Purchase this article with an account.
Qingying Jin, Thomas Ebert, Anke Tönjes, Kerstin Wirkner, Mengyu Wang, Dian Li, Neda Baniasadi, Cornelia Enzenbach, Matthias Blueher, Michael Stumvoll, Joachim Thiery, Markus Loeffler, Christoph Engel, Franziska G. Rauscher, Tobias Elze; Association between serum lipid parameters and retinal nerve fiber layer characteristics. Invest. Ophthalmol. Vis. Sci. 2019;60(9):5579.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Dyslipidaemia is often associated with the metabolic syndrome. Secondary Hyperlipidaemia can be caused by hypercaloric nutrition, obesity and diabetes mellitus, which, in turn, may be reflected in the nervous system. Here, we associate circumpapillary retinal nerve fiber (RNF) layer thickness (cpRNFLT) and fiber bundle peak locations (Fig.1) with lipids and lipoprotein serum levels in the age and gender stratified population-based Leipzig Research Centre for Civilization Diseases - LIFE Adult cohort.
10,677 eyes of 6,184 subjects (20-79 years) were analyzed as only subjects without lipid lowering drugs and with reliable (≥50 B-scan repetitions, quality ≥20 dB, ≤5% missing A-scans) spectral domain optical coherence tomography cpRNFLT scans (768 A-scans, diameter: 12°) were selected. Lipids, HDL- and LDL-cholesterol (HDL-C, LDL-C), apolipoprotein B (ApoB) and A1 (ApoA1) were analyzed in serum using standardized clinical chemical methods. To investigate associations of lipid levels with RNF geometry, the model with the optimal combination of age, radius and RNF geometry was determined by full Bayesian model comparison. To associate the spatial cpRNFLT profile, each of the 768 A-scans was separately used as regressor.
Elevated HDL-C and ApoA1 levels were associated with more temporal superior RNF peaks, elevated LDL-C, ApoB and triglyceride levels were associated with larger RNF interpeak angles (Fig.1B). Fig.2 shows the spatial cpRNFLT effects unexplained by age and RNF bundle geometry: Between 0% (triglycerides) and 49% (HDL-C) of the retinal locations were significantly related to serum lipid parameters (p<0.05 after adjustment for multiple comparisons; red in Fig.2). In subjects with higher HDL-C and ApoA1 levels, nasal areas are thicker and temporal areas thinner. For LDL-C and ApoB, the relationship is in opposite direction.
HDL-C, LDL-C, ApoB and ApoA1 were significantly associated with differences in RNF bundle geometry. Our results do not support the hypothesis that adverse lipid parameters are related to thinner RNFLT. Instead, specific thinner as well as thicker cpRNFLT areas were found which occur in opposite directions for HDL-C and LDL-C as well as the concomitant apolipoproteins. This suggests further and more specific eye anatomical associations with dyslipidaemia, which may help to relate specific ocular nerve fiber anatomy to unhealthy life style.
This abstract was presented at the 2019 ARVO Annual Meeting, held in Vancouver, Canada, April 28 - May 2, 2019.
This PDF is available to Subscribers Only