LESS FREQUENT THAN IN NON-DIABETIC PREGNANCIES MICHAEL E. FOLEY, LESLIE DALY, MARY COFFEY, RICHARD FIRTH, BRENDAN KINSLEY, COLM O’HERLIHY, University College Dublin, Obstetrics and Gynaecology, Dublin, Ireland, University College Dublin, Public Health Medicine & Epidemiology, Dublin, Ireland, National Maternity Hospital, Dublin, Ireland, Ireland OBJECTIVE: To determine the incidence and rate of recurrence of fetal macrosomia (>4500 g) at term (>37 weeks) in singleton gestational diabetic (GDM) and non-diabetic pregnancies. STUDY DESIGN: GDM was diagnosed by selective screening (50 g glucose challenge), indicated by risk factors and performed at 29-32 weeks gestation; positive cases (>8.3 mmol/L) underwent a 100-g glucose tolerance test (considered positive if at least two abnormal values). GDM was audited prospectively (1998 to 2002) and the incidence of fetal macrosomia and the rate of its recurrence were compared with non-diabetic pregnancies. RESULTS: The institutional incidence of GDM was 1% (388/37,406); 3.3% of the antenatal population were screened. Among 388 GDM patients (multiparas 77%), the incidence of macrosomia at 7.9% (31/388) was significantly greater than among non-diabetic pregnancies (3.7%, 535/14,461; P ! .0001). Mean gestational age of GDM pregnancies at delivery (39.5 wk) was significantly less than among non-diabetics (40.6 wk; P ! .001). Of 535 non-diabetic women with a macrosomic infant, 34% (182/535) had undergone negative screening for GDM. The recurrence rate of macrosomia in GDM at 26%( 13/50), was half that in non-diabetics (59%, 166 /279; P ! .001). Among multiparas whose previous infant was not macrosomic, significantly more GDM mothers delivered macrosomic infants than non-diabetics (5.3% vs. 2.6%; P = .003). Best predicators of macrosomia in GDM pregnancies were previous macrosomic infant (29%), maternal age >40 years (17%), and weight >100 kg (16%). Among insulin-dependent diabetic mothers (>37 weeks), fetal macrosomia occurred in 13% (21/161) with a recurrence rate of 47% ( 8/17). CONCLUSION: Rate of recurrence of macrosomia in GDM is half that in nondiabetic pregnancies, suggesting that dietary intervention can influence birthweight outcome. Dietary modification might be considered in all women with a history of macrosomia. 159 SCREENING FOR GESTATIONAL DIABETES: DIFFERENT CUT-OFFS FOR DIFFERENT ETHNICITIES? TANIA ESAKOFF, YVONNE CHENG, AARON B. CAUGHEY, University of California, San Francisco, Department of Obstetrics, Gynecology and Reproductive Sciences, San Francisco, California OBJECTIVE: To examine whether screening guidelines for gestational diabetes should be modified based on ethnicity. STUDY DESIGN: This is a retrospective cohort study of 14,565 pregnancies screened for gestational diabetes between January 1988 and December 2001. Values of the 50-g glucose-loading test (GLT) were examined at 5-point increments from 130 mg/dL to 150 mg/dL. The sensitivity and false positive rates (FP) of various GLT cut-offs were compared among Caucasians, African Americans, Hispanics and Asians. Outcomes were compared using the c test. RESULTS: The overall prevalence of gestational diabetes in the study population was 6.3%. The prevalence of gestational diabetes when stratified by ethnicity was 4.1% in Caucasians, 4.3% in African Americans, 7.0% in Hispanics, and 9.7% in Asians. Sensitivities and false positive rates at GLT cutoffs ranging from 130 to 150 were stratified by ethnicity below (Table). CONCLUSION: To maximize the sensitivity and to minimize the false positive rate of the GLT, it may be reasonable to consider varying the cut-off based on ethnicity. For example, if the goal is to achieve maximum sensitivity but achieve approximately a 10% false-positive rate, we would recommend 135 mg/dL for African Americans, 140 mg/dL for Caucasians and Hispanics, and 145 mg/dL for Asians. Alternatively, if the goal is to achieve 95% sensitivity, the screening threshold would need to be 135 mg/dL for African Americans and 130 mg/dL for the other ethnicities.