SEARCH

SEARCH BY CITATION

Keywords:

  • Diabetes;
  • diabetes mellitus;
  • pregnancy;
  • risk

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Commentary on ‘Diabetes mellitus following gestational diabetes: role of subsequent pregnancy’
  9. Acknowledgements
  10. References

Objective  This study was designed to determine the rate of diabetes up to 13 years after pregnancies complicated by gestational diabetes and to identify risk factors for developing diabetes. The role of a subsequent pregnancy, with and without gestational diabetes, was also examined.

Design  This was a retrospective cohort study of women with gestational diabetes.

Population and setting  Women who had gestational diabetes in their first pregnancy between 1989 and 2002 were identified through a population-based perinatal database in Nova Scotia, Canada.

Methods  Subsequent diagnoses of diabetes, up to 13 years after the first pregnancy, were obtained from physician billing and hospital discharge databases. Cox proportional hazards regression models were used to estimate adjusted relative risks (RR) and 95% confidence intervals.

Main outcome measures  Diagnosis of diabetes after pregnancy.

Results  Of the 1401 nulliparous women with gestational diabetes, 251 women (17.9%) developed diabetes in the follow-up period. The cumulative incidence at 1, 5, and 10 years was 5.9, 14.8, and 22.2%, respectively. Factors significantly associated with an increased risk of developing diabetes mellitus included a pre-pregnancy weight of ≥86 kg (RR = 1.8, 95% CI 1.2–2.9), insulin therapy during the index pregnancy (RR = 4.1, 95% CI 2.1–7.9), neonatal hypoglycaemia (RR = 2.6, 95% CI 1.6–4.2), and a subsequent pregnancy with gestational diabetes (RR = 2.3, 95% CI 1.6–3.4).

Conclusion  Indicators of the severity of gestational diabetes, defined by insulin use, neonatal hypoglycaemia, and recurrent gestational diabetes in a subsequent pregnancy, are important in predicting a subsequent diagnosis of diabetes. Our findings do not support the theory that subsequent pregnancy, per se, increases the risk of developing diabetes.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Commentary on ‘Diabetes mellitus following gestational diabetes: role of subsequent pregnancy’
  9. Acknowledgements
  10. References

Gestational diabetes mellitus (GDM) is defined as ‘carbohydrate intolerance of varying degrees of severity with onset or first recognition during pregnancy’.1 The prevalence of GDM in Canada is approximately 3.5% in the non-Aboriginal population.2 The recurrence rate for GDM in subsequent pregnancies has been estimated at 36% in the Nova Scotia population.3 GDM is associated with an increased risk of subsequent type II diabetes mellitus (DM).4–9 In the general female population of Nova Scotia in 2002/03, the prevalence of diabetes is 0.86% for women of age 20–29 years, 1.74% for women of age 30–39 years, and 3.67% for women of age 40–49 years.10 These rates are slightly higher than prevalence rates among white women less than 44 years of age in the USA.11

Cumulative incidence rates of DM among women with a history of GDM vary widely depending on the length of follow up and the underlying risk of DM in the population.12 Among women with a history of gestational diabetes, it is generally accepted that race, age, parity, family history of diabetes, pre-pregnancy weight, postpartum obesity, and weight gain are risk factors for developing DM.4,6–9,13 Other suspected risk factors include smoking, physical inactivity, diet, and drugs that adversely affect glucose metabolism.6 It has been suggested that increased insulin resistance associated with pregnancy may accelerate the depletion of beta cells.5 If this is the case, it would be expected that increased parity or subsequent pregnancies would increase the risk for development of subsequent diabetes among women with a history of gestational diabetes. However, findings related to the roles of parity and subsequent pregnancies are conflicting.6

This study was conducted to determine the rate of developing DM up to 13 years after delivery and to identify risk factors for DM among a population-based cohort of women who had GDM during their first pregnancy. In particular, the role of subsequent pregnancy as a potential risk factor for subsequent DM was investigated.

Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Commentary on ‘Diabetes mellitus following gestational diabetes: role of subsequent pregnancy’
  9. Acknowledgements
  10. References

The Nova Scotia Atlee Perinatal Database (NSAPD) was used to establish a cohort of Nova Scotia residents who delivered their first infant between 1989 and 2002 and who had gestational diabetes during the pregnancy. Clinical information on the pregnancy and newborn was obtained from the NSAPD, which includes information on all hospital deliveries in Nova Scotia for infants of birthweight ≥500 g or 20 weeks of gestation and over. The NSAPD contains extensive information on maternal lifestyle factors, previous pregnancy history, illnesses complicating pregnancy, labour and delivery information, and infant outcomes. Trained health records personnel abstract information, after delivery, from standardised clinical forms and hospital records. Continuing data quality assurance checks, including reabstraction studies and validation studies, show that the information in the Perinatal Database is reliable.14 Women were excluded if diabetes had been diagnosed before pregnancy.

Based on the recommendations from the Canadian Diabetes Association, it is recommended that all pregnant women in Nova Scotia be screened for GDM between 24 and 28 weeks with a 50-g oral glucose screen using a 1-hour plasma glucose (taken independently of time of last meal). Women with a value of ≥ 10.3 mmol/l are considered to have GDM present, and further testing is not indicated. Women with a value of ≥ 7.8 mmol/l (but less than 10.3 mmol/l) are recommended to have an oral glucose tolerance test (OGTT). For most of the study period, the 100-g OGTT was the standard diagnostic test for GDM in Nova Scotia. In the early 2000s, the 75-g OGTT began to be used by most healthcare facilities in the province for diagnosing GDM. For both tests, the plasma glucose values for diagnosing GDM for the 75-g and the 100-g test were based on published clinical practice guidelines of the Canadian Diabetes Association.2

The primary outcome, DM, was determined from the Canadian Institute for Health Information’s Discharge Abstract Database (for hospitalisations) and from the Nova Scotia Medical Insurance Database (for outpatient visits to physicians). These administrative databases are housed at the Population Health Research Unit of Dalhousie University and were linked to the NSAPD by Health card number. In this study, DM was defined by the occurrence of either two outpatient physician service claims or one hospitalisation with the diagnostic code for diabetes, within a 2-year period. This method (the 2-claim method) is the national standard for identifying women with diabetes in administrative data, based on the National Diabetes Surveillance System. A study from Ontario reported excellent validity for determining diabetes diagnoses using the 2-claim method, showing 86% sensitivity and 97% specificity.15 A registration file, developed in 1995, for tracking patient enrolment in Medical Services Insurance was used to identify the follow-up period after this date. Prior to 1995, follow up was determined by tracking dates of any physician visit or hospital admission. The total length of follow up was from the year of delivery from the first pregnancy until either the development of DM, end of study period (31 December 2002), or until the registration file indicated the subject no longer lived in Nova Scotia or had died.

Women diagnosed with DM within 1 year following a diagnosis of GDM may have had undiagnosed DM prior to pregnancy. In Nova Scotia, women with GDM are recommended to have a 75-g OGTT or a fasting plasma glucose test within 6 months of delivery. To avoid this misclassification and to provide results similar to previous studies,9 subjects diagnosed within the first year after the index pregnancy were excluded from the risk factor analysis. However, for the purposes of reporting rates of DM by follow-up time, the first year of follow up after the index pregnancy was included.

Initially, cumulative rates of new diagnoses of DM were determined for each year of follow up, accounting for censored observations. Most covariates analysed were dichotomous variables, but continuous variables (e.g. maternal delivery weight) were grouped into quartiles based on the distribution within the cohort. Information on the first subsequent pregnancy was coded as: no subsequent pregnancy, subsequent pregnancy without gestational diabetes, or subsequent pregnancy with gestational diabetes. This variable was time-varying and thus changed as subsequent pregnancies occurred. Subsequent pregnancy was analysed as a time-dependent covariate.

Proportional hazards regression was used to account for the varying length of follow up during the study period and for simultaneously modelling risk factors with DM as the outcome. Initially, unadjusted relative risks (RR), estimated from hazards ratios, and 95% confidence intervals were calculated for each potential risk factor. Backward stepwise regression was used to test the significance of each factor and to arrive at a final model. If the removal of a factor did not significantly affect the fit of the model, as tested by a likelihood ratio test (P < 0.05), the factor was deleted and the next factor was tested. Subjects with missing values for variables in the model were excluded but were reintroduced if the factor with missing values did not remain in the final model. For most variables, there were very few missing values.

A scoring system for predicting the development of DM was developed using established methods.16 A score was developed from the final proportional hazards regression model and assigned points to each perinatal risk factor. The points that were assigned to each factor were proportional to the corresponding regression coefficient. For each woman, a total score was calculated based on the sum of the points assigned to each perinatal factor. Estimates of the probability of the development of DM were calculated for all possible total scores. The probabilities were based on the coefficients obtained from the final regression model, the prevalence of each factor and the survival function at the mean value for each factor.

All analyses were conducted using SAS 8.2 software. This study received approval from the Research Ethics Board of the IWK Health Centre.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Commentary on ‘Diabetes mellitus following gestational diabetes: role of subsequent pregnancy’
  9. Acknowledgements
  10. References

The cohort consisted of 1401 nulliparous women with GDM who delivered an infant between 1989 and 2002. The average maternal age in the cohort was 28.1 years. Of the 1401 women with GDM, a total of 251 (17.9%) were identified with DM during the follow-up period. Figure 1 shows the cumulative incidence of DM by year of follow up after the first pregnancy. The number of women who developed DM and the number of women at risk of developing DM at the start of each year of follow up are also shown. The number of women at risk of developing DM drops off with each follow up year due to the shorter follow-up time among women with an index pregnancy during the later years of the study period. The cumulative incidence of DM at 1-, 5-, and 10-year intervals was 5.9, 14.8, and 22.2%, respectively. The cumulative incidence rate did not show signs of plateauing in the 13 years of follow up. There were 83 (5.9%) women who developed diabetes within the first year of follow up and 89 women who had less than 1 year of follow up were excluded from the risk factor analyses.

image

Figure 1. Cumulative incidence of diagnoses of DM by year following first pregnancy among women with gestational diabetes.

Download figure to PowerPoint

Table 1 shows the frequencies and unadjusted RR for some of the risk factors evaluated. As seen in Table 1, factors significantly associated with the development of diabetes in a univariate analysis included low 1-minute Apgar score, high pre-pregnancy weight, high maternal weight at delivery, low birthweight, insulin therapy during pregnancy, maternal endocrine disease, chronic hypertension, severe pregnancy-induced hypertension (which is defined as diastolic blood pressure ≥110 on at least two occasions with a 6-hour period with or without proteinuria), preterm delivery, neonatal hypoglycaemia, and subsequent pregnancy with gestational diabetes. An interval between first and second pregnancy of less than 2 years was associated with development of diabetes (RR = 4.2, 95% CI 1.2–15.0) (data not shown). The factors that were associated with a significant reduction in the risk of developing diabetes were pre-pregnancy weight of <60 kg, maternal delivery weight of <74 kg, and a subsequent pregnancy without gestational diabetes. Smoking during pregnancy, maternal anaemia, type of labour, and breastfeeding were not associated with the development of diabetes (data not shown).

Table 1.  Frequency and rates of DM and unadjusted RR for perinatal factors among women with a history of gestational diabetes in the first pregnancy
Factor of interest*No. of women** (n= 1229)No. of women with diabetes (n= 168) n (%)Unadjusted RR (95% CI)
  • HELLP, haemolysis, elevated liver enzymes and low platelet count.

  • *

    Pregnancy-specific information is captured at index pregnancy.

  • **

    Numbers do not always equal total due to missing values.

Maternal age (years)
<2535859 (16.5)1.2 (0.8–1.7)
25–2944761 (13.7)1.0
30–3430737 (12.1)1.0 (0.7–1.5)
35+11811 (9.3)0.9 (0.5–1.7)
Pre-pregnancy weight (kg)
<6032222 (6.8)0.5 (0.3–0.9)
60–7127733 (11.9)1.0
72–8526847 (17.5)1.6 (1.0–2.5)
86+24850 (20.2)2.2 (1.4–3.4)
Delivery weight (kg)
<7428821 (7.3)0.5 (0.3–0.9)
74–8426535 (13.2)1.0
85–9828840 (13.9)1.1 (0.7–1.8)
99+26256 (21.4)2.0 (1.3–3.0)
Drug therapy for diabetes
Insulin4812 (25.5)4.9 (2.7–8.9)
None1182156 (13.2)1.0
Endocrine disease
Yes226 (27.3)2.5 (1.1–5.8)
No1208162 (13.4)1.0
Chronic hypertension
Yes4513 (28.9)2.5 (1.4–4.4)
No1185155 (13.1)1.0
Pregnancy-induced hypertension (with or without proteinuria)
Not severe23541 (17.5)1.4 (1.0–2.0)
Severe/HELLP6116 (26.2)2.4 (1.4–4.0)
None934111 (11.9)1.0
Infant birthweight (g)
500–24998420 (23.8)2.0 (1.2–3.3)
2500–344949053 (10.8)0.7 (0.5–1.1)
3450–399942463 (14.9)1.0
4000+23232 (13.8)1.0 (0.6–1.5)
Gestational age at birth (weeks)
<34265 (19.2)2.3 (1.0–5.7)
34–368820 (22.7)2.0 (1.2–3.1)
37+1100141 (12.8)1.0
Apgar 1 minute
1–35213 (25.0)2.1 (1.2–3.8)
4–726636 (13.5)1.1 (0.7–1.5)
8–10901118 (13.1)1.0
Neonatal hypoglycaemia
Yes7019 (27.1)2.5 (1.6–4.1)
No1160149 (12.8)1.0
Subsequent pregnancies
GDM19542 (21.5)2.1 (1.5–3.1)
Non-GDM41128 (6.8)0.6 (0.4–1.0)
None62498 (15.7)1.0

Table 2 shows the adjusted results from the final model. Factors significantly associated with developing DM among this cohort of GDM women included a pre-pregnancy weight of ≥86 kg (RR = 1.8, 95% CI 1.2–2.9), insulin therapy during the index pregnancy (RR = 4.1, 95% CI 2.1–7.9), neonatal hypoglycaemia (RR = 2.6, 95% CI 1.6–4.2), and a subsequent pregnancy with GDM (RR = 2.3, 95% CI 1.6–3.4). Pre-pregnancy weight of <60 kg was associated with a reduced risk of developing DM compared with women with a pre-pregnancy weight of 60–71 kg (RR = 0.6, 95% CI 0.3–0.9). Among a subset of women who had a subsequent pregnancy, weight gain between pregnancies was associated with the development of DM. Compared with women with no weight gain between pregnancies, the adjusted RR associated with weight gain between the first and next pregnancy was 2.2 (95% CI 1.0–4.5) for weight gain between 0.1 and 1.5 kg, 2.6 (95% CI 1.2–5.5) for weight gain between 5.1 and 10 kg, and 2.4 (95% CI 1.1–5.6) for weight gain over 10.1 kg (results not shown).

Table 2.  Adjusted RR of subsequent diabetes for factors included in final model among women with a history of gestational diabetes
Factor of interest*Adjusted** RR (95% CI)
  • *

    Pregnancy-specific information pertains to first pregnancy.

  • **

    Adjusted results are from a model which includes other factors in the table.

Pre-pregnancy weight (kg)
<600.6 (0.3–0.9)
60–711.0
72–851.5 (1.0–2.4)
86+ kg1.8 (1.2–2.9)
Diabetic therapy
Insulin4.1 (2.1–7.9)
None1.0
Neonatal hypoglycaemia
Yes2.6 (1.6–4.2)
No1.0
Subsequent Pregnancies
GDM2.3 (1.6–3.4)
Non-GDM0.8 (0.5–1.2)
None1.0

Using the results of the final proportional hazards model, a point scale was developed for the risk of developing DM over a 10-year period after an index pregnancy with GDM (Tables 3 and 4). Although the range of the possible point totals was −4 to +19, the results for 9 points and above were combined into one risk category since the accuracy is diminished at the extremes of the distribution. Thus, the estimated 10-year risk for a woman with a history of GDM who had a pre-pregnancy weight of ≥86 kg, received insulin during the index pregnancy, and had GDM in a subsequent pregnancy was over 87% (total point count ≥9). In contrast, the estimated 10-year risk for a woman with a history of GDM who was <60 kg with no subsequent pregnancies, who was not on insulin therapy during pregnancy and whose infant did not have hypoglycaemia was 17% (total point count −3).

Table 3.  Ten-year probability of developing DM associated with each risk score
Point total10-year probability (%)
−414
−317
−220
−124
029
134
240
346
453
561
668
775
882
≥9≥87
Table 4.  Point scales based on factors associated with gestational diabetes
Factor of interestPoints
Pre-pregnancy weight (kg)
<60−3
60–710
72–852
86+3
Diabetic therapy
Insulin7
None0
Neonatal hypoglycaemia
Yes5
No0
Subsequent pregnancies
GDM4
Non-GDM−1
None0

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Commentary on ‘Diabetes mellitus following gestational diabetes: role of subsequent pregnancy’
  9. Acknowledgements
  10. References

Results from a systematic review showed that the cumulative incidence of type II diabetes, among a cohort of women with a history of gestational diabetes, increases markedly until 5 years postpartum and plateaus after 10 years.12 In our study, the rate of DM increased linearly with follow-up time among women with a history of GDM, with no evidence of a plateau in incidence of 10 years after the first pregnancy. The cumulative incidence rates we observed at each of the follow-up time points were similar to estimates from other population-based studies of predominately white women.4,8,17 Coustan et al.18 found that only 11% of women with a history of GDM had diabetes or impaired glucose within 10 years of pregnancy, however, most of the subjects had less than 2 years of follow up after the index pregnancy.

In the risk factor analyses, women with a diagnosis of DM in the first year after the index pregnancy and women with less than 1 year of follow up were excluded. As well, follow up began 1 year after the index delivery and, therefore, women are assumed to have normal glucose tolerance in the initial postpartum period. A number of factors were statistically significant in the univariate analysis but only four factors remained as independent risk factors in the multivariate model. High pre-pregnancy weight is a risk factor for DM among all women, regardless of GDM status. Insulin therapy during pregnancy and neonatal hypoglycaemia are indicators of GDM severity.

Among women with a subsequent pregnancy, it has been shown previously that weight gain between pregnancies is associated with an increased risk.8,9 This observation was also noted in this study, but the association was not significant. Pre-pregnancy weight before the first pregnancy was found to be statistically significantly associated with development of DM, with evidence of a dose-response relationship with increasing weight. Pre-pregnancy weight of <60 kg was associated with reduced risk of development of DM. It has been suggested that the progression to diabetes among women with a history of GDM may be accelerated with obesity and inactivity.6 A recent Danish study found that the incidence of diabetes among women with previous diet-treated GDM had more than doubled over a 10-year period, presumably due to increasing obesity in the population.19

It has been hypothesised that pregnancy contributes to the decline of beta cell function in women with compromised beta cell reserves (such as among women with gestational diabetes) and that additional pregnancies may impair the ability to maintain normal glucose tolerance after pregnancy.6,9,20 In our study, subsequent pregnancy itself was not found to be significantly associated with the development of DM but was only a risk factor among women who had gestational diabetes in a subsequent pregnancy. In women with GDM in the subsequent pregnancy, the adjusted RR of developing DM was increased more than two-fold compared with women who did not have a subsequent pregnancy. Only 6.5% of women without gestational diabetes in a subsequent pregnancy developed diabetes during the follow-up period compared with 10% of women without a subsequent pregnancy and 23% of women with GDM in the subsequent pregnancy. Our results are similar to those found in an Australian cohort where 3% of the women who did not have GDM in a subsequent pregnancy and 30% of women who did have GDM in a subsequent pregnancy developed DM.7 Results from a Spanish cohort of women with a history of gestational diabetes also found that additional pregnancies did not increase the risk of future diabetes.21 In contrast, a study of primarily Latino women found that subsequent pregnancy increased the risk of developing DM more than three-fold among women with a history of gestational diabetes. However, in this Latino population, the vast majority of women had recurrent gestational diabetes in their subsequent pregnancy, whereas in our population gestational diabetes recurred in only 32% of the subsequent pregnancies. Our results suggest that subsequent pregnancy, per se, does not confer an increased risk for developing DM.

Implications

The simple risk score derived from the risk factors identified in the final model allows clinicians to estimate the 10-year risk of a woman developing DM after a pregnancy with GDM, given the presence or absence of the four factors identified in this study. Although GDM itself is a risk factor for DM and postpartum screening is indicated, screening is especially important in women who are at very high risk. Whether or not interventions to reduce modifiable risk factors would be effective is not known. Because subsequent pregnancy is a time-dependent covariate, a woman’s risk status will potentially change after her index pregnancy.

Progression to diabetes following a GDM pregnancy appears to be accelerated among non-white ethnic groups.6 Ethnicity is not captured in the databases used in this study, but according to the 1991 Canadian census, the Nova Scotia population is predominately Caucasian, with 3.1% of the population comprising visible minorities.22 Therefore, the results of this study are most relevant for Caucasian populations.

Strengths and limitations

A limitation of this study pertains to the potential misclassification of the diagnoses of DM using administrative data. Validation work (comparing the 2-claim method with established methods for diagnosing diabetes) in several Canadian jurisdictions indicates that the specificity is very good and the sensitivity varies between 69 and 95%, depending on the method used to define a gold standard.23 Although it is reassuring that our cumulative incidence rates of diabetes, by year of follow up, were similar to other studies that used OGTT measures for diagnosing diabetes, the fact that we did not observe a plateau in the rates after 10 years of follow up may be due to measurement error with the 2-claim rule. There is some risk that this method may accumulate false-positive cases over time, due to coding error in the claims data.23

In Canada, the recommendation for women with GDM is that they be ‘re-evaluated within 6 months of delivery with a 2-hour plasma glucose in a 75-g OGTT (preferred test) or a fasting plasma glucose test and be counselled on a healthy lifestyle’.2 Although the ICD-9 code for screening for diabetes is a separate code from a diabetes diagnosis (V77.1 versus 250), there are concerns that the 2-claim rule (e.g. two outpatient physician service claims with a diabetes code over a 2-year period) might overestimate the number of true DM cases if women are screened annually.

To assess whether the 2-claim rule was overestimating the number of women who developed diabetes, an analysis was conducted to determine the number of claims among those diagnosed with DM. Approximately 85% of the women identified with DM had more than the minimum of two physician claims in 2 years, and 75% had more than three claims. Since most women in our study with DM diagnosed had more than two physician visits with a diabetes code over a 2-year period, we feel that bias due to annual screening or coding error is small. As well, based on compliance in other jurisdictions, it is suspected that compliance rates in Nova Scotia are low for screening for diabetes among women with a history of gestational diabetes.24

It was not possible to distinguish type I and type II diabetes from our data. However, it is presumed that the vast majority of diabetes in this population would be type II diabetes. Information on OGTT results during pregnancy would have provided an indication of severity of GDM, but this was not captured in the database. However, insulin treatment for GDM was used as an indicator of severity and was found to be significantly associated with subsequent development of DM. As well, we were limited to analysing maternal weight, rather than body mass index, as height was not available in the database.

Conclusion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Commentary on ‘Diabetes mellitus following gestational diabetes: role of subsequent pregnancy’
  9. Acknowledgements
  10. References

The findings from this study do not support the hypothesis that subsequent pregnancy increases the risk of developing DM. Rather, the results suggest that in this population of primarily Caucasian women, severity of GDM is an important indicator of future development of DM. Indicators of severity, such as insulin use during pregnancy, neonatal hypoglycaemia, or recurrent GDM in a subsequent pregnancy, were associated with increased risk of DM.

Commentary on ‘Diabetes mellitus following gestational diabetes: role of subsequent pregnancy’

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Commentary on ‘Diabetes mellitus following gestational diabetes: role of subsequent pregnancy’
  9. Acknowledgements
  10. References

Gestational diabetes mellitus (GDM), often known as ‘pre-diabetes’, predisposes affected women to both type I and type II diabetes. Insulin resistance during the second half of pregnancy transiently tips vulnerable women over a threshold that defines diabetes. Delivery corrects this adverse metabolic milieu until an age-related decline in pancreatic beta-cell function, and usually, age-related weight gain tips the woman back into a diabetic state. It has been proposed that repeated pregnancies make repeated demands of pancreatic beta-cell function that eventually reduces their future capacity to synthesise insulin; a question addressed by Russell et al.

In a retrospective cohort study, 1401 nulliparous women who had gestational diabetes had medical records reviewed up to 13 years after the index pregnancy. Within this time period, 22% of women (42/195) who had recurrent GDM developed future diabetes compared with 7% of women (28/411) who did not have recurrent GDM and 16% of women (98/624) who did not become pregnant again. A tentative conclusion from this study is that women who did not have a subsequent pregnancy were not protected from future diabetes and having a subsequent pregnancy predisposed to future diabetes only if affected by recurrent GDM. This is a tentative conclusion because women identified late in the study period only had limited follow up. Longer follow up of the whole cohort, especially of multiparous women, may have identified beta-cell dysfunction, especially following repeated pregnancies. Furthermore, the absence of a population of nulliparous women who did not have GDM prevented comparison with a group perceived to be at low risk of future diabetes.

Only 32% of women (195/606) in Russell’s study who became pregnant again developed recurrent GDM. The majority of studies estimate a recurrence rate of GDM around 60%.1 A relatively short follow up is again the likely explanation, as contrary to some observations, the incidence of diabetes after GDM continues to increase with maternal age, as of course it does in the general population.

An increased risk of future diabetes was also found to be associated with a pre-pregnancy weight >86 kg, and two indicators of GDM severity are as follows: (1) the need for insulin therapy and (2) neonatal hypoglycaemia. These are not novel observations. More than 40% of women of age 25–44 years in the USA and UK are overweight or obese (body mass index > 25 kg/m2), and the incidence of GDM is increasing. Early identification of women at risk of type II diabetes provides an opportunity for lifestyle changes and drug therapies before the onset of diabetes-induced cardiovascular morbidity. Postpartum follow up of women with a history of GDM is however woeful. Affected women usually slip through the postpartum responsibilities of obstetrician, endocrinologist and primary care physician. Even in the developed world, only about one-third of women who have had GDM attend for postpartum follow up.2 Maybe it would help if we went back to calling GDM, ‘pre-diabetes’, because for most women and indeed their offspring, it is not over when the pregnancy ends!

David Williams

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Commentary on ‘Diabetes mellitus following gestational diabetes: role of subsequent pregnancy’
  9. Acknowledgements
  10. References

This study was funded by a grant from the IWK Health Centre. C.R. received studentship support from the Nova Scotia Health Research Foundation. L.D. is supported by Clinical Research Scholar Award from Dalhousie University. K.S.J. is a Peter Lougheed New Investigator and L.D. is a New Investigator of the Canadian Institutes of Health Research. The authors are grateful to the Reproductive Care Programme of Nova Scotia and the Population Health Research Unit for data access. Although this research is based, in part, on data obtained from the Population Health Research Unit, the observations and opinions expressed are those of the authors and do not represent those of the Population Health Research Unit.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Commentary on ‘Diabetes mellitus following gestational diabetes: role of subsequent pregnancy’
  9. Acknowledgements
  10. References
  • 1
    Metzger BE, Coustan DR. Summary and recommendations of the fourth international workshop-conference on gestational diabetes mellitus. Diabetes Care 1998;21(Suppl 2):B1617.
  • 2
    Canadian Diabetes Association Clinical Practice Guidelines Expert Committee. Gestational diabetes mellitus. Can J Diab 2003;27(Suppl 2):S99105.
  • 3
    MacNeill S, Dodds L, Hamilton DC, Armson BA, VandenHof M. Rates and risk factors for recurrence of gestational diabetes. Diabetes Care 2001;24:65962.
  • 4
    Aberg AE, Jonsson EK, Eskilsson I, Landin-Olsson M, Frid AH. Predictive factors of developing diabetes mellitus in women with gestational diabetes. Acta Obstet Gynecol Scand 2002;81:1116.
  • 5
    Damm P, Kuhl C, Hornnes P, Molsted-Pedersen L. A longitudinal study of plasma insulin and glucagon in women with previous gestational diabetes. Diabetes Care 1995;18:65465.
  • 6
    Dornhorst A, Rossi M. Risk and prevention of type 2 diabetes in women with gestational diabetes. Diabetes Care 1998;21(Suppl 2):B439.
  • 7
    Henry OA, Beischer NA. Long-term implications of gestational diabetes for the mother. Baillieres Clin Obstet Gynaecol 1991;5:46183.
  • 8
    Linne Y, Barkeling B, Rossner S. Natural course of gestational diabetes mellitus: long term follow up of women in the SPAWN study. BJOG 2002;109:122731.
  • 9
    Peters RK, Kjos SL, Xiang A, Buchanan TA. Long-term diabetogenic effect of single pregnancy in women with previous gestational diabetes mellitus. Lancet 1996;347:22730.
  • 10
    Diabetes Care Program of Nova Scotia. Diabetes in Nova Scotia 2002/03: Prevalence, Incidence, Mortality, Morbidity, and Utilization Using the National Diabetes Surveillance System. Diabetes Care Program of Nova Scotia. Halifax. 2005 [http://www.diabetescareprogram.ns.ca/statistics/NDSS/NDSS.pdf]. Accessed 9 April 2007.
  • 11
    CDC National Center for Chronic Disease Prevention and Health Promotion. Data and Trend, National Diabetes Surveillance System. Prevalence of Diabetes. 2005 [http://www.cdc.gov/diabetes/statistics/prev/national/tprevwfage.htm]. Accessed 9 April 2007.
  • 12
    Kim C, Newton KM, Knopp RH. Gestational diabetes and the incidence of type 2 diabetes: a systematic review. Diabetes Care 2002;25:18628.
  • 13
    Catalano PM, Vargo KM, Bernstein IM, Amini SB. Incidence and risk factors associated with abnormal postpartum glucose tolerance in women with gestational diabetes. Am J Obstet Gynecol 1991;165:91419.
  • 14
    Fair M, Cyr M, Allen AC, Wen SW, Guyon G, MacDonald RC. An assessment of the validity of a computer system probabilistic record linkage of birth and infant death records in Canada. Chronic Dis Can 2000;21:813.
  • 15
    Hux JE, Ivis F, Flintoft V, Bica A. Diabetes in Ontario: determination of prevalence and incidence using a validated administrative data algorithm. Diabetes Care 2002;25:51216.
  • 16
    Sullivan LM, Massaro JM, D’Agostino RBS. Presentation of multivariate data for clinical use: the Framingham study risk score functions. Stat Med 2004;23:163160.
  • 17
    Damm P, Kuhl C, Bertelsen A, Molsted-Pedersen L. Predictive factors for the development of diabetes in women with previous gestational diabetes mellitus. Am J Obstet Gynecol 1992;167:60716.
  • 18
    Coustan DR, Carpenter MW, O’Sullivan PS, Carr SR. Gestational diabetes: predictors of subsequent disordered glucose metabolism. Am J Obstet Gynecol 1993;168:113944; discussion 1144–5.
  • 19
    Lauenborg J, Hansen T, Jensen DM, Vestergaard H, Molsted-Pedersen L, Hornnes P, et al. Increasing incidence of diabetes after gestational diabetes: A long-term follow-up in a Danish population. Diabetes Care 2004;27:11949.
  • 20
    Buchanan TA, Metzger BE, Freinkel N, Bergman RN. Insulin sensitivity and B-cell responsiveness to glucose during late pregnancy in lean and moderately obese women with normal glucose tolerance or mild gestational diabetes. Am J Obstet Gynecol 1990;162:100814.
  • 21
    Albareda M, Caballero A, Badell G, Piquer S, Ortiz A, De Leiva A, et al. Diabetes and abnormal glucose tolerance in women with previous gestational diabetes. Diabetes Care 2003;26:1199205.
  • 22
    Statistics Division, Nova Scotia, Department of Finance. 2001 Census of Canada, Nova Scotia Perspective. 2003. Nova Scotia Department of Finance: Halifax, Nova Scotia.
  • 23
    Kephart G, Casey J, Ranger R, Dunbar P, Karlovic Z. The Development and Validation of an Alternative Case Definition for the National Diabetes Surveillance System. Report to the Data Validation Working Group, National Diabetes Surveillance System. 2003. Diabetes Care Program of Nova Scotia: Halifax, Nova Scotia.
  • 24
    Smirnakis KV, Chasan-Taber L, Wolf M, Markenson G, Ecker JL, Thadhani R. Postpartum diabetes screening in women with a history of gestational diabetes. Obstet Gynecol 2005;106:1297303.
  • 1c
    Bottalico JN. Recurrent gestational diabetes: risk factors, diagnosis, management and implications. Semin Perinatol 2007;31:17684.
  • 2c
    Kim C, Tabaei BP, Burke R, et al. Missed opportunities for type 2 diabetes mellitus screening among women with a history of gestational diabetes mellitus. Am J Public Health 2006;96:16438.