sources, and supportive coverage environments. On the one hand, North America and Oceania initiated mHealth improvement comparatively early, benefiting from substantial technological and monetary benefits []. On the opposite hand, faculties in these areas typically possess mature well being promotion programs and well-established moral evaluation mechanisms, facilitating the implementation of intervention trials. In addition, larger ranges of well being consciousness and larger digital acceptance in Western cultures additional contribute to this sample. However, the generalizability of those findings could also be restricted when extrapolated to low- and middle-income nations. For instance, resource-constrained settings could encounter infrastructural and hardware-related boundaries, resembling uneven community protection and low charges of digital machine possession. Furthermore, substantial cross-cultural variations exist in perceptions of privateness, the position of household involvement, and prevailing well being communication practices. In the long run, cross-cultural validation and localized adaptation of DHIs needs to be strengthened [,], significantly in resource-constrained settings. Moreover, the event of low-cost, low-threshold DHI fashions needs to be explored to advance world well being fairness [,].

In phrases of inhabitants construction, present analysis on DHIs has predominantly targeted on typically wholesome faculty college students, a spotlight attributable to this group’s modifiable well being behaviors and susceptibility to environmental influences. However, some research have prolonged to particular populations, together with faculty college students with obese or weight problems, people in danger for consuming issues [,], and college students experiencing sleep issues or psychological stress [,]. This differentiation technique is partly motivated by the truth that particular populations face larger well being dangers, thereby growing the potential advantages and scientific worth of interventions [,]. It additionally aligns with the necessity for precision interventions and stratified administration. However, current analysis has not but sufficiently examined variations in engagement ranges amongst college students from numerous sociodemographic backgrounds. Limited consideration to components resembling socioeconomic standing and entry to digital units could lead to disproportionate advantages for college students with larger monetary or digital sources, whereas these experiencing financial constraints or restricted machine entry could also be marginalized within the intervention course of. Accordingly, future analysis is prone to advance alongside 2 complementary instructions: first, persevering with large-scale research focusing on basic undergraduate populations to evaluate the generalizability of interventions; and second, strengthening focused interventions for high-risk teams whereas prioritizing the discount of participation boundaries amongst college students from numerous backgrounds []. Such efforts could drive the event of DHIs towards larger refinement, fairness, and personalization.

Discussion on Intervention Implementation

In phrases of intervention aims, bodily exercise and weight loss program are the two life-style conduct classes receiving probably the most analysis consideration [], whereas sedentary conduct and sleep are comparatively underrepresented. Both single-behavior and multibehavior mixed interventions coexist. This sample is primarily influenced by a number of components. First, bodily exercise and weight loss program are straight related to weight administration, vitality steadiness, and metabolic well being—core variables that have an effect on faculty college students’ bodily health and power illness threat []. Related measures (eg, step depend, vitality consumption) are extra simply quantifiable and standardized, making them extra doubtless targets for intervention. By distinction, sedentary conduct and sleep, regardless of their acknowledged significance [], pose technical and operational challenges for DHIs, together with measurement complexity and delayed suggestions on intervention results [], contributing to a relative paucity of analysis [,]. Current DHIs reveal restricted effectiveness in lowering sedentary time amongst faculty college students [], whereas sleep interventions, though promising, stay understudied and predominantly give attention to insomnia aid [,]. Second, intervention methods replicate researchers’ evaluation of behavioral variability: bodily exercise and weight loss program exhibit a large window for controllability and modification, whereas sedentary conduct usually happens in tutorial or leisure contexts, complicating speedy adjustment by way of a single approach. The prevalence of single-behavior interventions is attributed to their suitability for early exploratory phases, permitting simpler management of variables and validation of intervention results. Conversely, the rise in multibehavior interventions displays the aggregation of life-style dangers amongst faculty college students, which complicates reaching enough well being advantages by means of modifications in a single conduct. Notably, the mix of bodily exercise and weight loss program is probably the most frequent, reflecting the need for built-in interventions focusing on weight administration and vitality metabolism []. Overall, future analysis is predicted to more and more undertake multibehavior approaches, integrating behavioral science theories and technological instruments to develop synergistic interventions that deal with the complexity of life-style dangers.

In phrases of intervention modalities, an evolutionary pattern is obvious, progressing from single to a number of codecs and from low to excessive ranges of interplay, pushed by the mixed forces of technological development, consumer demand, and intervention science. In early research, SMS textual content messages and emails had been the predominant types of DHIs [], owing to their low technological threshold, ease of deployment, and minimal value, which made them appropriate for speedy implementation in resource-limited contexts [,]. However, these approaches primarily concerned 1-way data supply, lacked personalization and real-time interplay, and had been inadequate in sustaining consumer engagement. With the widespread adoption of smartphones and the maturation of the app ecosystem, cell apps have steadily turn into the mainstream type of DHIs. These apps are extremely built-in and interactive, able to incorporating a number of capabilities resembling goal-setting, suggestions, reminders, and knowledge monitoring [], aligning with faculty college students’ high-frequency cell utilization habits and considerably enhancing the intervention expertise and engagement []. Web platforms retain benefits in scalability however are considerably much less user-friendly and fewer efficient in delivering push notifications in contrast with apps [,]. In addition, the mixing of social media and wearable units will increase the interactivity and contextual adaptability of DHIs [], additional enhancing behavioral monitoring and the availability of speedy suggestions [-]. Future developments are anticipated to emphasise technological convergence and clever improvement. On the one hand, mixtures of a number of codecs (eg, apps, social platforms, and gamification) will turn into more and more prevalent to handle the multidimensional wants of behavioral interventions []. On the opposite hand, personalised interventions leveraging synthetic intelligence, digital teaching, and immersive experiences (eg, augmented actuality/digital actuality) are anticipated to emerge as key analysis instructions [,], shifting DHIs from being information-driven to experience-driven and in the end facilitating sustained conduct change.

In phrases of intervention capabilities, present DHIs are predominantly characterised by schooling, steering, monitoring, and prompting parts [], indicating that these interventions primarily emphasize data supply and fundamental conduct administration. This design strategy is partly pushed by the substantial demand for well being information and abilities amongst faculty college students, with schooling and steering capabilities facilitating enhancements in cognition and self-efficacy. Concurrently, monitoring and prompting capabilities leverage expertise to allow knowledge recording and behavioral reinforcement, thereby selling the initiation of goal behaviors within the brief time period []. However, high-engagement options resembling suggestions, interplay, and gamification-based incentives stay underutilized [,], suggesting that DHIs usually lack deep personalization and social help parts [], which can be a important issue limiting long-term consumer engagement and intervention effectiveness. In phrases of future developments, the convergence of conduct change theories (eg, Capability-Opportunity-Motivation-Behavior [COM-B], behavioral economics) with clever algorithmic purposes is predicted to drive the evolution of DHI performance towards larger personalization, interactivity, and emotional engagement []. For instance, synthetic intelligence–pushed real-time suggestions might improve intervention adaptability, digital communities might strengthen social help, and gamification mechanisms coupled with reward programs might foster intrinsic motivation. Such developments are prone to not solely improve intervention engagement but additionally considerably enhance behavioral upkeep, fostering a gradual shift from data supply–oriented DHIs to approaches that place larger emphasis on consumer expertise and social interplay.

Discussion on Intervention Effectiveness

Overall, 31 of 46 (67%) research reported efficient outcomes (sure), indicating the excessive feasibility and appreciable potential of DHIs in bettering the life-style behaviors of school college students [-,]. However, a subset of research yielded insignificant (no) or restricted (restricted) results, which may be examined from a number of analytical dimensions. First, inadequate refinement and lack of theoretical underpinning in intervention design signify key components constraining effectiveness. In a number of circumstances, interventions lacked express theoretical frameworks for conduct change, relying predominantly on data supply. Such approaches usually did not sufficiently stimulate participant motivation or reinforce conduct upkeep, resulting in short-term positive factors that had been troublesome to maintain []. Second, current intervention research typically lack strong validation of long-term results. Most research are restricted to durations of 8-16 weeks and embrace inadequate follow-up, which constrains the power to confirm the sustainability and stability of behavioral modifications []. As a consequence, the long-term worth and sturdiness of DHIs stay troublesome to evaluate adequately. Third, intervention effectiveness seems to be strongly influenced by participant adherence. Analyses of engagement-related metrics point out that larger ranges of consumer engagement, compliance, and intervention consistency are typically related to extra favorable behavioral and scientific outcomes. By distinction, research characterised by excessive dropout charges usually rely predominantly on 1-way data supply, with restricted alternatives for suggestions and interplay. Fourth, the kind of goal conduct and related measurement challenges additionally contribute to those outcomes. Compared with bodily exercise, the intervention results on weight loss program, sedentary conduct, and sleep had been extra susceptible to exterior environmental influences (eg, tutorial workload, dietary contexts), and measurement instruments relied predominantly on self-reporting, thereby growing bias and uncertainty. Taken collectively, variations in intervention design, technological software, and behavioral traits collectively contribute to the substantial heterogeneity noticed in intervention outcomes [].

To achieve a deeper understanding of variations in intervention effectiveness, the COM-B framework may be utilized as a scientific analytical device []. (1) Within the “Capability” dimension, most interventions primarily enhanced faculty college students’ health-related information by means of instructional content material and steering supplies. Examples included the availability of weight loss program pointers, train plans, and sleep regulation methods designed to extend members’ consciousness of the significance of wholesome behaviors. However, these enhancements usually remained on the cognitive stage, with restricted emphasis on the event of sensible behavioral abilities. Specific parts, resembling weight loss program substitution choices, situational coping methods, or versatile train planning, had been regularly absent. In addition, some research didn’t present satisfactory help for knowledge interpretation, which restricted members’ capacity to translate behavioral monitoring knowledge into actionable steps []. (2) Within the “Opportunity” dimension, DHIs typically depend on digital platforms to create enabling behavioral circumstances, resembling purpose monitoring, reminder capabilities, and on-line useful resource sharing, which can theoretically cut back psychological boundaries to conduct enactment. However, the structuring of alternatives inside real-world contexts stays insufficiently optimized. Some interventions don’t adequately account for the distinctive time pressures and contextual constraints skilled by faculty college students on campus. For instance, methods aimed toward lowering sedentary conduct usually stay restricted to generic standing reminders, with out adaptation to classroom environments or frequent examine areas, thereby constraining alternatives for sustained conduct change. Furthermore, though some interventions try to include social help mechanisms (eg, neighborhood interactions or peer challenges), the depth and high quality of participant engagement are typically restricted. These interactions regularly contain 1-way data transmission, with restricted capability to foster emotional connection or efficient behavioral modeling. (3) Within the “Motivation” dimension, current interventions primarily emphasize the stimulation of extrinsic motivation by means of short-term incentives, resembling point-based rewards and job completion reminders. While such methods could promote preliminary engagement, they often lack mechanisms for the sustained cultivation of intrinsic motivation. Specifically, many interventions haven’t successfully supported faculty college students in growing a way of self-worth derived from continued engagement in wholesome behaviors. In addition, methods aimed toward enhancing optimistic emotional experiences are not often integrated. For instance, gamification designs usually stay confined to superficial point-based programs, with restricted capability to stimulate members’ sense of exploration, mastery, or accomplishment. Additionally, inadequate personalization of suggestions seems to considerably constrain the upkeep of motivation over time. Participants usually obtain generic informational messages moderately than well timed, individualized suggestions intently aligned with their precise behavioral efficiency.

In abstract, present DHIs predominantly undertake a “technology-driven” or “utility-oriented” design logic, with a major emphasis on practical implementation and surface-level engagement metrics. Because of the restricted integration of conduct change principle, such interventions are likely to exhibit constrained effectiveness in sustaining long-term outcomes. By distinction, theory-driven interventions—resembling these grounded within the COM-B framework—lengthen past short-term conduct initiation, emphasizing the synergistic improvement and dynamic help of functionality, alternative, and motivation. Through structured and phased behavioral help methods, such interventions could facilitate the institution of putting up with foundations for sustained change throughout cognitive, skill-based, environmental, and emotional dimensions []. As a consequence, long-term conduct upkeep could turn into extra attainable []. Future analysis ought to additional place conduct change principle as a central guideline in intervention design, transferring past the view of expertise as a standalone device and as a substitute embedding it organically inside help programs centered on conduct change mechanisms.

Strengths and Limitations

This examine is among the many first English-language evaluations to systematically combine a number of types of DHIs and a number of life-style conduct domains inside a core inhabitants of school college students, and it presents the next strengths. First, the examine design strictly adheres to PRISMA 2020 and was preregistered on PROSPERO. The systematic search spanned 10 main worldwide databases, guaranteeing the comprehensiveness and representativeness of the proof base. Second, by specializing in faculty college students as “digital natives,” this examine systematically analyzes intervention traits throughout 4 well being conduct domains—bodily exercise, sedentary conduct, weight loss program, and sleep—thereby addressing limitations of prior evaluations that emphasised a single conduct or device. Third, drawing on the COM-B framework, this examine examines the mechanisms of DHIs throughout the Capability, Opportunity, and Motivation dimensions and identifies key bottlenecks in intervention methods—resembling restricted technological performance, suboptimal ecological adaptability, and inadequate motivational activation—thereby offering each theoretical help and sensible steering for the long run design and optimization of DHIs for school college students.

Although this evaluation endeavored to include the prevailing literature as comprehensively as attainable, a number of limitations stay. First, the geographical distribution of the included research was uneven, with a heavy focus in high-income nations—significantly North America and Australia—which constrains the worldwide applicability of the findings; particularly, their generalizability to varsity college students in low- and middle-income nations requires empirical verification. Second, many research employed small samples, brief intervention durations, and restricted follow-up, and a few lacked strong management teams or satisfactory randomization, thereby weakening the steadiness of impact estimates and the power of causal inference. In addition, DHIs had been usually comparatively homogeneous, with restricted multidimensional interactivity and personalization; blinding procedures had been troublesome to implement; and dangers of bias arose in adherence evaluation and final result measurement. Therefore, future analysis ought to strengthen pattern representativeness, improve intervention refinement, and enhance methodological rigor to extend the exterior validity and sensible utility of the findings.

Implications and Recommendations

Recommendations for Policy and Practice

To totally understand the potential of DHIs whereas guaranteeing the sustainability and broad accessibility of intervention results, systematic enhancements in coverage design and implementation pathways are required. First, faculty college students needs to be explicitly integrated into nationwide and regional digital well being methods to facilitate a shift from conventional well being schooling towards built-in digital platforms, and higher-education establishments needs to be inspired to develop or undertake scientifically grounded, standardized instruments with clearly articulated mechanisms of motion. Second, localized improvement of intervention content material and performance needs to be supported, with consideration to adaptability throughout behavioral domains, cultural contexts, and scholar wants, thereby advancing refined, human-centered design with respect to technological thresholds, knowledge safety, and personalised suggestions. Third, intervention observe ought to strengthen college students’ energetic engagement and set up feedback-driven, behavior-reinforcing, and peer-support mechanisms to boost sustained use and intrinsic motivation. In parallel, cross-departmental cooperation mechanisms needs to be established on the faculty stage, and well being interventions needs to be embedded inside curricula, psychological help programs, and campus service sources to kind a synergistic help community. Finally, on the coverage stage, moral oversight and effectiveness analysis of DHIs applications needs to be strengthened, and an evidence-based analysis framework for DHIs needs to be established to make sure fairer, extra adaptable, and simpler interventions for school college students.

Recommendations for Future Research

This examine signifies that present analysis on DHIs for school college students stays constrained by unrepresentative samples, single-focus intervention content material, and unclear technological mechanisms; future work needs to be refined and deepened within the following respects. First, geographical and cultural range needs to be expanded, prioritizing research from low- and middle-income nations, diversified higher-education establishment varieties, and numerous social teams to boost the exterior validity of the findings. Second, the design and analysis of multibehavior-integrated interventions needs to be strengthened by transferring past single-behavior paradigms and inspecting behavioral synergies and optimum mixtures of intervention parts. Third, higher-quality examine designs—resembling QRCTs, MMSs, and long-term follow-up—needs to be employed to strengthen causal inference and the sustainability of intervention results. Fourth, theoretical improvement and empirical testing of intervention mechanisms needs to be strengthened by grounding analyses in conduct change principle to make clear how expertise enhances Capability, Opportunity, and Motivation, and to advance DHIs from merely offering technical capabilities to making a supportive ecosystem conducive to sustained conduct change. Finally, future research ought to emphasize the evaluation of intervention fairness, systematically account for potential moderators resembling gender, socioeconomic standing, and psychological standing, and establish subgroups with restricted responsiveness, thereby offering a strong proof base for establishing a extra inclusive and adaptive DHI mannequin for school college students.

Conclusions

This evaluation addresses a niche within the literature by focusing particularly on faculty college students, a gaggle usually neglected in analysis that usually facilities on broader grownup populations. Unlike prior evaluations that primarily look at a single life-style conduct, this examine adopts a extra holistic strategy by integrating a number of behaviors and evaluating a variety of DHIs with numerous modalities and functionalities. These findings present invaluable insights for refining future DHIs focusing on faculty college students and contribute to the event of simpler well being promotion methods in larger schooling. Although DHIs present potential for bettering life-style behaviors, their long-term effectiveness stays unsure. Current interventions face a number of limitations, together with a slim behavioral focus, fundamental technological performance, and restricted adaptability to numerous contexts, all of which can prohibit long-term engagement and personalised responsiveness. Moreover, many interventions don’t totally account for variations in useful resource entry and particular person conduct change pathways, doubtlessly limiting their applicability and fairness. Future analysis ought to prioritize integrating a number of behaviors, enhancing consumer engagement, bettering contextual adaptability, and increasing technological accessibility. Long-term research and equity-focused evaluations are important for strengthening the proof base and guaranteeing the sustainability and inclusivity of well being conduct change amongst faculty college students.