Abstract
Purpose::
To describe gender, ethnic and iris colour variability of retinal vessel calibre measurements in a population-based study of young children.
Methods::
The Sydney Childhood Eye Study (2003-4) examined 1740 6-year-old children from 34 randomly selected schools in Sydney. Eye examination included autorefraction, ocular biometry (Zeiss IOLMaster) and digital retinal photography. Retinal vessel calibre was measured from retinal photographs using a standardised computer-assisted method. Associations of retinal vascular calibers with gender, ethnicity and iris colour (blue, hazel/green, tan/brown, dark brown) were analysed.
Results::
Of 1740 children, 1612 had gradable retinal photographs. Mean retinal arteriolar (CRAE) and venular (CRVE) calibre was normally distributed, with means (standard deviation) of 163.2 (14.0) µm and 227.3 (18.3) µm, respectively. Boys had slightly narrower retinal arteriolar and venular calibre than girls (both p<0.0001), but the difference became non-significant after adjustment for age, ethnicity, birthweight, height, body mass index, mean arterial blood pressure, axial length. Both arteriolar and venular calibres, however were substantially narrower in Caucasian (mean CRAE 160.5 µm, mean CRVE 222.4 µm) than in East Asian children (mean CRAE 171.5 µm, mean CRVE 240,5 µm), p<0.0001 for both after multivariate adjustment. Among Caucasian children darker iris colour was associated with wider arteriolar and venular calibres. Mean CRAE increased from 157.5 µm in blue eyes to 169.2 µm in dark brown eyes (p for trend <0.0001). Mean CRVE increased from 218.4 µm in blue eyes to 230.0 µm in dark brown eyes (p for trend <0.0001). In subgroup analyses of children with dark brown eyes there was no significant difference in CRAE between Caucasian and East Asian children, but CRVE was still approximately 10 µm wider in East Asian than Caucasian children.
Conclusions::
Darker iris colour was associated with wider measured retinal vascular calibre. Iris colour, as a presumed surrogate measure of retinal pigmentation, appears to be an important confounder in measuring retinal vessel calibre. This may be due to a difference in contrast in eyes with darker retinal background. If confirmed, this effect may need to be taken into account when assessing the calibre of retinal vessel in subjects with different ethnic background.
Keywords: clinical (human) or epidemiologic studies: biostatistics/epidemiology methodology • retina • clinical (human) or epidemiologic studies: systems/equipment/techniques