Endocrine Abstracts

JOINT1077

Introduction: Overweight and obesity are a global health concern, affecting one in five schoolchildren in Switzerland. Although body height and weight are routinely measured during clinical visits and recorded in electronic health records (EHRs), these data are rarely used for research. We assessed the prevalence of overweight and obesity in Swiss children with use of EHRs.

Methods: SwissPedHealth is a national data stream of the Swiss Personalized Health Network (SPHN). The nested project SwissPedGrowth analyses data from children visiting hospitals in Basel, Bern, Geneva, Lausanne, Zurich, Luzern, or St. Gallen between 2017–2023. Clinical data warehouses extracted EHRs of in- and outpatients and sent anonymized data to the BioMedIT server for analysis. We included children aged 2–17 years, excluding those with ICD-10 coded diagnoses affecting BMI (e.g. intestinal infections, malignant neoplasms). We cleaned anthropometric data using the growthcleanr algorithm of Daymont et al. and calculated BMI using height and weight measurements taken within 30 days of each other. We computed BMI z-scores based on the International Obesity Task Force (IOTF) references, excluding implausible z-scores <-5 or >5. We calculated the prevalence of obesity among children based on IOTF cut-offs using both the highest z-score of a child (“obese ever”) and the mean z-score of a child (“obese on average”).

Results: Initial data from Basel, Geneva, and Lausanne contained 38, 588 children (54% boys) aged 2-17 years, with 49, 880 heights and 97, 789 weights. After excluding patients with BMI-affecting diseases (1%), without height and weight (63%), and with implausible BMI values (0.1%), we included 13, 708 children (36%), with a median of 2 (IQR: 1, 5) BMI measurements per child. Based on the highest BMI z-score, 1, 165 (8%) children were classified as obese ever, 2, 305 (17%) as overweight, and 10, 238 (75%) had a normal BMI throughout. Based on the mean z-score, 860 (6%) children were classified as obese on average, 1, 892 (14%) as overweight, and 10, 956 (80%) had a normal BMI. Older children were more often obese than young children (7-17-year-olds: 8% vs 2-7-year-olds: 4%). There was no difference between boys and girls. The prevalence of childhood obesity and overweight was consistent with previous Swiss studies.

Conclusion: SwissPedGrowth demonstrates the feasibility of using EHR data to investigate childhood obesity. Further analyses will explore related factors and examine how obesity is diagnosed and managed in hospitals and primary care (SwissPedHealth-PREPP). SwissPedGrowth provides a framework for future studies on growth and obesity in Swiss children.