Childhood obesity constitutes a major risk factor for future adverse health conditions. Multicomponent parent–child interventions are considered effective in controlling weight. Τhe ENDORSE platform utilizes m-health technologies, Artificial Intelligence (AI), and serious games (SG) toward the creation of an innovative software ecosystem connecting healthcare professionals, children, and their parents in order to deliver coordinated services to combat childhood obesity. It consists of activity trackers, a mobile SG for children, and mobile apps for parents and healthcare professionals. The heterogeneous dataset gathered through the interaction of the end-users with the platform composes the unique user profile. Part of it feeds an AI-based model that enables personalized messages. A feasibility pilot trial was conducted involving 50 overweight and obese children (mean age 10.5 years, 52% girls, 58% pubertal, median baseline BMI z-score 2.85) in a 3-month intervention. Adherence was measured by means of frequency of usage based on the data records. Overall, a clinically and statistically significant BMI z-score reduction was achieved (mean BMI z-score reduction −0.21 ± 0.26, p-value < 0.001). A statistically significant correlation was revealed between the level of activity tracker usage and the improvement of BMI z-score (−0.355, p = 0.017), highlighting the potential of the ENDORSE platform.

Leucocyte telomere length (LTL) is a robust marker of biological aging and is associated with obesity and cardiometabolic risk factors in childhood and adolescence. We investigated the effect of a structured, comprehensive, multidisciplinary, personalized, lifestyle intervention program of healthy diet and physical exercise on LTL in 508 children and adolescents (239 males, 269 females; 282 prepubertal, 226 pubertal), aged 10.14 ± 0.13 years. Participants were classified as obese (n = 267, 52.6%), overweight (n = 174, 34.2%), or of normal BMI (n = 67, 13.2%) according to the International Obesity Task Force (IOTF) cutoff points and were studied prospectively for one year. We demonstrated that LTL increased significantly after 1 year of the lifestyle interventions, irrespective of gender, pubertal status, or body mass index (BMI). Waist circumference was the best negative predictor of LTL at initial assessment. The implementation of the lifestyle interventions also resulted in a significant improvement in clinical (BMI, BMI z-score and waist to height ratio) and body composition indices of obesity, inflammatory markers, hepatic enzymes, glycated hemoglobin (HbA1C), quantitative insulin sensitivity check index (QUICKI), and lipid profile in all participants. These findings indicate that the increased LTL may be associated with a more favorable metabolic profile and decreased morbidity later in life.

Childhood obesity is a serious public health problem worldwide. The ENDORSE platform is an innovative software ecosystem based on Artificial Intelligence which consists of mobile applications for parents and health professionals, activity trackers, and mobile games for children. This study explores the impact of the ENDORSE platform on metabolic parameters associated with pediatric obesity and on the food parenting practices of the participating mothers. Therefore, the metabolic parameters of the 45 children (mean age: 10.42 years, 53% girls, 58% pubertal, mean baseline BMI z-score 2.83) who completed the ENDORSE study were evaluated. The Comprehensive Feeding Practices Questionnaire was used for the assessment of food parenting practices. Furthermore, regression analysis was used to investigate possible associations between BMI z-score changes and changes in metabolic parameters and food parenting practices. Overall, there was a statistically significant reduction in glycated hemoglobin (mean change = −0.10, p = 0.013), SGOT (mean change = −1.84, p = 0.011), and SGPT (mean change = −2.95, p = 0.022). Emotional feeding/food as reward decreased (mean change −0.21, p = 0.007) and healthy eating guidance increased (mean change = 0.11, p = 0.051). Linear regression analysis revealed that BMI z-score change had a robust and significant correlation with important metabolic parameters: HOMA-IR change (beta coefficient = 3.60, p-value = 0.046), SGPT change (beta coefficient = 11.90, p-value = 0.037), and cortisol change (beta coefficient = 9.96, p-value = 0.008). Furthermore, healthy eating guidance change had a robust negative relationship with BMI z-score change (beta coefficient = −0.29, p-value = 0.007). Conclusions: The Endorse digital weight management program improved several metabolic parameters and food parenting practices.

In clinical practice, differences in glucocorticoid sensitivity among healthy subjects may influence the outcome and any adverse effects of glucocorticoid therapy. Thus, a fast and accurate methodology that could enable the classification of individuals based on their tissue glucocorticoid sensitivity would be of value. We investigated the usefulness of untargeted plasma metabolomics in identifying a panel of metabolites to distinguish glucocorticoid-resistant from glucocorticoid-sensitive healthy subjects who do not carry mutations in the human glucocorticoid receptor (NR3C1) gene. Applying a published methodology designed for the study of glucocorticoid sensitivity in healthy adults, 101 healthy subjects were ranked according to their tissue glucocorticoid sensitivity based on 8:00 a.m. serum cortisol concentrations following a very low-dose dexamethasone suppression test. Ten percent of the cohort, i.e., 11 participants, on each side of the ranking, with no NR3C1 mutations or polymorphisms, were selected, respectively, as the most glucocorticoid-sensitive and most glucocorticoid-resistant of the cohort to be analyzed and compared with untargeted blood plasma metabolomics using gas chromatography–mass spectrometry (GC–MS). The acquired metabolic profiles were evaluated using multivariate statistical analysis methods. Nineteen metabolites were identified with significantly lower abundance in the most sensitive compared to the most resistant group of the cohort, including fatty acids, sugar alcohols, and serine/threonine metabolism intermediates. These results, combined with a higher glucose, sorbitol, and lactate abundance, suggest a higher Cori cycle, polyol pathway, and inter-tissue one-carbon metabolism rate and a lower fat mobilization rate at the fasting state in the most sensitive compared to the most resistant group. In fact, this was the first study correlating tissue glucocorticoid sensitivity with serine/threonine metabolism. Overall, the observed metabolic signature in this cohort implies a worse cardiometabolic profile in the most glucocorticoid-sensitive compared to the most glucocorticoid-resistant healthy subjects. These findings offer a metabolic signature that distinguishes most glucocorticoid-sensitive from most glucocorticoid-resistant healthy subjects to be further validated in larger cohorts. Moreover, they support the correlation of tissue glucocorticoid sensitivity with insulin resistance and metabolic syndrome-associated pathways, further emphasizing the need for nutritionists and doctors to consider the tissue glucocorticoid sensitivity in dietary and exercise planning, particularly when these subjects are to be treated with glucocorticoids.