Gestational Diabetes Mellitus Among Pregnant Women Attending Ante-natal Clinic at a Secondary Care Health Facility in Haryana, India

Introduction: Gestational diabetes has serious health effects during pregnancy and childbirth. We estimated the occurrence of gestational diabetes mellitus (GDM) among pregnant women in a secondary care hospital in Haryana. Methods: It was a hospital-based cross-sectional study, done in an ante-natal clinic (ANC) at a sub-district hospital (SDH), Faridabad district of Haryana, India. Eligible pregnant women attending the ANC clinic were recruited. An oral glucose tolerance test (OGTT) with 75 g of glucose was done with a collection of blood for fasting blood sugar (FBS) and two-hour post-OGTT blood glucose. A pre-tested semi-structured interview schedule was administered. Both the modified International Association of the Diabetes and Pregnancy Study Groups criteria (IADPSGC) and the Diabetes in Pregnancy Study Group of India (DIPSI) criteria were used. Data were presented as percentages, means, standard deviation, and 95% confidence interval (CI). Bi-variable and multi-variable logistic regressions were done. The level of significance was set at 0.05. Results: Of the 623 eligible participants, 66.1% were within the 20-25 age group. The GDM was found in 14.1% (95%CI: 11.5-17.1) participants as per modified IADPSG criteria and 6.7% (95%CI: 4.9-9.0) participants as per DIPSI criteria, respectively. Increasing age [adjusted odds ratio (AOR): 1.24 (95% CI: 1.05-1.47), p=0.008] and increasing years of schooling [AOR: 1.19 (1.01-1.41), p=0.032] were significantly associated with GDM by DIPSI criteria. Family history of DM was also found to have an increased odds with GDM using modified IADPSG criteria [AOR 2.87 (95% CI: 1.09-7.54), p=0.032]. Conclusion: Considerable proportion of pregnant women were found to have GDM in a Sub-district hospital at Ballabgarh in north India. The study highlighted the need and generated evidence about the feasibility of GDM screening utilizing routine staff in a secondary care facility.


Introduction
Gestational diabetes mellitus (GDM) is defined as glucose intolerance with onset or first recognition during pregnancy, which is not diagnostic of overt diabetes [1]. Although asymptomatic in its clinical course, GDM, if not timely identified and adequately managed, may lead to serious adverse obstetric and neonatal outcomes [2,3]. In the long run, both mother and child have higher chances of developing type 2 diabetes mellitus, and babies are more likely to become obese later in life [4,5].
The global prevalence of GDM varies widely, depending on population characteristics (e.g., maternal age, socioeconomic status (SES), race/ethnicity, or body composition), screening methods, and diagnostic criteria. As per the International Diabetes Federation, an estimated 21.1 million (16%) live births had some form of hyperglycemia in pregnancy in the year 2019, with the highest prevalence (28%) estimated in South East Asia (SEA) [6]. As per a systematic review, the prevalence of GDM was found to be 10.1% in SEA [7].
The International Association of Diabetes and Pregnancy Groups (IADPSG) proposed a set of criteria to diagnose GDM. According to these criteria, the diagnosis of GDM is made if at least one value of plasma glucose concentration is equal to or exceeds the thresholds of 92, 180, and 153 mg/dl for fasting, one-hour, and two-hour post glucose load glucose values, respectively, after performing a 75-g oral glucose tolerance test (OGTT) [8]. The World Health Organization accepted these criteria for screening for GDM [9]. The Diabetes in Pregnancy Study Group of India (DIPSI) criteria were used in guidelines on diagnosis and management of gestational diabetes mellitus by the Government of India and it recommended OGTT at 24-28 completed weeks of gestation to diagnose GDM [10]. As per DIPSI criteria, two-hour OGTT ≥ 140 mg/dl is considered GDM [11].
Increasing mechanization and urbanization of rural populations have been associated with an increased occurrence of type 2 diabetes in rural areas of India [12]. Undetected and untreated GDM will add to the increasing burden of non-communicable diseases in the country [13]. However, adequate information is not available about the study area to take necessary action to prevent and minimize short-term and long-term maternal and fetal morbidity associated with GDM. The current study was done to ascertain the occurrence of GDM and study the socio-demographic and other factors associated with it among pregnant women with 24-28 completed weeks of gestation (POG) attending the ante-natal clinic (ANC) in a secondary healthcare facility in Haryana, India by modified IADPSG and DIPSI criteria. We additionally compared the results obtained through the two criteria.

Study design
This is a hospital-based cross-sectional study.

Study participants
Pregnant women with 24-28 completed weeks of gestation attending an antenatal clinic (ANC) at a a subdistrict hospital (SDH) in Ballabgarh, Faridabad, Haryana were recruited for the study. Pregnant women who were suffering from chronic renal, pancreatic, and other severe illnesses; receiving steroids, nicotinic acid, or other medication likely to cause dysglycemia; and those with previously diagnosed diabetes mellitus were excluded from the study.

Study setting
The study was conducted in an ANC at Ballabgarh SDH. This facility is a 50-bed secondary care hospital, which was part of the

Methodology
Eligible pregnant women attending the ANC clinic at SDH Ballabgarh were given an appointment for an OGTT test and asked to come after overnight fasting. A nurse posted at the ANC clinic explained the whole procedure in detail to all eligible study participants and obtained consent for participation. On the scheduled dates when the study participants came for OGTT, a pre-tested semi-structured interview schedule was administered to all eligible pregnant women to collect information on their socio-demographic profile, relevant medical and obstetric history, previous history of GDM, family history of diabetes mellitus, and awareness regarding GDM. A modified BG Prasad socio-economic scale was used to assess the socioeconomic status of study participants [14]. Anthropometric measurements including body weight (to the nearest 0.1 kg; Equinox Personal Weighing Scale-Digital EQ-EB-9300) and height (to the nearest 0.1 cm; Seca 213 portable Stadiometer) were recorded. Blood pressure was recorded thrice at an interval of five minutes in a seating position in the non-dominant arm (to the nearest 02 mmHg) using a standard digital sphygmomanometer (Omron HEM 8711). The mean of the three readings was taken as the blood pressure. These measurements were done by a single trained nurse, and standardization, along with quality assurance measures, was undertaken through supportive supervision by the study team.

Sample size
A minimum of 616 pregnant women was required in the study assuming a prevalence of GDM of 13.9% [15], a two-tailed alpha error of 5%, power of 90%, and 20% of relative precision. All pregnant women of 24-28 weeks of gestation, presenting with antenatal OPD of SDH Ballabgarh during the study period, were offered to be part of the study.

Oral glucose tolerance test
All eligible study participants were subjected to an OGTT with 75-g anhydrous glucose powder. The glucose powder was dissolved in 250-300 ml of water and pregnant women were asked to consume it in five minutes. Two millilitres (ml) of blood samples were taken aseptically from the ante-cubital vein twice for estimation of fasting and two-hour post-glucose blood sugar levels in a biochemical analyzer (Biolis 24i Premium) and HbA1C in another automated analyzer (Adams Arkray, Model HA-8180, Japan that estimates using highperformance liquid chromatography). If the pregnant woman vomited within 30 minutes of consumption of glucose, the OGTT test was repeated on the next ANC day of the clinic or the following week. This was allowed a maximum of three attempts. In the case of vomiting after 30 minutes, the OGTT test was completed on the same day, as per the guidelines of the government of India [10]. Instruments were calibrated regularly for the purpose of quality assurance.

Modified IADPSG Criteria
Any one value of plasma glucose concentration equal to or exceeding the thresholds of 92 and 153 mg/dl (for fasting and two-hour post glucose load values, respectively), after performing a 75 g OGTT, was considered as GDM.

DIPSI Criteria
A blood glucose level of ≥140 mg/dl two hours after the consumption of 75 mg of anhydrous glucose was considered GDM.

Ethics
Ethical approval was obtained from the Institutional Ethics Committee, AIIMS, New Delhi (vide Letter Reference No. IEC/649/11/2018). Informed written consent was taken from each study participant after reading and explaining the participant information sheet to them in their local language. Pregnant women diagnosed with GDM were managed as per national guidelines of the government of India by the obstetrician in our study hospital [10].

Statistical analysis
The data were entered in a Microsoft Excel sheet (Microsoft® Corp., Redmond, WA) and analyzed with the help of STATA 12 statistical software (StataCorp LLC, TX, USA). Categorical data were presented as percentages (%). Means and standard deviations were used to present normally distributed data. Binary logistic regression analysis (stepwise method) was used to evaluate the independent associations of various factors with the prevalence of GDM. The multivariable logistic analysis was done for all the variables after the bi-variable analysis. A "p" value of less than 0.05 was considered statistically significant.

Results
A total of 623 out of 690 pregnant women were recruited for the study and invited to participate during the study duration, with a response rate of 90%. The mean (SD) age of the study participants was 23.8 (3.5) years (95%CI: 23.5-24.1). Two-thirds of the study participants were in the age group of 20-25 years (66.1%). The mean (SD) number of years of schooling among study participants was 10.0 (4.6) years (95%CI: 9.6-10.4). Almost all (95%) of the study participants were housewives. The median (interquartile range) per capita family income was INR 3400 (2500-5000). Details of the socio-demographic characteristics of study participants have been described in Table 1.  Almost one-third of the study participants were primigravida (36.9%). The mean (SD) period of gestation of study participants was 25.6 (1.7) weeks. Among multigravida study participants (n=393), nine study participants (2.3%) reported stillbirths in the last pregnancy. Around one in every ten study participants (11.2%) had undergone cesarean delivery in their last pregnancy. Almost 10% of the study participants reported a history of diabetes mellitus among first-degree family relatives. A total of 280 (44.9%) study participants reported at least one high-risk factor (as mentioned in the study appendix, like age <18 or >35 years, weight <45 kg, height <145 cm, grand multipara, pre-eclampsia, severe anaemia, antepartum haemorrhage, multiple pregnancies, abnormal presentation, etc.) in the current pregnancy, whereas 26.0% of pregnant women reported having at least one high-risk factor in the last pregnancy. Only two pregnant women reported that they were diagnosed with a case of GDM in an earlier pregnancy. A total of 295 pregnant women recalled the birth weight of the youngest babies. The mean (SD) weight of the youngest babies was 2772.0 g (593.5; n=295). A history of big babies (birth weight ≥ 3500 g) was reported by 53 (8.5%) study participants (n=295). The mean height (SD) and weight (SD) of study participants were 150.8 (6.3) cm and 54.9 (9.3) kg, respectively. Pre-pregnancy weights were available for 502 study participants. So, we calculated pregnancy weight gain (current weight -pre-pregnancy weight) for these participants. The median (IQR) pregnancy weight gain by study participants was 5.0 (3.0-7.1) kg.
Out of a total of 623 participants, 77 (12.4%) had fasting plasma glucose more than equal to 92 mg/dl with a mean value of 96. 6   In the multivariable analysis (

Discussion
A total of 623 pregnant women were screened to determine the prevalence of GDM and its associated factors. The prevalence of GDM was 14.1% by modified IADPSG criteria and 6.7% by DIPSI criteria. A wide range of prevalence of GDM (1-42%) has been reported by studies conducted in different parts of India [16].
Major reasons for such a wide variation in reported prevalence were the different diagnostic criteria used in the studies and differences in the socio-demographic profile of the study population. The American Diabetic Association (ADA), the IADPSG, and the DIPSI are various professional bodies that have suggested different cut-off values for diagnosing GDM in pregnant women. A study conducted in Gujarat reported that the estimated prevalence of GDM using capillary blood was higher in comparison to the estimated prevalence of GDM using venous blood (20.4% vs 11.5%) [17]. In our study, we used a venous blood sample for the estimation of blood glucose. This might be a reason for the low prevalence of GDM in our study. Our prevalence estimate of 14% using IADPSG criteria matches with the prevalence estimate reported in another hospital-based study from a South Indian setting utilizing the same criteria [18].
Both the IADPSG and DIPSI criteria have been used in this study. Internationally, IADPSG criteria are more widely accepted, though the government of India has accepted DIPSI guidelines for use in programmatic settings due to feasibility and logistical concerns. In our study, a higher prevalence of GDM was obtained using IADPSG criteria. A meta-analysis of Indian studies that pooled estimates for GDM using IADPSG criteria reported a relatively higher prevalence than other criteria at 19.2% (95% CI: 15.5, 23.6) [16]. The same study reported the pooled estimate emanating from studies that utilized DIPSI criteria as 7.4% (95% CI: 5.2, 10.2). Another systematic review that included studies determining the prevalence of GDM in South Asian countries also reported a higher prevalence estimate of GDM using IADPSG criteria (20.9%, 95% CI: 17.3, 24.6) compared to studies that used DIPSI criteria (8.3%, 95% CI: 5.7, 10.9) [19]. In our study also, we found that the prevalence of GDM was higher with IADPSG criteria than with DIPSI. The level of agreement between the two criteria in the study was found to be poor (kappa = 0.31). A study of 144 South Indian women reported sensitivity and specificity as 45% and 87%, respectively [20].
We found that pregnant women of increasing age had a higher chance of having GDM. Higher maternal age is a known factor associated with GDM. Studies conducted in different parts of India also reported similar findings [16,21]. With increasing age, individuals may become less physically active and develop other risk factors as well. This might be the possible reason for such an association.
There was a statistically significant association between a positive family history of diabetes among firstdegree relatives and increased odds of GDM on bi-variable and multivariable analysis. A family history of diabetes has some genetic predisposition to the development of GDM [22,23]. Our study result is consistent with the previous study done by Rajput et al. in a tertiary care hospital in Haryana among 607 ANC mothers from 24 to 28 weeks of gestation where the crude OR was 2.35 (95% CI: 0.99-5.6, p<0.05) though it was not significant in multivariable analysis [24]. A meta-analysis was done by Lee et al. among the Asian population, which also showed an OR of 2.77 (95% CI: 2.22-3.47), which is similar to our study findings [19].
Some Indian studies reported an increased chance of GDM among study participants with higher socioeconomic status [24], mostly due to a more sedentary lifestyle and a higher pre-pregnancy weight among study participants from higher SES. Our study did not find this association significant. One study done in China did not find any association between SES and GDM [25]. In fact, one study from China reported an inverse association between SES and GDM [26]. In many countries, people from higher socioeconomic strata are found to be more physically active and lead a healthy lifestyle [27]. This might be the possible explanation for an inverse association between SES and GDM in their studies.
Years of schooling were found to be associated with an increased odds of GDM. Though the longer duration of education may also be associated with a higher probability of higher maternal age, a modern lifestyle, and less physical activity among study participants, our study did not find the association. A study conducted in Rotterdam, the Netherlands, discovered that a lower educational level was associated with a higher prevalence of GDM. However, the study also reported that a large part of the observed increased risk of GDM in the lower educated group was due to relatively higher rates of overweight and obesity among those women [28].
This was a hospital-based study, yet in a secondary care facility that was closer to community settings. Blood was collected by a trained nurse; this added to the strength of our study. Venous blood was collected for blood glucose estimation. All quality control measures were followed for laboratory investigations, which minimized the measurement errors. Our study had a few limitations also. This was a single-centre study in a peri-urban setting in the northern part of the country. The findings might not be generalizable to other settings, and thus a multi-centric study would be needed to establish prevalence in different regions of the country. We did not measure blood glucose during the first trimester of pregnancy and kept a single visit during the last 24-28 weeks as the basis for estimating prevalence. Also, our study was not powered sufficiently to ascertain associations with all factors, as primarily our study was designed to ascertain the occurrence of GDM among pregnant women attending a secondary care hospital. We also did not include anthropometry and blood pressure measurements for determining associations with GDM and future studies in this regard can study these factors in-depth and measure these associations in a larger sample of pregnant women.

Conclusions
We have reported the levels of GDM using two criteria-one that is more internationally accepted and the other that is recommended by the Government of India for incorporation in the programmatic settings. It has been recommended now that an OGTT be performed for all pregnant women visiting ANC clinics in routine district health systems within India. However, OGTT tests are not being done routinely, even in secondary care hospitals. With such a high burden of GDM among pregnant women, all health care settings need to include OGTT in essential obstetric care packages. Our study has demonstrated that it is feasible to perform this in a government secondary care facility. In our sample, 14% and 7% of pregnant women attending the ANC clinic at SDH Ballabgarh had GDM as per the IADPSG and DIPSI criteria, respectively. The latter criteria under-reports the prevalence of GDM. Increasing age, higher years of education, and a positive family history of diabetes were associated with an increased risk of GDM. Future studies should also track prospectively the maternal and neonatal outcomes in these women detected with GDM.