Analysis Of Mineral Elements In Plasma Of Diabetes Mellitus Patients And Establishment Of The Discrimination Model

Objective To establish a simple, accurate and effective method for detection ofmineral elements in plasma by inductively coupled plasma mass spectrometry; and toexplore the relationship between plasma elements with diabetics and to establish thediscrimination model of diabetics by using partial least squares discriminant analysis.Methods Using inductively coupled plasma mass spectrometry and atomicabsorption spectrometry to measure the content of aluminum(Al), vanadium(V), chromium(Cr), manganese(Mn), iron(Fe), cobalt(Co), Nickel(Ni), copper(Cu), zinc(Zn), arsenic(As),selenium(Se), rubidium(Rb), strontium(Sr), molybdenum(Mo), cadmium(Cd), antimony(Sb), barium(Ba), lead(Pb), sodium(Na), magnesium(Mg), potassium(K), calcium(Ca) intotal22mineral elements in plasma in200diabetic patients and200age and sex matchedhealthy controls. SAS9.2statistical software was performed for data description andstatistical analysis. Paired sample Wilcoxon signed rank test was applied to compare thedifferences of those elements between diabetes mellitus and healthy controls, using theSpearman rank correlation analysis to test the correlation of the elements with fastingplasma glucose, partial least squares discriminant analysis was applied to establish thediscriminant model of diabetes mellitus, SIMCA-P11.5software was performed to drawrelevant results of PLS-DA. Results The detection limits of inductively coupled plasma mass spectrometry forelements in the plasma was0.001μg/L(Sb)-0.159μg/L(Zn), recoveries of elements wasbetween86%-108%, the relative standard deviation of short-term stability was1.10%-5.64%, and the relative standard deviation of long-term stability was2.03%-7.15%.The measured value of all the elements was within the standard range excepted V. Thismethod had a high sensitivity, precision, accuracy, and consumed less plasma, thepre-treatment was simple, the measuration was relatively easy and effective. Wilcoxonsigned rank test showed none of those22elements between diabetic which had normalplasma glucose value and healthy controls was statistical differences. However those had aabnormal plasma glucose diabetics displayed a significantly higher content of Cr, Mn, Ni,Zn, Se, Rb, Sb, Ba, Ca than healthy controls, while Mg was lower. Spearman rankcorrelation analysis revealed that Mn, Ni, Zn, As, Se, Rb, Sb, Ba, Pb, Ca was positivelycorrelated with glucose, Mg, Mo was negatively correlated with glucose. Using PLS toextracted factors, two extracted factors showed the most fitted model. The PLS-DA scoresplots of the two components drawed by SIMCA-P11.5software showed a good separationof diabetics and healthy controls. Then added fisher linear discriminant to establishdiscrimination model, using jackknife and prospective evaluation to estimate the efficiencyof the discrimination model, the correct rate of the model was88.50%and87.50%,respectively, the sensitivity of prospective evaluation was85.00%, specificity was90.00%.The figure of the variable importance for projection uncovered that Mg, Ca, Ni, V, Ba, Mn,Rb, Pb, Sb, As had a relatively large contribution to the model.Conclusion This study creates a simple, rapid and accurate measuring method byplasma inductively coupled plasma mass spectrometer, the method is suited forepidemiological researches with a large sample size. The difference of elements metabolismbetween healthy person and diabetics with regular plasma glucose isn’t significant, whilethose who have abnormal plasma glucose show metabolic disorder, the discriminationmodel builded by PLS-DA reveals a ideal discrimination efficiency.