Mobile App Users Do Better Losing Weight
Overview
With the rise in global obesity rates and related health problems, there is a growing need for accessible weight management solutions. Mobile applications provide such accessibility and can serve as effective alternatives to traditional healthcare models for weight management.
This study aims to evaluate the effectiveness of the SIMPLE mobile application in facilitating weight loss (WL) through time-restricted eating.
User data from January 2021 to January 2023 was analyzed. In-app activity was measured as the proportion of active days over 12, 26, and 52 weeks, with a day considered active if it included at least one in-app action (e.g., logging weight, food intake, fasting, or physical activity). Users were classified into four activity levels: inactive (in-app activity <33%), medium activity (33%-66%), high activity (66%-99%), and maximal activity (100%). Weight change across these activity groups was evaluated at 12, 26, and 52 weeks.
Among 53,482 users, a positive correlation was observed between the use of the SIMPLE app and weight loss. Users with higher app engagement lost more weight compared to less active users. Median weight loss for active users was 4.20%, 5.04%, and 3.86% at 12, 26, and 52 weeks, respectively. Up to 50.26% of active users achieved clinically significant weight loss (≥5%), a higher percentage than inactive users. A dose-response relationship was identified, showing that a 10% increase in app activity corresponded with additional weight loss of 0.43 kg, 0.66 kg, and 0.69 kg at 12, 26, and 52 weeks, respectively, after adjusting for gender, age, and initial Body Mass Index.
The findings indicate that the SIMPLE app effectively supports weight loss, with greater engagement leading to better outcomes. Mobile health applications provide a viable and effective solution for weight management and should be considered in strategies to support adults in losing weight.
Introduction
The prevalence of overweight and obesity has surged in recent decades, becoming a global public health concern. Obesity is linked to early mortality and a reduced quality of life due to its association with various chronic diseases. Early interventions, such as improved nutrition, increased physical activity, and behavior change strategies, can effectively prevent or slow the progression of these diseases, often through weight loss (WL). Intermittent fasting is one such intervention. This approach alternates periods of fasting, characterized by minimal or no calorie intake, with periods of eating. A popular form of intermittent fasting is time-restricted eating (TRE), which limits the hours during which food and calorie-containing drinks are consumed. TRE schedules vary, commonly including patterns like 12:12, 14:10, 16:8, 18:6, or 20:4, where the first number indicates fasting hours and the second the eating window.
Research has demonstrated that TRE can lead to WL and may also promote overall health through various physiological processes activated during fasting. Intermittent fasting has been associated with reduced risks of chronic diseases such as Type 2 diabetes and heart disease, and it may improve metabolic health through mechanisms linked primarily to WL. Additionally, intermittent fasting can become a sustainable lifestyle habit that aids in maintaining WL.
Mobile applications offer a more accessible and less resource-intensive alternative to in-person care, with evidence suggesting their effectiveness in facilitating WL. A study by Torres et al. (2022) showed significant WL among participants using an intermittent fasting app. Similarly, Valinskas et al. (2023) found that higher engagement with an intermittent fasting app correlated with greater WL in participants with obesity. However, further research with larger participant groups is needed to validate these findings and extend the evidence to individuals with overweight.
This study aimed to investigate the effectiveness of a mobile app in supporting adult WL and to examine the relationship between app engagement levels and WL success.
Methods
The study utilized data from SIMPLE, a free mobile application with several premium features, developed by Simple.Life Apps Inc. The SIMPLE app, available in multiple languages including English, French, German, Italian, Portuguese, and Spanish, is designed to support users in achieving and maintaining weight loss (WL) goals by promoting time-restricted eating (TRE), nutrition, and lifestyle habits. It includes trackers for fasting periods, food intake, beverages, physical activity, and weight, and can integrate data from third-party platforms like Apple Health, Google Fit, smart scales, and wearable devices. The app also provides users with reminders, an extensive library of educational content on intermittent fasting (IF), healthy eating, workouts, and personalized feedback based on in-app behaviors.
Weight, height, and upper BMI limits were determined based on the dataset’s 1st and 99th percentiles to include a wide range of participants while excluding outliers. Missing data was not an issue due to the filtration process. Data collected included self-reported gender, height, age, baseline weight, and continuous logs of food, drink, activity, fasting, and weight throughout the study period. Weight records showing changes exceeding 5% within seven days were excluded as outliers. Weight change was evaluated at 12, 26, and 52 weeks, calculated as the difference between the baseline weight and the recorded weight closest to the observation period’s end.
User engagement, defined as in-app activity, was the study’s independent variable of interest. An active day was one where users logged weight, physical activity, fasting (started or finished a fast), food, or drink. In-app activity was expressed as a proportion of active days within an observation period, ranging from 0.0% to 100.0%. Users were categorized into groups based on their in-app activity: inactive (0-33%), medium (33-66%), high (66-100%), and maximal (100%). Active Users were defined as those who logged activity on at least 33% of days within an observation period.
Weight change was assessed by calculating the absolute and percentage difference between baseline and final recorded weights within each observation period. Additionally, the proportion of users achieving clinically significant WL (≥5% and ≥10% of body mass) in each period was determined, along with the proportion of users who either lost weight or maintained their weight (defined as a weight change within ±1% of their baseline body weight).
Inclusion Criteria
Participants were included in the study if they began using the SIMPLE app between January 1, 2021, and January 1, 2023, met the following criteria: aged 18-70 years; baseline body weight of 45-250 kg; height of 150-250 cm; baseline Body Mass Index (BMI) of 25-60 kg/m2; selected a TRE schedule from 12:12 to 23:1; recorded body weight at least three times during the study; and had at least 12, 26, or 52 active days within 12, 26, or 52 weeks of observation, respectively, with at least one weight entry within 90 days of the baseline.
Exclusion Criteria
Exclusion criteria included pregnancy, breastfeeding, or reported eating disorders.
Statistical Analysis
The analysis utilized Python v3.9.13, along with the SciPy v1.9.1 and statsmodels v0.13.5 libraries. The Kruskal-Wallis test assessed statistical significance for differences in continuous metrics between activity groups, while the Chi-squared test was used for categorical metrics.
Linear regression examined the relationship between user in-app activity and weight loss (WL), considering both WL in kilograms and WL percentage. Logistic regression evaluated the association between user in-app activity and achieving ≥5% or ≥10% WL.
For each observation period, five models were created for each outcome of interest using the following independent variables: overall in-app activity (percentage of active days during the observation period); sex (male or female); age (years in decades); and initial BMI (kg/m²). To illustrate the effects of age and in-app activity on WL, age in decades and activity percentage in tens were utilized. A Bonferroni correction was applied for multiple comparisons, setting the alpha level of significance at 0.00079 for all tests.
The statistical significance of each coefficient was determined using a t-statistic. Model quality was evaluated using F-statistics, examining multicollinearity among independent variables, analyzing the distribution of residuals, and verifying linearity between dependent and independent variables. Additionally, skewness and kurtosis metrics were used to assess the normality of the residual distribution.
Ethical approval for this retrospective study was obtained from the Independent Ethical Review Board (WCG IRB) (Study Number: 1348695) in February 2023.
Result
Following the eligibility assessment and the removal of outliers, the study included data from 36,950 users at 12 weeks, 22,090 users at 26 weeks, and 14,240 users at 52 weeks, with a total of 53,482 unique participants.
The distribution of in-app activity exhibited a bimodal pattern across all three observation periods, highlighting a significant proportion of highly active app users.
The study observed significant weight loss (WL) among Active Users over 52 weeks. Active Users, defined as those with medium to maximal in-app activity, consistently demonstrated notable WL. At the 12-week mark, over 42% of Active Users achieved at least a 5% WL.
At 12 weeks, 20.24% (95% CI: 19.43, 21.05) of inactive users and 53.15% (95% CI: 51.88, 54.41) of maximum in-app activity users achieved at least 5% WL. Linear trends in WL relative to in-app activity levels were also evident at 26 and 52 weeks, with the maximal activity group consistently showing the highest median WL.
After adjusting for sex, age, and baseline BMI using regression analysis, the increase in WL with increased app usage remained statistically significant across all observation periods. A higher baseline BMI was consistently associated with greater WL, while older participants tended to lose less weight at both the 12th and 26th weeks. Male participants showed better WL performance only during the first 12 weeks, after which female app users tended to lose more weight. All linear and logistic regression assumptions were satisfied.
Linear regression analysis estimated that every additional 10% of in-app activity (equivalent to logging an action on three additional days within a 30-day period) increased absolute WL by 0.43 kg (95% CI 0.41, 0.44) after 12 weeks, by 0.66 kg (95% CI 0.63, 0.69) after 26 weeks, and by 0.69 kg (95% CI 0.65, 0.74) after 52 weeks. Conversely, every additional 10 years of age was associated with a decreased absolute WL of 0.15–0.25 kg.
Logistic regression analysis indicated that every additional 10 years of age decreased the likelihood of achieving minimal clinically significant WL by 1.08–1.12 times, depending on the observation period. Increased app usage, with every extra 10% of in-app activity (logging an action on three additional days within a 30-day period), increased the likelihood of achieving ≥5% WL by 1.2–1.23 times.
Conclusion
This study demonstrates the effectiveness of using the SIMPLE mobile application for achieving weight loss (WL) in adults with overweight and obesity in a real-world setting. Users of the app across various activity levels (low, medium, high, and maximal) managed to attain clinically significant WL of ≥5%. The median WL and the proportion of users achieving ≥5% WL increased with more frequent app activity.
Achieving ≥5% WL can benefit individuals with obesity by improving related comorbidities such as reducing the risk of Type 2 diabetes, improving glycemic control in those with Type 2 diabetes, enhancing lipid profiles, and mitigating hepatic steatosis. In this study, 42.20% (95%CI: 42.95, 45.44) of active users lost ≥5% of their body weight after 52 weeks. Although the study did not specifically measure health outcomes, these results suggest that the SIMPLE app could be an effective tool for enhancing overall health and supporting the national objective of reducing diet-related diseases in the American population.
Interestingly, the study’s findings align with existing literature, indicating that WL peaks at around six months of follow-up. The greatest median WL occurred at 26 weeks across all in-app activity groups. Although WL at 52 weeks remained substantial, no additional WL was observed, with a slight decrease in median WL likely due to users experiencing a WL plateau. These results underscore the importance of the initial six months for maximizing WL outcomes in individuals with overweight and obesity, with mobile applications providing a valuable tool due to their accessibility and interactive features, such as AI chatbots and gamification.
Demographic variations in WL were observed based on baseline BMI, age, and sex. Users with a higher baseline BMI experienced greater absolute and relative WL across all observation periods. Older app users lost less weight at 12 and 26 weeks but this age-related difference was no longer significant at 52 weeks. Gender differences were noted, with women initially losing less weight than men in the first 12 weeks but outperforming men in WL by 26 and 52 weeks. These findings highlight important sex-specific differences in WL outcomes.
A significant finding is the dose-response effect of in-app activity on WL outcomes. Increased in-app activity consistently resulted in greater WL, with the effect persisting and even slightly increasing between weeks 26 and 52. This has been observed in other studies as well, underscoring the impact of app activity on WL.
The study had several strengths, including the large amount of data collected, which provided sufficient statistical power for sound conclusions. The use of regression analysis allowed the separation of the influence of demographic factors from the variable of interest—the overall level of in-app activity on WL. Additionally, the study provides real-world data, enhancing the generalizability of the results.
However, there were limitations. Real-world data, while advantageous, may introduce biases not present in randomized controlled trials. Many app users did not meet the eligibility criteria due to insufficient data, and much of the weight data was self-reported, which could not always be validated. Despite these limitations, extensive outlier removal was conducted to ensure result accuracy.
Given the low accessibility to traditional weight management services, mobile health applications like SIMPLE offer a cost-effective and broadly accessible alternative, which is crucial for managing the growing global obesity epidemic. Real-world evidence is valuable for evaluating digital tools designed to support WL and should be considered when assessing the viability of such interventions.
Encouraging increased app usage and long-term commitment is fundamental to achieving WL, particularly since WL appears to peak around six months. Techniques such as gamification and advanced behavior change strategies could enhance user engagement and WL success and should be considered in the development of mobile health applications.
While it was beyond the scope of this study to evaluate the effects of different time-restricted eating (TRE) schedules on WL, or the individual effects of various in-app activities and their interactions, future research should explore these aspects. The SIMPLE mobile application has proven to be a successful WL tool, with a clear dose-response relationship between in-app activity and WL. Incorporating elements that encourage app usage can help more users achieve significant WL. Mobile applications supporting WL should be considered viable treatment options, and real-world observational studies should be used to gather evidence for treatment decisions for individuals with overweight and obesity.
Oncology Related Tools
Other
Latest Research
