Purchase this article with an account.
Barbara E K Klein, Kerri P Howard, Ronald Klein; Relationships of Serum Total Cholesterol (Total-C) and High Density Lipoprotein Cholesterol (HDLC) with Risk of Proliferative Diabetic Retinopathy (PDR) and Diabetic Macular Edema (DME) in Persons with Long Term Type 1 Diabetes (T1D). The Wisconsin Epidemiologic Study of Diabetic Retinopathy (WESDR). Invest. Ophthalmol. Vis. Sci. 2014;55(13):5350.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To examine the relationships of Total-C, HDLC and statin use to the prevalence and incidence of PDR and DME in persons with T1D participating in the WESDR.
Serum Total-C and HDLC were measured at the third examination (1990-92) and 4 subsequent follow-up examinations. Self-reported medication use was recorded at each examination. Retinopathy status was assessed by standardized retinal photographs obtained after pharmacologic pupil dilation and graded by a modified Early Treatment Diabetic Retinopathy Study protocol at each examination except the 2000-01 examination. Mixed modeling was used to examine prevalence across all examinations and incidence between examinations, accounting for unequal spacing between examinations and correlation between visits. All models were minimally adjusted for diabetes duration, glycated hemoglobin (A1c) and statin use.
In total, 1842 (885 women and 957 men) person-visits contributed data to these analyses for at least 1 outcome in this pilot study. The mean durations of diabetes were 21.8 years and 42.9 years, mean A1c was 9.3% and 7.7% and regular statin use was reported by 2.0% and 73.5% at the 1990-92 examination and most recent examination, respectively. The estimated incidence of PDR was 1.94% per person-year and for DME it was 1.71%. In modeling the associations, Total-C level was not associated with prevalent or incident PDR or DME. Higher levels of HDLC were associated with a significant protective effect for prevalent DME (odds ratio [OR] 0.62 for HDLC >60 vs. 40-60 mg/dL, P=0.02) and prevalent PDR (OR 0.77 for HDLC >60 vs. 40-60 mg/dL, P=0.02; OR 0.63 for HDLC 40-60 vs. ≤40 mg/dL, P=0.002) but was not associated with incident DME or incident PDR. Statin use was not related to prevalent or incident PDR or DME (data not shown).
HDLC but not Total-C was related to prevalent PDR and DME. Neither were related to incident PDR or DME in this older cohort with T1D. Statin use was not significantly related to either outcome. These data suggest that serum lipid levels and statin use have a limited role in the pathogenesis of these severe diabetic complications in persons with long-term T1D.
This PDF is available to Subscribers Only