| Objective: The purpose is using stepwise regression analysis method to investigate the effect of gravida to the fetus. In the many investigated agents, find out the significant to the fetus' growth and development. To offer data for health service in the perinatal stage, to formulate corresponding intervention study and the scientific accordance for the growth and diseases prevention of the fetus. Method: Draw out at random 77 gravies from the 4th hospital affiliated to HeBei Medical University . use the method of simple random sampling. Investigate personal history(X1～X6), menstrual history (X7～X11), marriage and foster history(X12～X14), past history(X15～X16), family history(X17),Look at Table 3. And record B ultrasound examination results, Look at Table 1, tracking and following-up meanwhile. Measure body length , body weight, head circumference, chest circumference, abdomen circumference, femur length, feet length in the six hours as soon as the neonates have been born . Look at Table 2. And proceed datadealing with SAS 6.12, SPSS 10.0 statistics software. The following are the analytical methods:1 Simple linear correlation analysis: Do linear correlation analysis to every two variables of all B ultrasound examination results(Y1～Y21). Do linear correlation analysis to every two variables of all neonate's markers(L1～L7). To screen the variables which are proceeded principal component analysis. 2 Principal component analysis: To proceed principal component analysis to every markers of B ultrasound examination markers(Y1～Y10, Y13～Y15, Y17～Y21)and every markers of neonate(L1～L7) separately, get the first principal component, second principal component. To offer variables for stepwise regression analysis.3 Stepwise regression analysis: To investigate the relative agents about gravida with the first principal component, the second principal component of B ultrasound examination, and the first principal component, the second principal component of neonate's markers. Regard F=3.98 as the variable limit value to be selected in and regard F=2.90 as the variable limit value to be refused. Use the investigated markers of gravida as independent variables , and use Z1, Z2, H1, H2 as the dependent variables. Found stepwise regression equation. Screen the agents which have significant effect on the dependent variables. Results: 1 Linear correlation analysis: Except the markers of embryocardia Y11 , fetus pneusis Y12 , amniotic fluid Y16 , between every two markers of the rest are assumed to be linear correlation.(Umbilical fluctuation exponent Y17 , umbilical resistant value Y18 , umbilical systole topmost bloodstream rate/relaxing period minimum rate Y21, these three markers are assumed to be positive linear correlation.) Look at Table 4. Between every two markers of all neonate's markers are assumed to be positive linear correlation, look at Table 5.2 Principal component analysis shows: The eigenvalue of Z1 is 11.3453, the dedication rate is 63.03%, so Z1 is the comprehensive marker which reflects the growth and the development of fetus. The eigenvalue of Z2 is 2.6741,the dedication rate is 14.86%, so Z2 is the markers which reflects the fetus's existents environment in the uterus. The eigenvalue of H1 is 3.7755, the dedication rate is 53.94%, so H1 is the comprehensive marker which reflects the development degree of neonate. The eigenvalue of H2 is 0.8578, the dedication rate is 12.25%, so H2 is the marker which reflects symonetry degree of neonate's development.3 Stepwise regression analysis forward method: Use F=3.87 as the variable limited value which can be selected in, and use F=2.72 as the variable limited value which can be refused. Gestational age X2, painful menstruation X11, civilization degree X5 and mother's weight X4 enter regression equationof Z1. Look at Table 10. Body length X3, gravidity X12 and menarche age X7 enter regression equation of H1. Look at Table 11. Conclusion: The agent of gestational age...
|
PDF Full Text Request