PhD Defence Eclaire Hietbrink | Tailored eHealth Strategies for Lifestyle Support in Type 2 Diabetes Care

Tailored eHealth Strategies for Lifestyle Support in Type 2 Diabetes Care

Eclaire Hietbrink is a PhD student in the department Biomedical Signals and Systems. (Co)Promotors are prof.dr. G.D. Laverman; prof.dr. M.M.R. Hutten and dr. A. Middelweerd from the faculty of Electrical Engineering, Mathematics and Computer Science (EEMCS), University of Twente.

Lifestyle modification is a cornerstone in the management of type 2 diabetes (T2D). Improvements in diet, physical activity, and sedentary behavior can significantly enhance glycemic control, insulin sensitivity, and overall metabolic health. Despite these benefits, long-term adherence to lifestyle change remains challenging. eHealth interventions provide new opportunities for supporting self-management by offering continuous monitoring and tailored behavioral support. In this way, they can help individuals track their lifestyle patterns, receive tailored advice, and stay motivated between clinical visits.

The acceptability and effectiveness of eHealth depends on how well it is tailored to users’ needs and circumstances. Traditional tailoring approaches often rely on static personal characteristics, whereas dynamic tailoring adapts support over time based on behavioral, contextual, and physiological data. Such approaches may improve engagement and effectiveness, particularly when integrated into routine healthcare. However, questions remain regarding how dynamically tailored eHealth should be designed and implemented in clinical practice.

The overall aim of this thesis was to explore how eHealth strategies can be dynamically tailored to support healthy lifestyles in people with T2D and how such strategies may be integrated into diabetes care.

Chapter 2 examined the current implementation of lifestyle counseling in secondary T2D care from the perspective of diabetes nurses. Semi-structured interviews revealed that while nurses consider lifestyle an essential aspect of care, they experience multiple factors that influence discussing it effectively. These include limited lifestyle expertise, lack of practical guidelines, time constraints, the patient-nurse relationship, and insufficient patient motivation. Nurses emphasized their role in raising awareness and providing brief, personalized advice rather than comprehensive lifestyle coaching. They also expressed the need for interdisciplinary collaboration and clear referral pathways to lifestyle professionals. Although eHealth was not yet widely used, nurses recognized its potential to complement lifestyle counseling and to strengthen patients’ self-management between visits. The study highlighted the importance of practical guidance, professional training, and organizational support to enhance lifestyle care in secondary diabetes settings.

Chapter 3 described the systematic development of E-Supporter 1.0, a dynamically tailored digital coach designed to promote physical activity and healthy eating among individuals with T2D. The intervention was developed following a structured process including the definition, development, and formative evaluation phases. Using the Health Action Process Approach and theories of behavior maintenance as a theoretical foundation, key behavioral determinants were identified and linked to specific behavior change techniques (BCTs). Intervention options included motivational messages, behavioral feedback, and supportive exercises, which were tailored based on user goals, progress, and barriers. Data for tailoring were collected via app-based monitoring, wearable data, and ecological momentary assessments (EMAs). Usability testing among 9 participants indicated positive experiences and general acceptability, although participants noted the need for greater personalization and flexibility. The study demonstrated how behavioral theory can be systematically translated into dynamic eHealth strategies and identified areas for further refinement.

Chapter 4 assessed the acceptability and preliminary effects of E-Supporter 1.0 in a nine-week feasibility study among 20 individuals with T2D. Participants received daily motivational messages and weekly feedback on their physical activity and diet. Overall, the intervention was perceived as useful and easy to use. Participants appreciated the daily reminders but expressed preferences for combining digital support with face-to-face coaching. The majority found the duration and intensity of the intervention acceptable, though some felt it was too short for lasting behavior change. The exploratory quantitative analyses indicated statistically significant improvements in daily step counts and self-efficacy, as well as improved adherence to the Dutch dietary guidelines. No statistically significant changes were observed in the behavioral change phase, likely because most participants were already in the maintenance stage. The findings support the acceptability of E-Supporter 1.0, while underscoring the need for further tailoring to individual circumstances and long-term support mechanisms.

To better align tailoring strategies with personal and contextual circumstances, Chapter 5 explored the acceptability of a just-in-time adaptive intervention (JITAI) that provided lifestyle support based on daily EMAs. Eight participants used the system for two weeks, receiving tailored SMS coaching messages triggered by their self-reported context, including current activity, location, physical or mental condition, mood, weather, and cravings. Participants valued the contextual relevance of the messages but reported variability in the perceived personalization and intervention intensity. The low-burden EMA approach was considered feasible, though some participants noted that prompts captured only brief “snapshots” and lacked deeper context. While the intervention succeeded in increasing perceived motivation for some, others found the support less relevant. The findings indicate that JITAIs can deliver context-sensitive lifestyle support but also highlight the need for individualized tailoring strategies, improved assessment methods, and improved integration with other behavioral strategies.

Chapter 6 presented a systematic review synthesizing 61 dynamically tailored eHealth interventions to support physical activity and a healthy diet in individuals with chronic diseases, described in 117 papers. The review examined tailoring strategies, theoretical foundations, delivery formats, and reported outcomes. Tailoring variables most frequently targeted physical activity and diet, but many interventions also integrated contextual, emotional, or physiological factors. Most interventions used rule-based approaches, with a smaller proportion employing machine-learning techniques. Behavior change theory was applied in two-thirds of studies, most commonly through the BCTs goal setting, self-monitoring, and feedback. Although within-group improvements were frequently reported, evidence for superiority over control groups was limited due to heterogeneity in study designs and reporting. The review concluded that methodological progress in this field requires greater transparency in reporting, multidisciplinary collaboration, and innovative evaluation designs to disentangle complex, multicomponent interventions.

In Chapter 7, the work in this thesis was discussed in the context of leveraging tailored eHealth within diabetes care. The findings highlight three key implications. First, tailoring should be multidimensional, aligning not only content but also timing, intensity, and delivery to individual needs. Second, eHealth can bridge clinical and daily life, requiring new roles for both patients and professionals, and closer integration across primary, secondary, and social care. Third, sustainable progress calls for greater transparency, interoperability, and reuse of intervention components to build cumulative knowledge.

In conclusion, this thesis provides a foundation for evidence-based and dynamically tailored eHealth strategies to support lifestyle change in the context of T2D care. By integrating behavioral theory, user data, and contextual adaptation, it demonstrates how eHealth interventions can become more tailored to individual needs. The findings lay the groundwork for future research to refine multidimensional tailoring, promote inclusivity across diverse populations, and explore sustainable integration into routine care. Continued progress in this field will determine whether dynamically tailored eHealth can evolve from a promising innovation into a structural component of diabetes care.