519 - Child BMI in Central Appalachia in the COVID Era
Saturday, April 23, 2022
3:30 PM – 6:00 PM US MT
Poster Number: 519 Publication Number: 519.239
Paige A. Poffenberger, West Virginia University, Morgantown, WV, United States; Rachel A. Kohler, West Virginia University Children's Hospital, Canonsburg, PA, United States; Christa Lilly, West Virginia University, Morgantown, WV, United States; Lee Pyles, Lee Pyles, Morgantown, WV, United States
MS3 West Virginia University Morgantown, West Virginia, United States
Background: Cross sectional studies have documented that the Covid-19 Pandemic has created an environment in which many children have faced weight gain and BMI fluctuations. These fluctuations have not been quantified longitudinally nor have possible confounding factors contributing to these changes in BMI been identified. We investigated serial BMI in a rural health system in central Appalachia.
Objective: We expect to see significant increases in BMI longitudinally over the time period with confounding variables to be sex and age.
Design/Methods: 277,854 children under 18 were found in WVU Health System via Epic EHR report writer query. 137,526 subjects had at least a single BMI determination. Change in BMI was calculated in 130,000 WV children across 4 time periods. 4 BMI data time periods represented pre and post onset of pandemic 3/1/2020: (1) 2018-06/30/2019, (2) 07/01/2019-2/29/2020, (3) 03/01/2020-12/31/2020, (4) 01/01/2021-10/01/2021. A mixed model analysis for repeated measures included age and sex plus interactions as factors and addressed missing data. See table 2 for results. Children under age 2 were not included in ANOVA.
Results: WV BMI is above normal at baseline as with previous assays. Mixed model analysis demonstrated striking increase in BMI over time that accelerated in early COVID (time 3). Age was a significant factor and females had a higher BMI at baseline, but the increase was equivalent for M/F. Conclusion(s): BMI Percentile increased across the COVID period in children. For future improvement, many children had no BMI measure in the entire observation period. Selection bias for BMI measurement is possible but about ⅓ of WV children (over 300,000 total) had a BMI in WVU Health System in the period of interest.. PAS entry CV.pdf Mixed Model Analysis of BMI PercentileRepeated measures model includes sex and age factors.