Comparison of Ocular Monochromatic Higher-order Aberrations in Normal Refractive Surgery Candidates of Arab and South Asian Origin.

PURPOSE: To compare the ocular monochromatic higher-order aberration. (HOA) profile in normal refractive surgery candidates of Arab and South Asian origin.
METHODS: This cross-sectional, observational, comparative study was performed in the cornea department of a specialty hospital. Normal refractive surgery candidates with no ocular morbidity except refractive error were recruited. Refractive surgery candidates underwent a preoperative evaluation, including wavefront aberrometry with the iDesign aberrometer (AMO, Inc., Santa Ana, California, United States). The HOA from right eyes were analyzed for HOA signed, absolute, and polar Zernike coefficients.
RESULTS: Two hundred Arab participants (group 1) and 200 participants of South-Asian origin (group 2) comprised the study sample. The age and refractive status were comparable between groups. The mean of the HOA root mean square (RMS) was 0.36 ± 17 μ and 0.38 ± 18 μ for Arab and South-Asian eyes, respectively (P < 0.05, rank sum test [RST]). Of the 22 higher order signed Zernike modes, only Z3 (-3), Z3 (-1),3 (1), Z4 (-4), Z4 (-2), Z4 (0), Z4 (4), and Z5 (-5) were significantly different from zero (one sample t-test, P < 0.002, with a Bonferroni correction of 0.05/22). All the signed and absolute Zernike terms were comparable between groups (RST, P > 0.002 [0.05/22]). The polar coefficients for coma, trefoil, spherical aberration, and tetrafoil were comparable between groups (P > 0.05, RST). Combined RMS values of third, fourth, fifth, and sixth order also were comparable between groups (P > 0.05, RST).
CONCLUSIONS: Preoperative whole eye HOA were similar for refractive surgery candidates of Arab and South-Asian origin. The values were comparable to historical data for Caucasian eyes and were lower than Asian (Chinese) eyes. These findings may aid in refining refractive nomograms for wavefront ablations.

INTRODUCTION

Ocular aberrometry has significantly increased our understanding of the optics of the eye and objective visual quality.123 Wavefront analysis has expanded beyond excimer ablation and has been applied to wavefront customized lenses, visual quality assessment after ocular and multiple ocular pathologies.456789

Wavefront analyzers measure ocular aberrations using the principles of Hartmann–Shack aberrometry, ray tracing, dynamic skiascopy, or Tscherning aberrometry.10111213 These devices measure the distortions created by the refractive error of the eye in a wavefront of light traveling through the ocular system. The total aberration profile of the eye can be decomposed into Zernike polynomials and presented as lower order aberrations and higher-order aberrations (HOA), based on the radial orders. For example, radial orders 0–2 are lower order aberrations and 3 and higher are HOA. In The American National Standards Institute double index scheme is the preferred method of denoting Zernike polynomials. In this indexing method, Zmn indicates m is the radial order and n is the angular frequency for the polynomial Z.

Lower order aberrations correlate with conventional refraction. HOA are not measurable by conventional refraction, and thus are the major topic of interest in wavefront analysis. Measurement of the HOA profile can be affected by multiple factors including pupil size, cycloplegia, age, ethnicity, ocular surface status, and the device used.141516171819202122232425262728

Previous studies that document HOA for different demographic subgroups groups have reported interesting results. The HOA profiles of Caucasian, Asian Chinese, Asian Indian, and Brazilian eyes have been documented in the literature.22232425262728 However, other than a study measuring corneal spherical aberration, there are no studies in the literature to evaluate the total ocular wavefront profile in Arab eyes.29 In this study, we evaluate the ocular (whole eye) wavefront profile of refractive surgery candidates of Arab origin and compare it to candidates of South Asian (Indian sub-continent) origin. To the best of our knowledge, this is the first study evaluating and comparing the HOA profile of refractive surgery candidates of Arab origin.

METHODS

Study population

This cross-sectional, comparative, observational study was performed at a specialized hospital. All the candidates gave informed consent for their participation. The study had the approval of the Institutional Review Board and followed the tenets of the Declaration of Helsinki. The candidate pool consisted of patients interested in refractive surgery. Subjects who wore soft contact lenses were advised not to wear them from 2 weeks prior to the tests. None of the subjects in our study wore hard contact lenses. All candidates underwent detailed history taking, slit lamp evaluation and Scheimpflug topographic analysis (Sirius, Costruzione Strumenti Oftalmici, Italy), pupillometry, tear volume, and film stability assessment, dilated anterior segment and indirect ophthalmoscopic retinal evaluation as a part of the initial assessment. Exclusion factors were dry eyes, any previous surgical intervention in the eye or corneal scars/irregularity, cataract, or any other ocular morbidity including corneal ectasia or scarring.

Wavefront aberrometer and measurements

The iDesign advanced Wavescan studio (Abbott Medical Optics, Santa Ana, CA, USA) was used for wavefront measurement in all cases. This device is a Hartmann–Shack based aberrometer that is a new, upgraded version of the Wavescan aberrometer (Abbott Medical Optics, Santa Ana, CA, USA). It has 5 times greater resolution, encompassing over 1250 data point for a 7.0-mm pupil compared to 240 points for the Wavescan. In addition to the higher and lower order aberration data, this device also provides a Hartmann–Shack based topography, white-to-white corneal diameter, and scotopic pupil diameter. The range of measurement is from −16 D to +12 D of sphere, up to 8 D of cylinder, and up to 8 μm of root mean square (RMS) HOA.30 The wavefront outcomes and the repeatability of the keratometric function have been found to be satisfactory in two recent studies.3132

Wavefront acquisition was performed after dark adaptating the pupil for 30 min. No dilating drops or lubricants were instilled prior to wavefront measurement. The wavefront acquisition was carried out in dark room conditions. After properly focusing the eye, the candidates were advised to blink a few times before the actual wavefront capture. The candidates were advised to close the eyes between measurements for 1 min. The software then analyzed the images for iris registration, wavefront data, and corneal topography data. The review screen showed a green icon for all three when the measurements were of appropriate quality for treatment or analysis. Cases were included in the study only if the wavefront capture was ≥6 mm. Three consecutive “good” measurements were taken. The total attempts were limited to a maximum of five, i.e., if useful data were not captured for three out of five attempts; the subject was excluded from the study. The machine automatically selected the wavefront scan with the best quality for the treatment. This scan was also chosen for analysis in the study. The data were noted and analyzed for a 6-mm pupil diameter. For HOA, the machine used Fourier’ transforms for analysis and presented Zernike polar coefficients and standard (signed) Zernike coefficients RMS values [Figure 1]. The complete HOA values RMS (HOARMS) value in microns was used for analysis as well as the data of individual polynomials up to sixth order. Two hundred consecutive cases were included in both groups (group 1: Arab origin; group 2: South Asian (Indian sub-continental) origin.

Figure 1

Screenshot of the wavefront capture output by the iDesign (AMO, CA) aberrometer. Polar and signed Zernike coefficients are presented in a tabular fashion, along with the total higher-order aberration root mean square and other data. Identifiers have been darkened out

Statistical analysis

All the data were exported as JPEG images and manually entered into an MS Excel (Microsoft Corp., Richmond, VA, USA) sheet. The data was then transferred to SPSS 16.0 (IBM Corp., Armonk, New York, USA) for analysis. Only the right eyes were used for analysis. The rank sum test (RST) was used to analyze the difference between means. Bar charts were used to illustrate the spread of Zernike polynomials/modes. The data was analyzed for polar as well as signed coefficient. Polar coefficients were computed as the RMS values of signed coefficients. Coma was computed as RMS of Z1−1 and Z11; trefoil as RMS of Z1−3 and Z13; secondary astigmatism as RMS Z4−2 and Z42; tetrafoil as RMS Z4−4 and Z44. RMS function is always positive denoting only the magnitude but not sign, hence, spherical aberration was converted to its absolute value when comparing to other polar modes.

RESULTS

Demographic, refractive error, and total higher-order aberration root mean square

The mean age, sphere, cylinder, spherical equivalent, and total HOARMS were comparable (P > 0.05, rank rum test), [Table 1]. The gender distribution was also comparable (female: male = 112: 88 (group 1) and 126: 74 (group 2); χ2 = 2.03, P = 0.15, Chi-square test).

Table 1

Age, refractive error, and total HOARMS distribution

Distribution of individual Zernike modes

Of the 22 higher order signed Zernike, only Z3−3, Z3−1, Z31, Z4−4, Z4−2, Z40, Z44, and Z5−5 were significantly different from zero (one sample t-test, P < 0.002 after Bonferroni correction [0.05/22]). Therefore, significantly higher terms from the third and fourth order were different from zero (7/9 terms) compared to only 1/13 terms from the fifth and sixth order (P = 0.001, Fischer's exact test). However, all the absolute Zernike modes were significantly different from zero (one sample t-test, P < 0.002 after Bonferroni correction [0.05/22]).

The mean and standard values for signed and absolute Zernike modes are presented in Table 2. All the signed and absolute Zernike terms were comparable between groups (Wilcoxon RST, P > 0.002 after Bonferroni correction (0.05/22), [Table 2].

Table 2

Distribution of Zernike coefficients

Polar coefficients and root mean square of entire orders

The polar coefficients of coma, trefoil, spherical aberration, and tetrafoil were comparable between groups (P > 0.05, RST), [Table 3]. The frequency distribution patterns of coma, tetrafoil, and spherical aberration are presented in Figure 2. The combined RMS values of third, fourth, fifth and sixth order were comparable (P > 0.05, RST), [Table 3]. The frequency distribution pattern of the four orders (3rd to 6th) is presented in Figure 3.

Table 3

Comparison between various radial orders and polar Zernike modes

Figure 2

Frequency distribution polygon showing the distribution and frequency of various components of aberration profile in group 1 (Arab origin) and group 2 (South Asian origin). (a) For total higher-order aberration root mean square, (b) for Coma, (c) for Trefoil, and (d) for Spherical aberration. The root mean square magnitude scale is similar for the four graphs in order to demonstrate their mean spreads with respect to each other

Figure 3

Frequency distribution polygon showing the distribution and frequency of the four higher orders studies in group 1 (Arab origin) and group 2 (South Asian origin). (a) For 3rd order, (b) for 4th order, (c) for 5th order, and (d) for 6th order. The root mean square magnitude scale is similar for the four graphs in order to demonstrate their mean spreads with respect to each other

Within group analysis indicated the mean RMS values of individual orders was seen in the following sequence from largest to smallest: 3rd, 4th, 5th followed by 6th order (Kruskal–Wallis test, P < 0.001 for both group 1 and group 2). Furthermore, 3rd order was significantly higher than the 4th order (RST, P < 0.001 for both groups 1 and 2). Coma, trefoil, and spherical aberration were the three major Zernike modes in both the groups. Coma was significantly higher than trefoil and spherical aberration in both groups (RST, P < 0.001 for both groups 1 and 2 for comparisons between coma and spherical aberration, and between coma and trefoil).

DISCUSSION

In this study, we have evaluated the total ocular monochromatic wavefront profile of normal candidates from Arab origin for the 1st time in published literature. The data was compared to a South Asian cohort in the same setting. The selection of South Asian origin for comparison was based on the population distribution profile in our cornea and refractive surgery services, where these two groups constitute the largest majority of patients.

We found that there was no significant difference in the age, gender, and refractive error between groups. This observation indicates the comparability of these groups. The Zernike modes such as signed coefficients, absolute coefficients, and polar modes were compared separately. Salmon and Van de Pol have previously explained the rationale behind evaluating signed, absolute, and polar coefficients separately.33 We found that the total HOARMS value, signed and absolute values of Zernike modes were comparable between Arab and South Asian eyes. Our result were similar to previous reports that in normal, unoperated eyes, the magnitude of fifth order, and higher HOA are very small and not significantly different from zero.2234 This pattern occurs dues to variation of signed Zernike coefficients around zero and is seen typically in normal or unoperated eyes, which are typically less distorted than eyes that have undergone surgery or pathological eyes. When only absolute values were considered, all coefficients were significantly different. This justifies using all the Zernike modes in computing the RMS of all Zernike coefficients.

Ethnic variations in HOA have been noted previously.22232425262728 However, there are very few studies comparing two different ethnic populations in the same setting. Using the same setting can control for variability of factors such as type and make of aberrometer, population cohort, and age. We found that the mean HOARMS value in our study was 0.36 ± 17 µ and 0.38 ± 18 µ for Arab and South Asian eyes, respectively. There was no difference in mean HOARMS value in the two groups. Nakano et al. compared Asian (Chinese) eyes with non-Asian eyes in a Brazilian setting.26 They found that in spite of the higher refractive error in the Asian eyes, the HOARMS values for various orders and the total HOARMS were comparable between groups. They did not comment on individual Zernike polynomials. Cervino et al. found that the HOARMS values for Caucasian and British Asians were similar.25 With the exception of Z33 and Z40, all other Zernike modes in their study were comparable.26 The only previous study specifically for South Asian origin was from Northern India and was published 6 years ago.28 The authors showed that the HOA profile of the South Asian (Indian) eyes was similar to the literature for Caucasian eyes compared to those with Chinese eyes.28 The mean HOARMS for Asian Indian eyes (0.36 ± 0.26 µ) from this previous study is similar to that seen our study. Furthermore, the mean HOARMS in both groups were also comparable to that seen the largest study on normative data of wavefront aberration by Salmon and van de Pol.33 They evaluated 2560 eyes of mixed origin from multiple centers with multiple aberrometers and found the mean HOARMS was 0.33 ± 0.13 µ at 6 mm pupil diameter).34 In the current study, the mean HOARMS was lower than that seen in Chinese Asian eyes reported in 2 studies (0.49 ± 0.16 µ for Wei et al.,27 0.51 ± 0.71 µ for Nakano et al.26) for a 6 mm pupil.

The most prominent order of Zernike polynomials in our study was the 3rd order (0.28 ± 14 µ for Arab eyes, 0.29 ± 16 µ for South Asian eyes). This was comparable to the value reported by Salmon and van de Pol (0.25 ± 0.12 µ), Wang and Koch (0.22 ± 0.09 µ) and Prakash et al. (0.23 ± 0.15 µ) at 6 mm pupil for mixed population of, Caucasian and South Asian patients respectively.2833 The average value for third order was higher in Chinese Asian eyes (0.37 ± 0.16 µ) compared to our values for both groups.27

Coma, trefoil, and spherical aberration were the largest aberrations in both groups. This trend was comparable to previous studies, which report a similar predominance of these three aberrations in normal nonsurgical eyes.26272833

There are drawbacks to our study. The sample size is relatively small. Second, even though this aberrometer is a newer, more advanced version of an already established aberrometer (Wavescan, AMO, CA), which has been shown to have good repeatability,35 a new study of repeatability would be helpful.

To conclude, we found that the HOA profile similar in normal, unoperated eyes of Arab and South Asian origin. Our data concur with a previous study on eyes of South Asian patients. The values were comparable to historical data for Caucasian eyes and were lower than that seen for Chinese Asian eyes. These findings can have useful implications for nomogram creation and adjustments for refractive surgery candidates of Arab origin undergoing wavefront-guided LASIK.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.