Hyperglycemia and adverse pregnancy outcome study (HAPO)

Jun 23 2007

Gestational diabetes may place a much higher percentage of pregnant women -- and their unborn babies -- at risk for adverse outcomes than previously believed, according to a major international study presented at the American Diabetes Association's 67th Annual Scientific Sessions.

"We found that the risk of having a large baby, a first-time Cesarean delivery, low blood glucose levels in the newborn requiring treatment, and high blood insulin levels in the baby that may signal problems ahead, all increased as the mother's blood glucose level during pregnancy increased," said Boyd E. Metzger, MD, Professor of Medicine, Division of Endocrinology, Northwestern University Feinberg School of Medicine, Chicago, and Principal Investigator of the study, in a recent interview. "These relationships were continuous over the entire range of blood glucose levels found in over 23,000 pregnancies, even in ranges previously considered to be within the normal range for pregnant women." Nonetheless, the higher the mother's blood glucose, the higher the risk that these problems will occur.

The Hyperglycemia and Adverse Pregnancy Outcome (HAPO) Study was a seven- year international study that recruited approximately 25,000 pregnant women at 15 centers in 9 countries to achieve a major advance in knowledge on levels of blood glucose during pregnancy that place the mother, fetus, and newborn at increased risk for adverse outcomes. Ultimately, 23,325 women completed the study. HAPO was a basic epidemiologic investigation designed to clarify unanswered questions on the association of various levels of glucose intolerance during the third trimester of pregnancy and risk of adverse outcomes.

Nearly 21 million Americans have diabetes, a group of serious diseases characterized by high blood glucose levels that result from defects in the body's ability to produce and/or use insulin. Diabetes can lead to severely debilitating or fatal complications, such as heart disease, blindness, kidney disease, and amputations. Diabetes is the fifth leading cause of death by disease in the U.S.

Pregnant women who have never had diabetes before but who have high blood glucose levels during pregnancy are said to have gestational diabetes (GDM). GDM affects about 4% of all pregnant women -- about 135,000 cases of gestational diabetes in the United States each year. GDM usually goes away after pregnancy, but chances are 2 in 3 that this condition will return in future pregnancies. In some women, however, pregnancy uncovers type 1 or type 2 diabetes. Obstetricians normally test women for GDM around the 28th week of pregnancy. Treatment for GDM involves diet and physical activity, and may also include daily blood glucose testing and insulin injections.

GDM is similar to type 2 in that it involves insulin resistance (an inability of the body to use its insulin properly) and an inability to make sufficient insulin -- so blood glucose levels rise. If left untreated or poorly controlled, GDM pours excess glucose across the placenta to the fetus. This causes the fetal pancreas to make extra insulin to get rid of the blood glucose. Since the fetus is getting more energy than it needs, the extra energy is stored as fat, which can lead to macrosomia -- a large baby. Babies with macrosomia face health problems of their own, including damage to their shoulders during birth. Because of the extra insulin made by the pancreas, newborns may have very low blood glucose levels at birth and may also be at higher risk for breathing problems. Babies with excess insulin may be at risk for obesity in childhood and, in adulthood, at risk for type 2.

For the last 40 years, the diagnosis of GDM has been based on criteria that predict the risk of the mother developing diabetes in the future. But GDM also carries a risk for the baby. However, the point at which maternal blood glucose elevation carries risk for the fetus has been unknown, in part due to such confounding factors as hypertension, overweight, and older age in the mother, which also may contribute to risk. Consequently, some experts have long said that the diagnosis of GDM is not made often enough, while others believe that mild elevations of blood glucose have no adverse effects. Further, physicians have not had guidelines to use to make a diagnosis of GDM related to fetal outcome.

At about 28 weeks of gestation (range 24 to 32), three blood samples were taken from each pregnant woman: a morning fasting blood glucose (FPG); an oral glucose tolerance test (OGTT) involving one sample taken one hour after drinking 75 grams of glucose; and another sample taken a second hour later. Medical caregivers were "blinded" to the test results, except those that exceeded pre-defined cutoff values requiring treatment, in which case -- for safety and ethical reasons -- the woman was removed from the study and treated. Otherwise, pregnancies proceeded and then outcomes were observed and tabulated.

"We found major independent effects of the mother's blood glucose level on each of the outcomes -- the size of the baby, the need for a first Cesarean delivery, low blood glucose requiring treatment, and high insulin levels in the newborn," said Metzger.

The definition of a "large" baby differed from center to center, reflecting the local population, and was defined as being in the largest 10% of the population.

"Low blood glucose levels in a newborn that require treatment are thought to be a consequence of exposure to higher than normal blood glucose levels in the mother, leading to a more difficult transition from the womb to surviving independently -- a classical medical problem in newborns of mothers with known diabetes," he explained.

In the HAPO study, insulin levels in newborns were measured from a blood sample taken from the umbilical cord at delivery. "While high insulin levels are not a problem per se, this indicator is considered a primary outcome because of a longstanding hypothesis and much evidence that many of the problems that these babies develop arise from high insulin levels, and our goal is to document that their presence and such problems are related to these levels," said Metzger.

"It is probable that the level of maternal blood glucose at which a diagnosis of gestational diabetes is made will soon be lowered based on the findings of this study," he said.

Nonetheless, Metzger acknowledged that, although their objective was to learn where elevated glucose clinically becomes a problem, they still cannot immediately say exactly where along the range of the blood glucose spectrum the effects of higher levels are clinically important. "We found that some problems occurred even in ranges previously considered within the normal range for pregnant women," he said.

For example, using just one of the three blood tests taken -- the fasting blood glucose -- results ranged from 75 to 105 mg/dl, and within that range 95 to 100 mg/dl is currently considered the upper range of normal in a pregnant woman. However, the chances of having a big baby increased four to six times over that spread from the lowest to the highest blood glucose results seen in the study. Similarly, the chances of the baby having a high insulin level at delivery increased by as much as 10 times over the range of low to high blood glucose levels in the mother.

"Because these relationships are continuous over the entire range of blood glucose levels, it's not immediately obvious where we should call the value abnormal -- at what level it is important to intervene to normalize blood glucose," he explained. "That cannot immediately be decided from looking at the data." Further, none of the blood glucose measurements used could be identified as best because all provided different but complementary information.

"We are not making diagnostic recommendations," he emphasized. "Those decisions need a consensus translation of a complex set of results and input from everyone interested in the impact, such as researchers, clinicians who care for patients -- including obstetricians, diabetologists, family physicians, pediatricians, and nutritionists -- as well as patients, patient advocates, and third-party payers." A conference to translate the results into new clinical criteria is in the planning stages.

Other presenters at the symposium were: Lynn P Lowe, PhD, Research Assistant Professor, Department of Preventive Medicine, Northwestern University and Project Manager of the study, who spoke on the design and implementation of HAPO; Moshe Hod, MD, Clinical Professor of Obstetrics and Gynecology, Tel Aviv University, who spoke on the consequences of fetal hyperinsulinism; Bengt Persson, MD, PhD, Associate Professor of Pediatrics, Karolinska Institute, Stockholm, who spoke on neonatal morbidity; and Jeremy N Oats, MB BS, DM, FRANZCOG, Adjunct Professor, School of Public Health, La Trobe University, Melbourne, who spoke on maternal outcome.

The study was supported by grants from the National Institutes of Health and the American Diabetes Association.

The American Diabetes Association is the nation's leading voluntary health organization supporting diabetes research, information and advocacy. Founded in 1940, the Association has offices in every region of the country, providing services to hundreds of communities. For more information, please call the American Diabetes Association at 1-800-DIABETES (1-800-342-2383) or visit http://www.diabetes.org/. Information from both these sources is available in English and Spanish.