Central corneal thickness in a Jordanian population and its association with different types of Glaucoma: cross-sectional study

Glaucoma is the second leading cause of blindness in the world, after cataracts, and is the leading cause of blindness among African-Americans [1]. It is generally classified into open-angle and closed-angle glaucoma, and both can be either primary or secondary. It is estimated that the number of people with primary open-angle glaucoma (POAG) worldwide will reach 58.6 million by 2020, while 21 million will be affected by primary angle-closure glaucoma (PACG) [1]. Of those, 5.9 and 5.3 million will be blind from irreversible optic nerve damage associated with POAG and PACG, respectively [1]. This means that early detection and identification of risk factors are key elements in controlling the disease and preventing its progression. Secondary open-angle glaucoma is another entity with diverse types, but Pseudoexfoliation glaucoma (PXFG) is currently the leading cause of secondary open-angle glaucoma with a prevalence reaching 40% in patients over the age of 80 [2], and is highly dependent on race and ethnicity [3].

Evidence in the recent literature has shown the importance of central corneal thickness (CCT) in relation to several ocular conditions. Most notably, thinner CCT has been identified as a risk factor for open-angle glaucoma [4]. Goldmann and Schmidt discussed the association between the corneal thickness and the intra-ocular pressure (IOP) in their publication in 1957 [5]. Fourteen years later, Hansen and Ehlers demonstrated the presence of a positive linear correlation between CCT and IOP [6]. More recently, a meta-analysis of worldwide CCT literature proposed a correction factor of 2.5 mmHg for each 50 μm change in CCT [7].

CCT can be very variable between different ethnic groups, as shown in studies comparing mean CCT in both normal and glaucomatous eyes between Caucasian, Hispanic, African American, and Asians, where the thinnest corneas being more prevalent among African American ethnicity [8, 9]. Other studies also showed significant variation in CCT between ethnic sub-groups, as in a study that included different Asian populations; Chinese, Japanese, Korean, Filipino, Pacific Islander, and South Asian, and showed that the thinnest corneas are found in South Asian populations [10]. Moreover, several studies have shown that CCT varies among individuals with different types of glaucoma (POAG, PACG, and PXFG) and normal eyes, as in Bechmann et al. study that showed thinner corneas in patients with PXFG and POAG compared with patients with PACG and normal eyes [11]. In Jordanian population; an Arabian, Middle-eastern ethnicity, previous studies showed that glaucoma is one of the leading causes of blindness [12], and at the same time, the level of knowledge about glaucoma among Jordanians is poor (around 75% have either no or low level of knowledge) compared to other parts of the world [13], leading to delayed diagnosis of this silent disease. Ophthalmologists in Jordan usually depend on studies that were done on other ethnicities upon treating their glaucoma patients, due to lack of studies on the characteristics of glaucoma among Jordanians. Given the importance of early detection of risk factors of glaucoma, and as corneal thickness is one of the main risk factors for glaucoma and is the one associated with ethnicity, we aim to assess the corneal thickness in different types of glaucoma and in normal controls in a Jordanian population. We further aim to explore the relationship between CCT and the severity of glaucomatous optic neuropathy in different types of glaucoma.

We included a total of 68 Jordanian patients in this study, with a mean age of 65.9 years (SD = 8.9). There were 42 (62%) men and 26 (38%) women. From the sample included, 119 eyes met our inclusion criteria and were included in our study, they were 61 (51.3%) right eyes and 58 (48.7%) left eyes (51 bilateral and 17 unilateral eyes). Eyes with glaucoma were categorized according to type of glaucoma into: POAG (54 eyes), PXFG (31 eyes) and PACG (34 eyes), as shown in (Table 1). We also included 29 control participants with a mean age of 55 years. They were 14 men and 15 women and a total of 57 eyes.

Upon analyzing the correlation between gender discrepancy and type of glaucoma developed, we found a significant difference (p < 0.001) as men were more likely to develop POAG (55.4%) compared to 28.9% for women. However women were more likely to develop PACG (44.4%) compared to only 18.9% for men. No significant difference was found between gender discrepancy and the development of PXFG. We found no statistically significant difference between both genders regarding other parameters. Mean values of different glaucoma-related parameters for both genders are detailed in (Table 2).

Regarding central corneal thickness (CCT), our results showed the following mean measurements: 538.31 μm (SD = 36.30) in eyes with POAG, 544.45 μm (SD = 28.57) in eyes with PXFG, 506.91 μm (SD = 34.55) in eyes with PACG, and 549.63 μm (SD = 42.9) in the control group. First, we studied the difference between CCT of each glaucoma type and the control group, we found that subjects with PACG had significantly thinner corneas than the control group (p = 0.021). No statistically significant difference was found between CCT of control group and any of POAG or PXFG groups (p = .702). Upon comparing mean CCT between glaucoma groups, we found a significant difference between the groups in general (p < 0.001), so we performed a post-hoc analysis to find the group differences as shown in (Table 3).

Upon studying the relation between CCT and parameters of optic nerve damage (CDR, MD, and RNFL) in patients with glaucoma, we found that CCT is negatively and significantly correlated with CDR (p = 0.012, r = − 0.231), so that the thinner the cornea, the higher the CDR. We also found a significantly positive correlation between CCT and MD (p < 0.001, r = 0.327), so that the thinner the cornea the lower the MD (more advanced visual field loss). Finally, we found a positive correlation between CCT and RNFL (p = .007, r = .283) meaning that the thinner the cornea, the thinner the RNFL. No significant correlation was found between CCT and either IOP or pattern standard deviation (PSD).

We studied factors predicting visual field mean deviation (MD), and found that it is significantly associated with both CDR (p < 0.001) and RNFL (p = 0.003). Upon analyzing the relation between type of glaucoma and our measured parameters, we found several significant relations; our results are summarized in (Table 2). No statistical significance was found between either the age or laterality of the involved eye (left or right) on one hand and type of glaucoma or other measured parameters on the other hand.

This study was done on a Jordanian population, a poorly studied population regarding glaucoma and its risk factors. The mean central corneal thicknesses (CCT) of normal adults and patients with 3 different types of glaucoma was measured by corneal topography. We also established that CCT in patients with PACG is significantly thinner than in other glaucoma types and in the control group. As visual field defect is an important outcome in glaucoma, we found that both cup to disc ratio (CDR) and retinal nerve fiber layer (RNFL) thickness, as measured by Optical Coherence Tomography, are the main factors associated with lower mean deviations on visual field testing. Gender discrepancy in glaucoma type preferences found to be in concordance with some previous reports where men are more likely to develop POAG [14], and women are at higher risk for PACG, likely due to anatomical predisposition [15]. (Fig. 1) shows images for a patient with advanced chronic angle closure glaucoma and with significantly thin central corneal thickness, where the visual field test demonstrating a superior arcuate scotoma splitting fixation and ocular coherence tomography of the retinal nerve fiber layer thickness (OCT RNFL) shows significant thinning.

Previous studies, that were done on western populations, aimed to find the relation between CCT and other parameters, including visual field defects, have mainly included POAG eyes, and they found significantly more advanced visual field defects for lower CCT compared to thicker ones [16, 17]. Here, we studied different types of glaucoma patients and control participants from a Jordanian population, and we found similar correlations where patients with lower CCT had significantly more advanced visual field defects, and thinner RNFL. We also found that patients with PACG had the lowest CCT compared to POAG, PXFG, and controls. An Iranian study, that included both POAG and PACG patients, found that POAG patients have lower CCT compared to their PACG counterparts [18]. Regarding PXFG, a Caucasian study found a significantly lower CCT in PXFG patients compared to POAG patients [19], this emphasizes the profound difference in CCT in different ethnic groups and the importance of CCT as a risk factor for progression of glaucoma.

The strong relation between CCT and ethnicity is well-studied. A large study that was done on a Chinese population to study the relation between CCT and around 20 variables found that ethnicity is one of the important determinants of CCT [20]. A worldwide meta-analysis found that the average CCT for normal eyes derived from various racial groups, regardless of the device used, was 544 μm [21], a value that is thinner than that for the Jordanian population we studied (550 μm). Comparing CCT values of our Jordanian population with Caucasian ethnicity, the most represented ethnicity in ethnicity-comparing studies [22], Caucasian ethnicity mean CCT was similar to Jordanian’s (550 ± 30.5 μm) [23]. On the other hand, African ethnicity has a lower CCT than that for Jordanians (514.77 ± 31.86) [23]. It should be noted that the technique used to measure CCT can affect the results [24], and although the results from commonly used devices mostly correlate, the direct comparison between them might not be accurate [22], so that we will only include studies that used similar techniques for direct value comparison purposes. Regarding CCT difference between different glaucoma types, only PACG eyes have thinner corneas compared to POAG, PXFG, and control eyes, This comes in accordance with what a recent study suggested, where no difference in corneal thickness between POAG, PXFG, and control eyes was found [25].

Other reports generally have results that are comparable to ours, except for a study by Tolesa and Gessesse that was done on Ethiopian patients [26], and found thicker CCT in patients with POAG (520 ± 38.95) compared to PXFG patients (507 ± 35).

Several previous studies have reported the association between larger CDR and glaucomatous progression [27]. Here, we found that patients with larger CDR have worse mean deviation (MD) compared to those with smaller ratios. In Ocular Hypertension Treatment Study [28], they found that cup to disc ratio is also a predictor of the development of POAG in normal subjects. A recent study found that for every 0. 1-unit increase in the CDR, there is an increase in the decay rate of visual field by 23% [29]. We also found that the largest CDR is found in PACG patients, followed by POAG, and finally PXFG.

Similarly, we found that RNFL thickness measured by Optical Coherence Tomography (OCT) is also a good indicator for the severity of visual field loss represented by the mean deviation (MD). In a 12-year study, the risk estimate for moderate and severe visual field defects found to suggest a seven to eight times greater likelihood for future field loss for eyes with initial atrophy [30]. Where previous reports found an association between IOP and MD [30], we think that this effect is indirect through affecting the CDR, as the relation between IOP and MD did not reach significance when we corrected for CDR, probably because these were treated IOP readings.

Several limitations of this study exist. We advice future studies to include only one eye for each included subject (independence of observation) to overcome correlated observation issue as explained in a previous study, although it can be corrected using certain statistical tests [31]. Moreover, a longitudinal approach to follow Middle Eastern patients would provide a better insight on the outcome of different CCT. Although this is a multicenter study, future studies from other Arabian countries should consider including larger sample size to confirm the generalizability of our results.

We found that PACG patients have the thinnest CCT compared to other types of glaucoma, namely POAG and PXFG. We also demonstrated that glaucomatous eyes with thinner corneas will probably have more advanced glaucomatous optic neuropathy. This emphasizes the importance of making CCT measurement a part of standard care in all glaucoma patients since it can predict patients with worse prognosis, and therefore can indicate the need for more aggressive management and earlier diagnosis. Finally, we showed that cup to disc ratio, retinal nerve fiber layer thickness and mean deviation on visual field testing usually are in concordance regarding the severity of glaucoma.