Shenaz Ramtoola, Specialist/Senior Specialist, Dr A.G. Jeetoo Hospital, Port Louis, Mauritius. Correspondence: Consultant Physician, Royal Infirmary, Blackburn BB2 2LR;
Tel 01254 263555; Fax 01254 294423; email: sramtoola@mcmail.com

Philip Home, Consultant Physician and Professor of Diabetes Medicine, Human Diabetes and Metabolism Research Centre, University of Newcastle upon Tyne.
Correspondence: Department of Medicine, Framlington Place, Newcastle upon Tyne, NE2 4HH;
Tel 0191 222 7019; Fax 0191 222 0723; email: Philip.Home@newcastle.ac.uk

Hassen Damry, Specialist/Senior Specialist, Dr A.G. Jeetoo Hospital, Port Louis, Mauritius. Correspondence: 12, Reverund Lebrun St, Mauritius; Tel 00 230 454 5928

Anwar Husnoo, Specialist/Senior Specialist, Dr A.G. Jeetoo Hospital, Port Louis, Mauritius. Correspondence: Dr A.G. Jeetoo Hospital, Port Louis, Mauritius; Tel 00 230 212 3201

Stephen Ah-Kion, Specialist/Senior Specialist, Dr A.G. Jeetoo Hospital, Port Louis, Mauritius Correspondence: Dr A.G. Jeetoo Hospital, Port Louis, Mauritius; Tel 00 230 212 3201.

Abstract

Mauritius has a prevalence of diabetes of 13.2 % and of Impaired Glucose Tolerance of 15.6 %. Outcome of diabetic pregnancy in Mauritius was evaluated prospectively over a period of three years at Dr A.G. Jeetoo Hospital, Port Louis, one of four regional hospitals of Mauritius. Between October 1993 and September 1996, 294 diabetic pregnancies were registered, of which 110 were in women who had pregestational diabetes; diabetes was newly diagnosed in pregnancy in 86 women, and IGT in a further 98 women. Perinatal mortality in 267 pregnancies which resulted in live birth or proceeded to at least 28 weeks gestation was 124/1000 for pregestational diabetes and 116/1000 for gestational diabetes, compared to the population rate of 26/1000 for 1993-1995 (RR 4.7 (95 % CI 2.7-8.2) and RR 4.4 (2.5-7.9), respectively). There was no excess perinatal mortality for gestational IGT at 22/1000. Limited resources should be concentrated on improving the care of pregnant women with known diabetes and on the early diagnosis and treatment of new or unknown diabetes in pregnancy. The value of identification and treatment of IGT in pregnancy remains doubtful.

Key words: diabetes; pregnancy; Gestational Diabetes Mellitus; Impaired Glucose Tolerance; perinatal mortality; caesarean section; birth weight; screening; Mauritius

Key Messages

Perinatal mortality in pregnancy with established diabetes remains at 4 to 5 times background rate in most developing as well as some advanced countries; lack of pre-pregnancy care and late presentation in pregnancy with poorly controlled diabetes may be more likely with Type 2 diabetes.

Newly diagnosed diabetes in pregnancy carries essentially the same risks as established diabetes. In this study, excess perinatal mortality was exclusively in women with frank diabetes with fasting hyperglycaemia ³ 7.8 mmol/l at diagnosis.

IGT diagnosed in pregnancy did not carry excess perinatal mortality risk. However, women with pregnancy diagnosed IGT had a three-fold increase in caesarean section rate.

In a population with high Type 2 diabetes and IGT prevalence, as well as a high prevalence of low birth weight infants, the value of identifying and treating IGT in pregnancy as for diabetes remains doubtful. It may be that in such a population, IGT in pregnancy may confer a survival advantage to the neonate from increased birth weight.

The oral glucose tolerance test may not be the most appropriate test for screening for diabetes selectively rather than for total glucose intolerance in pregnancy; fasting plasma glucose may more effectively select women at highest risk for perinatal mortality.

INTRODUCTION

Mauritius, a multi-ethnic island in the Indian Ocean, is known to have a high prevalence of diabetes and impaired glucose tolerance (IGT). The prevalence rates in women aged 25-34 and 35-44 years at the time of the present study were 3.0 and 10.5 % for diabetes, and 10.5 and 19.9 % for IGT, respectively.¹ Little is known about pregnancy with diabetes or IGT in these women. There are no data on the prevalence of pregnancy with diabetes, nor on pregnancy outcome in diabetic pregnancy. At the second national Non-Communicable Disease survey in 1992, 8.3 % of women who had newly diagnosed diabetes and 8.7 % of women with known diabetes at the previous 1987 survey reported one or more stillbirths in the intervening 5-year period, as did 4.7 % of those with IGT and 5.5 % of women with normal glucose tolerance at baseline.¹ In an analysis of perinatal mortality at one regional hospital,² of 162 cases of perinatal mortality recorded over one year (June 1991 to May 1992), 3.7 % of the mothers had been recorded as having diabetes. The full extent of the effects of maternal glucose intolerance on pregnancy outcome in these circumstances needed to be further evaluated.

Study design, site and population

The maternity services of the Ministry of Health in Mauritius are provided on a regional basis through a network of mobile clinics, Area and Community Health Centres, and district and regional hospitals. As for all other government health services, they are available to all on a non-contributory, non-paying basis; geographical accessibility in a small island with good roads is no bar to the use of services. Most pregnant women (over 95 %) therefore receive antenatal care by qualified midwifery and medical personnel.³ The four regional hospitals at the time of study catered for over 80 % of all deliveries on the island, a further 15 % taking place in private nursing homes. Less than 2 % of women have home deliveries with assistance by traditional birth attendants.⁴ Dr A.G Jeetoo Hospital, a 420-bedded unit, is the regional hospital for Port Louis, the capital city, and its surrounding suburban and rural areas. The catchment population is estimated at around 250 000, 20-25 % of the island population. Annual deliveries of 4500 represent 22 % of all deliveries on the island.

Procedures

Ongoing and new cases of diabetic pregnancy seen at the Dr A.G. Jeetoo Hospital were registered from October 1993, at the time of the establishment of a dedicated antenatal diabetes clinic led jointly by an obstetrician and a physician. The majority of pregnant women with diabetes were seen in this clinic at some time during the pregnancy; diabetic pregnancy in-patients were notified to and as far as possible managed by the same named physician and named obstetrician; obstetric requests for outpatient glucose tolerance tests (OGTT) were carried out by the antenatal diabetes clinic nurses and the results reviewed by the clinic medical staff, while inpatient antenatal OGTT results were also generally assessed by the antenatal diabetes clinic physician; hence a comprehensive prospective register of diabetic pregnancies was maintained. Regular visits were made to the antenatal, labour, and postnatal wards to ensure that all cases of diabetic pregnancy, whether or not the women had received treatment in pregnancy, and irrespective of the attending physician or obstetrician, were registered.

Maternal data collected included, among others, maternal demographics, diabetes history, obstetric history, and history of the present pregnancy. The required information was obtained by direct history taking from the patient and examination of any pre-pregnancy hospital medical notes, current pregnancy records, as well as antenatal co-operation cards of any pre-hospital antenatal care. Further information regarding patients’ diabetes history was sought from health centres and other hospitals where relevant.

Patients were classified into three groups, pregestational diabetes, gestational diabetes and gestational IGT. Diagnostic criteria for the newly diagnosed patients were based on the WHO (1985)⁵ recommendations. Not all patients needed an OGTT. Diagnosis and initiation of therapy had sometimes been based on fasting and/or random plasma glucose (FPG and RPG) only. Minimum criteria for inclusion in the gestational diabetes and IGT groups were as follows: for gestational diabetes, 2-h plasma glucose (2-h PG) ³  11.1 mmol/l on 75 g OGTT, or any two of FPG ³ 7.8 mmol/l and RPG ³ 11.1 mmol/l; for gestational IGT, 2-h PG ³  7.8 mmol/l and < 11.1 mmol/l, or two RPG ³  7.8 mmol/l and < 11.1 mmol/l.

Pregnancy outcome was assessed as miscarriage, live birth or stillbirth, and first week neonatal mortality. All pregnancy losses including intrauterine deaths before 28 weeks gestation were classed as miscarriage, and from 28 weeks as stillbirth, to allow comparison with published local perinatal mortality data. Background stillbirth, neonatal and perinatal mortality statistics used for comparison were those of 1993-1995 as published in the annual Ministry of Health’s Report of the Principal Medical Statistician (Health Statistics Annual).⁴

Infants of diabetic mothers were admitted to the nursery under the care of a paediatrician for observation. Data on mode of delivery, birth weight, neonatal condition at birth, first week neonatal survival and progress were retrieved from maternal and neonatal records after delivery. Neonatal survival in cases of early discharge was ascertained from the infant’s hospital follow-up attendance notes, from the mothers at outpatient clinics, by telephone or by writing to the mothers for information on the infant’s progress. Caesarean section rates were compared with the hospital rate obtained from routinely collected hospital data over the 3-year study period October 1993 to September 1996. Background population birth weight data for comparison were obtained from computerised labour ward book records over one year 1995-1996.

Data analysis

Data were analysed using Epi Info 6 software (Centers for Disease Control & Prevention, USA, and World Health Organisation, Geneva, Switzerland, 1994). Categorical data were analysed with the Fisher exact and Chi Square tests. Continuous data were analysed by one way analysis of variance. A p value of less than 0.05 was considered statistically significant.

Population and diabetes prevalence/incidence

During the 3 years October 1993 to September 1996, 294 pregnancies were registered in 270 women with diabetes or IGT at Dr A.G. Jeetoo Hospital. Of these, 110 cases were of pregestational onset, five women giving a clear history of ketoacidosis and insulin requirement from diagnosis, a further four having persistent glucose intolerance since a previous diagnosis of GDM, while the remaining women were clinically in the broad Type 2 diabetes class. Mean age of women with pregestational diabetes was 32.8 ± 5.8 years. Duration of diabetes was less than 5 years in 79 % of women, and only 6 % had had diabetes for 10 years or more. Two women had established diabetic nephropathy with proteinuria, hypertension and anaemia, and had laser-requiring retinopathy diagnosed in pregnancy. One additional patient was diagnosed with proliferative retinopathy in pregnancy. Pre-pregnancy, 57 women (52 %) were taking oral hypoglycaemic agents, and 19 (17 %) were on insulin, most women subsequently receiving insulin treatment in pregnancy, although in 21 % of these 97 women this was not until after 28 weeks gestation.

Diabetes was newly diagnosed in pregnancy in 86 women, 43 (50 %) of whom had frank diabetes with fasting hyperglycaemia of ³  7.8 mmol/l at presentation; IGT was diagnosed in a further 98 women. Diagnosis of glucose intolerance was made before 26 weeks gestation in 54 women (29 %), ten (5 %) of whom were diagnosed before 14 weeks. In ten (5 %) women, the diagnosis was only made at term. A significantly higher proportion of women with gestational diabetes, 41 %, were diagnosed before 26 weeks, compared to 19 % of those with IGT (p<0.01). Diagnosis most commonly (77 %) followed the finding of glycosuria; the presence of risk factors accounted for 11 % being tested for glucose intolerance; 5 % of cases were found through small ad hoc screening projects; 2 % had symptoms suggestive of diabetes; miscellaneous or unknown reasons accounted for the remaining 5 %. Mean age of women was 30.9 ± 5.4 years for gestational diabetes, and 29.5 ± 5.6 years for gestational IGT.

The prevalence of established and newly diagnosed diabetes in pregnancy in Mauritius was estimated at five and six per thousand deliveries respectively, gestational IGT being recognized in a further six per thousand deliveries.

Outcome by diabetes/IGT group

Six women were lost to follow-up, and two women who had not delivered by the end of February 1997 were excluded from the pregnancy outcome analyses, as was the one case of maternal mortality in the established diabetes group.

Of the remaining 285 pregnancies, 267 resulted in live birth or proceeded to at least 28 weeks gestation. The pregnancy losses before 28 weeks gestation included one case of anencephaly. There were 14 stillbirths, and, of the 253 live births, nine first week neonatal deaths were recorded (Table 1).

Perinatal fetal outcome according to diagnostic category in the 267 pregnancies resulting in live birth or stillbirth at or above 28 weeks gestation is shown in Table 2. Perinatal mortality was 124/1000 for pregestational diabetes and 116/1000 for gestational diabetes, compared to the population rate of 26/1000 for 1993-1995 (RR 4.7 (95 % CI 2.7-8.2) and RR 4.4 (2.5-7.9) respectively). There was no excess perinatal mortality in women with gestational IGT at 22/1000.

The excess perinatal mortality in women with pregestational diabetes was accounted for by a similar excess of both stillbirths (67 vs 14 per 1000; RR 4.8 (2.2-10.3)) and early neonatal mortality (60 vs 12 per 1000; RR 4.9 (2.1-11.6)). In women with gestational diabetes, the excess perinatal mortality was related particularly to an excess of stillbirths (81/1000; RR 5.7 (2.8-11.7)); the apparent difference in early neonatal mortality compared to the background rate (38/1000; RR 3.1 (1.0-9.4)) did not quite reach statistical significance (p=0.07).

All the cases of stillbirth and two of the three cases of first week neonatal death in gestational diabetes occurred in the 43 women who had frank diabetes with raised FPG ³  7.8 mmol/l at diagnosis.

Analysis of fetal outcome

Major congenital malformation accounted for three (13 %) of the cases of perinatal mortality, two of which were in women with newly diagnosed diabetes in pregnancy.

Eight out of the nine cases of first week neonatal mortality were premature infants. Prematurity was cited as a cause of death in six cases, associated with respiratory complications in five cases, including one neonate who also had multiple malformation, and associated with intraventricular haemorrhage in one case. Three of the cases were of severe prematurity (30 weeks or less). Eight of the fourteen stillbirths occurred before 37 weeks, including four at 28-30 weeks, and two others of birth weight 1500 and 1700 respectively.

Fetal maturity, birth weights and prevalence of low birth weight (< 2500 g) and macrosomia (³  4000 g) is shown in Table 3. Mean birth weight was significantly higher in neonates of mothers with gestational diabetes (3293 ± 714 g) and IGT (3083 ± 603 g) than mean population birth weight (2953 ± 567 g, p<0.001 and p<0.05 respectively). Mean birth weight in pregestational diabetes (3059 ± 641 g) was not significantly different from the population mean. The highest prevalence of macrosomia occurred in the gestational diabetes group (16 %), a highly significant increase over the background rate of 3.3 % (p<0.001), with a relative risk of 4.9 (95 % CI 3.0-8.1). The prevalence of macrosomia was also significantly increased over the background rate in women with pregestational diabetes (8 %, p<0.05), with a relative risk of 2.4 (95 % CI 1.2-4.9). In women with gestational IGT, the prevalence of macrosomia was not significantly different from the population prevalence, despite mean gestational age at delivery being highest in this group at 38.5 ± 1.8 weeks compared to 37.7 ± 2.3 weeks 37.0 ± 2.4 weeks in gestational diabetes and pregestational diabetes (p=0.01 and p<0.001 respectively). There were four infants with birth weight ³ 4500 g (2.3 %) among mothers with pregestational or gestational diabetes, representing a statistically significant increase over the 0.2 % population prevalence (relative risk 9.2, 95 % CI 3.0-29). There were no cases of birth weight ³  4500 g in the IGT group.

Delivery and infant health

Only 86 women (32%) had a spontaneous labour and normal delivery, elective delivery (induced labour and/or elective caesarean section) accounting for 56% of all deliveries, with no significant difference among the three diagnostic groups. The caesarean section rate in pregnancy with diabetes (57 %) or IGT (43 %) were significantly higher than the background hospital rate of 16 % (RR 3.5 (3.1-4.0), p<0.001 for diabetes; RR 2.7 (2.1-3.4), p<0.001 for IGT). There was no significant difference in caesarean section rate between the two diabetes groups (58 and 57 %). Of all caesarean section deliveries, 31 % were repeat sections.

Hypoglycaemia (< 1.7 mmol/l) detected by capillary blood glucose monitoring was recorded in 4 % of infants of mothers with gestational IGT group compared to 21 % of infants of mothers with pregestational diabetes (p<0.001) and 14 % of infants of mothers with gestational diabetes (p<0.05). There was no statistical association of hypoglycaemia with macrosomia. Hyperbilirubinaemia was seen in over 35 and 39 % of infants of mothers with pregestational and gestational diabetes respectively, the incidence in each of these being significantly higher than in infants of mothers with gestational IGT (21%, p < 0.05). No case of birth trauma was recorded among the macrosomic neonates of mothers with diabetes or IGT, nor in the five cases of shoulder dystocia recorded in deliveries of live born infants of mothers with diabetes. There were no cases of shoulder dystocia in gestational IGT.

A four- to five- fold increase in perinatal mortality in diabetic pregnancy, with rates of 100/1000 or more, appears to be the norm rather than the exception in most developing countries with mainly Type 2 diabetes in pregnancy.^6-10 Although the absolute rates are lower, in line with lower background population perinatal mortality rates of less than 10 %, the recent reports from the North Western and North Eastern regions in the UK of diabetic pregnancy outcome in the 1990’s^11,12 suggest that, sadly, a four- to five- fold increased perinatal mortality risk may hold true in advanced countries as well. Some authorities have expressed the opinion that the scope for further improvements in diabetic pregnancy outcome is limited.¹³ It has been suggested that patient access to and use of specialised care from early in pregnancy, if not before, is probably the determining factor for fetal losses in diabetic pregnancy, in advanced as in developing countries.^7,14

The absolute diabetic pregnancy perinatal mortality rates of over 100/1000 in Mauritius as in most developing countries to a large extent reflect the general and reproductive health of the population, and the overall quality of the existing systems of health care delivery, such that background perinatal mortality rates remain high. Pregnancy diabetes care cannot be addressed in isolation from the wider context of general and reproductive health care of all women and of the neonatal care facilities available, nor from the context of the provision of general diabetes care.

In developing countries perinatal mortality may account for a larger proportion of total fetal losses. Antenatal diagnosis of congenital anomaly followed by therapeutic termination only occurs exceptionally. Pregnancy losses for this reason may amount to 50-60/1000 in the UK, for example.^11,14 Neonatal survival of severely premature or congenitally malformed infants is also likely to be less prolonged in developing countries.

Gestational diabetes mellitus (GDM) is defined as "any degree of glucose intolerance with onset or first recognition during pregnancy".¹⁵ The recent report of the WHO Consultation has reiterated the term "gestational diabetes" to encompass both the gestational IGT and gestational diabetes groups.^16,17 The WHO (1985 and 1999) recommendations are that women with IGT should be managed in the same way as women with diabetes.^5,16,17

It is generally assumed that women diagnosed as frankly diabetic during pregnancy carry essentially the same risks as women with pregestational diabetes. This is confirmed in the present study. It is likely that in this population with high Type 2 diabetes prevalence, a significant proportion of women with diabetes diagnosed in pregnancy may have had undiagnosed pre-existing diabetes.

The implications of lesser degrees of glucose intolerance than clear diabetes on pregnancy outcome remain controversial. Whatever be the specific criteria applied for the diagnosis of GDM, worldwide, most of the cases found on screening in pregnancy are of IGT rather than diabetes.^18-22 Whilst the prevalence of total glucose intolerance in women aged 20-39 exceeds 10 % in one-third of the world populations, diabetes itself is uncommon in women of this age group, affecting less than 1 % in most populations.²³

The findings of the present study clearly indicate that diabetes diagnosed in pregnancy has significantly different pregnancy outcome implications from IGT, particularly as regards perinatal mortality. Perinatal mortality in "true" diabetes diagnosed during pregnancy was virtually identical to that in established diabetes, contributing half of all perinatal mortality in diabetic pregnancy. Pregnancy "IGT" had no implication of excess perinatal mortality, as was also evident from the reports of Roberts et al (1993)²⁰, Deerochanawong et al (1996)²¹, Li et al (1987)¹⁸, and the Diabetic Pregnancy Study Group of the European Association for the Study of Diabetes.²⁴ The only study in recent years to have shown a continuous trend towards increasing rates of perinatal mortality with increasing levels of glucose intolerance in the third trimester of pregnancy in previously non-diabetic women is that of Pettitt et al (1980) in Pima/Papago Indians.²⁵ The Pima/Papago Indians are a unique population with respect to their prevalence of glucose intolerance, and other genetic factors may be implicated.

In the present study, although mean birth weight was increased in gestational IGT, any increase in the prevalence of macrosomia did not reach statistical significance. This may simply be due to the small numbers involved. Macrosomia, the main "adverse" outcome of gestational IGT, remains difficult to influence in routine clinical practice. Whilst it has been questioned whether the increased birth weight associated with IGT is necessarily pathological,²⁶ in general low birth weight is definitely strongly associated with neonatal mortality. In developing countries, low birth weight remains the primary health concern for neonates and facilities for management of the associated problems are often deficient. It is conceivable that an increase in birth weight associated with gestational IGT, rather than being an "adverse" outcome, could confer a survival advantage to the neonate in these circumstances. In Mauritius, where mean birth weight is < 3000 g, only 3.3 % of infants weigh ³  4000 g at birth, and only 0.2 % ³  4500 g. On the other hand, the low birth weight rate (< 2500 g) is of 13 %, and "slow fetal growth, fetal malnutrition and immaturity" is the leading cause of neonatal (40.5 %) and infant (31.2 %) mortality. There is a risk of increasing the problem of low birth weight by diagnosing and treating large numbers of women with gestational IGT. It has been shown that glycaemic control that is too strict may result in intra-uterine growth retardation.²⁷

Further, studies demonstrating an association of low birth weight with later diabetes and other chronic disorders of adult life^28-32 imply that perhaps prevention of low birth weight should be prioritised over treatment lesser degrees of gestational glucose intolerance in reproductive health care in countries with high prevalence of Type 2 diabetes concomitantly with a high prevalence of low birth weight infants. It might even be argued that identification and treatment of women with mild glucose intolerance could paradoxically fuel transgenerational escalating diabetes prevalence by increasing the risk of fetal malnutrition and low birth weight infants.

Identification and treatment of glucose intolerance during pregnancy has not been shown to reduce or prevent caesarean section delivery; it may be the opposite.³³ In the present series, it is a matter of utmost concern that 43 % of women with gestational IGT, the majority of whom had maintained normoglycaemia in pregnancy, underwent caesarean section, almost three times the average hospital caesarean section rate of 16 % rate.

It appears to us that there is insufficient evidence to justify the use of limited resources in developing countries with high Type 2 diabetes prevalence to embark on screening programmes to identify women with IGT, who may represent more than 10 % of the antenatal population, and treat them in pregnancy as per the WHO recommendations; nor can it be excluded that such programmes could in some ways be detrimental. However, early identification and treatment of undiagnosed diabetes in pregnancy is clearly important. The OGTT, which will primarily identify large numbers of women with IGT may not be the most appropriate test for this purpose. The possible use of fasting plasma glucose with an appropriate cut-off level for diagnosis of diabetes in pregnancy needs to be studied further.

1. International Diabetes Institute (Australia). Mauritius Non-Communicable Disease Intervention Program. Report on 1992 disease and risk factor prevalence and incidence study, Melbourne: International Diabetes Institute, 1993.
2. Lanièce C. Descriptive analysis of perimortality survey. Port Louis: Ministry of Health, Mauritius, 1993
3. UNICEF. Situation analysis of women and children in Mauritius. Port Louis: UNICEF, 1994.
4. Principal Medical Statistician (Mauritius). Island of Mauritius Health Statistics Annual, 1995. Port Louis: Ministry of Health, Republic of Mauritius, 1996.
5. WHO Study Group. Diabetes mellitus. World Health Organization Technical Report Series 727. Geneva: WHO, 1985.
6. Kadiki OA, Rama Subba Reddy M, Sahli MA, Shawar H, Rao S. Outcome of pregnant diabetic patients in Benghazi (Libya) from 1984 to 1991. Diab Res Clin Pract 1993; 21: 39-42.
7. Huddle K, England M, Nagar N. Outcome of pregnancy in diabetic women in Soweto, South Africa 1983 to 1992. Diabetic Med 1993; 10: 290-294.
8. Johnstone FD, Nasrat AMR, Prescott RJ. The effect of established and gestational diabetes on pregnancy outcome. Br J Obstet Gynaecol 1990; 97: 1009-1015.
9. Bassaw B, Ataullah I, Roopnarinesingh S, Sirjusingh A. Diabetes in pregnancy. Int J Gynecol Obstet 1995; 50: 5-9.
10. Chirenje MZ. The effects of established and gestational diabetes on pregnancy outcome at Harare Maternity Hospital. Central African Journal of Medicine 1992; 38: 179-181.
11. Casson IF, Clarke CA, Howard CV, McKendrick O, Pennycook S, Pharoah POD, et al. Outcomes of pregnancy in insulin dependent diabetic women: results of a five year population cohort study. BMJ 1997; 315: 275-278.
12. Hawthorne G, Robson S, Ryall EA, Sen D, Roberts SH, Ward Platt MP. Prospective population based survey of outcome of pregnancy in diabetic women: results of the Northern Diabetic Pregnancy Audit, 1994. BMJ 1997; 315: 279-281.
13. Gregory R, Scott AR, Mohajer M, Tattersall RB. Diabetic pregnancy 1977-1990: have we reached a plateau? J R Coll Physicians Lon 1992; 26: 162-166.
14. Hawthorne G, Snodgrass A, Tunbridge M. Outcome of diabetic pregnancy and glucose intolerance in pregnancy : an audit of fetal loss in Newcastle General Hospital 1977-1990. Diab Res Clin Pract 1994; 25: 183-190.
15. American Diabetes Association. Expert Committee on the diagnosis and classification of diabetes mellitus. Report of the Expert Committee on the diagnosis and classification of diabetes mellitus. Diabetes Care 1998; 21 (Suppl 1): S5-S19.
16. Alberti KGMM, Zimmet PZ. Definition, diagnosis and classification of diabetes mellitus and its complications. Part 1: Diagnosis and classification of diabetes mellitus. Provisional Report of a WHO Consultation. Diabetic Med 1998; 15: 539-553.
17. World Health Organization. Definition, diagnosis and classification of diabetes mellitus and its complications. Geneva: World Health Organization, 1999.
18. Li DFH, Wong VCW, O’Hoy KMKY, Yeung CY, Ma HK. Is treatment needed for mild impairment of glucose tolerance in pregnancy? A randomized controlled trial. Br J Obstet Gynaecol 1987; 94: 851-854.
19. Lam KSL, Li DF, Lauder IJ, Lee CP, Kung AWC, Ma JTC. Prediction of persistent carbohydrate intolerance in patients with gestational diabetes. Diabetes Res Clin Pract 1991; 12: 181-186.
20. Roberts RN, Moohan JM, Foo RLK, Harley JMG, Traub AI, Hadden DR. Fetal outcome in mothers with impaired glucose tolerance in pregnancy. Diabetic Med 1993; 10: 438-443.
21. Deerochanawong C, Putiyanun C, Wongsuryrat M, Serirat S, Jinayon P. Comparison of National Diabetes Data Group and World Health Organization criteria for detecting gestational diabetes mellitus. Diabetes Care 1996; 399: 1070-1073.
22. Lao TT, Lee CP. Gestational "impaired glucose tolerance": should the cut-off be raised to 9 mmol/l? Diabetic Med 1998; 15: 25-29.
23. World Health Organization Ad Hoc Diabetes Reporting Group. Diabetes and impaired glucose tolerance in women aged 20-39 years. World Health Statistics Quarterly 1992; 45 (4):321-327.
24. Lind T, Philips PR. A prospective multicentre study to determine the influence of pregnancy upon the 75-g oral glucose tolerance test. In: Carbohydrate metabolism in pregnancy and the newborn. Sutherland HW, Stowers JM, eds. London: Springer-Verlag, 287-294, 1989.
25. Pettitt DJ, Knowler WC, Baird HR, Bennett PH. Gestational diabetes: Infant and maternal complications of pregnancy in relation to third-trimester glucose tolerance in Pima Indians. Diabetes Care 1980; 3: 458-464.
26. Dornhorst A, Chan SP. The elusive diagnosis of gestational diabetes. Diabetic Med 1998; 15: 7-10.
27. Langer O, Levy J, Brustman L, Anyaegbunam A, Merkatz R, Divon M. Glycaemic control in gestational diabetes mellitus-how tight is tight enough: small for gestational age versus large for gestational age? Am J Obstet Gynecol 1989; 161: 646-653.
28. Hales CN, Barker DJP, Clark PMS, Cox LJ, Fall C, Osmond C, et al. Fetal and infant growth and impaired glucose tolerance at age 64 years. BMJ 1991; 303: 1019-1022.
29. Hales CN, Barker DJP. Type 2 (non-insulin-dependent) diabetes mellitus: The thrifty phenotype hypothesis. Diabetologia 1992; 35: 595-601.
30. Barker DJP, Hales CN, Fall CHD, Osmond C, Phipps K, Clark PMS. Type 2 (non-insulin-dependent) diabetes mellitus, hypertension and hyperlipidaemia (syndrome X): relation to reduced fetal growth. Diabetologia 1993; 36: 62-67.
31. McCance DR, Pettitt DJ, Hanson RL, Jacobson LTH, Knowle WC, Bennett PH. Birth weight and non-insulin-dependent diabetes: thrifty genotype, thrifty phenotype, or surviving small baby phenotype? BMJ 1994; 308: 942-945.
32. Phipps K, Barker DJP, Hales CN, Fall CHD, Osmond C, Clark PMS. Fetal and impaired glucose tolerance in men and women. Diabetologia 1993; 36: 225-228.
33. Naylor CD, Sermerr M, Chen E, Sykora K. Caesarean delivery in relation to birthweight and gestational glucose tolerance: pathophysiology or practice style? JAMA 1996; 275: 1165-1170.

Table 1. Pregnancy outcome in 285 diabetic pregnancies registered at Dr A.G.Jeetoo Hospital, Port Louis, Mauritius, 1993-1996.
 

n (%)
^a Includes intrauterine death before 28 weeks gestation
^b At least 28 weeks gestation

Table 2. Stillbirth, early neonatal, and perinatal mortality rates according to diagnostic category
 

Data are rate/1000, relative risks and 95% CI compared to background population rates
^a Per 1,000 total live and stillbirths
^b Per 1,000 live births

Table 3. Fetal maturity and birth weight in women with diabetes or IGT
 

mean ± SD or n (%)
^a p<0.001, ^b p<0.05, compared to background population
^c p<0.05,^d p<0.001, compared to pregestational diabetes
^e p<0.05 compared to gestational diabet