Abstract First part Unique features and survival analysis of young breast cancer Purpose The median for age at diagnosis is between age 45 and age 49 within breast cancer population in Taiwan, while the breast cancer population in western world show the median for that is at age 62. Such a big shift deserves to be investigated for finding the unique features of breast cancer in a relative young Asian population via epidemiological study. Methods The patients in a cohort for this study were randomly sampled from breast cancer women diagnosed at National Taiwan University Hospital (NTUH) (N=1257) and at two medical centers in middle and south of Taiwan (N=369) from 1990 to 2008. Young breast cancer was defined as patients with diagnosed age less than 50 years old. Patients at age 50 or greater are in a group as the control for this study. The analysis based on the differences in both groups for clinical features, pathological features and life style. We applied Cox Regression Model to analyze the risk factors involved in early onset of breast cancer. The patient survival analysis made for both groups was performed. Results Young breast cancer patients have smaller BMI (22.803.48 vs. 24.374.01, p<0.0001), earlier menarche (13.831.44 vs. 14.781.85 year old, p<0.0001), fewer number of children (p<0.0001), less breast feeding (p<0.0001). They are higher educated (p<0.0001) and give birth at older age (26.664.28 vs. 24.894.55 year old, p<0.0001). The familial tendency is also significant in young breast cancer (p=0.0154). In clinical indices, the young breast cancer characterizes as smaller tumor size (2.742.07 vs. 2.982.23 cm, p=0.0316), ER and PR positivity (p=0.0003 and P<0.0001), less lymphovascular invasion (p=0.0028) and less axillary lymph node metastasis (p=0.0198) as well as aggressive adjuvant treatment. In Cox regression model for the risk factors on early onset of breast cancer are as follows: BMI < 18.5 (HR=1.293); high school level (HR=1.674); early age at menarche (HR=1.603); later age at first child birth (HR= 1.0143; giving birth twice or greater versus zero (HR=0.873-0.430); with family history of breast cancer (HR=1.402). The general survival analysis shows no significant difference between two groups. In subgroup analysis of stage specific survival, the young patient reveals a borderline survival advantage at stage III patients (5 year survival 71.3% vs. 59.8%, p=0.0674). Conclusion Young breast cancer patients have many unique demographic and clinic pathological characteristics. In Cox regression model, smaller BMI, higher education, early menarche, family history are risk factors attributed to early onset of breast cancer. The general survival analysis shows no significant difference between two groups except borderline survival advantage at stage III patients for young patients. Second part The assessment for uprising incidence of breast cancer in Taiwan-an analysis based on age at diagnosis, period and birth cohort model Purpose Breast cancer has become a growing concern in Taiwanese cancer society in the past twenty years. The incidence of such disease has raised four folds between 1980 and 2003. To overall evaluate the major causes, an Age-Period-Cohort model was chosen in this study. Methods A cohort consists of 70,211 breast cancer women according to the records provided by Department of Health, Bureau of Health Promotion in Taiwan. Three aspects: age at diagnosis, period and birth cohort was sub grouped in five years interval. The influence derived from individual variable to the incidence of breast cancer was evaluated. The multivariate analysis was performed using Age-Period-Cohort model to globally evaluate the major causes for increasing incidence of breast cancer. Results From period perspective, the breast cancer incidence increases yearly from 21.81 / 105 population in 1981 to 76.27 in 2005. Regardless of any birth cohort of women, the incidence of breast cancer always increases with age, while the later-born generations of women, the incidence increases much more speedy. The observed peak incidence of breast cancer in about 45-59 years is the summation of period and birth cohort effect. Therefore, the age-specific breast cancer incidence in Taiwan increases by age from the 20-year-old (0.47 / 105 population) until the age of 48. The peak incidence is 91.40 per 105 populations. From the 48 to 88 year old (29.54 / 105 population), the incidence was reduced by years of age. In the Age-period-cohort model analysis: With age group of 20-24 years old as the reference group, the relative risk of breast cancer of all other age groups are much higher from 4.19 times to 60.07 times. From birth cohort viewpoint, the relative risk of breast cancer increases with later cohort. The relative risk of 1966-1959 birth cohort is 3.26 times the risk of 1902-1905 baseline cohort. Conclusion In Taiwan, the overall breast cancer incidence increased by year, because the relative risk of breast cancer increases with later birth cohort. Regardless of any birth cohort of women, the incidence of breast cancer always increases with age, while the later-born generations of women, the incidence increases much more speedy. The observed peak incidence of breast cancer in about 45-59 years is the summation of period and birth cohort effect.