A doctoral dissertation by a Professor at the Catholic University of Valencia "San Vicente Mártir" (UCV) finds that every cigarette a mother smokes a day during the third quarter of pregnancy reduces the baby's birth weight in 20 grams. Professor Rafael Vila, Faculty of Nursing, received his PhD with the distinction Cum Laude for his dissertation entitled "Anthropometric, haematological, obstetric and toxic variables that influence birth weight: a predictive model".
Rafael Vila, who also works for the Health Department of La Ribera (Alzira-Valencia), Carlet health centre and La Ribera Hospital (Alzira-Valencia), studied 140 pregnant women. "If a mother smokes for example between 5 and 10 cigarettes a day from the 25th week of pregnancy, the child's weight may be reduced between 100 and 200 grams," assured the Professor. This would be worrying in cases where the baby could be affected by underweight.
However if a mother stops smoking during the first or second quarter, "the foetus is not affected by weight loss." For this reason he stressed "the importance of mothers not to smoke during pregnancy, especially at the end."
This data comes from a new predictive model designed by Vila that includes multiple variants and for the normal-weight category improves the accuracy of ultrasound scanning in predicting birth weight by an average of 10'48 percent.
The projected weight at 40 weeks calculated in the third trimester scan overestimates birth weight in 300'89 grams, "which is an average error of 10.75 percent" he said. Meanwhile, his predictive model only has an average error of 0.54 percent.
The multivariate regression model, as proposed by Vila, uses the routine ultrasound that is performed in the third quarter as a basis and adds other variables besides smoking, such as gestational age, fundal height and serum ferritin levels. According to the dissertation, the fundal height is the variable which is most related with accuracy in predicting birth weight, even surpassing ultrasound scanning.
Vila said that increasingly the predictions are solely based on ultrasound scans in the third quarter of pregnancy when it is known to have a precision error of around 11%. This accuracy can be improved by adding the other variables, especially fundal height. In this sense, Vila's predictive model "is feasible, it is not time-consuming and does not require sophisticated resources, since the data is already collected by the midwife at the health centre."
