Densitometry Versus Bioimpedance for Modeling Vitamin D-Endocrine and Metabolic Associations in Pediatric Obesity: A Cross-Sectional Parallel-Modality Analysis.

DXA systematically measured higher fat percentage than BIA in this pediatric obesity cohort.

• DXA fat % was 46.6% versus BIA fat % of 36.7%.
• The cohort consisted of 165 children with simple obesity and no vitamin D prophylaxis.
• Data were collected between July 2022 and July 2025.
• Bland-Altman analysis was used to assess DXA-BIA agreement.
• This finding is consistent with prior literature showing BIA systematically underestimates adiposity compared to DXA.

25(OH)D correlated inversely with adiposity by both modalities in univariate analysis, but adiposity was not a significant determinant of 25(OH)D after covariate adjustment.

• Univariate inverse correlations: DXA rho = -0.16, BIA rho = -0.19.
• After adjustment for season and age, adiposity was not a significant determinant of 25(OH)D with either modality.
• The cohort median 25(OH)D level was 21.9 ng/mL.
• Partial correlations and linear regression were used for adjustment analyses.

Seasonal and age factors dominated 25(OH)D variability rather than adiposity in this homogeneous pediatric obesity cohort.

• Age and season were recorded and used to adjust for potential covariates.
• Adiposity contributed little to within-group 25(OH)D variability once season and age were accounted for.
• The cohort median age was 13 years.
• The sample was restricted to children with simple obesity, creating a homogeneous within-group adiposity distribution.

No mediation of vitamin D's associations with metabolic outcomes via adiposity was detected with either measurement modality.

• Mediation-style models were tested linking vitamin D to metabolic outcomes through adiposity.
• Metabolic outcomes included glucose, insulin, HOMA-IR, and lipids.
• The null mediation finding was consistent across both DXA and BIA adiposity estimates.
• Parallel mediation analyses were conducted for both modalities.

The vitamin D-PTH-calcium endocrine axis was robust and consistent across both DXA and BIA measurement modalities.

• PTH, calcium, and phosphate were measured as part of the biochemical panel.
• The vitamin D-PTH-calcium axis associations were detected regardless of whether DXA or BIA was used to measure adiposity.
• Vitamin D showed endocrine but not metabolic associations in this cohort.
• This endocrine axis consistency was observed across parallel analyses including partial correlations and linear regression.

DXA did not outperform BIA for physiologic modeling of vitamin D associations in this pediatric obesity cohort.

• Both modalities yielded similar conclusions for all tested associations: adiposity-25(OH)D correlations, mediation models, and the vitamin D-PTH-calcium axis.
• Despite DXA measuring higher absolute fat percentage (46.6% vs. 36.7%), the direction and significance of physiologic associations were equivalent.
• Parallel analyses were designed specifically to compare the two modalities side-by-side.
• The authors concluded that within a homogeneous pediatric obesity cohort, the choice of modality does not materially affect physiologic modeling results.