- Email: [email protected]
The effect of pregnancy on retinal hemodynamics in diabetic versus nondiabetic mothers
The Effect of Pregnancy on Retinal Hemodynamics in Diabetic Versus Nondiabetic Mothers LISA S. SCHOCKET, MD, JUAN E. GRUNWALD, MD, AMY F. TSANG, MA, AND JOAN DUPONT, CRCC
● PURPOSE:
To investigate retinal circulatory changes that occur during the third trimester of pregnancy in diabetic patients and control subjects. ● METHODS: Bidirectional laser Doppler velocimetry and monochromatic fundus photography were used to assess the retinal circulation in seven pregnant diabetic patients and 13 age-matched pregnant control subjects. Retinal venous diameter (D), maximum erythrocyte velocity (Vmax), and retinal volumetric blood flow rate (Q) were measured in one eye of each subject during the third trimester of pregnancy (DPREG, VmaxPREG, and QPREG, respectively). These measurements were repeated during the postpartum period for both diabetic patients (11 ⴞ 7 weeks postpartum) and control subjects (16 ⴞ 6 weeks postpartum; P ⴝ .203; DPOST, VmaxPOST, and QPOST). ● RESULTS: In control subjects, DPREG was significantly reduced by ⴚ4.5% ⴞ 4.4% (mean percent difference ⴞ 1 standard deviation; paired t test, P ⴝ .006) relative to DPOST. In diabetic women, DPREG was also significantly reduced by ⴚ8.1% ⴞ 3.2% compared with DPOST (P ⴝ .001), a change that was significantly larger than that seen in control subjects (unpaired t test; P ⴝ .035). Compared with QPOST, QPREG was reduced by ⴚ7.1% ⴞ 14.2% (P ⴝ .123), in control subjects. In diabetic women, QPREG was significantly decreased by ⴚ18.4% ⴞ 9.3% compared with QPOST (P ⴝ .012). This reduction in QPREG was significantly greater in diabetic patients than in nondiabetic control subjects (unpaired t test, P ⴝ .040). No significant differences between VmaxPREG and VmaxPOST were observed in either diabetic patients (ⴚ3.1% ⴞ Accepted for publication June 8, 1999. From the Department of Ophthalmology, Scheie Eye Institute, School of Medicine, University of Pennsylvania, Philadelphia, Pennsylvania. This study was supported by the Vivian Simkins Lasko Retinal Vascular Research Fund, Research to Prevent Blindness, New York, New York, the Nina C. Mackall Trust, and grants EYO3243 and EY11479 from the National Institutes of Health, Bethesda, Maryland. Reprint requests to Juan E. Grunwald, MD, Scheie Eye Institute, University of Pennsylvania, 51 N 39th St, Philadelphia, PA 19104; fax: (215) 662-8025; e-mail: [email protected] 0002-9394/99/$20.00 PII S0002-9394(99)00234-2
©
1999 BY
12.9%; P ⴝ .400) or control subjects (ⴙ1.9% ⴞ 14.4%; P ⴝ .787). ● CONCLUSIONS: Retinal venous diameter is decreased during the third trimester of pregnancy in both diabetic and nondiabetic mothers. This decrease is significantly larger in diabetic than in nondiabetic mothers. In addition, we observed a reduction in retinal volumetric blood flow in diabetic patients during pregnancy that was significantly larger than that present in nondiabetic women. This fall in retinal volumetric blood flow in diabetic patients may exacerbate retinal ischemia and hypoxia and thus may be associated with the progression of diabetic retinopathy. (Am J Ophthalmol 1999;128: 477– 484. © 1999 by Elsevier Science Inc. All rights reserved.)
D
IABETIC RETINOPATHY, A MICROVASCULAR COM-
plication of diabetes, is the leading cause of blindness between the ages of 24 and 64 in the United States and the United Kingdom.1,2 Investigators have shown abnormalities in retinal vascular diameters and retinal blood flow in both the early3– 4 and late stages of diabetes.5– 8 In addition to blood flow abnormalities, patients with diabetic retinopathy also have an impaired ability to adjust retinal blood flow in the face of metabolic and environmental challenges.9 –13 In the past, pregnant diabetic women with evidence of diabetic retinopathy were often advised to terminate their pregnancies.14 Advances in obstetrics now allow an increasing number of women with diabetes to experience pregnancy, so we need to develop a better understanding of the relationships between pregnancy and the complications of diabetes mellitus. When studying pregnancy, it is important to note the physiologic changes that occur. In addition to the hormonal, metabolic, and immunologic alterations frequently described, significant differences are also present in the cardiovascular and hematologic systems. These changes include increased cardiac output, decreased vascular resistance, lowered arterial blood pressure, and increased
ELSEVIER SCIENCE INC. ALL
RIGHTS RESERVED.
477
resting pulse rate. Blood volume also increases approximately 45% above nonpregnancy levels.15 Several investigators have demonstrated that pregnancy adversely effects diabetic retinopathy.16 –22 Axer-Siegel and associates16 reported that of 22 pregnant, diabetic women with diabetic retinopathy at conception, 77.5% experienced progression of retinopathy and 22.5% developed proliferative diabetic retinopathy. Similarly, Klein and associates17 studied the rate of progression of diabetic retinopathy in 171 pregnant and 298 nonpregnant diabetic women. After accounting for glycemic control, blood pressure, and duration of diabetes, they discovered that current pregnancy was a major risk factor for the progression of retinopathy. The physiologic changes of pregnancy combined with a compromised retinal vascular regulatory capacity may play a role in the observed progression of diabetic retinopathy during pregnancy. Chen and associates23 studied retinal blood flow in pregnant women with and without diabetes and found a 14% to 19% increase in retinal blood flow in diabetic patients whose retinopathy progressed. In this study, we further investigated the changes in retinal blood flow that occur in diabetic patients and nondiabetic control subjects during the third trimester of pregnancy compared with the postpartum period.
TABLE 1. Characteristics of Diabetic Patients and Control Subjects
Characteristics
Age (yrs) 33 ⫾ 5† Duration of diabetes (yrs) 13.8 ⫾ 10.0 34 ⫾ 2 Gestation (wks)‡ 11 ⫾ 7 Postpartum (wks)‡ BPm during pregnancy (mm Hg) 85 ⫾ 11 BPm postpartum 85 ⫾ 8
32 ⫾ 4 NA 34 ⫾ 3 16 ⫾ 6 88 ⫾ 11 86 ⫾ 16
.557 NA .967 .203 .541 .857
superior and inferior retinal veins close to the optic disk. Each measurement site was marked on the fundus photograph. After the bidirectional laser Doppler velocimetry recordings, a fundus camera was used to take fundus photographs in monochromatic light at 570 nm. The diameter of the retinal vessel at the site of the bidirectional laser Doppler velocimetry recordings was assessed from projected photographic negatives. For each measurement site, six photographs were analyzed. Each of the six photographs was measured five times, and the results were averaged to obtain the retinal venous diameter (D). A second trained examiner then determined the maximum erythrocyte velocity (Vmax) and calculated the mean velocity of whole blood (Vmean) according to the equation, Vmean ⫽ C 䡠 Vmax, where C ⫽ 1/1.6.24 An average of Vmean for both the superior and inferior major retinal veins was used to calculate the average retinal volumetric blood flow rate (Q) according to the relation, Q ⫽ Vmean 䡠 D2/4. On completion of bidirectional laser Doppler velocimetry and monochromatic fundus photography determinations, baseline blood pressure and heart rate were immediately obtained. The mean brachial artery blood pressure was estimated with the standard formula, BPm ⫽ BPd ⫹ 1/3 (BPs ⫺ BPd). In each subject, measurements of vessel diameter, blood velocity, and blood flow were performed during the third trimester of pregnancy (DPREG, VmaxPREG, and QPREG). These measurements were repeated during the postpartum period for both diabetic patients (11 ⫾ 7 weeks postpartum) and control subjects (16 ⫾ 6 weeks postpartum; P ⫽ .203; DPOST, VmaxPOST, and QPOST). Determinations of Vmax and D were performed by separate examiners, who were masked in regard to the diagnosis and the timing of the measurements. Third-trimester measurements were compared with those obtained postpartum, and changes noted between
THE STUDY POPULATION INCLUDED SEVEN PREGNANT DIA-
betic women and 13 age-matched, nondiabetic pregnant control subjects recruited as volunteers by Scheie Eye Institute at the University of Pennsylvania. Eligibility criteria for control subjects included the following: current pregnancy, no history of ocular disease, normal ocular examination, and no systemic or topical medications. For the diabetic patients, eligibility criteria included current diagnosis of diabetes requiring insulin and no other systemic or topical medications. All subjects had deep anterior chambers and pupillary dilation greater than 5 mm. Ages of the diabetic women ranged from 28 to 38 years (mean, 33 ⫾ 5 years; ⫾1 SD), and the duration of their diabetes varied from 4 months to 26 years (mean, 14 ⫾ 10 years). Similarly, the ages of the nondiabetic women ranged from 26 to 37 years (mean, 32 ⫾ 4 years). Three of the seven diabetic patients had background diabetic retinopathy and four had normal funduscopic examinations. Our Institutional Review Board had approved the study, and all subjects gave informed consent to participate. Bilateral pupillary dilation was attained with tropicamide 1.0% and phenylephrine hydrochloride 1.5%. A Polaroid color fundus photograph (Polaroid Corporation, Cambridge, Massachusetts) of the optic disk was then obtained to document the measurement sites. By use of bidirectional laser Doppler velocimetry, we measured the centerline erythrocyte velocity (Vmax) in the largest AMERICAN JOURNAL
Control Subjects P Value*
BPm ⫽ mean blood pressure; NA ⫽ not applicable. *Unpaired Student t test. † Mean ⫾ 1 SD. ‡ Number of weeks’ gestation or postpartum at the time measurements were collected.
METHODS
478
Diabetic Patients
OF
OPHTHALMOLOGY
OCTOBER 1999
TABLE 2. Retinal Circulatory Parameters Parameters
D (m) Diabetic patients Control subjects Vmax (cm/s) Diabetic patients Control subjects Q (l/min) Diabetic patients Control subjects
Postpartum
Pregnancy
% Difference
P Value*
P Value†
153 ⫾ 12‡ 152 ⫾ 11
140 ⫾ 8 145 ⫾ 9
⫺8.1 ⫾ 3.2 ⫺4.5 ⫾ 4.4
.001 .006
.035
1.84 ⫾ 0.34 1.83 ⫾ 0.23
1.75 ⫾ 0.19 1.85 ⫾ 0.33
⫺3.1 ⫾ 12.9 ⫹1.9 ⫾ 14.4
.400 .787
.504
13.0 ⫾ 3.8 12.6 ⫾ 2.8
10.3 ⫾ 2.1 11.7 ⫾ 2.7
⫺18.4 ⫾ 9.3 ⫺7.1 ⫾ 14.2
.012 .123
.040
*Comparison of postpartum and pregnancy values in each subject using paired Student t tests. † Comparison of percent differences between diabetic patients and control subjects using unpaired t tests. ‡ Mean ⫾ 1 SD.
FIGURE 1. In each nondiabetic mother, the percent difference in venous diameter during the third trimester of pregnancy compared with the postpartum value. Average venous diameter is significantly reduced by ⴚ4.5% (paired t test, P ⴝ .006).
FIGURE 2. In each diabetic mother, the percent difference in venous diameter during the third trimester of pregnancy compared with the postpartum value. Average venous diameter is significantly reduced by ⴚ8.1% (paired t test, P ⴝ .001).
these two periods were contrasted between the two study populations. Paired and unpaired Student t tests were used for the statistical analysis, and P values of less than .05 were considered statistically significant
the measurements during pregnancy and after delivery was also similar between groups. Table 2 shows the retinal circulatory parameters obtained during postpartum and pregnancy for both diabetic patients and control subjects. Relative to the postpartum values, DPREG in control subjects was significantly reduced by ⫺4.5% ⫾ 4.4% (paired Student t test, P ⫽ .006; Figure 1). Diabetic women also demonstrated a significant decrease in DPREG of ⫺8.1% ⫾ 3.2% (P ⫽ .001; Figure 2) compared with DPOST. The decrease in DPREG in diabetic mothers was significantly greater than that seen in control subjects (unpaired t test, P ⫽ .035; Figure 3).
RESULTS TABLE 1 SUMMARIZES THE CHARACTERISTICS OF DIABETIC
patients and control subjects, demonstrating no significant differences in age or mean blood pressure during pregnancy (BPmPREG) and postpartum (BPmPOST). The timing of VOL. 128, NO. 4
PREGNANCY
AND
RETINAL HEMODYNAMICS
479
significant differences were observed in either QPOST (unpaired t tests, P ⫽ .829) or QPREG (unpaired t tests, P ⫽ .246) when comparing diabetic patients with control subjects. A significant negative correlation was noted between BPmPOST and QPOST in diabetic mothers (R ⫽ ⫺0.790; P ⬍ .05; Figure 8, left) but not in nondiabetic mothers (R ⫽ ⫹0.252; P ⬎ .05; Figure 8, right). No significant correlation was noted between BPmPREG and QPREG in either group.
DISCUSSION DESPITE TIGHT GLYCEMIC CONTROL, DIABETIC RETINOPA-
thy presents a major problem during childbearing years. Although investigators have documented the acceleration of diabetic retinopathy during pregnancy, the pathogenesis remains unclear. We analyzed retinal circulatory parameters in pregnant diabetic patients and nondiabetic pregnant control subjects to further identify any circulatory changes that may be related to the worsening of diabetic retinopathy during pregnancy. Our study shows a vasoconstriction of ⫺4.5% ⫾ 4.4% in DPREG compared with DPOST in control subjects (P ⫽ .006). This decrease in retinal venous diameter may be a manifestation of the retina’s autoregulatory response to the physiologic changes of pregnancy, including increased cardiac output and increased pulse rate. The diabetic mothers demonstrated a vasoconstriction of ⫺8.1% ⫾ 3.2% (P ⫽ .001) in DPREG, which was significantly greater than that observed in control subjects (unpaired; P ⫽ .035). In addition, a significant negative correlation between DPOST and BPmPOST was observed in diabetic patients (R ⫽ ⫺0.786; P ⬍ .05), suggesting that higher blood pressures are associated with smaller vessel diameters. A similar correlation was not observed in control subjects (R ⫽ ⫹0.317; P ⬎ .05), nor was a significant correlation present between BPmPREG and DPREG in either group. Chen and associates23 also studied retinal circulatory parameters in pregnant diabetic and nondiabetic mothers. They found that nondiabetic pregnant women had smaller vessel diameters in each trimester of pregnancy compared with postpartum and concluded that the decreased diameter in control subjects was a response to increased velocity, representing an autoregulatory response. They also noted that during the first trimester, vessel diameters were narrower in the diabetic mothers whose diabetic retinopathy did not progress compared with postpartum (P ⫽ .003). The diameters during the second and third trimesters, however, were not significantly different from the postpartum value (P ⬎ .05) nor were the diameters of patients whose diabetic retinopathy progressed. This supports other studies that have shown a failure of autoregulation in patients with diabetic retinopathy.
FIGURE 3. In nondiabetic and diabetic mothers, the average percentage difference in venous diameter (ⴞ95% confidence intervals) during the third trimester of pregnancy. Compared with postpartum values, there is a ⴚ4.5% average decrease in venous diameter during the third trimester of pregnancy in nondiabetic women (paired t test, P ⴝ .006), and a ⴚ8.1% decrease in diabetic women (P ⴝ .001). The decrease in diabetic mothers is significantly greater than that seen in nondiabetic mothers (unpaired t test, P ⴝ .035).
When comparing diabetic patients and control subjects, no significant differences were observed in DPOST (unpaired t tests, P ⫽ .829) or DPREG (unpaired t tests, P ⫽ .214). In addition, a significant negative correlation between DPOST and BPmPOST was observed in diabetic women (R ⫽ ⫺0.786; P ⬍ .05; Figure 4, left) but not in control subjects (R ⫽ ⫹0.317; P ⬎ .05; Figure 4, right). No significant correlation was noted between BPmPREG and DPREG in either diabetic or nondiabetic mothers. In contrast to the diameter findings, no significant differences between VmaxPOST and VmaxPREG were noted in either diabetic patients (⫺3.1% ⫾ 12.9%; P ⫽ .400) or control subjects (⫹1.9% ⫾ 14.4%; P ⫽ .787). Again, comparisons of diabetic patients and control subjects showed no significant differences between VmaxPOST (unpaired t tests, P ⫽ .934) or VmaxPREG (unpaired t tests, P ⫽ .415). Table 2 also indicates that compared with QPOST, QPREG in control subjects was reduced by ⫺7.1% ⫾ 14.2%, but this difference was not statistically significant (P ⫽ .123; Figure 5). In diabetic women, however, QPREG was significantly decreased by ⫺18.4% ⫾ 9.3% compared with QPOST (P ⫽ .012; Figure 6). Similar to the results found for DPREG, the reduction in QPREG in diabetic patients was significantly larger than that detected in control subjects (unpaired t test, P ⫽ .040; Figure 7). No 480
AMERICAN JOURNAL
OF
OPHTHALMOLOGY
OCTOBER 1999
FIGURE 4. Venous diameter postpartum (DPOST) compared with mean blood pressure postpartum (BPmPOST) in diabetic women (left) and control subjects (right). There is a significant negative correlation between DPOST and BPmPOST in diabetic women (R ⴝ ⴚ0.786; P < .05; left), but not in control subjects (R ⴝ ⴙ0.317; P > .05; right).
FIGURE 5. In comparison with postpartum, percentage difference in average volumetric blood flow during the third trimester of pregnancy in nondiabetic women. Average volumetric blood flow in nondiabetic mothers is reduced by ⴚ7.1% during the third trimester of pregnancy, but this is not statistically significant (paired t test, P ⴝ .123).
Differences between the study patients of Chen and associates23 and our study may explain some of the differences observed in vessel diameter. In the study by VOL. 128, NO. 4
PREGNANCY
AND
FIGURE 6. In comparison with postpartum, percentage difference in average volumetric blood flow during the third trimester of pregnancy in diabetic women. Average volumetric blood flow in diabetic mothers is significantly reduced by ⴚ18.4% during the third trimester of pregnancy (paired t test, P ⴝ .012).
Chen and associates,23 12 of the 22 diabetic patients had background diabetic retinopathy, four had proliferative diabetic retinopathy, and the remaining six had no diabetic retinopathy. Fifty percent of the 22 patients
RETINAL HEMODYNAMICS
481
mainly because of the reduction in vessel diameter. Chen and associates23 also found no significant change in volumetric blood flow in control subjects between any of the trimesters and postpartum, but they found a nonsignificant trend toward higher flow rates during pregnancy (P ⬎ .09). In contrast to our findings in diabetic mothers, Chen and associates23 observed a significant increase in flow during the second (P ⫽ .04) and third (P ⫽ .04) trimesters of pregnancy compared with postpartum. Again, it is possible that the disparity between our results and those of Chen and associates23 may reflect differences in blood flow that may be caused by discrepancies in the severity and rate of progression of diabetic retinopathy in both studies. Hellstedt and associates25 used blue-field entopic simulation to study macular capillary blood flow during pregnancy and postpartum in diabetic patients with minimal diabetic retinopathy. In contrast to Chen and associates,23 who observed increased retinal blood flowduring pregnancy, Hellstedt and associates25 found no increase in macular capillary flow velocity in diabetic mothers during pregnancy compared with postpartum. Several authors have suggested mechanisms for the progression of diabetic retinopathy during pregnancy. Chen and associates23 proposed that this progression may partly result from increased retinal blood flow caused by hyperperfusion. The ability of the retina to autoregulate its blood flow is impaired in diabetic retinopathy,9 –12 and thus in a diabetic patient, the physiologic changes of pregnancy impose an added stress on an already compromised retinal circulation. Phelps and associates26 found a significant positive correlation between the progression of diabetic retinopathy during pregnancy and the degree of improvement in glycemia control. They therefore suggested that this rapid improvement in glycemia control might have a deleterious effect on the retina, causing progression of retinopathy. Our results suggest that in both diabetic and nondiabetic mothers, retinal venous diameter is decreased during the third trimester of pregnancy, with more marked reductions observed in diabetic patients. In addition, we observed a reduction in retinal volumetric blood flow in diabetic patients during pregnancy. The decrease in retinal volumetric blood flow in diabetic patients may exacerbate retinal ischemia and hypoxia and thus may be associated with the progression of diabetic retinopathy. It is also possible that the hemodynamic changes associated with diabetic retinopathy lead to differing responses to the physiologic changes of pregnancy, depending on the severity of diabetic retinopathy. This may explain the discrepancies between the results reported by Chen and associates23 and those observed in this study. Further studies are needed to clarify why diabetic retinopathy tends to progress during pregnancy and whether patients with more severe disease respond differently to the added physiologic stress of pregnancy.
FIGURE 7. Average percentage difference in volumetric blood flow (ⴞ95% confidence intervals) during the third trimester of pregnancy compared with postpartum in nondiabetic and diabetic mothers. Compared with postpartum values, there is an average ⴚ7.1% nonsignificant decrease in volumetric blood flow during the third trimester of pregnancy in nondiabetic women (paired t test, P ⴝ .122) and a significant decrease of ⴚ18.4% in diabetic women (paired t test, P ⴝ .012). The decrease in diabetic mothers is significantly larger than that seen in nondiabetic mothers (unpaired t test, P ⴝ .040).
had progression of diabetic retinopathy. In contrast, our patient population had significantly less disease and more homogeneity in the extent of their diabetic retinopathy, with only three patients with background diabetic retinopathy and the remaining patients with no detectable funduscopic changes. None of the diabetic patients in our study showed any worsening of diabetic retinopathy during pregnancy. Also in contrast to the results reported by Chen and associates,23 we did not detect significant differences in VmaxPREG compared with VmaxPOST in either group. Compared with postpartum values, Chen and associates23 detected significant increases in retinal blood velocity as pregnancy progressed in both diabetic patients and control subjects. Similar to our vessel diameter findings, a nonsignificant ⫺7.1% ⫾ 14.2% decrease in QPREG compared with QPOST (P ⫽ .123) in control subjects was observed. Diabetic mothers exhibited a ⫺18.4% ⫾ 9.3% reduction in QPREG compared with QPOST (P ⫽ .012). This decrease in flow during pregnancy in diabetic mothers was significantly larger than that observed in nondiabetic mothers (unpaired t test, P ⫽ .040). The decrease in QPREG compared with QPOST in both groups was 482
AMERICAN JOURNAL
OF
OPHTHALMOLOGY
OCTOBER 1999
FIGURE 8. Volumetric blood flow postpartum (QPOST) compared with mean blood pressure postpartum (BPmPOST) in diabetic women (left ) and control subjects (right). There is a significant negative correlation between QPOST and BPmPOST in diabetic women (R ⴝ ⴚ0.790; P < .05; left) but not in control subjects (R ⴝ ⴙ0.252; P > .05; right).
REFERENCES 1. Kohner EM, Porta M. Protocols for screening and treatment of diabetic retinopathy in Europe. Eur J Ophthalmol 1991; 1:45–54. 2. Foulds WS, MacCuish A, Barrie T, et al. Diabetic retinopathy in the West of Scotland: its detection, and prevalence, and the cost-effectiveness of a proposed screening programme. Health Bull 1983;41:318 –326. 3. Grunwald JE, DuPont J, Riva CE. Retinal haemodynamics in patients with early diabetes mellitus. Br J Ophthalmol 1996;80:327–331. 4. Bursell S-E, Clermont AC, Kinsley BT, Simonson DC, Aiello LM, Wolpert HA. Retinal blood flow changes in patients with insulin dependent diabetes mellitus and no diabetic retinopathy: a video fluorescein angiography study. Invest Ophthalmol Vis Sci 1996;37:886 – 897. 5. Grunwald JE, Riva CE, Sinclair SH, Brucker AJ, Petrig BL. Laser Doppler velocimetry study of retinal circulation in diabetes mellitus. Arch Ophthalmol 1986;104:991–996. 6. Patel V, Rassam SMB, Newson RSB, Wiek J, Kohner EM. Retinal blood flow in diabetic retinopathy. BMJ 1992;305: 678 – 684. 7. Feke GT, Buzney SM, Ogasawara H, et al. Retinal circulatory abnormalities in type I diabetes. Invest Ophthalmol Vis Sci 1994;35:2968. 8. Grunwald JE, Riva CE, Baine J, Brucker AJ. Total retinal volumetric blood flow rate in diabetic patients with poor diabetic control. Invest Ophthalmol Vis Sci 1992;33:356 – 363. 9. Grunwald JE, Riva CE, Brucker AJ, Sinclair SH, Petrig BL. Altered retinal vascular response to 100% oxygen breathing in diabetes mellitus. Ophthalmol 1984;91: 1447–1452.
VOL. 128, NO. 4
PREGNANCY
AND
10. Rassam SMB, Patel V, Kohner EM. The effect of experimental hypertension on retinal autoregulation in humans: a mechanism for the progression of diabetic retinopathy. Exp Physiol 1995;80:53– 68. 11. Ernest JT, Goldstick TK, Engerman RL. Hyperglycemia impairs retinal oxygen autoregulation in normal and diabetic dogs. Invest Ophthalmol Vis Sci 1983;24:985–989. 12. Grunwald JE, Riva CE, Petrig BL, et al. Strict control of glycaemia: effects on blood flow in the large retinal vessels and in the macular microcirculation. Br J Ophthalmol 1995;79:735–741. 13. Grunwald JE, Riva CE, Martin DB, Quint AR, Epstein PA. Effect of an insulin-induced decrease in blood glucose on the human diabetic retinal circulation. Ophthalmology 1987;94: 1614 –1620. 14. White P. Diabetes mellitus in pregnancy. Clin Perinatol 1974;1:331–347. 15. Cunningham, FG. Williams Obstetrics. Stamford: Appleton and Lange, 1997:1–1448. 16. Axer-Siegel R, Hod M, Fink-Cohen S, et al. Diabetic retinopathy during pregnancy. Ophthalmology 1996;103: 1815–1819. 17. Klein BE, Moss SE, Klein R. Effect of pregnancy on progression of diabetic retinopathy. Diabetes Care 1990; 13:34 – 40. 18. Moloney JB, Drury MI. The effect of pregnancy on the natural course of diabetic retinopathy. Am J Ophthalmol 1982;93:745–756. 19. Dibble CM, Kochenour NK, Worley RJ, Tyler FH, Swarz M. Effect of pregnancy on diabetic retinopathy. Obstet Gynecol 1982;59:699 –704. 20. Klein BE, Klein R. Gravidity and diabetic retinopathy. Am J Epidemiol 1984;119:564 –569.
RETINAL HEMODYNAMICS
483
a mechanical image streaking velocimeter. Microvasc Res 1979;17:330 –332. 25. Hellstedt T, Kaaja R, Teramo K, Immonen I. Macular blood flow during pregnancy in patients with early diabetic retinopathy measured by blue-field entopic simulation. Graefe’s Arch Clin Exp Ophthalmol 1996;234:659 – 663. 26. Phelps RL, Sakol P, Metzger BE, Jampol LM, Freinkel N. Changes in diabetic retinopathy during pregnancy. Arch Ophthalmol 1986;104:1806 –1810.
21. Best RM, Chakravarthy U. Diabetic retinopathy in pregnancy. Br J Ophthalmol 1997;81:249 –251. 22. Soubrane G, Canivet J, Coscas G. Influence of pregnancy on the evolution of background retinopathy. Int Ophthalmol Clin 1985;8:249 –255. 23. Chen HC, Newsom RSB, Patel V, et al. Retinal blood flow changes during pregnancy in women with diabetes. Invest Ophthalmol Vis Sci 1994;35:3199 –3208. 24. Damon DN, Duling BR. A comparison between mean blood velocities and center-line red cell velocities as measured with
Authors Interactive威 We encourage questions and comments regarding this article via the Internet on Authors Interactive威 at http://www.ajo.com/ Questions, comments, and author responses are posted.
484
AMERICAN JOURNAL
OF
OPHTHALMOLOGY
OCTOBER 1999
● PURPOSE:
To investigate retinal circulatory changes that occur during the third trimester of pregnancy in diabetic patients and control subjects. ● METHODS: Bidirectional laser Doppler velocimetry and monochromatic fundus photography were used to assess the retinal circulation in seven pregnant diabetic patients and 13 age-matched pregnant control subjects. Retinal venous diameter (D), maximum erythrocyte velocity (Vmax), and retinal volumetric blood flow rate (Q) were measured in one eye of each subject during the third trimester of pregnancy (DPREG, VmaxPREG, and QPREG, respectively). These measurements were repeated during the postpartum period for both diabetic patients (11 ⴞ 7 weeks postpartum) and control subjects (16 ⴞ 6 weeks postpartum; P ⴝ .203; DPOST, VmaxPOST, and QPOST). ● RESULTS: In control subjects, DPREG was significantly reduced by ⴚ4.5% ⴞ 4.4% (mean percent difference ⴞ 1 standard deviation; paired t test, P ⴝ .006) relative to DPOST. In diabetic women, DPREG was also significantly reduced by ⴚ8.1% ⴞ 3.2% compared with DPOST (P ⴝ .001), a change that was significantly larger than that seen in control subjects (unpaired t test; P ⴝ .035). Compared with QPOST, QPREG was reduced by ⴚ7.1% ⴞ 14.2% (P ⴝ .123), in control subjects. In diabetic women, QPREG was significantly decreased by ⴚ18.4% ⴞ 9.3% compared with QPOST (P ⴝ .012). This reduction in QPREG was significantly greater in diabetic patients than in nondiabetic control subjects (unpaired t test, P ⴝ .040). No significant differences between VmaxPREG and VmaxPOST were observed in either diabetic patients (ⴚ3.1% ⴞ Accepted for publication June 8, 1999. From the Department of Ophthalmology, Scheie Eye Institute, School of Medicine, University of Pennsylvania, Philadelphia, Pennsylvania. This study was supported by the Vivian Simkins Lasko Retinal Vascular Research Fund, Research to Prevent Blindness, New York, New York, the Nina C. Mackall Trust, and grants EYO3243 and EY11479 from the National Institutes of Health, Bethesda, Maryland. Reprint requests to Juan E. Grunwald, MD, Scheie Eye Institute, University of Pennsylvania, 51 N 39th St, Philadelphia, PA 19104; fax: (215) 662-8025; e-mail: [email protected] 0002-9394/99/$20.00 PII S0002-9394(99)00234-2
©
1999 BY
12.9%; P ⴝ .400) or control subjects (ⴙ1.9% ⴞ 14.4%; P ⴝ .787). ● CONCLUSIONS: Retinal venous diameter is decreased during the third trimester of pregnancy in both diabetic and nondiabetic mothers. This decrease is significantly larger in diabetic than in nondiabetic mothers. In addition, we observed a reduction in retinal volumetric blood flow in diabetic patients during pregnancy that was significantly larger than that present in nondiabetic women. This fall in retinal volumetric blood flow in diabetic patients may exacerbate retinal ischemia and hypoxia and thus may be associated with the progression of diabetic retinopathy. (Am J Ophthalmol 1999;128: 477– 484. © 1999 by Elsevier Science Inc. All rights reserved.)
D
IABETIC RETINOPATHY, A MICROVASCULAR COM-
plication of diabetes, is the leading cause of blindness between the ages of 24 and 64 in the United States and the United Kingdom.1,2 Investigators have shown abnormalities in retinal vascular diameters and retinal blood flow in both the early3– 4 and late stages of diabetes.5– 8 In addition to blood flow abnormalities, patients with diabetic retinopathy also have an impaired ability to adjust retinal blood flow in the face of metabolic and environmental challenges.9 –13 In the past, pregnant diabetic women with evidence of diabetic retinopathy were often advised to terminate their pregnancies.14 Advances in obstetrics now allow an increasing number of women with diabetes to experience pregnancy, so we need to develop a better understanding of the relationships between pregnancy and the complications of diabetes mellitus. When studying pregnancy, it is important to note the physiologic changes that occur. In addition to the hormonal, metabolic, and immunologic alterations frequently described, significant differences are also present in the cardiovascular and hematologic systems. These changes include increased cardiac output, decreased vascular resistance, lowered arterial blood pressure, and increased
ELSEVIER SCIENCE INC. ALL
RIGHTS RESERVED.
477
resting pulse rate. Blood volume also increases approximately 45% above nonpregnancy levels.15 Several investigators have demonstrated that pregnancy adversely effects diabetic retinopathy.16 –22 Axer-Siegel and associates16 reported that of 22 pregnant, diabetic women with diabetic retinopathy at conception, 77.5% experienced progression of retinopathy and 22.5% developed proliferative diabetic retinopathy. Similarly, Klein and associates17 studied the rate of progression of diabetic retinopathy in 171 pregnant and 298 nonpregnant diabetic women. After accounting for glycemic control, blood pressure, and duration of diabetes, they discovered that current pregnancy was a major risk factor for the progression of retinopathy. The physiologic changes of pregnancy combined with a compromised retinal vascular regulatory capacity may play a role in the observed progression of diabetic retinopathy during pregnancy. Chen and associates23 studied retinal blood flow in pregnant women with and without diabetes and found a 14% to 19% increase in retinal blood flow in diabetic patients whose retinopathy progressed. In this study, we further investigated the changes in retinal blood flow that occur in diabetic patients and nondiabetic control subjects during the third trimester of pregnancy compared with the postpartum period.
TABLE 1. Characteristics of Diabetic Patients and Control Subjects
Characteristics
Age (yrs) 33 ⫾ 5† Duration of diabetes (yrs) 13.8 ⫾ 10.0 34 ⫾ 2 Gestation (wks)‡ 11 ⫾ 7 Postpartum (wks)‡ BPm during pregnancy (mm Hg) 85 ⫾ 11 BPm postpartum 85 ⫾ 8
32 ⫾ 4 NA 34 ⫾ 3 16 ⫾ 6 88 ⫾ 11 86 ⫾ 16
.557 NA .967 .203 .541 .857
superior and inferior retinal veins close to the optic disk. Each measurement site was marked on the fundus photograph. After the bidirectional laser Doppler velocimetry recordings, a fundus camera was used to take fundus photographs in monochromatic light at 570 nm. The diameter of the retinal vessel at the site of the bidirectional laser Doppler velocimetry recordings was assessed from projected photographic negatives. For each measurement site, six photographs were analyzed. Each of the six photographs was measured five times, and the results were averaged to obtain the retinal venous diameter (D). A second trained examiner then determined the maximum erythrocyte velocity (Vmax) and calculated the mean velocity of whole blood (Vmean) according to the equation, Vmean ⫽ C 䡠 Vmax, where C ⫽ 1/1.6.24 An average of Vmean for both the superior and inferior major retinal veins was used to calculate the average retinal volumetric blood flow rate (Q) according to the relation, Q ⫽ Vmean 䡠 D2/4. On completion of bidirectional laser Doppler velocimetry and monochromatic fundus photography determinations, baseline blood pressure and heart rate were immediately obtained. The mean brachial artery blood pressure was estimated with the standard formula, BPm ⫽ BPd ⫹ 1/3 (BPs ⫺ BPd). In each subject, measurements of vessel diameter, blood velocity, and blood flow were performed during the third trimester of pregnancy (DPREG, VmaxPREG, and QPREG). These measurements were repeated during the postpartum period for both diabetic patients (11 ⫾ 7 weeks postpartum) and control subjects (16 ⫾ 6 weeks postpartum; P ⫽ .203; DPOST, VmaxPOST, and QPOST). Determinations of Vmax and D were performed by separate examiners, who were masked in regard to the diagnosis and the timing of the measurements. Third-trimester measurements were compared with those obtained postpartum, and changes noted between
THE STUDY POPULATION INCLUDED SEVEN PREGNANT DIA-
betic women and 13 age-matched, nondiabetic pregnant control subjects recruited as volunteers by Scheie Eye Institute at the University of Pennsylvania. Eligibility criteria for control subjects included the following: current pregnancy, no history of ocular disease, normal ocular examination, and no systemic or topical medications. For the diabetic patients, eligibility criteria included current diagnosis of diabetes requiring insulin and no other systemic or topical medications. All subjects had deep anterior chambers and pupillary dilation greater than 5 mm. Ages of the diabetic women ranged from 28 to 38 years (mean, 33 ⫾ 5 years; ⫾1 SD), and the duration of their diabetes varied from 4 months to 26 years (mean, 14 ⫾ 10 years). Similarly, the ages of the nondiabetic women ranged from 26 to 37 years (mean, 32 ⫾ 4 years). Three of the seven diabetic patients had background diabetic retinopathy and four had normal funduscopic examinations. Our Institutional Review Board had approved the study, and all subjects gave informed consent to participate. Bilateral pupillary dilation was attained with tropicamide 1.0% and phenylephrine hydrochloride 1.5%. A Polaroid color fundus photograph (Polaroid Corporation, Cambridge, Massachusetts) of the optic disk was then obtained to document the measurement sites. By use of bidirectional laser Doppler velocimetry, we measured the centerline erythrocyte velocity (Vmax) in the largest AMERICAN JOURNAL
Control Subjects P Value*
BPm ⫽ mean blood pressure; NA ⫽ not applicable. *Unpaired Student t test. † Mean ⫾ 1 SD. ‡ Number of weeks’ gestation or postpartum at the time measurements were collected.
METHODS
478
Diabetic Patients
OF
OPHTHALMOLOGY
OCTOBER 1999
TABLE 2. Retinal Circulatory Parameters Parameters
D (m) Diabetic patients Control subjects Vmax (cm/s) Diabetic patients Control subjects Q (l/min) Diabetic patients Control subjects
Postpartum
Pregnancy
% Difference
P Value*
P Value†
153 ⫾ 12‡ 152 ⫾ 11
140 ⫾ 8 145 ⫾ 9
⫺8.1 ⫾ 3.2 ⫺4.5 ⫾ 4.4
.001 .006
.035
1.84 ⫾ 0.34 1.83 ⫾ 0.23
1.75 ⫾ 0.19 1.85 ⫾ 0.33
⫺3.1 ⫾ 12.9 ⫹1.9 ⫾ 14.4
.400 .787
.504
13.0 ⫾ 3.8 12.6 ⫾ 2.8
10.3 ⫾ 2.1 11.7 ⫾ 2.7
⫺18.4 ⫾ 9.3 ⫺7.1 ⫾ 14.2
.012 .123
.040
*Comparison of postpartum and pregnancy values in each subject using paired Student t tests. † Comparison of percent differences between diabetic patients and control subjects using unpaired t tests. ‡ Mean ⫾ 1 SD.
FIGURE 1. In each nondiabetic mother, the percent difference in venous diameter during the third trimester of pregnancy compared with the postpartum value. Average venous diameter is significantly reduced by ⴚ4.5% (paired t test, P ⴝ .006).
FIGURE 2. In each diabetic mother, the percent difference in venous diameter during the third trimester of pregnancy compared with the postpartum value. Average venous diameter is significantly reduced by ⴚ8.1% (paired t test, P ⴝ .001).
these two periods were contrasted between the two study populations. Paired and unpaired Student t tests were used for the statistical analysis, and P values of less than .05 were considered statistically significant
the measurements during pregnancy and after delivery was also similar between groups. Table 2 shows the retinal circulatory parameters obtained during postpartum and pregnancy for both diabetic patients and control subjects. Relative to the postpartum values, DPREG in control subjects was significantly reduced by ⫺4.5% ⫾ 4.4% (paired Student t test, P ⫽ .006; Figure 1). Diabetic women also demonstrated a significant decrease in DPREG of ⫺8.1% ⫾ 3.2% (P ⫽ .001; Figure 2) compared with DPOST. The decrease in DPREG in diabetic mothers was significantly greater than that seen in control subjects (unpaired t test, P ⫽ .035; Figure 3).
RESULTS TABLE 1 SUMMARIZES THE CHARACTERISTICS OF DIABETIC
patients and control subjects, demonstrating no significant differences in age or mean blood pressure during pregnancy (BPmPREG) and postpartum (BPmPOST). The timing of VOL. 128, NO. 4
PREGNANCY
AND
RETINAL HEMODYNAMICS
479
significant differences were observed in either QPOST (unpaired t tests, P ⫽ .829) or QPREG (unpaired t tests, P ⫽ .246) when comparing diabetic patients with control subjects. A significant negative correlation was noted between BPmPOST and QPOST in diabetic mothers (R ⫽ ⫺0.790; P ⬍ .05; Figure 8, left) but not in nondiabetic mothers (R ⫽ ⫹0.252; P ⬎ .05; Figure 8, right). No significant correlation was noted between BPmPREG and QPREG in either group.
DISCUSSION DESPITE TIGHT GLYCEMIC CONTROL, DIABETIC RETINOPA-
thy presents a major problem during childbearing years. Although investigators have documented the acceleration of diabetic retinopathy during pregnancy, the pathogenesis remains unclear. We analyzed retinal circulatory parameters in pregnant diabetic patients and nondiabetic pregnant control subjects to further identify any circulatory changes that may be related to the worsening of diabetic retinopathy during pregnancy. Our study shows a vasoconstriction of ⫺4.5% ⫾ 4.4% in DPREG compared with DPOST in control subjects (P ⫽ .006). This decrease in retinal venous diameter may be a manifestation of the retina’s autoregulatory response to the physiologic changes of pregnancy, including increased cardiac output and increased pulse rate. The diabetic mothers demonstrated a vasoconstriction of ⫺8.1% ⫾ 3.2% (P ⫽ .001) in DPREG, which was significantly greater than that observed in control subjects (unpaired; P ⫽ .035). In addition, a significant negative correlation between DPOST and BPmPOST was observed in diabetic patients (R ⫽ ⫺0.786; P ⬍ .05), suggesting that higher blood pressures are associated with smaller vessel diameters. A similar correlation was not observed in control subjects (R ⫽ ⫹0.317; P ⬎ .05), nor was a significant correlation present between BPmPREG and DPREG in either group. Chen and associates23 also studied retinal circulatory parameters in pregnant diabetic and nondiabetic mothers. They found that nondiabetic pregnant women had smaller vessel diameters in each trimester of pregnancy compared with postpartum and concluded that the decreased diameter in control subjects was a response to increased velocity, representing an autoregulatory response. They also noted that during the first trimester, vessel diameters were narrower in the diabetic mothers whose diabetic retinopathy did not progress compared with postpartum (P ⫽ .003). The diameters during the second and third trimesters, however, were not significantly different from the postpartum value (P ⬎ .05) nor were the diameters of patients whose diabetic retinopathy progressed. This supports other studies that have shown a failure of autoregulation in patients with diabetic retinopathy.
FIGURE 3. In nondiabetic and diabetic mothers, the average percentage difference in venous diameter (ⴞ95% confidence intervals) during the third trimester of pregnancy. Compared with postpartum values, there is a ⴚ4.5% average decrease in venous diameter during the third trimester of pregnancy in nondiabetic women (paired t test, P ⴝ .006), and a ⴚ8.1% decrease in diabetic women (P ⴝ .001). The decrease in diabetic mothers is significantly greater than that seen in nondiabetic mothers (unpaired t test, P ⴝ .035).
When comparing diabetic patients and control subjects, no significant differences were observed in DPOST (unpaired t tests, P ⫽ .829) or DPREG (unpaired t tests, P ⫽ .214). In addition, a significant negative correlation between DPOST and BPmPOST was observed in diabetic women (R ⫽ ⫺0.786; P ⬍ .05; Figure 4, left) but not in control subjects (R ⫽ ⫹0.317; P ⬎ .05; Figure 4, right). No significant correlation was noted between BPmPREG and DPREG in either diabetic or nondiabetic mothers. In contrast to the diameter findings, no significant differences between VmaxPOST and VmaxPREG were noted in either diabetic patients (⫺3.1% ⫾ 12.9%; P ⫽ .400) or control subjects (⫹1.9% ⫾ 14.4%; P ⫽ .787). Again, comparisons of diabetic patients and control subjects showed no significant differences between VmaxPOST (unpaired t tests, P ⫽ .934) or VmaxPREG (unpaired t tests, P ⫽ .415). Table 2 also indicates that compared with QPOST, QPREG in control subjects was reduced by ⫺7.1% ⫾ 14.2%, but this difference was not statistically significant (P ⫽ .123; Figure 5). In diabetic women, however, QPREG was significantly decreased by ⫺18.4% ⫾ 9.3% compared with QPOST (P ⫽ .012; Figure 6). Similar to the results found for DPREG, the reduction in QPREG in diabetic patients was significantly larger than that detected in control subjects (unpaired t test, P ⫽ .040; Figure 7). No 480
AMERICAN JOURNAL
OF
OPHTHALMOLOGY
OCTOBER 1999
FIGURE 4. Venous diameter postpartum (DPOST) compared with mean blood pressure postpartum (BPmPOST) in diabetic women (left) and control subjects (right). There is a significant negative correlation between DPOST and BPmPOST in diabetic women (R ⴝ ⴚ0.786; P < .05; left), but not in control subjects (R ⴝ ⴙ0.317; P > .05; right).
FIGURE 5. In comparison with postpartum, percentage difference in average volumetric blood flow during the third trimester of pregnancy in nondiabetic women. Average volumetric blood flow in nondiabetic mothers is reduced by ⴚ7.1% during the third trimester of pregnancy, but this is not statistically significant (paired t test, P ⴝ .123).
Differences between the study patients of Chen and associates23 and our study may explain some of the differences observed in vessel diameter. In the study by VOL. 128, NO. 4
PREGNANCY
AND
FIGURE 6. In comparison with postpartum, percentage difference in average volumetric blood flow during the third trimester of pregnancy in diabetic women. Average volumetric blood flow in diabetic mothers is significantly reduced by ⴚ18.4% during the third trimester of pregnancy (paired t test, P ⴝ .012).
Chen and associates,23 12 of the 22 diabetic patients had background diabetic retinopathy, four had proliferative diabetic retinopathy, and the remaining six had no diabetic retinopathy. Fifty percent of the 22 patients
RETINAL HEMODYNAMICS
481
mainly because of the reduction in vessel diameter. Chen and associates23 also found no significant change in volumetric blood flow in control subjects between any of the trimesters and postpartum, but they found a nonsignificant trend toward higher flow rates during pregnancy (P ⬎ .09). In contrast to our findings in diabetic mothers, Chen and associates23 observed a significant increase in flow during the second (P ⫽ .04) and third (P ⫽ .04) trimesters of pregnancy compared with postpartum. Again, it is possible that the disparity between our results and those of Chen and associates23 may reflect differences in blood flow that may be caused by discrepancies in the severity and rate of progression of diabetic retinopathy in both studies. Hellstedt and associates25 used blue-field entopic simulation to study macular capillary blood flow during pregnancy and postpartum in diabetic patients with minimal diabetic retinopathy. In contrast to Chen and associates,23 who observed increased retinal blood flowduring pregnancy, Hellstedt and associates25 found no increase in macular capillary flow velocity in diabetic mothers during pregnancy compared with postpartum. Several authors have suggested mechanisms for the progression of diabetic retinopathy during pregnancy. Chen and associates23 proposed that this progression may partly result from increased retinal blood flow caused by hyperperfusion. The ability of the retina to autoregulate its blood flow is impaired in diabetic retinopathy,9 –12 and thus in a diabetic patient, the physiologic changes of pregnancy impose an added stress on an already compromised retinal circulation. Phelps and associates26 found a significant positive correlation between the progression of diabetic retinopathy during pregnancy and the degree of improvement in glycemia control. They therefore suggested that this rapid improvement in glycemia control might have a deleterious effect on the retina, causing progression of retinopathy. Our results suggest that in both diabetic and nondiabetic mothers, retinal venous diameter is decreased during the third trimester of pregnancy, with more marked reductions observed in diabetic patients. In addition, we observed a reduction in retinal volumetric blood flow in diabetic patients during pregnancy. The decrease in retinal volumetric blood flow in diabetic patients may exacerbate retinal ischemia and hypoxia and thus may be associated with the progression of diabetic retinopathy. It is also possible that the hemodynamic changes associated with diabetic retinopathy lead to differing responses to the physiologic changes of pregnancy, depending on the severity of diabetic retinopathy. This may explain the discrepancies between the results reported by Chen and associates23 and those observed in this study. Further studies are needed to clarify why diabetic retinopathy tends to progress during pregnancy and whether patients with more severe disease respond differently to the added physiologic stress of pregnancy.
FIGURE 7. Average percentage difference in volumetric blood flow (ⴞ95% confidence intervals) during the third trimester of pregnancy compared with postpartum in nondiabetic and diabetic mothers. Compared with postpartum values, there is an average ⴚ7.1% nonsignificant decrease in volumetric blood flow during the third trimester of pregnancy in nondiabetic women (paired t test, P ⴝ .122) and a significant decrease of ⴚ18.4% in diabetic women (paired t test, P ⴝ .012). The decrease in diabetic mothers is significantly larger than that seen in nondiabetic mothers (unpaired t test, P ⴝ .040).
had progression of diabetic retinopathy. In contrast, our patient population had significantly less disease and more homogeneity in the extent of their diabetic retinopathy, with only three patients with background diabetic retinopathy and the remaining patients with no detectable funduscopic changes. None of the diabetic patients in our study showed any worsening of diabetic retinopathy during pregnancy. Also in contrast to the results reported by Chen and associates,23 we did not detect significant differences in VmaxPREG compared with VmaxPOST in either group. Compared with postpartum values, Chen and associates23 detected significant increases in retinal blood velocity as pregnancy progressed in both diabetic patients and control subjects. Similar to our vessel diameter findings, a nonsignificant ⫺7.1% ⫾ 14.2% decrease in QPREG compared with QPOST (P ⫽ .123) in control subjects was observed. Diabetic mothers exhibited a ⫺18.4% ⫾ 9.3% reduction in QPREG compared with QPOST (P ⫽ .012). This decrease in flow during pregnancy in diabetic mothers was significantly larger than that observed in nondiabetic mothers (unpaired t test, P ⫽ .040). The decrease in QPREG compared with QPOST in both groups was 482
AMERICAN JOURNAL
OF
OPHTHALMOLOGY
OCTOBER 1999
FIGURE 8. Volumetric blood flow postpartum (QPOST) compared with mean blood pressure postpartum (BPmPOST) in diabetic women (left ) and control subjects (right). There is a significant negative correlation between QPOST and BPmPOST in diabetic women (R ⴝ ⴚ0.790; P < .05; left) but not in control subjects (R ⴝ ⴙ0.252; P > .05; right).
REFERENCES 1. Kohner EM, Porta M. Protocols for screening and treatment of diabetic retinopathy in Europe. Eur J Ophthalmol 1991; 1:45–54. 2. Foulds WS, MacCuish A, Barrie T, et al. Diabetic retinopathy in the West of Scotland: its detection, and prevalence, and the cost-effectiveness of a proposed screening programme. Health Bull 1983;41:318 –326. 3. Grunwald JE, DuPont J, Riva CE. Retinal haemodynamics in patients with early diabetes mellitus. Br J Ophthalmol 1996;80:327–331. 4. Bursell S-E, Clermont AC, Kinsley BT, Simonson DC, Aiello LM, Wolpert HA. Retinal blood flow changes in patients with insulin dependent diabetes mellitus and no diabetic retinopathy: a video fluorescein angiography study. Invest Ophthalmol Vis Sci 1996;37:886 – 897. 5. Grunwald JE, Riva CE, Sinclair SH, Brucker AJ, Petrig BL. Laser Doppler velocimetry study of retinal circulation in diabetes mellitus. Arch Ophthalmol 1986;104:991–996. 6. Patel V, Rassam SMB, Newson RSB, Wiek J, Kohner EM. Retinal blood flow in diabetic retinopathy. BMJ 1992;305: 678 – 684. 7. Feke GT, Buzney SM, Ogasawara H, et al. Retinal circulatory abnormalities in type I diabetes. Invest Ophthalmol Vis Sci 1994;35:2968. 8. Grunwald JE, Riva CE, Baine J, Brucker AJ. Total retinal volumetric blood flow rate in diabetic patients with poor diabetic control. Invest Ophthalmol Vis Sci 1992;33:356 – 363. 9. Grunwald JE, Riva CE, Brucker AJ, Sinclair SH, Petrig BL. Altered retinal vascular response to 100% oxygen breathing in diabetes mellitus. Ophthalmol 1984;91: 1447–1452.
VOL. 128, NO. 4
PREGNANCY
AND
10. Rassam SMB, Patel V, Kohner EM. The effect of experimental hypertension on retinal autoregulation in humans: a mechanism for the progression of diabetic retinopathy. Exp Physiol 1995;80:53– 68. 11. Ernest JT, Goldstick TK, Engerman RL. Hyperglycemia impairs retinal oxygen autoregulation in normal and diabetic dogs. Invest Ophthalmol Vis Sci 1983;24:985–989. 12. Grunwald JE, Riva CE, Petrig BL, et al. Strict control of glycaemia: effects on blood flow in the large retinal vessels and in the macular microcirculation. Br J Ophthalmol 1995;79:735–741. 13. Grunwald JE, Riva CE, Martin DB, Quint AR, Epstein PA. Effect of an insulin-induced decrease in blood glucose on the human diabetic retinal circulation. Ophthalmology 1987;94: 1614 –1620. 14. White P. Diabetes mellitus in pregnancy. Clin Perinatol 1974;1:331–347. 15. Cunningham, FG. Williams Obstetrics. Stamford: Appleton and Lange, 1997:1–1448. 16. Axer-Siegel R, Hod M, Fink-Cohen S, et al. Diabetic retinopathy during pregnancy. Ophthalmology 1996;103: 1815–1819. 17. Klein BE, Moss SE, Klein R. Effect of pregnancy on progression of diabetic retinopathy. Diabetes Care 1990; 13:34 – 40. 18. Moloney JB, Drury MI. The effect of pregnancy on the natural course of diabetic retinopathy. Am J Ophthalmol 1982;93:745–756. 19. Dibble CM, Kochenour NK, Worley RJ, Tyler FH, Swarz M. Effect of pregnancy on diabetic retinopathy. Obstet Gynecol 1982;59:699 –704. 20. Klein BE, Klein R. Gravidity and diabetic retinopathy. Am J Epidemiol 1984;119:564 –569.
RETINAL HEMODYNAMICS
483
a mechanical image streaking velocimeter. Microvasc Res 1979;17:330 –332. 25. Hellstedt T, Kaaja R, Teramo K, Immonen I. Macular blood flow during pregnancy in patients with early diabetic retinopathy measured by blue-field entopic simulation. Graefe’s Arch Clin Exp Ophthalmol 1996;234:659 – 663. 26. Phelps RL, Sakol P, Metzger BE, Jampol LM, Freinkel N. Changes in diabetic retinopathy during pregnancy. Arch Ophthalmol 1986;104:1806 –1810.
21. Best RM, Chakravarthy U. Diabetic retinopathy in pregnancy. Br J Ophthalmol 1997;81:249 –251. 22. Soubrane G, Canivet J, Coscas G. Influence of pregnancy on the evolution of background retinopathy. Int Ophthalmol Clin 1985;8:249 –255. 23. Chen HC, Newsom RSB, Patel V, et al. Retinal blood flow changes during pregnancy in women with diabetes. Invest Ophthalmol Vis Sci 1994;35:3199 –3208. 24. Damon DN, Duling BR. A comparison between mean blood velocities and center-line red cell velocities as measured with
Authors Interactive威 We encourage questions and comments regarding this article via the Internet on Authors Interactive威 at http://www.ajo.com/ Questions, comments, and author responses are posted.
484
AMERICAN JOURNAL
OF
OPHTHALMOLOGY
OCTOBER 1999