SEARCH

SEARCH BY CITATION

Keywords:

  • vasoactive hormones;
  • natriuretic peptides;
  • diabetic retinopathy;
  • diabetic pregnancy;
  • renin-angiotensin-system

Abstract.

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Material and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Purpose: To evaluate the role of various vasoactive hormones in the evolution of diabetic retinopathy during pregnancy and postpartum.

Methods: Retinopathy was graded from fundus photographs of 45 pregnant women with type 1 diabetes and seven pregnant women without diabetes in a prospective study. Markers of renin-angiotensin-system (RAS), plasma renin activity (PRA), angiotensin II (AngII), aldosterone, natriuretic peptides (ANP, BNP, CNP) and adreonomedullin (AM) were measured during the first and third trimesters and at 3 months postpartum. The women with diabetes were grouped by progression of retinopathy during pregnancy and postpartum.

Results: Levels of PRA (p = 0.001) and ANP (p = 0.03) were significantly lower in diabetes than in non-diabetes subjects throughout pregnancy and postpartum. No significant differences appeared in levels of AngII, aldosterone, AM, BNP or CNP between the two groups. In multivariate logistic regression analyses with retinopathy progression by the third trimester as the dependent variable, only duration of diabetes qualified in the model (p = 0.027, R = 0.227, Exp(B) = 1.28).

Conclusions: Diabetic pregnancy is associated with lower levels of PRA and ANP compared to non-diabetic pregnancy. Lowered RAS activity may contribute to the hyperdynamic blood flow and progression of DR during diabetic pregnancy. Within the power of this study no clear associations between the vasoactive hormones and progression of retinopathy could be detected.


Introduction

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Material and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Although pregnancy does not generally have a deleterious longterm effect on the evolution of diabetic retinopathy (DR) (Kaaja et al. 1996), at least a temporary increase in retinal lesions occurs in many women during pregnancy (Soubrane et al. 1985; Hellstedt et al. 1996). This is especially so if the metabolic control and ophthalmic follow-up have been suboptimal before pregnancy, in which case a dramatic progression in DR can occur over a short period of time (Laatikainen et al. 1987).

The mechanisms linking the metabolic changes in diabetes to DR have still not been completely clarified. According to one hypothesis, elevated glucose levels lead to loss of pericytes in the retinal capillaries, thereby impairing the autoregulation of the retinal capillary blood flow. The resulting hyperdynamic circulation damages endothelial cells and leads to the appearance of early lesions of DR. In support of this hypothesis, we recently found that retinal capillary blood flow is elevated in diabetic compared to non-diabetic pregnancy (Loukovaara et al. 2003). In order to determine why the retinal capillary blood flow responds differently to pregnancy in diabetic compared to non-diabetic women, we measured circulating systemic vasoconstrictive and vasodilating hormones during pregnancy in diabetic and non-diabetic women, and related their levels to the severity of DR.

Material and Methods

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Material and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

The study was carried out with the approval of the local institutional review board, in accordance with the tenets of the Declaration of Helsinki. Written informed consent was obtained from all participants.

Pregnant diabetic women

From November 1998 to January 2002, 72 women with type 1 diabetes were recruited to the study at the Department of Obstetrics and Gynaecology, Helsinki University Hospital, as soon as their pregnancy was diagnosed (usually between 5 and 10 weeks of gestation). The subjects were examined at the Department of Ophthalmology during gestation weeks 12–14 and 34–36, and at 3 months postpartum. All of them were treated with long-acting insulin 2–3 times a day. Short-acting insulin was given before meals.

Nine women were excluded because of obstetric complications or co-existing eye disease: one because of spontaneous abortion, one for an induced abortion due to a high glycosylated haemoglobin level (12–13%), six for pre-term delivery and one for retinitis pigmentosa. After these nine women had been excluded, 63 diabetic women were available for clinical characteristics. However, various technical reasons (inadequate quantity of samples, broken/lost tubes, etc.) reduced the total number of subjects in the statistical analyses. Finally, complete follow-up data of all measurements of vasoactive hormones were available for only 45 diabetic women. Seven diabetic women had nephropathy defined as an albumin excretion rate ≥ 0.3 g/24 hours on two consecutive visits (Thomas et al. 2001). The creatinine clearance was normal in all these women.

Non-diabetic pregnant women

Fifteen non-diabetic women attending the Department of Obstetrics and Gynaecology for monitoring of normal pregnancy and in the same age range as the diabetic women were recruited as controls. Inability to attend examinations at all time-points and lost/insufficient samples reduced the total number of control women to seven in the final statistical analyses. All non-diabetes subjects had visual acuities of 6/6 or better and a normal ocular examination. None of them smoked or used any medication.

Ophthalmic examination

All subjects underwent a complete ophthalmic examination including measurement of visual acuity, biomicroscopy, indirect ophthalmoscopy and fundus photography. Intraocular pressure was below 21 mmHg in all participants.

Fundus photography

Fundus photography of both eyes was performed through dilated pupils (two drops of tropicamide, 5 mg/ml) by a trained operator using a Topcon TRC 50IA retinal camera (Topcon Corporation, Tokyo, Japan) and Kodak Elitechrome 100 film (Eastman Kodak, Rochester, New York, USA). The severity of DR was assessed using two 50-degree colour slides, one centred at the macula, and the other at the optic nerve head.

The photographs were evaluated by a retinal specialist (IJI) who was masked to all clinical information. Retinopathy was graded using a modification of the non-stereoscopic grading system defined by the Early Treatment Diabetic Retinopathy Study (ETDRS) Research Group (1991). For each eye, the maximum grade of lesions was determined to produce the overall level of severity for that eye (RP level) (Loukovaara et al. 2003). All the non-diabetes subjects were graded as not having any retinopathy in the masked grading. Finally, the retinopathy levels from both eyes were combined to give a score of retinopathy severity for each patient, using a scale of 1–11 according to the Diabetes Control and Complications Trial (DCCT) Research Group (1993).

Progression of retinopathy was classified as follows:

  • 1
    no progression when there was no change or a decrease in DCCT score during pregnancy, and
  • 2
    progression if the DCCT score increased by one or more levels during pregnancy.

Measurement of serum glycosylated haemoglobin concentration

Serum glycosylated haemoglobin (HbA1c) concentrations were measured by ion-exchange high performance liquid chromatography (Diamat, Bio-Rad Laboratories, Hercules, California, USA). Four values of HbA1c were used to represent the glucose control in the diabetes subjects: the mean values of all three HbA1c measurements (%) taken during the first and third trimesters, as well as the mean value for the whole pregnancy. Pre-gestational HbA1c values were available for 21 (47%) of the diabetic women (Table 1). The reference value for HbA1c was 4–6%.

Table 1.  Baseline characteristics of type 1 diabetic women and non-diabetic control subjects.
 Diabetic womenControlsp
 (n = 45)(n = 7) 
  • ND = not done.

  • Values are mean ± SD or n%.

  • *

    Pre-pregnancy HbA1c (%) values were available for 21 (47%) of the diabetic women.

Age (years)30.2 ± 4.532.4 ± 3.20.13
Duration of diabetes (years)16.7 ± 7.5  
Pre-pregnancy body mass index (kg/m2)24.8 ± 3.722.9 ± 2.00.05
Parity (%)
 Nulliparous22 (49)3 (43)1.0
 Parous23 (51)4 (57) 
Smoking (%)9 (20.0)0 (0)0.33
Duration of pregnancy (days)257.0 ± 6.4284.0 ± 6.1< 0.0001
Birth weight of the infant (g)  3669 ± 6793868 ± 4920.56
Blood pressure (mmHg)
 Systolic/1st trimester (max) 124 ± 13122 ± 90.81
 Diastolic/1st trimester (max) 75 ± 1177 ± 70.92
 Systolic/3rd trimester (max) 143 ± 24122 ± 90.01
 Diastolic/3rd trimester (max) 84 ± 1481 ± 90.49
HbA1c (%)
 Pre-pregnancy*7.9 ± 0.9ND 
 First trimester7.3 ± 0.9ND 
 Third trimester6.8 ± 0.9ND 
 Mean during whole pregnancy6.9 ± 0.7ND 
Nephropathy (%)7 (15.6)0 (0)0.57

Measurement of plasma vasoactive hormones

Venous samples were drawn at gestational weeks 12–14, 34–36 and at 3 months postpartum into chilled tubes with the subjects in sitting position after 15 min rest at noon. The samples were centrifuged at 7000 g for 10 min at 4 ° and the plasma immediately stored at − 70 ° until assayed. The samples were transported on dry ice to the endocrine laboratory in Christchurch Hospital, New Zealand for analysis. All samples from each subject were measured in a single assay to avoid interassay variability. Intra-assay coefficients of variation were between 5% and 9%. Radioimmunoassay (RIA) was used for measurements of plasma renin activity (PRA) (Dunn & Espiner 1976), angiotensin II (AngII) (Nicholls & Espiner 1976), aldosterone (Lun et al. 1983), atrial (A-type) natriuretic peptide (ANP) (Yandle et al. 1986), brain (B-type) natriuretic peptide (BNP) (Yandle et al. 1993), C-type natriuretic peptide (CNP) (Hunt et al. 1994) and adrenomedullin (AM) (Lewis et al. 1998).

Statistical analyses

Statistical analyses were performed using spss Version 9.0 for Windows (SPSS Inc., Chicago, Illinois, USA). Variables were tested for normality before applying parametric tests. The Mann–Whitney U-test was used for non-normally distributed continuous variables. For categorical variables, Fisher's exact test was used. Repeated measures anova with Bonferroni corrections was performed to compare temporal changes in vasoactive hormones (PRA, aldosterone, AngII, ANP, BNP, CNP, AM) during pregnancy and postpartum between the study groups. Multivariate logistic regression analyses with progression of retinopathy as the dependent variable was performed. P < 0.05 was considered statistically significant. Data are presented as mean and SD or n (%).

It has previously been reported that diabetic women have lower values of PRA than non-diabetic women (Bojestig et al. 2000). Our sample sizes of 45 diabetic women and seven controls gave this study 78% power to detect a difference of 2.5 pmol/l/hour in PRA levels between the null hypothesis that both group means were 5.0 pmol/l/hour and the alternative hypothesis that the mean of the control group was 2.5 pmol/l/hour with known group SDs of 2.7 pmol/l/hour and 1.0 pmol/l/hour and with a significance level of p = 0.05 using a one-sided, two-sample t-test. Additionally, with group sample sizes of 31 diabetic women without retinopathy progression and 14 with progression, this study had 93% power to detect a difference of 1.0 pmol/l/hour in PRA levels between the null hypothesis that both group means were 3.0 pmol/l/hour and the alternative hypothesis that the mean of the progression group was 2.0 pmol/l/hour with known group SDs of 1.0 pmol/l/hour and 1.0 pmol/l/hour and with a significance level of p = 0.05 using a one-sided, two-sample t-test.

Results

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Material and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

The baseline characteristics of the diabetic and non-diabetic women are summarized in Table 1. Differences between the groups were statistically significant for pre-pregnancy body mass index, duration of pregnancy and systolic blood pressure in the third trimester.

Retinopathy status

Twenty-four (53.3%) women had mild retinopathy (DCCT score ≤ 3) and 21 (46.7%) had more severe retinopathy (DCCT score > 3) in the first trimester. By the third trimester 15 (33.3%) of the women had mild retinopathy, and 30 (66.7%) had more severe retinopathy. During pregnancy, 31 (68.9%) of the diabetic women had no progression in the DCCT score and 14 (31.1%) progressed. Out of these 14 women, seven progressed more than two DCCT steps.

Nine (20.0%) diabetic women had received laser treatment prior to the current pregnancy; five (11.1%) had received panphotocoagulation because of proliferative diabetic retinopathy, and four (8.9%) had received local laser treatment. Four (8.9%) diabetic women developed proliferative changes during the study that required subsequent laser treatment. In addition, one diabetic woman received laser treatment because of leaking microaneurysms, and one because of local vitreous traction and haemorrhage from an avulsed temporal vein.

Nine diabetic women were smokers. They had been smoking before the current pregnancy and continued to smoke during pregnancy. The daily consumption of cigarettes was less than 20. However, there was no difference in retinopathy progression during pregnancy between the non-smoking (n = 36) and smoking (n = 9) diabetic women (p = 1.0, Fisher's exact test).

The eye status of the non-participating diabetic women (n = 27), who were excluded from the final analyses, did not differ from that of those included. Fourteen (51.9%) of the non-participating women had mild retinopathy and 13 (48.1%) had more severe retinopathy in the first trimester. One (3.7%) of the excluded patients had received laser treatment prior to the current pregnancy and two (7.4%) developed proliferative changes during the study.

Vasoactive hormones

Plasma concentrations of markers of renin-angiotensin-system (RAS) and AM between the study groups are summarized in Table 2 and markers of natriuretic peptides in Table 3.

Table 2.  Plasma concentrations of markers of renin-angiotensin-system and adrenomedullin during pregnancy in type 1 diabetic women without (stable DCCT, n = 31) or with (worsening DCCT, n = 14) progression of diabetic retinopathy and in non-diabetic controls (n = 7).
 1st trimester3rd trimester3 months postpartump (diabetic women versus controls)p (between diabetic women: progression versus no progression)
  1. AngII = angiotensin II; PRA = plasma renin activity; AM =adrenomedullin.

  2. Values are mean ± SD.

  3. P-values calculated with repeated measures anova with Bonferroni correction.

AngII (pmol/l)
All diabetes subjects10.7 ± 4.819.5 ± 16.74.4 ± 2.70.092 
 No progression11.7 ± 4.521.4 ± 19.54.1 ± 2.40.4471.0
 Progression10.6 ± 5.515.3 ± 6.95.1 ± 3.20.192 
Controls11.9 ± 4.832.0 ± 20.64.7 ± 2.4  
PRA (nmol/l/hour)
All diabetes subjects2.7 ± 0.92.5 ± 1.00.9 ± 0.50.001 
 No progression2.6 ± 0.82.5 ± 1.10.8 ± 0.40.0021.0
 Progression2.7 ± 1.22.4 ± 1.01.0 ± 0.60.009 
Controls3.0 ± 0.95.0 ± 2.71.2 ± 0.8  
Aldosterone (pmol/l)
All diabetes subjects715 ± 3311989 ± 963187 ± 1270.055 
 No progression698 ± 3562041 ± 1135175 ± 1340.2131.0
 Progression751 ± 2761874 ± 376214 ± 1080.233 
Controls1055 ± 2572439 ± 1373342 ± 155  
AM (pmol/l)
All diabetes subjects5.0 ± 2.56.8 ± 4.24.8 ± 1.40.665 
 No progression4.8 ± 2.27.3 ± 4.74.7 ± 1.31.01.0
 Progression5.3 ± 3.25.9 ± 2.75.0 ± 1.61.0 
Controls5.0 ± 0.98.2 ± 2.54.5 ± 1.0  
Table 3.  Plasma concentrations of natriuretic peptides during pregnancy in type 1 diabetic women without (stable DCCT, n = 31) or with (worsening DCCT, n = 14) progression of diabetic retinopathy and in non-diabetic controls (n = 7).
 1st trimester3rd trimester3 months postpartump (diabetic women versus controls)p (between diabetic women: progression versus no progression)
  1. ANP = atrial natriuretic peptide; BNP = brain natriuretic peptide; CNP = C-type natriuretic peptide.

  2. Values are mean ± SD.

  3. P-values calculated with repeated measures anova with Bonferroni correction.

ANP (pmol/l)
All diabetes subjects6.6 ± 2.58.6 ± 5.98.1 ± 3.50.031 
 No progression6.3 ± 2.77.8 ± 4.78.1 ± 3.70.0460.519
 Progression7.1 ± 1.710.4 ± 7.88.3 ± 3.10.592 
Controls9.7 ± 3.79.0 ± 3.812.0 ± 6.2  
BNP (pmol/l)
All diabetes subjects3.1 ± 1.33.1 ± 1.84.0 ± 1.80.676 
 No progression3.1 ± 1.42.9 ± 1.73.6 ± 1.51.00.457
 Progression2.9 ± 1.13.4 ± 1.94.9 ± 2.21.0 
Controls3.4 ± 1.52.6 ± 0.64.8 ± 2.5  
CNP (pmol/l)
All diabetes subjects1.1 ± 0.51.4 ± 1.31.1 ± 0.50.629 
 No progression1.0 ± 0.31.3 ± 0.41.0 ± 0.31.00.544
 Progression1.2 ± 0.91.7 ± 2.51.3 ± 0.70.856 
Controls0.9 ± 0.21.2 ± 0.51.0 ± 0.3  

Levels of PRA, AngII and aldosterone were lower in diabetes than in control subjects throughout pregnancy and postpartum, although this difference was statistically significant only for PRA (p = 0.001). Levels of PRA, AngII and aldosterone decreased during the postpartum period in both the diabetic and non-diabetic women (Table 2). However, there were no differences in the levels of AngII, aldosterone or AM between the diabetic and non-diabetic women when grouped by progression of retinopathy (Table 2).

Levels of ANP were significantly lower in the diabetic than in the non-diabetic women, but there was no significant difference between those who progressed compared to those who did not (Table 3). Other natriuretic peptides did not differ between the study groups. Neither were there any differences in levels of vasoactive hormones in diabetic women with nephropathy (n = 7) to those without (n = 38).

No differences in vasoactive hormones were found in the subgroups according to mean HbA1c. Likewise, vasoactive hormone levels did not differ between diabetic women with proliferative changes (n = 4) during the study period compared to those women without. Additionally, the exclusion of diabetic women with previous proliferative diabetic retinopathy does not change the results (data not shown).

Multivariate logistic regression analyses

A multivariate logistic regression analyses was performed to assess vasoactive hormones of interest and clinical variables (duration of diabetes, age of patient, level of retinopathy before current pregnancy, systolic and diastolic blood pressure in the third trimester, mean HbA1c during pregnancy) associated with retinopathy progression. With retinopathy progression by the third trimester as the dependent variable, only duration of diabetes qualified in the model (p = 0.027, R = 0.227, Exp(B) = 1.28). The vasoactive hormones of interest were dropped from the model. P = 0.2 was used as an entry criterion.

Discussion

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Material and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Hyperdynamic retinal capillary blood flow and consequent endothelial cell damage have been implicated as possible mechanisms in the pathogenesis of early DR. Systemic circulation is increased during pregnancy as indicated by elevated cardiac output and decreased peripheral resistance (Atkins et al. 1981). We have recently found retinal blood flow during pregnancy to be higher in diabetic than in non-diabetic women, potentially contributing to the progression of DR during pregnancy (Loukovaara et al. 2003).

Tissue circulation is under the regulation of both systemic and local mediators. In this study, we wanted to evaluate the possible role of systemic circulatory factors on the differences in the circulatory response to pregnancy between diabetic and non-diabetic women, and especially correlations to progression of DR.

As shown previously, the renin-angiotensin-aldosterone axis is activated during pregnancy (Immonen et al. 1983). However, all the measured parameters of the RAS showed either a trend to be lower or were significantly lower in diabetic compared to non-diabetic women during pregnancy. To our knowledge, this comparison during pregnancy has not been published previously, although Bojestig et al. (2000) reported the RAS to be suppressed in non-pregnant diabetes subjects.

Angiotensin II, the main mediator of the RAS, is primarily a potent vasoconstrictor. Thus, lowered activity of this system in diabetic women would be expected to increase tissue blood flow, and in the case of the diabetic retina contribute to the hyperdynamic blood flow and progression of retinopathy. However, as there was no correlation to DR progression in our patients, suppression of the RAS cannot be considered as a triggering factor for DR progression during pregnancy. Circulation in the normal retina is under efficient autoregulatory control. In diabetes, these autoregulatory mechanisms are weakened, but it is unknown to what extent systemic mediators regulate blood flow in the diabetic retina. To make things even more complicated, a local RAS has also been described in the retina, but the interplay between the systemic and local RAS is unknown (Danser et al. 1994; Wagner et al. 1996; Williams 1998; Wilkinson-Berka et al. 2001; Strain & Chaturvedi 2002). Some researchers have reported heightened vascular reactivity to AngII in type 1 diabetes subjects (Drury et al. 1984), including those with retinopathy (Christlieb et al. 1976).

Of the vasodilatory mediators studied, ANP also showed suppressed values in diabetic compared to normal pregnancy, whereas BNP, CNP and adrenomedullin showed no correlation to diabetes. The actions of natriuretic peptides include natriuresis, vasodilatation and inhibition of the renin-angiotensin-aldosterone (RAAS) axis (Yano et al. 1998). There are no previous reports on ANP during diabetic pregnancy, but increased or constant levels have been reported in normal pregnancy (McCance et al. 1990; Castro et al. 1994; Yoshimura et al. 1994; Sala et al. 1995; Minegishi et al. 1999). An elevation of ANP levels has been described in diabetes (Dussaule et al. 1988; Opocher et al. 1989), with higher levels of ANP in patients with DR (Yano et al. 1998).

To what extent the lower ANP levels compensate for the relative decrease in the activity of the RAS, detected here in diabetic pregnancy, cannot be deducted from our results. The biology of the other natriuretic peptides studied, BNP and CNP, is less documented, but shows similarities with that of ANP (Yoshimura et al. 1994; Stepan et al. 1998). In our study, these peptides did not show any correlation to either diabetic pregnancy or DR, however.

Adrenomedullin is another vasodilatory peptide produced by the endothelium and smooth muscle cells (Lainchbury et al. 1997) and with a postulated role in DR (Turk et al. 2000; Pio et al. 2001). Like BNP and CNP, AM levels in our study were not differentiated by diabetes during pregnancy or by progression of DR, a finding that agrees with research by Di Iorio et al. (2001).

Our study has some limitations that need to be mentioned. The pathophysiological importance of hormones may not always be reflected accurately by their levels in plasma. This is especially true for AM, which is seen primarily as a paracrine hormone acting within blood vessels. Likewise, tissue levels of some hormones including AngII and ANP, especially in the eye and within blood vessels, might exert potent actions that may not be reflected by their levels within the circulation. We must also point out that lack of evidence of the role of systemic vasoactive hormones on DR progression during pregnancy can be related to good glycaemic control with only mild progression of retinopathy during pregnancy in many patients. Progression of retinopathy is known to be associated with suboptimal glycaemic control before pregnancy and rapid normalization of hyperglycaemic blood glucose levels during pregnancy (Laatikainen et al. 1987; Lauszus et al. 2000). In our study, these mechanisms were less likely to cause progression as almost half of the diabetic women had taken part in pre-pregnancy planning, and reached fairly optimal glycaemic control before current pregnancy.

In summary, the diabetic women had lower levels of PRA and ANP than controls during pregnancy and postpartum. Lowered RAS activity may contribute to the hyperdynamic blood flow and progression of DR. However, within the power of this study we were unable to discern any close statistical associations between the vasoactive hormones measured and the progression of DR.

Acknowledgements

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Material and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

This study was presented in part at the Association for Research in Vision and Ophthalmology (ARVO), Annual Meeting in Fort Lauderdale, Florida, USA, 25–29th April 2004. The research was supported by the Finnish Eye Foundation, Helsinki; the Finnish Eye and Tissue Bank Foundation, Helsinki; Friends of the Blind, Helsinki; the Oskar Öflund Foundation, Helsinki; the Maud Kuistila Foundation, Helsinki; the Foundation for Diabetes Research, Tampere; the South Carelian Society of Doctors and Duodecim of Viborg, Lappeenranta, and HUCH Clinical Research Grant (TYH1325), Helsinki, Finland.

References

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Material and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References