Maternal smoking, birth weight, infant death, and the self-selection problem

Abstract

Several aspects of the relationship between maternal smoking and birth weights of infants are discussed. No satisfactory explanation for Yerushalmy's results has been given other than that low birth weight appears to relate more to the smoker than to the smoking. Recent studies by Silverman support this position. The possibility that nicotine may induce a physiologic response that serves to alleviate bioenergetic deficiency in some individuals should not be overlooked. In this view, both smoking and low birth weight are symptoms of deficient maternal bioenergetic systems.

PIP: The literature is reviewed in terms of 4 questions of importance to the issue of maternal smoking--infant birth weight and health studies: 1) inference of causality from correlation; 2) the constitutional hypothesis; 3) maternal and fetal constitutional factors associated with low birth weight and perinatal death; and 4) environmental factors associated with low birth weight and perinatal death. Only a few maternal smoking-infant health studies used the invalid method of inference of causality from correlation. There is an underlying biological rationale supporting the constitutional hypothesis that appears to explain lower mean birth weights of children of women who smoke, compared with birth weights of children of nonsmokers. Examination of the National Institute of Neurological Diseases and Stroke findings shows the following maternal constitutional factors, on the average, among those delivered of infants with low birth weights: low prepregnant weight, shortness of stature, and, tentatively, hyperthyroidism. Maternal factors associated with elevated perinatal mortality rates include diabetes mellitus, higher prepregnant weight ranges and taller maternal heights, advanced maternal age, low maternal age, birth weight of less than 2501 grams for last prior child, and male sex of infant. Environmental factors associated with low birth weight and detrimental health characteristics include elevated levels of carbon monoxide and ionization radiation, low socioeconomic index of the family, alcohol consumption, and high altitude. As yet there has not been a definitive analysis of the relationship between maternal smoking and low birth weight and perinatal death that simultaneously considers constitutional factors such as maternal height and weight and other factors such as parity and birthweight of less than 2500 grams for a preceding child.