Effect of a Smartphone App on Weight Change and Metabolic Outcomes in Asian Adults With Type 2 Diabetes

Key Points Question What is the effect of a culturally contextualized smartphone-based lifestyle intervention on weight change and metabolic outcomes in Asian adults who have overweight or obesity and type 2 diabetes compared with usual care? Findings In this randomized clinical trial of 204 adults with type 2 diabetes, the use of a smartphone app tracking personal health data and using integrated behavioral modification strategies led to significantly greater reductions in weight and hemoglobin A1c, along with a significantly greater proportion of patients with a reduction in diabetes medication dosages compared with usual care at 6 months. Meaning These findings suggest that a mobile health lifestyle intervention has the potential to improve weight and glycemic outcomes among individuals who have overweight or obesity in an Asian population with type 2 diabetes.


23
Lifestyle intervention programs commonly incorporate diet, exercise, behaviour modifications, 24 consultations with a dietitian or exercise therapist (or both), and a customized diet or exercise 25 plan (or both). It has been shown that success of the treatment is associated with the intensity 26 of lifestyle measures and can lead to greater weight loss in those who are overweight or obese, 27 and a reduction in new-onset diabetes. [4][5][6][7][8] However, studies showing positive results were mostly 28 conducted in controlled environments which may or may not be transferable to the local 29 (Singapore) context. 30 31 Lim SL, Ong KW, Johal J, et al. Effect of a smartphone app on metabolic outcomes in Asian adults with type 2 diabetes: a randomized clinical trial. JAMA Netw Open. 2021;4(6):e2112417. doi:10.1001/jamanetworkopen.2021 Although lifestyle interventions such as diet and exercise programs are considered fundamental 32 in the prevention and management oftype 2 diabetes, 9 current methods of delivering them are 33 not scalable to the at-risk population as they are resource-intensive, costly, time-consuming, 34 have high default rate and limited reach. 35

36
In 2014, the mobile population penetration rate in Singapore was 148%, as it was not 37 uncommon for individuals to own two or more mobile phones, and additionally 87% of 38 households had access to broadband. 10,11 These numbers will only increase with time. Modern technology-enabled interventions such as health-related mobile apps have been 53 increasingly used in facilitating weight loss. However, this technology-enabled modality of 54 treatment is currently fragmented with its weakest link being the diet aspect where compliance 55 is poor due to it being user-unfriendly, not targeted in its advice and not being able to provide 56 individualized diet according to culture and pre-existing medical conditions. 19 A few studies 57 using mobile technologies to facilitate weight loss have been done overseas and mostly in 58 Caucasian populations, but no such studies have been carried out on type 2 diabetes patients. 59 A study using this modality of intervention and conducted in the local context will provide new 60 knowledge on the feasibility and effectiveness of mobile technologies to facilitate lifestyle interventions in people with diabetes and provide guidance for the application of this modality of 62 treatment to other chronic diseases in future. 63

64
We propose to use a smartphone app and remote intervention programs to deliver lifestyle 65 interventions to individuals with diabetes, with the appropriate targeted and individualized 66 support, led and supervised remotely by dietitians. We will use a user-friendly smartphone app 67 to facilitate lifestyle changes (diet and activity) with elements of behavioral modifications 68 embedded, -, including remote coaching. The proposed study will design, implement and 69 evaluate behavioral and lifestyle interventions, using combined smartphone mobile app and 70 remote dietetic coaching program, specifically for effective lifestyle management of diabetes in 71 Singapore. 72

73
The study hypothesis is that a locally-developed and targeted mobile app, combined with 74 remote coaching, will be able to effect positive changes in health-related behaviors and lifestyle 75 choices, manage diabetes, and ultimately improve health outcomes and save healthcare costs. 76

77
Objective 78 The objective of the Diabetes Lifestyle Intervention using Technology Empowerment (D'LITE) 79 study is to compare the effectiveness of a weight loss lifestyle intervention, delivered via the 80 nBuddy Diabetes app and in-app coaching by dietitians, with usual care, on body weight and 81 metabolic profiles among overweight or obese non-insulin-treated Asians with type 2 diabetes. 82 83

Study Design 84
The D'LITE study is a parallel multicenter 2-arm randomized clinical trial (RCT). Participants will 85 be from health screening facilities, government polyclinics, general practitioner clinics and 86 hospital outpatient clinics in Singapore. 87 Eligible participants were randomized to either control or intervention group in a 1:1 allocation 109 ratio via block randomization stratified by gender, BMI (<27.5 or ≥27.5 kg/m 2 ), and age (<50 110 years or ≥50 years), which was changed from previous <40 years or ≥40 years 2 months post-111 recruitment due to a noticeably larger number of older participants. Participants were allocated 112 to either group by drawing personally from sealed stratified opaque envelopes, each containing 113 an equal proportion of intervention and control group assignments. To ensure high quality 114 envelope concealment, a third party personnel not involved in the study prepared the envelopes 115 in advance using matched block method. 116

Intervention 117
All control and intervention participants received a single 45 to 60 minutes session of diet and 118 physical activity advice, as per American Dietetic Association (ADA) guidelines, from a 119 registered research dietitian in the clinic at baseline. All participants were issued a standardized 120 digital weighing scale (Omron HN-289, Japan), and continued to receive standard diabetes care 121 from their usual health care providers.
Participants assigned to the intervention group were required to use the nBuddy Diabetes app 123 for 6 months to track weight twice weekly, diet and physical activity daily and to communicate 124 regularly with the research dietitians via the app. To encourage self-blood glucose monitoring, 125 the intervention group was also provided with a glucometer. Based on the participants' input, the 126 dietitian will provide personalized diet and lifestyle coaching. They will be guided to achieve their 127 individualized goals using in-built evidence based behavior modification tools via the app. These 128 are implemented as daily tips, prompters and decision-support system to motivate them towards 129 their weight goal and blood glucose control. Participants can also review their progress via the 130 weight, calorie intake, physical activity and blood glucose charts in the app. Educational videos 131 will be uploaded for participants weekly via the application to enhance their knowledge in 132 diabetes management. 133 134

Outcomes 135
The primary outcome of this study is weight loss, measured by a standardized digital weighing 136 scale at 3-month and 6-month post randomisation. Secondary outcomes were changes in body 137 weight post 3 months, metabolic profiles (HbA1c, fasting blood glucose (FBG), blood pressure, 138 total cholesterol, triglycerides, low-density and high-density lipoprotein levels), creatinine and 139 dietary intake. These measurements will be re-measured at 1-year and 2-year to assess long 140 term sustainability of the outcomes. Blood samples will be processed at accredited laboratories 141 -National University Hospital Referral laboratories and National Healthcare Group Diagnostics, 142 with technicians blinded to the group allocation. 143 144 Participants in both groups will be required to complete simple ethics-approved surveys at 145 baseline, 3-month, 6-month, 1-year and 2-year. Questions include participants' experience in 146 using mobile apps, study expectations, stage of change, previous diet advice, dietary patterns, 147 physical activity, medication changes and healthcare costs. At year 1 and 2, participants will 148 also be required to complete a lifestyle and behavior questionnaire to better understand their 149 weight loss goals, self-efficacy, physical activity, sedentary hours, factors leading to their weight 150 outcomes and eating behavior. 151 A 2-day food diary, reflecting one weekday and one weekend dietary patterns, will be 152 administered at baseline and subsequent visits. Participants are given basic instructions to list 153 out foods or drinks consumed and to estimate the portion sizes with standard household measures. To assess the change in energy and other macronutrients intake from baseline, 2-155 day food diary will be analyzed using the nBuddy Diabetes app's nutrient analysis database at 156 every time point, which incorporates Singapore Energy & Nutrient Composition of Food, 157 Malaysian Food Composition and USDA food databases, along with nutritional information from 158 food packaging, and nutrient analysis of recipes. 159

Sample Size 160
The sample size is calculated based on assumption of at least a moderate Cohen's effect size 161 of 0.5 for the difference in weight loss between groups at 6 months post randomization. A 162 minimum sample size of 85 participants per group would provide 90% power at 0.05 level of 163 significance (two-sided). A total sample size of 190 participants (95 per group) was planned, 164 factoring a 10% attrition rate. 165

Statistical Analysis 166
All analyses will be performed using SPSS for Windows software (SPSS Inc, Chicago, IL, USA). 167 Descriptive statistics for normally distributed numerical variables will be presented as mean 168 (standard deviation), otherwise median (interquartile range) will be presented. Categorical data 169 will be expressed as frequencies and percentages. Parametric tests (and the appropriate non-170 parametric) will be used to compare numerical variables between the control and intervention 171 groups upon normality & homogeneity assumptions checking. For categorical variables, the chi-172 square or Fisher Exact tests will be used. Between-group differences in the primary outcomes 173 (weight changes) and secondary outcomes (changes in HbA1c, fasting blood glucose, blood 174 lipids, creatinine levels, blood pressure, calorie, carbohydrate, sugar and other nutrients intake, 175 and physical activity) will be investigated using the Generalized Linear Mixed Model analysis, 176 adjusted for baseline measurements and relevant covariates, and accounting for clustering 177 effect of recruitment sources. Post hoc analysis using Generalized Poisson Mixed Model will be 178 conducted to determine the relative risks for diabetic medication reduction. Statistical 179 significance is set at p < 0.05. In the quantitative study, regression models will be used to test 180 for significant treatment effects after controlling for confounders. 6) The primary outcome of this study is weight loss at 6 months, measured in the clinic 215 using a calibrated digital weighing scale.