The World Health Organization (WHO) versus The International Association of Diabetes and Pregnancy Study Group (IADPSG) diagnostic criteria of gestational diabetes mellitus (GDM) and their associated maternal and neonatal outcomes

Introduction

Gestational diabetes mellitus (GDM) is defined as any degree of glucose intolerance first detected during pregnancy [1]. It is a common medical complication in pregnancy and is associated with adverse maternal and perinatal outcomes [2]. It is of great importance to diagnose this condition early and manage adequately to prevent poor outcome.

Local studies reported that the prevalence of GDM in Malaysia was 18.3% [3] and 24.9% [4]. Currently there are various diagnostic tests and criteria used to diagnose GDM but there is lack of uniformity in making a diagnosis of GDM across the world. In the Malaysian setting, screening for GDM is done selectively for patients with one or more risk factors using the one-step 75 g oral glucose tolerance test (OGTT) and diagnosis is made using World Health Organization (WHO) diagnostic criteria. OGTT is commonly done between 24 and 28 weeks as it is the most diabetogenic period during pregnancy. However, in patients with high risk factors, it is done earlier during pregnancy for early detection and management and, if the result is negative, the test is repeated at around 24–28 weeks or at any time a new risk factor arises.

The Hyperglycemic and Adverse Pregnancy Outcome (HAPO) study was conducted to determine the level of glucose intolerance in pregnancy that is associated with adverse outcomes [5]. The study showed linear relationship between maternal hyperglycemia and increased frequency of primary and secondary outcomes which included birth weight above the 90th percentile, primary cesarean delivery, premature delivery, shoulder dystocia and pre-eclampsia [6]. The results from the HAPO study were reviewed by the International Association of Diabetes and Pregnancy Study Group (IADPSG) in order to propose a new diagnostic criteria that can be used internationally. The IADPSG recommends using 75 g, 2 h glucose tolerance test with new cut-off values for fasting, 1 h and 2 h plasma glucose as shown in Table 1 [7].

As the WHO diagnostic criteria in 1999 was not evidence based and over 10 years old, it was proposed [8] to be restructured and updated in response to the new data available. A systematic review by Wendland et al. [9] after the criteria proposed by the IADPSG was published showed both the WHO and the IADPSG criteria had similar increase in adverse pregnancy outcomes in terms of large for gestational age babies, cesarean delivery and pre-eclampsia [9]. The Atlantis Diabetes in Pregnancy Program conducted in Ireland revealed that the prevalence of GDM in a European population increased to 12.4% when using the IADPSG criteria as compared to 9.4% when using the WHO criteria [10]. There were statistically significant adverse pregnancy outcomes in the IADPSG group as compared to the WHO group [10]. A retrospective study in China comparing the ADA and the IADPSG diagnostic criteria revealed that women who were diagnosed normal using the ADA criteria but GDM using the IADPSG criteria has poor pregnancy outcome, as they were not diagnosed as GDM and managed accordingly [11]. It is expected that this group would have better pregnancy outcome if they were to be diagnosed and managed as GDM, hence they concluded that the IADPSG criteria was more suitable to be used in their population [11]. Study in New South Wales comparing characteristics and outcomes of GDM mothers in different ethnic groups showed that women from South East Asia are often diagnosed as GDM based on their elevated 2 h glucose level while their fasting sugar is the lowest compared to other ethnic groups [12]. Hence it is expected that fewer women from this region will be diagnosed with GDM when using the IADPSG criteria as the fasting sugar level was lower whilst the 2 h glucose level was higher [12].

The National Institute for Health and Care Excellence (NICE) in 2015 published their own diagnostic criteria, with fasting glucose ≥5.6 mmol/L and 2 h plasma glucose ≥7.8 mmol/L [13] which differed from the proposed criteria by the IADPSG. The Malaysian Clinical Practice Guideline (CPG) on diabetes published in 2015 adopted new diagnostic criteria, with fasting glucose ≥5.1 mmol/L and 2 h plasma glucose ≥7.8 mmol/L [14]. However, this is a mixture of different guidelines and may not represent the most suitable diagnostic criteria in our population. It is yet to be practiced universally among obstetricians in this country. The scientific research that formed the basis of the international or foreign guidelines were performed in well-resourced countries. Thus the evidence cannot be extrapolated for use in our country, which is a middle income country with a high disease burden.

At present, there is no published study in this region to compare the two diagnostic criteria. We conducted this study to establish the best diagnostic criteria that would serve to prevent adverse outcomes related to GDM and at the same time does not increase our healthcare burden.

Methodology

This was a prospective randomized study conducted within the antenatal clinic of Universiti Kebangsaan Malaysia Medical Centre from 1st February 2015 until 31st September 2017 (see Figure 1). The study was approved by the Medical Research and Ethics Committee UKMMC (Research Code: FF-2015-067).

Figure 1:

Flowchart of study procedure.

Results

A total of 520 patients were recruited during the study period. They were assigned into either the WHO or the IADPSG group, respectively. Participants in both arms were statistically comparable for age, parity, ethnicity and BMI at booking. The baseline characteristics of the participants were shown in Table 2. The mean age for the WHO and the IADPSG groups were 31.9 ± 4.57 years old and 31.1 ± 4.15 years old, respectively. Most of the mothers were in their 30s, which made up 61.2% of the subjects. About 3.8% women were more than 40 years old while the rest were in their 20s (35%). Most were Malay (78.1%), followed by Chinese (15.4%), Indian (5%) and others (1.5%) who comprised of Bajau, Iban and Kadazan. There was no significant difference in the booking BMI between the two groups. The majority were multiparous, consisting of 65.9% among mothers in the WHO group and 57.1% in the IADPSG group.

In the WHO group, 37.9% had GDM compared to 38.6% in the IADPSG group as shown on Table 3. Table 4 showed the majority required diet control followed by insulin or OHA. Only 1% of subjects in the WHO and 3% in the IADPSG group required both insulin and OHA.

Table 5 shows comparison of maternal and neonatal outcomes in patients with GDM in both groups. There was no statistically significant difference in the incidence of all measured outcomes.

Table 6 shows the incidence of maternal and neonatal outcomes among GDM and non GDM women in the WHO group. There was a significantly higher incidence of gestational hypertension or pre-eclampsia and neonatal hypoglycemia among women with GDM compared to non GDM women in this group.

Table 7 shows the incidence of maternal and neonatal outcomes among GDM and non GDM women in the IADPSG group. There was a significantly higher incidence of primary cesarean section and macrosomia in GDM women compare to non GDM women in this group.

Table 8 shows the number of participants diagnosed with GDM using the IADPSG criteria against the WHO criteria.

Table 9 shows positive and negative predictive values of the IADPSG criteria when the WHO criteria is used as the standard criteria.

Discussion

GDM has been associated with increased risk of adverse maternal and neonatal outcomes. Hence, it is important to screen and identify such women to improve the outcomes. During the first trimester, fasting blood glucose is lower and as the pregnancy progresses, there will be increased in insulin resistance [15]. This is contributed by placenta hormones excretion such as human placenta lactogen (HPL), estrogen, progesterone, cortisol and human placental growth hormone. As a result of hyperglycemia, the fetus is at risk of fetal macrosomia, birth injury at delivery particularly due to shoulder dystocia and neonatal hypoglycemia. Consequently, women with GDM are at higher risk of cesarean delivery compared to non-diabetic women. It is also associated with the development of gestational hypertension, pre-eclampsia and risk of preterm delivery.

We performed the usual two steps 75 g OGTT and women were chosen based on risk factors. Although three step blood glucose taking was suggested by the IADPSG, we chose to perform the fasting and 2 h blood glucose level to equalize the commonly used WHO criteria and prevent further increase in healthcare cost.

We found that the prevalence of GDM in our population has increased to more than 30% compared to the previously reported 18% [3] and 24.9% [4] by local studies as shown in Table 3. This is likely due to changing in our screening program where all women age 25 years and above are considered high risk. Hence, we are detecting more GDM among our women. As our center is a tertiary hospital and referral center, this might have contributed to the increased prevalence of GDM in this study as well. There was no significant difference in the prevalence of women diagnosed with GDM when IADPSG and WHO diagnostic criteria was compared. This is in contrast with most studies done in Ireland, China and India which showed higher GDM prevalence in the IADPSG group compared to the WHO group [10], [16], [17].

All GDM women received the same modalities of treatment as shown in Table 4. Although not statistically significant, there was a higher number of GDM women requiring both OHA and insulin in the IADPSG group. This could be contributed to by higher than 2 hours post prandial cut-off value rendering the GDM women to have more unsatisfactory blood sugar profile therefore requiring medication. Table 5 shows that there was no significant difference in all adverse maternal and neonatal outcomes among GDM women between the two groups. This is likely because these GDM women have received similar treatment, hence no significant difference is seen between the groups. Although the findings were not significantly different, there was a trend towards higher incidences of gestational hypertension, preterm delivery and neonatal hypoglycemia among GDM women in the WHO group. On the other hand, GDM women in the IADPSG group has a higher trend towards fetal macrosomia and primary cesarean section. There does not seem to be any difference between the two diagnostic criteria in our population. Any decision to use one over the other must rest upon other considerations, such as cost, convenience, etc.

Referring to Table 6, GDM women in the WHO group has statistically significantly increased incidences of gestational hypertension and pre-eclampsia compared to non GDM women, as expected. This may lead to higher iatrogenic preterm delivery, which can be seen in this table although it is not statistically significant. This could also contribute to the higher frequency of neonatal hypoglycemia. The significantly higher occurrences of gestational hypertension and pre-eclampsia can be considered as points in favor of the WHO criteria as a diagnostic tool, as it screens out a significantly high risk group of women besides screening for GDM.

Nevertheless, this does not happen in GDM mothers in the IADPSG group. Table 7 shows that there was no significant difference in incidence of gestational hypertension and pre-eclampsia, preterm delivery or neonatal hypoglycemia. On the other hand, the incidence of fetal macrosomia was statistically higher in the IADPSG group. This leads to the higher cesarean section rate among GDM women in this group. Absence of shoulder dystocia or birth injury was possibly because of the lower threshold to deliver via cesarean section instead of vaginal delivery. These findings are in agreement with meta-analysis which showed lowering the threshold for detection and treatment of GDM in particular the postprandial glucose level, the birth weight of babies could be reduced [18].

From the above findings, the higher incidence of fetal macrosomia in IADPSG group is likely contributed by a late diagnosis as it uses a higher 2 h post prandial value. Women were not diagnosed timely and treated as they should be. Their babies might have been affected even prior to the diagnosis. Previous studies had shown that postprandial glucose level significantly influence the birthweight of babies compared to fasting glucose level [19], [20]. Subjecting a woman to a cesarean section equals to higher rate of complications, higher cost of health care, and jeopardies the woman’s future obstetric performance. On the other hand, if birth injury occurs during an attempted vaginal delivery with a fetal macrosomia, a medicolegal lawsuit may ensue, with even higher healthcare cost. If these women were diagnosed with GDM earlier using a lower postprandial level, fetal macrosomia could have been prevented. As a result, the number of cesarean sections will be reduced, so will the incidence of shoulder dystocia and birth injury. To prevent all these, women with GDM should have been diagnosed earlier before any adverse outcome set in. It is possible that a lower postprandial value is more suitable as a diagnostic criterion in our population.

In contrast to findings in the WHO group (Table 6), although the incidence of gestational hypertension or pre-eclampsia was higher, it does not have a direct causal relationship with GDM. Other factors contributing to this include genetic factors, previous history of hypertensive disease in pregnancy and social history. Steps can be taken to reduce or delay the onset by using prophylaxis with aspirin and calcium. This will indirectly reduce the rate of preterm delivery and neonatal hypoglycemia.

When the WHO criteria is considered as standard diagnostic criteria and compared to the IADPSG criteria, it shows that the IADPSG has lower sensitivity (66%) in detecting GDM (Table 9). The negative predictive value (NPV) of 82.3% signifies there would be about 17.7% women who would be missed from being diagnosed as GDM when using the IADPSG criteria. Thus, the adoption of the IADPSG criteria among the Malaysian population is controversial as some cases of GDM may be missed. This may result in increased adverse outcomes among undiagnosed and untreated women.

This study confirmed the higher risk of women with GDM of developing various adverse maternal and neonatal outcomes compared to non GDM mothers. We consider fetal macrosomia and cesarean section as being more detrimental among all the adverse outcomes studied. No doubt it places a higher cost to health care, both in the current and in future pregnancies. By adopting the current criteria for screening (all women aged ≥25 years), more women will be screened at a lower detection rate of GDM complications if the IADPSG criteria is used.