Prospective Cohort Studies Of Stroke Incidence And Prognosis

Stroke is one of the important diseases that threat the population health. Strokeranks to the second leading cause of death and first leading cause of disabilityworldwide. It is characterized with high morbidity, mortality, relapse rate and disabilityrate. Our study is divided into following three parts in order to investigate therelationship between risk factors with stroke incidence and prognosis. Part: Toexplore the independent and cumulative effects of hypertension and drinking, smokingand heart rate on stroke incidence based on a10-year prospective cohort study. Part:To explore the effects of baseline serum25(OH)D level and traditional risk factors onischemic stroke patients prognosis during hospitalization and3months. Part: Wemake a meta-analysis of cohort studies to explore the association between psoriasis andstroke.Part IObjectiveTo explore the independent and cumulative effects of hypertension and drinking,smoking and heart rate on stroke incidence based on a10-year prospective cohort study.Subjects and MethodsWe conducted a survey in32villages in2adjacent townships located in thecounties of Kezuohou Banner and Naiman Banner in Tongliao City of Inner Mongoliabetween2002and2003. A total of2589individuals were included in this study. Writteninformed consent was obtained for all study participants. They were administeredinterview using a standard questionnaire including demographic characteristics, familyhistory of hypertension, smoking and alcohol consumption. The physical examination of blood pressure, height, weight, waist circumference and hip circumference weremeasured. The blood samples were collected and fasting plasma glucose(FPG),triglyceride(TG), total cholesterol(TC), high-density lipoprotein cholesterol(HDL-C)were tested and low-density lipoprotein cholesterol(LDL-C) were also calculatedcorrespondingly. We conducted a follow-up for the baseline2589participants andcollected the stroke incidence information in2008,2009,2010and2012, respectively.We analyzed risk factors of stroke in multivariable Cox regression model.Participants were categorized into four subgroups by blood pressure/drinking status.The cumulative risk of stroke incidence among the four subgroups was estimated usingKaplan–Meier curves and compared by log-rank test. In addition, we used multivariableCox regression models to compute hazard ratios (HRs) of stroke across the other threesubgroups compared with nonhypertension/nondrinkers. We also assessed the predictionvalue of blood pressure/drinking status on stroke incidence by computing the area underreceiver operating characteristic curves (AUC) and comparing a model including onlyconventional risk factors with a model including blood pressure/drinking status inaddition to conventional risk factors. Similarly, Participants were categorized into foursubgroups by smoking status/heart rate. The cumulative risk of ischemic strokeincidence among the four subgroups was estimated using Kaplan–Meier curves andcompared by log-rank test. In addition, we used multivariable Cox regression models tocompute hazard ratios (HRs) of ischemic stroke across the other three subgroupscompared with nonsmokers with heart rate<80. We also assessed the prediction value ofsmoking status/heart rate on ischemic stroke incidence by computing the area underreceiver operating characteristic curves (AUC) and comparing a model including onlyconventional risk factors with a model including smoking status/heart rate in addition toconventional risk factors.ResultsAge increasing, male and hypertension were independent risk factors of strokewhile drinking was not associated with stroke incidence. Cumulative incidence rates ofstroke among non-hypertension/nondrinkers, non-hypertension/drinkers,hypertension/nondrinkers and hypertension/drinkers subgroups were1.5%,2.8%,7.4% and12.5%, respectively (P<0.001). Compared with non-hypertension/nondrinkerssubgroup, the risk (HRs,95%confidential intervals) of stroke innon-hypertension/drinkers, hypertension/nondrinkers and hypertension/drinkerssubgroups were1.02(0.47-2.21)2.61(1.43-4.75) and2.78(1.49-5.21), respectively.Drinkers with hypertension were at a highest risk of stroke. The area under the receiveroperating characteristic curve (AUC) for the model including only the conventional riskfactors was0.660. After adding hypertension and drinking status subgroup, the AUCwas0.684(P=0.005).Smoking was an independent risk factor of ischemic stroke while heart rate wasnot associated with ischemic stroke. Cumulative incidence rates of ischemic strokeamong nonsmokers with heart rate <80, nonsmokers with heart rate80, smokers withheart rate <80and smokers with heart rate80subgroups were1.41%è1.98%è3.97%and5.77%, respectively (P<0.001). Compared with nonsmokers with heart rate <80subgroup, the risk (HRs,95%confidential intervals) of ischemic stroke in nonsmokerswith heart rate80, smokers with heart rate <80and smokers with heart rate80subgroups were1.42(0.62-3.28)2.11(1.06-4.23) and2.86(1.33-6.14), respectively.Smokers with heart rate80bpm were at a highest risk of ischemic stroke. The areaunder the receiver operating characteristic curve (AUC) for the model including onlyconventional risk factors was0.739. After adding smoking status and heart ratesubgroup, the AUC was0.755(P=0.018).ConclusionOur study showed that age increasing, male and hypertension were independentrisk factors of stroke. Smoking was an independent risk factor of ischemic stroke.Drinking probably amplified the effect of hypertension on stroke risk to some degreeand high heart rate probably amplified the effect of smoking on ischemic stroke risk tosome degree. Blood pressure/drinking status was valuable in increasing stroke incidencepredictive efficiency and smoking status/heart rate was also valuable in increasingischemic stroke incidence predictive efficiency. Part IIObjectiveTo explore the effects of baseline serum25(OH)D level and traditional risk factorson ischemic stroke patients prognosis during hospitalization and3-month.Subjects and MethodsOur study population was from “China Antihypertensive Trial in Acute IschemicStroke”. Among them,3002ischemic stroke patients with complete information whowas followed up in3-month and tested baseline serum25(OH)D concentrations wereincluded in this study. We collected demographic information, lifestyle, clinicalcharacteristics, laboratory tests and disease history. Serum25(OH)D concentrations inadmission24hours were tested by automatic chemiluminescent analyzer LIAISON. Weevaluated patients during hospitalization and3-month by modified Rankin Scale andNational Institute of Health stroke scale, recording death, cardiovascular events andstroke recurrence. Death, major disability (MRs3), cardiovascular events and strokerecurrence were defined as the outcome of this study. Multivariable cox model was usedto analyze the association between serum25(OH)D and traditional risk factors withdeath and vascular events during hospitalization and3-month and HR(95%CI) wascalculated. Multivariable logistic model was used to analyze the association betweenserum25(OH)D and traditional risk factors with major disability and compositeoutcome during hospitalization and3-month and OR(95%CI) was calculated. We alsomade a linear trend test between serum25(OH)D, traditional risk factors and compositeoutcome in logistic model. Finally, we made a subgroup analysis to analysis serum25(OH)D and composite outcome in different stratums.ResultsIn hospitalization, compared with the patients less than55years, the HR(95%CI)of death and vascular events among patients older than55years was1.84(0.64-5.32),the ORs(95%CIs) of major disability and composite outcome among patients older than 55years were1.55(1.22-1.98) and1.56(1.23-1.98). Compared with white bloodcell<8.5×109/L patients, white blood cell8.5×109/L patients were at higher risks ofdeath and vascular events2.58(1.09-6.12), major disability1.32(1.06-1.65) andcomposite outcome1.36(1.09-1.69). Compared with blood glucose<7.0mmol/L patients,blood glucose7.0mmol/L patients were at higher risks of death and vascular events3.64(1.56-8.5), major disability1.44(1.16-1.77) and composite outcome1.46(1.18-1.80). Compared with25(OH)D20ng/ml patients, the HR(95%CI) of deathand vascular events among25(OH)D<20ng/ml patients was0.82(0.29-2.34), theORs(95%CIs) of major disability and composite outcome among25(OH)D<20ng/mlpatients were1.09(0.84-1.40) and1.09(0.85-1.40). A dose response relationship wasdetected between age, white blood cell èblood glucose and adverse outcome inhospitalization (P<0.05). No dose response relationship was detected between serum25(OH)D and adverse outcome in hospitalization (P>0.05).In3-month, compared with the patients less than55years, the HR(95%CI) ofdeath and vascular events among patients older than55years was1.72(1.03-2.85), theORs(95%CIs) of major disability and composite outcome among patients older than55years were1.61(1.24-2.10) and1.65(1.28-2.12). Compared with white bloodcell<8.5×109/L patients, white blood cell8.5×109/L patients were at higher risks ofdeath and vascular events2.18(1.54-3.09), major disability1.35(1.07-1.69) andcomposite outcome1.48(1.19-1.83). Compared with blood glucose<7.0mmol/L patients,the HR(95%CI) of death and vascular events among blood glucose7.0mmol/L patientswas1.28(0.90-1.84), the ORs(95%CIs) of major disability and composite outcomeamong blood glucose7.0mmol/L patients were1.30(1.04-1.63) and1.28(1.04-1.59).Compared with systolic blood pressure<160mmHg patients, systolic bloodpressure160mmHg patients were at higher risks of death and vascular events1.53(1.07-2.20), major disability1.33(1.07-1.65) and composite outcome1.33(1.08-1.63). Compared with25(OH)D20ng/ml patients, the HR(95%CI) of deathand vascular events among25(OH)D<20ng/ml patients was1.08(0.69-1.70), theORs(95%CIs) of major disability and composite outcome among25(OH)D<20ng/mlpatients were1.16(0.88-1.52) and1.17(0.90-1.51). A dose response relationship was detected between age, white blood cell èsystolic blood pressure, blood glucose andadverse outcome in3-month (P<0.05). No dose response relationship was detectedbetween serum25(OH)D and adverse outcome in3-month (P>0.05). Subgroup analysisindicated that in male patients, dyslipidemia patients and smoking patients,25(OH)D<20ng/ml patients were at higher risks of adverse outcomes compared with25(OH)D20ng/ml patients. The ORs (95%CIs) were1.49(1.09-2.05),1.65(1.08-2.51)and1.64(1.03-2.61).ConclusionOur study indicated that age increasing, baseline high blood glucose and whiteblood cell made ischemic stroke patients at a higher risk of poor prognosis inhospitalization and3-month. Baseline high systolic blood pressure made ischemicstroke patients at a higher risk of poor prognosis in3-month. A dose response exitedbetween these factors and ischemic stroke prognosis. To all participants, baseline25(OH)D level was not associated with ischemic stroke prognosis in hospitalization and3-month. Subgroup analysis indicated that in male patients, dyslipidemia patients andsmoking patients,25(OH)D deficiency may lead to worse ischemic stroke prognosis in3-month. Part IIIObjectiveTo evaluate the association of psoriasis with stroke by conducting a meta-analysisof cohort studies.Materials and MethodsCohort studies evaluating the association of psoriasis with stroke were searched inMEDLINE (Pubmed), EMBASE and Cochrane Library from their inception to October2013. A random-effects model was used to calculate the overall combined riskestimates. ResultsFive cohorts were finally included in this meta-analysis. Three were prospectivecohorts, one was a retrospective cohort and one was a prospective–retrospective mixedcohort. The pooled RR and95%CI was1.18(1.02-1.37) in the random-effects model.No evidence of publication bias was observed (Begger test: P=1.00Egger test:P=0.778).ConclusionThis meta-analysis of cohort studies suggests that psoriasis signi cantly increasesstroke risk. Psoriasis is likely to be a new risk factor of stroke independent ofconventional cardiovascular risk factors including hypertension and diabetes.