Introduction

The global prevalence of diabetes continues to rise, currently affecting an estimated 537 million people worldwide and projected to reach 783 million by 2045 [1]. Alongside increasing prevalence, the health and economic burden of diabetes is substantial and growing. In 2021, an estimated 6.7 million deaths were attributed to diabetes-related causes, with global healthcare costs reaching USD 966 billion [1]. A significant proportion of these costs results from secondary diseases such as depression or anxiety, which are associated with inadequate glycemic control and challenges in diabetes management [2].

Although type 1 diabetes was historically considered a childhood disease, recent evidence indicates a growing number of diagnoses occurring in adulthood [3,4]. Adults with type 1 diabetes therefore represent an increasingly relevant population in both clinical practice and research, and are the focus of the present study.

Technological advances have substantially transformed diabetes management. Automated insulin delivery (AID) systems integrate continuous glucose monitoring (CGM) with continuous subcutaneous insulin infusion (CSII) in a closed-loop architecture that enables algorithm-driven insulin adjustment. In contrast to sensor-augmented pump therapy or other technology-supported regimens without automated insulin modulation, AID systems continuously adapt insulin delivery based on real-time glucose data. By partially automating insulin delivery and glycemic control, they aim to improve metabolic stability while reducing the cognitive and behavioral demands of daily self-management. At the same time, their use introduces technological complexity, including setup and ongoing interaction requirements, which may pose challenges for some users.

The evaluation of diabetes technologies has predominantly focused on glycemic outcomes. HbA1c (reflecting average blood glucose levels over the preceding three months) and Time in Range (TIR, indicating the percentage of time glucose remains within the target range) have become central indicators of metabolic control [5]. An HbA1c value of approximately 5.5–6.5% (i.e., percentage of total hemoglobin that is glycosylated) is generally considered within the optimal physiological range for long-term glycemic control [6,7]. Existing evidence indicates that technological advances in insulin delivery and glucose monitoring – ranging from insulin pens to CSII and CGM – consistently improve glycemic control and treatment adherence compared to conventional injection-based approaches. These findings are supported by multiple randomized controlled trials and meta-analyses, which collectively demonstrate superior HbA1c outcomes among users of these technologies [8–11].

More recent evaluations of AID systems indicate additional improvements in glycemic outcomes beyond those achieved with CGM and CSII alone. Meta-analytic evidence from randomized controlled trials demonstrates consistent reductions in HBA1c and increased TIR among AID users [12]. However, many of these studies were conducted in controlled clinical settings, often excluding participants with comorbidities or suboptimal glycemic control, which may limit the generalizability of the findings. To address this limitation, real-world analyses encompassing large and more diverse populations have been conducted. These studies consistently report clinically meaningful improvements in both HbA1c and TIR across broad patient groups, confirming the effectiveness of AID systems outside of tightly controlled trial environments [13].

While these findings highlight the clinical benefits of AID systems, diabetes management extends beyond metabolic outcomes. Unlike many other chronic conditions, diabetes requires continuous daily self-management, placing a substantial cognitive and emotional burden on those affected [14]. Additionally, the adverse effects of poor glycemic control often develop gradually, which can reduce the perceived urgency of maintaining strict glycemic control. Consequently, for many individuals, the psychological demands of intensive therapy can rival or even outweigh the burden associated with suboptimal glucose control [13].

Accordingly, patient-centered perspectives have gained increasing importance in both research and clinical practice [15]. Beyond improving metabolic outcomes, reducing treatment burden and enhancing quality of life are now recognized as central goals. From a health economics standpoint, this is highly relevant, as the German Diabetes Society’s 2024 Health Report notes that patient time costs (i.e., the monetized value of time spent managing the disease) now exceed direct medical costs [2].

Patient-reported outcomes (PROs) provide a means of capturing these broader impacts. Although studies suggest that technology-supported therapies may improve PROs [12,16], such outcomes are frequently treated as secondary endpoints, and comparability across studies remains limited due to heterogeneous measurement approaches. Moreover, many of these measures focus on technology acceptance or specific concerns like hypoglycemia anxiety, while more comprehensive psychosocial constructs are less frequently examined [16].

One such psychosocial key construct is diabetes distress, defined as the emotional burden associated with the ongoing demands of diabetes self-management [17]. Distinct from depression [14], diabetes distress reflects a core aspect of the lived experience of the disease and is reported by approximately 20–40% of individuals with type 1 diabetes [18,19]. Importantly, higher levels of diabetes distress are associated with poorer glycemic control [20], suggesting potential bidirectional relationships between psychological burden and metabolic control.

Diabetes distress in type 1 diabetes encompasses multiple dimensions, including feelings of powerlessness, management-related strain, hypoglycemia-related concerns and social distress [18,21]. These dimensions provide a differentiated framework for examining how diabetes technologies may influence psychosocial well-being.

Given their automation and potential to reduce cognitive demands, AID systems may mitigate specific aspects of diabetes distress. By automating insulin delivery, they can reduce the burden of day-to-day disease management, minimize dosing errors, and lower the frequency of extreme glycemic events. Qualitative research supports this assumption. Suttiratana et al. [22] reported that AID users experienced lower mental stress, felt more supported in managing their condition, and reported fewer hypoglycemic episodes, especially at night. Similarly, Polonsky et al. [23] observed significant reductions in the corresponding dimensions of the Type 1 Diabetes Distress Scale, that is in Powerlessness, Management Distress, and Hypoglycemia Distress, after three months of AID use. Additionally, a comparative study reported differences in the Powerlessness dimension between AID users and sensor-supported CSII users [24].

Another advantage mentioned by patients in the qualitative study by Suttiratana et al. [22] was the discretion offered by AID systems. This suggests that AID users may also report lower scores in the Negative Social Perception dimension of Diabetes Distress compared to non-users.

However, current evidence on AID systems remains primarily focused on glycemic outcomes, while their impact on diabetes distress – particularly across its multidimensional structure – remains insufficiently understood in adult populations with type 1 diabetes. Existing studies often examine isolated psychosocial outcomes or rely on controlled trial settings, limiting comparability and real-world application.

The present study addresses this gap by comparing adults using AID systems with non-users in terms of both glycemic control and diabetes distress. By jointly assessing a clinical indicator and patient-reported outcomes, the study aims to provide a more comprehensive and patient-centered evaluation of diabetes technology. In doing so, it responds to calls for integrating lived experiences into the assessment of technological interventions in diabetes care [16].

Based on the literature outlined above, the study tested the following hypotheses:

H1: A higher degree of automation is associated with improved glycemic control.

H2: A higher degree of automation is associated with lower overall diabetes distress.

H3: Differences between groups are more pronounced in the dimensions Management Distress, Hypoglycemia Distress, Powerlessness, and Negative Social Perception than in other dimensions of diabetes distress.

To complement the quantitative analyses, qualitative interviews were conducted to gain deeper insight into group differences in diabetes distress. In particular, the study explores whether the automation of glucose management may free cognitive resources that can be redirected toward other aspects of diabetes self-management, reflecting a potential spillover effect.

Method

The presented hypotheses employed a sequential explanatory mixed-method design, in which quantitative data collection and analysis were followed by qualitative interviews to further explore and contextualize the quantitative findings. The data were nested, meaning that the interviewed participants were drawn from the same sample that participated in the quantitative survey. The criteria for selecting interview participants are described in the qualitative methods section.

Results

Quantitative results

Sample characteristics.

By definition, all individuals in the AID group used CSII and CGM systems, whereas in the non-AID group approximately one third (34.8%) used CSII and the majority relied on insulin pens (65.2%). Despite these differences in insulin delivery methods, CGM use was also highly prevalent in the non-AID group (94.4%), with only 5.6% relying on blood glucose test strips. Overall, both groups demonstrated a high level of diabetes technology use.

As shown in Table 1, the AID and non-AID groups were broadly comparable with respect to age, gender, and duration of diabetes (p > .05). However, significant differences emerged in educational characteristics, with AID users having a higher level of education than non-users (p = .008).

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Sample characteristics.

https://doi.org/10.1371/journal.pone.0353125.t001

Descriptive statistics and intercorrelations of the study variables.

As depicted in Table 2, the internal consistencies of the Diabetes Distress total scale and most subscales were good to excellent (Cronbach’s α = 0.78–0.93) and comparable to previous German and English validation studies [18,21]. The subscale Eating Distress showed somewhat lower internal consistency (Cronbach’s α = 0.67) which is likely related to the smaller number of items and their potentially greater heterogeneity compared to the other subscales.

Download:

• PNG
  larger image
• TIFF
  original image

Table 2. Means, standard deviations, internal consistencies, and intercorrelations among study variables.

https://doi.org/10.1371/journal.pone.0353125.t002

Correlation analysis indicated that age was negatively associated with several distress variables and HbA1c, suggesting lower distress and better glycemic control with increasing age. Gender was significantly associated with the subscales Powerlessness and Eating Distress suggesting that women experience more distress than men. In addition, education was negatively correlated with HbA1c, the total distress score, and several subscales, with higher education being associated with better glycemic control and lower distress. Based on these empirical associations as well as prior evidence indicating that sociodemographic factors are related to both diabetes outcomes and the uptake of diabetes technologies [28], age, gender, and education were included as covariates in subsequent analyses to account for potential confounding. To facilitate inclusion in the models and ensure sufficient group sizes, education was dichotomized (university entrance qualification or lower, n = 90, vs. university degree or higher, n = 79).

Group differences: AID vs. non-AID group.

Results of the ANCOVA analyses are summarized in Table 3. With regard to glycemic control, the AID users showed significantly lower HbA1c values (M = 6.68, SD = 0.74) compared to non-users (M = 7.21, SD = 1.38), indicating better glycemic control (F(164) = 5.81, p = .017, partial η² = .034).

Download:

• PNG
  larger image
• TIFF
  original image

Table 3. ANCOVA results for glycemic control and diabetes distress (adjusted for age, gender, and education).

https://doi.org/10.1371/journal.pone.0353125.t003

AID users also reported significantly lower overall diabetes distress (M = 2.42, SD = 0.78) than non-users (M = 2.99, SD = 0.92; F(164) = 14.40, p < .001, partial η² = .081). Across subscales of the T1-DSS, a consistent pattern emerged, with AID users reporting significant lower distress levels in all domains. The largest effects were found for Management Distress (η² = .072) and Powerlessness (η² = .064), while the smallest group differences were observed for the dimensions Hypoglycemia Distress (η² = .024), Eating Distress (η² = .027) and Negative Social Perception (η² = .033). To control for multiple testing, a Benjamini-Hochberg correction (FDR) was applied. All group differences remained significant after correction.

Qualitative part

Interview approach.

Semi-structured interviews were conducted using an interview guide that was developed based on the quantitative findings, with the aim of further exploring and explaining observed group differences in diabetes distress. Questions were formulated in an open and group-independent manner to avoid leading participants toward predefined assumptions. Rather than explicitly referring to AID systems, the interview focused more broadly on the role of the diabetes-related technologies in daily management.

Following initial questions on technology use and its general impact on their diabetes management, the interview was structured around the dimensions of diabetes distress that had shown relevant patterns in the quantitative analyses. For each dimension, open-ended questions were asked and followed by prompts for clarification or elaboration where necessary. Examples of such open questions included “How would you describe your relationship with your doctor?,” “Do you feel that your friends and family make a bigger issue out of your condition than you would like?,” and “How do you feel when you think about hypoglycemia?.”

If diabetes technologies were not mentioned spontaneously, participants were specifically asked whether and how the respective experience had changed due to their use. For selected dimensions (e.g., Negative Social Perception and Management Distress), additional follow-up questions explored perceived changes over time and the role of age. The order of topics was flexible and adapted to the flow of the conversation.

Integration of quantitative and qualitative findings.

Table 4 presents a joint display integrating quantitative and qualitative findings across diabetes distress dimensions. Overall, the interviews provided contextual explanations for the observed quantitative group differences in diabetes distress and glycemic control, thereby enabling integrated interpretation across methods.

Download:

• PNG
  larger image
• TIFF
  original image

Table 4. Joint display integrating quantitative and qualitative findings on AID use and non-use across diabetes distress dimensions.

https://doi.org/10.1371/journal.pone.0353125.t004

Across dimensions, participants’ narratives reflected the quantitative patterns observed between AID users and non-users. Lower Powerlessness in the AID group was reflected in descriptions of automated correction of glucose fluctuations and reduced persistence of dysglycemic episodes. Reduced Management Distress was mirrored in accounts of increased flexibility and a lower cognitive load in daily diabetes management. With regard to Hypoglycemia Distress, participants in both groups described a continuous glucose monitoring as an important source of safety, particularly due to alarm functions and trend information. AID users additionally emphasized the benefit of automated insulin adjustments in preventing hypoglycemic events. Lower Eating Distress among AID users was reflected in reports of reduced need for precise carbohydrate calculations and greater flexibility in food intake. In contrast, non-AID users described a higher cognitive burden related to continuous planning and calculation. For Negative Social Perception, participants across both groups reported that modern diabetes technologies can reduce the visibility of the condition in everyday life, particularly through the use of CGM and insulin pumps. However, AID systems were not specifically identified as contributing additional advantages in this regard. Family and Friends Distress was reflected in participants’ perceptions that improved glycemic management reduced the perceived burden on close others and decreased their involvement in daily disease management. Finally, Physician Distress, was characterized by heterogeneous experiences, ranging from low perceived relevance of routine consultations to positive evaluations of supportive healthcare providers.

Additional emerging themes.

Beyond the predefined distress dimensions, several additional themes emerged inductively during the interviews. When a topic was mentioned multiple times and considered interesting, it was explored in the subsequent interviews.

One recurring topic was the perceived dependence on diabetes technologies. While most participants did not describe their use in terms of dependency, many emphasized the substantial benefits and expressed that they would not want to return to previous treatment methods. For example, interviewee nine (AID user) said: “So the system itself gives me so much added value in every respect physiologically, mentally, in terms of quality of life and so on, that I would find it very, very, very, very difficult to do without it, whether I can describe it as an addiction, I don’t know. I can certainly cope with pen therapy again from one day to the next, but it would just make a lot of things more complicated, more complex, more difficult.”

Another theme related to the reduced visibility of diabetes due to increasing technologization. While this was often experienced as relieving, some participants reported that their condition received less attention or understanding from others. For example, when asked “Do you feel that your friends and family make a bigger deal of your illness than you would like?” several participants responded that they actually felt the opposite. Respondent five, for instance, said: “No, not really, on the contrary, they don’t really realize it and I often think a bit of support would be appropriate..” In this context, participants expressed a desire for greater recognition of the ongoing effort required to manage the disease even when using AID systems.

Participants also reported challenges associated with the technologies, including technical issues (e.g., signal loss, occlusion) and physical reactions such as skin irritations. In case of interviewee 12 (AID user), dissatisfaction with a specific system was described, including perceived negative effects on glycemic control, leading to changes in devices. Interviewee four (non-AID user) reported frequent occlusions and supplemented pump therapy with long-acting insulin as an additional safety measure.

Finally, continuous glucose monitoring was described as highly beneficial across both groups. However, some participants reported that the constant availability of glucose data initially led to increased stress and over-monitoring, which decreased over time as they adopted to the technology. For example, interviewee six (AID user) explained: “I also had to learn how to deal with the fact that you constantly saw the value and also constantly saw the progression, which you didn’t see otherwise. [...] that stressed me out a bit at the beginning and then I also tended to overreact at the start when it shot up. [...] when I learnt to stay a bit more relaxed, I would say the sensor was the biggest relief.”

Discussion

The increasing technologization of diabetes treatment over the past 25 years offers opportunities to reduce both the disease burden for those affected and the rising economic costs of treatment. AID systems represent a recent development in diabetes technologies, enabling automated insulin delivery by algorithmically linking CSII and CGM systems. The aim of this study was not only to examine glycemic control outcomes associated with AID use, but also to focus on psychosocial factors that have been largely neglected in previous research. Therefore, the main emphasis was placed on diabetes distress and its dimensions. The present study is one of the few to examine differences between AID users and non-users across the seven dimensions of diabetes distress. Furthermore, it is among the first to apply a mixed methods approach to explore and contextualize these differences, combining quantitative data with follow-up interviews conducted within the same sample. This approach aimed to provide a deeper understanding and interpretation of observed differences in distress dimensions. Importantly, given the high prevalence of CGM use (~94%) in the non-AID group, the comparison in the present study should be interpreted as AID with automated insulin delivery versus CGM-based diabetes management without automation, rather than technology versus no technology use.

In the present study, lower HbA1c levels were observed in the AID group compared to the non-AID group. This finding is consistent with previous research. The between-group difference of approximately 0.5 percentage points exceeds the average difference of 0.2 percentage points reported in the meta-analysis by Godoi et al. [12] and falls within the range of the 0.1 to 0.9 described in a review by Considine and Sherr [13]. A difference of this magnitude is generally considered clinically meaningful, as each 1% increase in HbA1c is associated with a substantially higher risk of micro- and macrovascular complications [29]. This is particularly relevant given the commonly recommended long-term HbA1c target of 7%, as small differences around this threshold may determine whether patients meet or exceed recommended glycemic targets [30]. In the present sample, mean HbA1c in the AID group was below this threshold, whereas it was above 7% in the non-AID group. Therefore, the observed difference of 0.5% may be considered clinically relevant in our population. However, these differences should be interpreted as associations rather than causal effects given the cross-sectional design. It is also possible that self-selection into AID use contributed to these differences, for example through differences in baseline glycemic control, self-management capacity or healthcare access.

Moreover, lower overall diabetes distress was observed in AID users compared to non-users. This pattern is consistent with previous meta-analyses, reviews, and cohort studies [12,13]. The effect size found in the present study (Cohen’s d ≈ 0.6) is larger than effects reported in previous studies [12]. One possible explanation is that clinical studies often recruit subjects with already lower baseline distress levels. Additionally, the outpatient context provides subjects with strong support, which might not be available in other settings, potentially leading to smaller effects than those observed in real-world contexts [13].

As only one comparative study [24] has examined diabetes distress subdimensions, our findings are also discussed in relation to studies on related constructs, such as Hypoglycemia Anxiety [31], as well as findings from cohort studies [23] and qualitative studies on diabetes distress [22]. Consistent with the findings of Kudva et al. [24], lower values in the Powerlessness dimension were observed in the AID group compared to the non-AID group. In contrast to this previous study reporting effects limited to this dimension, our study observed significant differences between AID users and non-users across all sub-dimensions, with the effect size being one of the largest for the Powerlessness dimension. The qualitative part of our study suggests that AID systems may reduce exposure to extreme glycemic fluctuations, which participants linked to reduced feelings of helplessness. In the study by Suttiratana et al. [22], one participant also described diabetes management as a demoralizing daily task, but noted that the use of the system made it less demoralizing for the first time since childhood.

The largest group difference was observed for the Management Distress dimension, with AID users showing a medium-sized advantage. Qualitative data from previous research [22] and the present study suggest that automated insulin delivery may be associated with a reduced cognitive and behavioral burden of diabetes management. For example, AID users described a greater sense of freedom and flexibility in terms of physical activity, which the system enables. This sense of freedom and flexibility, often unfulfilled in daily life, was expressed as a major desire by participants in the non-AID group. A further negative correlation was observed between participants’ age and their distress levels in this diabetes dimension. In the qualitative part of the study, acceptance of one’s illness and age were often jointly discussed. A previous longitudinal study identified acceptance as a significant predictor of reduced diabetes distress over time [32]. Therefore, the observed association may reflect age-related changes in acceptance, although alternative explanations such as disease duration and related adaptation processes cannot be excluded. Importantly, adjusting for age in the ANCOVA models did not materially change group differences.

Moreover, lower Hypoglycemia Distress was observed in the AID group, which is consistent with four of the five studies included in the meta-analysis by Godoi et al. [12]. However, the effect size was smaller than expected. One possible explanation is the widespread use of CGM systems in the non-AID group, which participants described as helpful in managing hypoglycemia. Previous studies have shown that CGM systems are associated with reductions in hypoglycemia anxiety and a lower frequency of hypoglycemic events compared to glucose measurement with test strips [16]. Thus, differences between groups may be attenuated by high baseline levels of technological support in both groups. Nevertheless, AID systems may differ from CGM alone in their ability to actively prevent hypoglycemia events, which may explain the small to moderate group differences and which is illustrated in the following quote: “Well, there were really three worlds in which I lived. I would say that the time before using a CGM system was often characterized by the fact that my wife really had to look after me at night with orange juice and cola to get me out of a hypo, so to speak, or in some cases even ambulance support was necessary. That was super stressful. That’s super stressful. It’s a really stressful situation for a partner or family. The use of a CGM system had reduced this very, very significantly; even when there was the rare case of hypos, there was an alarm. It’s totally annoying, but of course it requires less action than if someone develops massive hypoglycemia and since now, I’ve had the AID system since 2016, there are really only alarm situations in situations where technical failures occur, such as battery failure or sensor errors. Otherwise, the nights go by so that you can really sleep through the night and wake up relaxed the next morning. It really makes such a massive difference; you can’t even imagine it. It’s really incredibly valuable and enriching and I’m very, very grateful.”

As in the other dimensions, significant differences between AID users and non-users were found in the Negative Social Perception sub-dimension, although the effect size was small. This contrasts with the cohort study by Polonsky et al. [23], in which no significant differences were found between baseline and follow-up. Qualitative findings by Suttiratana et al. [22], however, suggest that discretion in everyday life may represent an advantage of AID systems. Participants reported feeling more comfortable managing their diabetes via mobile phone rather than using more visible traditional devices, such as glucose meters or insulin pumps. In the qualitative part of the present study, participants similarly described greater discretion when using diabetes technologies, such as CSII and CGM systems. However, AID systems themselves were not specifically identified as providing additional advantages in this regard. Furthermore, quantitative analyses indicated an effect of age on this dimension. Consistent with this, participants in the qualitative interviews reported experiencing less social stress with increasing age. For example, diabetes devices were worn more openly and less hidden as participants grew older. This finding aligns with the qualitative study by Haynes et al. [33], in which children reported wearing their insulin pumps with pride rather than hiding them as they got older.

Eating distress was also lower in the AID group than in the non-AID group, although the effect size was small. This is partly consistent with the cohort study of Polonsky et al. [23], in which eating distress showed one of the greatest reductions after three months of follow-up. The interview data suggest that perceived reliance on the automated glucose regulation may reduce food-related cognitive burden and concerns about miscalculations. The small effect, observed in the present study should, however, be interpreted with some caution, as the Eating Distress subscale showed slightly lower internal consistency compared to the other subscales, although it remained within an acceptable range for short scales.

Moreover, lower Physician Distress was observed in the AID group. This effect was small to moderate and is consistent with the small improvement observed on this dimension after a three-month follow-up in the cohort study by Polonsky et al. [23]. In the interviews conducted, two types of interactions between AID users and their respective practitioners were identified. The first type of interaction was not considered valuable, but also not stressful, as the interviewees stated that they were able to manage their disease effectively with the support of the system. In the second type of interaction, some participants reported trust and gratitude towards healthcare providers who had encouraged them to adopt the new technologies. This may indicate that the observed association is influenced by relational factors such as perceived support or satisfaction with care, rather than technology use per se. These characteristics have been identified in previous studies as predictors of a good practitioner-patient relationship [34].

In the present study, values on the distress dimension Family and Friends Distress were lower in the AID group than in the non-AID group, whereas no differences were reported by Polonsky et al. [23]. In the interviews, AID users described their social environment as less burdened, referring to fewer extreme hypo- and hyperglycemic events and reduced involvement of others in diabetes management. They further noted that acute glycemic fluctuations were less visible to those around them. This may reflect a reduced salience of the illness in everyday interactions, which in turn may be associated with lower distress, particularly when individuals no longer perceive themselves as a source of stress for others. Tsiouli et al. [35] reported a bidirectional association between family conflicts and suboptimal glycemic control in pediatric type 1 diabetes. In contrast, participants in the AID group did not link this dimension of distress to glycemic control in the qualitative data. This suggests that the current conceptualization of this sub-dimension may not fully capture relevant sources of distress in AID users, warranting further investigation. The concept of shared illness appraisal (i.e., understanding the illness as a shared rather than individual responsibility) may provide a useful framework in this context, as it has been associated with improved well-being for both partners [36]. The expressed desire for greater attention and recognition from family members, which emerged as a pattern in the interviews, aligns with this perspective. Although AID systems may reduce the visibility of diabetes, they do not eliminate its presence.

The obstacles and challenges related to diabetes technologies identified in the qualitative section are consistent with previously reported limitations of CSII and CGM systems. Common CSII-related issues include occlusions and skin irritation [10,37], whereas CGM systems are frequently associated with warm-up phases, signal loss, and occasional and inaccurate or missing glucose values [38]. In the present study, however, such technology-related challenges were not consistently attributable to AID use status or specific device components, but rather reflected general experiences with diabetes technologies across participants. These issues highlight the need for ongoing improvements in diabetes technologies. Addressing such limitations remains an important consideration for manufacturers to support more reliable and less burdensome use. In addition, improving accessibility and reducing barriers to use may be relevant for broader uptake. Improvements in these areas may contribute to enhancing the potential benefits of these technologies for individuals and society.