Physical activity (PA), dietary behaviors, and sleep are collectively termed “the big three pillars of an individual’s health” [1]. Regular engagement in PA, healthy eating habits, and high-quality sleep are critical for enhancing brain growth and the central nervous system [2,3], promoting cognitive performance [4], improving musculoskeletal health [5], reducing symptoms of mental disorders (eg, anxiety and depression) [6], and lowering the likelihood of developing obesity and noncommunicable diseases (eg, cardiovascular diseases) [7]. Nonetheless, people’s lifestyles have become increasingly characterized by physical inactivity and sedentary behaviors [6], unhealthy dietary habits [8], and disrupted sleep routines [9], all of which can adversely affect lifelong health trajectories [10]. This underscores the importance of establishing healthy lifestyle behaviors from an early age (ie, the preschool years), when behavior patterns are still developing and more likely to persist throughout the lifespan [11].

Given the high attendance rate of preschoolers (approximately 50%-90%) in Early Childhood Education and Care (eg, long day care, nurseries, kindergartens, and family day care), along with the substantial amount of time most children spend in these settings (on average 25-35 hours/week), they represent an ideal context for delivering healthy lifestyle strategies for preschoolers [12]. A recent study indicated that a school-based intervention effectively increased preschoolers’ moderate-to-vigorous physical activity (MVPA) and total PA, while decreasing sedentary time at the 8-week posttest [13]. The results of a systematic review evaluating the effectiveness of childcare-based interventions on preschoolers’ healthy eating showed that most of the reviewed studies (8 out of 14) reported significant improvements, including increased intake of fruits and vegetables and decreased total fat consumption [14]. A meta-analysis further reported that school-based interventions targeting preschoolers’ 24-hour movement behaviors reduced sedentary behavior outcomes and increased sleep duration [15]. Arguably, healthy lifestyle interventions implemented in Early Childhood Education and Care settings could be an effective approach to promoting preschoolers’ healthy lifestyles.

However, recent evidence has highlighted the weaknesses of interventions carried out in Early Childhood Education and Care settings. First, kindergartens demonstrated a lack of fidelity in intervention delivery and faced difficulties in implementation due to high staff turnover rates [16]. In addition, teachers were reluctant to participate, as the intervention diverted attention from other educational activities [17]. Finally, differences in individual teaching styles and interpretations of the intervention led to inconsistent adherence to intervention protocols [18].

In response, extensive studies have highlighted that parental involvement in interventions may be highly valuable for shaping preschoolers’ PA, dietary behaviors, and sleep [19]. In the early years of life, preschoolers are still developing autonomy and rely on parental supervision [20], while parents strongly influence children’s PA, dietary behaviors, and sleep through their regulations and recommendations [21]. However, due to the challenges associated with face-to-face parent-based interventions (eg, time consumption and travel costs) [22], eHealth interventions—defined as the use of the internet and mobile technologies to deliver health-related information [23]—offer appealing alternatives because of their accessibility and lower costs [24].

A previous systematic review highlighted that existing literature examining the impact of parent-based eHealth interventions on preschoolers’ PA, dietary behaviors, and sleep either focused on a single variable among these behaviors or did not balance the intervention content modules (ie, diet-related modules were more extensive than those related to PA and sleep) [25]. Additionally, the sequence in which modules were delivered was often overlooked. Familiarity with later-distributed modules at the posttest may have led to more favorable outcomes for these modules compared with earlier ones [25]. The systematic review further indicated that empirical relationships identified in one culture may stem from its specific context and may not be transferable to other countries [25]. Considerable uncertainty still exists regarding the effectiveness of parent-based eHealth interventions on preschoolers’ PA, dietary behaviors, and sleep in China. Therefore, high-quality, robustly designed parent-based eHealth interventions with balanced modules and well-structured content delivery are needed to determine their effectiveness on Chinese preschoolers’ PA, dietary behaviors, and sleep.

This study aimed to investigate the effectiveness of a parent-based eHealth intervention on PA, dietary behaviors, and sleep among preschoolers raised in a Chinese cultural context.

The protocol for this study has been published [26]. This was a single-blinded randomized controlled trial with 2 parallel arms, consisting of a 12-week intervention followed by a 12-week follow-up. It was designed to promote preschoolers’ PA, dietary behaviors, and sleep by providing parents with relevant health information and recommendations through an eHealth modality (ie, videos and WeChat [Tencent Holdings Limited], which is similar to WhatsApp) and by motivating parents to create a family environment with healthy regulations in which their children live. Data were collected at baseline, 12 weeks postbaseline, and 24 weeks postbaseline in 4 kindergartens in Guiyang, Guizhou Province, Mainland China, with assistance from Sun Yat-sen University and Guiyang Preschool Education College, between October 2023 and July 2024. Multimedia Appendix 1 presents the CONSORT (Consolidated Standards of Reporting Trials) checklist.

The entire recruitment process was conducted within the school setting. Kindergarten teachers sent messages to parents in their respective WeChat class groups, providing a brief introduction to the study’s objectives and significance. Parents were then asked to respond with either “yes” or “no” to indicate their willingness to participate. Face-to-face presentations outlining the objectives and procedures of the intervention were conducted in the 4 kindergartens for parents who agreed to participate (ie, those who replied “yes” in the WeChat group). The inclusion criteria were as follows: (1) parents were over 21 years old and had children aged 3-6 years; (2) parental commitment to participate in the full 24-week intervention; (3) access to mobile technology; (4) parents and children were required to be healthy (defined as a state of physical, mental, social, intellectual, and emotional well-being, and the absence of disease) and able to engage in normal PA; and (5) parents had to reside with the participating child for at least 4 days/week.

Children were excluded from the study if they (1) had a diagnosed neurological or physical disability; (2) had a parent with a serious physical or psychological condition that would make participation too demanding for the family; or (3) had special dietary requirements or allergies that would necessitate tailoring the intervention or could be adversely affected by it.

This intervention study was approved by the Research Ethics Committee of Hong Kong Baptist University (approval number SOSC-SPEH-2022-23_115) and registered in the ClinicalTrials.gov Protocol Registration and Results System under identifier NCT06025019. All preschoolers and their parents were fully informed of the potential risks of the experiment. Eligible participants were asked to add the researchers on WeChat and to sign the informed consent form at the end of the parent meetings. The consent form outlined the points detailed in Textbox 1.

After recruitment, eligible participants were randomized into either the eHealth intervention group or the control group at a 1:1 ratio, with allocation concealment ensured using a computer-generated sequence (SPSS version 27; IBM Corp). The entire allocation process was overseen by the statistician. To maintain strict adherence to the study protocol, only the 2 researchers (ZP and PWCL) responsible for implementing the intervention were informed of group allocation, while the participants remained blinded.

A priori analysis was conducted using G*Power 3.1.9.6, with an F-test ANOVA repeated-measures, within-between interaction design [27]. A recent systematic review and meta-analysis reported a small yet statistically significant effect size for interventions aimed at enhancing PA in children under 5 years old [28]. Additionally, intervention studies targeting improvements in diet and sleep quality in young children have shown small to medium effect sizes [29,30]. Therefore, a small effect size was adopted in this study. A minimum sample size of 82 parent-child dyads/group (intervention and control) was required to detect a small effect size (Cohen d=0.3) with a power of 0.80 and an α level of .05. Considering a potential 20% attrition rate [31], a total of 206 parent-child dyads were needed (103 in each group).

The intervention was implemented in accordance with the previously published protocol [26]. Briefly, it was grounded in social cognitive theory, which posits that learning is a social phenomenon in which individuals acquire knowledge by observing, imitating, and modeling the behaviors of others [32]. For preschool-age children, observing and mimicking parental behaviors is a fundamental way of learning, as they absorb information from the family environment to develop social, cognitive, and emotional skills [20]. Therefore, a key component of this intervention was to motivate parents to model healthy PA, dietary behaviors, and sleep routines, while fostering a home environment with supportive regulations in these areas.

Social cognitive theory proposes that an individual’s behavior can be modified through the interaction of personal, environmental, and behavioral factors [33]. Personal factors refer to an individual’s self-efficacy in carrying out a behavior, shaped by personality, knowledge, beliefs, self-perceptions, and expectations. Environmental factors are defined as supportive contexts that facilitate the performance of a behavior. Behavioral factors refer to an individual’s response once a behavior has been performed. Together, these 3 factors outline the key intervention components needed to induce behavioral change.

Additionally, social cognitive theory outlines 4 processes for behavior change: attention, retention, reproduction, and motivation [34]. If any of these processes are absent, learning and adoption of new behaviors will not occur [34]. These 4 processes inform the sequence that interventions should follow to induce behavioral change.

The eHealth intervention comprised 12 interactive modules focused on preschoolers’ PA, dietary behaviors, and sleep, with each theme consisting of 4 modules. These 12 modules were rotated weekly with a theme over the 12-week period. Participants were organized into 26 WeChat groups, each containing 8 members, based on the optimal group size for interaction [35]. Trained facilitators—kindergarten teachers with expertise in PA, diet, and sleep—were recruited for each group and blinded to the study objectives. Their responsibilities included sending videos, facilitating interactions, answering questions, helping parents set goals for their children, and providing feedback, as well as revising goals. Parents received 2 self-monitoring SMS text messages/week: the first, sent midweek, requested feedback on goal progress. A response of “N” triggered a skill-building suggestion, while a response of “Y” received positive reinforcement. The second message, sent on weekends, assessed goal attainment before introducing the new weekly theme.

Based on the sequence outlined in social cognitive theory, participants received the following content each week: (1) Attention—matched with personal factors—emphasizes that individuals cannot learn unless they pay attention to what is happening around them [36]. To address this, attention was promoted by providing printed educational materials and videos through WeChat [37]. At the beginning of each week, parents received a pre-prepared poster and a 3-minute video related to the weekly theme. These videos, recorded by the researchers and structured around behavior change techniques, included preschoolers’ lifestyle recommendations from the World Health Organization, benefits of the target behaviors, common barriers, goal setting, and feedback. (2) Retention—aligned with environmental factors—emphasizes that individuals must not only recognize observed behaviors but also remember them for future application [36]. Therefore, retention was addressed through interactions with other participants in WeChat groups, which helped parents recall the content of the videos and educational materials. Facilitators encouraged communication within these groups, where parents shared their children’s activities, meal photos, and strategies for reducing sedentary behavior. The groups were monitored by the researchers to ensure that discussions aligned with evidence-based guidelines. Parents who did not participate in group exchanges received private reminder messages from the facilitator via WeChat. (3) Reproduction, which is matched with behavioral factors, refers to the individual’s ability to be physically and mentally capable of performing the observed behaviors [36]. Reproduction was supported through goal setting, action planning, and problem-solving around barriers. Facilitators worked with parents privately to establish individualized goals related to PA, dietary behaviors, and sleep for their children, and to refine these goals using the SMART (Specific, Measurable, Achievable, Relevant, and Time-bound) framework. Parents also had regular opportunities to interact with facilitators to address challenges and adjust their plans as needed. (4) Motivation, which is matched with behavioral factors, refers to the individual’s interests and sense of achievement in relation to the tasks and activities being implemented [36]. Motivation was fostered by creating cognitive dissonance, prompting parents to reflect on their children’s current behaviors and to identify the positive outcomes and expectations associated with adopting the planned behaviors. Facilitators also provided 2 types of individualized feedback—based on both group interactions and progress toward goal achievement. First, graphical visualizations of reported PA, diet, and sleep were shared to motivate compliance with World Health Organization recommendations. Second, a “traffic light” poster was used to indicate compliance (green: met; yellow: close; and red: not met). For goals not met, facilitators provided tailored solutions, tips, and reinforcement strategies privately via WeChat.

Baseline characteristics of the intervention and control groups were compared using the chi-square test for categorical variables (preschoolers’ gender, number of children with overweight and obesity, and parents’ gender, educational level, income range, marital status, number of children in the household, and number of parents with overweight and obesity) and independent t tests for continuous variables (preschoolers’ and parents’ age, height, weight, and BMI). An intention-to-treat analysis was conducted. Data from the wGT3X-BT ActiGraph (ie, preschoolers’ PA, sleep quality, and sleep duration) and parent-reported questionnaires (ie, preschoolers’ dietary behaviors, sleep problems, and screen time) were analyzed using generalized estimating equations. Group and time were included as main effects, with demographic variables as covariates (preschoolers’ gender, age, height, weight, BMI; parents’ age, height, weight, gender, educational level, income range, marital status, number of children in the household, and BMI). This analysis examined the effects of the parent-based eHealth intervention on outcomes, including preschoolers’ PA, sedentary behaviors, dietary behaviors, sleep problems, sleep latency, sleep efficiency, sleep duration, and screen time. Pairwise comparisons with Bonferroni adjustment were conducted to examine differences between the intervention and control groups whenever a significant group × time interaction was observed. Missing values were not imputed, as generalized estimating equations provide a regression framework that appropriately handles missing data and does not require the assumption of normally distributed data [47]. The magnitude of the parent-based eHealth intervention on outcomes was expressed as effect size (Cohen d), calculated using effect size calculators based on the formula provided by Sullivan and Feinn [48]. Effect sizes were interpreted as small (0.2), medium (0.5), and large (0.8) [48].

Figure 1 shows the flow of participants through the study. Baseline data for the parent-child dyads are presented in Table 1. A total of 327 parents attended the 4 face-to-face meetings, of whom 237 were eligible, signed informed consent, and were randomly assigned to the intervention (n=120) or control group (n=117). In the intervention group, the mean age of preschoolers was 4.51 (SD 0.72) years, and 67 (55.8%) children were boys. The mean age of parents in the intervention group was 34.28 (SD 4.74) years, and 93 parents (77.5%) were mothers. In the control group, preschoolers had a mean age of 4.31 (SD 0.70) years, with 69 (59%) children being boys. Parents in the control group had a mean age of 34.88 (SD 3.99) years, and 78 (66.7%) parents were mothers. All demographic characteristics were similar between the 2 groups, except for parents’ height (P=.005) and the number of parents with overweight/obesity (P=.045).

Descriptive information and changes over time are presented in Tables 2 and 3. Differences between the 2 groups are shown in Table 4.

Compared with baseline, both the intervention and control groups showed significant within-group changes in LPA, MPA, VPA, total MVPA, and average MVPA at posttest. The trends indicated a significant decrease in time spent in LPA (intervention: −1003.38 minutes/week, 95% CI −1150.10 to −856.66, P<.001; control: −745.95 minutes/week, 95% CI −936.82 to −555.09, P<.001) and MPA (intervention: −255.34 minutes/week, 95% CI −330.57 to −180.10, P<.001; control: −156.10 minutes/week, 95% CI −256.44 to −55.76, P<.001), alongside an increase in MVPA (intervention: 164.48 minutes/day, 95% CI 142.50-186.45, P<.001; control: 150.19 minutes/day, 95% CI 119.67-180.71, P<.001) and VPA (intervention: 1174.72 minutes/week, 95% CI 1008.57-1340.86, P<.001; control: 1030.86 minutes/week, 95% CI 809.58-1252.13, P<.001).

Compared with the control group, preschoolers in the intervention group showed a significant reduction in time spent in LPA (adjusted mean difference −151.61 minutes/week, 95% CI −418.18 to −96.67, P=.002, Cohen d=0.28) and an increase in time spent in VPA (adjusted mean difference 138.47 minutes/week, 95% CI 117.61-183.10, P=.03, Cohen d=0.23) at posttest. At follow-up, there were no significant between-group differences in LPA (adjusted mean difference −1.96, 95% CI −3.23 to 23.07, P=.11) or VPA (adjusted mean difference 10.01, 95% CI −10.8 to 17.01, P=.66). No significant between-group differences were observed for other PA-related outcomes at either posttest or follow-up (MPA: P=.06; TMVPA: P=.67; and AMVPA: P=.36).

No significant between-group differences in dietary behaviors were observed at either time point (enjoyment of food: P=.38; emotional overeating: P=.58; satiety responsiveness: P=.74; slowness in eating: P=.23; desire to drink: P=.61; food fussiness: P=.22; emotional undereating: P=.14; and food responsiveness: P=.50). Within-group analyses showed substantial changes in enjoyment of food and satiety responsiveness at posttest and follow-up in the intervention group, whereas no changes were observed in the control group. Specifically, the mean score for enjoyment of food in the intervention group increased from 3.30 (SD 0.07) at baseline to 3.42 (SD 0.08) at posttest (adjusted mean difference 0.12, 95% CI 0.10-0.34, P=.049) and further to 3.51 (SD 0.08) at follow-up (adjusted mean difference 0.21, 95% CI 0.02 to 0.41, P=.023). By contrast, the mean score for satiety responsiveness decreased from 3.06 (SD 0.06) at baseline to 2.86 (SD 0.06) at posttest (adjusted mean difference −0.21, 95% CI −0.39 to −0.01, P=.03) and to 2.84 (SD 0.06) at follow-up (adjusted mean difference −0.21, 95% CI −0.40 to −0.03, P=.007).

Significant within-group changes in preschoolers’ sleep-related outcomes were observed in both groups at posttest, except for sleep duration. Both the intervention and control groups showed a reduction in sleep problem scores at posttest (intervention: adjusted mean difference −2.38, 95% CI −4.76 to −0.0006, P=.049; control: adjusted mean difference −3.27, 95% CI −5.42 to −1.12, P<.001), with no significant between-group differences at either assessment (P=.67).

Although both the intervention and control groups showed a decreasing trend in sleep latency at posttest (intervention: adjusted mean difference −27.80 minutes/day, 95% CI −33.07 to −22.53, P<.001; control: −26.05 minutes/day, 95% CI −31.22 to −20.89, P<.001) and follow-up (intervention: −26.63 minutes/day, 95% CI −32.05 to −21.20, P<.001; control: −18.43 minutes/day, 95% CI −25.65 to −11.20, P<.001), the intervention group exhibited a significantly greater reduction than the control group at both posttest (adjusted mean difference −21.04 minutes/day, 95% CI −16.07 to −6.00, P=.005, Cohen d=0.78) and follow-up (adjusted mean difference −9.86 minutes/day, 95% CI −15.02 to −4.71, P<.001, Cohen d=0.55).

Both the intervention and control groups showed a significant increase in sleep efficiency at posttest (intervention: adjusted mean difference 8.37%, 95% CI 1.47-6.27, P=.006; control: 3.87%, 95% CI 1.47-6.27, P<.001) and follow-up (intervention: 6.89%, 95% CI 4.32-9.47, P<.001; control: 2.94%, 95% CI 0.79-5.08, P=.001). The improvement in sleep efficiency was more pronounced in the intervention group at both posttest (adjusted mean difference 4.61%, 95% CI 4.29-9.72, P<.001, Cohen d=0.34) and follow-up (adjusted mean difference 4.07%, 95% CI 1.53-6.61, P<.001, Cohen d=0.30) compared with the control group.

Considering within-group changes, preschoolers in the intervention group showed a decline in screen time on both weekdays (adjusted mean difference –17.23 minutes, 95% CI –33.44 to –1.02, P=.027) and weekends (adjusted mean difference –60.72 minutes, 95% CI –89.23 to –32.20, P<.001). By contrast, preschoolers in the control group showed minimal changes in screen time on weekdays (adjusted mean difference –0.01 minutes, 95% CI –15.34 to 15.36, P>.99) and weekends (adjusted mean difference –0.19 minutes, 95% CI –29.85 to 29.48, P>.99). Although the intervention group maintained a reduced screen time at follow-up on weekdays (adjusted mean difference –4.52 minutes, 95% CI –20.82 to 11.78, P>.99) and weekends (adjusted mean difference –9.86 minutes, 95% CI –36.79 to 17.08, P=.55), these changes were not statistically significant compared with baseline.

Compared with the control group, preschoolers in the intervention group showed a decrease in screen time at posttest on weekdays (adjusted mean difference –16.42 minutes, 95% CI –30.83 to –2.01, P=.01, Cohen d=0.25) and weekends (adjusted mean difference –73.88 minutes, 95% CI –98.48 to –49.28, P<.001, Cohen d=0.46). At follow-up, the intervention group showed reductions in screen time on weekdays (adjusted mean difference –7.67 minutes, 95% CI –30.21 to 14.88, P=.34) and weekends (adjusted mean difference –27.57 minutes, 95% CI –66.33 to 11.19, P=.55) compared with the control group; however, these changes were not statistically significant.

This intervention successfully targeted parents to improve preschoolers’ PA, sleep, and screen time compared with the control group, with no significant effects on dietary behaviors. Regarding PA, preschoolers in the intervention group engaged in less LPA and more VPA at the 12-week posttest compared with the control group; however, these changes were not maintained at the 12-week follow-up. For sleep-related outcomes, the intervention group demonstrated reduced sleep latency and improved sleep efficiency at both the 12-week posttest and the 12-week follow-up compared with the control group. Additionally, preschoolers’ screen time in the intervention group significantly decreased on both weekdays and weekends at the posttest compared with the control group, but this reduction was not maintained at the 12-week follow-up.

Both groups exhibited changes in PA intensity from baseline to posttest, characterized by a decline in LPA and MPA, along with an increase in MVPA and VPA. These findings are consistent with previous studies [49-53], which suggested that LPA shows a decreasing trend from age 3 onward, while MVPA and VPA increase from age 3 to 5, plateau thereafter, and decline from age 7 to 9 [53]. Previous studies have explained that children explore their surroundings and develop self-awareness by demonstrating varying levels of competence in basic motor skills such as catching, rolling, running, and galloping [49,52]. Preschoolers often overestimate their motor competence, believing they are more skilled than they actually are; this inflated perception can motivate them to practice and master higher-level motor skills [54]. When children feel skilled, they are more likely to persist in improving their abilities [55]. Children with higher levels of actual motor competence and a greater repertoire of motor skills were more likely to engage in higher levels of PA, including increased MVPA and VPA [52]. Additionally, the “ActivityStat” hypothesis suggests that an imposed increase or decrease in PA in a single domain may lead to compensatory changes in another domain, helping to maintain stable overall PA and energy expenditure over time [56]. Accordingly, preschoolers in the intervention group maintained their total PA at a relatively constant level by adapting various mechanisms, such as altering the frequency, intensity, or duration of their PA [57]. Of particular note, this finding contrasts with the results reported by the systematic review of Zhou et al [25], which found no significant between-group differences in PA-related outcomes, as objectively measured by GT3X, among preschoolers following a parent-based online intervention. This discrepancy may be attributed to the greater number of PA-related modules in our study compared with the studies included in that review (n=4 vs n=1). It is plausible that the additional PA-related modules—including PA knowledge, group interaction, goal setting, tailored feedback, and weekly goal revision—captured more attention from parents in the intervention group, potentially motivating them to support their children in engaging in higher levels of PA.

Notably, parents in the intervention group were consistently educated on the recommendation that young children should engage in a minimum of 3 hours of PA/day, with a minimum of 1 hour consisting of MVPA. This targeted parental education and promotion of PA guidelines likely served as a key strategy driving the observed differences in LPA and VPA compared with the control group [58,59].

From baseline to follow-up, preschoolers in the intervention group showed a significant increase in “enjoyment of food” (general interest in food) alongside a decline in “satiety responsiveness” (ability to stop eating in response to fullness cues). It is possible that the diet modules, regularly delivered to the intervention group, provided parents with a better understanding of balanced nutrition and healthy eating. This enhanced knowledge may have prompted parents to offer their children a wider variety of foods, contributing to increased food-approaching behaviors, including enjoyment of food [60].

In addition, many Chinese parents often misinterpret their children’s larger body weight and size as indicators of healthy growth, associating being overweight with affluence [61]. Parents may prefer their children to be chubbier and underestimate their status of overweight or obesity, making them more likely to pressure children to eat, an approach that has been associated with higher levels of enjoyment of food [62]. Chinese parental feeding practices are generally rooted in cultural values emphasizing not wasting food and ensuring children’s nutritional needs are met [63]. Parents often serve large portions of food as a way of expressing love and care for their children [60]. They tend to encourage their children to finish all the food on their plates during mealtimes, regardless of the child’s hunger and fullness cues [63]. Over time, these practices may reinforce the motivation to eat beyond physiological needs, contributing to higher enjoyment of food and lower satiety responsiveness in children [60].

Additionally, it is common in China for grandparents to co-reside with their adult children and serve as primary caregivers for their grandchildren [64]. Many grandparents, having lived through periods of food scarcity (such as the Great Famine of the 1950s), often associate increased food intake with successful caregiving and worry that their grandchildren may experience hunger [65]. As a result, grandparents may encourage excessive eating to prevent malnutrition, pressuring their grandchildren to eat even when not hungry, which can increase children’s enjoyment of food while reducing their responsiveness to satiety cues [65].

In previous parent-based eHealth intervention studies, the focus was primarily on preschoolers’ dietary consumption (such as noncore foods, fruits and vegetables, and sugar intake) rather than on eating behaviors [25]. Consequently, there is limited scope for direct comparison. The allocation of preschoolers in our study resulted in an unequal distribution of eating behaviors between the intervention group (eg, enjoyment of food, 47/120, 39.2%; satiety responsiveness 14/120, 11.7%) and the control group (eg, enjoyment of food, 37/117, 31.6%; satiety responsiveness 13/117, 11.1%) at baseline. Such uneven allocation may have influenced the overall intervention effects on preschoolers’ dietary behaviors. Future studies with a more balanced distribution of eating behaviors are therefore warranted.

Additionally, eating behaviors, which develop during early childhood and reflect individual appetitive traits associated with the regulation of food intake, have been found to be influenced by parenting practices and styles. For instance, feeding practices that involve encouragement and rewards can enhance food-approaching behaviors (such as enjoyment of food) and may even reduce satiety responsiveness in preschoolers [66]. High parental self-efficacy in feeding has also been linked to healthier eating behaviors in preschoolers [67]. In addition, an authoritarian parenting style has been associated with increased food responsiveness, emotional overeating, enjoyment of food, and satiety responsiveness [68]. Future studies exploring the mediating effects of parenting practices and styles are warranted, as mediation analysis can provide a better understanding of the factors that influence the effectiveness of parent-based eHealth interventions on preschoolers’ eating behaviors.

Improvements in sleep problems, sleep latency, and sleep efficiency among preschoolers in the intervention group were observed at posttest following the parent-based eHealth intervention, with gains in sleep latency and sleep efficiency sustained at follow-up. These enhancements may be explained by the role of PA in promoting several sleep variables, including sleep-onset latency, sleep efficiency, and total sleep duration [69-71]. According to energy conservation and body restoration theories, sleep duration and the amount of slow-wave sleep increase as a function of higher energy expenditure [72]. Specifically, PA depletes energy reserves, particularly glycogen stores in muscles and the liver. As energy expenditure increases and these stores are reduced, the body signals a need for recovery, enhancing the drive for sleep. This can lead to longer sleep durations and a greater proportion of slow-wave sleep, which is particularly crucial for physical recovery and is where many restorative processes occur [72].

Previous studies further reported that PA triggers the release of neurotransmitters such as serotonin and norepinephrine, which are important for mood regulation and relaxation, potentially contributing to improved sleep onset and maintenance [73]. When daytime MVPA was increased by 10%, sleep latency decreased by 12.15 minutes/day [74]. Higher levels of MVPA were also significantly associated with better sleep efficiency and reduced sleep fragmentation [70]. Two studies showed that MVPA improved sleep quality and alleviated sleep problems by decreasing sleep latency, increasing total sleep time, and reducing presleep anxiety [75,76]. Preschoolers in the intervention group exhibited a notable increase in total MVPA, from approximately 812 minutes at baseline to 1732 minutes at posttest, and then 824 minutes at follow-up, which may have contributed to the observed improvements in sleep problems, sleep latency, and sleep efficiency.

Beyond the biological perspective, parents in the intervention group were regularly provided with educational knowledge related to sleep, including adequate sleep and optimal sleep environments. This enhanced knowledge likely helped parents better understand the factors contributing to good sleep hygiene and the importance of maintaining consistent bedtimes. Regular bedtime routines improve sleep quality—such as reducing sleep latency and night wakings in young children—and help synchronize the internal biological clock, promoting self-regulation of the sleep-wake cycle and facilitating the natural sleep process [77,78]. Children with shorter sleep latency are able to spend a greater proportion of time in actual bed sleep and are less likely to experience nighttime disruptions and awakenings, which reduces interruptions and enhances sleep efficiency [79].

The insignificant difference between the 2 groups in sleep problems may be attributed to 2 factors. First, the high proportion of preschoolers with severe sleep problems (ie, total scores above 41) at baseline in both the intervention (109/120, 90.8%) and control groups (107/117, 91.5%) suggests that the improvements observed in parent-reported sleep problems at posttest may have been influenced by a regression-to-the-mean effect, rather than solely by the impact of the intervention [80]. Individuals initially identified with high values on the outcome measure are likely to show lower values upon remeasurement, even in the absence of any intervention [80]. Thus, it is possible that their sleep issues may have naturally improved over time, regardless of whether they received an intervention. Additionally, unlike sleep latency and efficiency, which were objectively assessed, preschoolers’ sleep problems were evaluated using parent-reported measures. This introduces the potential for parents to intentionally or unintentionally misreport information [81].

The intervention group showed a significant reduction in screen time on both weekdays and weekends compared with the control group. This finding is consistent with previous studies [82-84]. Parent-child attachment, defined as a mutually satisfying relationship, may influence screen time, as lower attachment levels have been associated with increased screen time due to reduced parental monitoring [82]. Stronger father-child and mother-child attachment was correlated with decreased screen time, with ego resilience mediating this relationship [83]. Meaningful interactions, such as engaging in enjoyable activities together, can strengthen parent-child attachment [84]. In the intervention group, parents were encouraged to participate in PA with their children, fostering bonding and positive memories, which likely contributed to the reduction in screen time.

This study addressed weaknesses identified in previous research, such as balancing the intervention content and minimizing the impact of the sequence in which the content was delivered on the outcomes. Integrating the entire intervention process with 3 critical components and 4 processing steps based on social cognitive theory provided a consolidated rationale for the intervention design. Objectively assessed data on preschoolers’ PA and sleep enhanced the accuracy and reliability of the results.

The findings of this study should be interpreted in light of certain limitations. First, measurements of preschoolers’ dietary behaviors and sleep problems relied on subjective questionnaires, which involved self-reported data. Consequently, parents may have misreported this information, either intentionally or unintentionally, potentially affecting the results in both groups. Second, because the posttest coincided with the Chinese Lunar New Year holiday, many participants were unable to return to school to complete the measurements within the proposed time frame, resulting in a high attrition rate (181/237, 76.4%). Third, a greater-than-expected proportion of children were classified as having an active lifestyle (engaging in 180 minutes of PA daily, with MVPA meeting 60 minutes), which may have diluted the intervention’s effect on MVPA. Finally, as participants were mainly from underdeveloped regions in China, the results may not generalize to preschoolers in developed regions.

Chinese parents often feel a strong sense of responsibility to guide and correct their children’s behavior, which can result in a high level of controlling behavior from an early age [85]. Interventions aimed at improving the lifestyle behaviors of Chinese preschoolers should not only target the children’s behavior but also give greater attention to parenting practices and styles. Future research is essential to examine how these parenting approaches mediate the effectiveness of parent-based eHealth interventions on preschoolers’ PA, diet, and sleep. Moreover, this intervention has potential for scalability and broad reach. Studies with larger sample sizes are needed to translate this effective parent-based eHealth intervention to other age groups and regions of China.

Within the Chinese cultural context, educating parents about healthy lifestyles via social media effectively improved preschoolers’ VPA, sleep latency and efficiency, and screen time compared with the control group. The insights gained from this study may help address the unhealthy lifestyle behaviors commonly observed in contemporary young children. Considerably more work is needed to leverage cutting-edge technological advancements to help families across broader regions create healthy living environments for their children. Such efforts can not only enhance individual health and family well-being but also reduce the societal financial burden associated with treating diseases resulting from poor lifestyle habits.