Association Between Ischemic Stroke And Serum Amyloid A Protein

ObjectiveThe purpose was to explore the association of serum amyloid Aprotein(SAA) and risk of ischemic stroke (IS) with its various subtypes, severity,and infarct volume. SAA and high-sensitivity C-reative protein (hsCRP) levelswere united to predict risk of IS. It was aim to evaluate clinical value of SAAabout predicting IS with its various subtypes, severity, infarct volume and hsCRP.The purpose was in order to provide the reliable clinical research conclusionwhich could be used for early prevention and treatment of IS.MethodsAccording to inclusion and exclusion standards to strictly collect data,case-control study was used to continuously collected130IS patients in casegroup and145healthy people in control group without history of related vasculardiseases which were collected at some time with matching community and age.All subjects were recorded demographic characteristics, IS related risk factors,SAA related factors, serum hsCRP level and so on.According to2007modified The Trial of Org10172in Acute StrokeTreatment（TOAST）criteria, case group was divided in79atherothrombosisstroke(AT),20cardioembolism (CE),26small artery disease (SAD) and5Stroke of Undetermined Etiology (SUD)；According to The National Institutes of Health Stroke Scale (NIHSS)，ATpatients were divided in mild group22(＜4points), medium group30(4~15points), severe group27patients (＞15points)； According to Magnatic Resonance Imaging（MRI）or computed tomography（CT）of the head imagingand following formula length×width×number of layers, AT patients were dividedin small infarct group20(volume＜4cm~3), middle infarct group35（volume4~16cm~3） and large infarct group24(volume＞16cm~3).The serum SAA and hsCRP levels were respectively detected byenzyme-linked immunosorbent assay (ELISA) and clincal routineimmunoturbidimetry. Univariate and multiple analyses were used to evaluate theassociation of SAA level and risk of IS with its various subtypes and infarctvolume which only used univariate analysis. Then Logistic regression of multipleanalysis was used to analyse the association between SAA level and NIHSSand of hsCRP level with risk of IS. It was further to unite SAA and hsCRP levelsto predict risk of IS by Logistic regression. Spearman analysis was used toevaluate the association between SAA level and NIHSS, infarct volume, hsCRPlevel.Results1, The case group SAA level was higher than control group(10.15±5.21ng/mL vs8.83±3.82ng/mL), and the difference was statisticallysignificant(P=0.017); The highest quartile group(＞10.77ng/mL) SAA level hadincreased the risk of IS (OR=1.83,95%CI1.13～2.18) as compared with lowestquartile group（≤6.33ng/mL）, and the difference was statistically significant(P=0.025). Then after adjustment of related risk factors, including gender,history of hypertension, diabetes mellitus, smoking, low density lipoprotein（LDL）and history of family stroke the difference was still statistically significant(OR=1.72,95%CI1.23～2.61, P=0.046）.2, Subtype analysis: AT group SAA level was higher than control group，thedifference was statistically significant (OR=1.62,95%CI1.05～1.89, P=0.029),and after adjustment of related risk factors, including gender, history ofhypertension,diabetes mellitus, smoking, LDL and history of family stroke the difference was still statistically significant (OR=1.43,95%CI1.01~1.71,P=0.034). SAA level had no associations with CE or SAD.3, SAA level of the severe group of NIHSS was both higher than mild andmedium groups, and the difference were both statistically significant (both P＜0.05); SAA level and NIHSS were positive correlation (r=0.534, Spearmancorrelation); In NIHSS＞15points AT patients the highest quartile group SAAlevel had increased the severity of IS (OR=2.13,95%CI1.34～3.69) ascompared with the lowest quartile group, and the difference was statisticallysignificant(P=0.043). SAA level and NIHSS＞15points were significantlypositive correlation (r=0.782, Spearman correlation).4, To respectively compare SAA level of large infarct group with smallinfarct group and middle infarct group, the differences all had no statisticssense(all P＞0.05). SAA level and infarct volume had positive correlation(r=0.452, Spearman correlation). SAA level and hsCRP level had positivecorrelation (r=0.481, Spearman correlation).5, The case group hsCRP level was higher than control group(6.07±3.40mg/L vs3.38±1.87mg/L), and the difference was statisticallysignificant(P=0.002); After adjustment of related risk factors, including gender,history of hypertension, diabetes mellitus, smoking, LDL and history of familystroke hsCRP level＞10mg/L group risk of IS was3.79times than＜3mg/Lgroup (95%CI1.10～13.05, P=0.035).6, SAA level＞10.77ng/mL and hsCRP level＞10mg/L group risk of IS was10.75times than SAA level≤10.77ng/mL and hsCRP level＜3mg/L group（95%CI3.50-19.04, P=0.000）.Conclusions1, SAA and hsCRP both have associations with IS;.2, SAA can be used to identify AT in IS patients;3, The higher SAA level, the more severe IS in AT patients; 4, At present，SAA level can not predict the infarct volume;.5, SAA level and NIHSS, hsCRP level had positive correlations;6, The unity SAA level and hsCRP level can exactly predict the risk of ISwhich can have complementary effect.