The report by Foster and Johnson in this issue of theBJO (p 1277) points to dramatic developments in our knowledge of glaucoma worldwide. Firstly, they continue a trend endorsed by an international committee to use the term “glaucoma” only for people who have suffered injury to the optic nerve as judged by visual field abnormality, combined with enlargement of the cup/disc ratio outside statistical limits for the population studied. This approach is intended to differentiate between those with a true disease and those suspects who are at increased risk—for example, those with higher than normal eye pressure or those with narrow angles. In both the developed and the developing world, we must concentrate limited medical resources on those at greatest risk for loss of function and decreased quality of life. In addition, we must avoid treatment of those who would never have lost vision in their lives, and who may be more threatened by treatment than by their natural course.

With respect to angle closure glaucoma (ACG) specifically, Foster and Johnson summarise recent data that build on previous population research. The pioneering work of Alsbirk,1Hu,2 Shiose,3 and Salmon4 had reinforced the suggestions that people derived from north China, including native peoples of North America and Greenland, had a high prevalence of ACG. The studies among urban Chinese in Singapore and Mongolian rural villagers confirm that ACG prevalence is higher among these people than among any other populations in the world. Not only do Chinese people represent one billion citizens of China proper, but they are a substantial proportion of those in other south east Asian countries, as well as large minorities in developed countries worldwide. At the same time, we must not assume that there is a uniform rate of ACG among all those who may be referred to as Chinese, since diversity is clearly present in so large a group.

In fact, we might question the assumptions that ACG prevalence in Singapore may be applied to the urban People's Republic of China generally and that Mongolian prevalence of ACG may be generalised to all rural Chinese. Clearly, Foster and Johnson recognise that their estimates are an important, but only an initial, step in moving towards more and more accurate estimates of the prevalence of glaucoma in a diverse population.

The ethnic diversity of ACG is further underscored by other recent population based surveys. It had been taught in many centres that African derived people had little ACG and that ACG was very common among Hispanic people. Yet, two surveys suggest that the prevalence of ACG is similar among some African5 and Hispanic6 populations as among Europeans—that is, about one fifth the rate of open angle glaucoma (OAG) in these populations.

In a meta-analysis of glaucoma prevalence data in 1996,7we suggested that the rate of ACG in Chinese people might be three times higher than that for OAG. The Foster/Johnson data analysis concludes that this may nearly be true for Mongolians, but that the relative prevalence for urban Chinese people is closer to 1:1 (actually 2:3). Each study finds that OAG is equally prevalent among Chinese and European people (here African and Hispanic people clearly differ in having more OAG). Worldwide glaucoma figures may need revision in the light of new information.

ACG therefore deserves greater emphasis as it comprises a higher proportion of all those with glaucoma. The lack of past attention may derive from its infrequency in developed countries. This attitude deserves immediate change for two important reasons. ACG blinds many more people than does OAG, and it may be more easily treated. The morbidity estimates for ACG were more clearly defined by Foster and Johnson, indicating that it blinds 10 times more people than OAG in absolute terms, even though the estimated numbers of those with ACG in China are only 30% higher than those with OAG.

What can be done to prevent visual loss from ACG? Firstly, we must develop screening methods that identify those likely to lose vision. Simply identifying narrow angles isn't enough, since the proportion of those with occludable angles is nearly 10 times the proportion of those with damage to their optic nerve from glaucoma. We need innovative and inexpensive screening tools which define features of the anatomical and particularly the physiological attributes of eyes that develop ACG. Simple measures of anatomical size (chamber, axial length, etc) have failed to differentiate effectively those who will get ACG. There are not 10 times more small eyes among Chinese people on a population basis compared to Europeans; hence, there must be, as yet, unmeasured features of their eyes that further increase risk for the disease among smaller Chinese eyes compared to smaller European eyes.

Secondly, iridotomy is the present standard approach to initial treatment of ACG. But, it is subject to abuse if too many eyes are treated which would never develop the disease. The improvements in screening should lead to clinical trials in which longitudinal follow up determines the relative value of iridotomy in various populations based on estimated risk at baseline. At the other end of the spectrum, those who have already suffered severe attacks may often require more than iridotomy, including trabeculectomy surgery.8 For these people, research should identify how frequently iridotomy eliminates further need for treatment and who might best be provided with more extensive treatment at diagnosis.

While there has been much emphasis on separate diagnostic testing for ACG, the work of Foster and Johnson, as well as others, points out that only 20–25% of those with ACG develop acute attacks, while the remainder are found to have and continue to have a chronic, asymptomatic disorder. Albeit, starting from a different mechanism, ACG of the more chronic form, and OAG are, therefore, both identifiable in communities by a similar diagnostic approach involving evaluation of the optic nerve and/or visual field. This, to some degree, simplifies the design of public health measures that seek to perform initial identification of all those with visual field loss from glaucoma.

OAG remains an important problem and the present emphasis on ACG is appropriate only because of past inattention. Each glaucoma subtype, OAG or ACG, is a greater cause of visual morbidityindividually than any other ocular disease, except cataract and trachoma. Our resources should be directed in ways that prevent these avoidable conditions from adding to the world's blindness burden.