Objective: To test the hypothesis that progression of diabetic retinopathy in pregnancy is associated with reduced fetal growth and related neonatal morbidity.
Methods: Women with type 1 diabetes (n = 205) were enrolled before 14 weeks' gestation in a prospective study of diabetes in pregnancy and treated with intensive insulin therapy. They had serial ophthalmologic evaluations before 20 weeks' gestation and in late gestation or postpartum. Subjects were divided into two groups based on whether retinopathy progressed (progression group) or remained unchanged (no progression group).
Results: Retinopathy progressed in 59 of 205 women (29%) and was associated with advanced White classification (P = .001): three (5%) were class B, 14 (23%) class C, 24 (41%) class D, and 18 (30%) class F-RF. Reduced fetal growth was associated with progression of retinopathy. Mean birth weight was lower (P = .02), and more infants were small for gestational age (P = .02) and had low birth weights (P = .02) in the progression group. More large-for-gestational-age infants were noted in the no-progression group (P = .04). Birth weight percentile distributions showed a shift of the curve to the left in the progression group (P = .03). There were no differences in gestational age at delivery, macrosomia, preterm delivery, respiratory distress syndrome, neonatal hypoglycemia, or neonatal death. Small for gestational age was associated with chronic hypertension (odds ratio [OR] 6.4; 95% confidence interval [CI] 1.5, 27.9) and retinopathy progression (OR 4.7; 95% CI 1.2, 23.8).
Conclusion: Development and progression of diabetic retinopathy during pregnancy were associated with reduced fetal growth manifested as increased rate of small-for-gestational-age and low-birth-weight infants.
Diabetic retinopathy is a progressive microvascular complication of diabetes with subsequent threat of blindness. Appearance and severity of diabetic retinopathy are related to poor glycemic control, hypertension, and rapid institution of glucose normalization.1–4 In pregnancy, progression is associated with severity of prepregnancy retinopathy at rates of 17%, 36%, and 47%, based on whether baseline examination shows a normal eye examination, background retinopathy, or proliferative retinopathy, respectively.5,6 As when nonpregnant, retinopathy progression in pregnancy is more likely in women with poor glycemic control, rapid institution of strict glycemic control, or hypertensive disease.7,8
Progression of retinopathy (including development) might indicate general worsening of diabetic microvascular disease, and vascular changes are not specific to the retina, so theoretically the process that leads to retinopathy progression also might lead to decreased uteroplacental perfusion, resulting in reduced fetal growth. In a study reported only in abstract form (Kovilam O, Rosenn O, Miodovnik M, Khoury J, Kranias G, Lipman M, et al. Is proliferative retinopathy a risk factor for adverse pregnancy outcome in women with IDDM? [Abstract] J Soc Gynecol Invest 1997;4:152A), we showed that without diabetic nephropathy, retinopathy was not a risk factor for adverse pregnancy outcome; however, there appeared to be reduced fetal growth. To investigate the potential association between reduced fetal growth and diabetic retinopathy, we specifically compared pregnant women with diabetes and nonprogressing (stable) retinopathy with pregnant women with diabetes who developed or had progressive (unstable) retinopathy during pregnancy. We tested the hypothesis that progression of diabetic retinopathy in pregnancy is associated with reduced fetal growth and related neonatal morbidity.
Progression of diabetic retinopathy during pregnancy was associated with reduced fetal growth.
Department of Obstetrics and Gynecology, University of California-Davis, Davis, California; Stabilimento Ospedaliero Thiene, Thiene, Italy; Department of Obstetrics and Gynecology, St. Luke's- Roosevelt Hospital Center, New York, New York, Division of Biostatistics, and the Department of Pediatrics, University of Cincinnati College of Medicine, Cincinnati, Ohio.
Address reprint requests to: Sherrie Smith McElvy, MD, Department of Obstetrics and Gynecology, Division of Maternal-Fetal Medicine, University of California, Davis, 4860 Y Street, Suite 2500, Sacramento, CA 95817. E-mail: firstname.lastname@example.org
Supported in part by the Diabetes in Pregnancy Program Project HD 11275.
Received August 7, 2000. Received in revised form November 15, 2000. Accepted December 7, 2000.
Materials and Methods
From 1978, pregnant women with type 1 diabetes were enrolled prospectively in a program at the University of Cincinnati. The institutional review board approved the study, and patient informed consent was obtained before enrollment. The current study is a secondary analysis of that prospective interdisciplinary study of diabetes in pregnancy. We included pregnant women who enrolled before 14 weeks' gestation, had type 1 diabetes, White classification of B-RF, and at least two ophthalmologic examinations (one before 20 weeks and one in late pregnancy or up to 12 weeks postpartum).
Treatment of women in our program has been described.9 Each subject was supplied with a blood glucose reflectance meter and instructed to check blood glucose concentrations 4–8 times a day. Long-term glucose control was assessed by measurement of glycohemoglobin A1 concentrations at enrollment and each trimester. Management and goals of glycemic control were fasting and preprandial capillary blood glucose concentration below 100 mg/dL and a 90-minute post prandial capillary blood glucose concentration below 140 mg/dL. Daily insulin dosages were divided into three or four injections of short-acting and intermediate-acting insulin, and adjustments were made as needed. At enrollment, nephropathy was evaluated with 24-hour urine collection for quantitative estimations of proteinuria. Nephropathy was defined by proteinuria at least 500 mg/24-hour urine collection before 16 weeks' gestation.
Women were instructed to have ophthalmologic evaluations by one of two ophthalmologists at enrollment, every trimester, and at 6–12 weeks postpartum. For this analysis, an examination in early pregnancy (before 20 weeks) and during the third trimester or postpartum up to 12 weeks was required. Without a third-trimester evaluation (because of early delivery or noncompliance), a postpartum evaluation that showed progression of retinopathy beyond that of the initial evaluation was considered indicative of progression during pregnancy. The ophthalmologic examination included documentation with color photographs and fluorescein angiography (during nonpregnant state). Retinopathy was graded according to a predefined scale (grades 0–6).10 Progression of retinopathy was defined as its development in one or both eyes de novo, or upgrade from first or second trimester to third trimester or postpartum. Women with grade 6 retinopathy and previous photocoagulation who had changes that required repeat photocoagulation were classified as progressed in their retinopathy. The definitions for chronic hypertension and preeclampsia have been published.11
Neonatal outcome data included a ponderal index defined as the cubed root of body weight times 100 divided by height. To compare birth weights between groups, gestational age at delivery and chronic hypertension status were included in the model and an adjusted birth weights were compared.
Data were analyzed with SAS (SAS Institute, Inc., Cary, NC). χ2 test was used for categorical variables. t test or Wilcoxon rank-sum was used to compare ordinal and continuous variables depending on normality assumption. A logistic regression model was used for multivariable associations with reduced fetal growth. Independent variables significant at P < .15 were considered for initial inclusion in the logistic model. A general linear modeling procedure was used to analyze the effect of retinal change on growth alteration. P < .05 was considered statistically significant.
There were 522 pregnancies in women with type 1 diabetes. One hundred forty women were excluded because the pregnancy did not extend beyond 20 weeks' gestation, 82 were excluded because they did not enroll before 14 weeks' gestation, and 95 were excluded because they did not have at least two ophthalmologic examinations. Two hundred five pregnancies met inclusion criteria, and development or progression of retinopathy occurred in 59 (29%). Advanced White classification was associated with increased progression (P = .001). Rates of progression in women with class B, C, D, and F-RF were 3 (5%), 14 (23%), 24 (41%), and 18 (30%), respectively. The eighteen subjects in the progression group with White classification F-RF included 12 women with nephropathy (alone or with proliferative retinopathy). Among 12 with retinopathy, two had infants that were small for gestational age (SGA) and seven had infants with low birth weight. Maternal characteristics for the two groups are presented in Table 1. Women in the progression group were more likely to be younger at diagnosis of diabetes (P = .01) and have preeclampsia (P = .01). There was no difference between groups in cesarean or preterm birth rates before 34 and 37 weeks' gestation. Entry and first-trimester blood glycohemoglobin A1 concentrations were higher, and magnitude of their decrease from first to second trimester was greater in women with progression of retinopathy (Table 2). A summary of changes in retinopathy in our population is given in Table 3. The median week for the early ophthalmologic examination was the 11th week with an interquartile range of the eighth to 15th weeks. The median week for the late examination was the 32nd week with an interquartile range of the 27th to 36th weeks. The median week for the postpartum eye examination was the sixth week with an interquartile range of 3–8 weeks. There were 70 subjects who did not have postpartum ophthalmologic examinations.
Neonatal outcome data are shown in Table 4. Birth weight (P = .02) and adjusted birth weight (adjusted for gestational age and chronic hypertension, P = .01) were lower in the progression group. Reduced fetal growth (below tenth percentile, P = .02) and low birth weight (below 2500 g, P = .02) occurred at higher rates in the progression group, whereas the rate of large-for-gestational-age infants was significantly lower (P = .04) in that group. The birth weight percentile distributions are illustrated in Figure 1, showing a shift of the curve to the left for the progression group (P = .03). There was no difference between groups with regard to macrosomia (weight above 4000 g) or ponderal index (above 90th percentile). More polycythemia (P = .01) was noted for infants in the progression group. In the no-progression group, there was one stillbirth and one neonatal death, each with congenital malformations (transposition of the great vessels and anencephaly, respectively) and low birth weight.
Nine infants were SGA, six born to women in the progression group. Two of six infants had mothers with retinopathy progression and nephropathy. Twenty-seven infants had low birth weights, 13 in the progression group. Seven of 13 infants had mothers with retinopathy progression and nephropathy.
In the no-progression group, three infants were SGA, two of whose mothers had proliferative retinopathy (grade 6) at their first-trimester ophthalmologic examination and chronic hypertension. The third infant had a congenital malformation and was born to a mother with benign retinopathy (grade 3) and no hypertension. Six infants whose mothers were in the progression group were SGA, and none had congenital malformations. The first-trimester ophthalmologic examinations of those six mothers showed three with proliferative retinopathy (grade 4), two with benign retinopathy (grade 2), and one with no retinopathy. Three of those mothers had chronic hypertension, one with superimposed preeclampsia.
Logistic regression analysis was done to evaluate the independent effect of retinopathy progression on reduced fetal growth (Tables 5 and 6). Independent variables analyzed were progression of retinopathy (change in grade), gestational age at delivery, initial glycohemoglobin A1, change in glycohemoglobin A1 from first to second trimester, chronic hypertension, preeclampsia, and nephropathy. Progression of retinopathy (odds ratio [OR] 4.7; 95% confidence interval [CI] 1.2, 23.8) was significantly and independently associated with reduced fetal growth after accounting for chronic hypertension (OR 6.4; 95% CI 1.5, 27.9).
We evaluated risk of reduced fetal growth in women who developed or had progression of diabetic retinopathy during pregnancy because retinal deterioration might indicate worsening microvascular disease not specific to the retina. We hypothesized that the mechanism leading to progression of retinopathy also might lead to reduced fetal growth by compromising utero-placental blood flow. In a study on adverse pregnancy outcomes in women with type 1 diabetes and proliferative retinopathy at initial ophthalmologic examinations, we found that without nephropathy, proliferative retinopathy was not a risk for adverse pregnancy outcomes. However, there was a 325-g lower overall birth weight and an 18% incidence of low birth weight in the group with retinopathy compared with 8% incidence of low birth weight in the group without retinopathy. Those findings prompted further investigation of the association between fetal birth weight and progression of retinopathy during pregnancy.
In the current study, we showed reduced fetal growth with a twofold increase in low birth weight, fivefold more SGA infants, and fewer large-for-gestational-age infants in the progression group. There also was a 268-g reduction in overall birth weight in the progression group. Not only were the extremes (small and large for gestational age) affected, but the entire distribution of growth (Figure 1) was shifted to the left. Small-for-gestational-age infants are at increased risk of adverse outcomes, and women with progressive retinopathy should have fetal growth monitored closely.
Most reports on diabetic retinopathy in pregnancy did not address neonatal outcomes. Among those that studied neonatal outcomes, most focused on the most advanced stage, proliferative diabetic retinopathy, without defining whether subjects had stable or progressive disease. Reece et al12 reported neonatal outcomes in 20 pregnant women with proliferative retinopathy (at least half progressed during pregnancy): 15% of those pregnancies ended in abortion or stillbirth and the perinatal survival rate was 94%. They concluded that most pregnancies resulted in live births of appropriate-for-gestational-age infants. Nine of 17 (53%) births had no neonatal complications and three (18%) had congenital anomalies. Klein et al13 reported pregnancy outcomes in 179 women with diabetes. Among maternal factors, severity of diabetic retinopathy was the most significant predictor of adverse pregnancy outcome, defined as abortion, perinatal death, and severe congenital malformations. Those studies did not differentiate stable from progressive retinal disease, so we were unable to compare them with our current findings. A weakness in our study was lack of smoking history in the logistic regression model. Data on smoking history were not collected after 1988; however, incidence of smoking in women enrolled before 1988 was 30%.
Logistic regression supported an independent association between progression of retinopathy and reduced fetal growth. We found an association of reduced fetal growth with chronic hypertension and change in blood glycohemoglobin A1 concentration (from first to second trimester). Previous studies8,14,15 showed an association of chronic hypertension and change in blood glycohemoglobin A1 concentration with progression of retinopathy. The association of chronic hypertension with SGA infants is well established16; however, association between change in glycohemoglobin A1 concentration and SGA is not. This investigation found that the mechanism in which rapid normalization of glucose control leads to progression of retinopathy might also lead to reduced fetal growth. The association of retinopathy progression with advanced White classification was consistent with other studies.17
1. Alvarsson ML, Grill VE. Effect of long-term glycemic control on the onset of retinopathy in IDDM subjects. A longitudinal and retrospective study. Diabetes Res 1989;10:75–80.
2. Klein R, Klein BEK, Moss SE, Davis MD, DeMets DL. Glycosylated hemoglobin predicts the incidence and progression of diabetic retinopathy. JAMA 1988;260:2864–71.
3. Brinchmann-Hansen O, Dahl-Jorgensen K, Sandvik L, Hanssen KF. Blood glucose concentrations and progression of diabetic retinopathy: The seven-year results of the Oslo study. BMJ 1992;304:19–22.
4. Dahl-Jorgensen K, Brinchmann-Hansen O, Hanssen KF, Sandvik L, Aagenaes O. Rapid tightening of blood glucose leads to transient deterioration of retinopathy in insulin-dependent diabetes mellitus: The Oslo study. BMJ 1985;290:811–5.
5. Rosenn BM, Miodovnik M. Medical complications of diabetes mellitus in pregnancy. Clin Obstet Gynecol 2000;43:17–31.
6. Star J, Carpenter MW. The effect of pregnancy on the natural history of diabetic retinopathy and nephropathy. Clin Perinatol 1998;25:887–916.
7. Lovestam-Adrian M, Agardh CD, Aberg A, Agardh E. Preeclampsia is a potent risk factor for deterioration of retinopathy during pregnancy in type 1 diabetic patients. Diabetic Med 1997;14:1059–65.
8. Rosenn B, Miodovnik M, Kranias G, Khoury J, Combs CA, Minouni F, et al. Progression of diabetic retinopathy in pregnancy: Association with hypertension in pregnancy. Am J Obstet Gynecol 1992;166:1214–8.
9. McElvy SS, Miodovnik M, Rosenn B, Khoury JC, Siddiqi T, Dignan PSJ, et al. A focused preconceptional and early pregnancy program in women with type 1 diabetes reduces perinatal mortality and malformation rates to general population levels. J Matern Fetal Med 2000;9:14–20.
10. Diabetic Retinopathy Study Research Group. A modification of the Airlie House classification of diabetic retinopathy. Invest Ophthalmol Vis Sci 1981;21:210–26.
11. Combs CA, Rosenn B, Kitzmiller JL, Khoury JC, Wheeler BC, Miodovnik M. Early-pregnancy proteinuria in diabetes related to preeclampsia. Obstet Gynecol 1993;82:802–7.
12. Reece EA, Lockwood CJ, Tuck S, Coulehan J, Homko C, Wiznitzer A, et al. Retinal and pregnancy outcomes in the presence of diabetic proliferative retinopathy. J Reprod Med 1994;39:799–804.
13. Klein BE, Klein R, Meuer SM, Moss SE, Dalton DD. Does the severity of diabetic retinopathy predict pregnancy outcome? J Diabetic Compl 1988;2:179–84.
14. Phelps RL, Sakol P, Metzger BE, Jampol LM, Freinkel N. Changes in diabetic retinopathy during pregnancy. Arch Ophthalmol 1986; 104:1806–10.
15. The Diabetes Control and Complications Trial Research Group. Effect of pregnancy on microvascular complications in the diabetes control and complications trial. Diabetes Care 2000;23:1084–91.
16. Easterling TR, Benedetti TJ, Carlson KC, Brateng DA, Wilson J, Schmucker BS. The effect of maternal hemodynamics on fetal growth in hypertensive pregnancies. Am J Obstet Gynecol 1991; 165:902–6.
17. Serup L. Influence of pregnancy on diabetic retinopathy. Acta Endocrinol 1986;277:122–4.