Biological variation and inter-relation between key calcium homeostasis biomarkers

Area of Science:

• Biochemistry and Endocrinology
• Clinical Laboratory Science
• Biostatistics

Background:

• Calcium homeostasis is regulated by complex interactions between calcium, phosphate, vitamin D, and parathyroid hormone (PTH).
• Understanding the biological variability of these biomarkers is crucial for accurate interpretation of clinical laboratory results.
• Existing reference intervals may not fully capture the dynamic interplay between these co-regulated analytes.

Purpose of the Study:

• To determine the within- and between-subject biological variation (CV_I, CV_G) of key calcium homeostasis biomarkers.
• To develop and validate a predictive model for parathyroid hormone (PTH) concentrations based on calcium, phosphate, and 25-OH vitamin D levels.
• To assess the clinical utility of the predictive model in identifying non-physiological results in routine practice.

Main Methods:

• Weekly blood sampling over 10 weeks from 30 healthy volunteers, with duplicate analysis of biomarkers.
• Estimation of biological variation using ANOVA and a Bayesian model, including outlier removal.
• Development of a linear mixed-effects model (LMM) to predict PTH based on calcium, phosphate, and 25-OH vitamin D.

Main Results:

• Consistent estimates for within-subject biological variation (CV_I) were obtained for PTH (approx. 18%), calcium (approx. 1.7%), phosphate (approx. 9.3%), 25-OH vitamin D (approx. 4.7-7.2%), and 1,25-OH vitamin D (approx. 16.2-17.1%).
• The LMM identified calcium as the primary predictor for PTH, yielding the equation: PTH (pmol/L)=exp(3.46-2.67·ln(Ca (mmol/L)) ± Z·0.29). Phosphate and 25-OH vitamin D did not significantly improve the prediction interval.
• The predictive PTH-calcium model flagged 56% of routine results compared to 41.6% flagged by conventional reference intervals, indicating enhanced sensitivity.

Conclusions:

• Two distinct statistical methods yielded comparable estimates for biological variation in calcium homeostasis biomarkers.
• A predictive model incorporating calcium effectively forecasts PTH concentrations, demonstrating the co-regulation of these analytes.
• This model holds potential for improving the detection of physiologically aberrant results and increasing clinician awareness of calcium metabolism disturbances.