Ocular and General Determinants of Intraocular Pressure: The Two-Continent Eye Study.

Purpose

The purpose of this study was to assess associations of intraocular pressure (IOP) readings.

Methods

This was a population-based studies conducted in Russia, China, and India. The project included the population-based investigations of the Beijing Eye Study (BES; n = 3139 participants, age = 40+ years), Ural Eye and Medical Study (UEMS; n = 5514, age = 40+ years), Ural Very Old Study (UVOS; n = 522, age = 85+ years), and Ural Children Eye Study (UCES; n = 4294, age = 6+ years), and Central India Eye and Medical Study (CIEMS; n = 4508, age = 30+ years). Pneumo-tonometry or applanation tonometry were performed.

Results

The study included 35,199 eyes (17,977 individuals, mean age = 46.9 ± 23.2 years, range = 6-100 years, axial length = 23.2 ± 1.0 mm, range = 18.22-34.20 mm). In the test group (BES, UVOS, and UCES), higher IOP-readings were associated (r2 = 0.37) with younger age, female sex, rural habitation region, higher prevalence of diabetes mellitus, higher systolic blood pressure, thicker central corneal thickness (CCT) and smaller corneal curvature radius (i.e. steeper cornea), longer axial length, lower prevalence of cataract surgery, and higher prevalence of pseudoexfoliation. IOP values, corrected for these associations, correlated with the IOP readings in the external validation groups of UEMS (r2 = 0.85) and CIEMS (r2 = 0.80). In Bland-Altman analysis, the mean bias was 0.80 millimeters of mercury (mm Hg; 95% confidence interval [CI] = 0.76-0.84 or 5.8%, 95% CI = 5.5-6.1) and 1.51 mm Hg (95% CI = 1.47-1.56 or 10.2%, 95% CI = 9.9-10.5), in the UEMS and CIEMS, respectively. The corrected IOP compared to the measured IOP was 8.1 mm Hg lower and 10.0 mm Hg higher, respectively, in 2 clinical examples. A real IOP value of 21 mm Hg corresponded to a measured IOP readings ranging from 29.1 mm Hg to 11.0 mm Hg.

Conclusions

Among numerous ocular and systemic factors influencing IOP readings (with a relative low correlation coefficient), corneal curvature radius was one of the main ocular determinants. Correction of measured IOP readings led to a large interindividual variation in the upper limit of the "normal" IOP range.