ANíMATE mHealth app for guidance on weight loss in adults with obesity: A randomized controlled trial

Abstract

Aim

To assess the effectiveness of a specific mobile health application (ANíMATE) in promoting weight loss in Spanish adults with obesity.

Materials and methods

We conducted a 4-month exploratory randomized controlled trial (ClinicalTrials.gov registration number: NCT05236881; registration date: January 24, 2022). Thirty-six participants with class I/II obesity were randomized to usual care (3 face-to-face visits) or the intervention group (ANíMATE in addition to 3 face-to-face visits). The primary outcome was total weight loss percentage (%TWL) at 4 months. Secondary outcomes included patient-reported outcome measures, additional weight-related variables, adherence, and satisfaction with the application.

Results

At 4 months, no differences were observed in the primary outcome, %TWL: 2.73% (0.86; 5.06) in the intervention group versus 1.23% (−0.23; 3.21) in the control group (P = .283). Differences were also not observed in secondary outcomes, with the exception of physical activity at 4 months, which was higher in the control group (2721 [1049; 5940] vs 923 [−8081; 639] MET-min; P = .020), and the frequency of weekly self-weighing at 2 months, also higher in the control group (76.9% vs 36.4%; P = .042). Mean user satisfaction with the application, evaluated using an ad hoc questionnaire, was high (16.5 out of 20).

Conclusions

In people with obesity, intervention with the ANíMATE mobile application added to usual care did not demonstrate additional benefits in %TWL, the primary outcome, in a study with limited statistical power. Further long-term research is needed to develop tailored mobile health strategies for weight management interventions that promote active user engagement.

Resumen

Objetivo

Evaluar la eficacia de una aplicación específica de salud móvil (ANíMATE) para promover la pérdida de peso en adultos españoles con obesidad.

Material y métodos

Llevamos a cabo un ensayo clínico aleatorizado exploratorio de 4 meses de duración (número de registro en ClinicalTrials.gov: NCT05236881; fecha de registro: 24/01/2022). Treinta y seis participantes con obesidad tipo I/II fueron asignados aleatoriamente a atención habitual (3 visitas presenciales) o al grupo de intervención (uso de ANíMATE además de 3 visitas presenciales). La variable principal fue el porcentaje de pérdida total de peso (%PTP) a los 4 meses. Las variables secundarias incluyeron medidas de resultados reportadas por los pacientes, variables adicionales relacionadas con el peso, adherencia y satisfacción con la aplicación.

Resultados

A los 4 meses, no se observaron diferencias en la variable principal, el %PTP: 2,73% (0,86; 5,06) en el grupo de intervención frente a 1,23% (−0,23; 3,21) en el grupo control (p = 0,283). Tampoco se observaron diferencias en las variables secundarias, con la excepción de la actividad física a los 4 meses, que fue mayor en el grupo control (2721 [1049; 5940] frente a 923 [−8081; 639] MET-min; p = 0,020), y de la frecuencia de automonitorización semanal de peso a los 2 meses, también superior en el grupo control (76,9% frente a 36,4%; p = 0,042). La satisfacción media de los usuarios con la aplicación, evaluada mediante un cuestionario ad hoc, fue elevada (16,5 sobre 20).

Conclusión

En personas con obesidad, la intervención con la aplicación móvil ANíMATE añadida a la atención habitual no mostró beneficios adicionales en el %PTP, la variable principal, en un estudio con una potencia estadística limitada. Se necesita más investigación a largo plazo para desarrollar estrategias personalizadas de salud móvil dirigidas a intervenciones para control del peso que fomenten la participación activa de los usuarios.

Introduction

Obesity is a chronic, complex, and multifactorial condition characterized by excessive fat accumulation. It is the most prevalent metabolic disease in developed countries and a major epidemic of the 21st century. In Spain, more than one-third of adults are overweight, and 14.1% are obese.¹ Obesity increases the risk of chronic conditions, including type 2 diabetes, high blood pressure (HBP), dyslipidemia, obstructive sleep apnea (OSA), cardiovascular disease, certain cancers, and mental health disorders, while also impairing bone health, reproduction, quality of life (QoL), and overall survival.²

Even moderate weight loss (5%–10%) is associated with clinically significant improvements in obesity-related conditions.³ However, only a minority of individuals with obesity receive targeted treatment through lifestyle interventions, medication, or bariatric surgery. Barriers include limited recognition of obesity as a disease, insufficient training of healthcare professionals (HCPs), and lack of accessible and cost-effective lifestyle programs.⁴ Telemedicine offers opportunities to overcome these barriers, although evidence regarding the effectiveness of electronic health interventions for weight loss maintenance or prevention of weight gain remains limited.⁵

In this context, mobile health (mHealth) applications (apps) are increasingly used to establish nutritional goals, monitor physical activity (PA), and support lifestyle modification. Among individuals with chronic medical conditions, mHealth apps are associated with increased rates of health-promoting behaviors⁶ and may reduce primary care burden, lower costs, and improve quality of care.⁷ Reviews and meta-analyses synthesizing the available evidence on mHealth interventions for weight-related issues conclude that apps may facilitate weight loss, although the evidence remains inconclusive.^8–10 A recent Cochrane review of 18 studies including 2703 participants evaluated the effects of apps for adults with overweight or obesity compared with no intervention or minimal intervention. A modest reduction in body mass index (BMI) of −2.6 kg/m² was observed at 6–8 months, although this effect was not sustained at 12–24 months. The authors concluded that current evidence is limited and does not demonstrate a clear benefit.⁸

In 2008, our group launched the PREDIRCAM project, a web-based platform focused on promoting healthy lifestyles and reducing cardiometabolic risk.¹¹ In 2015, a randomized controlled trial (RCT) was conducted to evaluate PREDIRCAM2, a web-based support system for obesity treatment and diabetes prevention, showing comparable weight loss results between the intervention and usual care groups.¹²

Subsequently, the PREDIRCAM2 web platform was adapted into an mHealth tool, ANíMATE (Adherence, Nutrition, Physical Activity, Motivation, MedicAtion, Technology, and Empowerment),¹³ enabling continuous monitoring of variables such as diet, weight, and PA, while personalizing notification content according to reported behaviors and habits.

The aim of the present study was to evaluate the impact of the ANíMATE mHealth app on weight loss, weight-related variables, and patient-reported outcomes among Spanish adults with obesity.

Methods

Design and study population

This study was an exploratory, 4-month, open-label, parallel-group, single-center RCT conducted at a tertiary care hospital in Barcelona, Spain. Participants were enrolled during routine follow-up visits in the Endocrinology and Occupational Health Departments.

Inclusion criteria were adults (≥18 years) with a BMI between 30 and 39.99 kg/m² who owned a compatible smartphone (Android 7.0 or higher) and had access to a weighing scale.

Exclusion criteria included the presence of major severe obesity-associated comorbidities (HBP treated with >3 medications, diabetes treated with insulin, dyslipidemia treated with PCSK9 inhibitors, OSA requiring continuous positive airway pressure, coronary heart disease, or stroke), use of medications, apps, or any treatment affecting weight within 6 months prior to study enrollment (Supplementary Table 1), previous bariatric surgery, pregnancy or planned pregnancy, and any medical condition likely to interfere with the study protocol.

Sample size

The study was designed as exploratory; therefore, no formal sample size calculation was performed. Sample size was established at 18 patients per group based on recruitment projections and to ensure that at least 15 patients per group would remain available for analysis, considering an anticipated dropout rate of 15%.

Randomization

Eligible participants, after reviewing the information sheet and signing the informed consent form, were randomly assigned in a 1:1 ratio to the intervention group (IG) or active control group (CG) by one investigator (GC) using a sequence generated at https://www.sealedenvelope.com/ by another investigator (RC). To ensure allocation concealment, the sequences were stored in opaque envelopes. Block randomization stratified by obesity class (I or II) was implemented to balance baseline characteristics between the 2 groups.

Data collection and outcomes

Age and sex data were collected at baseline from electronic health records. Participants were also asked about their highest educational level and employment status. Additionally, an ad hoc questionnaire regarding knowledge and use of technology was completed during the first visit (Supplementary Table 2). At baseline and at all follow-up visits, measurements of weight, BMI, waist circumference (WC), systolic blood pressure (SBP), and diastolic blood pressure (DBP) were obtained. HbA1c was measured using a point-of-care method (Roche® Cobas B101) at baseline and at 2 and 4 months.

The primary endpoint was total weight loss percentage (%TWL) at 4 months. Secondary endpoints included %TWL at 2 months, excess body weight loss percentage (%EBWL) at 2 and 4 months, and changes in BMI, WC, SBP, DBP, and HbA1c from baseline to 2 and 4 months between groups.

At baseline and at the 4-month visit, participants completed the following questionnaires: Mediterranean Diet Adherence Screener (MEDAS), Spanish version; International Physical Activity Questionnaire-Long Form (IPAQ-LF), Spanish version; Attitudes Toward Change in Eating Disorders (ACTA); and the 36-Item Short Form Health Survey (SF-36), Spanish version (Supplementary Tables 3–6).

MEDAS assesses dietary habits characteristic of the Mediterranean diet, with scores ≤6, 7–8, and ≥9 corresponding to low, acceptable, and high adherence, respectively.¹⁴ IPAQ-LF provides a comprehensive evaluation of daily PA and demonstrates good reliability in the Spanish population.¹⁵ ACTA is a self-reported instrument designed to assess attitudes toward change in eating disorders.¹⁶ SF-36 was used to assess QoL, with low scores indicating poor health status.¹⁷

Other secondary endpoints included changes from baseline to endpoint in dietary and exercise behaviors, QoL, and adherence, assessed through visit attendance rate, uptake of proposed recommendations, and maintenance of weight and food records.

At each follow-up visit, CG participants were required to provide a food diary from the preceding week, as well as a weekly weight record. IG participants, who documented these records within the ANíMATE app, were required to complete a food diary covering at least 2 working days and 1 nonworking day from the previous week and maintain a minimum of 75% weekly weight records. Additionally, IG participants were asked whether they continued using the app 2 months after study completion.

Finally, satisfaction with the ANíMATE app was assessed at the end of the study using an ad hoc questionnaire (Supplementary Table 7).

ANíMATE app: development, content, and use of the intervention

ANíMATE is an Android-compatible medical app developed by the Universidad Politécnica de Madrid in collaboration with healthcare professionals (HCPs). The app is based on social cognitive theory and incorporates key behavioral change techniques, including knowledge enhancement, goal setting, barrier identification, self-monitoring, and feedback.

The app provides several functionalities organized into 3 modules:

• •

  “Home”: Provides general information and facilitates self-monitoring. ANíMATE integrates with the Fatsecret® app for dietary and weight tracking and with Google® Fit for automatic PA registration or manual entry of specific activities. Patients can also access previous notifications, reports, and goals.

• •

  “Progress”: Displays graphs tracking the patient's weight and energy balance progress.

• •

  “Extra”: A library containing general health information, recipes, and PA recommendations.

Fig. 1 shows a screenshot of the ANíMATE app.

Fig. 1.

ANíMATE screenshots of the “Home,” “Progress,” and “Extra” sections.

ANíMATE encourages patients through various push notifications, including alerts, recommendations, congratulatory messages, reminders, and tracking questions designed to identify unrecorded foods and habits. These notifications provide support and advice regarding healthy habits, behavioral change strategies, motivational encouragement, and reminders to help users remain engaged.

Intervention

At baseline, the recommended caloric intake for weight loss was calculated. Resting metabolic rate was estimated using the Harris–Benedict equation, with target weight corresponding to a BMI of 25 kg/m² and adjusted according to activity factor (1.3 for sedentary/light activity; 1.7 in men and 1.6 in women for moderate activity; and 2.1 in men and 1.9 in women for vigorous activity). To achieve a weight loss of 0.5–1 kg per week, a daily caloric deficit of 2092 Kj¹⁸ was recommended, assuming that approximately half would be achieved through exercise and half through reduction of the recommended daily caloric intake.

Participants in the CG received standard support based on a personalized Mediterranean diet tailored to their recommended caloric intake (40%–55% carbohydrates, 15%–25% protein, and 25%–35% fat).¹⁹ They were also provided with a sample weekly meal plan, received PA counseling, and were instructed to perform weekly self-weighing and complete a diet and exercise diary during the week preceding each appointment.

Participants in the IG received assistance with downloading, installing, and using the app, which was configured according to the recommended caloric intake and PA objectives. They were instructed to log weight, dietary intake, and exercise data within the app.

Both groups attended 3 face-to-face visits (baseline, 2 months, and 4 months) with an endocrinologist. At each follow-up visit, reinforcement of dietary and PA counseling was provided.

Statistical analysis

Because of the small sample size, nonparametric tests were used. Categorical data were expressed as frequencies and percentages, whereas continuous variables were expressed as median and IQR, P25–P75. Differences were assessed using the Mann–Whitney U test for ordinal and interval variables and the χ² test for categorical variables. Data were analyzed according to the intention-to-treat principle. Statistical analyses were performed using IBM SPSS version 29.0 (Chicago, Illinois).

To evaluate differences within the IG according to participants’ level of app use, participants were clustered into 2 groups based on the functional form of app usage over time (Supplementary Statistical Methods).

Ethics

The study protocol, consent forms, and patient information sheets were approved by the Institutional Review Board of Hospital de la Santa Creu i Sant Pau (IIBSP-ANI-2020-10). All participants provided written informed consent before study procedures were initiated. The study was registered at ClinicalTrials.gov (NCT05236881; registration date, January 24, 2022).

The Consolidated Standards of Reporting Trials (CONSORT) reporting guidelines were followed (Supplementary Table 8).

Results

Recruitment began in April 2022 and concluded in June 2024, marking completion of participant enrollment; the study ended in January 2025. Thirty-six participants were assessed for eligibility and randomized, with 18 assigned to each group. Nine participants (25%) withdrew from the study: 3 from the CG and 6 from the IG, with no significant difference between groups (P = .245). In 6 cases, withdrawal was due to loss of interest (Supplementary Fig. 1).

Baseline participant characteristics are summarized in Table 1. Median age was 53 years, median BMI was 32.9 kg/m², and 31 of 36 participants were women. Significant differences were observed in sex distribution, with the CG composed exclusively of women (vs 72.2% women in the IG; P = .005), and in diastolic blood pressure (DBP), which was lower in the IG (76 vs 85 mmHg; P = .012).

Table 1.

Baseline characteristics of study participants. Data are expressed as n (%) or median (IQR, P25–P75).

	Alln = 36	ANíMATE groupn = 18	Control groupn = 18	P value
Age (years)	53.5 (49.3–59.0)	55.0 (52.3–59.3)	52.5 (43.0–56.0)	.18
Sex (female)	31 (86.1)	13 (72.2)	18 (100)	.01*
Education				.41
Elementary school	1 (2.9)	0 (0)	1 (5.6)
Middle school	3 (8.3)	1 (5.6)	2 (11.1)
High school	3 (8.3)	2 (11.1)	1 (5.6)
Vocational school	12 (33.3)	8 (44.4)	4 (22.2)
University	17 (47.2)	7 (38.9)	10 (55.5)
Employment status				1.00
Student	2 (5.5)	1 (5.5)	1 (5.5)
Employed	28 (77.8)	14 (77.8)	14 (77.8)
Unemployed	0 (0)	0 (0)	0 (0)
Retired	6 (16.7)	3 (16.7)	3 (16.7)
Weight (kg)	88.2 (79.6–94.1)	83.3 (77.4–96.8)	88.6 (85.5–91.7)	.55
BMI (kg/m^2)^a	32.9 (31.8–34.6)	32.9 (31.5–34.9)	32.9 (32.1–34.6)	.67
WC (cm)^a	109.0 (103.5–113.5)	108.0 (99.5–118.0)	109.0 (104.5–113.0)	.67
SBP (mmHg)^a	123 (112–133)	117 (111–130)	129 (113–135)	.18
DBP (mmHg)^a	81 (76–88)	76 (74–84)	85 (78–91)	.01*
HbA1c (%)^a	5.5 (5.3–5.8)	5.7 (5.3–5.9)	5.4 (5.2–5.8)	.16
Technology use
Smartphone use^b (0–5)	5 (5–5)	5 (5–5)	5 (5–5)	.87

BMI, body mass index; WC, waist circumference; SBP, systolic blood pressure; DBP, diastolic blood pressure; HbA1c, glycated hemoglobin.

* indicates p value < 0.05.

a

Sixteen in the IG, 17 in the CG.

b

Twelve in the IG, 15 in the CG.

Weight-related outcomes

The primary endpoint, %TWL at 4 months, did not differ between groups: IG, 2.73 (IQR, 0.86–5.06) versus CG, 1.23 (IQR, −0.23 to 3.21); P = .283. No differences were observed in additional weight-related outcomes (Table 2).

Table 2.

Study outcomes: weight-related variables, blood pressure, and glycated hemoglobin. Data are expressed as median (IQR, P25–P75).

	2 monthsIG n = 12CG n = 13	4 monthsIG n = 12CG n = 15
Total weight loss (%)
IG	2.87 (0.13, 3.68)	2.73 (0.86, 5.06)
CG	0.71 (−0.83, 2.74)	1.23 (−0.23, 3.21)
P	.21	.28
Excess body weight loss (%)
IG	10.65 (0.55, 13.88)	7.95 (3.75, 25.97)
CG	2.66 (−3.04, 12.09)	4.44 (−0.64, 16.04)
P	.30	.31
Δ BMI (kg/m^2)
IG	−0.97 (−1.40, −0.04)	−0.99 (−1.58, −0.27)
CG	−0.24 (−0.89, 0.29)	−0.41 (−1.00, 0.09)
P	.19	.24
Δ Waist circumference (cm)
IG	−1.0 (−4.75, 0.75)	−1.0 (−6.5, 1.5)
CG	−2.0 (−2.5, 1.0)	−1.0 (−2.0, 2.0)
P	.76	.51
Δ Systolic blood pressure (mmHg)
IG	−1.5 (−9.3, 9.0)	−2.5 (−7.8, 4.3)
CG	−5.0 (−16.0, 9.5)	0.0 (−9.0, 6.0)
P	.53	.63
Δ Diastolic blood pressure (mmHg)
IG	0.5 (−4.25, 6)	1.5 (−4.5, 4.0)
CG	−4.0 (−6.5, 0.5)	−4.0 (−6.0, 4.0)
P	.11	.27
Δ HbA1c (mmol/mol)
IG	0.0 (−0.1, 0.1)	0.05 (−0.2, 0.1)
CG	−0.1 (−0.2, 0.2)	0.0 (−0.2, 0.2)
P	.91	.90

BMI, body mass index; HbA_1c, glycated haemoglobin.

Dietary and exercise behaviors and QoL

At baseline, no differences were observed in MEDAS score, PA levels, or ACTA score. Regarding QoL, the CG reported worse baseline scores on 3 SF-36 subscales (physical functioning, pain, and health change).

After 4 months, no significant between-group differences were observed in MEDAS, ACTA, SF-36 overall score, or SF-36 subscales (Table 3). The only difference was a greater increase in PA in the CG (2721 [IQR, 1049–5940] vs 923 [IQR, −8081 to 639] MET-min; P = .020), although no differences were found in stratified PA levels, which were high in both groups.

Table 3.

Patient-reported outcome measures: Mediterranean diet, physical activity, eating disorder, and quality-of-life scores. Data are expressed as median (IQR, P25–P75).

Questionnaire (min–max score)	Baseline		Δ 4 months
	Score	n	Score	n
MEDAS (0^–14)
IG	9.0 (8.0–11.0)	11	1.0 (1.0–2.0)	8
CG	7.0 (5.0–9.0)	11	2.0 (−1.5 to 4.5)	9
P value	.06		.81
IPAQ-LF
Physical activity (MET-min/week)
IG	8794 (1812–12,270)	10	923 (−8081 to 639)	6
CG	6546 (2639–11,067)	9	2721 (1049–5940)	8
P value	.74		.02*
Physical activity (0 = low, 1 = moderate, 2 = high)
IG	3.0 (2.0–3.0)	10	0.0 (−1.0 to 0.25)	6
CG	3.0 (2.0–3.0)	9	0.0 (0.0–0.75)	8
P value	1.00		.23
Sedentary (min/week)
IG	2040 (1380–3540)	10	−360 (−1065 to −30)	6
CG	2280 (960–4170)	9	−120 (−3960 to 180)	7
P value	.84		.89
ACTA (0–40^)
Precontemplation
IG	8.5 (5.8–12.5)	10	−1.0 (−6.0 to 4.0)	7
CG	5.0 (2.8–9.0)	10	1.0 (−3.0 to 3.0)	8
P value	.26		.68
Contemplation
IG	15.9 (7.0–20.2)	10	−0.91 (−8.2 to 4.5)	7
CG	15.9 (7.7–17.5)	10	−0.46 (−4.3 to 2.7)	8
P value	.94		.69
Preparation
IG	10.9 (4.3–23.9)	10	−6.4 (−9.1 to 0.91)	7
CG	12.7 (1.6–20.5)	10	3.2 (−3.0 to 6.1)	8
P value	.79		.11
Action
IG	24.5 (5.5–29.0)	10	0.0 (−7.0 to 10.0)	7
CG	26.0 (3.0–28.0)	10	4.5 (−2.3 to 9.0)	8
P value	.88		.64
Maintenance
IG	22.5 (20.8–26.3)	10	1.0 (−2.0 to 2.0)	7
CG	22.5 (19.0–26.3)	10	1.0 (−3.8 to 3.8)	8
P value	.70		.86
Relapse
IG	9.3 (2.9–19.3)	10	−2.9 (−8.57 to 0.0)	7
CG	13.6 (5.7–19.3)	10	−1.4 (−7.5 to 3.2)	8
P value	.62		.60
SF-36 (0–100^)
Physical functioning
IG	95.0 (83.8–100)	10	0.0 (0.0–5.0)	7
CG	80.0 (67.5–90.0)	10	−5.0 (−5.0 to 12.5)	9
P value	.03*		.52
Role physical
IG	100 (68.8–100)	10	0.0 (−100 to 0.0)	7
CG	100 (68.8–100)	10	0.0 (−25.0 to 12.5)	9
P value	1.00		.44
Role emotional
IG	100 (33.3–100)	10	0.0 (0.0–0.0)	7
CG	100 (83.3–100)	10	0.0 (0.0–0.0)	9
P value	.65		.59
Energy
IG	60.0 (45.0–68.8)	10	5.0 (−5.0 to 15.0)	7
CG	40.0 (18.8–67.5)	10	5.0 (−10.0 to 22.5)	9
P value	.18		1.00
Emotional well-being
IG	76.0 (61.0–89.0)	10	8.0 (−12.0 to 12.0)	7
CG	72.0 (63.0–93.0)	10	−8.0 (−10.0 to 4.0)	9
P value	.94		.40
Social functioning
IG	81.3 (59.4–100)	10	0.0 (−25.0 to 12.5)	7
CG	87.5 (71.8–100)	10	0.0 (−31.3 to 0.0)	9
P value	.61		.48
Pain
IG	90 (58.8–100)	10	−10.0 (−22.5 to 0.0)	7
CG	35.0 (22.5–80.0)	10	0.0 (−11.3 to 23.8)	9
P value	.02*		.23
General health
IG	65.0 (53.8–77.5)	10	0.0 (−15.0 to 10.0)	7
CG	52.5 (43.8–66.3)	10	5.0 (−5.0 to 25.0)	9
P value	.10		.34
Health change compared with 1 year earlier
IG	62.5 (50.0–81.3)	10	0.0 (0.0–25.0)	7
CG	50.0 (25.0–50.0)	10	25.0 (0.0–25.0)	9

* indicates p value < 0.05.

^ indicates worse questionnaire results.

CG, control group; IG, intervention group; MEDAS, Mediterranean Diet Adherence Screener; IPAQ-LF, International Physical Activity Questionnaire-Long Form; ACTA, Attitudes Toward Change in Eating Disorders; SF-36, 36-Item Short Form Health Survey.

Adherence and satisfaction

Visit adherence was 96.7% in the CG and 100% in the IG. Uptake of physician recommendations regarding diet and/or exercise exceeded 90% at 2 months and 75% at 4 months, with exercise demonstrating the lowest adherence. More than 50% of participants provided food records at 2 months and approximately 40% at 4 months, with no differences between groups. Finally, adherence to weekly self-weighing was significantly higher in the CG at 2 months but not at 4 months (Table 4).

Table 4.

Adherence to visits, physician recommendations, food records, and weight records. Data are expressed as n (%).

	2 months		4 months
	n (%)	n	n (%)	n
Appointments attended
IG	12 (100%)	12	12 (100%)	12
CG	13 (87%)	15	15 (100%)	15
P value	.19		1.00
Uptake of recommendations
IG	None: 1 (8.3%)Diet: 4 (33.3%)Exercise: 2 (16.7%)Both: 5 (41.7%)	12	None: 3 (25.0%)Diet: 5 (41.7%)Exercise: 1 (8.3%)Both: 3 (25.0%)	12
CG	None: 1 (7.7%)Diet: 7 (53.8%)Exercise: 1 (7.7%)Both: 4 (30.8%)	13	None: 4 (26.7%)Diet: 3 (20.0%)Exercise: 0 (0%)Both: 8 (53.3%)	15
P value	.47		.39
Food record completion
IG	6 (54.5%)	11	5 (45.5%)	11
CG	9 (69.2%)	13	6 (40.0%)	15
P value	.46		.78
Weight record completion
IG	4 (36.4%)	11	4 (36.4%)	11
CG	10 (76.9%)	13	11 (73.3%)	15

* indicates p value < 0.05.

CG, control group; IG, intervention group.

Participants in the IG rated the ANíMATE app 16.5/20 (IQR, 13.5–20), and 2 months after study completion, 3 of 12 IG participants continued using the app.

No important harms or unintended effects were identified in any participant.

Differences within the IG according to level of app use

Participant trajectories were grouped into 2 clusters using the k-means algorithm (Supplementary Fig. 2). No differences in weight loss were observed according to the intensity of app use (Supplementary Table 9).

Discussion

In this exploratory open-label RCT, individuals with obesity using the ANíMATE app showed no differences in the primary outcome (%TWL at 4 months) or in most secondary outcomes.

Despite randomization, some baseline differences were observed between groups regarding sex distribution, DBP, and 3 SF-36 subscales, overall suggesting better health status in the IG. Because most outcomes were assessed as changes from baseline, we decided not to adjust for baseline differences.

Both groups showed reductions in weight and WC, with the IG demonstrating a nominally greater %TWL (2.87% at 2 months and 2.73% at 4 months) compared with the CG (0.71% and 1.23%, respectively), accompanied by parallel changes in %EBWL and BMI. A recent systematic review concluded that smartphone apps may have some effect on weight reduction, although these effects are generally small and of limited clinical significance, particularly in the short- to medium-term.⁸ The effects on weight change observed in the present study are comparable to those reported in similar mHealth programs. For example, the EVIDENT3 study in a Spanish population reported weight reductions of 2.05% in the IG and 1.1% in the CG (minimal intervention) at 3 months.²⁰ Additionally, evidence comparing apps with usual care remains very limited, with only 1 study reporting participant satisfaction data but not weight-related outcomes.²¹ The present RCT provides novel evidence comparing app-based intervention with an active CG.

Regarding dietary behavior, both groups showed higher MEDAS scores at study completion, without significant between-group differences. Because the study population already demonstrated acceptable-to-high adherence to the Mediterranean diet at baseline, the potential for further improvement was limited.

Furthermore, both the IG and CG increased PA levels. As demonstrated in former studies, mHealth interventions may promote small-to-moderate increases in PA, with effects maintained long term although diminishing over time.²² Although the net increase in PA at 4 months was greater in the CG, no significant differences were observed in stratified PA levels, which remained high in both groups. We currently have no clear explanation for this difference, especially considering that the CG reported more pain in the SF-36 subscale assessment. Limited app use may have contributed to this finding.

Although baseline scores for the physical functioning, pain, and health change subscales of the SF-36 were worse in the CG, no significant differences in QoL changes between groups were observed over time. In obesity, evidence regarding QoL benefits derived from lifestyle-based interventions is inconsistent and inconclusive,²³ and only 1 previous app-based study assessed QoL,²⁴ showing little or no difference at 12 months. Several factors may explain these findings in our study, including the use of a generic QoL questionnaire, which may not adequately capture obesity-related QoL.

Adherence to follow-up visits and physician recommendations was generally good. Low adherence to PA recommendations in both groups is consistent with known physical and psychosocial barriers among individuals with obesity (e.g., pain, stigma, and safety concerns).²⁵ Recording dietary intake showed a downward trend, and by study completion, food records were provided by fewer than 50% of participants. One possible explanation is that this task may be burdensome and time-consuming, potentially promoting nonadherence and underestimation of actual intake through omission or modification of consumed foods. Consistent with this observation, weekly self-weighing adherence was significantly higher in the CG at 2 months. One possible explanation is that IG participants had the additional task of logging weight into the app to receive feedback. Because record keeping is closely associated with weight loss,²⁶ this additional burden may have contributed to the absence of clear differences between groups in weight-related outcomes. To explore this possibility, differences in weight loss according to app use intensity were analyzed, with no significant changes observed in outcomes. This occurred despite the potential of mHealth apps to improve adherence by providing flexibility, support for self-monitoring through food databases, wireless scales and PA trackers, and streamlined processes designed to make tracking faster and simpler. These findings highlight the importance of continuously improving app design to promote user engagement.

Satisfaction with the ANíMATE app was high (mean score, 16.5/20), with users particularly valuing the design, usability, and access to historical food, PA, and weight data (Supplementary Table 10). Nevertheless, only 25% of participants continued using the app 2 months after study completion, underscoring the challenge of maintaining engagement, especially given the close relationship between self-monitoring and successful weight loss.²⁷ This finding further emphasizes the need to develop designs that facilitate sustained participant interaction.

Weight loss studies frequently encounter high dropout rates.²⁷ In our study, dropout rates were nominally higher than expected (25% vs 15%), although no significant differences were observed between groups. These results are consistent with dropout rates reported in other mHealth studies, ranging from 7% to 47%.²⁸ Possible explanations include ambitious weight loss goals and difficulty maintaining self-monitoring behaviors.

This study has several strengths. The principal strength is the evaluation of a custom-developed app created by engineers in collaboration with HCPs. Additional strengths include accurate and secure data management, high-quality health information, incorporation of established behavior change tools, and detailed outcome analyses. Furthermore, user experience and satisfaction were comprehensively evaluated, providing valuable insights. Future versions incorporating user feedback to improve usability and simplify meal logging may enhance retention and adherence.

The main limitation of our study is its sample size. The small sample and participant dropout reduced statistical power, although this is consistent with the exploratory design. In addition, the short duration of the intervention prevented evaluation of sustained effects, which is especially relevant in a chronic disease such as obesity. Furthermore, despite the use of blocked randomization, baseline differences between groups were observed. Sex distribution was not well balanced, and most participants were women despite the higher prevalence of obesity among men,¹ a pattern commonly observed in weight loss studies²⁹ and possibly related to sex-based differences in perception of obesity-related implications. Differences in DBP and QoL were also identified; however, analysis based on changes from baseline should minimize their impact. Finally, app satisfaction was assessed using a nonvalidated ad hoc questionnaire rather than a standardized instrument.³⁰

Conclusions

In this exploratory study, the ANíMATE app-based obesity treatment produced modest weight loss and improvements in obesity-related health outcomes. Compared with the CG, PA at 4 months and self-weighing adherence at 2 months were lower. Given its potential to address gaps in obesity care, additional long-term research is needed to develop tailored mobile health strategies for weight management that promote adherence, retention, and effectiveness, as no clear benefit over the active CG was demonstrated.

Authors’ contributions

Gemma Cuixart: Conceptualization, methodology, validation, formal analysis, investigation, resources, writing – original draft, writing – review and editing.

Antonio Cobo: Software, formal analysis, data curation, writing – review and editing.

José M. Iniesta-Chamorro: Software, writing – review and editing.

José Tapia-Galisteo: Software, writing – review and editing.

Gema García-Sáez: Software, writing – review and editing.

Rosa Corcoy: Conceptualization, methodology, validation, formal analysis, writing – review and editing, supervision.

M. Elena Hernando: Software, writing – review and editing.

Cintia González: Conceptualization, methodology, validation, formal analysis, writing – review and editing, supervision.

All authors agree with the manuscript and declare that the content has not been published elsewhere.

Funding

None declared.

Conflicts of interest

The authors certify that they have no affiliations with or involvement in any organization or entity with any financial or nonfinancial interest in the subject matter or materials discussed in this manuscript, including honoraria, educational grants, participation in speakers’ bureaus, membership, employment, consultancies, stock ownership or other equity interests, expert testimony, patent-licensing arrangements, personal relationships, professional relationships, affiliations, knowledge, or beliefs.

Acknowledgments

The authors thank the participants who agreed to take part in this study, as well as the healthcare professionals who provided their assistance. The authors also thank the Clinical Epidemiology Department for guidance in data analysis and Carmen Pérez-Gandía for her contribution to the application design and development.

This work was supported by the Research Institute of Electrical Communication, Tohoku University and the CIBER-BBN ANíMATE Early-Stage Project.