Objective
We sought to study long-term (neuro)developmental and behavioral outcome of pregnancies complicated by intrauterine growth restriction at term in relation to induction of labor or an expectant management.
Study Design
Parents of 2-year-old children included in the Disproportionate Intrauterine Growth Intervention Trial at Term (DIGITAT) answered the Ages and Stages Questionnaire (ASQ) and Child Behavior Checklist (CBCL).
Results
We approached 582 (89.5%) of 650 parents. The response rate was 50%. Of these children, 27% had an abnormal score on the ASQ and 13% on the CBCL. Results of the ASQ and the CBCL for the 2 policies were comparable. Low birthweight, positive Morbidity Assessment Index score, and admission to intermediate care increased the risk of an abnormal outcome of the ASQ. This effect was not seen for the CBCL.
Conclusion
In women with intrauterine growth restriction at term, neither a policy of induction of labor nor expectant management affect developmental and behavioral outcome when compared to expectant management.
Intrauterine growth restriction (IUGR) at term is associated with increased perinatal morbidity and mortality. Long-term morbidity is also increased in pregnancies complicated by IUGR. Studies have reported learning difficulties, defects in speech, neurological deficits, and behavioral problems to occur more frequently in term neonates born small for gestational age.
The Disproportionate Intrauterine Growth Intervention Trial at Term (DIGITAT) compared the effect of induction of labor in pregnancies complicated by IUGR with an expectant monitoring policy. The results of this study showed no important differences in adverse neonatal outcome between the 2 randomized groups. However, in the induction group, more neonates were admitted to intermediate care after induction than neonates in the expectant monitoring group (48% vs 36%). After a policy of expectant management, a larger percentage of neonates were born with a birthweight <10th centile when compared to neonates in the induction group (13% vs 31%, mean difference –18%; 95% confidence interval, 12–24%). In both groups, neonatal admissions as well as Morbidity Assessment Index (MAIN) score for newborns were lower after 38 weeks’ gestational age.
The objectives of this study were to: (1) study the long-term effects on (neuro)developmental and behavioral outcome of pregnancies complicated by IUGR at term; and (2) compare the influence of induction of labor to an expectant management policy on these long-term outcomes.
Materials and Methods
Participants
The study population consisted of children born to mothers who participated in the DIGITAT trial. From November 2004 through November 2008, pregnant women with a singleton fetus in cephalic presentation and suspected IUGR between 36 +0 -41 +0 weeks were recruited. Suspected IUGR was defined as a fetal abdominal circumference or an estimated fetal weight <10th percentile, or deceleration of the fetal abdominal circumference growth in the third trimester. Consenting women were randomly allocated to either induction or expectant monitoring. Participants allocated to the expectant monitoring group were strictly monitored until the onset of spontaneous labor. Details of the DIGITAT trial have been described elsewhere.
Baseline and neonatal characteristics
Data such as maternal characteristics around the time of randomization, gestational age at birth, birthweight, composite adverse neonatal outcome, and MAIN score were recorded in the original trial. Composite adverse neonatal outcome was defined as neonatal death, 5-minute Apgar score <7, umbilical artery pH <7.05, or admission to neonatal intensive care. The MAIN score is a validated numeric index outcome of early neonatal outcomes of prenatal care and adverse prenatal exposures in babies delivered >28 weeks’ gestational age and was calculated for all the neonates based on the characteristics recorded around birth. A MAIN score >0 indicates the presence of neonatal morbidity (ranging from mild to severe morbidity).
Developmental assessment: Ages and Stages Questionnaire
The Ages and Stages Questionnaire (ASQ) is a screening questionnaire designed to detect developmental delay in children. It contains questions to be answered by parents about 5 areas of development of their child: communication, gross motor, fine motor, problem solving, and personal-social. For each area, a mean score is calculated. The higher the score, the more abnormal the outcome is. An abnormal score is a score of ≥2 SD below the expected mean of a reference population, adjusted for age, and indicates a delay in development and a need for further assessment.
Child Behavior Checklist
The Child Behavior Checklist (CBCL) consists of 100 items concerning behavioral problems, on the basis of which a total problem score can be computed. It also informs on 7 narrowband syndrome scales (emotionally reactive, anxious/depressed, somatic complaints, withdrawn, sleep problems, attention problems, and aggressive behavior) and 2 broadband scales (internalizing and externalizing behavior). For each scale a standardized t score is calculated and a score >97th percentile falls into the clinical range that indicates serious behavior problems. The higher the t score, the more serious the behavioral problems.
Procedure
Parents of children randomized in the DIGITAT trial (n = 650) were requested to fill out the 2 questionnaires about the development of their child when their child was between 23-26 months of age. Research nurses contacted the parents by telephone and subsequently sent out the questionnaires by post. If the parents had not responded to the questionnaires, they were contacted again by the research nurses.
Statistical analysis
The number of children with abnormal scores for the ASQ and the CBCL were compared for the 2 groups with a policy of induction of labor or expectant management using the χ 2 test. For both questionnaires, the mean scores per area were compared between the 2 groups using t tests. Univariate analyses were performed using χ 2 for categorical values or t tests for means to identify factors of influence on the ASQ and CBCL by comparing children with an abnormal outcome to those without developmental problems. Factors with a P value < .10 were entered in a logistical regression model, either as continuous or as categorical variables, to assess the joint influence on the outcome of the ASQ and CBCL test. Software (SPSS, version 16.0; SPSS Inc, Chicago, IL) was used.
Results
Of the 650 parents of children, 582 (89.5%) randomized in the original trial were approached ( Figure ). Two parents were not approached for follow-up because their children were born with serious congenital abnormalities and caregivers of another child were not approached as the mother died postpartum of unknown causes. The response rate within the approached group was 54% (n = 158) in the induction group and 46% (n = 133) in the expectant monitoring group ( P = .02). In both groups, a small number (n = 24) of questionnaires were discarded because they were incomplete or filled in when the child was age <23 or >26 months.
Baseline characteristics
The baseline characteristics of the 2 management groups, as well as of the nonrespondents and nonapproached participants are shown in Table 1 . Similar to the findings of the primary trial, children in the induction group were lighter at birth and had a lower gestational age than children in the expectant management group. Baseline characteristics of the respondents were also compared with the nonrespondents/nonapproached. The responding mothers were older, less likely to smoke, and more frequently Caucasian than the nonrespondents. When comparing the approached group to the nonapproached group, we found that women in the nonapproached group were more likely to smoke.
| Characteristic | Approached, n = 582 | Nonapproached, n = 68 | Difference in percent or mean (95% CI) | Respondents, n = 292 | Nonrespondents, n = 290 | Difference in percent or mean (95% CI) | Induction of labor, n = 158 | Expectant management, n = 134 | Difference in percent or mean (95% CI) |
|---|---|---|---|---|---|---|---|---|---|
| A | B | A-B | C | D | C-D | E | F | E-F | |
| Maternal age, y | 27.2 (23.4–31.3) | 24.9 (21.7–30.5) | 1.0 (–0.3 to 2.4) | 28.1 (25.1–31.9) | 26.2 (22.2–30.6) | 1.8 (1.0–2.6) e | 28.0 (25.1–32.1) | 28.2 (24.7–31.5) | –0.1 (–1.3 to 0.9) |
| BMI at study entry a | 22.1 (19.7–25.5) | 22.1 (19.6–25.5) | –0.4 (–1.7 to 1.0) | 22.1 (19.9–25.7) | 22.0 (19.4–25.4) | 0.1 (–0.7 to 0.9) | 22.1 (19.9–25.8) | 23.2 (20.2–25.3) | 0 (–1.2 to 1.2) |
| Maternal smoking b | 229 (39.3) | 36 (52.9) | –13.6 (–26.1 to –1.1) d | 96 (32.9) | 133 (45.9) | –13.0 (–20.9 to –5.1) e | 53 (33.5) | 43 (32.1) | –1.4 (–9.3 to 12.2) |
| Caucasian c | 451 (77.5) | 56 (82.4) | –4.9 (–14.5 to 14.3) | 253 (86.6) | 198 (68.3) | 18.4 (11.7–24.9) e | 139 (88.0) | 114 (85.1) | 2.9 (–4.9 to 10.8) |
| Education | |||||||||
| Lower professional school | 303 (52.1) | 35 (51.5) | 0.6 (–11.9 to 13.1) | 147 (50.3) | 156 (53.8) | –3.5 (–11.6 to 4.6) | 83 (52.5) | 64 (47.8) | 4.8 (–6.7 to 16.3) |
| Higher professional school | 58 (10.0) | 5 (7.4) | 2.6 (–4.0 to 9.3) | 35 (12.0) | 23 (7.9) | 4.1 (–0.8 to 8.9) | 16 (10.1) | 19 (14.2) | –4.1 (–11.6 to 3.5) |
| Gestational age at birth, d | 270.5 (263.2–278.7) | 268.9 (263.8–279.8) | 0.2 (–2.3 to 2.7) | 269.9 (263.9–278.7) | 271.5 (262.9–278.9) | –0.3 (–1.8 to 1.2) | 266.4 (261.5–271.2) | 277.5 (269.8–283.6) | –9.9 (–11.8 to –7.9) |
| Birthweight, g | 2485.0 (2233.8–2750.0) | 2487.5 (2250.0–2911.3) | –17.8 (–115.2 to 79.5) | 2490.0 (2215.0–2755.0) | 2482.5 (2259.0–2745.0) | –23.7 (–83.5 to 36.2) | 2435.0 (2173.8–2660.0) | 2600.0 (2230.0–2850.0) | –133.8 (–221.2 to –46.5) e |
| Birthweight <10th centile | 407 (69.9) | 44 (64.7) | 5.2 (-6.7 to 17.2) | 205 (70.0%) | 202 (69.8) | 0.5 (–6.9 to 8.0) | 103 (65.2) | 102 (76.1) | –10.9 (–21.3 to –0.6) d |
| Intermediate level of care admission | 245 (42.1) | 30 (44.1) | –2.0 (–14.5 to 10.4) | 130 (44.5) | 115 (39.7) | 4.8 (–3.1 to 12.9) | 77 (48.7) | 53 (39.6) | 9.1 (–2.2 to 20.6) |
| MAIN score >0 | 128 (22.0) | 17 (25.0) | –3.0 (–13.8 to 7.8) | 69 (23.6) | 59 (20.3) | 3.3 (–3.4 to 10.0) | 40 (25.3) | 29 (21.6) | 3.7 (–6.1 to 13.4) |
| Composite adverse neonatal outcome | 29 (5.0) | 8 (11.8) | –6.7 (–14.6 to 1.0) | 14 (4.8) | 15 (5.2) | –0.4 (–3.9 to 3.2) | 7 (4.4) | 7 (5.2) | –0.8 (–5.7 to 4.2) |
| Randomization | |||||||||
| Induction | 292 (50.2) | 30 (44.1) | 5.9 (–6.6 to 18.4) | 158 (54.1) | 133 (45.9) | 8.2 (0.15–16.3) d | NA | NA | NA |
| Expectant management | 290 (49.8) | 38 (55.9) | –5.8 (–18.4 to 6.6) | 134 (45.9) | 157 (54.1) | –8.2 (–16.3 to –0.15) d | NA | NA | NA |
a n = 506 for approached, n = 61 for nonapproached, n = 263 for respondents, n = 246 for nonrespondents, n = 141 for induction, n = 122 for expectant;
b n = 535 for approached, n = 66 for nonapproached, n = 267 for respondents, n = 272 for nonrespondents, n = 144 for induction, n = 123 for expectant;
c n = 545 for approached, n = 67 for nonapproached, n = 278 for respondents, n = 271 for nonrespondents, n = 151 for induction, n = 127 for expectant;
Ages and Stages Questionnaire
For the ASQ, 25% (n = 38) of the children in the induction group and 29% (n = 35) of the children in the expectant management group had an abnormal score in ≥1 areas of development ( Table 2 ). The mean scores per problem area were calculated for induction and expectant management. No significant differences were found in the mean scores ( Table 3 ) or in the number of children with abnormal scores ( Table 2 ) for a policy of induction compared to expectant management.
| Questionnaire | Induction of labor, n (%) a | Expectant management, n (%) b | Difference in percentage (95% CI) |
|---|---|---|---|
| ASQ | 38 (25) | 35 (29) | –4 (–14 to 7) |
| CBCL | 21 (14) | 13 (11) | 3 (–5 to 11) |
a n = 152 for ASQ, n = 147 for CBCL;
| Variable | Induction | Expectant management | P value |
|---|---|---|---|
| Problem area ASQ | ASQ (n = 152) | ASQ (n = 122) | |
| Communication | 50.9 (11.7) | 51.2 (13.1) | .8 |
| Gross motor | 53.7 (13.4) | 52.3 (10.2) | .3 |
| Fine motor | 48.7 (9.3) | 47.9 (11.2) | .5 |
| Problem solving | 42.3 (10.4) | 44.1 (12.5) | .2 |
| Personal social | 46.7 (11.0) | 47.3 (11.6) | .7 |
| Syndrome scale CBCL | CBCL (n = 122) | CBCL (n = 118) | |
| Emotionally reactive | 52.9 (5) | 52.6 (4.5) | .6 |
| Anxious/depressed | 51.3 (2.9) | 50.9 (2.0) | .2 |
| Somatic complaints | 54.3 (7.1) | 54.1 (6.3) | .8 |
| Withdrawn | 53.0 (5.4) | 52.3 (4.0) | .2 |
| Sleep problems | 53.0 (5.9) | 52.2 (5.5) | .3 |
| Attention problems | 54.1 (5.2) | 53.7 (5.0) | .5 |
| Aggressive behavior | 53.9 (5.9) | 53.4 (4.9) | .5 |
| Internalizing | 45.5 (10.7) | 44.7 (9.2) | .5 |
| Externalizing | 50.2 (9.2) | 48.2 (9.9) | .1 |
| Total problem score | 47.6 (9.9) | 45.6 (9.8) | .1 |
Child Behavior Checklist
For the CBCL, 14% in the induction group and 11% in the expectant management group had an abnormal score in ≥1 areas of the CBCL ( Table 2 ). There were no differences between the mean t scores for a policy of induction of labor compared to expectant management ( Table 3 ).
Table 4 shows that 43% of children with a birthweight <2.3rd centile had an abnormal outcome of the ASQ, and that lower percentages with abnormal scores were found in higher birthweight centiles ( P < .001). Thirty-five percent of children with a MAIN score >0 had an abnormal outcome of the ASQ compared to 22% of children with a MAIN score equal to 0 ( P = .04). None of the 4 children admitted to the intensive care had a poor outcome of the ASQ. However, of the children admitted to an intermediate level of care, 34% had an abnormal outcome of the ASQ, significantly higher than the 20% abnormal scores found in children not admitted or admitted to the maternal ward ( P = .005). No significant correlation was found between gestational age at birth, composite adverse neonatal outcome at birth, management policy, maternal smoking during pregnancy, or education level of mother and an abnormal outcome of the ASQ. We could not identify any factors that were significantly related to the outcome of the CBCL ( Table 4 ).
