(P11-046-23) Optimum Sensitivity and Specificity Cutoffs for Conventional Biomarkers for Inadequate Vitamin B12 Status Based on the New Combined Indicator of Status, cB12

Objectives: Diagnostic accuracy of conventional markers for vitamin B₁₂ (B₁₂ deficiency depends on the marker and cutoff range used. Using NHANES data we previously showed that a single marker may overestimate the prevalence of inadequate B₁₂ status compared to the combined indicator (3cB₁₂), an index that incorporates serum B₁₂, methylmalonic acid (MMA), and total homocysteine (tHcy). Given practical limitations to use multiple markers, our goal was to determine new diagnostic cutoffs for each conventional marker that corresponded to inadequate B₁₂ status as classified by 3cB₁₂ ≤-0.5.

Methods: We applied ROC analysis to cross-sectional data from NHANES 1999-2004 for adults ≥20 y (n=12,341) and further for adults with normal renal function (n=9,186) who had inadequate B₁₂ status (n=367 and n=175, respectively) to determine for each conventional marker an optimum decision cutoff based on a combination of optimum sensitivity and specificity.

Results: The diagnostic power to predict inadequate B₁₂ status based on area under the curve (AUC [95% CI]) was similar for the conventional markers: MMA 0.951 (0.927, 0.974), B₁₂ 0.948 (0.934, 0.962), and tHcy 0.935 (0.918, 0.953). For all adults, the ROC derived optimum decision cutoffs (95% CI) [sensitivity/specificity] for detecting inadequate B₁₂ status were: B₁₂ < 240 (205, 241) pmol/L [87.8/82.2], MMA >248 (226, 269) nmol/L [87.8/93.4], and tHcy >10.7 (10.5, 12.4) mmol/L [87.6/82.5]. Similarly, for adults with normal renal function the optimum decision cutoffs (95% CI) [sensitivity/specificity] were: B₁₂ < 212 (204, 239) pmol/L [87.8/88.6], MMA >240 (200, 260) nmol/L [86.1/94.3], and tHcy >10.5 (10.5, 10.7) mmol/L [88.9/85.3].

Conclusions: Of the 3 conventional markers, MMA had the highest sensitivity to predict inadequate B₁₂ status, regardless of an individual’s renal status, with similar specificity as B₁₂ and tHcy. The optimum cutoff was similar between all adults and those with normal renal function for MMA and tHcy.

Funding Sources: This work was supported by direct appropriations from the US Congress.