Reply

Abstract

The caveats Ellison and Harris offer seem to boil down to this: Because weight gain is probably not causally related to birth weight in less-than-famished women, and because this lack of relationship has not been demonstrated by intervention studies, then the studies must have been done wrong. We should, it is implied, keep doing the studies, with tighter, more expensive, and impossible-to-implement designs until the “right” answer is obtained. If we fail to do this, objections to results on methodologic grounds will be used to dismiss the findings and to question the efficacy of weight gain and dietary counseling of pregnant women.If we waited for “perfect” data to take action, decision making would be paralyzed and improvements in pregnancy outcome limitedIn reality, limitations inherent to human studies leave room for criticism of the most meticulous study designs. If we are to improve outcomes of pregnancy, action has to be taken on the basis of scientific consensus derived from less-than-perfect data. If we waited for “perfect” data to take action, decision making would be paralyzed and improvements in pregnancy outcome limited. We can’t wait for that.A substantial body of evidence supports the scientific consensus on relationships among weight gain, birth weight, and infant health ((1)National Academy of Sciences.Nutrition During Pregnancy. I. Weight Gain. II. Nutrient Supplements. National Academy Press, Washington, DC1990Google Scholar). The consistency, specificity, strength, and linearity of the relationships between weight gain and birth weight identified in an abundance of studies qualifies weight gain as a causal factor. Relationships between weight gain and birth weight are strengthened when the confounding effects of prepregnancy weight status (or, more likely, when central body fat stores ((2)Brown JE, Potter JD, Jacobs DR Jr, Kopher RA, Rourke MJ, Barosso GM, Hannan PJ, Schmid LA. Maternal waist-to-hip ratio as a predictor of newborn size: results of the Diana Project. Epidemiology. In press.Google Scholar) on birth weight are taken into account. Weight gain in pregnancy is associated with clinically significant differences in birth weight in samples of women in the United States ((3)Goldenberg R.L. Davis R.O. Oliver S.P. Cutter G.R. Hoffman H.J. Dubard M.B. Copper R.L. Maternal risk factors and their influence on fetal anthropometric measurements.Am J Obstet Gynecol. 1993; 168: 1197-1205Google Scholar, (4)Brown J.E. Berdan K.W. Splett P. Robinson M. Harris L.J. Prenatal weight gains related to the birth of healthy-sized infants to low-income women.J Am Diet Assoc. 1986; 86: 1679-1683Google Scholar) as well as in severely malnourished women. In addition, it is clear that dietary interventions to restrict weight gain in pregnancy are successful. The reduced weight gains are associated with substantial reductions in birth weight ((5)King A.G. Free-feeding pregnant women.Am J Obstet Gynecol. 1949; 58: 299-307Google Scholar, (6)Naeye R.L. Blanc W. Paul C. Effects of maternal nutrition on the human fetus.Pediatrics. 1973; 52: 494-503Google Scholar). Low maternal weight gain increases the risk that infants will have neurologic abnormalities ((7)Singer J.E. Westphal M. Niswander K. Relationships of weight gain during pregnancy to birth weight and infant growth and development in the first year of life. A report from the collaborative study of cerebral palsy.Obstet Gynecol. 1968; 31: 17-423Google Scholar). The effect of weight gain on fetal growth appears to precede the third trimester when the greatest amount of fetal weight is gained ((3)Goldenberg R.L. Davis R.O. Oliver S.P. Cutter G.R. Hoffman H.J. Dubard M.B. Copper R.L. Maternal risk factors and their influence on fetal anthropometric measurements.Am J Obstet Gynecol. 1993; 168: 1197-1205Google Scholar).Use of knowledge about weight gain and birth weight is adversely affected by obfuscations such as “net weight gain.” Net weight gain fails to account for the direct effects of prepregnancy weight as obese women tend to have smaller babies and to retain more of the weight gained for themselves. Obese women gaining little or no weight are likely to have sizeable infants and a net weight loss. (The negative net weight shown in Ellison's figure likely represent obese women who gained little weight in pregnancy yet delivered sizeable infants). It is no wonder that the use of net weight gains weakens conclusions about the association between weight gain and birth weight.It is time to end the debate about weight gain per se and to go on to investigate pressing issues about which we know far less, such as the effects of timing of nutrient deprivation during pregnancy on subsequent disease outcomes in offspring. Given the knowledge we have, it would be irresponsible not to implement current recommendations for weight gain and dietary assessment and counseling services for women. The caveats Ellison and Harris offer seem to boil down to this: Because weight gain is probably not causally related to birth weight in less-than-famished women, and because this lack of relationship has not been demonstrated by intervention studies, then the studies must have been done wrong. We should, it is implied, keep doing the studies, with tighter, more expensive, and impossible-to-implement designs until the “right” answer is obtained. If we fail to do this, objections to results on methodologic grounds will be used to dismiss the findings and to question the efficacy of weight gain and dietary counseling of pregnant women.If we waited for “perfect” data to take action, decision making would be paralyzed and improvements in pregnancy outcome limited In reality, limitations inherent to human studies leave room for criticism of the most meticulous study designs. If we are to improve outcomes of pregnancy, action has to be taken on the basis of scientific consensus derived from less-than-perfect data. If we waited for “perfect” data to take action, decision making would be paralyzed and improvements in pregnancy outcome limited. We can’t wait for that. A substantial body of evidence supports the scientific consensus on relationships among weight gain, birth weight, and infant health ((1)National Academy of Sciences.Nutrition During Pregnancy. I. Weight Gain. II. Nutrient Supplements. National Academy Press, Washington, DC1990Google Scholar). The consistency, specificity, strength, and linearity of the relationships between weight gain and birth weight identified in an abundance of studies qualifies weight gain as a causal factor. Relationships between weight gain and birth weight are strengthened when the confounding effects of prepregnancy weight status (or, more likely, when central body fat stores ((2)Brown JE, Potter JD, Jacobs DR Jr, Kopher RA, Rourke MJ, Barosso GM, Hannan PJ, Schmid LA. Maternal waist-to-hip ratio as a predictor of newborn size: results of the Diana Project. Epidemiology. In press.Google Scholar) on birth weight are taken into account. Weight gain in pregnancy is associated with clinically significant differences in birth weight in samples of women in the United States ((3)Goldenberg R.L. Davis R.O. Oliver S.P. Cutter G.R. Hoffman H.J. Dubard M.B. Copper R.L. Maternal risk factors and their influence on fetal anthropometric measurements.Am J Obstet Gynecol. 1993; 168: 1197-1205Google Scholar, (4)Brown J.E. Berdan K.W. Splett P. Robinson M. Harris L.J. Prenatal weight gains related to the birth of healthy-sized infants to low-income women.J Am Diet Assoc. 1986; 86: 1679-1683Google Scholar) as well as in severely malnourished women. In addition, it is clear that dietary interventions to restrict weight gain in pregnancy are successful. The reduced weight gains are associated with substantial reductions in birth weight ((5)King A.G. Free-feeding pregnant women.Am J Obstet Gynecol. 1949; 58: 299-307Google Scholar, (6)Naeye R.L. Blanc W. Paul C. Effects of maternal nutrition on the human fetus.Pediatrics. 1973; 52: 494-503Google Scholar). Low maternal weight gain increases the risk that infants will have neurologic abnormalities ((7)Singer J.E. Westphal M. Niswander K. Relationships of weight gain during pregnancy to birth weight and infant growth and development in the first year of life. A report from the collaborative study of cerebral palsy.Obstet Gynecol. 1968; 31: 17-423Google Scholar). The effect of weight gain on fetal growth appears to precede the third trimester when the greatest amount of fetal weight is gained ((3)Goldenberg R.L. Davis R.O. Oliver S.P. Cutter G.R. Hoffman H.J. Dubard M.B. Copper R.L. Maternal risk factors and their influence on fetal anthropometric measurements.Am J Obstet Gynecol. 1993; 168: 1197-1205Google Scholar). Use of knowledge about weight gain and birth weight is adversely affected by obfuscations such as “net weight gain.” Net weight gain fails to account for the direct effects of prepregnancy weight as obese women tend to have smaller babies and to retain more of the weight gained for themselves. Obese women gaining little or no weight are likely to have sizeable infants and a net weight loss. (The negative net weight shown in Ellison's figure likely represent obese women who gained little weight in pregnancy yet delivered sizeable infants). It is no wonder that the use of net weight gains weakens conclusions about the association between weight gain and birth weight. It is time to end the debate about weight gain per se and to go on to investigate pressing issues about which we know far less, such as the effects of timing of nutrient deprivation during pregnancy on subsequent disease outcomes in offspring. Given the knowledge we have, it would be irresponsible not to implement current recommendations for weight gain and dietary assessment and counseling services for women.