Post-acute dyslipidemia and abnormal BMI in children and adolescents with COVID-19: An EHR Cohort Study from the RECOVER Initiative

Yuqing Lei, Ting Zhou, Bingyu Zhang, Dazheng Zhang, Huilin Tang, Jiajie Chen, Qiong Wu, Lu Li, L. Charles Bailey, Michael J. Becich, Saul Blecker, Dimitri Christakis, Daniel Fort, Sharon J. Herring, Wenke Hwang, Amrik Singh Khalsa, Susan Kim, David M. Liebovtiz, Abu Saleh Mohammad Mosa, Suchitra Rao, Soumitra SenGupta, Song Xing, Yacob G. Tedla, Ravi Jhaveri, Caren Mangarelli, Christopher B. Forrest, Yong ChenmedRxiv 2025 · DOI

TL;DR

A massive study of over 1.6 million children reveals that SARS-CoV-2 infection is linked to a significantly higher risk of developing dyslipidemia (abnormal cholesterol/triglycerides) and high BMI in the months following the acute illness. These metabolic changes occur even in children who were not previously obese, suggesting that COVID-19 may directly trigger long-term metabolic dysfunction in the pediatric population.

What Problem Does This Paper Address?

While the acute respiratory symptoms of COVID-19 are often milder in children than in adults, the medical community is increasingly concerned about "Post-Acute Sequelae of SARS-CoV-2" (PASC), or Long COVID. In adults, researchers have already documented increased risks of diabetes and heart disease following infection. However, data for children has remained scarce and often relied on subjective diagnosis codes rather than objective laboratory data.

This paper addresses a critical gap: Does a COVID-19 infection predispose children and adolescents to metabolic "silent killers" like high cholesterol and obesity? Understanding this is vital because metabolic issues acquired in childhood often persist into adulthood, potentially amplifying lifetime cardiovascular risks for an entire generation.

Methods

The researchers conducted a retrospective cohort study using the RECOVER pediatric Electronic Health Record (EHR) database. They analyzed two parallel groups:

  1. Dyslipidemia Cohort: Children aged 0–21.
  2. BMI Cohort: Children aged 2–21.

To ensure the results weren't just a fluke of the "pandemic lifestyle" (like gym closures or poor diet), they compared COVID-positive children against contemporaneous COVID-negative peers.

The study used a rigorous statistical approach called propensity score-stratified Poisson regression, adjusting for hundreds of variables including age, sex, race, pre-existing conditions, and even how often a child visited the doctor. Most importantly, they used empirical calibration with 36 "negative control outcomes" (conditions like ingrown toenails that shouldn't be caused by COVID) to filter out any residual systematic bias in the data.

Datasets

The study leveraged the RECOVER (Researching COVID to Enhance Recovery) database, which aggregates EHR data from 25 US children’s hospitals and health institutions.

  • Timeframe: March 2020 to September 2023.
  • Scale: The analysis included 384,289 COVID-positive patients for the lipid analysis and 285,559 for the BMI analysis, compared against over 1 million negative controls.
  • Metric: Unlike previous studies, this one used objective laboratory thresholds (actual cholesterol numbers and measured BMI percentiles) rather than just doctor-assigned diagnosis codes.

Key Insights and Ideas

  • Objective Evidence: The risk isn't just a "paperwork" increase in diagnoses; it is reflected in actual blood work. COVID-positive kids showed higher rates of abnormal LDL (bad cholesterol), low HDL (good cholesterol), and high triglycerides.
  • The "Infection Effect": Because the researchers compared infected kids to non-infected kids during the same time period, the results suggest the metabolic changes are likely a result of the virus itself, not just general pandemic-related stress or inactivity.
  • Biological Triggers: The authors hypothesize that SARS-CoV-2 may disrupt lipid homeostasis by binding to ACE2 receptors or by triggering systemic inflammation (like Interleukin-6), which can interfere with how the liver processes fats.
  • Independence from Baseline Weight: Interestingly, the increased risk of abnormal BMI was consistent regardless of whether the child was obese before the infection, suggesting the virus might be a "great equalizer" in terms of metabolic risk.

Key Results

The study found that children with documented COVID-19 had a significantly higher risk of metabolic issues in the post-acute period (28–179 days after infection):

  • Overall Dyslipidemia: A 23% higher risk (aRR 1.23) of any abnormal lipid result.
  • Specific Lipid Fractions:
    • 28% higher risk for high triglycerides.
  • 24% higher risk for low HDL (good cholesterol).
  • 19% higher risk for high LDL (bad cholesterol).
  • Abnormal BMI: A 15% higher risk (aRR 1.15) of developing a BMI in the 95th percentile or higher.
  • Incidence: While the absolute percentages are small (e.g., 0.9% of COVID+ kids developed dyslipidemia vs. 0.7% of controls), the population-level impact is substantial given how many millions of children have been infected.

Limitations and Future Work

The authors acknowledge several limitations:

  • Under-ascertainment: Children who had asymptomatic COVID-19 or tested at home might have been misclassified as "negative."
  • Indication Bias: Kids with more severe symptoms might have been more likely to receive lab tests, potentially skewing the data toward more "visible" cases.
  • Short-term Window: The study only looked at the first six months post-infection. It remains unclear if these metabolic changes are permanent or if they resolve over years.
  • Future Directions: The researchers call for routine lipid and BMI monitoring in pediatric follow-up care for COVID-19 survivors to enable early intervention and prevent long-term heart disease.

Summary updated

Abstract

Children and adolescents with documented COVID-19 were associated with higher risks of new-onset dyslipidaemia and abnormal BMI in the post-acute period compared with COVID-19-negative peers. Associations were consistent across lipid fractions, remained after empirical calibration, and were similar after accounting for baseline obesity.
BibTeX
@misc{lei2025post,
  author = {Yuqing Lei and Ting Zhou and Bingyu Zhang and Dazheng Zhang and Huilin Tang and Jiajie Chen and Qiong Wu and Lu Li and L. Charles Bailey and Michael J. Becich and Saul Blecker and Dimitri Christakis and Daniel Fort and Sharon J. Herring and Wenke Hwang and Amrik Singh Khalsa and Susan Kim and David M. Liebovtiz and Abu Saleh Mohammad Mosa and Suchitra Rao and Soumitra SenGupta and Song Xing and Yacob G. Tedla and Ravi Jhaveri and Caren Mangarelli and Christopher B. Forrest and Yong Chen},
  title = {{Post-acute dyslipidemia and abnormal BMI in children and adolescents with COVID-19: An EHR Cohort Study from the RECOVER Initiative}},
  year = {2025},
  doi = {10.1101/2025.09.17.25336020},
  howpublished = {medRxiv},
}
Figure