Tailored video-based lifestyle intervention for overweight or obese adults in Phnom Penh, Cambodia: a pilot cluster randomized controlled trial

Abstract

Physical activity plays a crucial role in preventing and managing obesity and its related comorbidities, with minimal side effects. Similarly, consuming a balanced diet is essential for maintaining good health and adequate nutrition. However, most adults fail to meet population-based dietary and physical activity guidelines. In this pilot cluster-randomized controlled trial, we aimed to evaluate the effectiveness of a video-based physical activity and dietary intervention among adults aged 30–59 years in Phnom Penh, Cambodia. A total of 63 adults participated in the 12-week program, with 31 assigned to the intervention group and 32 to the control group. The intervention group received a video-based program delivered via Telegram and YouTube, complemented by a 2-h in-person session focusing on technical orientation and safety training. Standard care was provided to both the intervention and control groups and consisted of a single 60-min in-person lifestyle education session and a printed booklet commonly used in Cambodian health center settings. The intervention group demonstrated greater improvements in physical activity adherence, healthy eating adherence, exercise self-efficacy, healthy eating self-efficacy, systolic blood pressure, body weight, body mass index, waist circumference, fasting blood glucose, and high-density lipoprotein cholesterol levels compared with the control group (P < .05). Given its high accessibility and ease of implementation, video-based content appears to be an effective approach and policy for improving health behaviors, even in resource-limited settings. These findings support the use of video and digital platforms as potential strategies and evidence-based policy for preventing and managing non-communicable diseases in low-resource settings. Trial registration ISRCTN11839050 (retrospectively registered).

Key messages.

• Video programs improve physical activity, healthy eating, and metabolic markers.

• Telegram and YouTube enable digital health interventions in low-resource settings.

• Boosted self-efficacy and peer support help sustain healthy behaviors.

• Affordable remote methods support non-communicable disease prevention and care.

Introduction

Overweight and obesity are persistent public health concerns in both developing and developed countries, with their prevalence steadily increasing over the past several decades despite numerous prevention efforts (Saghafi-Asl et al. 2020, Roslim et al. 2021). Globally, in 2022, ∼2.5 billion (43%) of adults aged 18 years or older were overweight, of whom 890 million (16%) were living with obesity (World Health Organization (WHO) 2024). In Cambodia, approximately one in five adults (19.4%) aged 18–69 years are overweight, and 4.3% are obese (Ministry of Health 2023).

Individuals with overweight and obesity are more likely to accumulate persistent non-communicable disease (NCD) risk factors, experience long-term weight gain, and develop adverse psychosocial outcomes, all of which contribute to increased morbidity and premature mortality (Shaban Mohamed et al. 2022). The impact of overweight and obesity in Cambodia is not trivial, reducing both health and economic outcomes. Estimates suggest that NCDs cost the Cambodian economy 6%–7% of gross domestic product, inclusive of direct health care costs and productivity losses. Obesity-related diseases are key modifiable drivers of such productivity losses due to early-onset disability, absenteeism, and premature mortality (Kulikov et al. 2019). Importantly, preventing progression from overweight to obesity represents an important opportunity to decrease productivity losses and health system costs in resource-limited settings (Zhang et al. 2021). In a low-resource, high out-of-pocket spending health system, prevention of progression from overweight to obesity and reducing the burden of obesity-linked complications are economic and clinical imperatives. This highlights the pressing need for scalable, affordable interventions.

Physical activity plays a crucial role in preventing and managing obesity and related comorbidities, with minimal side effects (Wiklund 2016, Choi et al. 2023). It increases energy expenditure through enhanced metabolic activity and contributes to a negative energy balance, while also improving mental well-being and musculoskeletal health (Baker et al. 2016, Kim et al. 2017). Regular physical activity is further associated with reduced risks of premature mortality and chronic conditions, including cardiovascular disease, diabetes, cancer, hypertension, obesity, depression, and osteoporosis (Naureen et al. 2022, Stensel 2023). Similarly, adherence to a balanced and nutritious diet is essential for maintaining overall health and reducing the incidence and mortality risk of cardiovascular disease (Taylor et al. 2024).

However, a substantial proportion of adults worldwide do not meet the WHO’s recommendation of 150–300 min of moderate-intensity aerobic physical activity per week, along with muscle-strengthening activities (WHO 2020). In Cambodia, both physical activity and dietary practices remain important targets for NCD prevention. National data indicate that 9.3% of adults have insufficient physical activity, while dietary inadequacy is highly prevalent, with 80.7% failing to consume at least five daily servings of fruits and vegetables (Ministry of Health 2023). Research in Cambodia and similar settings has reported systemic barriers to lifestyle modification, such as long working hours, lack of access to safe and affordable exercise facilities, high cost and low availability of healthy foods, competing family responsibilities, health literacy (HL) limitations (in health knowledge or skills), and low self-efficacy for the maintenance of new behaviors (Steinman et al. 2020a, Wamsiedel et al. 2025).

Cambodia, a lower-middle-income country undergoing rapid urbanization and lifestyle transitions, is experiencing a rising burden of overweight and obesity, particularly among middle-aged adults (Ministry of Health 2023). Adults aged 30–59 years are disproportionately affected, as this group is more likely to engage in sedentary occupations, maintain established but often unhealthy dietary patterns, and accumulate cardiometabolic risk factors while facing substantial economic and caregiving responsibilities (Whitaker et al. 2018, Park et al. 2020). In contrast, younger adults (18–29 years) are in more transient life stages with evolving educational and employment contexts, whereas older adults (≥60 years) commonly experience multimorbidity and functional limitations that require more individualized and clinically supervised interventions (Hu et al. 2024, Bourke, Brown, and Kwan, 2025). These considerations informed the focus of the present study on adults aged 30–59 years with overweight or obesity, a high-risk population for whom lifestyle modification is both feasible and critical in Cambodia (Kulikov et al. 2019, Ministry of Health 2023).

Traditional obesity management, such as face-to-face counseling and clinic-based education, is often constrained by high costs, the need for specialized facilities, and limited scalability in resource-limited settings like Cambodia (Chham et al. 2023, Wang et al. 2024). To overcome these barriers, smartphone-based video interventions offer a cost-effective and wide-reaching alternative (Gans et al. 2015). Unlike text-based methods, tailored video content is significantly more effective in promoting dietary and physical activity changes (Cheung et al. 2017). Recent evidence highlights that such digital technologies facilitate behavioral self-monitoring and result in superior weight-loss outcomes compared to conventional non-digital approaches (Carter et al. 2013, Protano et al. 2024).

In Cambodia, mobile phone ownership and use of platforms such as Facebook and Telegram are rapidly growing, which is encouraging for digitally delivered interventions that have the potential to reach a large proportion of the population at low marginal cost (Li et al. 2025). Yet, no formal video-based program focusing on both physical activity and diet has been tested among the 30–59 years old Cambodian adults with overweight or obesity. Therefore, this pilot cluster-randomized controlled trial aimed to evaluate the feasibility, acceptability, and preliminary effects of a tailored, video-based community lifestyle intervention among Cambodian adults aged 30–59 years with overweight or obesity. This study contributes to addressing an implementation gap in the use of digital platforms in low-resource settings by providing context-specific evidence on feasibility, engagement, and early effects, thereby informing preventive policy strategies.

Materials and methods

Study design

This study was a community-based, single-blind, pilot cluster-randomized controlled trial designed to evaluate the feasibility and preliminary effectiveness of a digital platform-based physical activity and dietary intervention for adults with overweight or obesity in Phnom Penh, Cambodia. Due to the nature of the behavioral intervention, blinding of participants was not feasible. However, to minimize potential bias, participants were not informed about the specific study hypotheses or the alternative intervention conditions.

Participant eligibility

Participants were eligible if they met the following inclusion criteria: (i) aged 30–59 years; (ii) permanent residents of Phnom Penh; (iii) body mass index (BMI) ≥ 25 kg/m²; (iv) not pregnant or within 6 months postpartum; (v) no known conditions limiting physical activity; (vi) internet-enabled smartphone and Telegram access; (vii) literacy in Khmer (reading and writing); and (viii) no concurrent participation in other weight-loss programs, use of weight-loss medications, or plans for bariatric surgery during the study period. Participants were also required to agree to random group assignment, provide written informed consent, and commit to the study requirements. In the present study, an international BMI cutoff of ≥25 kg/m² was applied to define eligibility in order to identify adults with clearly elevated adiposity-related risk and to maintain comparability with national and international datasets.

Sample size estimation

The target sample size for the full Randomized Controlled Trial (RCT) was calculated using G*Power 3.1, based on an expected effect size of 0.33 (Rotunda et al. 2024), with α set at 0.05 and power (1−β) at 0.80. This resulted in a total of 230 participants. Following Hertzog’s (2008) recommendation that pilot studies typically utilize 10%–25% of the sample size required for a full-scale trial, we initially targeted 25% of the individual-based power calculation. We applied an anticipated 10% attrition rate, setting the final target at 32 participants per group (Total N = 64). To account for the cluster randomization design, the design effect (DE) was further considered. Given the lack of prior ICC (intracluster correlation coefficient) data for lifestyle interventions in Cambodian villages, a median ICC of 0.005, reported across various health outcomes in primary care settings, was used as a conservative assumption (Adams et al. 2004). With an average cluster size of 32, the estimated DE is 1.155, indicating that a fully powered definitive trial would require ∼266 participants in total. The final target of 64 participants represents ∼24.1% of the cluster-adjusted full trial sample, which meets the criteria for a pilot trial.

Participant recruitment

Participants were recruited through multiple methods. First, local village health support group leaders in Prek Pnov and Kouk Roka communes assisted with recruitment after obtaining permission from commune and health center representatives. Second, advertisements were posted on Facebook within the community to reach potential participants. Third, individuals who expressed interest were asked to register and complete a survey either offline or online, with assistance provided by the research team as needed. Finally, participants meeting the eligibility criteria were invited to participate in the baseline assessment. The study villages were located in the Prek Pnov operational district, which had been previously identified by local health authorities during a KOICA-funded public health project. However, all participants in the current study were newly recruited from the community and had not participated in the prior KOICA project.

Informed consent

Trained research assistants individually explained the study’s purpose and procedures to potential participants, allowing ∼10 min for discussion. Participants were informed that their participation was voluntary, their anonymity and confidentiality would be maintained, and they could withdraw at any time without penalty. Written informed consent was obtained from all participants prior to participation.

Randomization and allocation

This study employed a cluster RCT. Two villages were randomly assigned in a 1:1 ratio using a simple random draw (lottery method) conducted by the research team prior to participant recruitment. The village was used as the unit of randomization to minimize contamination between groups. Participants were subsequently enrolled based on their residence within the assigned village. Neither participants nor field investigators had control over group assignment, thereby reducing potential allocation bias.

Video development

To support the intervention and promote behavioral engagement, structured video content on physical activity and dietary education was developed by a multidisciplinary team comprising certified fitness trainers, clinical dietitians, NCD specialists, and public health professionals. The content was tailored to the cultural context and health literacy levels of adults residing in Phnom Penh, Cambodia.

Development of a tailored physical activity video program

The video program was designed to increase daily physical activity at home, incorporating motivational elements such as goal setting, sharing daily adherence, and group messaging. It followed a structured, progressive, and culturally adapted format, delivered through five video modules for use in a 5-day cycle. Each module, lasting ∼60 min, was led by a certified Cambodian trainer with Khmer-language voice-over instructions to ensure accessibility and engagement. The content was reviewed by a university professor specializing in physical therapy in the United States (US). To enhance relatability and encourage participation, two local residents demonstrated the exercises alongside the trainer. The program required minimal space and equipment, making it feasible for participants in urban Phnom Penh. Each session included three components: (i) warm-up and aerobic exercises, (ii) resistance and muscle-strengthening training, and (iii) cool-down and stretching. Exercises targeted all major muscle groups (e.g. Day 1: legs, Day 2: chest) and were designed to progressively challenge participants while minimizing injury risk. A timer was displayed in the upper-left corner of each video to ensure accurate adherence to the exercise set. Additionally, a 12-min orientation video was created to explain the structure and content of all exercise modules in detail (Table 1).

Table 1.

Weekly structure and content of the physical activity program.

Day	Focus	Contents	Duration (Min)
1	Leg	Focused on lower body strength and endurance, incorporating exercises such as squats, sumo squats, hamstring workouts, and lunges. Aerobic activities included plank jacks, mountain climbers, high knees, butt kicks, and jumping jacks. Abdominal exercises and a comprehensive stretching routine concluded the session.	60
2	Chest	Emphasized upper body and chest muscle development through knee push-ups, dumbbell presses and flys (using household items if necessary), overhead triceps extensions, and front raises. Cardiovascular segments included full-body aerobic exercises similar to Day 1, followed by abdominal work and stretching.	60
3	Back	Targeted the back, shoulders, and arms with exercises such as dumbbell deadlifts, underhand rows, horizontal lat pulls, arm curls, side raises, and bent-over side raises. The aerobic and stretching routines were consistent with previous days, ensuring a balanced approach.	60
4	Aerobic/cardio	Prioritized cardiovascular fitness and fat burning through a series of aerobic exercises, including arm walking, slow burpees, plank jacks, mountain climbers, high knees, butt kicks, jumping jacks, and scissor jumps. Abdominal strengthening and a full-body stretching sequence completed the session.	60
5	Cardio Tabata day	Featured high-intensity interval training (Tabata) with short bursts of full-body aerobic exercises performed in rapid succession, interspersed with brief rest periods. The session concluded with muscle stretching to promote flexibility and recovery.	60

For safety, each session began with a 5-min warm-up (e.g. running in place, dynamic stretching, and joint mobilization) to prepare the body for more intense activity. The main exercise sets included multiple repetitions with short rest intervals to optimize both strength and cardiovascular benefits. No specialized equipment was required, and modifications were provided for participants with varying fitness levels. Participants were encouraged to follow the video modules 5 days per week, with rest or light activity on the remaining days, and to share their daily activity records and experiences via the project’s Telegram group for additional motivation and peer support.

Development of a tailored dietary education video program

A 40-min dietary education video was developed to improve participants’ understanding of obesity and encourage healthier eating behaviors. Its primary goal was to motivate participants to reduce fat and sugar intake while increasing fruit and vegetable consumption. The video content covered the definition and health risks of obesity, principles of balanced diet planning, and practical methods for meal preparation. To enhance comprehension, the video incorporated real-life demonstrations using local food items, illustrating how to select healthy ingredients and prepare simple, affordable meals in accordance with dietary guidelines. Topics included portion control, reading nutrition labels, and identifying hidden sugars and fats in common Cambodian foods. The sessions were designed to be visually engaging and culturally appropriate, featuring live cooking demonstrations, food models, and Khmer-language narration.

Prior to releasing the intervention materials, drafts of the educational and exercise videos were shared with potential participants to obtain feedback on the feasibility of performing the exercises as demonstrated. Based on this feedback, minor adjustments were made to ensure that the content was clear, understandable, and executable for the target population.

Intervention

The intervention was delivered over 12 weeks, which is a duration considered sufficient for health behavior change and meaningful health improvements, while remaining feasible in low-resource settings (Evans et al. 2017, Kim et al. 2022, Ma et al. 2023). Tailored physical activity and dietary education video materials were uploaded to a YouTube channel and distributed to participants in the intervention group (Fig. 1).

Figure 1.

YouTube channel with 5-day exercise and diet education videos.

Access links were shared via both group and private Telegram messages, allowing participants to view and repeat the sessions at their convenience. Each participant received a pedometer and was encouraged to achieve a daily step count of at least 8000 steps. Supplementary material, including an exercise mat, a pair of dumbbells, and athletic shoes, were provided to support physical activity engagement. Participants were instructed to complete 1 h individual physical activity per day using the five videos at home or in any convenient location. At baseline, a 2-h face-to-face session focusing on technical orientation and safety training (e.g. posture, exercise intensity, progression, warm-up, and cool-down) was delivered by the exercise trainer who developed the videos to ensure participants could correctly follow the program. To promote adherence, reminders and motivational messages were sent regularly via Telegram. A closed Telegram group was created to facilitate peer support, where participants shared daily updates such as photographs of their physical activity, step counts, and meals. This peer interaction aimed to reinforce accountability and foster group motivation.

Standard care

To isolate the effect of the video-based intervention, standard care, consisting of lifestyle education using a printed booklet, was provided to both the intervention and control groups. This approach reflects routine health education practice in Cambodian community health centers and is typically delivered by nurses as part of standard NCD prevention activities. The booklet included BMI classification, key risk factors and health consequences of overweight and obesity, and general recommendations for a healthy diet and physical activity. The booklet was developed by the research team in consultation with public health and nutrition experts and was based on publicly available health education resources, including World Health Organization recommendations and national health promotion guidelines commonly used in Cambodia. The booklet was distributed during a 1-h in-person education session, and participants were encouraged to read and follow the guidance independently. No additional face-to-face counseling sessions were provided during the 12-week follow-up period in either group. The only between-group difference was the provision of structured video materials and Telegram-based reminders and peer support in the intervention group. After completing the study, control group participants were granted access to the same physical activity and nutrition videos provided to the intervention group to minimize ethical concerns related to withholding health education.

Daily monitoring of intervention adherence

To monitor and enhance adherence to the intervention protocol, a structured daily tracking system was implemented throughout the 12-week study period. For physical activity, participants submitted photographs or videos of themselves performing the exercise routines featured in the intervention videos, serving as visual confirmation of session completion. They were also required to submit a daily photo of their pedometer screen displaying their step count, with a target of at least 8000 steps per day, as outlined in the physical activity guidelines delivered during the intervention. For dietary adherence, participants shared images of their main meals, typically breakfast, lunch, or dinner, enabling the research team to assess the quality and quantity of food consumed based on the dietary education provided (Fig. 2).

Figure 2.

Sample photos shared daily by participants via Telegram.

The research team reviewed these daily uploads to track compliance, provide personalized encouragement, and maintain participant motivation. Adherence was defined as the proportion of days in which a participant submitted at least one of the required items—meal photo, exercise photo, or pedometer photo—relative to the total number of tracking days. This tracking system ensured accurate documentation of engagement while fostering accountability and peer support through real-time group interactions.

Outcome measures

Study outcomes were assessed at two time points: baseline and 12 weeks postintervention. The primary outcomes were body weight, waist circumference (WC), and BMI. Secondary outcomes included blood pressure, fasting blood glucose, glycosylated hemoglobin (HbA1c), lipid profile [triglycerides, total cholesterol (TC), LDL-C, and HDL-C], self-reported physical activity, dietary behaviors, self-efficacy, and health-related quality of life (HRQoL).

Anthropometric and clinical measurements

Anthropometric and clinical measurements were conducted using standardized protocols to ensure reliability. WC was measured using a flexible measuring tape at the midpoint between the lowest rib and the iliac crest, with participants in a standing position. Blood pressure was measured after at least 5 min of seated rest. Participants with blood pressure ≥ 140/90 mmHg underwent a repeat measurement within 5 min using a digital sphygmomanometer (A&D Medical, Tokyo, Japan, manufactured in China). Fasting venous blood samples were collected by trained nurses and clinical laboratory technicians at X Hospital following an overnight fast of at least 8 h. Samples were analyzed for fasting plasma glucose, HbA1c, and lipid profile parameters—including triglycerides, TC, LDL, and HDL—using an automated chemistry analyzer (BK-200, Biobase, China). HbA1c levels were measured using a fully automated glycated hemoglobin analyzer (BK-HbA1c, Biobase, China). All samples were securely stored and disposed of in accordance with the hospital’s biosafety and biomedical waste disposal regulations after analysis.

Behavioral and psychosocial outcome measures

Questionnaire-based assessments were used to evaluate the behavioral and psychosocial outcomes. Physical activity and healthy eating adherence were measured using relevant subscales of the Health-Promoting Lifestyle Profile II, including six items for physical activity and nine items for healthy eating. Responses to each item were summed, with each scored on a 4-point scale (1 = never, 2 = sometimes, 3 = often, 4 = routinely), where higher scores indicated better adherence (Walker and Hill-Polerecky 1996). General self-efficacy, reflecting confidence in lifestyle management, was assessed using the New General Self-Efficacy Scale, which comprises eight items scored on a 4-point Likert scale (1 = not at all true to 4 = exactly true). Higher scores indicate greater perceived self-efficacy (Chen et al. 2001). Domain-specific self-efficacy was also evaluated. Healthy eating self-efficacy was assessed using the Eating Self-Efficacy Brief Scale, which measures an individual’s confidence in regulating eating behaviors under social and emotional pressures (Lombardo et al. 2021). The scale consists of eight items scored on a 6-point Likert scale (0 = not easy at all to 5 = completely easy), with summed scores indicating higher self-efficacy. Exercise self-efficacy was measured using the Self-Efficacy for Exercise Scale, which consists of nine items scored on a 10-point Likert scale (0 = not confident to 10 = very confident), with summed scores reflecting greater confidence in exercising (Resnick and Jenkins 2000). HRQoL was assessed using the 12-Item Short Form Health Survey (SF-12) (Ware et al. 1996). Physical component summary (PCS) and mental component summary (MCS) scores were calculated using US standard weights, following the SF-12 scoring manual, and transformed into norm-based T-scores [mean = 50, standard deviation (SD) = 10], with higher scores indicating better health status (Ware et al. 1996). Owing to the absence of Cambodian population norms, US weights were applied, consistent with prior international studies (Gandek et al. 1998). The average time required to complete the self-report questionnaires was ∼20 to 30 min.

Ethical considerations

Ethical approval was obtained from the National Ethics Committee for Health Research and the Institutional Review Board of the authors’ institute prior to the commencement of the study. All participants were thoroughly informed of the study’s purpose, procedures, and potential risks, including muscle soreness and fatigue. Measures were implemented to mitigate these risks and manage any adverse events. Participants were encouraged to report any discomfort or concerns and were allowed to request modifications to the program as needed. The safety and well-being of all participants were prioritized and safeguarded throughout the study.

Clinical trial registration

This trial was retrospectively registered with the ISRCTN registry (ISRCTN11839050). The authors transparently report this and confirm that the study protocol, eligibility criteria, and outcomes were not modified in response to study findings. A detailed study protocol has not been published; however, the study procedures and outcomes are fully described in the Methods section.

Statistical analysis

Homogeneity of participants’ characteristics between the intervention and control groups was assessed using the chi-square test or Fisher’s exact test, as appropriate. The normality of continuous outcome variables at baseline was evaluated within each group using the Shapiro–Wilk test. A per-protocol (PP) analysis approach was employed to assess the effectiveness of the intervention among participants who completed the 12-week program. Independent t-tests were used for variables with a normal distribution, whereas the Mann–Whitney U test was applied when normality was violated in either group. Statistical significance was set at P < .05. All statistical analyses were performed using SPSS version 21.0. Standardized effect sizes (Hedges’ g) were calculated for between-group differences in change scores to quantify the magnitude of intervention effects.

Results

Participant flow and baseline characteristics

A total of 107 individuals were screened for eligibility. After excluding those who did not meet the inclusion criteria (n = 25), refused to participate (n = 11), or declined for other reasons (n = 5), 34 participants were allocated to the intervention and 32 to the control group. In the intervention group, three participants dropped out within the first 4 weeks of the intervention: one due to her son being hospitalized for an extended period, and two due to work-related commitments. Their baseline characteristics did not differ meaningfully from those of participants who completed the study in terms of sex, age, body mass index, and presence of chronic diseases. Given that participants in the intervention group withdrew early, prior to meaningful exposure to the intervention, and did not provide postintervention outcome data, an intention-to-treat analysis was not feasible. Consequently, 31 in the intervention group and 32 in the control group were included in the final PP analysis (Fig. 3).

Figure 3.

Participant flow diagram.

The mean age was 43.0 years in the intervention group and 41.84 years in the control group. Men comprised 67.7% of the intervention group and 50.0% of the control group. The majority of participants were married, with 93.5% in the intervention group and 87.5% in the control group. Participants with middle or higher education levels accounted for 58.1% in the intervention group and 53.1% in the control group. Regarding occupation, formal employment was the most common category, representing 38.7% and 37.5% of the intervention and control groups, respectively. For household income adequacy, “not sufficient” was the most common response in both groups, with 51.6% in the intervention group and 71.9% in the control group. The prevalence of diagnosed chronic diseases was 25.8% in the intervention group and 28.1% in the control group, whereas medication use for chronic diseases was reported by 19.4% and 15.6% of participants in the intervention and control groups, respectively. The homogeneity test confirmed no significant differences between the intervention and control groups across these characteristics (Table 2).

Table 2.

Homogeneity of participants’ characteristics between the intervention and control groups.

Characteristic	Categories	Intervention Group (n = 31)	Control Group (n = 32)	Z/x 2	P
		n(%) or M ± SD	n(%) or M ± SD
Age		43.00 ± 8.93	41.84 ± 8.19	−0.565a	.572
Gender	Male	21(67.7)	16(50.0)	2.045	.153
	Female	10(32.3)	16(50.0)
Spouse	Yes	29(93.5)	28(87.5)	0.669b	.672
	No	2(6.5)	4(12.5)
Education	Low (no schooling/primary)	13(41.9)	15(46.9)	0.218	.897
	Middle (secondary/high school)	12(38.7)	12(37.5)
	High (university/college)	6(19.4)	5(15.6)
Occupation	Unemployed/informal (none, day laborer)	9(29.0)	11(34.4)	0.237	.888
	Self-employed (small business)	10(32.3)	9(28.1)
	Formal employment (company staff, civil servant)	12(38.7)	12(37.5)
Household income adequacy	Sufficient	15(48.4)	9(28.1)	2.741	.081
	Not sufficient	16(51.6)	23(71.9)
Diagnosis of chronic disease	Yes	8(25.8)	9(28.1)	0.043	.836
	No	23(74.2)	23(71.9)
Medication use for chronic disease	Yes	6(19.4)	5(15.6)	0.152	.697
	No	25(80.6)	27(84.4)

M, Mean; SD, standard deviation. ^aMann–Whitney U test; ^bFisher’s exact test.

Intervention engagement and adherence

Among participants in the intervention group, 88.2% (30/34) uploaded at least one photo or video documenting physical activity or dietary adherence during the intervention period. The majority of these participants uploaded content on average 3–5 days per week, whereas a smaller proportion uploaded content 1–2 days per week, indicating lower levels of engagement. During the first 2 weeks of the intervention, participants primarily accessed exercise videos via YouTube streaming. In later weeks, many participants reported downloading the videos to their devices in order to reduce mobile data usage, minimize disruptions due to unstable internet connectivity, and facilitate repeated viewing to reinforce activity routines. These engagement patterns suggest acceptable adherence and highlight the feasibility of delivering digital health interventions in low-resource settings with variable internet access.

Homogeneity of outcome variables between the intervention and control groups at baseline

The homogeneity test results for baseline outcome variables showed no significant differences between the intervention and control groups in physical activity adherence, healthy eating adherence, general self-efficacy, healthy eating self-efficacy, MCS of HRQoL, systolic blood pressure (SBP), diastolic blood pressure, body weight, BMI, WC, TC, fasting blood glucose, LDL-C, triglycerides, and HbA1c. However, healthy eating self-efficacy was significantly higher in the control group than in the intervention group (25.28 ± 9.29 vs. 20.61 ± 9.23; P = .023), whereas the PCS of HRQoL was significantly higher in the intervention group than in the control group (53.16 ± 8.42 vs. 46.94 ± 10.57; P = .032). Additionally, HDL-C was significantly higher in the intervention group than in the control group (61.97 ± 11.20 vs. 59.78 ± 30.03; P= .020) (Table 3).

Table 3.

Homogeneity of outcome variables between the intervention and control groups (baseline).

Variables	Intervention group (n = 31) M ± SD	Control group (n = 32) M ± SD	t or Z	P
Physical activity adherence	13.81 ± 4.42	12.06 ± 2.17	−1.231a	.218
Healthy eating adherence	20.77 ± 3.44	20.94 ± 1.85	−0.792a	.428
General self-efficacy	24.48 ± 2.97	24.66 ± 2.09	−0.508a	.611
Healthy eating self-efficacy	20.61 ± 9.23	25.28 ± 9.29	−2.269a	.023
Exercise self-efficacy	55.52 ± 18.26	61.41 ± 17.89	−1.492a	.136
PCS (HRQoL)	53.16 ± 8.42	46.94 ± 10.57	−2.145a	.032
MCS (HRQoL)	48.61 ± 8.85	51.34 ± 10.97	−1.084b	.282
SBP (mmHg)	134.52 ± 22.08	127.29 ± 20.92	−1.451a	.147
DBP (mmHg)	81.00 ± 7.51	85.03 ± 9.96	−0.942a	.346
Body weight (kg)	76.07 ± 6.89	75.25 ± 9.91	−1.074a	.283
BMI (kg/m2)	27.77 ± 2.40	28.18 ± 3.10	−0.021a	.984
WC (cm)	91.13 ± 6.18	90.94 ± 7.82	−0.510a	.610
TC (mg/dl)	206.52 ± 34.15	208.78 ± 76.16	−1.100a	.271
FBS (mg/dl)	100.32 ± 22.40	115.50 ± 51.92	−0.220a	.826
HDL-C (mg/dl)	61.97 ± 11.20	59.78 ± 30.03	−2.332a	.020
LDL-C (mg/dl)	102.16 ± 40.99	91.82 ± 6.85	−0.646a	.518
TG (mg/dl)	211.94 ± 126.38	285.91 ± 377.82	−0.564a	.573
HbA1C (%)	5.71 ± 0.54	6.57 ± 2.23	−1.748a	.081

M, mean; SD, standard deviation; PCS, physical component summary; MCS, mental component summary; HRQoL, health-related quality of life; SBP, systolic blood pressure; DBP, diastolic blood pressure; BMI, body mass index; WC, waist circumference; TC, total cholesterol; FBS, fasting blood sugar; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; TG, triglyceride; HbA1C, glycated hemoglobin. ^aMann–Whitney U test; ^bIndependent t-test.

Differences in outcome variables between the intervention and control groups

Pre- and postintervention comparisons revealed significant improvement in the intervention group for physical activity adherence (P < .001), healthy eating adherence (P < .001), general self-efficacy (P = .028), exercise self-efficacy (P < .001), healthy eating self-efficacy (P = .012), SBP (P = .001), body weight (P = .001), BMI (P = .001), and WC (P = .003). In the control group, significant changes were observed only in physical activity adherence (P < .001), body weight (P = .003), BMI (P = .003), and fasting blood glucose (P = .008). Between-group comparisons revealed that the intervention group experienced significantly greater improvements in physical activity adherence (P = .035), healthy eating adherence (P = .034), healthy eating self-efficacy (P = .038), SBP (P = .017), body weight (P = .022), BMI (P = .019), WC (P = .029), fasting blood glucose (P = .021), and HDL-C (P = .049) than those observed in the control group. No significant differences were found between the groups for general self-efficacy, exercise self-efficacy, PCS and MCS of HRQoL, TC, LDL-C, triglycerides, or HbA1c. Although the intervention group showed a significant within-group increase in exercise self-efficacy, this difference was not statistically significant compared to the control group (P = .071). Between-group differences were additionally quantified using standardized effect sizes (Hedges’ g). Effects sizes suggested improvements in physical activity adherence (g = 0.54), healthy eating adherence (g = 0.54), and healthy eating self-efficacy (g = 0.42), as well as reductions in systolic blood pressure (g = −0.62), body weight (g = −0.25), BMI (g = −0.27), and waist circumference (g = −0.56), compared with the control group (Table 4).

Table 4.

Differences in outcome variables between the intervention and control groups.

Variables	Group	Pretest M ± SD	Posttest M ± SD	Difference (Post–pre) M ± SD	Within-group t or Z (P)	Between-group t or Z (P)	Effect size (Hedges’ g)
Physical activity adherence	Int.	13.81 ± 4.42	18.68 ± 3.21	4.87 ± 4.06	−6.673(<.001)a	−2.157(.035)a	0.54
	Cont.	12.06 ± 2.17	14.78 ± 3.40	2.72 ± 3.85	−3.990(<.001)a
Healthy eating adherence	Int.	20.77 ± 3.44	24.03 ± 3.56	3.26 ± 4.35	−4.169(<.001)a	2.171(.034)a	0.54
	Cont.	20.94 ± 1.85	22.06 ± 2.86	1.13 ± 3.41	−1.869(.071)a
General self-efficacy	Int.	24.48 ± 2.97	25.90 ± 2.62	1.42 ± 3.42	−2.309(.028)a	−1.305(.192)b	0.44
	Cont.	24.66 ± 2.09	24.78 ± 1.43	0.13 ± 2.27	−0.492(.623)b
Exercise self-efficacy	Int.	55.52 ± 18.26	71.32 ± 14.91	15.81 ± 17.69	−4.975(<.001)a	1.838(.071)a	0.46
	Cont.	61.41 ± 17.89	67.53 ± 12.88	6.13 ± 23.61	−1.468(.152)a
Healthy eating self-efficacy	Int.	20.61 ± 9.23	25.81 ± 7.35	5.19 ± 10.85	−2.664(.012)a	−2.072(.038)b	0.42
	Cont.	25.28 ± 9.29	26.13 ± 6.86	0.84 ± 9.58	−0.041(.967)b
PCS (HRQoL)	Int.	53.16 ± 8.42	54.63 ± 9.66	1.47 ± 13.11	−0.626(.536)a	0.847(.400)a	0.21
	Cont.	46.94 ± 10.57	45.51 ± 8.21	−1.43 ± 14.06	0.575(.569)a
MCS (HRQoL)	Int.	48.61 ± 8.85	50.05 ± 11.54	1.44 ± 12.35	−0.412(.681)b	−0.619(.536)b	0.23
	Cont.	51.34 ± 10.97	49.95 ± 8.44	−1.39 ± 11.55	0.682(.500)a
SBP (mmHg)	Int.	134.52 ± 22.08	124.30 ± 16.74	−10.73 ± 17.91	−3.256(.001)b	−2.396(.017)b	−0.62
	Cont.	127.29 ± 20.92	127.56 ± 18.08	−0.16 ± 15.49	0.058(.954)a
DBP (mmHg)	Int.	81.00 ± 7.51	82.03 ± 10.35	0.60 ± 13.73	−0.810(.418)b	−0.317(.751)b	0.15
	Cont.	85.03 ± 9.96	82.25 ± 10.05	−2.78 ± 28.73	−0.745(.456)b
Body weight (kg)	Int.	76.07 ± 6.89	74.33 ± 7.65	−1.74 ± 3.61	−3.177(.001)b	−2.297(.022)b	−0.25
	Cont.	75.25 ± 9.91	74.23 ± 9.48	−1.02 ± 1.80	3.203(.003)a
BMI (kg/m2)	Int.	27.77 ± 2.40	27.11 ± 2.47	−0.65 ± 1.28	−3.190(.001)b	−2.352(.019)b	−0.27
	Cont.	28.18 ± 3.10	27.81 ± 3.06	−0.37 ± 0.66	3.164(.003)a
WC (cm)	Int.	91.13 ± 6.18	88.16 ± 5.62	−2.97 ± 5.04	3.281(.003)a	−2.234(.029)a	−0.56
	Cont.	90.94 ± 7.82	90.84 ± 5.84	−0.09 ± 5.17	0.103(.919)a
TC (mg/dl)	Int.	206.52 ± 34.15	203.00 ± 32.99	−3.52 ± 20.95	0.935(.357)a	−1.547(.122)b	−0.20
	Cont.	208.78 ± 76.16	211.75 ± 51.14	2.97 ± 40.02	−1.509(.131)b
FBS (mg/dl)	Int.	100.32 ± 22.40	100.42 ± 15.55	0.10 ± 13.79	−0.039(.969)a	−2.304(.021)b	0.17
	Cont.	115.50 ± 51.92	109.84 ± 60.06	−5.66 ± 43.84	−2.638(.008)b
HDL-C (mg/dl)	Int.	61.97 ± 11.20	60.81 ± 11.49	−1.16 ± 5.45	1.186(.245)a	−1.969(.049)b	−0.02
	Cont.	59.78 ± 30.03	58.91 ± 11.74	−0.88 ± 23.21	−1.750(.080)b
LDL-C (mg/dl)	Int.	102.16 ± 40.99	100.60 ± 36.22	−1.56 ± 31.79	0.273(.787)a	−1.224(.221)b	−0.27
	Cont.	91.82 ± 6.85	104.59 ± 40.25	12.77 ± 66.25	−1.982(.047)b
TG (mg/dl)	Int.	211.94 ± 126.38	207.96 ± 157.24	−3.98 ± 125.51	−0.849(.396)b	−0.660(.509)b	0.15
	Cont.	285.91 ± 377.82	241.28 ± 194.06	−44.63 ± 350.01	−1.122(.262)b
HbA1C (%)	Int.	5.71 ± 0.54	5.66 ± 1.27	−0.05 ± 0.75	−1.154(.877)b	−1.432(.152)b	0.29
	Cont.	6.57 ± 2.23	6.27 ± 2.23	−0.30 ± 0.95	1.796(0.082)a

Int., intervention group; Cont., control group; M, mean; SD, standard deviation; PCS, physical component summary; MCS, mental component summary; HRQoL, health-related quality of life; SBP, systolic blood pressure; DBP, diastolic blood pressure; BMI, body mass index; WC, waist circumference; TC, total cholesterol; FBS, fasting blood sugar; HDL-C, high-density lipoprotein cholesterol; TG, triglyceride; HbA1C, glycated hemoglobin; LDL-C = low-density lipoprotein cholesterol. ^aPaired t-test; ^bMann–Whitney U test

Discussion

In this study, a video-based physical activity and dietary program was developed to address the high prevalence of NCDs and the associated mortality rate in Cambodia. The program was administered to the intervention group over 12 weeks, and their health outcomes were compared with those of the control group. The results revealed significantly greater improvement in the intervention group for physical activity adherence, healthy eating adherence, exercise self-efficacy, healthy eating self-efficacy, SBP, body weight, BMI, WC, fasting blood glucose, and HDL-C (P-values ranging from .02 to .04). In addition to statistical significance, standardized effect sizes (Hedges’ g) supported meaningful intervention impacts, particularly for physical activity and healthy eating adherence and reductions in systolic blood pressure and waist circumference.

The significant improvements in exercise and healthy eating self-efficacy in the intervention group suggest that the program positively influenced participants’ self-confidence and psychological factors related to behavioral change. Although the difference in exercise self-efficacy between the intervention and control groups was not statistically significant (P = .071), a significant increase was observed within the intervention group, which may be related to the sample size. Individuals with higher self-efficacy are more likely to initiate and maintain physical activities and tend to sustain healthy behaviors even in novel situations (Pekmezi et al. 2009). The intervention program likely increased participants’ self-efficacy by allowing them to observe successful peers, encourage each other, and experience small achievements—such as meeting step count goals and monitoring dietary intake and exercise videos daily via Telegram. This approach effectively leveraged the key sources of self-efficacy described in Bandura’s social cognitive theory: mastery experiences, vicarious experiences, and social persuasion (Bandura 2001). Similarly, a Malaysian study on weight control education based on the Health Belief Model (HBM) reported significant improvements in dietary and exercise self-efficacy. The authors attributed these improvements to positive behavioral changes driven by increased perceived benefits and reduced barriers, which are core components of the HBM (Raman et al. 2024). In the current study, participants learned about the risks of obesity and NCDs through educational videos and gained knowledge about the benefits of exercise and healthy eating. Barriers to adherence were reduced through coaching and personalized goal setting. These findings suggest that interventions grounded in theoretical frameworks like Social Cognitive Theory and HBM, which enhance self-efficacy and health beliefs, can be effective for managing obesity in developing countries. The finding of improved general self-efficacy in the intervention group, without a significant difference between groups (P = .192), indicates that self-efficacy improved in a domain-specific manner. In other words, the program enhanced confidence in the targeted exercise and dietary behaviors but did not significantly affect overall self-efficacy. These results demonstrate that the program was a practical intervention focused on lifestyle changes and suggest that improving self-efficacy in specific behaviors is sufficient to promote positive health behavioral changes. Participants in the intervention group showed significantly greater improvements in physical activity adherence and healthy eating adherence than those in the control group.

The high engagement and participation rates are believed to result from the production and distribution of educational video content tailored to the cultural context and health literacy of adults living in Phnom Penh, Cambodia. The videos were developed through collaboration among fitness trainers, clinical nutritionists, and experts in NCDs and public health and included narration in Khmer and familiar visual materials to enhance understanding and immersion. Notably, the program increased realism and relatability by featuring two local residents in demonstrations led by a certified Cambodian trainer. Previous studies have similarly reported that culturally appropriate video content is more effective than generalized materials in inducing healthy behavioral change (Catley et al. 2022). Additionally, studies have reported that video-based modeling can effectively facilitate the acquisition of healthy behaviors by inducing behavioral changes through observational learning based on actual demonstrations (Catley et al. 2022), which supports the strategy used in this study. The physical activity videos were systematically structured into expert modules, each lasting 50–60 min and provided 5 days a week, covering lower body, upper body, cardio, and Tabata workouts. These videos were designed for repeated viewing, featuring timers in the upper-left corner, clear movement demonstrations, and progressive difficulty adjustments to ensure accurate practice and reduce the risk of injury. To further support adherence and participation, daily reminder messages were sent, pedometers with an 8000 steps/day goal were provided, and basic exercise equipment, such as mats and dumbbells, was distributed. The dietary intervention included 40-min nutritional educational videos explaining the risk of obesity, principles of a balanced diet, healthy meal preparation, and how to read nutritional information. To increase practicality, the videos demonstrated simple and inexpensive recipes using ingredients commonly available in local markets. Such video-based exercise and nutrition sessions offer a valid and useful intervention that can be accessed at home or anywhere at a low cost, making them particularly effective for enhancing physical activity and healthy eating adherence in resource-limited settings.

The video-based physical activity and dietary program used in this study was effective in enhancing self-efficacy and adherence as well as in reducing metabolic markers such as blood pressure, BMI, fasting blood glucose, HDL-C, and WC. A recent study on video-based training for children with obesity demonstrated similar effectiveness on body composition and metabolic markers, including increased HDL-C and reduced uric acid, BMI, and body fat (Pedraza-Escudero et al. 2025). Additionally, a study on video-based smartphone app interventions among German adults reported a significant increase in physical activity over 8 weeks, measured by metabolic equivalent of task minutes per week (Fischer et al. 2022). These findings expand the evidence supporting non-face-to-face, home-based physical activity education as an effective approach for managing body weight and improving metabolism in adolescents and in adults with obesity, and in populations at risk of metabolic disorders, even in settings with limited cost and space. This supports the potential role of such videos as part of core strategies for the prevention and management of NCDs. Regarding the potential for sustained improvement in metabolic markers through continued intervention of at least 6 months, as suggested by Pedraza-Escudero et al. (2025), the present study highlights the need for future long-term follow-up research. Uploading the developed videos to platforms like YouTube and encouraging their widespread use offers a low-cost, effective dissemination method that could be strategically implemented at the government level (Boté-Vericad et al. 2025). Overweight individuals in India were interviewed in a study about exercise videos, and participants reported preferring them for convenience, cost savings, and time efficiency; however, they also noted the importance of including music, accurate form descriptions, and clear indications of repetitions and sets (Shishira et al. 2024). In the present study, we addressed these preferences and incorporated additional engaging features such as peer models, friendly instructors, high-quality videos, and bright lighting.

The digital platform-based intervention used in this study—consisting of a Telegram group and YouTube educational videos—was particularly significant because it facilitated healthy behavioral changes without traditional face-to-face education or meetings. Participants in the intervention group shared their physical activity progress and photos of their meals through the Telegram group, providing mutual encouragement and motivation. Such peer support and social networking are known to play crucial roles in sustaining healthy behaviors (Pekmezi et al. 2009). A meta-analysis also reported that social media interventions are effective for managing body weight, largely due to enhanced motivation through social support and interactions (Loh et al. 2023). Considering Cambodia’s context, this approach is especially practical and effective. Owing to geographic dispersion, poor transportation, and limited infrastructure, gathering regularly for traditional health education may be difficult for residents. However, mobile phone and internet availability are relatively high, and social media platforms like Facebook and Telegram are widely used, with ∼70% of the population active on these platforms (Kemp 2024). This inexpensive but highly efficient mHealth approach can serve as a sustainable strategy for NCD prevention in developing countries with limited health resources (Steinman et al. 2020b). Future research should explore sex- and age-specific digital literacy affecting access to digital platforms, as well as sociocultural factors influencing physical activity and dietary adherence, which may impact the effectiveness and sustainability of such interventions.

This study has several limitations. First, the intervention was limited to adults in an urban setting (Phnom Penh), potentially limiting the generalizability of the findings to rural populations or other regions. Second, the findings should be interpreted with caution, as the use of PP analysis may have overestimated the intervention effect. Finally, the statistical analysis did not account for the village-level cluster effect due to the small size of clusters (K = 2) in this pilot study, which may affect the precision of standard errors. Future definitive trials should utilize cluster-adjusted analytical approaches such as generalized estimating equations or mixed-effects models to ensure robust results. The results of this pilot study support the feasibility of scaling up the intervention. Before large-scale implementation, participatory co-design with community members and local health workers can help identify diverse user preferences, including intervention content, delivery modality, and intensity, as well as opportunities for peer interaction, to enhance acceptability and scalability.

Conclusion

This study demonstrated the feasibility and preliminary effectiveness of combining video-based physical activity and dietary education with digital social platforms to improve health behaviors and metabolic markers in Cambodian adults. The findings suggest that highly accessible and cost-effective video-based interventions can serve as sustainable strategies and evidence-based policy for NCD prevention and management in resource-limited settings. Future large-scale studies grounded in social cognitive theory and the HBM are required to evaluate the long-term effectiveness of these interventions and facilitate their integration into public health programs.

Author contributions

Conception or design of the work: Y.Y., S.Y., O.H., V.K. Data collection: S.Y. and O.H. Data analysis and interpretation: Y.Y. Drafting the article: Y.Y. Critical revision of the article: Y.Y., S.Y., O.H., and V.K. Final approval of the version to be submitted: Y.Y., S.Y., O.H., and V.K.

Funding

This work was supported by the Ministry of Education of the Republic of Korea and the National Research Foundation of Korea, NRF- 2023S1A5A2A03083762.

Data availability

The datasets generated and/or analyzed during the current study are available from the corresponding author on reasonable request for academic purposes only.

Reflexivity statement

The authorship team comprises researchers with complementary expertise and varied levels of seniority, contributing to a balanced and reflexive research process. All authors specialize in NCD research in Cambodia and have substantial experience conducting community-based and policy-relevant health research in low-resource settings. The first author is a global health researcher with extensive experience in research and teaching in global and public health, including long-term engagement in Cambodia.

Ethical approval

Ethical approval was obtained from the National Ethics Committee for Health Research (No. 465 NECHR) and the Institutional Review Board of Jeonbuk National University (File No. 2024-09-017-001) prior to the commencement of the study. Participants were required to agree to random group assignment, provide written informed consent, and commit to the study requirements.

Disclosure of AI assistance

During the preparation of this manuscript, the authors used ChatGPT (OpenAI) to assist with English language editing and improving readability. The authors reviewed and revised the content as needed and take full responsibility for the final manuscript.