Subjects and methods

From January 2009 to November 2011, the healthy right eye of a total of 292 adults (145 men, 147 women) from regular patients were recruited from the Department of Ophthalmology, Far Eastern Memorial Hospital, Taipei, Taiwan, China into our study. Subjects with previous ocular surgery, acute or chronic ocular trauma, glaucoma, diabetes mellitus, corneal or retinal disease,  or history of wearing contact lenses were excluded. The IOLMaster (Carl Zeiss, Jena, Germany) was used to measure AL，anterior chamber depth (ACD), and mean keratometric value (K). An ORA was used to measure the Goldmann-correlated intraocular pressure (IOPG) and CH, and the integrated handheld ultrasonic pachymeter of the ORA was used to measure CCT under topical anesthesia. Each subject was assigned to the IOLMaster measurement first and then the ORA measurement, because the former examination was noncontact and not requiring anesthesia. All these examinations were performed by a technician who was experienced in performing the study protocol. All measurements were performed at the same time of day to decrease diurnal effects. Demographics of the patients were also recorded. All patients signed informed consent forms agreeing to the examinations.
This research was carried out in accordance with the Declaration of Helsinki and after obtaining approval from the Institutional Review Board of Far Eastern Memorial Hospital (Clinical Trials.gov Identifier: NCT01235780).  The Pearson correlation coefficient was used to test correlations between CH and quantitative factors, including age, AL, ACD, K, CCT, and IOPG. The chi-square test was used to assess the significance of the correlation between CH and gender. The results are expressed as mean ± standard deviation, and p value less than 0.05 was considered statistically significant.

The mean CH was 10.16±1.55 mmHg (range:8-14). The age (mean 66.51±12.12 years; range, 33-87) and AL (mean 23.86±1.95 mm; range, 21.5-33.5) were significantly and negatively correlated with CH (the Pearson correlation coefficient [r]=-0.21, P=0.003 [Figure 1] and r=-0.27, P=0.0001 [Figure 2], respectively).

Figure 1 Graph showing a significantly negative correlation between the age and the value of corneal hysteresis (P=0.003)

The CCT (mean 537.77 ± 35.24 μm; range, 460-598) and IOPG (mean 14.82 ± 3.28 mmHg; range, 6.6-24.6) were significantly and positively correlated with CH (r=0.42, P=0.0001 [Figure 3] and r = 0.18, P=0.03 [Figure 4], respectively).
The ACD (mean 2.94 ±0.43 mm; range, 1.83- 4.13) and K (mean 44.23±1.48 diopters; range, 41-48.12) were not correlated with CH (r=-0.03, P=0.34 and r=0.12, P=0.18, respectively). The CH was comparable in male and female eyes (P=0.23).

Discussion

CH is determined during rapid motion of the cornea in response to a short-duration (20-ms) pulse of air pressure applied to the cornea¹. An electro-optical system detects the corresponding deformation of the cornea. The air pulse causes the cornea to move beyond the point of applanation to form a slight indentation. When the pressure of the air pulse decreases, the cornea returns to its normal configuration, passing for a second time the point of applanation. The difference in the two pressure values at the point of applanation, which is called the CH, depends on the viscoelastic properties of the corneoscleral shell. The average of the two applanation pressures is the IOPG.
Ortiz and associates reported lower CH values in older eyes; the difference in CH value between the youngest (9-14 years) and oldest (60-80 years) age groups was statistically significant⁶. Fontes and associates reported that CH in healthy adults was negatively associated with age⁹. It has also been demonstrated that the cornea stiffens with age as a result of glycation-induced cross-linking of collagen molecules¹⁵. Kamiya and coauthors demonstrated that aging significantly decreases CH without significantly changing CCT or IOP, suggesting that age-related structural changes due to collagen cross-linking may lead to a degeneration of corneal biomechanical properties independent of CCT or IOP⁸. These findings are in accordance with our present finding that older age is correlated with lower CH value.
Prior studies demonstrated that CH was positively associated with CCT measured by optical coherence tomography in normal Chinese^12,13. Among healthy Japanese and Caucasian eyes, higher CH was associated with thicker CCT measured by ultrasonic pachymetry^10,16. In normal eyes in Brazil, CH was significantly correlated with CCT measured using a Pentacam analyzer⁹. In the current study of normal Chinese eyes, CH was significantly and positively correlated with CCT measured using ultrasound. Thus, our findings are in agreement with those of previous studies, and CCT regardless of method of measurement is strongly associated with CH. It is hypothesized that CCT is an important determinant of corneal viscoelasticity in various races.
A significant and positive correlation between IOPG and CH was also found in cases of anisometropia¹³. In this article, eyes with increased CH were found to have higher IOPG, which was positively correlated with CCT (r=0.17, P=0.02). We postulated that reduction in IOPG may be caused by decrease in CCT. On the other hand, decreased IOPG may distort the cornea, change the physiological properties of the cornea, and alter CH value. For example, IOP readings are falsely low after LASIK surgery compared to normal IOP readings³. The reason may be corneal thinning due to ablation, or change in the biomechanical properties of the cornea as a result of flap creation, or both^6,17.
Xu and coauthors revealed, in patients with high anisometropia, that CH was lower in the myopic eye than in the contralateral control eye¹³. Shen and coauthors reported that CH was significantly lower in patients with high myopia (defined as spherical equivalent greater than-9 diopters) than in normal subjects¹². In the current study, CH was also lower in eyes with longer AL. Previous studies found that change in CCT does not depend on degree of myopia or AL^13,18-20. Lower CH in long eyes cannot be explained by changes of CCT, and may indicate that the biomechanical properties of the cornea have become altered. Scleral biomechanical properties such as stiffness decrease during the progression of myopia because of pathologic changes in the scleral collagen^21,22. The corneal stroma is composed of a collagenous extracellular matrix that is continuous with the sclera²¹. Elongation of AL can change the biomechanical properties of the sclera and possibly the cornea^12,13.
Kamiya and coauthors found no significant correlation between CH and gender in normal Japanese¹⁰. The present study also found no correlation. Howeve, Fontes and coauthors reported that CH was higher in the normal healthy eyes of Brazilian women than in the normal healthy eyes of Brazilian men⁹. In Italian population, men demonstrated a higher CH than women did²³. East Asians（Chinese and Japanese）were very different from South Americans and Italians in this respect． Therefore， racial difference may explain the gender difference．
Prior studies showed no significant correlation between CH and mean keratometric values measured by an autorefractometer and by a Pentacam^9,10. In the present study, CH was not associated with K measured by the IOLMaster. Therefore, K measured by different approaches was not associated with CH．
The association was not significant between CH and ACD, either measured by Pentacam⁹, anterior segment tomographer²⁴, or IOLMaster in our study．
CH values in healthy Taiwan-Chinese adults determined by the present study were comparable to those in children in our prior study⁵. Previous studies consisted of Hong Kong-Chinese²⁰ and SingaporeChinese²⁶ adults also having similar CH data in Taiwan-Chinese in this study. CH values in East Asians (Japanese¹⁰, Koreans²⁴, and Chinese^{20, 26}) were slightly lower than those in healthy Caucasians^1,16. Therefore, it is hypothesized that biomechanical properties of the cornea may be influenced by ethnicity^1,10,13,16. Actually, differences in CH values have been noted between healthy black and white subjects¹⁴.