Ocular Translation Due to Gravitational Acceleration

Purpose: We asked if globe position in the human orbit is influenced by gravitational head orientation while reclining. Methods: In 11 adult volunteers, surface coil magnetic resonance imaging (MRI) in 2mm thick axial and quasi-coronal planes MRI with a centered visual target was performed in supine position that aligns the cranial midline along the gravity vector, and repeated in right and left decubitus positions that orient gravity orthogonally. From coronal MRI we computed globe centroid position relative to the whole head. Results: Relative to supine position, in decubitus posture, the upper eye shifted 0.28±0.56mm (standard deviation) anteriorly, and 0.91±0.49mm medially (P<0.003 for both), without shifting superiorly. The lower eye shifted 0.88±0.59mm anteriorly (P<10-6) and 0.25±0.50mm superiorly (P=0.030), but did not shift medially. Vector magnitude of globe translation was significantly non-zero at 1.13±0.55mm for the upper and 1.16±0.59mm for the lower eye (P<10-7). Conclusions: Since lateral decubitus positioning eliminates direct gravitational effect on anteroposterior eye position, we conclude that in the supine position, gravity causes about 0.3-0.9mm enophthalmos, which is significant relative to intraorbital optic nerve length and would exert traction on the optic nerve of the downward positioned eye. The larger medial shift of the upper than lateral shift in the lower eye might reflect the stiffer medial than lateral rectus pulley suspension, or differential fluid shifts. Gravity thus has an appreciable effect on the human eye that should be considered when modeling the effect of eye rotation on the mechanical state of the optic nerve and other orbital tissues.