- Email: [email protected]
Modifiable risk factors for growth restriction in twin pregnancies
American Journal of Obstetrics and Gynecology (2005) 192, 1440–2
www.ajog.org
Modifiable risk factors for growth restriction in twin pregnancies Wade D. Schwendemann, MD,a John M. O’Brien, MD,b,* John R. Barton, MD,b Douglas A. Milligan, MD,b Niki Istwan, RNc Department of Obstetrics and Gynecology, University of Kentucky,a Division of Maternal-Fetal Medicine, Central Baptist Hospital,b Lexington, Ky, and Matria Healthcare, Marietta, Gac Received for publication September 2, 2004; revised November 29, 2004; accepted December 20, 2004
KEY WORDS Twin pregnancy Growth restriction Tobacco abuse
Objective: This study was undertaken to evaluate modifiable risk factors for adverse fetal growth in twin pregnancies. Study design: A large cohort study from a database of women with twin gestations identified at risk for preterm labor was performed. Examining each infant’s birth weight and gestational age at delivery, infants were classified as being average (AGA), large (LGA), or small (SGA) for gestational age, using the Alexander reference curve. Clinical and demographic factors were compared between patients delivering at least 1 SGA infant and AGA pairs using Pearson’s c2 Student t test statistics and logistic regression. Results: There were 11,827 twin pregnancies evaluated. Risk factors associated with SGA deliveries included tobacco abuse, poor weight gain, lean prepregnancy body mass index, African American race, and nonmarried. The logistic regression identified tobacco abuse as the single greatest risk for poor fetal growth, (odds ratio [OR] 1.95; 95% CI [1.68, 2.27]). Weight gain of less than one-half lb/wk also increased SGA risk (OR 1.35; 95% CI [1.16, 1.68]), whereas weight gain greater than 1 lb/wk decreased SGA risk (OR 0.77; 95% CI [0.68, 0.86]). Conclusion: Tobacco abuse and weight gain are the modifiable risk factors, which require intervention during a twin pregnancy. Patients should be encouraged to stop tobacco abuse and gain a minimum of one-half lb/wk in the later half of pregnancy to minimize the risk for growth restriction. Ó 2005 Elsevier Inc. All rights reserved.
Fetal growth restriction (FGR) is a significant cause of perinatal morbidity and mortality. Identification of maternal risk factors associated with FGR and small-
* Reprint requests: John M. O’Brien, MD, Perinatal Diagnostic Center, Central Baptist Hospital, 1740 Nicholasville Rd, Lexington, KY 40503. E-mail: [email protected] 0002-9378/$ - see front matter Ó 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.ajog.2004.12.071
for-gestational-age (SGA) infants provides the clinician more information to counsel patients and optimize their perinatal management. Although it has been generally assumed that the risk factors for FGR in twin pregnancies would be the same as those in singleton pregnancies, there is limited literature on the subject. The importance of fetal growth curves for the monitoring and clinical management of complicated pregnancies has long been recognized. In
Schwendemann et al assessing for FGR for twins, it is important to use plurality specific growth curves.1 We sought to determine what modifiable risk factors existed in a population of women who had at least 1 twin with FGR using appropriate growth curves.
Material and methods The population for this large cohort study was identified from a database containing clinical information on high-risk pregnant women receiving outpatient services for preterm labor from Matria Healthcare between December 1988 and December 2002. Women meeting the following criteria were included: twin gestation, enrolled in preterm labor surveillance program, documented maternal prepregnancy weight and weight at start of outpatient surveillance, and documented pregnancy outcome. Most patients enrolled into the program were symptomatic with excess uterine activity and therefore do represent a selected population of twins. Infants were classified as being average (AGA), large (LGA), or small (SGA) for gestational age using the Alexander reference curve for twin gestations. Women delivering AGA/AGA pairs (AGA/AGA group, n = 10,172) were compared with women with SGA/ SGA (n = 250) or SGA/AGA pairs (n = 1405). (At least 1 SGA group, n = 1655.) Women with any LGA infant were excluded from all analyses. Maternal weight gain was defined as the average per week over the observation period in the program. Clinical and demographic factors were compared between the patients delivering at least one SGA infant and AGA pairs with the use of Pearson’s c2 and Student t test statistics. Logistic regression was performed to assess multiple effects of significant modifiable factors.
Results Data for 11,827 twin pregnancies were evaluated. Table I demonstrates that the demographic characteristics of the women who did versus those who did not have at least 1 growth-restricted infant. The women in the study did not deliver at statistically different times and had a similar percentage of their deliveries before 35 weeks. Multiple risk factors were identified for delivering a SGA newborn including race, tobacco abuse, prepregnancy body mass index, marital status, and weight gain. A logistic regression analysis was performed on these modifiable risk factors as demonstrated in Table II. The risk factors with highest odds ratios, poor maternal weight gain, and tobacco abuse, were also found to be cumulative. Patients having both risk factors present had the highest incidence of at least one SGA infant at
1441 Table I
Maternal characteristics AGA/AGA (n=10,172)
Maternal age Black race Married Tobacco user* Lean prepregnancy BMI* Wt gain !1/2 lb/wk* Wt gain O1 lb/wk* GA at starty
At least 1 SGA infant (n=1,655)
29.3 G 5.5 29.3 G 5.9 13.3% 17.2% 82.9% 79.2% 8.6% 15.5% 20.7% 23.3 % 15.5% 19.9% 38.9% 32.2% 27.7 G 4.0 27.5 G 4.2
P value .699 ! .001 ! .001 ! .001 .021 ! .001 ! .001 .138
Data presented as mean G SD, or percentage, as indicated. * Modifiable factor. y Start of outpatient surveillance.
Table II
Logistic regression for modifiable risk factors P value
Prepregnancy BMI Wt gain !1/2 lb/wk Wt gain O1 lb/wk Tobacco user
! ! ! !
.001 .001 .001 .001
OR (95% CI) 1.03 1.35 0.77 1.95
(1.02-1.04) (1.16-1.56) (0.68-0.86) (1.68-2.27)
23.7%. Nontobacco users with poor maternal weight gain had an incidence of SGA of 18.5%.
Comment FGR is associated with increases in both short-term and long-term morbidity, as well as neonatal mortality. The mortality rate of SGA newborns has been shown to be 5 to 20 times greater in infants born between 1500 and 2500 g, and from 70 to 100 times higher in those infants born weighing less than 1500 g when compared with AGA infants.2 Short-term morbidities are also increased with FGR, including an increase in meconium aspiration, hypoglycemia, and birth asphyxia. Hypothermia, secondary to decreased fat stores, can lead to metabolic deterioration if not recognized appropriately. Finally, more recent literature has contradicted a previously held belief that infants with FGR had a lower incidence of respiratory distress syndrome.3 Long-term morbidities associated with FGR are still being identified. Infants born FGR, particularly those less than 2300 g, have a lag in mental development when compared with appropriately grown controls.4 Still other data have demonstrated a link between those infants born growth restricted and the risk of developing significant adult diseases, including cardiovascular disease, type 2 diabetes mellitus, and hyperlipidemia.
1442 Several different mechanisms have been proposed for these findings. Twin gestations are at a significantly higher risk for FGR than singleton gestations. Although our study involved a selected population with increased uterine activity, we believe these findings are generalizable because of our sample size. Although there are several risk factors, both maternal and fetal, that are outside the scope of influence of the physician, our study demonstrates there may be important risk factors for FGR that the clinician can work to eliminate.5 Based on our data, both tobacco abuse and poor maternal weight gain in the later half of pregnancy are behaviors that could elicit counseling for a patient with multiples and such an intervention may potentially improve infant outcomes.
Schwendemann et al
References 1. Alexander GR, Kogan M, Martin J, Papiernik E. What are the fetal growth patterns of singletons, twins, and triplets in the United States? Clin Obstet Gynecol 1998;41:115-25. 2. Williams RL, Creasey RK, Cunningham GC, Hawes WE, Norris FD, Tashiro M. Fetal growth and perinatal viability in California. Obstet Gynecol 1982;59:624-32. 3. Thompson PJ, Greenough A, Gamsu HR, Nicolaides KH. Ventilatory requirements for respiratory distress syndrome in small-forgestational-age infants. Eur J Pediatr 1992;151:528-31. 4. Low JA, Galbraith RS, Muir D, Killen H, Karchmar J, Campbell D. Intrauterine growth retardation: a preliminary report of longterm morbidity. Am J Obstet Gynecol 1978;130:534-45. 5. Kinzler WL, Aranth CV, Vintzrelos AM. Medical and economic effects of twin gestations. J Soc Gynecol Investig 2000; 7:321-7.
www.ajog.org
Modifiable risk factors for growth restriction in twin pregnancies Wade D. Schwendemann, MD,a John M. O’Brien, MD,b,* John R. Barton, MD,b Douglas A. Milligan, MD,b Niki Istwan, RNc Department of Obstetrics and Gynecology, University of Kentucky,a Division of Maternal-Fetal Medicine, Central Baptist Hospital,b Lexington, Ky, and Matria Healthcare, Marietta, Gac Received for publication September 2, 2004; revised November 29, 2004; accepted December 20, 2004
KEY WORDS Twin pregnancy Growth restriction Tobacco abuse
Objective: This study was undertaken to evaluate modifiable risk factors for adverse fetal growth in twin pregnancies. Study design: A large cohort study from a database of women with twin gestations identified at risk for preterm labor was performed. Examining each infant’s birth weight and gestational age at delivery, infants were classified as being average (AGA), large (LGA), or small (SGA) for gestational age, using the Alexander reference curve. Clinical and demographic factors were compared between patients delivering at least 1 SGA infant and AGA pairs using Pearson’s c2 Student t test statistics and logistic regression. Results: There were 11,827 twin pregnancies evaluated. Risk factors associated with SGA deliveries included tobacco abuse, poor weight gain, lean prepregnancy body mass index, African American race, and nonmarried. The logistic regression identified tobacco abuse as the single greatest risk for poor fetal growth, (odds ratio [OR] 1.95; 95% CI [1.68, 2.27]). Weight gain of less than one-half lb/wk also increased SGA risk (OR 1.35; 95% CI [1.16, 1.68]), whereas weight gain greater than 1 lb/wk decreased SGA risk (OR 0.77; 95% CI [0.68, 0.86]). Conclusion: Tobacco abuse and weight gain are the modifiable risk factors, which require intervention during a twin pregnancy. Patients should be encouraged to stop tobacco abuse and gain a minimum of one-half lb/wk in the later half of pregnancy to minimize the risk for growth restriction. Ó 2005 Elsevier Inc. All rights reserved.
Fetal growth restriction (FGR) is a significant cause of perinatal morbidity and mortality. Identification of maternal risk factors associated with FGR and small-
* Reprint requests: John M. O’Brien, MD, Perinatal Diagnostic Center, Central Baptist Hospital, 1740 Nicholasville Rd, Lexington, KY 40503. E-mail: [email protected] 0002-9378/$ - see front matter Ó 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.ajog.2004.12.071
for-gestational-age (SGA) infants provides the clinician more information to counsel patients and optimize their perinatal management. Although it has been generally assumed that the risk factors for FGR in twin pregnancies would be the same as those in singleton pregnancies, there is limited literature on the subject. The importance of fetal growth curves for the monitoring and clinical management of complicated pregnancies has long been recognized. In
Schwendemann et al assessing for FGR for twins, it is important to use plurality specific growth curves.1 We sought to determine what modifiable risk factors existed in a population of women who had at least 1 twin with FGR using appropriate growth curves.
Material and methods The population for this large cohort study was identified from a database containing clinical information on high-risk pregnant women receiving outpatient services for preterm labor from Matria Healthcare between December 1988 and December 2002. Women meeting the following criteria were included: twin gestation, enrolled in preterm labor surveillance program, documented maternal prepregnancy weight and weight at start of outpatient surveillance, and documented pregnancy outcome. Most patients enrolled into the program were symptomatic with excess uterine activity and therefore do represent a selected population of twins. Infants were classified as being average (AGA), large (LGA), or small (SGA) for gestational age using the Alexander reference curve for twin gestations. Women delivering AGA/AGA pairs (AGA/AGA group, n = 10,172) were compared with women with SGA/ SGA (n = 250) or SGA/AGA pairs (n = 1405). (At least 1 SGA group, n = 1655.) Women with any LGA infant were excluded from all analyses. Maternal weight gain was defined as the average per week over the observation period in the program. Clinical and demographic factors were compared between the patients delivering at least one SGA infant and AGA pairs with the use of Pearson’s c2 and Student t test statistics. Logistic regression was performed to assess multiple effects of significant modifiable factors.
Results Data for 11,827 twin pregnancies were evaluated. Table I demonstrates that the demographic characteristics of the women who did versus those who did not have at least 1 growth-restricted infant. The women in the study did not deliver at statistically different times and had a similar percentage of their deliveries before 35 weeks. Multiple risk factors were identified for delivering a SGA newborn including race, tobacco abuse, prepregnancy body mass index, marital status, and weight gain. A logistic regression analysis was performed on these modifiable risk factors as demonstrated in Table II. The risk factors with highest odds ratios, poor maternal weight gain, and tobacco abuse, were also found to be cumulative. Patients having both risk factors present had the highest incidence of at least one SGA infant at
1441 Table I
Maternal characteristics AGA/AGA (n=10,172)
Maternal age Black race Married Tobacco user* Lean prepregnancy BMI* Wt gain !1/2 lb/wk* Wt gain O1 lb/wk* GA at starty
At least 1 SGA infant (n=1,655)
29.3 G 5.5 29.3 G 5.9 13.3% 17.2% 82.9% 79.2% 8.6% 15.5% 20.7% 23.3 % 15.5% 19.9% 38.9% 32.2% 27.7 G 4.0 27.5 G 4.2
P value .699 ! .001 ! .001 ! .001 .021 ! .001 ! .001 .138
Data presented as mean G SD, or percentage, as indicated. * Modifiable factor. y Start of outpatient surveillance.
Table II
Logistic regression for modifiable risk factors P value
Prepregnancy BMI Wt gain !1/2 lb/wk Wt gain O1 lb/wk Tobacco user
! ! ! !
.001 .001 .001 .001
OR (95% CI) 1.03 1.35 0.77 1.95
(1.02-1.04) (1.16-1.56) (0.68-0.86) (1.68-2.27)
23.7%. Nontobacco users with poor maternal weight gain had an incidence of SGA of 18.5%.
Comment FGR is associated with increases in both short-term and long-term morbidity, as well as neonatal mortality. The mortality rate of SGA newborns has been shown to be 5 to 20 times greater in infants born between 1500 and 2500 g, and from 70 to 100 times higher in those infants born weighing less than 1500 g when compared with AGA infants.2 Short-term morbidities are also increased with FGR, including an increase in meconium aspiration, hypoglycemia, and birth asphyxia. Hypothermia, secondary to decreased fat stores, can lead to metabolic deterioration if not recognized appropriately. Finally, more recent literature has contradicted a previously held belief that infants with FGR had a lower incidence of respiratory distress syndrome.3 Long-term morbidities associated with FGR are still being identified. Infants born FGR, particularly those less than 2300 g, have a lag in mental development when compared with appropriately grown controls.4 Still other data have demonstrated a link between those infants born growth restricted and the risk of developing significant adult diseases, including cardiovascular disease, type 2 diabetes mellitus, and hyperlipidemia.
1442 Several different mechanisms have been proposed for these findings. Twin gestations are at a significantly higher risk for FGR than singleton gestations. Although our study involved a selected population with increased uterine activity, we believe these findings are generalizable because of our sample size. Although there are several risk factors, both maternal and fetal, that are outside the scope of influence of the physician, our study demonstrates there may be important risk factors for FGR that the clinician can work to eliminate.5 Based on our data, both tobacco abuse and poor maternal weight gain in the later half of pregnancy are behaviors that could elicit counseling for a patient with multiples and such an intervention may potentially improve infant outcomes.
Schwendemann et al
References 1. Alexander GR, Kogan M, Martin J, Papiernik E. What are the fetal growth patterns of singletons, twins, and triplets in the United States? Clin Obstet Gynecol 1998;41:115-25. 2. Williams RL, Creasey RK, Cunningham GC, Hawes WE, Norris FD, Tashiro M. Fetal growth and perinatal viability in California. Obstet Gynecol 1982;59:624-32. 3. Thompson PJ, Greenough A, Gamsu HR, Nicolaides KH. Ventilatory requirements for respiratory distress syndrome in small-forgestational-age infants. Eur J Pediatr 1992;151:528-31. 4. Low JA, Galbraith RS, Muir D, Killen H, Karchmar J, Campbell D. Intrauterine growth retardation: a preliminary report of longterm morbidity. Am J Obstet Gynecol 1978;130:534-45. 5. Kinzler WL, Aranth CV, Vintzrelos AM. Medical and economic effects of twin gestations. J Soc Gynecol Investig 2000; 7:321-7.