| Background and purposePrevalence rate of pancreatic cancer is shown rising worldwile. Its incidence rate has increased 3-fold in the recent 30 years in England,the United States and other west countries, and is the 4th leading cause of death from cancer in the United States and European, and also is the second biggest cause of death from tumor related to gastrointestinal tract, holding 20% death of tumor. With the rising of living standard and alteration of diet structure in our country, prevalence rate of pancreatic cancer has largely increased in the recent 30 years in the city, and become the fastest rised cause of death from tumor in the recent 10 years. Prognosis is extremely poor since diagnosis is usually late due to lack of symptoms. Median survival time, regardless of therapy, is 2-3 months after diagnosis. The 1-year survival rate is about 8% and 5-year survival rate only 3%. It is likely that in future decades, pancreatic cancer will continue to be a frequent cause of cancer mortality. In view of the high mortality for pancreatic cancer, an important stategy for reducing the burden of this cancer is to understand and, if possible, reduce exposure to known risk factors and increase exposure to known protective factors, which is to say primary prevention. This is a frequency matched case-control study based on hospital. Case group was 1031 cases totally, and was selected from in-patients of Han confirmed pancreatic cancer of Changhai, Changzheng and Dongfanggandang hospital from 6,1993 to 6,2001.Control group was 1299 cases totally, 200 of which was provided to select, and was randomly selected from non-neoplastic and non- endocrine in-patients matched with resident hospital, admission date, race, sex, resident area. We investigate cases and controls through patient himself and /or relatives known well his condition by questionnaire of unification, which mainly included resident area, smoking, alcohol drinking, medical history, surgical history, neoplastic and diabetes mellitus familial history, chest X-ray, abdominal B-type ultrasound,CT, ERCP, EUS, MRCP, MRI, pathological check, ABO blood groups, HB, PLT, AKP, GGT, ALB, CA-19-9, neoplastic area and body height, body weight, survival period, Helicobacter Pylori (HP) infection, female reproductive history of some patients. All sorts of factors were performed single factor analysis at first (to perform chi-square or t-test according to data type respectively and Cox regression on survival data), which was significant was entered to multiple factor analysis. The relationship between pancreatic cancer and all sorts of factors was suspected by odds ratio (OR), and the value of OR and its 95% confidence interval (CI) were tested by chi-square determination and Mantel-Haenszel test. OR=ad/bc, 95%CI= OR(1±1.96/), the chi-square of fourfold table was calculated by SAS software. The dosage-response relation was analyzed by linearity tendency test and continuous variance by t-test. Demixing analysis performed the factor whose number of exposure factors was few. If the difference of OR after delaminating was not so large, its total ORMH was equal to ∑(aidi/ti)/ ∑(bici/ti), and its totalχ2MH was equal to [∑ai-∑E(ai)]2/∑V(ai), of which ∑E(ai)= ∑m1in1I/ti, ∑V(ai)= ∑{m1imoin1inoi/[ti2(ti-1)]}. χ2 tendency test of grade data was equal to [T1-(n1T1/n)]2/V,of which V=n1n2(nT3-T22)/n2(n-1),T1=∑aiXi,T2=∑miXi,T3=∑miXi2, Xi=i。Results1． Life and diet habits: smoking, heavy alcohol drinking, BMI in male were associated with occurrence of pancreatic cancer, of which smoking was an independent related factor. The ORs and 95%CIs of smoking and alcohol drinking in male were 2.3(1.8-3.1) and 1.7(1.3-2.3) respectively. The p value of smoking in male by multiple factor analysis was 0.007.2． Blood group: group A and group B in male were associated with occurrence of pancreatic cancer, and theirs ORs and 95%CIs were 1.29(1.01-1.64) and 0.74(0.57-0.96) respectively.3． Surgical history: partial gastrectomy, cholecystectomy, appendectomy w...
|
PDF Full Text Request