Inflammatory Markers and Diabetic Retinopathy in Type 1 Diabetes

This was a cross-sectional prospective study comprised of subjects with type 1 diabetes (n = 154) matched by age and sex with nondiabetic control subjects in a 3:1 ratio. Eligible subjects (control subjects and type 1 diabetic patients) were between the ages of 14 and 42 years. Test subjects had type 1 diabetes for >5 years. Subjects were excluded if they had a blood pressure >140/90 mmHg, were tobacco users, were on medications known to affect CRP (ACE inhibitors, hydroxymethylglutaryl CoA reductase inhibitors, nonsteroidal anti-inflammatory agents, and aspirin), had other known concomitant illnesses, or were pregnant. Written informed consent was obtained from willing participants using institutional review board-approved consent forms and Health Insurance Portability and Accountability Act-compliant authorization forms.

Eye examinations included direct ophthalmoscopy (with pupils dilated) as well as digital retinal photographs that were obtained with a Canon CR6-45NM D60 camera. Three fields were photographed for each eye: optic nerve and macula, nasal retina, and temporal retina. A single ophthalmologist (W.E.J.) did the grading for the eye exams using the modified Airlie House classification (7).

All samples were spun in a refrigerated centrifuge at 2,000g for 10 min at 4°C. Plasma and serum were extracted and stored at −70°C until assayed. The plasma samples for prostaglandin E₁ (PgE₁) and prostaglandin E₂ (PgE₂) were stored in special tubes containing 100 μg indomethacin. The HbA_1C was measured with the DCA-2000 Meter (Bayer Laboratories, Elkhart, IN). IL-6, PgE₁, and PgE₂ were assayed using colorimetric sandwich enzyme-linked immunosorbent assay kits from R&D Systems (Minneapolis, MN). The high-sensitivity CRP assay was performed on a BNII Nephelometer (Dade Behring, Deerfield, IL).

Differences in the levels of the inflammatory markers were compared between the groups with different eye grades. The CRP levels were normalized using a logarithmic transformation. The comparisons were first made without any adjustment and were then repeated following adjustment for age, duration of diabetes, sex, and BMI. The variables were analyzed using a one-way ANCOVA, which gave P- values for comparisons among the three diabetic groups and among all diabetic and control subjects before and after adjustments for age, duration of diabetes, sex, and BMI as appropriate. The SAS statistical package was used (8).

There were a total of 115 type 1 diabetic subjects (51 males and 64 females) with a mean ± SD age and duration of diabetes of 26.7 ± 7.2 and 16.0 ± 7.9 years, respectively. Thirty-nine control subjects (18 males and 21 females) with a mean ± SD age of 26.6 ± 6.7 years were enrolled. Age (P = 0.001), BMI (P = 0.02), and duration of diabetes (P = 0.001) were significantly higher in the 39 diabetic subjects with grades 4–6 retinopathy compared with subjects with grades 2/3 retinopathy (42 subjects) or no retinopathy (37 subjects). There were no significant differences in HbA_1c levels (7.8–8.0%) among the three groups (P = 0.79).

PgE₂ levels were significantly higher in the diabetic population than in the control group (P < 0.001) (Table 1). This was true for all levels of retinopathy. This relationship remained significant after controlling for the effects of BMI. All other markers of inflammation were similar between the two groups.

A significant relationship was found (P = 0.01) between grades of retinopathy and CRP. However, after controlling for age, duration of diabetes, sex, and BMI, the significance of the relationship was lost (P = 0.42). None of the other markers of inflammation were significantly associated with retinopathy.

For the diabetic population, BMI was positively correlated with CRP and IL-6 before adjustment for age, sex, and duration of diabetes (P < 0.001 for both) and after adjustment (P = 0.04 and 0.02, respectively). Age (P = 0.002), BMI (P = 0.001), and IL-6 (P = 0.001) were the only significant predictors of CRP. The relative increase in CRP for each of these variables was adjusted for the other variables in the model. IL-6 and CRP were also positively correlated with each other.

The purpose of this study was to determine the contribution of inflammation to the pathogenesis of retinopathy in subjects with type 1 diabetes. Although CRP was significantly associated with the eye grades, the significance was lost after controlling for the effect of other covariables (age, duration of diabetes, and BMI). Other markers of inflammation (IL6, PgE₁, and PgE₂) did not vary significantly among groups of type 1 diabetic subjects classified according to the severity of retinopathy. Kilpatrick et al. (9) used multivariate analyses to show age, BMI, and HbA_1c to be associated with higher CRP concentrations in adults with type 1 diabetes. Although CRP was significantly associated with albumin excretion using univariate analysis, the association was lost using multiple regression.

As described previously (10), PgE₂ values for the type 1 diabetic subjects were consistently higher than for the control subjects. The PgE₂ levels, however, did not correlate with retinal grade. PgE₂ is believed to be involved in inflammation, apoptosis, angiogenesis, and increasing vascular permeability (11). The consistent elevation of PgE₂ in the diabetic population found in this study raises the possibility that this may be a factor in the etiology of microvascular complications.

In summary, this cross-sectional study of inflammatory markers and the retinal complications of type 1 diabetes showed that PgE₂ was the only marker significantly elevated in diabetic subjects compared with control subjects. However, PgE₂ was not significantly related to varying degrees of retinopathy among the diabetic subjects.

These findings were presented in part at the American Diabetes Association 64th annual meeting in Orlando, Florida, 4–8 June 2004.