| Colorectal cancer is one of the most common malignant cancers. Durinthe past decades, the incidence of colorectal cancer is increased all around thwor1d. In Jiashan CoUnty the mortality of colorectal cancer is aboL20/l00000 per year Which is the highest among China. The causes ccolorectal cancer are generally regarded as tWo aspects f heredity anenvironment. Heredity accouns for about 20 percent of death fOr colorectfcancer Environmeni factors involve pedicularly dietary habits, such as refish eating and drinking polluted waer All of these are thought to bassociated with colorectal cancerNested Case-Control StUdy (NCCS) is a method of analytica1epidemiological stUdies, a1so named case-control stUdy in cohort. The casesand controls are both se1ected from a Whole cohort (usually called CohortSet). For NCCS, exposure data are collected befOre disease, and the stUdyresults are certain in time consequence. It is suitable to the stUdies ofetiology for chronic diseases, such as cancer.1.Materials and MethodsA co1orectal cancer screeulng prograxn begun at ls, May l989 andended at 30th April l990 was conducted in ten coUlltries of Jiashan countyZhejiang province, China. 64,693 individuals Who were 30 years old or overwere enro1led in the screening prograxn, and then formed the cohort set ofthes stUdy By the Death Registfation System, Cancer Registration Systemand Colorectal Cancer Report System founded by the JiaSha-n coUny thecohort set were followed-uP for about l0 years. Up to 3l,, Dec l998,l96new colorectal cancer cases (l12 colon cancer cases and 84 rectal cancercases) were diagnosed and formed the case grouP of this study TWomethods were used tO se1ect colltrols in this stUdy Firstly, fOr each case, fivenon-colorectal cancer subjects were randomly selected from the cohort setas controls, resulting a final control grouP of 980 individua1s (mode11); then,fOr each case matched with sex, resident location and age, 980 controls (1 t5matched) were selected from the Who1e survival individuals as contro1s(mode12).The stUdy was based on the questiormaire of l989-1990 screeningprogram. The study conten comPosed of general characteristics, personalhabits and syInPtoms a-nd disease history maybe re1ated wtth colorectalcanceT, including sex, age, job, education, dietary habits, drinking watersources, alcohol consumption and cigarette smoking, changes of stool status,abdominal operation history intestina1 disease history asthIna history andallergy history family cancer history psychic stdrilation history and so on.SPSS and SAS wer used to analyze data in thes stUdy Uinvarateanalysis was the first anayzing step fOr both modell and model2. Then,vedables showing associations wtth the risk of colorectal cancer coloncancer a-nd rectal cancer at p<0.20 level were amer tested in forwardstepwise non-conditional and conditional Logistic regression models fOrmodell and model2, respective1y. The final model consisted of thOsevedables showing a significan level at p<0.05. OR vaiue and its 95%CIwere calculated for the factors.2. ResuItsAge is closely associated wtth colorectal canceT, the result showed thatthe incidence of co1orectal cancer increased with age, it reached the max atthe age of 55-60, then drop down with age. The result showed that theaverage diagnosed age of rectal cancer is 58, significanly younger than thatof colon cancer, Which was 62 years, (P=0.0l5). In the univariate analysis ofmodell, compared with the liteYate, the OR value of the i1literate forcolorectal cancer was 1 .766. After matched with age, sex, location in model2,the OR value did not show sighficance anymore. It suggested that theassociation between illiterate and colorectal cancer was confounded by ageand sex.As to the relationship betWeen drithe Waer tyPe and colorectal cancer,the results showed that drinking mixed WaeT, which means that subjectmostly drank river Waer and gutter water...
|
PDF Full Text Request