Introduction

Biomechanics is the study of the mechanical laws governing the structural components of organisms or objects1. The biomechanical properties of the cornea reflect the corneal deformation that occurs under external forces, which is a deciding factor of corneal shape2.

The measurement of corneal biomechanics is of great importance for the clinical diagnosis of many diseases, such as keratoconus3,4,5,6,7, Fuchs Dystrophy7,8, and diabetes2. It also plays an essential role in the preoperative evaluation of cataracts9,10, glaucoma, refractive surgery6,7, and other ophthalmic operations.

Because corneal biomechanics plays a vital role in the clinic, its study has always been a hot subject in ophthalmology. In fact, since the late 1970s, corneal biomechanics has attracted the attention of many researchers. Furthermore, the detection instruments used to assess corneal biomechanics are constantly improving with the advancement of technology. In recent years, various instruments have been developed for measuring corneal biomechanics in vivo, such as the ocular response analyzer (ORA, Reichert ophthalmic instrument) and the Corvis ST (Oculus Optikgera¨te GmbH, Wetzlar, Germany) instrument. Corvis ST is a noncontact metering system with an ultra-high-speed Scheimpflug camera that provides comprehensive corneal biomechanical parameters and biomechanically corrected IOP, overcoming ORA's limitations by offering detailed dynamic corneal deformation information7,11. In addition, to assess the effect of biomechanical parameters on the results obtained, the software version was updated several times and new parameters were added, including the SSI, CBI, and SPA1.

Corneal biomechanical parameters are good predictors of early corneal dilation5. Therefore, a timely and correct understanding of the changes in corneal biodynamic parameters is essential for evaluating diseases and monitoring surgical outcomes. Recently, several studies have found a significant association between age and various corneal biomechanical parameters12,13,14,15. Specifically, in the Indian population, significant differences in 19 corneal biomechanical parameters were found across different age groups16. In Caucasians, age was significantly negatively correlated with corneal hysteresis (CH) and corneal resistance factor (CRF), and significantly positively correlated with Goldmann-correlated IOP (IOPg) and corneal compensated IOP (IOPcc)17. Therefore, This highlights the importance of gaining biomechanical knowledge across different age groups to improve the decision-making process for individual patients. What’s more, Chua showed that CH was lower in Chinese than in Indians and that CH level was correlated significantly with genetic ancestry in Southeast Asian populations18. Vinciguerra et al. observed that three dynamic corneal response (DCR) parameters (SPA1, Ambrosio’s relational thickness horizontal (Arth) and SSI) measured by Corvis ST, were significantly different between Chinese and Caucasian healthy subjects19. These findings suggest that corneal biomechanics vary among different ethnic populations, and these differences may play a crucial role in disease diagnosis. However, few large-scale studies have been performed in the Chinese population on the correlation between age and the comprehensive corneal biomechanical parameters provided by Corvis ST. Additionally, few studies have simultaneously included new parameters such as SSI, CBI, and SPA1.Therefore, this study aims to explore the distribution of corneal biomechanical parameters among different age groups in healthy Chinese populations and to study their relationship with various demographic and ocular characteristics.

Methods

Subject recruitment

This cross-sectional study enrolled 864 eyes of 543 normal Chinese participants at the Beijing Tongren Hospital, Beijing, China, from August 2017 to January 2020. Although we initially recruited 543 healthy participants with the aim to examine both eyes of each participant, due to incomplete examination data for some participants, the final analysis included data from only 864 eyes.

The corneal biomechanics parameters of visually normal participants measured using Corvis ST were analyzed, and almost all participants (age range, 11–84 years) were stratified into five separate age groups, each spanning a decade (11–20, 21–30, 31–40, 41–50, and > 50 years). Participants who had a history of corneal or ocular surgery, ocular pathology, systemic diseases affecting the eye, or long-term use of topical medications were excluded. This study was approved by the Ethics Committee Office of Beijing Tongren Hospital and was in line with the principles of the Declaration of Helsinki.

Sample size

The sample size for this study was calculated based on the expected correlation between corneal biomechanical parameters and age. We used Spearman correlation analysis with an anticipated medium effect size (correlation coefficient, r = 0.3), a significance level (α) of 0.05, and a statistical power (1 − β) of 0.80. This resulted in an estimated minimum sample size of 95 participants per age group. Considering potential data loss and to ensure adequate representation across five age groups, we enrolled a total of 543 healthy participants, accounting for 864 eyes.

Ocular examinations

All participants underwent a complete series of eye examinations, including uncorrected vision assessment, slit-lamp microscopy, fundus examination, corneal biomechanics assessment, corneal tomography (Pentacam), and intraocular pressure examination (Corvis ST). All ophthalmic measurements were performed by two trained ophthalmologists between 08:00 and 17:00 on the same day.

Corvis ST measurement

The cornea biomechanical characteristics were measured using Corvis ST (software version 1.6r2015), a new noncontact tonometer with optical thickness measurement function that can use a ultra-high-speed Scheimpflug camera to record the cornea’s reaction under a defined air pulse. The instrument can capture over 4300 images per second, with each image capturing 576 points20. Several studies have described the details and principles of the Corvis ST instrument21,22. Our study assessed several demographic and ocular variables (including age, IOP, and CCT), as well as 35 parameters pertaining to corneal biomechanical properties from the Corvis ST assessment, especially several new parameters: SPA1, CBI, and SSI. Based on the Scheimpflug image, IOP and corneal thickness can be measured precisely. Measurements that showed a quality index of “OK” were exclusively included in the analysis.

Table 1 shows the short names, full names, and explanations of the 35 Corvis ST parameters. The parameters included 28 DCR parameters and seven Vinciguerra screening parameters23,24,25. The DCR parameters were divided into four phases: at first applanation (A1), at second applanation (A2), at highest concavity (HC), and at maximum (Max). The Vinciguerra screening provides are seven helpful parameters to distinguish between normal and keratoconus corneas.

Table 1 Description of the parameters measured using Corvis ST.
Full size table

Statistical analysis

Statistical analysis was performed using SPSS, version 26.0 (SPSS, Inc., Chicago, IL, USA). The Shapiro–Wilk test was used to estimate the distribution normality of the measured variables and the results indicated that the corneal biomechanical parameters were normally distributed. Moreover, the mean and standard deviation (M ± SD) of the continuous variables and the frequencies and percentages of the categorical variables were used for descriptive analyses. In addition, participants were distributed into different age groups on a 10-year basis, i.e., 11–20, 21–30, 31–40, 41–50, and > 50 years. The means were compared between the left and right eyes using a paired sample t-test26 and the means of the different age groups were compared using analysis of variance. Spearman’s correlation analysis was used to explore the correlation between 35 biomechanical parameters and demographic/ocular characteristics (including age, IOP, and CCT). Significance was set at P < 0.05. A stepwise multivariate linear regression analysis was performed using demographic/ocular characteristics (including age, IOP, and CCT) as independent variables and 35 biomechanical parameters as dependent variables, in order to avoid the possibility of Type I errors, the test level was then adjusted using the Bonferroni correction and was considered statistically significant when P < 0.0014(0.05/35 ≈ 0.0014).

Results

Demographic and ocular characteristics

This cross-sectional study was conducted on 864 (431 right and 433 left) eyes of 543 normal individuals. The demographic characteristics of the participants in each age group are described in Table 2. In the 864 eyes analyzed in the study, there were approximately equal numbers in each age group, and the CCT was matched across age groups.

Table 2 Demographic and ocular characteristics of the participants.
Full size table

Corneal biomechanical parameters

The analysis of the correlation between the left and right eyes and the 35 biomechanical parameters revealed that 13 of them were significantly correlated with the differences between the eyes (Table 3). Moreover, the deflection length of the cornea was significantly different between the left and right eyes, regardless of whether it pertained to the first or second applanation.

Table 3 The M ± SD values of the corneal biomechanical parameters in left and right eyes.
Full size table

In this study, we analyzed 35 corneal biomechanical parameters, including 28 DCR parameters and seven Vinciguerra screening parameters, with a particular emphasis on three new parameters (CBI, SSI, and SPA1). We investigated whether these parameters differed across age groups, as described in detail below.

Dynamic corneal response parameters in the different age groups

Out of the 28 DCR parameters under study, 22 were significantly different across the age groups (P < 0.05). At applanation 1, all parameters other than A1V differed significantly among the age groups (P < 0.05). At applanation 2, A2T was similar across the different age groups (P > 0.05), whereas the remaining parameters were significantly different across the age groups (P < 0.05). In turn, at the HC, all parameters were significantly different (P < 0.05), with the exception of HCdArcL and Radius (P > 0.05). Regarding the Max parameters, all parameters but MIR and WEMMT differed significantly across the age groups (P < 0.05). Table 4 reports the corneal biomechanical parameters in the various age groups.

Table 4 The mean ± SD values of dynamic corneal response parameters and Vinciguerra screening parameters in the different age groups.
Full size table

Vinciguerra screening parameters in the different age groups

Among the seven Vinciguerra screening parameters under study, only bIOP was similar among the age groups. In contrast, the remaining six parameters, i.e., deformation amplitude ratio (DARM) 2 mm and 1 mm, SPA1, integrated radius (INR), CBI, and SSI, were significantly different across the age groups (P < 0.05). The mean biomechanically-corrected Intraocular pressure (bIOP) value was lowest in the 21–30-year-old group and highest in the > 50-year-old group.

We analyzed the distribution of new parameters (including CBI, SPA1 and SSI) among different age groups emphatically (Fig. 1). The results showed that SPA1 had an upward trend after the age of 30, and SSI showed a significantly rising trend across the five age groups.

Fig. 1
figure 1

Mean values of new biomechanical parameters (including SPA1, CBI, and SSI) for all subgroups according to age and shows the variation by age (ANOVA).

Full size image

Spearman’s correlation and stepwise multivariate linear regression analyses

Table 5 lists the results of Spearman’s correlation analysis, which revealed that, out of the 35 parameters under analysis, 27 were significantly associated with age, 23 were significantly associated with IOP, and 24 were significantly associated with CCT.

Table 5 Correlation between demographic/ocular characteristics and corneal biomechanical parameters.
Full size table

The results of the stepwise multivariate linear regression analysis are reported in Table 6. We aimed to study the effect of demographic/ocular characteristics (including age, IOP, and CCT) on cornea biomechanical parameters. Therefore, the demographic/ocular characteristics mentioned above were used as the independent variables and the 35 biomechanical parameters were used as the dependent variables in the stepwise multivariate linear regression analysis, which revealed that 21 out of 35, 13 out of 35, and 11 out of 35 biomechanical parameters were significantly correlated with age, IOP, and CCT, respectively. 11 corneal biomechanical parameters were positively correlated with age and 10 were negatively correlated with age. CBI was significantly negatively correlated with IOP and CCT, SSI was significantly positively correlated with age and IOP, and SPA1 was positively correlated with IOP and CCT.

Table 6 Results of the stepwise multiple regression analysis between dynamic corneal response parameters and demographics/ocular characteristics.
Full size table

Discussion

Because of its significant value in clinical diagnosis, corneal biomechanics is highly valued by researchers. In recent years, the field of corneal biomechanics has developed rapidly, and the instruments that are used to measure corneal biomechanics have changed continuously, accompanied by the emergence of new parameters. ORA provides many biomechanical parameters and is an instrument that is widely used in the clinic; however, it cannot provide the DCR parameters4,20. Corvis ST, which is a new non-contact tonometry system that can use an ultra-high-speed Scheimpflug camera to record the cornea’s reaction under a defined air pulse27 was introduced in 2010. This instrument can display the dynamic deformation of the cornea in real time and directly describe the eye’s mechanical behavior. The Corvis ST instrument exhibits high repeatability and reproducibility in measuring DCR parameters in healthy eyes28. To improve the accuracy of the measurement of biomechanical parameters continuously, the versions of this software have been updated several times. In this study, the newer Corvis ST version (1.6r2015) was used to analyze 28 DCR parameters and seven Vinciguerra screening parameters, including new parameters, such as SPA1, SSI, and CBI. To the best of our knowledge, few studies have assessed these new parameters simultaneously. Moreover, the present work is one of the most extensive studies that analyzed the association between corneal biomechanical parameters (provided by the newer Corvis ST version) and different age groups in a healthy Chinese population.

Our chief observations were as follows: 1. most of the corneal biomechanical parameters were not significantly different between the right and left eyes; 2. among the 28 DCR parameters, 22 items differed significantly between age groups; 3. SPA1 showing an upward trend after the age of 30 and SSI displaying a significantly upward trend when comparing the five age groups ; and 4. most of the parameters were significantly associated with demographic/ocular characteristics (including age, IOP, and CCT). Regarding the new parameters, CBI was significantly negatively correlated with IOP and CCT, SSI was significantly positively correlated with age and IOP, and SPA1 was positively correlated with IOP and CCT. In the following paragraphs, we discuss these observations at length.

In the analysis,13 parameters significantly different between right and left eyes. Moreover, the deflection length of the cornea differed significantly between the left and right eyes, regardless of whether it pertained to the first or second applanation. There are few studies available in this area of research. We wonder whether the observed variation between the left and right eyes is related to human asymmetric behavioral traits, such as the dominant eye29. However, at present, most of the studies of the dominant eye addressed refractive error, cataract, and amblyopia29,30,31,32, and few reports pertain to corneal biomechanics. Therefore, further exploration is needed to determine the specific mechanisms underlying the differences in biomechanical parameters observed between the left and right eyes.

In the analysis of “DCR parameters and age groups”, we found that most of the DCR parameters varied significantly among the different age groups. Many previous studies have confirmed our view: Elsheikh et al.33 show that the HC radius is significantly related to age; whereas Nathaniel et al.34 believe that whole eye movement max length(WEMML), deformation amplitude ratio (DARM), and radius are the parameters that are most affected by age. Whole eye movements are highly correlated with age, which can be explained by the fact that the composition of fat behind the eye ball changes with age, possibly leading to modifications in eye displacement under the air puff. Conversely, the correlation between the radius and age as well as the ratio of deformation amplitude may indicate their ability to quantify corneal biomechanics. In a study conducted by the Instituto de Olhos Renato Ambrósio, the maximum concave time was positively correlated with age35. Marta et al.36 reported significant differences between age groups during the second applanation.

Experimental studies performed in vitro have shown that corneal collagen fiber properties change with age, which explain the changes in corneal biomechanical parameters that occur with age37,38. In addition, Sherrard et al. demonstrated that corneal stiffness increases with age39. We believe that, with the increase of age, the diameter of collagen fibers38 as well as the collagen crosslinking of the corneal matrix increase, resulting in decreased corneal viscosity and increased hardness38,40,41. Moreover, the rate of stiffness increases in a nonlinear trend, which is more significant in older age groups37. Regarding the specific mechanism underlying this phenomenon, some studies have elaborated that (1) the increase in the number of collagen molecules is the main factor causing the increase in collagen fiber diameter, with the increase in the Bragg spacing between collagen molecules being another causative factor38; (2) as age increases, glycosylation induces the expansion of the spacing between molecules, leading to an increase in the crosslinking of molecules37; (3) when nonenzyme crosslinking increases, the cross-sectional area of collagen fibers32 and the fibril molecules of the cornea increase as a result8,37; and (4) when the proteoglycan composition of the inter-fibrillary matrix changes, the interfibrillar spacing42 decreases. Moreover, with age, the increase in corneal hardness may be related to the progression of glaucoma (older, harder corneas are less deformed, rendering the sieve plate more susceptible to damage)43,44. Therefore, it is imperative to explore fully the correlation between corneal biomechanical characteristics and age.

Our analytical data showed that A1 time (A1T) and highest concavity time (HCT) were positively correlated with age, whereas A2 time (A2T) was similar across the different age groups. A study of 35 patients with diabetes12 reported that A2T can represent the time-related part of the corneal viscoelastic response and can be used as an indicator of corneal viscoelasticity. This parameter is not only affected by corneal resistance, but also related to the cornea’s viscous damping properties. Moreover, when corneal hydration increases, the viscosity of the cornea may also increase. This hydration change may affect the measuring repeatability of A2T at different times of the same day45.

We also explored the Vinciguerra screening parameters. Our results showed that these seven parameters, with the exception of bIOP, varied significantly among the different age groups. To the best of our knowledge, this is the largest study that explored the age-related variations in the new Corvis ST parameters (SPA1, CBI, and SSI) in Chinese population. Our study showed that these new parameters were significantly different across the five age groups (P < 0.05). SPA1 is a corneal hardness parameter obtained by generating the initial data measured by Corvis ST, followed by calculation using a specific equation that considers the effect of mixed factors, such as IOP and eye movement. SPA1 is expected to become an effective indicator of cornea’s ability to resist deformation10. The SSI algorithm, which uses the least square method, and finite element models, which simulate the effects of IOP and the Corvis ST air puff, can predict corneal behavior12. SSI is a parameter that describes the intrinsic material stiffness of the cornea, with an SSI value > 1.0 indicating that the cornea is stiffer and a value < 1.0 indicating a softer cornea46. A previous study noted that SSI is significantly associated with age, which is consistent with our findings. Using a stepwise multiple linear regression, we observed a significant positive correlation between SSI and age.We can conclude that, with aging, the hardness of the cornea increases, which may be attributed to age-related changes in the corneal microstructure and changes in collagen fiber strength42. Fully understanding the relationship between corneal hardness parameters and age can help track the progression of dilated corneal diseases, such as keratoconus, and provide a reference for the study of refractive surgery. In the study of healthy Indians, CBI was positively correlated with age24, whereas CBI had no correlation with age in my research. This discrepancy in results may be related to the small sample size and genetic ancestry differences.

In our study, Spearman’s correlation analysis and stepwise multiple linear regression analysis showed that many corneal biomechanical parameters were affected by age, IOP, and CCT, which is consistent with previous findings. Hun Lee et al.14 pointed out that IOP was positively correlated with radius, and negatively correlated with max inverse radius (MIR). In addition, Vinciguerra et al.33 showed that theradius and DARM were associated with CCT. Another study performed in Hong Kong showed that peak distance (PD) was positively correlated with IOP and CCT, which differed from our stepwise multivariate linear regression results, in which PD was not associated with CCT and exhibited a negative correlation with IOP. The study reported by Lisa Ramm et al.12 showed that SSI had no significant correlation with CCT and IOP, whereas in the present study, SSI was significantly associated with IOP. This difference may be attributed to ethnic differences and the effects of systemic diseases, such as diabetes, which are not taken into account, as hyperglycemia in diabetes mellitus may cause changes in corneal biomechanics12 resulting in significant differences.

Zhang et al.47 pointed out that corneal stiffness was positively correlated with CCT and IOP; similarly, our analysis showed that a corneal-hardness-related parameter (SPA1) was positively associated with IOP and CCT. Therefore, when the corneal thickness and the intraocular pressure increase, the hardness of the cornea increases accordingly. We can explain this phenomenon based on the mechanism of corneal changes in glaucoma or corneal edema. An increase in intraocular pressure causes a change in the diameter of the cornea, which leads to a variation in corneal radius and redistributes the mechanical force of the corneal tissue48. Concomitantly, as intraocular pressure increases and the cornea hardens, it is more difficult for Corvis ST, which uses air blowing, to deform the cornea, and easier for the cornea to regain its shape. Thus, when the intraocular pressure increases, the first applanation takes a longer time and the second applanation is shorter. Our findings confirmed this view—the stepwise multiple linear regression showed that A1T was positively correlated with IOP, and that A2T was negatively correlated with IOP.

One of the advantages of the present research was that we used the newer Corvis ST version (1.6r2015) to study the corneal biomechanical parameters, especially the correlation between age and new parameters, such as SPA1, SSI, and CBI. Moreover, we conducted single-factor and multi-factor analyses of three demographic/ocular characteristics and 35 corneal biomechanical parameters. However, the present study also had several limitations. First, the study of corneal biomechanical characteristics was not sufficiently comprehensive because it lacked some items, such as astigmia and mean keratometry. Second, this was an observational cross-sectional study, in which it was impossible to determine the mechanism via which the corneal biomechanics change with age. In addition, we did not consider the effects of hormone-related factors, such as the menstrual cycle and long-term use of prostaglandins, on the measurement parameters49,50, or the effects of systemic diseases, such as diabetes12.

Conclusion

In conclusion, we performed a cross-sectional study in a large, healthy Chinese population and examined the correlation between corneal biomechanical parameters (including 28 DCR parameters and seven Vinciguerra screening parameters) across age groups; we also explored the factors that affected the corneal biomechanical parameters. Most of the corneal biomechanical parameters were significantly different among the age groups; moreover, they were significantly correlated withIOP and CCT. These results may provide a basis for the application of the new version of Corvis ST to ocular diseases, as well as a reference for the study of corneal biomechanics in the Chinese population. The adequate study of biomechanical parameters in different age groups is essential for the decision-making process in individual patients.