Effects of Hyperlipidemia on Plasma Sodium, Potassium, and Chloride Measurements by an Indirect Ion-Selective Electrode Measuring System

We used indirect ion-selective electrode (ISE) methods (Roche/Hitachi Modular) to investigate the effect of hyperlipidemia (cholesterol plus triglycerides) on measurements of sodium, potassium, and chloride and the ability of published formulas to correct for the decrease in measured Na+, K+, and Cl−. As recommended by Kroll (1), we did not use enriched samples but instead used 21 patient samples (lithium-heparin plasma and serum) with triglyceride concentrations >15 mmol/L, total lipid concentrations (cholesterol plus triglycerides; Roche/Hitachi Modular) >20 mmol/L, a lipemic index (Modular) >3, and a total protein concentration within the reference interval (62–83 g/L). We analyzed samples before and after ultracentrifugation (Airfuge; Beckman Coulter; 15 min at 107 000g) by indirect ISE potentiometry (Modular) and direct ISE (Rapidlab 865; Bayer). The ultracentrifuged indirect ISE result (y) and nonultracentrifuged direct ISE result (z) were subtracted from the nonultracentrifuged indirect ISE result (x; see Fig. 1 in the Data Supplement that appears with the online version of this letter at http://www.clinchem.org/content/vol52/issue1/). The lipid concentration varied inversely, not only with Na+ but also with Cl− and (to a lesser degree) K+ concentrations, as measured by the indirect ISE system (see Fig. 1 in the online Data Supplement). A 10 mmol/L increase in total lipid concentration decreased the Na+ and Cl− concentrations by ∼1 mmol/L and K+ by ∼0.04 mmol/L. We developed formulas to calculate corrected Na+ (F3), Cl− (F5), and K+ (F7) results and compared the results with those from published formulae (listed below): F1: Corrected Na+ = Measured Na+ + [measured serum lipids/4.63 g/L (∼5.23 mmol/L)]2 F2: Corrected Na+ = Measured Na+ + [{[0.21 × triglycerides (g/L)] − 0.6} × (Na+/100)]3 F3: Corrected Na+ = Measured Na+ + (total lipids/10) F4: Corrected Cl− = Measured Cl− + [{[0.21 × triglycerides (g/L)] − 0.6} × (Cl−/100)]3 F5: Corrected Cl− = Measured Cl− + (total lipids/10) F6: Corrected K+ = Measured K+ + [{[0.21 × triglycerides (g/L)] − 0.6} × (K+/100)]3 F7: Corrected K+ = Measured K+ + (total lipids × 0.004) In a second set of 24 hyperlipidemic patient samples, analyzed prospectively, we observed effects of total lipids similar to those seen in the first set; we also demonstrated the utility of the formulas derived from the first set (Table 11 ). The published formula by Ionescu-Tirgoviste and Cheta (2) leads to overestimation of the decrease in sodium. The Steffes–Freier formula (3) is based on triglyceride concentration alone; thus, there is potential for error in cases in which the hyperlipidemia is caused predominantly by hypercholesterolemia or mixed hypertriglyceridemia/hypercholesterolemia. In samples from patients with confirmed hypercholesterolemia, electrolytes should be measured with a direct ISE system. The new formulas are simpler than published ones and allow operators to perform the calculations promptly. Thus, laboratories not equipped with computers, direct ISE systems, Airfuges, or lipid-clearing agents can estimate the true electrolyte concentrations. Care must be taken, however, if hyperlipidemia is complicated by hypoproteinemia or hyperproteinemia (4). In our organization, we have adopted a standardized approach to the processing of hyperlipidemic samples. We visually check turbid samples to determine whether it is possible to see through the sample. If not, the sample is analyzed for cholesterol, triglycerides, and lipemic index, and these results are reported irrespective of whether a lipid profile is requested. Triglycerides or cholesterol >20 mmol/L on first presentation are critical for proper patient management and are promptly phoned to the clinical unit. In cases in which the lipemic index is ≥3, the sample is optically turbid, and the total lipid concentration is >25 mmol/L, the processing varies depending on the laboratory. Smaller laboratories analyze the sample on a direct ISE system for Na+, K+, and Cl−, and then forward the remaining sample to the nearest laboratory in the network with the capacity for ultracentrifugation and analysis of the remaining requested test profile. In the larger laboratories equipped with Airfuges, the sample is ultracentrifuged, and all results, including Na+, K+, and Cl−, are reported from the ultracentrifuged sample. The results are reported with a comment indicating that the results were obtained after ultracentrifugation to remove lipids. No lipid-clearing agents are used. In summary, hyperlipidemia caused errors in indirect ISE electrolyte measurements. All 3 electrolytes (Na+, Cl−, and K+) determined by the indirect ISE system were affected, showing artifactual decreases as a result of hyperlipidemia. The Na+ and Cl− were decreased by ∼1 mmol/L and K+ by ∼0.04 mmol/L for each 10-mmol/L increase in total lipid concentration. When direct ISE methods and ultracentrifuges are unavailable to handle severely lipemic samples, corrective formulas can be used. Differences observed with the recommended and our correction formulas for decreases in the indirect ISE electrolytes with increasing total lipid concentration in the second subset of samples. F1, F2, F4, and F6 are published correction formulas. F3, F5, and F7 are correction formulas derived from the first subset in this study. Differences observed with the recommended and our correction formulas for decreases in the indirect ISE electrolytes with increasing total lipid concentration in the second subset of samples. F1, F2, F4, and F6 are published correction formulas. F3, F5, and F7 are correction formulas derived from the first subset in this study.