Australian and New Zealand Journal of Obstetrics and Gynaecology 2003; 43: 429–432

429

Blackwell Publishing Ltd. Original ArticleNeonatal morbidity in GDM and type 2 DM

Admissions to neonatal intensive care unit following pregnancies complicated by gestational or type 2 diabetes

Diana WATSON,1 Janet ROWAN,1,2 Lee NEALE1 and Malcolm R. BATTIN1,3

1National Women’s Hospital, Departments of 2Obstetric Medicine and 3Paediatrics, University of Auckland, Auckland, New Zealand

AbstractBackground: When gestational diabetes mellitus (GDM) is diagnosed in a population with a high prevalence ofunrecognised type 2 diabetes mellitus (type 2 DM), the rate of neonatal morbidity is not clear. There is also a paucityof data reporting neonatal outcome in women with recognised type 2 DM.

Aim: To describe, in a population with a high background prevalence of type 2 DM, neonatal morbidity in infantsof women with GDM and type 2 DM admitted to the neonatal intensive care unit (NICU).

Methods: A 2-year audit was carried out in a tertiary level obstetric hospital with a multi-ethnic delivery population.All infants admitted to the NICU whose mothers had GDM or type 2 DM were identified from the hospitaldatabase. The records of 136 infants were retrospectively reviewed and data collected on outcome measures includ-ing maternal diagnosis, macrosomia, mode of delivery, delivery complications, hypoglycaemia, respiratory distressand congenital anomalies.

Results: Admission to NICU occurred in 29% of GDM and 40% of type 2 DM pregnancies. Median gestationwas 37 weeks (range: 25–41), with 46% delivered preterm. Forty percent of infants were delivered by emergencyCaesarean section. Fifty-one percent of admissions had hypoglycaemia and 40% required support for respiratorydistress. Women with type 2 DM diagnosed either prepregnancy or post-partum were the highest risk group forneonatal morbidity, including congenital anomalies.

Conclusions: Neonatal morbidity is common in infants of women with type 2 DM and GDM in a population withhigh prevalence of type 2 DM.

Key words: gestational diabetes, neonatal morbidity, pregnancy, type 2 diabetes mellitus.

Background

Pregnancy is associated with increasing insulin resistance.1

Women who are unable to compensate for this develop ele-vated plasma glucose levels, termed gestational diabetesmellitus (GDM),1 which is associated with increased perinatalmorbidity.2 Before birth, the elevated maternal glucose andother nutrients cross the placenta and can lead to fetal hyper-insulinaemia with subsequent macrosomia.2,3 This is associatedwith an increased risk of shoulder dystocia or traumaticdelivery. After delivery, the infant is at increased risk of hypo-glycaemia, respiratory distress, and jaundice.2,3 In addition, anelevated glucose during embryogenesis, in previously un-recognised type 2 diabetes mellitus (type 2 DM), predisposesto congenital anomalies.4,5

Women diagnosed with GDM are a heterogeneous groupranging from women with minor glucose elevations to thosewith previously undiagnosed type 2 DM. At each centre, therate of neonatal complications is likely to be affected by theunderlying prevalence of type 2 DM as well as the criteria

for diagnosing GDM, which is not uniform.6 Treatment afterdiagnosis may also influence neonatal outcome.7,8

In New Zealand (NZ), the delivery population is multi-ethnic, including populations that have a high prevalence oftype 2 DM. At the National Women’s Hospital (NWH) thedistribution of maternal ethnicity is 52% Caucasian, 18%Pacific Island, 9.8% Chinese, 8% Maori, 4.3% Indian and4.8% other Asian.9 Using criteria based on recommendationsfrom the European Association for Study of Diabetes,10 therate of GDM ranges from 2% in Caucasian women to 7% inPacific Island and Indian women. Two-thirds of the womendiagnosed with GDM at NWH are from Pacific Island orIndian and other Asian populations.9 The threshold for

Correspondence: Dr Malcolm R Battin, Newborn Services,National Women’s Hospital, Private Bag 92189, Auckland, New Zealand. Email: malcolmb@adhb.govt.nz

Received 9 January 2003; accepted 16 July 2003.

D. Watson et al.

430 Australian and New Zealand Journal of Obstetrics and Gynaecology 2003; 43: 429–432

diagnosis of GDM is higher in NZ than other countries,6

so these women have a greater degree of plasma glucoseelevation. This is illustrated by the results of the 75 g glucosetolerance test (GTT) carried out 6 weeks post-partum.9 Ofthe 70% of women who had a GTT, 16% were reclassifiedas having impaired glucose tolerance (IGT) (2h plasmaglucose 7.8–11.0 mmol/L) and 13% have type 2 DM (fast-ing plasma glucose ≥7.0 mmol/L and/or 2 h plasma glucose≥11.1 mmol/ L ). There is a paucity of data on neonatalmorbidity in type 2 DM and GDM populations with a highprevalence of underlying type 2 DM, despite recognitionthat type 2 DM in pregnancy is increasing.11 Therefore,we carried out an audit of neonatal morbidity in infants ofwomen with GDM or type 2 DM within our hospital, look-ing at a composite of morbidity in the infants requiringadmission to the neonatal intensive care unit (NICU). Inwomen with GDM who had post-partum GTT, morbidityaccording to maternal post-partum diagnosis was analysed.

Methods

National Women’s Hospital is a tertiary referral centre with7500 deliveries per annum. Women are routinely screenedfor GDM between 24 and 28 weeks’ gestation with a non-fasting 50 g polycose drink. If the 1 h plasma glucose is≥7.8 mmol/L, a 75 g oral GTT is carried out. Gestationaldiabetes is diagnosed if the fasting glucose is ≥5.5 mmol/Land/or the 2 h glucose is ≥9.0 mmol/L.

Women with GDM or pre-existing diabetes are seen in asingle clinic served by a multidisciplinary team, includingobstetricians, physicians, midwives and dietitians. Advice isgiven about diet and exercise, and insulin treatment addedif glycaemia aims are not achieved. Home blood glucosemonitoring is carried out four times daily, with aims of afasting glucose <5.5 mmol/L and 90 min postprandial glucose<6.8 mmol/L.

Women in the clinic usually deliver their babies between38 and 40 weeks’ gestation, unless pregnancy complicationsrequire earlier delivery. Infants are admitted to the postnatalward unless problems requiring NICU admission are identi-fied. Neonatal blood glucose monitoring is carried out withinan hour of delivery and before each feed until two consecutiveglucose levels are ≥2.6 mmol/L. Hypoglycaemia was definedas a plasma or capillary glucose <2.6 mmol/L. Treatment isbased on our published guidelines.12 A single level below2.6 mmol/L is treated with supplemental feeding if possible.Recurrent glucose levels <2.6 mmol/L, or symptomatichypoglycaemia lead to admission to the NICU.

For the present audit, all pregnancies complicated by GDMor type 2 DM were identified from the NWH database fromJanuary 1999 to December 2000. For the infants admitted tothe NICU, medical records were reviewed and data collected.Maternal data included age, ethnicity, type of diabetes (ante-natally and post-partum) and route of delivery. Neonataldata included gestational age at delivery, birthweight, Apgarscore, cord pH and neonatal complications including detailsregarding hypoglycaemia, respiratory distress and congenital

anomalies. The Auckland Ethics Committee deemed thepresent project a clinical audit.

Respiratory distress was defined as requiring 4 h or overof respiratory support in a 24-h period. As our nursery hasboth level II and level III beds and the exact location ofadmission could depend on space or staffing issues, all neonataladmissions were combined and called NICU admissions.

Data are presented as mean ± standard deviation ifnormally distributed, or as median (range) as appropriate. Fordata analysis , χ2, Fisher’s exact test, Mann–WhitneyU-test or Student’s t-test were used as appropriate.

Results

For the study period, 427 pregnancies were identified inwomen with GDM or type 2 DM. In total, 382 babies wereborn to women with GDM and 60 to women with type2 DM. In the GDM group 112 (29%) were admitted toNICU and in the type 2 DM group 24 (40%) were admitted,giving a total of 136 infants reviewed for the present report.Three of these infants were transferred from other hospitals.There were eight sets of twins in the GDM group.

Maternal characteristics are outlined in Table 1. Datafor the GDM group are presented as the whole group and,when post-partum GTT result was documented, accordingto post-partum reclassification. Pacific Island women withGDM were most likely to have persistent glucose intolerancepost-partum, with Indian women also at increased riskcompared with Caucasian women.

Infant outcomes according to maternal diagnosis aresummarised in Table 2. In the total group, 58% were deliveredby lower segment Caesarean section (LSCS) at a mediangestation of 37 weeks (range: 25–41). Forty-six percent weredelivered preterm, with the highest rates in women withGDM who had IGT or type 2 DM post-partum. One-thirdof women with type 2 DM diagnosed antenatally or post-partum had infants weighing more than 4000 g. At delivery,neonatal resuscitation was not required in 64 (47%) infants,facial oxygen only was given in 29 (21%) and positivepressure ventilation in 43 (32%) of infants. The median(range) Apgar score at 1 and 5 min was 9 (1–10) and 10 (2–10), respectively. Cord pH was measured in 83 infants witha mean result of 7.23 (SD 0.1). Two term deliveries werecomplicated by shoulder dystocia. In one, the mother hadrecognised type 2 diabetes and the baby weighed 4245 g at37 weeks. The infant had a fractured humerus and Apgarscores were 3 and 8 at 1 and 5 min. In the other, the motherhad insulin-treated GDM with suboptimal control (and nopost-partum GTT). The infant weighed 4345 g at 40 weeksand had Apgar scores of 2 and 2 at 1 and 5 min withmeconium exposure at delivery. This infant was dischargedfrom the NICU after 1 day.

Hypoglycaemia was the most common complication, beingdocumented in 51% of infants admitted to NICU. Overall68% of admitted infants received intravenous dextrose,which may have prevented hypoglycaemia occurring in someinfants. In the non-hypoglycaemic cases the indication was

Neonatal morbidity in GDM and type 2 DM

Australian and New Zealand Journal of Obstetrics and Gynaecology 2003; 43: 429–432 431

coexisting morbidity such as respiratory distress or prematu-rity. Respiratory distress was the second most common short-term morbidity, occurring in 40% of infants admitted to theNICU. The rates of respiratory distress in the preterm andterm infants of 39% (26/67) and 43% (31/70), respectively,were not significantly different (P = 0.34) and the mean (SD)gestation of infants with respiratory distress was 36.1 weeks(3.2). No infants were discharged on oxygen. Five terminfants, four in the GDM group and one in the type 2 DMgroup, were described as having respiratory distress withmeconium exposure. Two infants, one in each group hadpulmonary hypertension treated with nitric oxide.

Congenital anomalies were detected in nine infants, ofwhich one was transferred from another centre. Details areshown in Table 3.

Discussion

Neonatal morbidity is common following a pregnancy com-plicated by either GDM or type 2 DM. In the present seriesNICU admission occurred in 29% of GDM and 40% of type2 DM pregnancies. Of infants admitted to NICU, 51% had hypo-glycaemia and 40% required support for respiratory distress.

Delivery by emergency Caesarean section and prematuritywere predictors for NICU admission. The emergency LSCSrate of 38% compares with a rate of 25% in the total groupof women with GDM and type 2 DM delivered during thestudy period (NWH clinic data). The elective LSCS rate wasthe same (20%). Almost half the admissions were in preterminfants, compared with a 19% preterm delivery rate in womenwith GDM and type 2 DM during the study period.9

Table 1 Maternal characteristics complicated by GDM or type 2 DM identified from the National Women’s Hospital database fromJanuary 1999 to December 2000

Antenatal diagnosis GDM post-partum diagnosis†Type 2 DM GDM Normal IGT Type 2 DM

Women (n) 24 104 27 12 19Maternal age mean (SD) 32.8 (4.8) 32.2 (5.3) 32.6 (4.7) 34.2 (5.1) 30.9 (4.2)*Ethnicity known (n) 19 91 22 11 19Pacific Island 12 (63%) 38 (42%) 2 (9%) 2 (18%) 13 (68%)European 3 (16%) 23 (25%) 10 (45%) 2 (18%) 0Indian 3 (16%) 17 (19%) 7 (32%) 2 (18%) 3 (16%)Maori 1 (5%) 5 (5%) 2 (9%) 0 3 (16%)Chinese 0 6 (7%) 0 4 (36%) 0Other 0 3 (3%) 1 (5%) 1 (11%) 0

†Post-partum diagnosis: 46 of 104 women (44%) did not carry out post-partum glucose tolerance test (GTT), or results not found. In the remaining, outcome is shown according to results of postpartum 75 g oral GTT. *Women with type 2 diabetes mellitus (DM) were younger than the women with impaired glucose tolerance (IGT) post-partum (P < 0.05). GDM, gestational diabetes mellitus.

Table 2 Outcomes of infants admitted to neonatal unit

Antenatal diagnosis GDM post-partum diagnosisType 2 DM GDM Normal IGT Type 2 DM

Infants (n) 24 112 30 14 19Delivery by LSCS 14 (59%) 65 (58%) 14 (52%) 8 (57%) 13 (68%)

Emergency 10 (42%) 42 (38%) 10 (37%) 7 (50%) 8 (42%)Elective 4 (17%) 23 (21%) 4 (15%) 1 (7%) 5 (26%)

Gestation (weeks) 37 36 37 35 36Median (range) (33–39) (25–41) (32–41) (33–40) (27–39)Preterm <37 weeks 10 (42%) 53 (47%) 12 (40%) 12 (71%) 12 (63%)Birthweight (g) 3601 3028 (1042) 2918 2452 3277*Mean (SD) (901) (703) (580) (1245)Weight >4000 g 8 (33%) 17 (15%) 2 (7%) 0 6 (32%)**Hypoglycaemia (<2.6) 14 (58%) 55 (49%) 15 (55%) 6 (50%) 10 (53%)Intravenous dextrose 18 (75%) 74 (66%) 18 (60%) 7 (50%) 15 (79%)Respiratory distress 12 (50%) 43 (38%) 8 (30%) 5 (42%) 9 (47%)Congenital anomaly 2 7 3 0 3Days in NICU 3 (1–14) 3 (1–93) 3 (1–25) 2 (1–14) 3 (1–93)Median (range)

Results expressed as number (percentage) unless otherwise stated. *Birthweight for type 2 diabetes mellitus (DM) significantly higher than for impaired glucose tolerance (IGT) group P < 0.05; **birthweight >4000 g significantly more common for type 2 DM than IGT or normal group P < 0.05. GDM gestational diabetes mellitus; LSCS, lower segment Caesarean section; NICU, neonatal intensive care unit.

D. Watson et al.

432 Australian and New Zealand Journal of Obstetrics and Gynaecology 2003; 43: 429–432

Birthweight over 4000 g was highest in the group ofwomen with type 2 DM diagnosed antenatally or post-partum. This may, at least in part, be because of the increasedpercentage of Pacific Island women in these groups. Theyhave a relatively large body habitus and one would expectthem to have larger babies than Caucasians and, in particular,Asian women. Despite the large infants, shoulder dystocia andneonatal complications because of traumatic delivery wereuncommon. Unfortunately it was not possible to correlatematernal glycaemic control with neonatal outcomes, as accur-ate maternal data were not available in the clinical records.

The main neonatal complications necessitating NICUadmission were hypoglycaemia and respiratory distress.For respiratory distress, the rate in term infants was greatlyincreased compared to a background rate of less than 0.5%in term infants admitted to neonatal units.13

The rate of major congenital anomaly, requiring NICUadmission, in the study period was 1.8%, excluding theinfant transferred from another centre. This was increased inwomen with type 2 DM diagnosed either prepregnancy orpost-partum. Our centre has already reported an increasedrisk of major anomaly in a larger series of women withprepregnancy type 2 DM (4.4% confidence intervals (CI)2.4–7.3) and postnatally diagnosed type 2 DM (4.6% CI2.3–8.2).4

The present report demonstrates that neonatal morbidityis high in women with type 2 DM and also GDM when there isa high background prevalence of type 2 DM. Pre-term deliveryis an important contributing factor. Careful assessment forcongenital anomalies should be made in offspring of womenwith GDM who are likely to have undiagnosed type 2 DM.

References

1 Yamashita H, Shao J, Friedman J. Physiologic and molecularalterations in carbohydrate metabolism during pregnancy and

gestational diabetes mellitus. Clin Obstet Gynecol. 2000; 43:87–98.

2 Kjos S, Buchanan T. Gestational diabetes mellitus. N Engl JMed. 1999; 341: 1749–1756.

3 Langer O. Is normoglycemia the correct threshold to preventcomplications in the pregnant diabetic patient? DiabetesReviews. 1996; 4: 2–10.

4 Farrell T, Neale L, Cundy T. Congenital anomalies in theoffspring of women with type 1, type 2 and gestational diabetes.Diabet Med. 2002; 19: 322–326.

5 Schaefer U, Songster G, Xiang A, Berkowitz K, Buchanan TA,Kjos SL. Congenital malformations in offspring of womenwith hyperglycemia first detected during pregnancy. Am JObstet Gynecol. 1997; 177: 1165–1171.

6 Coustan DR. Screening and diagnosis of gestational diabetes.Semin Perinatol. 1994; 18: 407–413.

7 Shushan A, Ezra Y, Samueloff A. Early treatment of gesta-tional diabetes reduces the rate of fetal macrosomia. Am JPerinatol. 1997; 14: 253–256.

8 Langer O, Rodriguez DA, Xenakis EM, McFarland MB,Berkus MD, Arrendondo F. Intensified versus conventionalmanagement of gestational diabetes. Am J Obstet Gynecol.1994; 170: 1036–1047.

9 Annual Clinical Report 2000. National Women’s Hospital, Auck-land, New Zealand. Auckland: NWH, 2001.

10 Lind T, Phillips PR. Influence of pregnancy on the 75gOGTT: a prospective multi-centre study. The DiabetesPregnancy Study Group of the European Association for theStudy of Diabetes. Diabetes. 1991; 40: 8–13.

11 Feig D, Palda V. Type 2 diabetes in pregnancy: a growingconcern. Lancet. 2002; 359: 1690–1692.

12 Newborn Services Medical Guidelines [Internet]. Auckland:Auckland District Health Board; [updated 16 July 2003; cited29 June 2003]. Available from: http://adhb.co.nz/newborn/MedicalGuidelines/HypoglycaemiaManagement.htm

13 Donoghue D and the ANZNN. The Report of the Australianand New Zealand Neonatal Network, 2000. Sydney: ANZNN,2002.

Table 3 Congenital anomaly detail

AnomalyAntenatal maternal

diagnosisHbA1c at booking

(gestation)Postnatal maternal

diagnosis

TGA Type 2 DM 10.3% (10 weeks) Type 2 DMArachnoid cyst with Type 2 DM No HbA1c Type 2 DMhydrocephalusBranch pulmonary stenosis GDM 8.3% (25 weeks) Type 2 DMDouble outlet left ventricle GDM 11.2% (35 weeks) Type 2 DMMulticystic kidney GDM 10.0% (36 weeks) Type 2 DMASD, VSD, PDA GDM No HbA1c NormalVSD GDM No HbA1c NormalVATER GDM No HbA1c NormalHypoplastic left ventricle GDM (transfer) 5.1% No data

ASD, atrial septal defect; HbA1c, haemoglobin A1c, PDA, patent ductus arteriosus; TGA, transposition of the great arteries; VATER, vertebral defects, imperforate anus, tracheoesophageal fistula, and radial and renal dysplasia sequence; VSD, ventricular septal defect.