The intrauterine ponderal index in relation to birth weight discordance in twin gestations

Objective: To establish the relationship between the fetal ponderal index and birth weight discordance in twins. Method: The fetal ponderal index (estimated fetal weight ÷ femur length³) was calculated in 86 pairs of twins delivered within 2 weeks of the last sonography and analyzed in relation to birth weight discordance. Results: A weak but significant correlation between fetal ponderal index and birth weight (r = 0.26, P < 0.0007) but no correlation with gestational age (r = 0.035, P = 0.65) were found. Members of concordant pairs (<15% birth weight difference) had a significantly higher fetal ponderal index compared with members of mildly (15–25%) discordant pairs (P < 0.02), but not as compared with members of severely discordant (>25%) pairs. Conclusion: The characteristics of the fetal ponderal index in twins are similar to those in singletons. Fetal size seems to be diminished in severe but not in mild discordants. However, in its present form, the fetal ponderal index is a poor predictor of discordant growth and therefore should be employed cautiously in twin gestations.     

A novel approach to intertriplet birth weight discordance

OBJECTIVE: This study was undertaken to evaluate intertriplet birth weight discordance. STUDY DESIGN: Birth weight discordance greater than 25% was evaluated in a cohort of 2804 US live-born triplets. Symmetric and high- and low-skew sets were defined by the rank of the middle triplet between the heavier and the lighter triplets. Frequencies of discordance level and type were analyzed by gestational age, parity, and total triplet birth weight. RESULTS: Discordance of 25.1% to 35% and greater than 35% was found in 19.4% and 9.5% of the triplets analyzed, respectively. Frequencies of greater than 25% discordant sets demonstrated polynomial relationship to gestational age (R (2) = 0.94, P <.001) total triplet birth weight deciles (R (2) = 0.97, P <.001). Frequencies of discordance type are unchanged throughout gestation. CONCLUSION: Birth weight discordance in triplets is frequent and large and implies exhaustion of fetal growth potential despite a uterine environment that appears to perform at maximal effort in these pregnancies.     

The ponderal index in triplets: I. Relationship to small for gestational age neonates

The aim of the study was to examine the ponderal index in small for gestational age (SGA) triplets. Prospectively collected data from a cohort of triplets born at 28 to 37 weeks were analyzed. A low neonatal ponderal index (birth weight/[length]3) was defined as less than 1 SD below the mean (2.0), and SGA was considered as birth weight below the 10th percentile by triplet standards. We studied 2181 sets of triplets. Triplets delivered at < or = 33 weeks have a lower mean ponderal index compared with those delivered at > 33 weeks. About 70% of SGA triplets do not have a low ponderal index, whereas 79.2% of infants with a low ponderal index are not SGA by triplet standards. Both the frequency of a low ponderal index and the frequency of infants with a low ponderal index who are not SGA decrease with increasing gestational age. We conclude that the majority of triplets with a low ponderal index might not be considered growth restricted, supporting the concept that reduced fetal weight of triplets is more likely a physiological rather than a pathological phenomenon.     

Birth weight discordance in multiple gestations: occurrence and outcomes.

This paper reviews the patterns of occurrence, measurement and the effect of birth weight discordance on fetal and neonatal mortality in multiple pregnancies (twins and triplets). Birth weight discordance is fairly common among multiple pregnancies, and about one-quarter of the twin deliveries are affected by a birth weight discordance of 15%, while nearly 5% of twin gestations experience severe discordance (>or= 35%). Factors influencing birth weight discordance are exaggerated in triplet deliveries. Approximately 20% of triplet sets experience a birth weight discordance of 25 - 35% and nearly 10% experience severe forms of discordance. Frequencies of discordant sets at >25% discordance along the range of birth weight deciles show an inverse logarithmic relationship in twins, while the best-fit function in triplets is polynomial. Birth weight discordance is significantly associated with both fetal and neonatal mortality. Neonatal demise among the severely discordant smaller twin is significantly more frequent than in a non-discordant smaller twin. The magnitude of the effect is greater when one or both of the discordant twins are concomitantly small for gestational age.     

The association between maternal characteristics and different types and levels of discordance in triplet pregnancies

OBJECTIVE: This study evaluated the association of maternal factors known to influence outcomes of triplets, different discordance levels (-25, 25.1-35, and )35%), and three types (according to the birth weight of the middle-sized triplet) of birth weight discordance in triplets. METHODS: We used data collected by the Women's Health Division of Matria Healthcare, Inc. (Marietta, GA). We analyzed a cohort of 2706 triplet sets, to calculate the frequencies of different levels and types of birth weight discordance by maternal age, parity, weight, height, body mass index and weight gain at 24 weeks of gestation. RESULTS: We found a positive association between maternal parity and birth weight discordance level but no clear association between the other maternal factors and the level of discordance as well as the type of discordance. However, a trend was seen whereby overweight women had a trend towards the low-skew (a set comprising one large and two small triplets) type of birth weight discordance and an opposite trend in underweight women. CONCLUSIONS: These results corroborate previous findings that nulliparity is associated with aberrant growth in triplet pregnancies.     

Mean arterial pressure in concordant and discordant triplets during the first week of life.

OBJECTIVE: To retrospectively determine mean arterial pressure (MAP) for stable concordant and discordant triplets during the first 7 days of life. BACKGROUND: Morbidity and mortality for prematurely born triplets is high, therefore, MAP monitoring during the first day of life is important for their clinical management. MAP reference values for special populations such as triplets have not been published. Recently, we reported that in stable discordant twins MAP values during the first day of life were significantly lower in the smaller than in their larger siblings. Comparable information for triplets is not available. DESIGN: Retrospective cohort study. METHODS: We studied 30 sets of concordant and 29 sets of discordant (birth weight difference > or =20%) consecutively born triplets. Stable patients were defined as those having umbilical cord hemoglobin > or =13 g/dl, normal blood gases, who were never treated for hypotension, and survived at least 7 days. MAP (torr) were measured by oscillometry in 3410, and by transducer via an umbilical arterial catheter in 1251 instances. RESULTS: Concordant and discordant triplets were similar in demographics, history of preterm labor (63 and 63%), chorioamnionitis (10 and 10%), pre-eclampsia (53 and 48%), cesarean delivery (100 and 100%), antenatal steroids (77 and 73%), cord hemoglobin (16 and 16 g/dl), combined triplets birth weight (4922 and 4732 g), gestational age (32 and 33 weeks), normal head ultrasounds or Grade I intracranial hemorrhage (96 and 100%) and neonatal mortality (2 and 1%), but were different in the number of infants requiring mechanical ventilation (57 and 31%). A total of 80 (89%) concordant triplets and 77 (88%) discordant triplets were stable according to our definition. Concordant stable triplets, whether small, medium or large, had similar MAP at birth. Their MAP values increased noticeably from birth to 24 hours and more subtlely to 7 days. Triplets of < or =32 weeks GA had lower MAP throughout than those of > or =33 weeks GA. Discordant stable triplets were divided into 27 small (1382 g), 26 medium (1683 g) and 27 large (1969 g); during the first 24 hours, medium and smaller triplets had MAP values that were lower than those of their larger siblings. From the second to the seventh day of life, all MAP values and trends were similar. Among discordant triplets, 86% of the smallest, 13% of the medium and 13% of the largest infants had asymmetrical intrauterine growth restriction. CONCLUSION: In stable concordant and stable discordant triplets, MAP correlates with birth weight, gestational age and postnatal age. MAP values increase noticeably during the first 24 hours and more subtlely during the next 7 days. Concordant or discordant, small, medium, and large triplets have similar MAP values and trends to that of their siblings. Small and medium discordant triplets have lower MAP values during the first day of life than their larger siblings but by the second day there MAP trends and values were no longer different.     

The ponderal index in triplets: II. Gestational age-related patterns of neonatal weights and lengths

In order to evaluate the determinants of high and low ponderal indices in triplets, we analyzed prospectively collected data from a cohort of 2181 triplet births. Low and high neonatal ponderal indices (birth weight/[length]3) were defined as below or above 1 SD from the mean. The mean ponderal index was of 2.4 +/- 0.4. At 30-31 weeks, there were significantly more infants with a low ponderal index; after 33 weeks, more infants were born with a high index. Birth weights of infants with a high index were significantly higher throughout the entire range of gestational ages, whereas their lengths were significantly smaller. Both birth weights and infant length had significant correlations with gestational ages for infants with a low (R2=0.97 and R2=0.94, respectively) as well as with high ponderal indices (R2=0.95 and R2=0.94, respectively). The regression analyses suggest, however, different patterns for infants with low or high ponderal indices.     

Intertwin birth weight discordance as a potential adaptive measure to promote gestational age

OBJECTIVE: To determine if discordant twins are delivered at a more advanced gestational age than concordant twins of the same total twin birth weight. STUDY DESIGN: We used the U.S. National Center for Health Statistics matched multiple birth data file containing all twin births in the United States from 1995 through 1997. Birth weight discordance was defined as > 25% difference between the heavier and lighter twin. We compared mean gestational age between concordant and discordant pairs at 250-g total birth weight intervals in the entire population and in the subgroups of primiparas and multiparas. RESULTS: The frequency of discordant pairs declined with increasing total twin birth weight, from 7.2% at 3,000-3,250 g, to 3% at 4,750-5,000 g (R2 = .94, P < .05), with no significant difference between primiparas and multiparas (all P values > .05). The mean gestational age of discordant pairs was significantly higher across the entire range of total birth weight intervals except for the uppermost interval (> 4,750 g). The effect of discordance on gestational age was modified by parity, with significant differences between concordant and discordant pairs among primiparas mainly at the lower birth weight strata (P < .05). CONCLUSION: In the total twin birth weight range of 3,000-5,000 g, birth weight discordant twins are delivered at a more advanced gestational age than concordant twins. Discordant growth may serve as an adaptive measure to promote maturity by reducing the inevitable uterine overdistension.     

Perinatal outcome, placental pathology, and severity of discordance in monochorionic and dichorionic twins

OBJECTIVE: To evaluate differences in pregnancy outcomes and placental findings among severely discordant monochorionic and dichorionic twins. METHODS: We studied retrospectively a cohort of 382 twin pregnancies with gestational ages that ranged from 24 to 40 weeks. Pregnancies were classified as dichorionic or monochorionic by histologic examination of placentas. Infants were subdivided into concordant (less than 5% difference in birth weight), mildly discordant (5-25% difference), and severely discordant groups (more than 25% difference), and their clinical characteristics and findings at placental examination were analyzed and compared. RESULTS: Severe discordance occurred significantly more often in monochorionic than in dichorionic twins and was associated with significantly more deliveries before 36 weeks and more newborns remaining more than 10 days in the neonatal intensive care unit. Severely discordant monochorionic and dichorionic twins had significantly worse perinatal mortality and morbidity than mildly discordant and concordant twins. The weight of the placenta of the smaller fetus in severely discordant dichorionic twins with separate placentas and the total placental weight in severely discordant monochorionic twins were significantly smaller than the weights of the placentas in their concordant and mildly discordant counterparts. The umbilical cords of the smaller fetuses in both dichorionic and monochorionic pregnancies exhibited significantly more velamentous insertions and single umbilical arteries than in concordant or mildly discordant twins of similar chorionicity. CONCLUSION: Severe discordance is more frequent and has greater morbidity in monochorionic than dichorionic twins. The most frequent findings in the placentas of severely discordant twins were small placental weight and umbilical cord abnormalities.     

Effect of maternal height on gestational age and birth weight in nulliparous mothers of triplets with a normal pregravid body mass index

OBJECTIVE: To study the effect of maternal height on gestational age and birth weight of triplets born to nulliparous women with a normal pregravid body mass index (BMI). STUDY DESIGN: A 1988-2000 prospective cohort of 1,219 U.S. live-born triplets was evaluated. Patients were grouped into stature categories of 5-cm intervals. Mean values for BMI, height, maternal age, gestational age at birth, total triplet birth weight and frequencies of births at < 28 weeks and with very low birth weight were calculated for each height category. RESULTS: Maternal heights were normally distributed (mean, 165.2 +/- 6.2 cm). We found a significant positive correlation (R2 = .95), different from a zero slope (P < .01), between mean total triplet birth weight and height category. Nulliparous women who were taller than 165 cm had age, BMI and gestational age characteristics similar to those of their shorter counterparts but delivered significantly heavier triplets and were at significantly lower risk of delivering very-low-birth-weight triplets. CONCLUSION: Taller women are more likely to deliver heavier triplets and are at lower risk of delivering very-low-birth-weight triplets. This information should be included in counseling women with the potential of conceiving triplets.     

Comparison between singleton- and triplet-specific "growth" curves to detect growth restricted triplet infants

OBJECTIVE: To evaluate whether triplet- or singleton-specific growth standards should be used to define the growth restricted triplet fetus/neonate. STUDY DESIGN: We retrospectively compared the predictive values of singleton vs. triplet-specific "growth" standards using the neonatal ponderal index as reference for growth restriction. RESULTS: A ponderal index <1 SD from the mean was found in 356 (14.4%) of 2477 triplet infants. A total of 686 (27.7%) infants were small for gestational age (SGA) by singleton standards, but only 168 (6.8%) by triplet standards. After 31 weeks' gestation, triplet standards are significantly better associated with a low ponderal index (OR 2.0, 95% CI 1.1, 3.4 at 32 weeks to OR 4.1, 95% CI 2.8, 6.3 at 36-37 weeks), resulting in better positive predictive value and higher specificity of triplet standards in predicting a low ponderal index. CONCLUSION: Triplet birth weight standards are better than singleton standards at >31 but not at < or =31 weeks' gestation in predicting a low neonatal ponderal index.     

The distribution of intra-twin birth weight discordance and its association with total twin birth weight, gestational age, and neonatal mortality

OBJECTIVE: To determine an appropriate cutoff value to differentiate physiologic and pathologic birth weight discordance, to determine the prevalence of birth weight discordance > or =25% among twin pregnancies in different sub-populations, and to examine its clinical significance. STUDY DESIGN: Historical cohort study of 147,262 twin sets registered in the United States between 1995 and 1997. RESULTS: The prevalence of birth weight discordance > or =25% among twin pregnancies was 8.6%. The prevalence of birth weight discordance > or =25% was significantly decreased with increasing total twin birth weight deciles, was more frequently found in twins with discordant gender (9.1%) than in those twins with concordant gender (8.3%) and in mothers whose age was 30 years or older (9.1%) than those of younger mothers (8.1%). Birth weight discordance > or =25% was associated with earlier gestational age at delivery (35.0 weeks versus 36.0 weeks) and higher neonatal mortality (5.4% versus 2.3%) as compared to twins with lower birth weight discordance. CONCLUSIONS: The prevalence of birth weight discordance > or =25% among twin pregnancies was 8.6%, which is associated with lower gestational age at delivery and higher neonatal mortality rates, and may represent a pathologic process.     

First trimester prediction of growth discordance in twin gestations

OBJECTIVE: The purpose of this study was to determine whether first trimester ultrasound scanning can identify twin gestations that are at risk for subsequent growth discordance. STUDY DESIGN: Ultrasound examinations of dichorionic twin pregnancies between 11 and 14 weeks of gestation were evaluated for growth discordance with crown-rump length. Pregnancies that were complicated by fetal death or anomalies were excluded from the analysis. Birth weight discordance was defined as >or=20% difference in birth weight, relative to the larger twin. Fisher's exact, Mann Whitney U, and Spearman rho tests were used for statistical analysis. RESULTS: Of 130 twin pregnancies, 16 pregnancies (12.3%) had discordant birth weight. Crown-rump length disparity was correlated positively with birth weight discordance (r=0.31; P<.001). Of pregnancies with a discrepancy of 3 days discrepancy (P=.004), which resulted in a likelihood ratio of 5.9 for having discordant birth weight. CONCLUSION: Twins who are ultimately discordant at birth may exhibit differences in growth as early as 11 to 14 weeks of gestation.     

Does birth weight discordance in preterm twins affect neonatal outcome?

OBJECTIVE: To compare neonatal morbidity and mortality between hospitalized discordant and concordant preterm twin pairs. STUDY DESIGN: This was a retrospective, hospital-based study of preterm twins, followed in neonatal intensive care unit. Twins were classified as discordant when the difference in birth weight was >15%. Morbidity and mortality rates were compared between discordant and concordant groups. RESULTS: Of 136 preterm twins in the study, 54 were discordant and 82 were concordant. In discordant twins hypoglycemia and cesarean delivery rate was higher than concordant pairs. The mortality rate did not differ between concordant and discordant groups. CONCLUSIONS: Hospitalized preterm discordant twin pairs have an increased risk of hypoglycemia compared to concordants. Discordant growth is not a risk factor for prolonged hospitalization and increased mortality.     

Placental transport rather than maternal concentration of amino acids regulates fetal growth in monochorionic twins: implications for fetal origin hypothesis.

OBJECTIVE: The cause of discordant growth in monochorionic twins is not clear. We hypothesize that growth restriction of 1 monochorionic twin is due to fetal under-nutrition. STUDY DESIGN: We measured plasma amino acid concentrations by high performance liquid chromatography in maternal venous and fetal umbilical cord venous blood samples that were collected at birth from gestational age-matched monochorionic twins (n = 14) with a birth weight discordance of > or =20%. The concordant monochorionic twins with birth weight differences of < or =10% constitute a control group. RESULTS: In the intrauterine growth-restricted twins, fetal concentrations of essential amino acids valine (P <.01), leucine (P <.01), isoleucine (P <.01), phenylalanine (P <.01), and L-arginine (P <.05) were lower than the co-twins and concordant monochorionic twin pairs. Fetal concentrations of the nonessential amino acids taurine (P <.001), serine (P <.01), glycine (P <.01), tyrosine (P <.01), and aspartic acid (P <.01) were lower in the intrauterine growth-restricted twin than the co-twin or concordant monochorionic twins. No such differences were found between concordant monochorionic twin pairs. Maternal amino acid concentrations were similar between discordant and concordant groups. CONCLUSION: Concentrations of certain essential and nonessential amino acids in the intrauterine growth-restricted twins were lower than the co-twins. These differences support the hypothesis that intrauterine growth-restriction that affects 1 of the monochorionic twins is due to the impaired placental transport of amino acids rather than intertwin transfusion of blood.     

Human versus animal insulin in the management of insulin-dependent diabetes: lack of effect on fetal growth

It is generally accepted that the human placenta is impermeable to free insulin and that insulin present in the fetus is entirely of fetal origin. A recent study suggested that antibody-bound animal insulin crosses the placental barrier and may exert direct effects on fetal growth. We hypothesized that mothers with insulin-dependent diabetes treated with animal insulin would have infants with higher birth weights and ponderal indices compared with mothers treated with human insulin. We studied 209 mothers with insulin-dependent diabetes who were enrolled in our program and who delivered after 28 weeks' gestation: 170 were treated with animal insulin and 39 with human insulin. There were no differences between the groups in the mean birth weight (adjusted by gestational age at delivery) or ponderal index of the infants. The rate of macrosomia (birth weight greater than the 90th percentile for gestational age or ponderal index above 2.85) was similar in both groups. The sample size was adequate to yield a power of 80% to detect a difference between groups of 179 g or more in birth weight and 0.1 g/cm3 in ponderal index. We suggest that the type of insulin (animal versus human) used by the pregnant insulin-dependent diabetic mother has no bearing on fetal weight gain.     

Multifetal pregnancy reductions of triplets to twins: comparison with nonreduced triplets and twins.

OBJECTIVE: Multifetal pregnancy reduction has been shown to improve survival rates in high-order multifetal pregnancies (>/=4). There is, however, some controversy as to whether multifetal pregnancy reduction improves pregnancy outcomes of triplets reduced to twins. The purpose of this study was to evaluate this issue by comparing outcomes of triplet gestations undergoing reduction to twins with outcomes of nonreduced twin gestations and expectantly managed triplet gestations. STUDY DESIGN: The study included 143 triplet pregnancies that underwent reduction to twins over a 10-year period at a single center. These were compared with 12 nonreduced triplet pregnancies from the Wayne State University Perinatal Database and with 2 groups of twin pregnancies: 605 from the Wayne State University Perinatal Database and 207 from the Quest Diagnostics Database. RESULTS: The miscarriage rate for expectantly managed triplets was 25%, compared with 6.2% for triplets reduced to twins. This rate was similar to the rates for both groups of nonreduced twins: 5.8% (Quest) and 6.3% (Wayne State University). Severe prematurity occurred in 25% of nonreduced triplets compared with 4. 9% of twins after reduction. This rate was also similar to that of nonreduced twins: 7.7% (Quest) and 8.4% (Wayne State University). The mean gestational age at delivery for expectantly managed triplets (32.9 +/- 4.7 weeks) was significantly shorter than for triplets reduced to twins (35.6 +/- 3.1 weeks). By comparison, nonreduced twins had a mean gestational age at delivery of 35.8 +/- 3.9 weeks for Quest and 34.4 +/- 3.6 weeks for Wayne State University. Mean birth weights were significantly lower in expectantly managed triplets as compared with triplets undergoing reduction to twins (1636 +/- 645 g vs 2381 +/- 602 g, respectively). Nonreduced twins had a mean birth weight of 2254 +/- 653 g for Quest and 2123 +/- 634 g for Wayne State University. Pregnancy loss rates, mean length of gestation, and mean birth weight did not vary significantly between triplets who underwent reduction to twins and nonreduced twins. CONCLUSIONS: Reduction of triplets to twins significantly reduces the risk for prematurity and low birth weight and may also be associated with a reduction in overall pregnancy loss. This suggests that multifetal pregnancy reduction of triplets to twins is a medically justifiable procedure not only from an actuarial viewpoint but also from the ethical perspective of supporting patients' autonomy and respect for patients' individual circumstances.     

Doppler velocimetry in concordant and discordant twin gestations.

Doppler blood flow studies in 63 pairs of concordant and 17 pairs of discordant twins were compared with those of 277 appropriate for gestational age singleton fetuses. Discordancy was defined as a more than 20% intra-pair actual birth weight difference. The data were divided into five groups: singletons and large and small concordant and discordant twins. Statistical comparisons of the regression lines for the large concordant and discordant twins and the singletons showed no significant differences among the three lines, either in slopes (P = .1) or intercepts (P = .08). Comparisons of the regression lines for small concordant and discordant twins and the singletons indicated a significant interaction among the three lines (P = .01). Additional analysis leads us to conclude that the systolic-diastolic ratio (S/D) decreases with advancing gestational age for all groups except small discordant twins. The S/D of small discordant twins differed significantly from that of singletons and tended to differ from that of small concordant twins.     

Frequency, distribution, and theoretical mechanisms of hematologic and weight discordance in monochorionic twins.

OBJECTIVE: The purpose of this study was to determine the frequency, distribution, and most likely etiology of hematologic and weight discordance in pathologically proven monochorionic twins, and to use this information to reevaluate the neonatally derived definition of the twin-twin transfusion syndrome. METHODS: We reviewed our experience with 97 pathologically proven monochorionic twin pregnancies. The frequency and distribution of weight and hemoglobin-hematocrit (hb-hct) discordance were determined for all twin pairs. Factors that may have contributed to the discordance were identified, and theoretical mechanisms were proposed. RESULTS: All combinations of weight and hb-hct discordance were observed. Thirty-four twin pairs (35%) were discordant for weight. In half of these (17 of 34), the hb and hct were concordant. In 18% (six of 34), the smaller twin had the higher hb-hct, and in 32% (11 of 34), the smaller twin had the lower hb-hct. Twenty-three of 63 size-concordant pairs (36%) were discordant for hb-hct. Ten infants were infected at birth, eight had malformations, and 25 likely suffered an acute transfusion event. CONCLUSIONS: Any combination of weight and hb-hct discordance can occur in monochorionic twins. Acute and chronic twin-twin transfusion, uteroplacental insufficiency, infection, malformations, or other factors may have accounted for the discordance observed. Thorough antenatal evaluation with invasive testing and marker studies (to identify a physiologically unbalanced placental anastomosis) may be necessary to establish an accurate diagnosis. We conclude that weight and/or hb-hct discordance is relatively common in monochorionic twins and in itself is not sufficient to diagnose twin-twin transfusion.     

Identification of monozygotic twins that are concordant for preeclampsia

Preeclampsia is thought to be discordant in monozygotic twins. While recruiting for a genetic study into preeclampsia, we identified 4 sets of twins and a triplet gestation; all were monozygous on deoxyribonucleic acid "fingerprinting." Two twins were concordant for preeclampsia, and 2 of the triplets had pregnancy-induced hypertension, although only 1 of them had proteinuria. Hence we confirm the existence of monozygotic twins that are concordant for preeclampsia.