Maternal Hypoglycemia and Placental Pathology

There have been several studies linking hyperglycemia during pregnancy to adverse perinatal outcomes such as macrosomia, fetal cardiac defects, and shoulder dystocia. However, there have only been a few studies addressing the effects of hypoglycemia on perinatal outcomes. Several of these studies have shown an increased risk of intrauterine growth restriction (IUGR) with maternal hypoglycemia particularly during the glucose tolerance test (GTT).3-6 However, many of these studies have not taken into account comorbid factors or have conflicting findings.3-6 IUGR is defined as infant birth weights below the tenth percentile when the baby is otherwise genetically normal. The purpose of this retrospective pilot study is to assess the association of maternal hypoglycemia during the one hour glucose tolerance test with decreased placental weight and identifiable placental pathology. We hypothesize that a decreased one hour glucose tolerance test is a risk factor for decreased placental weight, an increased fetal to placental weight ratio, and other identifiable placental pathology.

Abell et al. found that maternal hypoglycemia during the one hour GTT was associated with decreased birth weights and intrauterine growth restriction (IUGR). However, this study made no comment on the effect of confounding factors such as tobacco use, history of gestational diabetes, or maternal BMI. Both Sokol et al and Langer et al revealed a link between decreased maternal response during the third trimester GTT and IUGR, although in the Sokol study, each pregnancy had one known risk factor for gestational diabetes. Conversely, several more recent papers have reported that maternal hypoglycemia is not predictive of IUGR or poor perinatal outcomes. -8 Consequently, the effect maternal hypoglycemia during the oral GTT is unclear. The placenta is viewed as a reflection of the intrauterine environment during pregnancy. Therefore, we wish to examine hypoglycemia during the oral one hour GTT as a risk factor for decreased placental weight, increased fetal weight to placental weight ratio, and an increased incidence of abnormal placental pathology.

A normal term placenta weighs approximately 450 grams, which is on average, one sixth of the fetal weight.9 Increased maternal substrate availability of vital fetal nutrients, such as glucose, has been known to result in large for gestational age infants, such as occurs with gestational diabetics.10 This may also be reflected in the increased placental weights linked with large for gestational age infants.11 Conversely, IUGR and small for gestational age infants may have decreased placental weights, possibly due to decreased substrate availability secondary to maternal hypoglycemia. 12 Increased fetal to placental weight ratios of 1.5 or greater above the average value for a particular gestational age are also indicative of an IUGR infant.

Identifying a possible indicator for an increased risk of decreased placental weight, which may in turn indicate an increased risk for IUGR, is vitally important when considering the outcomes of IUGR infants. These infants are at great risk for hypoglycemia, sepsis, seizure, stillbirth, respiratory distress, and meconium aspiration. Long term outcomes include poor school performance, hyperactivity, hypertension, and cardiovascular disease.2 Study Population Inclusion/Exclusion Criteria Our study population will consist of women younger than 35, who delivered a term infant either by vaginal delivery or cesarean section at Tulsa Regional Medical Center between July 1, 2005 and July 31, 2007. This time range was selected in order to have an adequate number of recent charts to review. (Women 35 and older are considered "advanced maternal age" and may be at increased risk of placental pathology for that reason, so they are excluded.) We expect to collect approximately 350 usable charts, with perhaps only 35 of those from patients with blood sugars in the "low" or experimental category. A power analysis using these assumptions is provided at the end of this protocol. All placentas are sent to pathology on a routine basis. A term infant is defined as delivery at 37 weeks or greater.

The study population will consist of a typical resident clinic patient population from the OSU OB-Gyn clinic. The ethnicities of the study population include African-American, Caucasian, Hispanic, American Indian, and Asian women. The population included in the study must have a one hour glucose tolerance test result equal to or less than 88 mg/dl. These results are equal to a range in the fifth percentile or less.3 We used 50 grams of oral glucola at 24 weeks or greater for our glucose tolerance test.

We will exclude women with the following comorbid conditions, due to the fact that such conditions manifest placental pathology and may affect placental weight: insufficient gestational dates, multiple gestation, hypothyroidism, hyperthyroidism, heart disease, tobacco use, drug use, history of or current gestational diabetes, history of IUGR, chromosomal abnormalities, congenital malformations, preexisting diabetes, any autoimmune disease, hypertensive disorders, abnormal maternal screen, pulmonary disease, previous gastric bypass, any disease process that results in decreased gastrointestinal absorption, HIV, intrauterine amniotic infection, history of placental abruption, thrombophilia, body mass index (BMI) greater than 40 or BMI less than 19.8.13

Early Termination Criteria N/A

Methods/Procedures Diagnosis

The diagnosis of maternal hypoglycemia after a one hour glucola screen will be defined as a one hour plasma glucose measurement of equal to or less than 88 mg/dl. The glucose screen will be performed at 24 weeks gestation or greater, as calculated by last menstrual period and ultrasound. The one hour GTT results will be obtained from the patient's chart. A decreased placental weight will be defined as a weight equal to or below one standard deviation of the mean placental weight according to gestational age (Table 1). The control group of this population will have a one hour GTT result greater than 88mg/dl. In this study, we will assess placental weight, birth weight, the placenta to birth weight ratio, and placental pathology as described in Table 2. All placental findings were assessed by a single pathologist blinded to the patient's one hour GTT results.

Data Collection The collaborating investigators and research assistants will collect age, height, weight at initial visit, date of delivery, infant birth weight, one hour GTT result, smoking and drug use status, placental weight and pathology results, and any of the above mentioned exclusion criteria only from any patient chart. (Height and weight are needed to calculate BMI.) Research assistants will be utilized to aid in the collection of this data from clinic and hospital records. These assistants will be OSU medical students employed by the work study program and will have basic human research and HIPAA training.

Analysis The two populations will be compared for various response variables. Whenever the response variable is meaningfully numeric and continuous in nature, two population t-tests will be used to compare the means of the two groups. When the response is categorical in nature, contingency tables will be created and Chi Squares tests utilized to compare the proportion of responses for the two groups. All analyses will be conducted with the use of PC SAS Version 9 (SAS Institute, Cary, NC).

Confidentiality The records reviewed will be kept in the Houston Park medical records department or Tulsa Regional Medical Center health information management department. Both these facilities are locked or under security supervision at all times. All study records will be de-identified by conversion to numbers before review, analysis, or interpretation. J. Martin Beal, D.O., Sarah McCoy, Ph.D., and Elsa Vadakekut, D.O. will have access to these numbers which will be kept locked in the OSU Ob/Gyn office, which is separate from both medical records departments. All documents linking patient names to charts will be destroyed after the data collection is complete.