Purchase this article with an account.
F. Islam, T. Y. Wong, R. L. Mcintosh, J. J. Wang; Direct and Indirect Effects of Metabolic Risk Factors for Retinopathy in Persons With and Without Diabetes. Invest. Ophthalmol. Vis. Sci. 2009;50(13):3904.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
The relative contribution and importance of different metabolic risk factors to retinopathy in persons with and without diabetes are not fully understood. We aimed to quantify the direct and indirect associations between metabolic syndrome components and retinopathy.
The Singapore Malay Eye Study examined 3280 (78.7% response) of which 3264 participants had measured metabolic syndrome components: systolic blood pressure (SBP), body mass index (BMI), casual glucose and total and high density lipoprotein (HDL) cholesterol. Retinopathy was assessed from retinal photographs following the modified Airlie House Classification, and defined as level ≥12. Diabetes (n=763, 23% of the participants) was defined as casual glucose ≥11.1 mmol/L or self-reported physician diagnosed diabetes. Direct and indirect associations of retinopathy with metabolic-related variables were assessed using path analysis. Path coefficients (PC)) were quantified using Linear Structural Equation models.
Younger age was directly associated with retinopathy prevalence (PC=-0.10) in persons with diabetes but not in those without diabetes (PC=0.01). Increasing age was also indirectly associated with retinopathy via increasing SBP, with similar magnitude of association in both groups with (PC =0.065) and without diabetes (PC=0.063). Decreased BMI was directly associated with retinopathy in those with diabetes (PC= -0.11) but increased BMI directly associated with retinopathy in those without diabetes (PC= +0.10). Higher SBP and glucose levels were more strongly associated with retinopathy in persons with diabetes (PC=0.26 and 0.17, respectively) than in those without diabetes (PC=0.16 and 0.08, respectively). Total and HDL cholesterol were directly associated with retinopathy in persons without diabetes (PC=0.08 and -0.08, respectively) but not in those with diabetes (corresponding PC=0.01 and 0.03). Metabolic-related variables explained 11% of variations in retinopathy prevalence in persons with diabetes but only 6% of the variation in those without diabetes.
We found that the direct associations of retinopathy with age, BMI, serum lipids, SBP and glucose were different in persons with and without diabetes, findings suggesting likely different aetiologies for retinopathy in persons with and without diabetes.
This PDF is available to Subscribers Only