We are encouraged by the interest shown by Nowroozzadeh in our article describing the relationship of retinal vascular caliber with optic disc and macular structure. He has offered additional possible explanations for our findings that retinal vessel caliber varies systematically with morphologic differences in the optic nerve head, retina, and macula. We would like to offer our views on some of his observations.

He proposes that variations in retinal vascular caliber that we observed may be explained by tensile stress on the vessel wall. The magnitude of this stress is described by the equation for tensile stress in a sphere σ = pr /2 t , where σ refers to the tensile stress, p refers to the intramural pressure, and t refers to the wall thickness. For the sake of veracity, a blood vessel is closer in form to a cylinder than a sphere, in which case the appropriate formula for the circumferential stress on the wall, although qualitatively similar, could be defined as σ = pr / t . This formula also is applicable only to thin-walled vessels, which, by definition, adhere to the rule of r / t ≥ 10. This criterion applies fairly well to thin-walled venules, but is probably inappropriate for describing the behavior of arteries or arterioles. This could explain in part the generally stronger associations between venular caliber and optic disc and retinal nerve fiber layer (RNFL) parameters we reported. However, the suggestion that “any factors that increase the tensile stress of these structures can increase the mechanical stress on the vessels and consequently decrease the blood flow and vascular diameter” in the letter is difficult to accept for a number of reasons. That increased tensile stress reduces vascular diameter implies a pathologic, and probably chronic, response to increased tensile strength. Such a response is unlikely to be applicable to a healthy pediatric population such as ours or that in the Sydney Childhood Eye Study. Furthermore, any changes in vascular diameter are likely to be reflected in changes in the luminal diameter, rather than the external diameter that we measured. The simple mechanical formula quoted also does not take into account the biomechanical properties of the vessel wall. Corneal hysteresis (CH), a measure of the biomechanical properties of the cornea and possibly of the corneoscleral shell as well, has been shown to correlate with lamina cribrosa compliance, and we also have demonstrated correlations of corneal hysteresis with retinal vascular caliber in our cohort. We concurrently have reported that corneal hysteresis is not correlated with optic disc or RNFL parameters. Thus, the current evidence on the role of biomechanical factors in determining retinal vascular caliber suggests that a purely biomechanical explanation cannot adequately account for the associations between retinal vascular caliber and optic disc or RNFL measurements. As such, other mechanisms may play a role. For example, retinal vascular caliber may affect optic and RNFL measurements through an effect on nutritional support on these parameters, or retinal vessel caliber simply may vary systematically and physiologically with morphologic differences in the optic nerve head.