Our study evaluated the ocular biometric data characteristics in Shanghai urban and rural populations by using a partial coherence laser interferometry (PCI). PCI is widely used in clinical work due to its highly accurate and reproducible measurements. The accuracy of the biometric data leads to accurate calculations of IOL power. Previous studies have investigated the characteristics of biometric data of different race populations, such as southern Chinese , Latin American , Malay , Indian  and Western populations . To our knowledge, this is the first biometry study that focuses on urban and rural cataract patients in China.
We demonstrated that the biometry parameter, such as AL, ACD, K and CA, was distributed non-normally in the general investigated population (all participants), in the different location populations (urban versus rural) and in the gender populations (male versus female), respectively.
The AL data in our study was positively skewed and showed significant kurtosis, as reported in the Reykjavik Eye Study  and the studies by Fotedar  and Chen . In the combination/general population, mean AL was 23.98 ± 2.09 mm, which was similar to the results of the study in southern China by Cui et al. (24.07 ± 2.14mm) . But the mean AL in urban areas was about 0.8 mm longer than that in rural areas (24.35 ± 2.34 mm versus 23.58 ± 1.70 mm, respectively). Mean AL in Shanghai was similar to Yu’s study (24.38 ± 2.47 mm) from central China  and Huang’s study (24.32 ± 2.42 mm) from western China, but was clearly longer than other AL studies using PCI [8-10], which investigated other ethnicities, such as European (23.43 ± 1.51 mm) , Latin American (23.8mm) , Malay (23.55mm)  and Mongolian (23.13 ± 1.15 mm) .
Yu  and Huang  had investigated the inhabitants in Wuhan and Chengdu urban areas, respectively, which are similar in latitude to Shanghai. All of these cities belong to the reaches of the Yangtze River. For other Chinese inhabitants in urban southern China, on the reaches of Zhu River, and other Asian countries, mean AL is shorter than for those living in Shanghai. Obviously, AL parameters were related to ethnicity and environment. There was no previous study that involved the local urban and rural populations. We found the AL of the urban population to be longer than that of the rural population. We propose that the urban population may be exposed to more near work and become more myopic with a longer AL than the rural population. As for the gender factor, we found that men had a longer AL than women, which is consistent with findings from previous studies [6, 12, 15]. These disparities reflect the different physical conditions between the sexes.
We found that the ACD was deeper in a younger, male, urban population than it was in an older, female, rural population. The first two findings were similar to other studies [4, 12, 16]. The trend of ACD variance with sex was attributed to differences of male and female anatomy, particularly height. The depth of the ACD often decreased with an increase in age-related lens thickness, which can be observed in any gender or race [4, 12, 16]. In comparison to European (3.11 ± 0.43mm) , Austrian (3.10mm)  and Latin American studies (3.41± 0.35 mm) , the mean ACD depth is shallower in Shanghai and Jinshan. Previous studies found that elderly Asian women were more likely to develop acute angle-closure glaucoma due to factors of sex and race, which is similar to the results of Chinese population studies from Cui (3.01 ± 0.57mm)  and Chen (3.03 mm) . However, other investigators, like Huang (3.08 ± 0.47 mm) and Yu (3.15 ± 0.48mm), found a deeper ACD than we did. Long axial length myopia is often consistent with a deeper ACD, while short axial length hyperopia is consistent with a shallower ACD. The percentage of eyes with a long axial length in the population will influence the mean ACD; therefore, when comparing the mean ACD in different Chinese studies, researchers should consider adjusting for the effects of myopia.
As for the differences in ACD between the urban and rural populations, we presume it was also caused by the myopia ratio in these two groups. Xu et al. had demonstrated that myopic refractive error was significantly associated with younger age and an urban region . Because our study was based on cataract surgery candidates who had not been correctly evaluated with their refractive error status, no advanced refractive error corrected analysis about the relationship between and biometry was performed. In He’s study, however, refractive error was strongly correlated with axial length and anterior chamber depth by using the multivariate models . The differences in ACD may be strongly related to the AL distribution in the two populations.
In our study, we found that 44.18% of the overall population had 1.0 D or more corneal astigmatism, which is higher than other Chinese population results. Chen  reported that 41.3% of the eyes studied had presented a corneal astigmatism equal to or higher than 1.0 D，while Cui and Yu reported 43.9% and 43.5%, respectively [6, 14]. Two other Chinese studies showed a higher percentage of eyes with over 1.0 D astigmatism than our study (Yuan 47.27%, Guan 45.46%) [20, 21]. We speculate that the difference between these studies is caused by two factors: first, these studies focused on a larger age range than our study, and second, these studies recruited many cases that included both eyes, which may increase the statistical power of a test, thereby increasing the likelihood of detecting true significant effects.
In our study, we found that the proportion of eyes in urban areas that had a corneal astigmatism of 1.0 D or greater was 41.9%, which was less than the proportion in rural areas (45.9%). We also found the astigmatism axis turned to the ATR direction with age. This trend of ATR increasing as subjects grow older has been proven by many previous investigations. However, a higher percentage of ATR astigmatism was found in the rural population than in the urban population. In the population with 1.0 D or more astigmatism in urban areas, 42.25% of patients had WTR and 46.16% had ATR. In rural areas in the same population, 23.14% of patients had WTR and 67.05% had ATR. Previous studies have demonstrated a higher percentage of ATR astigmatism in an Asian population (49.7% in our study, 53.2% in Cui’s study and 62.2% in a Thailand population) [6, 22] than in Western populations. There are many factors that influence whether the astigmatism axis direction turns with-the-rule or against-the-rule, such as ethnicity, anatomy, eyelid morphology and the effects of intraocular pressure on the curvature of the cornea. The reasons that have led to the difference in prevalence of ATR in various locations still needs to be investigated further. Cataract surgeons should consider using more toric IOLs in rural populations than in urban populations.
Our study has some limitations. First, the study was clinic-based and may not be representative of the entire population. Second, the relationship between biometric features and refraction was not evaluated due to the cloudy crystalline lens of the cataract patients. There were also missing measurements, such as the white-to-white, central corneal thickness, lens thickness and vitreous chamber depth. Finally, social status, education and occupation were not recorded in our study.
In conclusion, we report biometry and astigmatism data in a large cohort of urban and rural subjects for the first time. In our study, the rural subjects were more likely to have a short AL, shallow ACD and an axis turned in an ATR direction. This profile of ocular biometric data and corneal astigmatism will be helpful in planning IOL power calculations and astigmatism correction in patients who live in different locations.