Editors Selection IGR 22-2

Risk Factors: Predictive Risk Factors for Angle Closure

Shan Lin

Comment by Shan Lin on:

95114 Development of angle closure and associated risk factors: The Handan eye study, Zhang Y; Zhang Q; Thomas R et al., Acta Ophthalmologica, 2021; 0:

Find related abstracts

Zhang et al. have evaluated the development of angle closure and its risk factors among a rural Chinese population with baseline open angles, over the course of five years. Patients were those enrolled in the Handan Eye Study. This population study originally enrolled 6830 subjects. Data gathered included demographic information, clinical eye exam findings, refraction, A-scan ultrasound results, and visual fields. A subset of participants was selected for gonioscopy. There were 457 subjects with baseline open angles who met follow-up criteria at the five-year visit. A substantial number - 150 (33%) - had development of any form of primary angle-closure disease (PACD). The vast majority (94.7%) were primary angle-closure suspect (PACS) and the remaining (5.3%) were primary angle closure (PAC). There were no cases of primary angle-closure glaucoma (PACG). On univariate analysis, risk factors for PACD included older age, female gender, lower income, hypertension, larger spherical equivalent, smaller anterior chamber depth, larger lens thickness, and smaller axial length. On multivariate logistic regression, significant risk factors were shallower anterior chamber depth (ACD) (P = 0.003) and narrow angle width on gonioscopy (P < 0.001). However, analysis of the receiver operator curve using these two factors as a determinant of PACD did not have strong predictive ability (area under the curve [AUC] was 0.703).

There are only a small number of prospective studies that describe the risk of converting from open angles to angle closure, although a strength of the present study is the population- based design. The findings in the present study that shallower ACD and narrower angle width are risk factors are consistent with the prior studies. The lack of predictive ability of these two factors as a determinant of PACD suggests there are potential interactions with other factors, warranting further studies which may include more advanced imaging techniques such as 3-dimensional anterior segment optical coherence tomography.

Issue 22-2

Change Issue