Background: Lipidomics, a branch of metabolomics, is an attractive technique to characterize bone marrow lipid composition, which may be associated with skeletal acquisition and homeostasis. However, the reliability of lipidomics-derived lipid composition of the bone marrow is unknown, especially for pediatric populations with bone fragility. The purpose of this study was to evaluate the intersite reliability and standard error of measurement (SEM) of vertebral bone marrow lipid composition at the thoracic (T11/T12) and lumbar (L1/L2) spine determined by targeted lipidomics among children with varying degrees of bone fragility undergoing routine orthopedic surgery.
Methods: Children aged between 12 and 19 years of age, with a confirmed diagnosis of adolescent idiopathic scoliosis or neuromuscular scoliosis and cerebral palsy, and undergoing routine posterior spinal fusion surgery at our institution were initially included in this study. Transpedicular vertebral body bone marrow samples were taken from thoracic (T) or lumbar (L) vertebrae. Further inclusion criteria involved having bone marrow extracted from both T11 and T12 (n = 24) or L1 and L2 (n = 19). Lipid composition was measured using a targeted lipidomics technique and examined as the saturated, monounsaturated, and polyunsaturated index and as individual fatty acids. Relative and absolute test-retest reliability was assessed using the intraclass correlation coefficient (ICC) and SEM.
Results: For the T11/T12 analysis: the ICC and SEM were 0.59 and 1.7% for the saturated index, 0.31 and 6.2% for the monounsaturated index, and 0.44 and 6.1% for the polyunsaturated index; the ICC showed a considerable range for individual fatty acids from 0.07 (fatty acid 20:2) to 0.82 (15:0) with 62.1% of the fatty acids having poor reliability (i.e., ICC < 0.50). For the L1/L2 analysis: the ICC and SEM were 0.50 and 2.4% for the saturated index, −0.12 and 6.0% for the monounsaturated index, and 0.00 and 4.9% for the polyunsaturated index; the ICC showed a considerable range for individual fatty acids from −0.34 (18:1_n-9) to 0.88 (15:0 and 18:3_n-3) with 79.3% of the fatty acids having poor reliability.
Conclusions: The intersite test-retest reliability was poor-to-moderate for index measures and generally poor for individual fatty acids for the thoracic and lumbar spine. At this time, it is not recommended to pool bone marrow adipose tissue across vertebral sites for bone marrow adiposity research or clinical monitoring for pediatric populations with bone fragility.