A-095 Deciphering the Physiological Distribution and Clinical Relevance of Magnesium Forms in a Large Patient Cohort

Abstract Background Magnesium (Mg²?) is a key electrolyte routinely measured in clinical chemistry laboratories to assess metabolic, cardiovascular, and renal function. Despite its biological importance, standard laboratory assessments primarily rely on total magnesium (tMg), which includes both protein-bound and free magnesium fractions. However, only ionized magnesium (iMg²?) represents the physiologically active form involved in enzymatic processes and cellular homeostasis. Studies indicate that tMg underestimates hypomagnesemia, particularly in critical illness, chronic kidney disease, and cardiovascular disorders, leading to potential mismanagement. This study is the first that evaluates the agreement between tMg and iMg²? in a large patient cohort and identifies biological determinants of discordance. Methods A total of 2,853 hospitalized patients were analyzed for tMg and iMg²?, alongside albumin, pH, creatinine, and estimated glomerular filtration rate (eGFR). Agreement between tMg- and iMg²?-based classifications was assessed using Cohen’s kappa coefficient. Multivariable logistic regression determined independent predictors of discordance, including albumin, eGFR, pH, and age. Sensitivity and specificity analyses assessed the clinical relevance of using iMg²? as a reference marker. Results The cohort included 2,853 hospitalized patients (mean age 59.2 ± 16.8 years; 52% male). Hypoalbuminemia (<30 g/L) was present in 42%, chronic kidney disease (eGFR <60 mL/min/1.73m²) in 35%, and metabolic acidosis (pH <7.35) in 28%. Based on iMg²? levels, 24% were hypomagnesemic, 66% normomagnesemic, and 10% hypermagnesemic. Overall, 36.7% of patients exhibited discordance between tMg and iMg²? classifications. Agreement was lower in hypoalbuminemic (61.0%) vs. normoalbuminemic (74.2%) patients. Logistic regression identified hypoalbuminemia as a strong predictor of discordance (OR = 0.96 per g/L increase in albumin, p < 0.001). Impaired renal function (eGFR <60 mL/min/1.73m²) increased discordance risk (OR = 1.008 per 10 mL/min decrease, p < 0.001). Younger age also correlated with discordance (OR = 0.99 per year, p = 0.005), while pH had no significant impact (p = 0.146). Sensitivity analysis confirmed iMg²? had superior predictive accuracy for clinically relevant hypomagnesemia compared to tMg. Conclusion This study highlights that tMg is an unreliable surrogate for biologically active magnesium, particularly in patients with hypoalbuminemia or impaired kidney function. Given the strong association between hypomagnesemia and adverse clinical outcomes, reliance on tMg alone may lead to misdiagnosis and inappropriate treatment. Routine implementation of iMg²? measurement could improve clinical decision-making, particularly in nephrology, cardiology, and intensive care settings. Measuring iMg²? instead of tMg could refine hypomagnesemia diagnosis, leading to more targeted and effective magnesium supplementation. This is particularly relevant in critically ill patients, where electrolyte imbalances impact morbidity and mortality. Integrating iMg²? assessments into biochemical panels may improve risk stratification for cardiovascular and renal complications. Future guidelines should consider incorporating iMg²? as a standard marker for clinical evaluation of magnesium homeostasis, optimizing patient outcomes through better electrolyte management.