Prevalence of gestational diabetes mellitus in India: The ICMR-INDIAB national study (ICMR-INDIAB-24)

Materials & Methods

The ICMR-INDIAB study is a nationwide, cross-sectional survey assessing metabolic non communicable diseases (NCDs) in a representative Indian sample^7-14. The study methodology of this cross-sectional study is provided elsewhere in detailed. Ethical approval was obtained from the Institutional Ethics Committee of the coordinating centre and the respective Ethics Committees of the nodal institutions implementing the ICMR-INDIAB study across different States. All participants provided written informed consent. The study was also registered with the Clinical Trials Registry of India (CTRI/2019/03/018095).

Results

Figure 1 shows the disposition of individuals in the study. The ICMR-INDIAB study included 113,043 adults, of whom 53.5 per cent were women. Of the women surveyed, 70 per cent were in the reproductive age group (20-49 yr) of whom 1,206 (2.9%) were pregnant at the time of survey. OGTT results (i.e., fasting and/or 2-h glucose measurements) were available for 1,032 pregnant women, excluding those with overt diabetes. The pregnant women studied were spread across all three trimesters. None of the pregnant women had pre-pregnancy diabetes, and only one woman reported pre-pregnancy hypertension. Supplementary table I compares the characteristics of pregnant women from the ICMR-INDIAB study with those from the NFHS-5 survey. No significant differences were observed in urban-rural distribution, literacy levels, socio-economic status, or trimester of pregnancy between the ICMR-INDIAB study and the NFHS-5 survey, confirming the representativeness of this subsample of the INDIAB survey.

Close

Fig. 1.

Flowchart depicting the participant recruitment in the Indian Council of Medical Research-India Diabetes (ICMR-INDIAB) Survey.

Export to PPT

Table I presents the prevalence (crude and weighted) of GDM in the Indian population. The weighted prevalence was 22.4 per cent, including 2.5 per cent self-reported GDM and 19.8 per cent, newly diagnosed GDM. Urban women had a slightly higher prevalence of GDM than rural women, but the difference was not statistically significant. Early GDM was diagnosed in 19.2 per cent and late GDM, in 23.4 per cent.

Supplementary table II presents the prevalence of GDM by different criteria. Using the IADPSG, TOBOGM (high glycaemic band), DIPSI criteria and NICE criteria, the prevalence rates of GDM were 36.3 per cent, 28.7 per cent, 11.4 per cent and 22.4 per cent, respectively. Trends across urban/rural areas and by Early GDM and Late GDM remained similar, irrespective of diagnostic criteria used.

The characteristics of women without GDM and those with GDM both Early GDM and Late GDM are presented in table II. The average age of the pregnant women was 26.4 yr. Those with GDM were significantly older than those without. The women with GDM were also taller, had better education (14.9% vs. 10.1%), were more professionally accomplished compared to those without GDM. Individuals with GDM had significantly higher blood pressure, glucose levels, and a greater prevalence of family history of diabetes. No significant differences were observed between those with Early GDM and Late GDM within this study population.

Region-wise prevalence of GDM in India is presented in figure 2. The highest prevalence of GDM was observed in central India (n=139, 32.9%), followed by north India (n=223, 31.4%), south India (n=121, 24.2%), east India (n=135, 17.9%), northeast India (n=334, 16.6%) and west India (n=80, 16%). The highest prevalence of Early GDM was reported in central India (34.2%) and the lowest in north India (9.2%), while the highest and lowest prevalence of Late GDM was reported in north India (39%) and northeast India (15.3%), respectively.

Close

Fig. 2.

Heatmap showing prevalence of overall gestational diabetes mellitus (GDM) in India by region. Regions details are as follows, North: Chandigarh, Delhi, Haryana, Himachal Pradesh, Punjab, Rajasthan; South: Andhra Pradesh (undivided), Karnataka, Kerala, Puducherry, Tamil Nadu; East: Bihar, Jharkhand, Odisha, West Bengal; West: Goa, Gujarat, Maharashtra; Central: Chhattisgarh, Madhya Pradesh, Uttarakhand, Uttar Pradesh; North East: Arunachal Pradesh, Assam, Manipur, Meghalaya, Mizoram, Nagaland, Sikkim, Tripura. The templates of the maps are constructed using MapChart (https://www.mapchart.net/india.html).

Export to PPT

An increasing trend in the prevalence of GDM was observed over time across the various phases of the ICMR-INDIAB study (Supplementary Table III). The GDM prevalence increased from 13.3 per cent in Phase I (2008-2010) to 30.2 per cent in Phase V (2019-2020) of the ICMR-INDIAB study. Similar pattern of increase in prevalence was observed for Early GDM (12% to 37.6%) and Late GDM (13.8% to 27.2%) over the same period.

Table III shows the multivariate logistic regression analysis of factors contribution to GDM. In the regression analysis, systolic blood pressure and family history of diabetes were found to be associated with higher risk of GDM. No differences were seen between Early and Late GDM.

Discussion

To our knowledge, this is the first nationally representative study assessing the prevalence of GDM in India. It is also the first to report on regional prevalence rates of GDM and also the first to report on prevalence of Early and Late GDM. We report the following findings (i) overall, 22.4 per cent of the pregnant women had GDM in India; (ii) the prevalence of Early GDM was 19.2 per cent and that of Late GDM, 23.4 per cent; (iii) there is heterogeneity in the prevalence of GDM in India with highest prevalence seen in central India and lowest in west India; (iv) there are no significant urban/rural differences in GDM prevalence; (v) during the various phases of the ICMR-INDIAB study done over a period of 12 yr, there appears to be an increase in prevalence of GDM and (vi) higher systolic blood pressure and family history of diabetes were associated with the increased risk of GDM.

The prevalence of GDM varies widely due to different diagnostic criteria, such as those from American Diabetes Association (ADA), WHO, IADPSG, DIPSI, NICE, and others. These criteria use distinct screening methods and thresholds, leading to inconsistent prevalence rates. This inconsistency underscores the need for standardised diagnostic approaches. Analysis of HAPO study centres showed that in Asians (mainly Chinese), NICE criteria detected more GDM, while in Whites, Hispanics, and Blacks, IADPSG identified more cases¹⁹. In this study we used NICE criteria as it aligned with OGTT fasting and 2-h glucose measures. However, for comparison, GDM prevalence was also provided using the IADPSG, the TOBOGM high glycaemic band, and the DIPSI criteria. The highest prevalence was with IADPSG (36.3%) vs. NICE (22.4%), consistent with other studies^19,21 showing IADPSG increases diagnosis rates by about 75 per cent²². DIPSI showed the lowest prevalence, likely due to lower sensitivity as reported in earlier studies^23-25.

A recent systematic review²⁶ of 110 Indian studies reported a pooled GDM prevalence of 13 per cent (urban 12%, rural 10%), but most data were hospital-based. Another review²⁷ showed a wide GDM range (0%-41.9%) with pooled prevalence by IADPSG (19.2%), WHO 1999 (10.1%), and DIPSI (7.4%) criteria. The National Family Health Survey (NFHS-5) (2019-21)²⁸ found weighted GDM prevalence of 4.2 per cent and age-adjusted prevalence of 5.4 per cent, where GDM was defined as elevated random blood glucose (≥200 mg/dL in nonfasting state and ≥92 mg/dL in fasting state). The NFHS survey²⁹ reported self-reported GDM prevalence of 0.53 per cent in 2015-16 and 0.80 per cent in 2019-20, with higher rates in some States like Goa (4.88%) and Kerala (3.06%), indicating regional variability. The stratification of risk of diabetes in early pregnancy (STRiDE)³⁰ study reported 19.2 per cent prevalence by OGTT; women in India with gestational diabetes mellitus strategy (WINGS)⁴ found to be 18.4 per cent by IADPSG and 14.6 per cent by WHO criteria; REVAMP³¹ reported 15.2 per cent prevalence by DIPSI. Definitions partly explain differences. Notably, ICMR-INDIAB data show a rise from 13.3 per cent (2008-10) to 30.2 per cent (2019-20), likely driven by maternal age, obesity, urbanisation, and lifestyle, marking GDM as a growing public health concern.

Although urban-rural differences in the prevalence of T2D among non-pregnant women are well-documented, our study found no significant difference in GDM prevalence between urban and rural populations. This finding raises the possibility that pregnancy itself may act as a dominant metabolic influence, potentially overriding the influence of conventional diabetes risk factors, such as lifestyle, diet, and adiposity, typically associated with urbanisation. This convergence in GDM prevalence in urban and rural areas warrants further investigation.

A meta-analysis³² of 84 asian studies identified key GDM risk factors including multiparity (≥2), prior GDM, family diabetes history, stillbirth, congenital anomalies, abortion, macrosomia, preterm birth, pregnancy-induced hypertension, polycystic ovary syndrome, age ≥25 yr, and BMI ≥25, with odds ratios from 1.90 to 8.42. Socioeconomic status also influences GDM risk, with higher prevalence in high SES groups³³. Studies like STRiDE³⁰ and WINGS⁴ have highlighted HbA1c, prior GDM, family history of diabetes, and age as significant predictors of GDM. In our study, only family history and higher systolic blood pressure were associated with increased GDM risk.

Few studies report on prevalence of Early and Late GDM^34-36. A global review³⁷ reported Early GDM ranging from 0.7-36.8 per cent. In INDIAB, Early GDM ranged 19.2-34 per cent by different criteria. Our study likely underestimates prevalence due to missing 1-h glucose data per IADPSG criteria. It is known that the pathogenesis of Early GDM differs from that of Late GDM³⁸. Late GDM is more common in Northern India, Early GDM in Central India, possibly reflecting differences in risk factor profiles. The TOBOGM trial^39,40 showed early screening (<20 wk) and intervention reduce complications, such as, preterm birth, neonatal hypoglycaemia, macrosomia, and birth trauma. In India, early detection is crucial for reducing maternal, neonatal risks, and future diabetes burden. Incorporating early screening in antenatal care can improve outcomes significantly.

The study’s main strength is its large, nationally representative, community-based sample with standardized cardiometabolic assessments throughout pregnancy, ensuring reliable data and capturing true regional GDM variations. Limitations include its cross-sectional design, preventing causal conclusions or long-term follow up, and missing body weight and BMI data (not collected as per ICMR-INDIAB protocol). Use of capillary blood glucose adds variability compared to venous plasma, though correlation is acceptable⁴¹. Conducting even two-sample OGTTs is challenging in large epidemiological studies, as demonstrated by the ICMR-INDIAB experience⁴².

In conclusion, our findings highlight the high prevalence of GDM among Indians, underscoring the need for monitoring for GDM throughout pregnancy, starting from the first trimester. With nearly one in four pregnant women in India having GDM, this represents a large population at risk of pregnancy complications and future chronic disease, necessitating urgent public health attention and resources. The high prevalence of Early GDM (one-fifth of the pregnant women) underscores the importance of early screening, while the regional differences highlight the need for tailored, region-specific interventions addressing local risk factors and healthcare access. Key risk factors like elevated systolic blood pressure and family diabetes history identify groups needing prioritized screening. Inadequate implementation of the GDM screening protocol during antenatal care can result in missed or delayed diagnosis, preventing timely interventions such as dietary adjustments, glucose monitoring, and medication when necessary. Poor implementation of screening protocols can delay diagnosis and intervention, increasing risks for mother and child. Overall, these findings stress the critical need for standardized, universal GDM screening and effective management integrated into maternal health programmes nationwide.