Associations between collagen X biomarker and linear growth velocity in a pediatric chronic kidney disease cohort

Abstract

Background: Collagen X biomarker (CXM) is a novel biomarker of linear growth velocity. We investigated whether CXM correlated with measured growth velocity in children with impaired kidney function.

Methods: We used data from children aged 2 through 16 years old enrolled in the Chronic Kidney Disease in Children (CKiD) study. We assessed the association between CXM level and growth velocity based on height measurements obtained at study visits using linear regression models constructed separately by sex, with and without adjustment for CKD covariates. Linear mixed-effects models were used to capture the between-individual and within-individual CXM changes over time associated with concomitant changes in growth velocity from baseline through follow-up.

Results: A total of 967 serum samples from 209 participants were assayed for CXM. CXM correlated more strongly in females compared to male participants. After adjustment for growth velocity and CKD covariates, only proteinuria in male participants affected CXM levels. Finally, we quantified the between- and within-participant associations between CXM level and growth velocity. A between-participant increase of 24% and 15% in CXM level in females and males, respectively, correlated with a 1 cm/year higher growth velocity. Within an individual participant, on average, 28% and 13% increases in CXM values in females and males, respectively, correlated with a 1 cm/year change in measured growth.

Conclusions: CXM measurement is potentially a valuable aid for monitoring growth in pediatric CKD. However, future research, including studies of CXM metabolism, is needed to clarify whether CXM can be a surrogate of growth in children with CKD. A higher resolution version of the Graphical abstract is available as Supplementary information.

Keywords: Biomarker; CKD; CXM; Growth; Pediatrics.

Figure 1.

Participant and biosample selection for final analyses. ¹ Age inclusion criteria: 2-16 years old; serum biosample inclusion criteria: available biosample at midpoint of 3 consecutive visits. 674 participants excluded ² CKiD study approved biosample quantity : 1001 samples. ³ Biosample vial <0.5mL (5 samples); missing biosamples (6 biosamples). 11 biosamples excluded ⁴ Height outliers identified using conditional studentized residuals (CSRs). 14 biosamples excluded. An additional 2 participants and 2 biosamples excluded for no longer meeting minimum criteria for number of biosamples. ⁵ Excluded due to the known association of abnormal growth in these patients; 2 participants and 7 biosamples excluded.

Figure 2.

CXM concentrations versus age across 976 person-visits stratified by a) females contributing 367 person-visits and b) males providing 609 person-visits. Blue lines represent locally weighted smoothing with 95% confidence intervals shown in as gray shaded region.

Figure 3.

Calculated growth velocity versus age across 976 person-visits stratified by a) females contributing 367 person-visits and b) males providing 609 person-visits. Blue lines represent locally weighted smoothing with 95% confidence intervals shown in as gray shaded region.

Figure 4.

Plot of individual trajectories of CXM (ng/mL) and growth velocity (GV; cm/year) by age for a) the first 7 females with ≥7 CXM samples and b) first 7 males with ≥9 CXM samples. CXM = solid dot. Measured GV = solid triangle.

Figure 5.

CXM concentrations versus calculated growth velocity across 976 person-visits stratified by a) females contributing 367 person-visits and b) males providing 609 person-visits. Blue lines represent locally weighted smoothing with 95% confidence intervals shown in as gray shaded region.