Abstract
A point-of-care (POC) device to measure mouse glucose and lipid profiles is an important unmet need for cost-effective, immediate decision making in research. We compared metabolic analyte profiles obtained using a human clinical POC device with those from a veterinary laboratory chemical analyzer (LCA). Unfasted terminal blood samples were obtained by cardiac puncture from C57Bl/6J mice used in a diet-induced obesity model of type 2 diabetes mellitus; age-matched C57Bl/6J controls; a transgenic mouse model of Alzheimer's disease on a C57BL/6J background (16 wk old); and aged C57BL/6J mice (24 to 60 wk old). Aliquots of the blood were immediately assayed onsite using the POC device. Corresponding serum aliquots were sent analyzed by LCA. Measures from the POC and LCA devices were compared by using the Bland-Altman and Passing-Bablok methods. Of a total of 40 aliquots, LCA results were within reported reference ranges for each model. POC results that fell beyond the device range were excluded from the analyses. The coefficient of determination and Passing-Bablok analysis demonstrated that POC glucose and HDL had the best agreement with LCA. The Bland-Altman analysis found no value-dependent bias in glucose and no significant bias in HDL. The remaining lipid analytes (cholesterol and triglyceride) showed significant bias. Until an improved, validated mouse POC device with lipid profile capability is available, the POC device that we tested appears adequate for screening glucose and HDL in mouse blood. Disadvantages of this clinical POC device are the narrow human ranges relative to ranges found in mice and its limited precision as compared with the LCA. This study demonstrates that when the samples are within the device range limits, this human POC device can accurately track metabolic syndrome and be used to compare patterns in glucose and HDL.