© JAPI • VOL. 52 • SEPTEMBER 2004www.japi.org707
Gestational Diabetes Mellitus in India
V Seshiah*, V Balaji**, Madhuri S Balaji**, CB Sanjeevi+, A Green++
Background : Glucose intolerance during pregnancy predisposes the offspring for increased risk of
developing glucose intolerance in the future. This vicious cycle is likely to influence and perpetuate the
incidence and prevalence of glucose intolerance in any population.
Aim : No data is available about the prevalence of glucose intolerance during pregnancy in our country
and hence a study was undertaken on this aspect.
Methods : This study was performed in the antenatal clinic of Government Maternity Hospital, Chennai,
India. As a pregnant woman in second or third trimester checks into the antenatal clinic, she was given
50 gm oral glucose load and blood sample was collected after one hour. This test was performed on
1251 pregnant women. They were requested to come after 72 hours for the 75 gm OGTT recommended
by WHO. Among the 1251 women, 891 responded. The blood sample was taken in the fasting state
and at 2 hours after 75 gm of oral glucose. Diagnosis was based on the WHO criteria for gestational
diabetes mellitus (GDM).
Results : The mean age of these pregnant women was 23 ± 4 years. There was a significant increase in
the prevalence of GDM in relation to gravida. The effect of BMI did not quite reach statistical significance
(χ2 (df=1) = 3.659, P = 0.055), but a model of linear trend was significant.
Of the 1251 women who underwent the 50 gm oral glucose challenge test, 670 (53.55%) had one hour
plasma glucose ≥ 130 mg/dl. Among the 891 pregnant women who had 75 gms OGTT, 168 (18.9%)
were diagnosed as GDM, taking both FPG ≥ 126 mg/dl and/or 2 hr PPG ≥ 140 mg/dl as cut-off values.
Taking only 2 hr plasma glucose for analysis, 144 (16.2%) had a value ≥ 140 mg/dl.
A similar study was conducted in different parts of the country taking only the 2 hr 75 gm post-glucose
value of ≥ 140 mg/dl as diagnostic criteria for GDM. Of the total number of pregnant women (n = 3674)
screened, 16.55% of them found to have GDM.
Conclusion : Our study has documented the increased prevalence of GDM in our population
necessitating universal screening for glucose intolerance in pregnancy. Using 2 hr plasma glucose ≥ 140
mg/dl as a one step procedure is simple and economical, particularly for the countries ethnically more
prone to high prevalence of diabetes. ©
to the need to direct special attention to this population,
especially in developing countries.1 The primary prevention
is likely to reverse or halt this trend. For this the need is to
focus at the intrauterine environment as the “preventive
medicine starts before birth”. Intrauterine exposure to
hyperglycemia during the critical period of fetal development
programmes the development of pancreas negatively and
affects the insulin secretory function.2 Hence this study was
undertaken to detect the glucose intolerance that occurs
during pregnancy as this metabolic disturbance predisposes
the offsprings for higher risk of developing glucose
intolerance in their later life.3 As of today we have no current
national data regarding the occurrence of abnormal glucose
tolerance in the pregnant women. The routine screening for
*Medical Director, Diabetes Unit, Apollo Hospitals, Chennai.
Chairman - Dr V Seshiah Diabetes Care and Research Institute,
Chennai. ** Consultant Diabetologist, Dr V Seshiah Diabetes Care
and Research Institute, Chennai. +Department of Molecular
Medicine, Karolinska Institute, Sweden. ++Department of
Epidemiology and Social Medicine, University of Aarhus, Denmark.
Received : 25.5.2004; Revised : 28.6.2004; Accepted : 15.7.2004
prevalence of diabetes in adults showed an expected total
rise of > 120% from 135 million in 1995 to 300 million in 2025.
These numbers also include GDM,and should alert physicians
he prevalence of diabetes is increasing globally and India
is no exception. The 1997 WHO estimates of the
© JAPI • VOL. 52 • SEPTEMBER 2004
glucose intolerance during pregnancy is not done in maternity
hospitals maintained by the Government, municipality or local
bodies that care for the majority of the pregnant women in
MATERIALS AND METHODS
This study was carried out in Raja Sir Ramaswamy
Mudhaliar lying hospital attached to the Government Stanley
Medical College and Hospital, Chennai, India during the
period February to December 2001. To obtain an unbiased
data this maternity hospital was chosen for the study as
pregnant women from different socioeconomic strata attend
this hospital for antenatal check up and confinement.
Consecutive 1251 pregnant women in the second or third
trimester attending the antenatal clinic were given a 50 gm
oral glucose load and the venous blood was collected after
one hour.4 Details of family history of diabetes, history of
previous pregnancies and the socio-economic status were
obtained. Blood pressure was recorded. The body mass index
(BMI) of the subjects was calculated and expressed in kg/m2.
All women were requested to have their regular diet and return
after 72 hours for the 75 gm oral glucose tolerance test
recommended by WHO.5 The blood sample was taken in the
fasting state and 2 hr after 75 gm of oral glucose. The plasma
glucose was estimated by GOD-POD method by using Bayer’s
kit. The result of the initial testing (50 gm-1 hr) was considered
positive if the plasma glucose ≥ 130 mg/dl. This cut-off value
was chosen to increase the detection rate.4 A woman was
considered to have GDM if the FPG ≥ 126 mg/dl and/or 2-hr
plasma glucose value ≥ 140 mg/dl with the 75 gm oral glucose.6
Associations were analysed with Mantel-Haenszel
technique, providing odds ratios stratified for potential
confounders and corresponding χ2-based statistics.
Confidence limits were estimated by means of the approximated
Poisson distribution. P-values <0.05 were considered
statistically significant. To standardise the diagnosis of GDM,
the WHO has proposed using a 2hr 75gm OGTT, with a
threshold plasma glucose concentration of greater than 7.8
mmol/L (140 mg/dl) at 120 minutes similar to that of IGT outside
pregnancy.5 Moses et al adopted WHO criteria in their study,
where they used a single 75gm OGTT and diagnosed GDM
with 2 hr PPG ≥ 140mg/dl.7 David Pettit in his editorial titled
‘The 75 gm oral glucose tolerance test in pregnancy’ favoured
WHO recommendation.8 More importantly GDM based on
2hr 75 gm OGTT defined by either WHO or ADA Criteria
predicts adverse pregnancy outcome.9 Hence we chose WHO
criteria due to the simplicity and the economical
considerations. Further, assuming that the effective treatment
is available, WHO criteria of 2 hr PPG ≥ 140mg/dl identifying
a large number of cases may have a greater potential for
A total of 1251 pregnant women had the initial 50 gm-1 hr
test. Of these, 669 (53.5%) were test-positive. For the
subsequent 75 gm-2 hr test 891 (71.2%) women responded.
Among those positive for the 50 gm-1 hr test, 548 (81.9%)
responded whereas 343 (58.9%) responded among the women
who were negative for the 50 gm-1 hr test (Table 1). The
positive association between a positive outcome of the 50
gm-1 hr test and taking part in the subsequent 75 gm-2 hr test
was statistically significant (Mantel-Haenszel odds ratio after
stratification for age: 3.14, χ2 (df=1) = 78.067, P<0.0001), and
there was no evidence of interaction with age (χ2 (df=4) for
heterogeneity = 1.770, P=0.778).
Of the 891 women who underwent the 75g OGTT, 24 of
them had FPG ≥ 126 mg/dl, 133 of them had 2 hr PG ≥ 140 mg/
dl and 11 had both FPG ≥ 126 mg/dl and 2 hr PG ≥ 140 mg/dl.
Overall, 168 women (18.9%) [95% confidence limits: 16.3%-
21.6%] were diagnosed as GDM.
The mean age of these pregnant women was 23 ± 4 years.
The prevalence proportion increased with age from 15.7%
(confidence limits: 8.6%-25.3%) in the age group 15-19 years
to 32.1% (confidence limits: 20.3%- 46.0%) for the age groups
30+ years (Fig. 1). With regard to the age effect, a model of
linear trend was statistically significant (χ2 (df=1) = 10.630,
Data on BMI was available for 664 (74.5%) and the
prevalence proportion of GDM increased with increasing BMI
(Fig. 2). This is evident from our study that the prevalence of
GDM with BMI ≤ 23 and ≥ 24 is 19.7% and 24.5% respectively.
The effect of BMI did not quite reach statistical significance,
but was approaching statistical significance (χ2 (df=1) = 3.659,
P re v a le n ce P e rc e n ta g e
1 5 .7 1 5 .6
2 3 .2
3 2 .1
1 5 - 1 92 0 - 2 4 2 5 - 2 9 3 0 +
A g e (in y e a rs )
Fig. 1: Prevalence percentage (with 95% confidence intervals) of
gestational diabetes by age groups (n=860 after exclusion of 31
subjects because of missing data on age).
Table 1 : Overview of results of initial 50 gm-1 hr test. In
parentheses: Number of women who responded to the
subsequent 75 gm-2 hr test
Result of initial 50 gm-1 hr test
Age groupPositiveNegative Total
15-1949 (40) 78 (43)127 (83)
20-24 361 (294) 309 (181)670 (475)
25-29196 (162)143 (84) 339 (246)
30-3432 (26) 22 (15)54 (41)
35+ 12 (9)8 (6) 20 15)
Age unknown 19 (17) 22 (14)41 (31)
Total 669 (548) 582 (343)1251 (891)
© JAPI • VOL. 52 • SEPTEMBER 2004www.japi.org709
P = 0.055) by chi square test of linear trend.
The prevalence proportion of GDM increased with gravida,
from 18.1% (confidence limits: 14.38% - 22.29%) in the primi
gravida to 25.8% (confidence limits: 11.86% - 44.61%) for the
gravidas ≥ 4. (Fig. 3).
We believe that the way the study subjects were
ascertained provides a representative and unbiased sample
of pregnant women, as the bulk of the pregnant women attend
hospital of this type for antenatal check-up and confinement
in our country.
Among 1251 pregnant women screened with glucose
challenge, 891 responded for 2 hr 75 gm OGTT. However,
neither positive family history nor parity influenced the
reporting for the 75 gm-2 hr test. This phenomenon of no
show rate appears to be universal. Magee et al reported in
their follow up, 91 of the 457 positive screen individuals failed
to undergo diagnostic test.10 Luiz et al also observed in their
study, that 23% of their screen positive women did not return
for OGTT.11 Our prevalence rate was 18.9%. We got this figure
as we took into consideration for analysis the FPG value also
as this was available besides 2hr PPG. Of the 891 women, 24
had FPG ≥ 126 mg/dl. They might have had pregestational
diabetes. We could not confirm this as A1c estimation is not
recommended as routine test for screening and hence we
also did not perform the test.12 If our estimated prevalence
proportions among those positive and negative, respectively,
for the 50 gm-1hr test were applied to total sample of 1251
women, the revised overall prevalence proportion would be
17.7% (confidence limits: 15.6%-19.9%) against the value of
18.9% found among the 891 women. Another outcome of this
study was that the fasting plasma glucose estimation is not
necessary for Universal screening as this procedure identified
negligible percentage of women with glucose intolerance
(2.69%). WHO is also not in favour of estimation of FPG for
An overall prevalence proportion of GDM at 17.7% in this
rather young population of pregnant women was
considerably high. Subsequent to the observation of the high
prevalence of GDM in one centre, a multicentre study was
initiated in different parts of the country in 2002 - 2003 taking
only the 2 hr 75 gm post glucose value of ≥ 140mg/dl as
diagnostic criteria for GDM. Almost a similar prevalence of
15% was obtained in another govt. maternity hospital
affiliated to Madras Medical College in the city of Chennai.
This trend of high prevalence of GDM was also found in
other parts of the country, 15% in Trivandrum, 21% in Alwaye,
12% in Bangalore, 18.8% in Erode and 17.5% in Ludhiana.
The total number of pregnant women screened in these
centres was 3674 and an overall GDM prevalence of 16.55%
was observed. This study documented a definite increasing
trend in the prevalence of GDM compared to that of 2% in
198213 and 7.62% in 1991.14 This trend is also seen in other
countries. For example in Australia at one hospital where the
same testing procedure and diagnostic criteria have been
used for more than 2 decades, the prevalence has more than
A recent national survey reported the prevalence of IGT
in the age group of 20-29 years and 30-39 years as 12.2% and
15.3% respectively.16 No gender difference was seen in the
prevalence of IGT.16 Further for a given population and
ethnicity the risk of diabetes and pregnancy, mirrors that of
the underlying frequency of type 2 DM in the general
population.17 We also observed in our study that the
prevalence of GDM is closer if not similar to the prevalence
Fig. 2: Prevalence percentage (with 95% confidence intervals) of
gestational diabetes by body mass index (n=664 after exclusion of
227 subjects because of missing data on body mass index).
Fig. 3: Prevalence percentage (with 95% confidence intervals) of
gestational diabetes by gravida (n= 891).
Table 2 : Results of the initial 50 gm-1hr test evaluated
against the confirmatory 75 gm-2 hr test. In
parentheses: 95% confidence intervals.
Result of 75 gm-2 hr test
+ GDM - GDMTotal Prevalence proportion
Positive for134 4145480.245 (0.209; 0.283)
Negative for34 309343 0.099 (0.070; 0.136)
Total168723891 0.189 (0.163; 0.216)
Sensitivity: 0.798 (0.727; 0.854)
Specificity: 0.427 (0.391; 0.464)
Predictive value of positive test:0.245 (0.210; 0.283)
Predictive value of negative test:0.901 (0.863; 0.929)
© JAPI • VOL. 52 • SEPTEMBER 2004
rate of IGT in our population. With this huge population of
reproductive age in India, a significant segment of women
with abnormal glucose tolerance during pregnancy needs
Our attempt and outcome of the screening for glucose
intolerance during pregnancy has given an insight for the
phenomenal increase in the prevalence of diabetes in India.
This view is substantiated by the observation of Dabelea et
al on Pima Indians. The gestational diabetes mellitus has a
far reaching consequence in predisposing the offsprings to
glucose intolerance has been documented in Pima Indians.
The children born in 1965 to women with gestational diabetes
were followed up till 2000. By the time they reached 35 years,
more than half of the group had diabetes.3 Hence as a policy
to identify GDM and its consequences on the infant a 75 gm
OGTT has been recommended to all women in the population
during the third trimester of pregnancy.3
In India, both undernutrition and overnutrition exist during
pregnancy. There are two reported studies in India which are
related to size at birth to future risk of type 2 diabetes. In
Mysore, low birth weight did not increase the risk of diabetes
but babies who were short and fat (higher body mass index,
BMI) at birth were at increased risk.18 Fall et al speculate that
the rise in type 2 diabetes in Indian urban populations may
have been triggered by mild obesity in mothers, leading to
glucose intolerance during pregnancy, macrosomic changes
in the fetus and insulin deficiency in adult life.18 Yet another
study by Yajnik et al attributes high prevalence of type 2 DM
and IGT in Indian people may be linked to poor fetal growth.19
Same author also suggests that type 2 DM in India may be
programmed in fetal life, hence diabetes prevention will have
to start in early life (in utero) and continue in later life.20 The
importance is that the intrauterine millieur interieur, whether
one of nutritional deprivation or one of nutritional plenty,
results in changes in pancreatic development and peripheral
response to insulin that may lead to adult - onset GDM and
type 2 DM.21
Our study favours one step procedure recommended by
WHO for screening instead of two step procedure using
preliminary screening with 50 gm one hr test .4 Table 2 shows
the results of the evaluation of the outcome of the 50 gm-1 hr
test, using the outcome of the 75 gm-2 hr test as standard.
The sensitivity of the 50 gm 1 hr test was 79.8% when
compared to the low specificity of 42.7%. Combined with a
relatively small number of false negative subjects, this yielded
reasonable high values of sensitivity (79.8%, confidence limits:
72.7%-85.4%) and predictive value of a negative 50 gm-1 hr
test (90.1%, confidence limits: 86.3%-92.9%). However, there
was a substantial number of subjects false positive for the 50
gm-1 hr test, resulting in poor values of specificity (42.7%,
confidence limits: 39.1%-46.4%) and predictive value of a
positive 50 gm-1 hr test (24.5%, confidence limits: 21.0%-
28.3%). Since the specificity of using 50 gm-1 hr test is low,
instead of performing screening test using 50 gm-1 hr test
and then 100 gm / 75 gm OGTT4, a one step procedure of
performing 75gm OGTT directly is ideal as we need to perform
universal screening for glucose intolerance in pregnancy.
In conclusion, our study has documented the increased
prevalence of GDM in our population and we venture to give
the following observations and recommendations. Universal
screening for glucose intolerance during pregnancy is
essential as Indian women have high prevalence of diabetes
and their relative risk of developing GDM is 11.3 times
compared to white women.22 Asian women are ethnically more
prone to develop glucose intolerance compared to other
ethnic groups.15 GDM based on 2hr 75gm OGTT defined by
WHO predicts adverse pregnancy outcome and warrants
treatment.9 A 2 hr 75 gm post plasma glucose ≥ 140mg/dl
serves both as screening and diagnostic criteria besides being
a simple and economical one step procedure. As the routine
screening for glucose intolerance during pregnancy is not
done, probably the undiagnosed glucose intolerance that
has been occurring in the past has resulted in the increased
prevalence of diabetes in India. The timely action taken now
in screening all pregnant women for glucose intolerance,
achieving euglycemia in them and ensuring adequate nutrition
may prevent in all probability, India becoming diabetes capital
of the World.
‘No single period in human development provides a greater
potential than pregnancy for long range pay off via relatively
short range period of enlightened metabolic manipulation’ -
We would like to thank Dr. C Ravindranath - Dean -
Stanley Medical College and Hospital, Prof. SK Rajan - Prof
and Head, Department of Medicine and Diabetology - Stanley
Medical College and Hospital, Professor Thirupurasundari -
Superintendent, Raja Sir Ramaswamy Mudhaliar lying Hospital
for the support in carrying out the study. Our profound
thanks to all the investigators who participated in the study.
Centres and Staff involved in the Study
Prof. RS Hariharan and Prof. C Anjalakshi, Department of
Diabetology, Madras Medical College, Chennai. Prof. Mary
John, Christian Medical College, Ludhiana: Prof. KP Paulose,
SUT Hospital, Trivandrum. Prof. Prasanna Kumar KM,
MS Ramaiah Hospitals, Bangalore. Dr. Shyam Mukundan,
Lakshmi Nursing Home, Alwaye. and Dr. R Aruyerchelvan,
SRC Diabetes Care Centre, Erode.
1. Avi Ben-Haroush, Yariv Yogev, Moshe Hod. Epidemiology of
gestational diabetes mellitus. Textbook of Diabetes and
2. Van Assche FA, Holemans K, Aerts L. Long term
consequences for offspring of diabetes during pregnancy. Br
Med Bull 2001;60:173-82.
3. Dabelea D, Pettitt DJ. Effect of diabetes on pregnancy and
offspring: follow up research in the Pima Indians. J Matern
Fetal Medicine 2000;9:83-8.
4.Expert Committee on the Diagnosis and Classification of
© JAPI • VOL. 52 • SEPTEMBER 2004www.japi.org711 Download full-text
Diabetes Mellitus Report of the Expert Committee on the
Diagnosis and Classification of Diabetes Mellitus. Diabetes
Care 1997;20 (Suppl. 1):1183-97.
5. Alberti K, Zimmett P, WHO Consultation. Definition,
diagnosis and classification of diabetes mellitus and its
complications, 1: diagnosis and classification of diabetes
mellitus. Diabet Med 1998;15:539-53.
6.WHO study group prevention of diabetes mellitus- Geneva.
World health Org (Tech Report Series 844), 1994.
7. Moses RG. The 75 gms glucose tolerance test in pregnancy a
reference range determined on a low risk population and
related to pregnancy outcome. Diabetes Care 1998;21:1807-
8.Pettitt DJ. 75 gms OGTT pregnancy. Diabetes Care
9. Schmidt MI, Duncan BB, Reichelt AJ, Branchtein L, Matos
MC, Costa e Forti A, Spichler ER, Pousada JMDC, Teixeira
MM, Yamashita T for the Brazilian Gestational Diabetes Study
Group. Gestational diabetes mellitus diagnosed with a 2-h
75 gm oral glucose tolerance test and adverse pregnancy
outcomes. Diabetes Care 2001;24:1151-5.
10. Magee S, Walden CE, Benedetti TJ, Knopp RH. Influence of
diagnostic criteria on the incidence of gestational diabetes
and perinatal morbidity. JAMA 1993;269:609-15.
11. Luiz Guilherme Kraemer De Aguiar, Haroldo Jose De Matos,
Marilia De Brito Gomes. Could fasting plasma glucose be
used for screening high-risk outpatients for gestational
diabetes mellitus? Diabetes Care 2001;24:954-5.
12. Aziz NL, Abdelwahab S, Moussa M, Georgy M. Maternal
fructosamine and glycosylated haemoglobin in the prediction
of gestational glucose intolerance. Clin Exp Obstet Gynecol
13. Agarwal S, Gupta AN. Gestational Diabetes. J Assoc Physicians
14. Narendra J, Munichoodappa C, Gurudas A, Ramprasad AV,
Madhav T, Vijayalakshmi, Nirmala, Keith. Prevalence of
glucose intolerance during pregnancy. Int J Diab Dev Countries
15. Beischer NA, Oats JN, Henry OA, Sheedy MT, Walstab JE.
Incidence and severity of gestational diabetes mellitus
according to country of birth in women living in Australia.
Diabetes 1991;40 Suppl 2:35-8.
16. Ramachandran A, Snehalatha C, Kapur A, Vijay V, Mohan V,
Das AK, Rao PV, Yajnik CS, Prasanna Kumar KM, Nair JD.
For the Diabetes Epidemiology Study Group in India (DESI).
17. Hilary King. Epidemiology of Glucose Intolerance and
Gestational Diabetes in women of childbearing age. Diabetes
Care 1998;21:Suppl 2, B9-B13.
18. Fall CH, Stein CE, Kumaran K, Cox V, Osmond C, Barker DJ,
Hales CN. Size at birth, maternal weight, and type 2 diabetes
in South India. Diabet Med 1998;15:220-7.
19. Yajnik et al. Fetal growth and glucose and insulin metabolism
in four year old Indian children. Diabetic Medicine 1995;12:330-
20. Yajnik CS. Fetal origins of diabetes in countries. Diabetes
21. Savona - Ventura C, Chircop M. Birth weight influence on the
subsequent development of gestational diabetes mellitus.
Acta Diabetol 2003;40:101-4.
22. Dornhost A, Paterson CM, Nicholls JS, Wadsworth J, Chiu
DC, Elkeles RS, Johnston DG, Beard RW. High prevalence of
GDM in women from ethnic minority groups. Diabetic Med
The office bearers of Association of Physicians of India Chhattisgarh State Chapter (2004-2005)
SK Tiwari, Bilaspur
PS Deshpande, Raipur
PSA Sharma, Bhilai
A Rai, Raipur
K Sengupta, Raipur
P Kalvit, Bilaspur
P Jain, Balod
GB Gupta, Raipur
A Verma, Bilaspur
D Ratnani, Bhilai
KC Debnath, Korba
V Kupatkar, Bilaspur
RBP Gupta, Jagdalpur