The sample size was estimated according to the statistical analysis of logistic regression. Considering that there were altogether 7 variables to be examined, based on the guideline of sample size for logistical regression with a ratio of 30 observations for each factor, the sample size was (30*7+50%*30*7), considering a 50% refusal rate that equaled 315 possible participants. Systematic random sampling method was used to select participants among all the first -time Chiness mother at 4 month postpartum. The list of all eligible participants was obtained from the hospital. In order to achieve the sample size of 315, they were selected by taking every 4 person from about 1600 eligible mother during the month surveyed. The first participant was chosen by a simple sampling method using a randomized figure table on the Internet. Two hundred and seventy two questionnaires were effectivery completed from four hundred delivered questionnaires were used for data analysis.