{
  "abstract": "Introduction Diabetes technologies may improve glycemic control and psychological well-being among adolescents and young adults (AYA) with type 1 diabetes. This cross-sectional study examines perceptions of automated insulin dosing (AID) systems and their association with glycemic and psychological outcomes compared with multiple daily insulin injections (MDI) and continuous subcutaneous insulin infusion (CSII).Research design and methods Participants were recruited from the largest diabetes camp for AYA in Germany. A total of 151 participants (70% female, mean age 20.7±2.9 years, 33% AID users) completed a questionnaire that included self-reported glycated hemoglobin A1c (HbA1c), global health status, emotional well-being (WHO-5), Generalized anxiety (GAD-7) and diabetes distress (PAID-5). AID users also rated their experiences with the system.Results AID users reported significantly lower HbA1c levels (7.3±1.0%) than CSII users (7.5±1.1%) and MDI users (8.4±2.0%, p=0.003). Approximately 75% of AID users viewed their current system as an improvement over previous therapy, reporting greater ease (84%), comfort (82%) and safety (80%). They reported higher treatment satisfaction than CSII users (p=0.044) and lower diabetes burden than MDI users (p=0.044) after controlling for age and gender. Treatment groups did not differ in well-being or anxiety. Better global health status was associated with the absence of other chronic health conditions (p=0.024), greater well-being (WHO-5; p<0.001), lower HbA1c (p=0.038) and fewer anxiety symptoms (GAD-7, p=0.007). Despite these positive indicators, a substantial proportion of participants reported symptoms of depression (18%), anxiety (30%), and diabetes distress (39%).Conclusions AID systems were associated with improved glycemic control and high satisfaction among AYA. However, psychological distress remained prevalent across all treatment modalities, underscoring a discrepancy between metabolic benefits and persistent mental health challenges. These findings highlight the need to integrate psychological support alongside technological advances in the care of AYA living with diabetes.",
  "authors": [
    {
      "affiliations": [
        "Medical Psychology, Hannover Medical School, Hannover, Germany"
      ],
      "name": "Gundula Ernst"
    },
    {
      "affiliations": [
        "Medical Psychology, Hannover Medical School, Hannover, Germany"
      ],
      "name": "Su-Jong Kim-Dorner"
    },
    {
      "affiliations": [
        "Medical Psychology, Hannover Medical School, Hannover, Germany"
      ],
      "name": "Madelaine Hampel"
    },
    {
      "affiliations": [
        "Medical Psychology, Hannover Medical School, Hannover, Germany"
      ],
      "name": "Henrike Fritsch"
    },
    {
      "affiliations": [
        "Medical Psychology, Hannover Medical School, Hannover, Germany"
      ],
      "name": "Karin Lange"
    }
  ],
  "full_text": "WHAT IS ALREADY KNOWN ON THIS TOPIC The benefits of advanced diabetes technologies, such as automated insulin dosing (AID) systems, on clinical outcomes are well documented; however, their broader impact on mental health and user experience remains unclear, particularly among adolescents and young adults (AYA).WHAT THIS STUDY ADDS AID users reported high satisfaction with their systems and significantly better glycemic control compared with those using continuous subcutaneous insulin infusion or multiple daily insulin injections. Despite these advantages, elevated levels of psychological distress, including anxiety, depression and diabetes distress, were observed across all treatment groups. Self-reported health status was closely associated with both glycemic control and psychological well-being.HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY This study demonstrated that while advanced diabetes technologies improve glucose control, their full potential for enhancing overall health in AYA will require integrated targeted psychosocial support.Introduction Type 1 diabetes (T1D) is the most common metabolic disease in children and adolescents. It requires daily, challenging self-management to achieve the goal of near-normal blood glucose levels. Since the introduction of intensified insulin therapy with multiple daily glucose measurements and corresponding insulin injections (multiple daily insulin injections; MDI), several technologies have been developed. To facilitate disease management, many young people use continuous subcutaneous insulin infusion (CSII) pumps, continuous glucose monitoring (CGM) sensors and, more recently, automated insulin delivery (AID) systems. AID systems comprise a CGM system, an algorithm that calculates the required insulin dose based on the measured glucose values, and an insulin pump that delivers this dose. These three components work together to automatically regulate insulin needs. Manual intervention is still required for meal boluses and adjustments for special events such as exercise in most systems. However, the automation of therapy has the potential to enhance both glycemic control and quality of life (QoL).Systematic reviews comparing different treatment modalities demonstrated that, on average, the use of these technologies resulted in lower glycated hemoglobin A1c (HbA1c) levels, more time in the target range (TiR: 70–180 mg/dL) and less time in hypoglycemia compared with therapy with MDI.1 2 Studies with longer follow-up have shown that AID systems are not only successful in improving glycemic control but also safe for daily use.3While these clinical benefits are well documented, it is equally important to consider the patient’s lived experience using person-reported outcomes (PROs) such as treatment satisfaction, QoL, disease-related distress, anxiety and fear of hypoglycemia. The study of PROs is becoming increasingly important in diabetes research as they have a significant impact on adherence, disease progression and emotional well-being.4 5 However, studies of the PROs of new diabetes technologies show mixed results.1–3Adolescents and young adults (AYA) are a particular group that merits close attention when considering the potential benefits of AID use. They represent a challenging population as it is more difficult to achieve treatment goals in this phase of life. Reduced impulse control, new interests, unsteady life rhythms and hormonal changes have been described to make diabetes management more challenging. These developmental factors frequently lead to suboptimal metabolic control in AYA.6In addition to glycemic challenges, psychological distress is also high at this stage of life. About 30% of AYA show high levels of depression, anxiety and diabetes distress.7 8 Whether diabetes technologies can help reduce these psychosocial burdens is unclear. On one hand, automation makes diabetes management easier, and this may reduce emotional distress. On the other hand, new challenges may arise due to the complexity of the technology such as overwhelming feeling due to technical issues, excessive demands (eg, calibrations) and a high number of alerts.9 10 In fact, a recent meta-analysis has shown that the use of AID has a beneficial effect on TiR and metabolic decompensation,11 but the effects on PROs remain largely unknown in most of the included studies.This study, therefore, explores the views of AYA with T1D on the use of current diabetes technologies in everyday life. The aim of this study is to assess (1) the association of diabetes technologies used (AID, CSII and MDI) with metabolic control and PROs and (2) how young people rate the use of AID systems compared with their previous diabetes therapy.Material and methods The cross-sectional study was conducted at a German diabetes camp for young people with diabetes aged 16–25 years. The camp was organized by a pharmaceutical company and supported by the German Diabetes Association (Deutsche Diabetes Gesellschaft, DDG) and volunteer diabetes experts. During the 4-day stay in 2022, the AYA could participate in workshops on current diabetes topics, try out sports and leisure activities and meet other AYA with diabetes and healthcare professionals. The camp aimed to encourage young people to live empowered lives with their disease. The camp was promoted through diabetes centers, support groups and social media. Participation was voluntary.Participants and procedure Prior to the camp, all 239 participants were invited to take part in the anonymous survey. To be included in the study, participants had to have T1D, be between 16 and 25 years old and be proficient in written German. They received a flyer with information about the aim and general conditions of the study. Informed consent for participation in the study and data processing was obtained from all participants. According to German law, no additional consent was required for individuals under the age of 18. Participants accessed the online questionnaire via a QR code or a web link. The study was approved by the Ethics Committee of Hannover Medical School (No. 10373_BO_K_2022) and complies with the Declaration of Helsinki.Questionnaire Sociodemographic and clinical data The questionnaire collected demographic information (age, sex, education, parental country of birth) and clinical data (duration of diabetes, treatment modalities, most recent HbA1c value and TiR from CGM data over the past 14 days). PROs included a single-item self-rated global health status question, a single-item treatment satisfaction measure and psychological health assessments (WHO-5, Generalized Anxiety Disorder (GAD)-7, Problem Areas in Diabetes (PAID)-5).Global health status Participants rated their current overall health using a single-item question on a 5-point Likert scale, ranging from ‘very good’ to ‘very bad’.Treatment satisfaction All participants answered a single-item treatment satisfaction question using a 7-point Likert scale ranging from ‘extremely unsatisfied’ to ‘extremely satisfied’, evaluating their satisfaction with their current therapy.Satisfaction with the AID system AID users rated their daily use of the AID compared with their previous treatment, using items specifically developed and validated for the evaluation of AID systems. 12 Eight items were used to assess various aspects on a 5-point Likert scale including satisfaction, safety and comfort. See the note section of figure 1 for the complete list of questions.Figure 1Evaluation of AID use in everyday life among 49 AID users. *Evaluation of AID system questions: (1) How are you doing with the AID system in your daily life compared with your previous treatment? (2) How satisfied are you with using the AID system? (3) Do you feel safe with your AID system? (4) How comfortable is the AID system for you in your daily life? (5) How has your sleep been since you started using an AID system compared with before? (6) How difficult is it for you to operate your AID system? (7) How much do you trust your AID system? (8) Has your confidence in your own diabetes management changed since you started using an AID system? †Response categories have been linguistically adapted. AID, automated insulin delivery.Well-being Psychological well-being was assessed using the WHO-5 questionnaire. 13 On a 6-point Likert scale (0=at no time to 5=all of the time), five positive statements are answered (eg, ‘Over the last 2 weeks I have felt calm and relaxed’). The summed raw score is generated ranging from 0 (no well-being) to 25 (maximum well-being), and then multiplied by 4 to establish a percentage score. A percentage score below 50 indicates poor mental well-being and may indicate depression, while a score ≤28 means that major depression is very likely.14Anxiety The German version of the GAD-7 15 was used to measure generalized anxiety. It consists of 7 items assessing how often respondents experienced symptoms of anxiety in the past 2 weeks (eg, ‘Feeling nervous, anxious or on edge’). It is scored on a 4-point Likert scale (0=not at all to 3=nearly every day). The total score can range from 0 to 21, with anxiety categorized as ‘no or low’ (total score: 0–4), ‘mild’ (total score: 5–9), ‘moderate’ (total score: 10–14) and ‘severe’ (total score: 15–21).Burden of diabetes Diabetes distress was assessed using the 5-item PAID-5 questionnaire. 16 A 5-point Likert scale (0=not a problem to 4=serious problem) is used to answer questions about the current burden of the disease in everyday life (eg, ‘Worrying about the future and the possibility of serious complications?’). A total score between 0 and 20 is used for scoring. A cut-off score of ≥8 indicates increased diabetes-related distress.Statistical analyses Descriptive statistics are presented as frequencies, means (M) and SD. The χ 2 tests were used to examine group difference for categorical variables. One-way analysis of variance (ANOVA) was used to examine differences in mean HbA1c levels with a priori contrasts comparing the AID group separately to the MDI and CSII groups.Group differences in PROs were analyzed in a multivariate analysis of variance (MANOVA) with the same a priori contrasts. Analysis of covariance (ANCOVA) and MANCOVA were employed in addition, to examine group differences while controlling for age, gender, diabetes duration and the presence of other chronic medical conditions. Bonferroni tests were used to follow-up on significant omnibus ANOVA results, specifically to explore differences between the MDI and CSII groups.All parameters were assessed for normality, and variables that could not be transformed to meet the assumptions of parametric tests were analyzed using non-parametric tests. HbA1c data were transformed using an inverse transformation (1/HbA1c), however, for clarity, untransformed data are presented in the table in their original units. Correlations among continuous variables were examined by using Pearson’s correlation coefficients, categorizing r=1 to 0.29 as small, 0.3 to 0.49 as moderate, and ≥0.5 as large.17 A significance level of α=0.05 was used for analyses. All data analyses were performed using the IBM SPSS V.27.0 (SPSS, Chicago, Illinois).This manuscript adheres to the Strengthening the Reporting of Observational Studies in Epidemiology checklist.Results Sample Of the 239 camp participants, 200 accessed the digital questionnaire (84% response rate) and data from 178 were available. Of these participants, two were excluded due to having a different type of diabetes other than T1D and further 25 due to being older than 25 years. The final sample consisted of N=151 AYA. The excluded 27 participants did not significantly differ from those included in the study in terms of gender and AID use. Table 1 presents characteristics of the entire sample and by type of diabetes therapy. The mean age of the respondents was 20.7±2.9 years and 70.2% were female. Four individuals identified as a diverse gender. Their data were included in all analyses except for gender-based analyses due to the small sample size. Fewer people had a migration background compared with German population (17.2% vs 28.7%) (https://de.statista.com).Table 1Sample characteristics by type of diabetes technologyDiabetes technologyNMean±SD or n (%)P valueAllN=151100%MDIn=4227.8 %CSIIn=5939.1 %AIDn=5033.1 %Gender, female†151106 (70.2%)‡27 (64.3%)42 (71.2%)37 (74.0%)0.563Age, years15120.7±3.020.1±2.820.6±3.121.4±2.90.083High school degree, yes§15195 (62.9%)26 (62.0%)35 (59.3%)34 (68.0%)0.638Parental migration background, yes15126 (17.2%)10 (23.8%)9 (15.3%)7 (14.0%)0.406Diabetes duration, years1519.9±5.37.5±6.310.5±4.1*11.2±4.9**0.002**Psychological treatment last year, yes14938 (25.5%)10 (23.8%)16 (27.6%)12 (24.5%)0.895Other chronic conditions, yes15049 (32%)10 (23.85)19 (32.8%)20 (40.0%)0.257*p < 0.05 and **p < 0.01 compared to MDI according to Bonferroni test.†Gender analysis excludes diverse gender of n = 4 (MDI = 1, CSID = 2, AID = 1).‡All percentages are calculated excluding missing values.§For minor participants currently in school (n = 31), the expected degree at the end of schooling is reported. χ2 test was used for categorical variables and ANOVA for continuous variables unless otherwise noted.AID, automated insulin dosing; ANOVA, analysis of variance; CSII, continuous subcutaneous insulin infusion; MDI, multiple daily injections.A total of 33.1% used an AID system (see online supplemental document). In our sample, both CSII and AID users differed significantly from MDI users for having a longer duration of diabetes. The proportion of female, age and immigration status did not significantly differ among treatment modalities.SP110.1136/bmjdrc-2025-005243.supp1Supplementary dataGlycemic control The mean self-reported HbA1c for the entire sample was 7.6±1.4% (see table 2). The treatment groups differed on HbA1c levels (F(2123)=6.04, p=0.003). Planned contrasts revealed that HbA1c was significantly lower in the AID group compared with the MDI group (mean difference=−1.14%, 95% CI (−1.76 to −0.52), t(123)=3.37, p=0.001). The CSII group also had a lower mean HbA1c compared with the MDI group according to the Bonferroni test (p=0.025). There was no significant difference in mean HbA1c levels between the AID and CSII groups (mean difference=−0.22%, 95% CI (−0.77 to 0.33), t(123)=0.90, p=0.37). Using ANCOVA, after controlling for age, gender, disease duration and the presence of other chronic conditions, the difference in HbA1c levels among technology groups remained significant (F(2114)=8.80, p<0.001). Specifically, AID users continued to show lower HbA1c levels compared with MDI users even after adjusting for the covariates (adjusted mean difference=−1.37%, 95% CI (−2.04 to −0.73), t(123)=−4.01, p<0.001). Similarly, the treatment groups differed in TiR (F(2131)=4.33, p=0.015). Planned contrasts showed that AID users had a significantly higher mean TiR compared with CSII users (10.81%, 95% CI (3.49 to 18.13), t(131)=2.92, p=0.004).Table 2Self-reported glucose values and person-reported outcomes by type of diabetes technologyDiabetes technologyNM±SD or n (%)Omnibus test pAllMDICSIIAIDHbA1c, %1267.6±1.48.4±2.0†**7.5 ± 1.1‡*7.3±1.00.003 ≤ 7.5% HbA1c, yes74 (58.7%)13 (40.6%)30 (57.7%)31 (73.8%)TiR, %13459.2±18.860.2±19.253.9±19.9†**64.7±15.60.015 ≥ 70% TiR, yes46 (34.3%)11 (31.4%)13 (24.5%)22 (47.8%)Treatment satisfaction (range 0–6)1464.05±1.394.05±1.363.81±1.54†*4.36±1.190.131Global health status (range 1–5)1513.83±0.743.98±0.683.71±0.773.86±0.760.204WHO-5 (range 0–100)14552.39±20.7051.90±19.6951.65±21.2153.70±21.310.966 High risk of depression (≤ 28)26 (17.9%)7 (17.1%)10 (17.5%)9 (19.1%)GAD-7 (range 0–21)1457.7±5.27.4±4.58.0±5.77.5±5.20.870 High risk of anxiety (≥ 10)44 (30.3%)11 (26.9%)19 (33.3%)14 (29.8%)PAID-5 (range 0–25)1446.7±4.77.4±4.76.5±4.96.1±4.40.423 High diabetes burden (≥ 8)56 (38.9%)19 (46.3%)22 (38.6%)15 (32.6%)*p < 0.05 and **p < 0.01.†Difference significant according to planned contrasts compared to AID.‡Difference significant compared to MDI users according to post-hoc Bonferroni test.AID, automated insulin delivery; CSII, continuous subcutaneous insulin infusion; GAD-7, Generalized Anxiety Disorder Scale; HbA1c, glycated hemoglobin A1c; MDI, multiple daily injections; PAID-5, Problem Areas in Diabetes Scale; TiR, time in range; WHO-5, WHO Wellbeing Index.A significant gender difference was found in glucose outcomes. A higher percentage of females achieved the recommended HbA1c cut-off compared with males (63.7% vs 41.9%, χ2 (1, N=122) = 4.51, p=0.034). A similar trend was observed for TiR with 39.8% of females and 21.6% of males reaching the target (χ2 (1, N=130)=3.86, p=0.049).Person-reported outcomes Table 2 shows the mean PRO scores by diabetes technology used. Global health status did not differ by insulin technology. Overall, 72.8% of participants reported their global health as either ‘good’ or ‘very good’ with no significant differences across technology groups (MDI=32 (76.2%); CSII=42 (71.2%); and AID=36 (72%), χ2 (2, N=151) = 0.34, p=0.845).Treatment satisfaction among treatment modalities did not differ, but planned contrasts showed that the overall treatment satisfaction was higher among AID users compared with CSII users (mean difference=0.55, 95% CI (0.01 to 1.09), t(143)=2.03, p=0.044).Well-being Overall, 44.1% of the sample had a cut-off score of <50 indicating poor psychological well-being and 17.9% were below the cut-off for clinical depression (≤28). The proportion of people with a score ≤28 was similar in the three technology user groups. There was no significant difference between gender groups.Anxiety The mean anxiety score for the entire sample was 7.7 (±5.2) out of 21. The level of anxiety did not vary by technology use. Overall, 17.2% of the participants were classified as moderately anxious and another 13.1% were classified as severely anxious. No significant gender difference was noted.Diabetes distress The average diabetes-related burden score was 6.7 (±4.7) out of 20. Of all respondents, 38.9% had a sum score ≥8, indicating an increased diabetes-related burden. The MDI group had the highest proportion of those with an increased burden with 46.3%, followed by the CSII group with 38.6% and the AID group with 32.6%. This is also reflected in the mean values of the groups although the difference was not statistically significant. In general, females reported a higher diabetes burden than males (7.3 (±4.7) vs 4.9 (± 4.2), respectively, t(138) = 2.83, p=0.005).Overall, MANOVA revealed no significant differences in psychological PROs across the different forms of insulin therapy (Pillai’s Trace=0.023, F(6280)=0.54, p=0.775), and no a priori contrasts were significant. MANCOVA was also not significant even after adjusting for age, gender, disease duration and presence of other chronic conditions (Pillai’s Trace=0.033, F(6262)=0.74, p=0.616). However, a priori contrasts within the MANCOVA framework indicated that AID users reported a small but significantly lower diabetes burden compared with MDI users (adjusted mean difference=−2.10, 95% CI (−4.14, −0.054), p=0.044). The covariates, age (Pillai’s Trace=0.079, F(3130)=3.72, p=0.013) and gender (Pillai’s Trace=0.078, F(3130)=3.72, p=0.014) had significant multivariate effects on the combined PROs. Parameter estimates indicated that increasing age was associated with greater diabetes burden (B=0.43 (SE=0.14), 95% CI (0.16 to 0.70), t=3.10, p=0.002) and male AYA reported significantly lower burden (B=−2.81 (SE=0.85), 95% CI (−4.50 to –1.13), t=−3.30, p=0.001).Relationship between HbA1c and PROs Table 3 shows that the PROs correlated moderately to strongly with each other. There was no significant correlation between the psychological PROs and HbA1c. However, treatment satisfaction and global health status showed a small inverse correlation with HbA1c.Table 3Pearson’s correlation of HbA1c with demographic variables and person-reported outcomesVariable1†2345671. HbA1c a--2. Age−0.088--3. Diabetes duration0.1010.299***--4. Treatment satisfaction−0.253**−0.055−0.158--5. WHO-5−0.084−0.070−0.0690.393***--6. GAD-70.0750.0510.149−0.370***−0.712***--7. PAID-50.1110.214*−0.029−0.404***−0.430***0.512***--8. Global health rating−0.208*−0.050−0.0820.333***0.530***−0.498***−0.392****p < 0.05, **p < 0.01 and ***p < 0.001.†Correlation was run with inverse HbA1c but the signs were reversed for clarity.GAD-7, Generalized Anxiety Disorder Scale; HbA1c, glycated hemoglobin A1c; PAID-5, Problem Areas in Diabetes Scale; WHO-5, WHO Wellbeing Index.Exploratory regression Predictors of HbA1c were explored using a backward linear regression model that incorporated variables identified in prior research. These variables included age, gender, diabetes duration, immigration status, treatment satisfaction, other chronic conditions, well-being (WHO-5), anxiety (GAD-7), diabetes burden (PAID-5) and the treatment modality (see table 4). In this model, the use of CSII and AID was associated with lower HbA1c. Additionally, being male was associated with higher HbA1c levels, while increasing treatment satisfaction was linked to lower HbA1c. A separate regression analysis was conducted to predict overall global health status. In this model, HbA1c was introduced in addition to the previously identified variables above. Better health status was predicted by the absence of other chronic conditions, lower HbA1c levels, higher well-being (WHO-5) and fewer anxiety symptoms.Table 4Regression analysis for model building of HbA1c and current health status in adolescents and young adults with T1DDependent variableIndependent variableR2βPHbA1c*0.243<0.001CSII use−0.3270.002AID use−0.360<0.001Gender, male0.2730.002Treatment satisfaction−0.329<0.001Global health status0.414<0.001Other chronic condition, yes−0.1670.024HbA1c*−0.1530.038WHO-5, well-being0.363<0.001GAD-7, anxiety−0.2730.007Variables included in the backward linear regression were as follows: for Inverse HbA1c—age, binary gender, diabetes duration, family migration background, other chronic medical condition, treatment satisfaction, WHO-5, GAD-7, PAID-5 and dummy-coded CSII and AID use; and for global health status—all of the above, plus inverse HbA1c.See the online supplemental document for complete analysis report using inverse HbA1c.*Analysis was performed using inverse HbA1c, but the signs of β are reversed for interpretability in its original unit.AID, automated insulin delivery; CSII, continuous subcutaneous insulin infusion; GAD-7, Generalized Anxiety Disorder Scale; HbA1c, glycated hemoglobin A1c; PAID-5, Problem Areas in Diabetes Scale; T1D, type 1 diabetes; WHO-5, WHO Wellbeing Index.Satisfaction with AID use On average, satisfaction with the AID system was high (see figure 1). Overall, 75.5% of respondents indicated that they felt better or much better with the AID system than with their previous form of therapy. However, 12.2% reported that their sleep had become much worse since they started using AID. For about half of the respondents (48.9%), the AID system had improved their confidence in their own diabetes management.Gender, age and disease duration were not significantly associated with total AID satisfaction scores. However, lower HbA1c levels were moderately correlated with greater satisfaction with AID use (r=0.336). Increased well-being and reduced anxiety were linked to higher AID satisfaction scores (r=0.316 and −0.260, respectively). Increasing diabetes burden had a strong negative association with AID satisfaction (r=−0.510).Discussion The aim of this survey was to examine how AYA experience the use of diabetes technologies in real-life settings. The focus was on PROs, as these are particularly important to the everyday life and well-being of this age group. As in other studies, 11 the use of AID systems was associated with improved glycemic control in AYA. Treatment satisfaction was highest among AID users, with the difference reaching statistical significance compared with those using CSII. Three-quarters of AID users reported positive experience with greater satisfaction with their AID system compared with their previous therapy. The 12% subgroup who reported a negative impact on sleep quality can be attributed to the use of first-generation sensors. Some systems used sensors that required regular calibration and often triggered alarms when the measurements were inaccurate (see online supplemental document for the types of AID used). This often occurred at night, disturbing sleep. The current generation of sensors requires little or no calibration, and these issues have now been resolved.These findings make a significant contribution to the ongoing debate surrounding the potential negative impact of new technologies, especially on AYA. Concerns often include the transparency of diabetes due to wearing the devices, changes in body image, loss of control with increasing automation and excessive demands due to the huge amount of diabetes-related data and technical problems.16 18 However, these concerns were not supported by our data. High treatment satisfaction among AID users and their appreciation for the convenience and relief provided by the technology point to a positive user experience in their daily lives. This finding is consistent with the data from the iDCL Trial,19 which reported that participants experienced a greater number of benefits than burdens as a result of using the AID system.Despite improvements in glycemic outcomes and treatment satisfaction, our study found no significant differences in depression and anxiety levels across treatment modalities. These findings are consistent with previous studies on AID use in children and adolescents19–22 as well as a meta-analysis including patients of all ages.3 A systematic review by Pease and colleagues2 found no clear superiority of any therapy in terms of QoL. Others have shown that diabetes technologies tend to have predominantly positive effects on diabetes-specific (eg, fear of hypoglycemia, diabetes distress, treatment satisfaction), rather than generic, PROs.23 In line with these findings, our sample showed lower levels of diabetes distress, among AID users compared with MDI users. However, this difference was small. In a recent study of an adult population, AID users showed significantly lower scores in the ‘eating distress’ subdimension, but not in the overall PAID score (compared with MDI).24 Regardless of their treatment modality, respondents expressed particular concern about potential complications and the exhaustion caused by the constant effort required to manage their diabetes. Overall, it seems that advanced technologies such as AID can help overcome some of the challenges of living with diabetes. However, they can only reduce the general burden of the condition to a limited extent.A recent meta-analysis examining diabetes burden by age group reported that adults experienced significant reductions in diabetes burden after initiating AID, however, not in children and teenagers.25 Notably, the mean age of pediatric samples in those studies ranged from 9 to 16 years, substantially younger than our AYA sample (20.7±3.0 years). Interestingly, parents of pediatric AID users reported reductions in their own diabetes burden. The authors suggested this may reflect the fact that diabetes burden and corresponding benefit of AID lie with the individual primarily responsible for diabetes care. Our analysis revealed the effects of treatment modality only after adjustments were made for age and gender. This suggests that the burden of diabetes increases during the transition to adulthood, due to a shift in self-management responsibilities.26 The benefits of AID may therefore become more apparent at this stage of development. These results reinforce the idea that the benefits of diabetes technology may not be uniform across age groups25 and highlight the need for further research focused on the unique challenges faced by AYA.Our exploratory regression analysis showed that HbA1 was significantly influenced by both treatment modality and treatment satisfaction, indicating the use of advanced technology and positive emotional response to treatment are key for achieving optimal metabolic outcomes. A separate regression analysis examining overall health status found that psychological PROs, alongside HbA1c, were significant predictors. Together, these findings demonstrate that while advanced diabetes technologies can improve clinical outcomes, overall health status may be shaped by broader domains, including mental health factors. This highlights the importance of addressing both metabolic control and psychological well-being in this population.Despite these results, it is concerning that our sample of AYA continues to report high levels of psychological distress.7 8 About 40% of our sample reported high diabetes-related distress, one third reported anxiety symptoms, and 44% showed increased depressive symptoms. This rate is four times higher than for metabolically healthy AYA (9.5%).27 Moreover, over 25% had received treatment for psychological issues within the past year. This may be partially explained by the self-selection of our sample (eg, high proportion of females, greater disease burden among camp participants) or the timing of the survey, which took place shortly after the COVID-19 pandemic, which has led to a deterioration in the mental health of AYA.28 Nevertheless, our findings confirm that quality of metabolic control and psychological health are key components of overall well-being. Given the high prevalence of psychological distress in this group, mental health support should be an integral part of diabetes management, as recommended by the ISPAD guidelines.29 AYA with T1D should be regularly screened for psychological distress and offered appropriate psychosocial support from their healthcare teams.30A key strength of this study was its real-world setting, capturing the lived experiences of AYA with T1D. However, our findings should be interpreted within the context of several limitations. First, participants were recruited from a voluntary diabetes camp, which may have introduced selection bias specific to this cohort. For instance, the sample was characterized by a disproportionately high number of female participants, limiting the interpretation of gender-specific analysis. As women score higher on measures of psychological distress,24 the over-representation of women must be considered when evaluating psychological outcomes. Similarly, four participants who identified as gender-diverse could not be included in the analysis due to the small subgroup size. To better investigate the influence of gender diversity, future research should make a greater effort to recruit individuals from this group. Another potential issue arising from our sampling method is that participants in a voluntary diabetes camp may be particularly committed to managing their condition. This may be reflected in the high proportion of AID users and participants with low HbA1c levels, particularly among female AYA. Conversely, participation in the camp may also have been motivated by greater challenges in diabetes management, which could explain the elevated prevalence of psychological distress observed in the sample. Therefore, caution is warranted, as the generalizability of the results may be limited. Second, all data, including glycemic measures, were self-reported; thus, recall and social desirability biases cannot be ruled out. Due to the anonymous nature of the study, participants’ medical records were not accessed. Nevertheless, given that the study focused on PROs, self-reporting was the most appropriate methodological approach. Finally, as this was a cross-sectional study, causal relationships among variables could not be established. For example, HbA1c levels and PROs prior to transitioning to an AID system were not assessed. A prospective longitudinal study is needed to examine the temporal relationships among technology use, glycemic control and psychological well-being in AYA with diabetes in order to establish causality.Our study demonstrated that AID use was associated with increased treatment satisfaction, reduction in diabetes-specific burden and improved metabolic control. However, general psychological distress, including symptoms of depression and anxiety, did not differ across treatment modalities and remained prevalent. As diabetes technology continues to evolve, improvements in both glucose control and user experience are anticipated. That said, to truly benefit AYA with T1D, these technological advances must be integrated into a holistic, patient-centered approach to care, which includes attention to mental health and overall well-being.",
  "title": "Disconnect between advanced diabetes technology and psychological well-being among young people: a cross-sectional analysis",
  "uid": "eca23558-56ab-5736-8cfd-aa438ff5999e"
}