Jean M. Lawrence1
(1)
Department of Research and Evaluation, Kaiser Permanente Southern California, 100 South Los Robles, Pasadena, CA 91101, USA
Jean M. Lawrence
Email: Jean.M.Lawrence@kp.org
Abstract
This chapter summarizes the methodological challenges inherent in the surveillance of gestational diabetes mellitus (GDM) and the comparisons of trends in GDM across various studies and populations. The results of selected studies that examined the prevalence or incidence of GDM using data from population-based studies are compared both in terms of methodology and study results. Population-based studies of the trends in GDM have been conducted using samples from national databases, from specific metropolitan areas and regions, states and provinces, and from managed health care organization enrollees. The way in which GDM is defined clinically, case definitions based on different blood glucose value thresholds, and differences in the types of data used to construct both the numerators and denominators to examine trends in GDM can contribute to differences in results across these population-based studies. The racial and ethnic composition of the populations being studied as well as the age distribution of the women giving birth also has a significant impact on the trends observed in these populations.
4.1 Introduction
Recent studies have shown that the prevalence of gestational diabetes mellitus (GDM) has increased by 10–100% over the past 20 years, with greater increases observed among women from racial and ethnic minority groups.1 Such an increase in GDM has important public health implications for these women as well as their offspring who are exposed to maternal hyperglycemia in utero. Women who develop GDM are at increased risk for GDM in subsequent pregnancies and are at increased risk for developing type 2 diabetes in the years after delivery.2, 3 Additionally, infants exposed to GDM in utero are at increased risk for obesity and future development of type 2 diabetes.4, 5 In this chapter, we describe the methodological challenges of conducting population-based studies of GDM and describe the trends in GDM prevalence over the past 20 years based on selected population-based studies from the United States, Canada, and Australia; chapter 6 provides an overview of the burden of GDM on developing countries.
4.2 Methodological Challenges in GDM Epidemiologic Research
Researchers designing epidemiologic studies to assess the trends in GDM face several methodological challenges, including how GDM is defined, the various blood glucose thresholds that are used to define women as having GDM, the proportion of the population screened for GDM during pregnancy, the sources of data that are available to identify women with GDM in population-based studies, and the terminology used (incidence vs. prevalence) when describing trends over time. Each of these challenges and how they affect the ability to compare study findings over time and across populations will be discussed in the sections that follow.
4.2.1 Definition of GDM
GDM is defined as carbohydrate intolerance of varying degrees of severity with onset or first recognition during pregnancy.6, 7 This definition applies regardless of treatment during pregnancy (insulin, glyburide, or dietary modification) and whether the condition persists after pregnancy. Based on this definition, women diagnosed with GDM include women with undiagnosed pregestational diabetes mellitus (PDM) as well as women with hyperglycemia induced by pregnancy. Women of childbearing age are not routinely screened for diabetes before pregnancy and large population-based studies of women screened for diabetes prior to pregnancy have not been undertaken. Additionally, screening for glucose intolerance that persists into the postpartum period among women with a history of GDM is variable, with few studies reporting postpartum testing with fasting plasma glucose or an oral glucose tolerance test in excess of 50% of women in recent periods.8–14 In a recent review of 13 studies that reported glucose abnormalities after pregnancies complicated by GDM,15 the prevalence of women with GDM who developed type 2 diabetes tested in periods ranging from 4 weeks to 1 year after delivery ranged from 2.5%16 to 16.7%.17 It is likely that some of these women had undiagnosed PDM that was identified during pregnancy.
4.2.2 Criteria Used to Define GDM
Various diagnostic criteria based on fasting and non-fasting blood glucose thresholds values are used to characterize women as having GDM (Table 4.1). These include definitions developed by the World Health Organization (WHO),18 the National Diabetes Data Group (NDDG),19 the American Diabetes Association (ADA),20 the Canadian Diabetes Association (CDA),21 the European Association for the Study of Diabetes (EASD),22 and the Australasian Diabetes in Pregnancy Society (ADIPS).23 Differences in defining thresholds that must be met or exceeded result in disparate estimates of GDM prevalence; prevalence estimates are particularly difficult to interpret when women with GDM are diagnosed based on different diagnostic criteria both within and between studies. A uniform approach to characterizing GDM, applied across multiple populations in the USA and other countries, would facilitate more direct comparisons of GDM prevalence within and between populations. Forthcoming results from the Hyperglycemia and Adverse Pregnancy Outcomes (HAPO) study may provide such a definition.24, 25 The HAPO study is discussed in detail in chapter 2.
Table 4.1
Various diagnostic criteria and authorities for gestational diabetes mellitus
Reference |
Authority |
Glucose load, g |
GCT or OGTT |
Number of results required for diagnosis (n ≥ mmol/L) |
Blood glucose values (≥ mmol/L) |
|||
f/nf |
Fasting |
1-h |
2-h |
3-h |
||||
Alberti and Zimmet18 |
WHO |
75 f |
OGTT |
1 |
7.0 |
7.8 |
||
Anonymous19 |
NDDG 1979 |
100 f |
OGTT |
2 |
5.8 |
10.5 |
9.2 |
8.0 |
Anonymous20 |
ADA |
50 nf |
GCT |
1 |
7.8a |
|||
Anonymous20 |
ADA |
75 f |
OGTT |
2 |
5.3 |
10.0 |
8.6 |
|
Anonymous20 |
ADA |
100 f |
OGTT |
2 |
5.3 |
10.0 |
8.6 |
7.8 |
Meltzer et al21 |
CDA |
50 nf |
GCT |
– |
7.8 |
|||
Meltzer et al21 |
CDA |
75 f |
OGTT |
2 |
5.3 |
10.6 |
8.9 |
|
Brown et al22 |
EASD |
75 f |
OGTT |
1 |
6.0 |
9.0 |
||
Hoffman et al23 |
ADIPS |
50 nf |
GCT |
– |
7.8 |
|||
Hoffman et al23 |
ADIPS |
75 nf |
GCT |
– |
8.0 |
|||
Hoffman et al23 |
ADIPS Australia |
75 f |
OGTT |
1 |
5.5 |
8.0 |
||
Hoffman et al23 |
ADIPS New Zealand |
75 f |
OGTT |
1 |
5.5 |
9.0 |
f/nf Fasting/non-fasting; GCT Glucose challenge test; OGTT Oral glucose tolerance test
a If the result on the glucose challenge test (screening) is ≥10.3 mmol/L, GDM is diagnosed without further testing
4.2.3 Population-Based Screening for GDM
The estimate of the prevalence of GDM in any population is contingent upon screening for GDM in that population since the diagnosis of GDM requires screening for hyperglycemia during pregnancy. Studies that include both unscreened and screened women in the denominator may draw a different conclusion than studies which are able to restrict their analysis to screened women and conduct sensitivity analysis to determine the impact of differential screening on the overall prevalence of GDM in their population. For example, studies of trends in GDM that have discussed screening in the populations under study have reported an increase in screening over time,26 a consistently high rate of screening across the study period,27, 28 and that screening was a component of usual care,9 while other studies have not reported on the proportion of women screened for GDM in their populations under study.29–32 In studies that do not limit their denominator to the population screened for GDM, some of the observed changes in the prevalence of GDM can be attributed to the increase in screening in the population over time, which would result in more complete case ascertainment over the years of the study. Studies that use hospital discharge diagnostic codes or infant birth certificates in the absence of information on prenatal screening for GDM include screened and unscreened women in their denominators when calculating trends in GDM. These studies are not able to make adjustments for inclusion of women who are not screened nor can they conduct sensitivity analysis to determine whether including women that are not screened has an impact on the results of their study. Keeping these limitations in mind, the results of these studies should be interpreted cautiously.
4.2.4 Sources of Information Used to Identify Persons with GDM
Population-based studies of GDM have used a variety of ways to identify women with GDM in their study samples. The information available to identify women as having GDM is often dependent on the populations under study. Studies from health plan-based populations,9, 26–28 for example, may be based on blood glucose test results from oral glucose tolerance tests and glucose challenge tests conducted during pregnancy and as such they can apply a consistent criteria throughout the study period to identify women with GDM that is independent of hospital discharge diagnostic codes or infant birth certificates. Studies that must rely solely on information from infant birth certificates or hospital discharge data, as is the case for most studies from state, regional, or national populations, cannot determine if a uniform criteria was used to identify women as having GDM throughout the study period.29–32In their study of trends in the prevalence of GDM from 1989 through 2004 using data from the National Hospital Discharge Survey (NHDS), the authors note that the change in the criteria used to identify women as having GDM from the NDDG definition (19) to the Carpenter and Coustan criterion (20) during the period of the study, the latter of which had lower thresholds at which women were considered to have GDM, may have been the most importantexplanation for the increase in GDM prevalence in recent years in their study.30
Studies have identified women with GDM based on infants’ birth certificates exclusively, infants’ birth certificates or hospital discharge codes, hospital discharge diagnostic codes exclusively, hospital discharge code or laboratory test results, laboratory test results exclusively, local clinical databases, and medical record abstraction (Table 4.2). The ability of these sources of information to accurately identify women as having GDM is variable. The 1989 version of the US birth certificate did not differentiate between PDM and GDM, while the 2003 revision (the most recent national version of the US birth certificate) captures information on each of these conditions in separate categories.33 The ability to exclude women with PDM from the denominator when studying trends in GDM is important as several recent studies have demonstrated that the prevalence of PDM in the populations giving birth is increasing.28, 34 In a recent review paper, Devlin and colleagues reviewed 12 studies that evaluated the reliability of US birth certificates and hospital discharge diagnostic codes to identify births complicated by maternal diabetes.35Eight of these studies distinguished between PDM and GDM. Six studies conducted between 1989 and 2007 used infant birth certificates or hospital discharge to identify women with GDM and linked these data to other sources for validation. The sensitivities for the four studies that used birth certificates validated against a variety of sources to identify GDM ranged from 46 to 83%, while the sensitivities for the two studies that used hospital discharge data validated against medical records were 71% and 81%.35 The identification of women with PDM using their infants’ birth certificates performed less well, with sensitivities ranging from 47 to 52%. The sensitivities were 78 and 95% to identify women with PDM using hospital discharge diagnostic codes.
Table 4.2
Selected studies of gestational diabetes mellitus in the USA, Canada, and Australia: 1985–2006
Population (geographic area) |
Sample |
N deliveries, race/ethnicity |
Study period (years)a |
GDM prevalence |
Relative changeb(%) |
Criteria for GDM definition |
Source of GDM information |
Age adjusted |
United States |
||||||||
USA30 |
Country |
58,922,266 Race distribution not reported for total sample |
1989–2004 |
1.9% (1989–1990) to 4.2% (2003–2004) |
122 |
Unknown |
National Hospital Discharge Survey, ICD-9 code |
Age-specific |
Montana38 |
State |
43,543 12% AI (below 88% White) 88% White; |
2000–2003 |
White: AI 2.4–2.9% (below 2.0–2.2%) 2.0–2.2% |
10 21 |
Unknown |
Infant birth certificates |
No |
Minnesota32 |
State |
130,671 77% NHW 4.2% Asian |
1993, 2003 |
2.6–3.5% 2.5–3.2% 1.9–5.5% |
35 28 189 |
Unknown |
Infant birth certificates |
Yes |
Los Angeles County31 |
County |
2,156,459 Race not reported |
1991–2003 |
1.5–4.8% |
220 |
Unknown |
Hospital discharge data, ICD-9 code |
Yes |
New York City39 |
City |
>1.5 million 43% Hispanic or Caribbean |
1990, 2001 |
2.6–3.8% |
46 |
Unknown |
Infant birth certificates |
Yes |
New York City29 |
City |
1,067,356 Many ethnic categories and sub-categories |
1995–2003 |
5.0–5.4% |
8 |
Unknown |
Infant birth certificates and hospital discharge data |
Yes |
Population (geographic area) |
Sample |
N deliveries, race/ethnicity |
Study period (years)a |
GDM prevalence |
Relative changeb(%) |
Criteria for GDM definition |
Source of GDM information |
Age adjusted |
Northern California26 |
MCO |
267,051 45% NHW |
1991–2000 |
3.7–6.6% 5.1–6.9% |
78 35 |
2000 ADA |
Laboratory test results Laboratory test results OR ICD-9 code |
Yes |
Colorado27 |
MCO |
36,403 61% NHW |
1994–2002 |
2.1–4.1% |
95 |
NDDG |
Clinical database |
Yes |
Southern California28 |
MCO |
209,287 52% Hispanic, 26% NHW |
1999–2005 |
7.5–7.4% |
−1 |
2000 ADA |
Laboratory test results |
Yes |
Oregon9 |
MCO |
36,251 60% White |
1999–2006 |
2.9–3.6% |
24 |
NDDG |
Laboratory test results OR ICD-9 code and pharmacy |
No |
Canada |
||||||||
Manitoba40 |
Province |
324,605 93.1% Not FN 6.9% FN |
1985–200 4 |
2.3–3.7% 1.8–3.0% 6.8–8.1% |
61 67 19 |
Changed over time |
Standard prenatal form |
No |
Ontario37 |
Province |
659,164 Race not reported |
1995–2002 |
3.2–3.6% |
12 |
Unknown |
Hospital Discharge data |
No |
James Bay, Ontario41 |
Region |
1,298 FN (Cree) |
1987–1995 |
8.5% |
N/A |
NDDG |
Medical record review |
N/A |
James Bay, Quebec47 |
Region |
579 FN (Cree) |
1995–1996 |
12.8% |
N/A |
NDDG |
Medical record review |
N/A |
Australia |
||||||||
New South Wales42 |
State |
370,703 51% Australian, 28% Asian |
1998–2002 |
4.0–5.1% |
27 |
(Primarily) ADIPS |
Inpatient statistics (Hospital), midwives data collection, ICDM-10AM (Australian modification) |
No |
New South Wales43 |
State |
956,738 75% Australia or New Zealand |
1995–2005 |
3.0–4.4% |
45 |
ADIPS |
Midwives data collection |
Yes |
South Australia45 |
State |
230,011 98% Non-aboriginal 2% Aboriginal |
1988–1999 |
≈1.7–3.2% ≈5.3–5.9% |
88 11.3 |
ADIPS or WHO |
Laboratory test results |
Yes |
MCO managed care organization; ADA American Diabetes Association; NDDG National Diabetes Data Group; WHO World Health Organization; OGTT oral glucose tolerance test; FPG fasting plasma glucose; ADIPS Australasian Diabetes in Pregnancy Society; NHW non-Hispanic White; AI American Indian; FN First Nation
a For years with dashes between them (example, 1991–2000), data from all years were included in the study. For years with commas between them (example, 1993, 2003) only data from those 2 years are included in the study
b Relative change is calculated based on the prevalence in first and last year (or period) of the study
In a study conducted in New South Wales Australia, two population-based data sources; the Midwives Data Collection (birth data), which included information on maternal characteristics, pregnancy, labor, delivery, and infant outcomes and the Admitted Patient Data Collection (hospital data), a census of all public and private inpatient hospital discharges, were compared against the medical record for approximately 1,200 women. The sensitivities of birth data and hospital data to identify women with GDM were 63.3 and 68.6%, respectively, while the sensitivities of these data sources to identify women with PDM were 45.1 and 100%, respectively.36 Thus, when the sensitivity of hospital discharge diagnostic codes were compared to the sensitivity of the birth certificate to identify women with PDM and GDM, hospital discharge codes were more likely to correctly identify women as having PDM and GDM than was information on the birth certificates. The differences in the sensitivity between the two sources of data were greater when identifying women with PDM than when identifying women with GDM.
4.2.5 Terminology – Incidence vs. Prevalence
Studies that describe trends in GDM have used the terms “prevalence” and “incidence” somewhat interchangeably, although the term prevalence has been more commonly used. Incidence is the number of cases of a disease or illness newly identified in a specific population in a specified period of time. Prevalence is the total number of persons having a disease or condition in a specific population during a specific period of time. To be characterized as having the condition in the period under study, a person may either develop the condition before the beginning of the study period or be newly diagnosed with the condition during the study period. In either case, to be included in the denominator the person with the condition must be part of the population under study during the specified period. Prevalence can be further categorized as point prevalence (a specific point in time), period prevalence (at any time during a specific period), and annual prevalence (at any time during the year).
Most studies of GDM examine the number of women giving birth after a specific point in the pregnancy (i.e., 20 weeks gestation) or whose pregnancies result in live birth during a specific period, often one calendar year, who develop GDM during that pregnancy. Given that the duration of pregnancy is, on average, 40 weeks and GDM is most often identified between 24 and 28 weeks, GDM may develop during the calendar year in which the delivery occurs or the previous calendar year, as would be the case for women who give birth early in the year. Ferrara distinguished between cumulative incidence and prevalence of GDM based on the composition of the denominator. Studies which limit the denominator to women with screened pregnancies, regardless of whether they resulted in a live birth, yield cumulative incidence rates while studies including only women with live births yield prevalence estimates.1 The composition of the denominator will vary based on the source of information used to identify women at risk for the outcome who are to be included in the denominator. Studies based on infant birth certificate data cannot include women who were pregnant but did not have a live birth, studies using hospital discharge diagnoses codes may include women who had late second trimester or third trimester fetal losses resulting in hospitalization, while clinical databases may include women with fetal losses beyond a specific time in pregnancy. Thus, when comparing trends in GDM across studies, it is important to take into account the composition of the sample that comprise both the numerator (women with GDM) and the denominator (the population at risk for GDM) to determine the comparability of finding across studies instead of solely relying on the terminology (incidence or prevalence) used to describe the methodological approach in any given study.
4.2.6 Other Methodological Issues Affecting Studies of GDM
4.2.6.1 Maternal Race and Ethnicity
A variety of other factors may impact our ability to compare trends in GDM across studies and to adjust for key risk factors that may impact these trends. The availability of accurate information to categorize maternal race/ethnicity is important as it impacts on the studies’ ability to reliably provide race/ethnicity-specific prevalence or incidence estimates and to adjust for changes in the racial/ethnic composition of the population when doing trend analyses. One large study conducted over a 7-year period in the province of Ontario in Canada which focused on the risk of developing diabetes after a GDM-affected pregnancy did not provide race/ethnicity specific GDM estimates since this study was conducted using administrative data which did not include this information.37 While this study provides information on the trend in the prevalence of GDM in Ontario, it does not provide information on how these trends may be differentially affected by women from the various racial and ethnic groups in that province over time. Studies that must rely on data from administrative sources to categorize the race/ethnicity of the women in the population have a significant amount of missing data and may define their categories less precisely than studies that include maternal race/ethnicity from infant birth certificates. For example, in an analyses of data from the NHDS, the authors reported that data on race was missing for up to 20% of births from 1995 to 2001 and up to 29% of births from 2002 to 2004. Additionally, Hispanic ethnicity was not included as a separate category from white or black race, and women of Asian or Pacific Islander race were excluded from the analyses due to a small number of annual births in this racial group.30 Thus, while this was the only study identified that provided national estimates of GDM prevalence by US geographic region and maternal age category, its race/ethnicity specific conclusions must be interpreted cautiously as the racial/ethnic group with the highest GDM prevalence, Asian and Pacific Islanders, was excluded from the analysis and women of Hispanic ethnicity, another group with a high GDM prevalence could not be analyzed separately, but were combined into the two main race categories, White and Black, most likely increasing GDM prevalence in these groups due to the proportion of women who were both Hispanic and white or black. Other studies have derived information on maternal race/ethnicity from infant birth certificates,9, 28, 29, 32, 38, 39 hospital records,26 self-reported First Nation status,40 band number to identify native American status,41 and country and region of maternal birth.42, 43
4.2.6.2 Maternal Prepregnancy Weight and Body Mass Index
Prepregnancy body mass index (BMI) is a risk factor for GDM, with overweight and obese women having a higher risk of GDM than women of healthy weight.44 Based on data from 75,403 women from 26 states and New York City who self-reported prepregnancy weight and height on the Pregnancy Risk Assessment Monitoring System (PRAMS) survey from 2004 and 2005, 23% of women giving birth were overweight and 19% were obese in 2004 and 2005.45 Maternal BMI has rarely been included in population-based studies of GDM given the limited availability of maternal height and weight in these study samples. In the study of trends in GDM prevalence in New York City, prepregnancy weight but not height was used as an adjustment factor in models presenting trends in GDM from 1995 to 2003 but was not discussed in the paper.29 In a study from a managed health care plan in Oregon, prepregnancy BMI was included based on information from the medical records, although BMI was discussed in the context of postpartum screening (the primary aim of the study) and not in relation to trends in GDM.9
As of 2007, 24 states and Puerto Rico included prepregnancy height and weight on their birth certificates (personal communication, Joyce Martin, National Center for Health Statistics). However, published reports which evaluate the validity of prepregnancy height and weight on the birth certificate compared to a gold standard such as a clinical database are lacking and the results would most likely be variable by location within states as well as between states. Electronic medical records, which are being implemented across the country in a variety of health plans and practices, may ultimately provide additional information on BMI for women giving birth. The availability of these data may increase our understanding of the contribution of maternal prepregnancy BMI as well as weight gain during pregnancy in the development of GDM, particularly as it may differ by racial or ethnic group. Another chapter in this book includes a more comprehensive discussion of the role of obesity in GDM.
4.3 Trends in Prevalence of GDM
Despite the methodological challenges in studying the epidemiology of GDM, we must rely on these population-based studies to further our understanding of the trends in GDM prevalence over time. We must consider the strengths and limitations of each study when comparing and contrasting estimates across studies. There are often significant trade-offs between the size and diversity of the populations available for study in terms of geographic diversity (counties, states, regions, and countries), insurance status (insured women only vs. all women regardless of insurance status), and the amount of detailed information available to identify women with GDM and characterize these trends in population-based studies. In the selected studies over the past 20 years, summarized in Table 4.2, almost all studies reported an increase in the prevalence of pregnancies affected by GDM, although the magnitude of the increase varied significantly by the populations studied and the methods use to undertake these studies.
4.3.1 Comparison of Trends Across Studies
The change in prevalence for the studies reviewed from the USA, Canada, and Australia varied significantly by region, data source, racial/ethnic composition of the population giving birth, and years included in the study but almost all demonstrated increases in GDM over the course of the period under study with one exception28 (Table 4.2). The greatest increase in prevalence of GDM was reported in a study of Los Angeles County births, which observed a 220% increase based on hospital discharge diagnostic codes from 1991 to 200331 while there was no significant change observed in a managed health care population in Southern California (a 6-county area which included Los Angeles County) based on laboratory-identified cases of GDM from 1999 to 2005.28 The prevalence reported by the Los Angeles County study from 1999 through 2003, the period during which these two studies overlap, ranged from approximately 4.2 to 4.5%31 based on hospital discharge diagnostic codes. This estimate was significantly lower than the prevalence reported by the managed health care population study for these same years, which ranged from 7.5 to 8.2% using laboratory test results.28 Differences in the results between the two studies may be based on several factors which include the difference in the composition of the denominator (screened plus unscreened women in the Los Angeles County study vs. screened women only in the health plan study) and the way that cases were identified (hospital discharge diagnostic codes in the Los Angeles County study vs. a consistent cut-point applied to blood glucose test results from the laboratory database in the health plan study).
When results were compared within one large managed health care plan across four different regions of the USA (northern and southern California [two studies], Oregon, and Colorado),9, 26–28 the study from the southern California region exhibited the highest prevalence of GDM based on laboratory-identified cases but no significant change in prevalence from 1999 to 2005 while the other studies reported an increase in prevalence, with a doubling reported in Colorado from 1994 to 2002, a 25% increase in Oregon from 1999 to 2006, and an increase of 35 or 78% depending on the case definition used in northern California from 1991 to 2000. The two California studies were the most comparable, as they both restricted their denominator to women with documented screening for GDM during pregnancy, and used a consistent laboratory cut-point to identify women with GDM throughout the study period. Southern California presented their results based solely on laboratory-identified cases while northern California reported results using this same laboratory-defined case definition in addition to a second case definition that included laboratory defined cases and cases identified using hospital discharge diagnostic codes.26, 28 Both California studies reported a higher prevalence or cumulative incidence of GDM than did the studies in the other regions. Non-Hispanic white women, who are at lower risk for GDM than women from other racial/ethnic groups, comprised about 60% of the populations giving birth in the samples from Oregon and Colorado but comprised only about one-quarter of women in southern California and less than half of the women in northern California. The Oregon and Colorado studies relied on the NDDG criteria to establish GDM clinically while the California studies used the ADA criteria, which require a lower blood glucose level to characterize the women as having GDM.
The single study that included the entire US population (almost 59 million births over 16 years) was based on results from the NHDS and reported a 122% increase in GDM from 1989–1990 to 2003–2004.30As previously noted, the results of this study could be affected by the differences in diagnostic criteria used during the study period, inclusion of Hispanic women in the specific race categories, and inclusion of unscreened women in the denominator. Among white women, the prevalence of GDM increased by 80% while a 172% increase was observed among Black women. Increases in GDM were observed in all geographic regions of the country.30 Studies using birth certificate data from Montana38 and Minnesota32 reported an increase of 10% over 3 years and a 28% increase over 10 years, respectively, among white women. The increase among American Indian women in Montana was 21%, about twice that of white women, while the increase among Asian women in Minnesota was 189%, or almost seven times the increase observed among white women in the same period. The two studies from New York City reported quite different prevalence estimates of GDM. In the study based exclusively on birth certificates, a 46% increase was observed in the year 2001 as compared to the prevalence of the year 199039 while in the study which used both birth certificates and hospital discharge data, an 8% increase was observed in the year 2003 when compared with the prevalence of the year 1995.29 However, as expected, the prevalence of GDM was higher in the study using birth certificates and hospital discharge data combined than in the study that used birth certificate data only.
In the two studies conducted in Canada which reported trends in prevalence of GDM, a 12% increase in the prevalence was reported in the province of Ontario from 1995 to 2002,37 while a 61% increase was observed among women in the province of Manitoba from 1985 to 2004.40 As we previously mentioned, the study conducted for the province of Ontario was not able to examine the differences in prevalence of GDM by race/ethnicity as this information was not included in their administrative data.37 However, in the study conducted in the province of Manitoba, the overall increase in prevalence was 61%, with a smaller increase observed among First Nation women (19%) compared to 67% among non-First Nation women.40 However, while the absolute increase was lower, the prevalence of GDM was much higher in First Nation women (8.1%) as compared to non-First Nation women (3.0%) in 2004, the last year of their study.
In the two studies conducted in the state of New South Wales, Australia, a 27% increase in the prevalence of GDM was reported from 1998 to 2002 based on hospital information, midwives data collection, and hospital discharge codes combined42 whereas a 45% increase was observed in the same state from 1995 through 2005, from a prevalence of 3.0% in 1995 to 4.4% in 2005.43 In the state of South Australia, the prevalence of GDM increased by 88% among non-Aboriginal women and 11.3% in Aboriginal women from 1988 to 1999 based on laboratory test results only.46 Studies that examined trends in GDM within European countries could not be identified for inclusion in this review.
4.3.2 Comparison of Prevalence by Racial and Ethnic Group
Studies of GDM conducted in the US have consistently demonstrated that women of Native American, Hispanic, and Asian race/ethnicity are at greater risk for developing GDM than are non-Hispanic white women,26–28, 30, 31while GDM prevalence for African American women has been reported as being about the same26, 28, 30, 31 or higher29, 32 than non-Hispanic white women, depending on the study reviewed. In several studies that had significant diversity in the racial and ethnic composition of their populations under study and sample sizes that were sufficient to produce stable estimates of prevalence,26, 28, 29, 31 women of Asian and Pacific Islander race/ethnicity have the highest prevalence of GDM. Ferrara et al26 reported that the cumulative incidence of GDM was 10.9% in 2000, the last year of their study; while Lawrence et al28 reported a prevalence of 11.8% among Asian/Pacific Islander women (the majority of whom were Asian) in 2005. Hispanic women had the next highest burden of GDM during pregnancy, reporting estimates of 7.6 and 8.5%, respectively, in these two studies. Non-Hispanic white and African American women had a similar burden of GDM in each study.
Few studies have had sufficient sample size to report on the prevalence of GDM among Native American women. Two studies describing the trends in the prevalence of GDM among First Nation Cree women in the provinces of Ontario and Northern Quebec in Canada reported that 8.5% (1987–1995) and 12.8% (1995–1996) of the women in their study samples, respectively, developed GDM during their pregnancies.41, 47 In Manitoba Canada, the prevalence of GDM rose from 6.8% (1985–1989) to 8.1% (2000–2004) among First Nation women.40 During the same period, the prevalence rose from 1.8 to 3.0% among non-First Nation women. GDM prevalence in American Indian women in Montana in 2004 was 2.9% compared to 2.2% in White women based on information on infant birth certificates.38
Within racial/ethnic groups that are often combined when reporting results from US studies, there are significant differences in the prevalence of GDM by subgroups of women included in these categories. Several studies have compared the prevalence of GDM across women from Asian groups using consistent methodologies for the comparison. Savitz et al reported a risk of GDM of 6.6% for East Asian women, ranging from 3.0% for Japanese women and 3.3% for Korean women to 7.7% for Taiwanese women and 9.9% for women from Hong Kong among women living in New York City.29 Among South Central Asian women, the risk was 14.3%, ranging from 4.6% for Iranian women to 16.2% for Pakistani women and 21.2% for Bangladeshi women. In this study, the only groups of Asian women that did not have a higher risk for GDM than non-Hispanic white women after adjustment for age, education, prepregnancy weight, parity, and smoking status were Japanese and Iranian women. Lawrence et al reported a similar difference in prevalence of GDM among Asian American women in southern California, with Japanese and Korean women having a prevalence of 6.8 and 7.8%, respectively, while Indian (12.7%), Filipina (12.6%), Vietnamese (12.2%), and Southeast Asian women (9.9%) all had a high prevalence of GDM.48 Rao et al found a significant difference in GDM prevalence among Japanese, Chinese and Filipina women, with a prevalence of GDM of 3.4, 6.5, and 6.1% respectively.49 In a study of GDM prevalence in Australia, women born in Northeast and Southeast Asia as well as South Asia had a prevalence of GDM of 9.4 and 10.5%, respectively, in comparison with 2.7% among women born in Australia and New Zealand.43
In the New York City study,29 women that would traditionally be grouped as black or African American in other studies, including women defined as African, from Sub-Saharan Africa, and non-Hispanic Caribbean women had risks of GDM of 4.3, 5.9, and 6.9%, respectively.29 Within these groups, subgroup differences were also observed, with the risk for Sub-Saharan African women ranging from 4.1 to 6.9% depending on their country of birth, although samples sizes in these groups became quite small. None of the other studies reported provided subgroup information on women in the group that were described as Black or African American.
4.3.3 Demographic Changes that may Affect Trends in GDM
Shifts in the demographics of the population giving birth can increase the overall prevalence of GDM in the population. Two such trends in the USA are the increasing maternal age at birth and the increase in proportion of births to women from minority populations, both of which are associated with the risk of GDM. The birth rate for women aged 35–39 years has increased each year since 1978 (19.0), rising by almost 50% since 1990 (31.7) to 47.3 births per 1,000 women in 2006. The birth rate for women age 40–44 years (9.4) increased each year since 2000, and has more than doubled since 1981 (3.8) while the birth rate for women aged 45–49 years increased to 0.6 births per 1,000 women in 2006; this rate more than doubled between 1990 (0.2) and 2000 (0.5), but was stable until 2005.50 The racial/ethnic composition of the population giving birth in the USA has also changed over time. In 2006, 54.1% of the 4,265,555 women giving birth were non-Hispanic white, compared to 64.7% of the 3,903,012 women giving birth in 1989, a decrease of about 20% in the proportion of women giving birth who were non-Hispanic white. In 2006, 24% of the women giving birth in the USA were Hispanic, while about 5.6% were Asian or Pacific Islander.49 For women in age categories 30–34, 35–39, and 40–44 years, the age-specific birth rates are consistently higher for Hispanic women and Asian or Pacific Islander women than non-Hispanic white women.33 The convergence of these two demographic trends may also contribute to the increasing prevalence of GDM over time.
4.4 Conclusions and Future Research
The increasing prevalence of GDM signals an impending surge in the number of women that will be affected by diabetes in the coming years. Both women who develop GDM during their pregnancies and their offspring who are exposed to the environmental influence of GDM in utero are at increased risk of developing diabetes in the future. In order to better characterize these populations and to evaluate future trends in GDM and PDM, the methodological challenges described in this chapter must be addressed. A more uniform case definition for GDM, better quality data to identify women as having GDM or PDM in large population-based samples, information on maternal height as well as maternal weight both prepregnancy and at the time of delivery and information to characterize their race/ethnicity in more precise and self-defined categories are needed.
References
1.
Ferrara A. Increasing prevalence of gestational diabetes mellitus: a public health perspective. Diabetes Care. 2007;30(suppl 2):S141-S146.PubMedCrossRef
2.
Kim C, Newton KM, Knopp RH. Gestational diabetes and the incidence of type 2 diabetes: a systematic review. Diabetes Care. 2002;25:1862-1868.PubMedCrossRef
3.
Kim C, Berger DK, Chamany S. Recurrence of gestational diabetes mellitus: a systematic review. Diabetes Care. 2007;30:1314-1319.PubMedCrossRef
4.
Hillier TA, Pedula KL, Schmidt MM, Mullen JA, Charles MA, Pettitt DJ. Childhood obesity and metabolic imprinting: the ongoing effects of maternal hyperglycemia. Diabetes Care. 2007;30:2287-2292.PubMedCrossRef
5.
Dabelea D, Mayer-Davis EJ, Lamichhane AP, et al. Association of intrauterine exposure to maternal diabetes and obesity with type 2 diabetes in youth: the SEARCH Case-Control Study. Diabetes Care. 2008;31:1422-1426.PubMedCrossRef
6.
Metzger BE, Buchanan TA, Coustan DR, et al. Summary and recommendations of the Fifth International Workshop-Conference on Gestational Diabetes Mellitus. Diabetes Care. 2007;30(suppl 2):S251-S260.PubMedCrossRef
7.
Metzger BE, Coustan DR. Summary and recommendations of the Fourth International Workshop-Conference on Gestational Diabetes Mellitus. The Organizing Committee. Diabetes Care. 1998;21(suppl 2):B161-B167.PubMed
8.
Almario CV, Ecker T, Moroz LA, Bucovetsky L, Berghella V, Baxter JK. Obstetricians seldom provide postpartum diabetes screening for women with gestational diabetes. Am J Obstet Gynecol. 2008;198(528):e1-e5.PubMed
9.
Dietz PM, Vesco KK, Callaghan WM, et al. Postpartum screening for diabetes after a gestational diabetes mellitus-affected pregnancy. Obstet Gynecol. 2008;112:868-874.PubMedCrossRef
10.
Ferrara A, Peng T, Kim C. Trends in postpartum diabetes screening and subsequent diabetes and impaired fasting glucose among women with histories of gestational diabetes mellitus. A report from the Translating Research Into Action for Diabetes (TRIAD) Study. Diabetes Care. 2009;32:269-274.PubMedCrossRef
11.
Hunt KJ, Conway DL. Who returns for postpartum glucose screening following gestational diabetes mellitus? Am J Obstet Gynecol. 2008;198:404-406.PubMedCentralPubMedCrossRef
12.
Russell MA, Phipps MG, Olson CL, Welch HG, Carpenter MW. Rates of postpartum glucose testing after gestational diabetes mellitus. Obstet Gynecol. 2006;108:1456-1462.PubMedCrossRef
13.
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;106:1297-1303.PubMedCrossRef
14.
Lawrence JM, Hsu JW, Chen W, Black MH, Sacks DA. Prevalence and timing of postpartum glucose testing and sustained glucose dysregulation after gestational diabetes mellitus. Diabetes Care. 2010;33:569-576.PubMedCrossRef
15.
Kitzmiller JL, Dang-Kilduff L, Taslimi MM. Gestational diabetes after delivery. Short-term management and long-term risks. Diabetes Care. 2007;30(suppl 2):S225-S235.PubMedCrossRef
16.
Costa A, Carmona F, Martinez-Roman S, Quinto L, Levy I, Conget I. Post-partum reclassification of glucose tolerance in women previously diagnosed with gestational diabetes mellitus. Diabet Med. 2000;17:595-598.PubMedCrossRef
17.
Jang HC, Yim CH, Han KO, et al. Gestational diabetes mellitus in Korea: prevalence and prediction of glucose intolerance at early postpartum. Diabetes Res Clin Pract. 2003;61:117-124.PubMedCrossRef
18.
Alberti KG, Zimmet PZ. Definition, diagnosis and classification of diabetes mellitus and its complications. Part 1: diagnosis and classification of diabetes mellitus provisional report of a WHO consultation. Diabet Med. 1998;15:539-553.PubMedCrossRef
19.
Anonymous. Classification and diagnosis of diabetes mellitus and other categories of glucose intolerance. National Diabetes Data Group. Diabetes. 1979;28:1039-1057.CrossRef
20.
Anonymous. Gestational diabetes mellitus. Diabetes Care. 2004;27(suppl 1):S88-S90.
21.
Meltzer S, Leiter L, Daneman D, et al. 1998 clinical practice guidelines for the management of diabetes in Canada. Canadian Diabetes Association. CMAJ. 1998;159(suppl 8):S1-S29.PubMed
22.
Brown CJ, Dawson A, Dodds R, et al. Report of the Pregnancy and Neonatal Care Group. Diabet Med. 1996;13:S43-S53.PubMedCrossRef
23.
Hoffman L, Nolan C, Wilson JD, Oats JJ, Simmons D. Gestational diabetes mellitus–management guidelines. The Australasian Diabetes in Pregnancy Society. Med J Aust. 1998;169:93-97.PubMed
24.
Anonymous. The hyperglycemia and adverse pregnancy outcome (HAPO) study. Int J Gynaecol Obstet. 2002;78:69-77.CrossRef
25.
Trujillo AL, Jovanovic L. Waiting for HAPO. Diabetes Metab Res Rev. 2008;24(suppl 2):S1-S2.PubMedCrossRef
26.
Ferrara A, Kahn HS, Quesenberry CP, Riley C, Hedderson MM. An increase in the incidence of gestational diabetes mellitus: Northern California, 1991-2000. Obstet Gynecol. 2004;103:526-533.PubMedCrossRef
27.
Dabelea D, Snell-Bergeon JK, Hartsfield CL, Bischoff KJ, Hamman RF, McDuffie RS. Increasing prevalence of gestational diabetes mellitus (GDM) over time and by birth cohort: Kaiser Permanente of Colorado GDM Screening Program. Diabetes Care. 2005;28:579-584.PubMedCrossRef
28.
Lawrence JM, Contreras R, Chen W, Sacks DA. Trends in the prevalence of preexisting diabetes and gestational diabetes mellitus among a racially/ethnically diverse population of pregnant women, 1999-2005. Diabetes Care. 2008;31:899-904.PubMedCrossRef
29.
Savitz DA, Janevic TM, Engel SM, Kaufman JS, Herring AH. Ethnicity and gestational diabetes in New York City, 1995-2003. BJOG. 2008;115:969-978.PubMedCrossRef
30.
Getahun D, Nath C, Ananth CV, Chavez MR, Smulian JC. Gestational diabetes in the United States: temporal trends 1989 through 2004. Am J Obstet Gynecol. 2008;198(525):e1-e5.PubMed
31.
Baraban E, McCoy L, Simon P. Increasing prevalence of gestational diabetes and pregnancy-related hypertension in Los Angeles County, California, 1991-2003. Prev Chronic Dis. 2008; 5:A77.PubMedCentralPubMed
32.
Devlin HM, Desai J, Holzman GS, Gilbertson DT. Trends and disparities among diabetes-complicated births in Minnesota, 1993-2003. Am J Public Health. 2008;98:59-62.PubMedCrossRef
33.
Martin JA, Kung HC, Mathews TJ, et al. Annual summary of vital statistics: 2006. Pediatrics. 2008;121:788-801.PubMedCrossRef
34.
Feig DS, Razzaq A, Sykora K, Hux JE, Anderson GM. Trends in deliveries, prenatal care, and obstetrical complications in women with pregestational diabetes: a population-based study in Ontario, Canada, 1996-2001. Diabetes Care. 2006;29:232-235.PubMedCrossRef
35.
Devlin HM, Desai J, Walaszek A. Reviewing performance of birth certificate and hospital discharge data to identify births complicated by maternal diabetes. Matern Child Health J. 2009;13(5):660-666. epub ahead of print.PubMedCrossRef
36.
Bell JC, Ford JB, Cameron CA, Roberts CL. The accuracy of population health data for monitoring trends and outcomes among women with diabetes in pregnancy. Diabetes Res Clin Pract. 2008;81:105-109.PubMedCrossRef
37.
Feig DS, Zinman B, Wang X, Hux JE. Risk of development of diabetes mellitus after diagnosis of gestational diabetes. CMAJ. 2008;179:229-234.PubMedCentralPubMedCrossRef
38.
Montana Department of Public Health and Human Services Chronic Disease Prevention and Health Promotion Program (2009) Trends in diabetes in pregnancy among American Indian and White mothers in Montana 1989-2003: An update. 1-7
39.
Thorpe LE, Berger D, Ellis JA, et al. Trends and racial/ethnic disparities in gestational diabetes among pregnant women in New York City, 1990-2001. Am J Public Health. 2005;95:1536-1539.PubMedCrossRef
40.
Aljohani N, Rempel BM, Ludwig S, et al. Gestational diabetes in Manitoba during a twenty-year period. Clin Invest Med. 2008;31:E131-E137.PubMed
41.
Godwin M, Muirhead M, Huynh J, Helt B, Grimmer J. Prevalence of gestational diabetes mellitus among Swampy Cree women in Moose Factory, James Bay. CMAJ. 1999;160:1299-1302.PubMedCentralPubMed
42.
Shand AW, Bell JC, McElduff A, Morris J, Roberts CL. Outcomes of pregnancies in women with pre-gestational diabetes mellitus and gestational diabetes mellitus; a population-based study in New South Wales, Australia, 1998-2002. Diabet Med. 2008;25:708-715.PubMedCrossRef
43.
Anna V, van der Ploeg HP, Cheung NW, Huxley RR, Bauman AE. Sociodemographic correlates of the increasing trend in prevalence of gestational diabetes mellitus in a large population of women between 1995 and 2005. Diabetes Care. 2008;31:2288-2293.PubMedCrossRef
44.
Chu SY, Callaghan WM, Kim SY, et al. Maternal obesity and risk of gestational diabetes mellitus. Diabetes Care. 2007;30:2070-2076.PubMedCrossRef
45.
Chu SY, Kim SY, Bish CL. Prepregnancy Obesity Prevalence in the United States, 2004-2005. Am J Obstet Gynecol. 2008;200(271):e1-e7.
46.
Ishak M, Petocz P. Gestational diabetes among Aboriginal Australians: prevalence, time trend, and comparisons with non-Aboriginal Australians. Ethn Dis. 2003;13:55-60.PubMed
47.
Rodrigues S, Robinson E, Gray-Donald K. Prevalence of gestational diabetes mellitus among James Bay Cree women in northern Quebec. CMAJ. 1999;160:1293-1297.PubMedCentralPubMed
48.
Lawrence JM, Contrera R. Prevalence of pre-gestational and gestational diabetes mellitus (GDM) among Asian American Women by Ethnic Subgroup, 1999-2005. Diabetologia. 2007;50:S54.
49.
Rao AK, Cheng YW, Caughey AB. Perinatal complications among different Asian-American subgroups. Am J Obstet Gynecol. 2006;194:e39-e41.PubMedCrossRef
50.
Martin JA, Hamilton BE, Sutton PD, et al. Births: final data for 2006. Natl Vital Stat Rep. 2009;57:1-104.