Our study of a case-manager program to enhance detection of postpartum glucose intolerance in women with GDM reveals two principle findings: first, those women who failed to return for post-partum glucose screening had more severe GDM than women who did not return. Women who failed to return were more likely to have previously had GDM, had higher pre-pregnancy weight and higher point estimates for all glucose levels at diagnosis of GDM, were more likely to require medication and specifically insulin to treat their GDM, and had less control of their GDM than women who returned for post-partum glucose screening. Our second principal finding was the high prevalence of isolated elevated 2-hour glucose levels (either IGT or diabetes) with normal fasting glucose values, even using the new lower cut-point for IFG of 100 mg/dl. Of the women with a post-partum OGTT with IGR, approximately one-third had abnormalities of the 2-hour value alone: 31% (4 of 13) of the women with diabetes and 35% (41 of 117) of those with IGR; therefore, if only fasting glucose levels were obtained a substantial proportion of women with IGT would not be identified.
Our results are similar to results recently published by Russell et al who report that only 45% of women with GDM in their multi racial/ethnic cohort of 344 women underwent postpartum glucose testing of whom 36% had IGR postpartum15
. However, in contrast to our results the only factor they identified as being strongly associated with postpartum glucose testing was attendance at the postpartum visit15
The prevalence of IGR immediately following GDM was high in our population; over one third of the women who returned for post-partum screening (fasting only or a complete OGTT) had IGR. Moreover, this likely underestimates the true prevalence of IGR following GDM in our population because markers for GDM severity were lower in women that returned than in women who failed to return, and an OGTT was not completed by all study participants. Explanations for the high prevalence of IGR in our study include the change in the cut-point for IFG from 110 mg/dl to 100 mg/dl as well as the high background risk of type 2 diabetes in our predominantly Mexican and Mexican American study population19
The high rate of IGR we found in the current study also appears to represent an increase over the previously detected rate of IGR within our own population 2
. From 1995–1997, approximately 28% of the women who returned to our clinic met the criteria for IGR based on 1997 ADA criteria20
. The cohort of women represented in that study was substantively different from the more recent cohort described in the current paper only in the degree of effort made to ensure follow-up testing, and therefore the resultant follow-up testing rate (18% versus 57% respectively). This would appear to corroborate our current finding that it is the women at highest risk who do not return for testing.
Recently, the Diabetes Prevention Program was one of several clinical trials to indicate that either through diet and exercise, or with the aid of a pharmacological agent, it was possible to lower the incidence or delay the onset of diabetes among individuals at high risk of the disease8–12
. Hence, the optimum way to reduce the risk associated with diabetes may be preventing diabetes itself, either by altering lifestyle, or by using pharmacologic agents. The burden of diabetes is especially high in women who acquire it at a young age, such as those with IFG or IGT in the early postpartum period following GDM.
Reducing the incidence of type 2 diabetes following GDM also reduces the inherent risks to future offspring of exposure to a diabetic intrauterine environment. Early in pregnancy uncontrolled type 2 diabetes is associated with increased risk of congenital malformations and spontaneous miscarriages. In a study in the UK, 9.9% of women with uncontrolled type 2 diabetes had an infant with a congenital malformation, an 11-fold greater risk than the national statistics report21
. In a study of Hispanic women who had NIDDM in southern California 11.7% of infants were born with major congenital anomalies compared to 2% in the background non-diabetic population22
. In that study, the only independent predictor of anomalies was the A1C value in the first trimester, emphasizing the importance of recognition of the diabetic state prior to and early in pregnancy. In addition, maternal diabetes during gestation has been associated with high birth weight, increased childhood and adult obesity, increased risk of type 2 diabetes, and earlier onset of type 2 diabetes in the offspring23–29
Moreover, children exposed in utero
to maternal diabetes are at higher risks of obesity and diabetes than their unexposed siblings, suggesting that the increased risk to the exposed offspring is not exclusively genetic, nor affected exclusively by post-natal exposures, such as family nutritional practices 30, 31
Our population is over 90% Mexican American women; therefore, the results may not apply to women in the general U.S. population. However, Mexican Americans comprise 60% of Hispanics in the U.S. and Hispanics are the fastest growing and largest ethnic minority group32
. Mexican Americans are not only more likely to have diabetes, but their diabetes is more severe33, 34
and they have higher cardiovascular mortality risk35–37
than non-Hispanic whites. Evidence from NHANES III, the San Antonio Heart Study, and a study of school aged children in San Antonio suggests that the prevalence of obesity and diabetes is rapidly increasing in Mexican Americans19, 38–42
. Hence, it is imperative that diabetes research include the Mexican American population to ensure generalizability to this fast growing segment of our population. Finally, because the Institutional Review Board approved prospective data collection on all patients with diabetes during pregnancy at the University Hospital in San Antonio, TX, our study does not suffer from volunteer bias which improves the generalizability of our study to its source population, namely low income Mexican American women.
Additional limitations of our study included the absence of information on socioeconomic status at the level of the individual and the absence of information directly addressing why women may not return for postpartum glucose testing. We have examined clinical factors and their relationship to whether or not a woman returns for postpartum glucose screening; however, to improve our ability to screen women postpartum it would be pertinent to also examine socio-cultural explanations. Interestingly, the single factor we had information on which may address logistically why women were unable to return for postpartum testing was associated in the expected direction with whether a woman returned. That is, women who failed to return on average had more living children than women who returned for postpartum glucose testing.
In summary, in a population largely composed of Mexican American women postpartum glucose screening including an OGTT following GDM identifies a significant proportion of the population with IGR. Standard clinical procedures should be developed and utilized to enable and encourage all women to return for postpartum glucose screening.