The possibilities of artificial intelligence and telemedicine in controlling the obesity epidemic

Obesity is a chronic, multifactorial disease associated with major cardiometabolic complications and increasing healthcare burden. Because long-term weight management is strongly limited by poor adherence and insufficient follow-up, telemedicine and digital health interventions have been proposed as tools to support continuous monitoring and behavioural reinforcement. In this thesis, artificial intelligence is considered as a technological component of modern telemedicine solutions, particularly in relation to automated self-monitoring and adaptive feedback. The aim of the study was to develop a scientifically grounded concept for optimizing dynamic health monitoring in obesity treatment using telemedicine and information and communication technologies (ICT). The study is based on secondary data analysis and combines a literature review, a SWOT analysis, and a meta-analysis of randomized controlled trials. A PRISMA-guided search was conducted in PubMed (2010–2024) and identified seven eligible randomized controlled trials in adults with overweight or obesity. Random-effects meta-analysis was performed using the REML estimator with Hartung–Knapp adjustment, with body weight change (kg) as the primary outcome. Three predefined time windows were analysed (approximately 3 months, 6 months, and 12 months), and available 24-month outcomes were examined separately. Across all time windows, pooled estimates consistently favoured digital and telemedicine-based interventions over usual care. At 3 months, the pooled mean difference was −1.69 kg (95% CI −3.81 to 0.42; p = 0.0837) with substantial heterogeneity. At 6 months, the pooled mean difference was −2.82 kg (95% CI −10.64 to 5.01; p = 0.1370) based on two studies. At 12 months (the primary interpretation time point), five studies showed a pooled mean difference of −0.73 kg (95% CI −1.52 to 0.06; p = 0.0627) with low heterogeneity (I² = 5.9%). Only one study reported 24-month outcomes, demonstrating a mean difference of −1.10 kg (95% CI −1.88 to −0.32; p = 0.0057). Subgroup analyses did not identify statistically significant differences between digital-only interventions and programs supplemented by structured human support. Overall, the findings suggest that digital and telemedicine-based interventions produce a modest but directionally consistent reduction in body weight and may function as scalable components of long-term obesity care. However, the limited number of trials and modest effect sizes indicate the need for larger and longer-term studies, as well as careful implementation within structured chronic care pathways.