Is It Possible to Detect Effects of the COVID-19 Pandemic on the Prevalence of Malnutrition-Anorexia Using Big Data Tools in a Pediatric Population?

This research investigates the consequences of the COVID-19 lockdown on pediatric health, with a specific focus on nutritional deficiencies potentially linked to disordered eating. Contemporary Big Data analytical techniques provide a powerful framework for detecting such population-level shifts and exploring their underlying drivers. Primary Aim: To determine if significant changes occurred in the prevalence of malnutrition, identified by a low body mass index (BMI), among children following confinement during COVID-19. Methodology: We conducted a cross-sectional analysis of anonymized data from digital health records. Key metrics—gender, age, weight, and height—were analyzed for a cohort of young people, comparing a pre-pandemic baseline (early 2020) with a post-confinement period (early 2022). Advanced computational models were applied to process these extensive datasets. The analytical strategy utilized the Cole-Green LMS algorithm with penalized likelihood, implemented via RefCurv 0.4.2 software, chosen for its efficacy with large-scale information. Selection of hyperparameters was guided by the Bayesian Information Criterion (BIC). Our specialists in mathematics endorsed this methodological pathway as the most robust for our objectives. Nutritional status was assessed by identifying individuals whose BMI fell more than 2.0 standard deviations below the age-adjusted population mean. Findings: The study included 66,975 clinical records from individuals under 16 years, analyzing over 1.2 million distinct data points. Results and comparative visualizations across different geographical districts are presented. A notable rise of 60 instances per 100,000 residents was observed following the pandemic. This increase was not uniform, showing distinct patterns: it was more marked in boys than girls, affected females more in rural settings, and males more in urban centers. Interpretation: Leveraging Big Data allows for highly efficient public health surveillance, pinpointing demographic groups that would most benefit from targeted support, thus ensuring optimal use of limited medical resources. Based on these results, proactive screening programs in specific urban zones should concentrate on male adolescents, while in certain rural areas, the focus should shift to female adolescents, who may constitute an under-identified at-risk group.