Methods

This prospective, observational study was conducted at the Department of Ophthalmology, Korea University College of Medicine. This study included 20 patients who had been diagnosed with bilateral OAG and treated only with latanoprost 0.001% ophthalmic solution in both eyes at the Korea University Ansan Hospital. Study approval was obtained from the Institutional Review Board of the Korea University Ansan Hospital. The study was conducted in adherence with the tenets of the World Medical Association’s Declaration of Helsinki. Written informed consent was obtained from each subject prior to participating in the study.

Each patient underwent comprehensive ophthalmic examinations of both eyes, including best-corrected visual acuity, refractive error, slit-lamp biomicroscopy, axial length (AL) with IOL Master (Carl Zeiss Meditec, Jena, Germany), central cornea thickness (CCT) with a noncontact specular microscope (sp-3000p; Topcon, Tokyo, Japan), gonioscopy, Goldmann applanation tonometry, Humphrey Swedish interactive threshold algorithm (SITA) 30-2 testing (Zeiss-Humphrey, San Leandro, California, USA), and dilated stereoscopic funduscopy.

To meet the criteria for a diagnosis of OAG, patients had to have a glaucomatous optic disc change, a reproducible glaucomatous visual field defect on the SITA 30-2 perimetry, and normal-appearing open angles on gonioscopy. Glaucomatous optic disc changes were defined as focal or diffuse neuroretinal rim thinning, localized notching, or retinal nerve fiber layer defects with corresponding visual field defects. Glaucomatous visual field defects were defined by 2 of the following 3 criteria: the presence of a cluster of 3 points on a pattern deviation probability plot with P < .05, one of which had P < .01; a pattern standard deviation with P < .05; or glaucoma hemifield test results outside normal limits.

Patients with concomitant ocular disease, best-corrected visual acuity worse than 20/40, refractive error of more than 9 diopters spherical equivalent (SE) or 2 diopters of astigmatism, use of contact lenses, a history of previous intraocular surgery, or a history of spine disease and its related symptoms were excluded from this study. Patients with advanced glaucoma (mean deviation ≤−12 dB) were also excluded.

To minimize the effect of diurnal variation, IOP measurements were performed by a single observer between 6:00 PM and 8:00 PM. IOP was measured using a single ICare rebound tonometer (ICare Pro; Tiolat Oy, Helsinki, Finland) in both eyes for each position. The right eye was always examined first. The subject was asked to look straight ahead and the IOP was measured by touching the transducer to the center of the subject’s cornea. ICare Pro has a built-in inclination sensor to measure the supine posture IOP; an arrow appears in the display indicating the vertical position of the tonometer. Three consecutive sets of measurements, with 6 measurements for each set, were made. The average of each set was made automatically, and the averaged values were used for the statistical analyses. To allocate a randomized sequence of IOP measurements in the lateral decubitus posture, each participant was asked to pick up 1 card, where the different positions of head and body were written on the back, one after another until all the 6 cards were open. Thus, each participant was informed of the sequence of changing body and head positions.

First, each participant was asked to sit on a backrest chair in a quiet room under dim light conditions. Five minutes later, IOP was measured in the sitting posture; and IOP was measured in the supine posture in another 5 minutes. Then, the subject was asked to maintain the right or left lateral decubitus posture on a bed with a soft pillow of different heights underneath the head, in the randomized sequence predetermined in the aforementioned manner. IOP was measured 5 minutes after assuming each posture. In the right and left lateral decubitus postures, IOP measurements were taken with 3 different head positions ( Figure 1 ): lateral decubitus with the head 30 degrees higher than the center of the thoracic vertebra; lateral decubitus with the head parallel to the center of the thoracic vertebra; lateral decubitus with the head 30 degrees lower than the center of the thoracic vertebra. Multiple pillows (2 or 3) were used for the 30 degrees head-up position (right and left lateral decubitus posture with high head position), whereas a smaller pillow (of lower height) or no pillow was used for the 30 degrees down position (right and left lateral decubitus with low head position). The degree of a lateral-ward bend of the neck was determined by assessment of the angle between the midlines of the face and the body. The subject was asked to maintain a straight gaze during assumption of the lateral decubitus posture so that his or her dependent eye was not compressed against the pillow.

Illustration of various head positions in lateral decubitus posture: (Left) Lateral decubitus with the head 30 degrees lower than the center of the thoracic vertebra; (Middle) lateral decubitus with the head parallel to the center of the thoracic vertebra; (Right) lateral decubitus with the head 30 degrees higher than the center of the thoracic vertebra.

For all patients, the better eye and the worse eye were defined based on the mean deviation value determined by the Humphrey visual field testing. The height and weight were recorded by patients’ self-report.

Results

Among 20 patients, 9 patients were male and the overall mean age was 52.6 ± 8.0 years. The mean height was 165.1 ± 7.6 cm and the mean weight was 61.1 ± 8.6 kg. There was no significant difference between right and left eyes in refractive error, mean AL, CCT, mean IOPs measured by Goldmann applanation tonometry, or mean deviation (all P > .05) ( Table 1 ). No significant inter-eye difference was found in IOPs measured by ICare Pro tonometry in the sitting and supine posture. Among the 20 patients, 8 patients had the worse visual field in the right eye and 12 patients had the worse visual field in the left eye. There was no significant difference between better and worse eyes in refractive error, mean AL, CCT, or mean IOPs measured by Goldmann applanation tonometer, but there was a significant difference for the mean deviation ( P < .001). No significant difference in IOPs was found between the better and worse eyes in the sitting and supine posture (all P > .05) ( Table 1 ).

Table 2 lists the IOP measurements obtained in the lateral decubitus postures with different head positions. The IOP of the dependent eye was significantly higher than that of the nondependent eye in the lateral decubitus postures with different head positions (all P ≤ .05).

Figure 2 shows the differences of IOP alterations between supine and lateral decubitus postures with different head positions. Compared with that of the ipsilateral eye in the supine posture, the IOP of the dependent eye was higher in lateral decubitus posture with head at 30 degrees lower position (all P < .003). The IOP of the dependent eye in right, left, or worse eye–dependent lateral decubitus posture with neutral head position was higher than that of the ipsilateral eye in the supine posture ( P = .003, P = .002, and P = .001, respectively), whereas such difference was not noted in the better eye–dependent lateral decubitus posture ( P = .011). Figure 3 shows comparisons of IOPs of the dependent eye in the lateral decubitus postures with different head positions. Among those measured in the lateral decubitus posture, the IOP of the dependent eye with 30 degrees lower head position was higher than that measured in the neutral or 30 degrees higher head position (all P < .001). Detailed P values and IOP differences are listed in Supplemental Table 1 (available at AJO.com ).

Intraocular pressure (IOP) differences in dependent eye between supine and lateral decubitus postures with different head positions. The IOP difference was determined by subtracting the IOP measured in the supine posture from that measured in the lateral decubitus posture. The squares and the lower and upper bars indicate 25th–75th percentile, minimum value, and maximum value, respectively. * Indicates P < .003 compared with the IOP measured in the supine posture by Wilcoxon signed rank test with adjustment for multiple comparisons by Bonferroni correction.

Comparisons of intraocular pressures of dependent eyes in the lateral decubitus postures with different head positions. The squares and the lower and upper bars indicate 25th–75th percentile, minimum value, and maximum value, respectively. * Indicates P < .003 by Wilcoxon signed rank test with adjustment for multiple comparisons by Bonferroni correction.

Figure 4 is the scatterplot of the individual changes in IOP in the supine and the lateral decubitus postures with different head positions compared with the sitting IOP. The magnitude of IOP elevations induced by recumbent postures varied among the OAG patients. When compared between the IOP of the dependent eye measured in the lateral decubitus postures with the head lowered and that measured in the supine position, 28 of 40 eyes (70%) showed IOP elevations of ≥3 mm Hg; 20 eyes (50%), ≥4 mm Hg; 11 eyes (28%), ≥5 mm Hg; and 4 eyes (10%), ≥6 mm Hg. Table 3 summarizes the magnitudes of changes in IOP induced by alterations of body or head position. No significant difference in the amount of posture-induced IOP changes was found between the better and worse eyes (all P > .05).

Scatterplot of individual changes in intraocular pressure (IOP) of the dependent eye compared with the IOP measured in the sitting posture.

Univariate and multivariate linear regression analysis ( Supplemental Tables 2 and 3 , available at AJO.com ) between the recumbent posture–induced IOP elevations and the demographic data found that AL was positively associated with the IOP elevations induced by neutral (dependent eye, B = 0.969; nondependent eye, B = 0.894; both P < .05) and high (dependent eye, B = 1.387; nondependent eye, B = 0.523; both P < .05) head position in lateral decubitus posture. The age correlated negatively with IOP elevations induced by lateral decubitus posture with low head position (dependent eye, B = −0.086, P = .037). The other factors including CCT, SE, height, and weight showed no association with the recumbent posture–induced IOP elevations.