Central vs. peripheral blood pressure components as determinants of retinal microvessel diameters

Background: We assessed association of retinal arteriolar and venular diameters with central and peripheral blood pressure (BP). Methods: We post-processed retinal photographs from 514 participants randomly recruited from a Flemish population (mean age, 50.6 years; 50.8% women), using IVAN software to generate retinal arteriolar (CRAE) and venular (CRVE) equivalents. We measured peripheral BP by mercury sphygmomanometry and central B. P by tonometry at the carotid artery. We applied multivariable-adjusted regression analysis. Results: For peripheral vs. central BP (mmHg) average levels were 126.6 vs. 122.1 systolic and 79.4 vs. 79.6 diastolic, and 95.1 vs. 97.9 and 47.2 vs. 42.5 for mean and pulse pressure, respectively. CRAE and CRVE averaged 153μm and 219μm. Effect sizes (μm) for CRAE for 1-SD increase in peripheral vs. central BP were -3.77 vs. -3.52 systolic, -3.16 vs. -3.13 diastolic, -3.84 vs. -3.64 for mean BP, and -2.07 vs. -1.83 for pulse pressure (p≤0.006). Models that included two BP components demonstrated that CRAE decreased (p≤0.035) with systolic (peripheral vs. central, -2.87 vs. -2.40) and diastolic (-1.58 vs. -1.80) BP. CRAE decreased with mean BP (-3.53 vs. -3.53; p<0.0001), but not with pulse pressure (p≥0.19). CRVE was not related to any peripheral or central BP component (p≥0.062). All CRAE regression slopes on corresponding peripheral and central BP components were similar (p≥0.28). Conclusion: Higher systolic, diastolic and mean BPs were associated with smaller CRAE, regardless of whether BP was measured centrally or peripherally. Central BP does not refine the inverse association of CRAE and CRVE with peripheral BP.