Linke, Stephan J.*; Baviera, Julio*; Munzer, Gur*; Fricke, Otto H.*; Richard, Gisbert*; Katz, Toam*
Several factors are known to influence the pupil size, including the level of retinal illuminance,1 the accommodative state of the eye,2 and various sensory and emotional conditions. The impact of age on the diameter of the pupil in the dark was described in 1958 by Kadlecova et al.3 Pupil size is assumed to be an important element of optical quality in refractive procedures, and various studies analyzed the relationship between the ablation zone and mesopic pupil size with respect to night vision problems after laser vision correction.4–10 Therefore, precise preoperative determination of pupil diameter in low light illumination conditions is desirable. The most frequently used devices for scotopic and low-mesopic pupil size measurement are digital (Procyon, London, UK) and handheld infrared (Colvard, Oasis, Glendora, CA) pupillometers, which allow simultaneous or closely separated measurements. Although there is an extensive scientific work on pupil diameter, knowledge on factors effecting scotopic/mesopic pupil size is limited. Cakmak et al.11 showed in a recent retrospective study of 412 refractive surgery candidates that age and magnitude of both spherical and cylindrical refractive error are the most determinative factors on mesopic pupil size. The aim of our large scale retrospective study on 13,959 refractive surgery candidates was to analyze the impact of age, spherical and cylindrical refractive error, average keratometry, minimal pachymetry, and gender on mesopic pupil size. Mesopic pupil size was measured with Colvard pupillometer.
This retrospective study was performed at Care Vision Eye Clinics in Germany and Austria between April 2006 and August 2010. Most of the subjects were candidates to undergo refractive surgery with Excimer laser, either LASIK or PRK. We examined in detail the medical records of all cases. There were 13,959 subjects. A detailed ophthalmological examination was performed for each case. In addition to general medical and ophthalmic histories, preoperative measurements included uncorrected distance visual acuity, best corrected distance visual acuity, manifest spherical refraction, cycloplegic refraction, tonometry, pupillometry, pachymetry, corneal topography, slit lamp examination of the anterior segment, and funduscopy. We converted all refractive data to minus cylinder form to prevent confusion during analysis. Informed consent for retrospective data analysis was obtained from refractive surgery candidates after explanation of the nature and possible consequences of the study and approval of local ethics committee (no. 2882) were achieved. The research adhered to the tenets of the Declaration of Helsinki.
Pupil Size Measurement
Because the level of dark adaptation is extremely difficult to standardize between subjects, all subjects were asked to wait in an area of dim illumination for 2 min before examination. Mesopic pupil size was determined with the Colvard pupillometer, with surrounding background room illumination of approximately 0.15 lux as measured with a luxmeter by well-trained optometrists.
The handheld Colvard pupillometer uses light amplification technology. The patient is asked to fixate on a red light produced by an infrared LED inside the device, and the examiner is able to focus the iris and pupil by moving the pupillometer slightly forward and backward. A millimeter ruler is superimposed by a reticule in the device over the image and allows direct measurement. The examiner was instructed to estimate the size of the horizontal pupil diameter to within 0.25 mm. The other eye was covered by the patient's hand during the measurement.
Pachymetry and Keratometry
Corneal thickness (thinnest point) and average keratometry readings were determined either with a scanning-slit optical device (Orbscan IIz; B&L, Rochester, NY) or a high-resolution rotating Scheimpflug imaging system (Pentacam; Oculus, Wetzlar, Germany).
After the data were compiled, they were entered into a spreadsheet program (Excel, Microsoft Corp, Redmond, WA) and further statistically analyzed with SPSS software (version 17.0; SPSS, Chicago, IL). There were five independent variables: gender, age, refractive state, keratometry, and pachymetry. Age, refractive state, average keratometry, and pachymetry were treated as continuous variables. All continuous variables were tested, and they showed a normal distribution. Thus, mean values were compared by the Student's t-test or univariate variance analysis (ANOVA) or covariance analysis (ANCOVA). In addition, the continuous variables, age and refractive state, were further divided into subgroups. Refraction groups were formed with respect to preoperative spherical equivalent and cylinder: (1) myopic astigmatism (SE <0.0 dpt and cylinder <3.00 dpt); (2) hyperopic astigmatism (SE >0.0 dpt and cylinder <3.00 dpt); and (3) high astigmatism (SE all and cylinder >3.0 dpt). Age was divided into three subgroups: group 1, <39 years; group 2, 40 to 54 years; and group 3, >55 years. One eye per subject was analyzed after being randomly selected using random generator of SPSS software. The possible differences in pupil size between right and left eyes were tested with Student's t-test. No significant difference was found. For statistical analysis, data description was based on arithmetic means and standard deviation of the respective pupil diameter measurements. Also medians and quartiles were shown for further descriptive comparison. Distributions were compared by ANOVA F-Test. Statistical analyses of individual factors' correlation with mesopic pupil size were done with Pearson correlation test. A p value < 0.05 was considered statistically significant. Multivariate regression analysis was performed to calculate the coefficient of determination R2 as an estimate of predicted pupil size. Only variables with a p < 0.05 were entered into, or allowed to remain in, the stepwise multivariate regression analysis.
This study included 13,959 eyes of 13,959 patients. Mean age was 36.07 years (ranging from 18 to 74 years). Mean overall mesopic pupil diameter (mean ± standard deviation) was 6.45 ± 0.82 mm. All eyes were treated independently because of the random selection of one eye per patient. Mean mesopic pupil size of the right eye was 6.44 ± 0.82 mm and of the left eye 6.46 ± 0.82 mm. The difference between the two eyes was statistically not significant (p = 0.197; t-test) as shown in Fig. 1.
Mean mesopic pupil size was 6.45 ± 0.84 mm for men and 6.44 ± 0.81 mm for women, and the difference between the two genders was not significant (p = 0.336; t-test) as shown in Fig. 2.
The eyes were divided into three groups according to their refractive states: myopic astigmatism (n = 11,721), hyperopic astigmatism (n = 1321), and (high) astigmatism (n = 917). Mean pupil diameter was 5.96 ± 0.8 in hyperopic astigmatism, 6.36 ± 0.83 mm in (high) astigmatism, and 6.51 ± 0.8 mm in myopic astigmatism. The descriptive data and results regarding mesopic pupil size in refractive subgroups are shown in Table 1 and Fig. 3.
The subjects were then divided into three age-groups: group 1 included age < 39 years (n = 8749); group 2, age 40 to 54 years (n = 4667); and group 3, age > 54 years (n = 543). Mean pupil diameter was 6.66 ± 0.79 for age < 39 years, 6.14 ± 0.74 mm for age 40 to 54 years, and 5.7 ± 0.74 mm for age > 54 years. The descriptive data and results are shown in Table 2 and Fig. 4, respectively.
The three refractive groups (myopic astigmatism/hyperopic astigmatism/(high) astigmatism) were further compared with regard to difference in mean age with ANOVA analysis. Mean age in hyperopia subgroup was 44.97 years, in myopia subgroup 35.03 years, and in astigmatism group 36.51 years. ANOVA showed that mean age of the three groups was significantly different (p < 0.001). This is summarized in Table 3 and Table 4.
In addition, ANCOVA was carried out, taking age as a covariate to show whether age was a probable factor causing the difference in mean scotopic pupil diameter among the three refractive groups. Pairwise ANCOVA revealed that the difference in mesopic pupil size between all refraction groups (p < 0.001) persisted even after age was taken as a covariate. The results of ANCOVA analysis with age as covariate are shown in Table 5.
Bivariate correlations between mesopic pupil size and spherical equivalent, age, average keratometry, and minimal pachymetry were performed via Pearson correlation analysis. Correlation coefficients and p values are shown in Table 6.
Age revealed the strongest correlation to mesopic pupil size in Pearson correlation analysis. Spherical equivalent had a moderate correlation with mesopic pupil size. Correlation between mesopic pupil size and average keratometry/minimal pachymetry was statistically (p < 0.001) significant, but this correlation is only weak (r = −0.064 for keratometry and r = −0.057 for pachymetry) and clinically insignificant. To analyze whether any of the variables that showed a statistically significant association with mesopic pupil size in univariate analysis contributed independently to the variability of these measurements, multivariate regression analyses were performed. Age, refraction (preoperative spherical equivalent), average keratometry, pachymetry, and ocular side were defined as dependent variables. The multivariate regression model (R2 = 0.179 and p < 0.001) showed that all four variables except ocular side had statistically significant effects on the model. Standardized beta coefficient reveals that average keratometry and pachymetry have only a marginal impact on mesopic pupil size, whereas age (beta coefficient −0.38) has the strongest impact on mesopic pupil size of the factors analyzed. Standardized and unstandardized coefficients of these variables are demonstrated in Table 7.
The scattergram in Fig. 5 reveals the relationship between mesopic pupil size and age controlling for refraction state (R2 linear = 0.171).
Since the beginnings of laser refractive surgery >20 years ago, steady improvements in Excimer technology and preoperative diagnostic tools have helped to optimize the treatment outcomes. The severity of quality of vision complaints with small optical zones (OZ) <5.5 mm, especially at night, led surgeons and manufacturers to increase the ablation zone diameter to >6.00 mm.12
Although several studies4,7,10 concluded that large pupil size does not positively correlate with any of postoperative visual symptoms 12 months after LASIK, the role of mesopic pupil size as a risk factor in predicting night vision complaints is still controversial. In general, optical models, such as point spread function calculations, suggest that pupil size should be an important determinant in selecting appropriate ablation zone size for treatment13,14 However, visual perception is multifactorial, and optical models cannot account fully for adaptive mechanisms. Some of these adaptive properties have been defined, such as the Stiles–Crawford effect (directional sensitivity of the retina), whereas others are less well understood. Bühren et al.15 demonstrated that not mesopic pupil size itself but the ratio between the planned OZ and the pupil size (FC) should be considered before treatment. Although large OZ diameters would be desirable, their usefulness is limited because of higher ablations depths required, which carry the risk for iatrogenic keratectasia.16 However, we and others6,15 think that it is counterintuitive to disregard pupil size as a potential risk factor regarding night vision complaints.
Most refractive surgeons agree that a standardized measurement of pupil size under defined background illumination conditions with good repeatability is of importance to screening patients eligible for refractive surgery.17,18 Most Excimer laser platforms base their experience and normograms on an optical zone of 6.0 or 6.5 mm, which correspond to scotopic pupil aperture of 5.5 to 6.0 mm. In our study, we used the Colvard pupillometer to determine mesopic pupil size, and to the best of our knowledge, we present the largest cohort of pupil size measurements with Colvard pupillometer (n = 13,959). Originally, the Colvard device was reported to measure scotopic pupil size,19 but recent literature shows best comparison with Procyon low mesopic pupil size measurement: Bootsma et al.20 showed a stronger correlation of Colvard pupillometry with Procyon measurements taken under mesopic-low light conditions than with Procyon measurements taken under scotopic conditions. Measurements of pupil size taken with the Colvard pupillometer are monocular, and the accuracy may therefore be affected by bilaterality because it is a handheld device.21 On the other hand, Kohnen et al.18 did not show that bilateral pupil size measurement with the Colvard device has a negative measuring bias and results are comparable with a digital (Procyon) device, but the authors summarize that digital pupillometers currently provide the most accurate measurement of the scotopic pupil. For this reason, we analyzed the right and left eye separately and found no significant difference in mean mesopic pupil size between right and left eye (p = 0.197; Student's t-test). In addition, no significant difference in mean mesopic pupil size between male and female subjects was found (p = 0.336; Student's t-test).
The overall mean mesopic pupil size in this large population of refractive surgery candidates was 6.45 ± 0.82 mm (mean age: 36.07 years; n = 13,959), and it is therefore in the middle of the range of the previous reports on scotopic/mesopic pupil size measurement with the Colvard pupillometer.
Kohnen et al. reported a mean Colvard pupil size of 5.78 ± 0.98 mm (1 min dark adaptation and 0.28 lux background illuminance) in a study of 100 eyes of subjects with a mean age of 38.8 years18 and of 5.68 ± 1.07 mm (2 min dark adaptation and background illuminance 18.33 lux) in another study of 100 eyes of subjects with a mean age of 36.8 years.22 In our opinion and according to literature,3,23 these mean values are smaller than expected in the reported age-groups. Bootsma et al. measured a mesopic (Colvard) pupil size of 5.86 ± 0.81 (background illuminance 0.15 lux) in an older population (n = 121) with a mean age of 48.8 years.20 Schmitz et al. reported a mean scotopic pupil size of 6.61 mm in a cohort of 56 subjects with a mean age of 23 years (2 min dark adaptation and background illuminance “complete darkness”),24 which is comparable with our mean mesopic pupil size of 6.66 mm in the <39 years subgroup. For all these pupil size measurements, the mean age of the study population must be taken into consideration. In addition, Bradley et al. reported that a steep and variable learning curve with the Colvard pupillometer is another source for potential bias.25 However, because of the large number of measurements performed in our study, this learning curve should not have a significant impact on the outcome. Schallenberg et al. found no significant change of pupil diameter measured with Colvard pupillometer after 2 min or more of dark adaptation.26
Recently, Cakmak et al. analyzed ocular factors that may affect mesopic pupil size.11 Pupil size measurements were performed in their study with a COAS Ocular Wavefront analyzer in mesopic conditions (0.6 lux). Although this is not a standard method to determine mesopic pupil size, previous studies reported that pupil size measurements with wavefront aberrometers were comparable with standard Colvard and Procyon pupillometers.20,24,27 Cakmak et al. reviewed records of 412 refractive surgery candidates and calculated mean mesopic pupil size (COAS Ocular Wavefront analyzer) to 6.18 mm (mean age: 31.85 years). Cakmak et al. showed that age and magnitude of both spherical and cylindrical refractive error are determinative factors on mesopic pupil size. Pupil size was larger in myopic patients than in hyperopic patients. As the magnitude of spherical refractive error increases, pupil size decreases. It was also confirmed that age was an important factor for mesopic pupil size with a smaller pupil size as the age advances.11
In our large-scale study, mean mesopic pupil size (Colvard pupillometer) was 5.96 ± 0.8 mm in hyperopia, 6.36 ± 0.83 mm in astigmatism, and 6.51 ± 0.8 mm in myopia. The differences in mesopic pupil size between all the refractive groups (myopic astigmatism, hyperopic astigmatism, and high astigmatism) were statistically significant (p < 0.001), and persisted even after age was taken as a covariate (ANCOVA). From the clinical point of view, it must be mentioned that although statistically significant, the difference between myopia and astigmatism regarding mesopic pupil size was only marginal. Correlation analysis showed that not only age (r = −0.405 p < 0.001) and refractive state but also average keratometry and pachymetry (thinnest point) had a statistically significant effect on mesopic pupil size, but this correlation must be considered clinically insignificant. We could show that the sum of all the factors (age, refractive state, keratometry, and pachymetry) analyzed can only predict the expected pupil size in <20% cases because R2 value was considerably low (0.179). On the other hand, >80% of the determinative factors and their impact on mesopic pupil size are not explained by our analysis. Last, but not least, pupil size is influenced by intra-individually variable factors such as alertness, diet, drug use, activity, sleep, and lifestyle factors.28 These effects appear to be limited because Robl et al. showed that using a standardized dark-adaptation protocol, refractive surgery candidates did not show a significant variation in scotopic pupil diameter for two testing sessions.29
To our knowledge, we present the largest study including 13,959 randomly selected eyes of refractive surgery candidates and analyzing the effect of ocular and general parameters on mesopic pupil size.
In summary, the present study has revealed and confirmed that age is the strongest determinative factor on mesopic pupil size in refractive surgery candidates. In addition, magnitude of refractive error and astigmatism also had a statistical significant impact on mesopic pupil size. Although statistically significant, the impact of average keratometry and minimal pachymetry on mesopic pupil size is only marginal and should be further analyzed in future studies.
However, the smaller the impact of individual factors on mesopic pupil size, the more important is the standardization of pupil size measurement with regard to dark adaptation and background illuminance. Prospective studies are warranted.
Stephan J. Linke
Department of Ophthalmology
The authors thank the staff and patients of Care Vision for their support in establishing the anonymized refractive data collection (Hamburg Refractive Data Base). The authors are particularly grateful to Vasyl Druchkiv for supporting and supervising the database and expert statistical analysis. The authors have no financial or proprietary interest in this study. No financial support was received.
No grant support or financial funding was received. All authors do not have, direct or indirect, commercial or proprietary, interest in any device, equipment, instrument, drug, or company mentioned in the article, or in any device competing against any device, equipment, instrument, drug, or company mentioned in the article.
1. Wyatt HJ, Musselman JF. Pupillary light reflex in humans: evidence for an unbalanced pathway from nasal retina, and for signal cancellation in brainstem. Vision Res 1981;21:513–25.
2. Marg E, Morgan MW Jr. The pupillary near reflex; the relation of pupillary diameter to accommodation and the various components of convergence. Am J Optom Arch Am Acad Optom 1949;26:183–98.
3. Kadlecova V, Peleska M, Vasko A. Dependence on age of the diameter of the pupil in the dark. Nature 1958;182:1520–1.
4. Schallhorn SC, Kaupp SE, Tanzer DJ, Tidwell J, Laurent J, Bourque LB. Pupil size and quality of vision after LASIK. Ophthalmology 2003;110:1606–14.
5. Haw WW, Manche EE. Effect of preoperative pupil measurements on glare, halos, and visual function after photoastigmatic refractive keratectomy. J Cataract Refract Surg 2001;27:907–16.
6. Salz JJ, Trattler W. Pupil size and corneal laser surgery. Curr Opin Ophthalmol 2006;17:373–9.
7. Chan A, Manche EE. Effect of preoperative pupil size on quality of vision after wavefront-guided LASIK. Ophthalmology 2011;118:736–41.
8. Fan-Paul NI, Li J, Miller JS, Florakis GJ. Night vision disturbances after corneal refractive surgery. Surv Ophthalmol 2002;47:533–46.
9. Lee YC, Hu FR, Wang IJ. Quality of vision after laser in situ keratomileusis: influence of dioptric correction and pupil size on visual function. J Cataract Refract Surg 2003;29:769–77.
10. Bailey MD, Mitchell GL, Dhaliwal DK, Boxer Wachler BS, Zadnik K. Patient satisfaction and visual symptoms after laser in situ keratomileusis. Ophthalmology 2003;110:1371–8.
11. Cakmak HB, Cagil N, Simavli H, Duzen B, Simsek S. Refractive error may influence mesopic pupil size. Curr Eye Res 2010;35:130–6.
12. Hassan Z, Lampe Z, Bekesi L, Berta A. Excimer laser photorefractive keratectomy with different ablation zones. Acta Chir Hung 1997;36:122–4.
13. Roberts CW, Koester CJ. Optical zone diameters for photorefractive corneal surgery. Invest Ophthalmol Vis Sci 1993;34:2275–81.
14. Endl MJ, Martinez CE, Klyce SD, McDonald MB, Coorpender SJ, Applegate RA, Howland HC. Effect of larger ablation zone and transition zone on corneal optical aberrations after photorefractive keratectomy. Arch Ophthalmol 2001;119:1159–64.
15. Bühren J, Kuhne C, Kohnen T. Influence of pupil and optical zone diameter on higher-order aberrations after wavefront-guided myopic LASIK. J Cataract Refract Surg 2005;31:2272–80.
16. Seiler T, Quurke AW. Iatrogenic keratectasia after LASIK in a case of forme fruste keratoconus. J Cataract Refract Surg 1998;24:1007–9.
17. Rosen ES, Gore CL, Taylor D, Chitkara D, Howes F, Kowalewski E. Use of a digital infrared pupillometer to assess patient suitability for refractive surgery. J Cataract Refract Surg 2002;28:1433–8.
18. Kohnen T, Terzi E, Buhren J, Kohnen EM. Comparison of a digital and a handheld infrared pupillometer for determining scotopic pupil diameter. J Cataract Refract Surg 2003;29:112–7.
19. Colvard M. Preoperative measurement of scotopic pupil dilation using an office pupillometer. J Cataract Refract Surg 1998;24:1594–7.
20. Bootsma S, Tahzib N, Eggink F, de Brabander J, Nuijts R. Comparison of two pupillometers in determining pupil size for refractive surgery. Acta Ophthalmol Scand 2007;85:324–8.
21. Khanani AM, Brown SM. Determining scotopic pupil size. J Cataract Refract Surg 2005;31:1266–7.
22. Kohnen T, Terzi E, Kasper T, Kohnen EM, Buhren J. Correlation of infrared pupillometers and CCD-camera imaging from aberrometry and videokeratography for determining scotopic pupil size. J Cataract Refract Surg 2004;30:2116–23.
23. Borgmann H. [Basic data for clinical pupillography. II. The age-dependence of the pupil diameter in darkness]. Albrecht Von Graefes Arch Klin Exp Ophthalmol 1972;184:300–8.
24. Schmitz S, Krummenauer F, Henn S, Dick HB. Comparison of three different technologies for pupil diameter measurement. Graefes Arch Clin Exp Ophthalmol 2003;241:472–7.
25. Bradley JC, Anderson JE, Xu KT, Brown SM. Comparison of Colvard pupillometer and infrared digital photography for measurement of the dark-adapted pupil diameter. J Cataract Refract Surg 2005;31:2129–32.
26. Schallenberg M, Bangre V, Steuhl KP, Kremmer S, Selbach JM. Comparison of the Colvard, Procyon, and Neuroptics pupillometers for measuring pupil diameter under low ambient illumination. J Refract Surg 2010;26:134–43.
27. Cheng AC, Lam DS. Comparison of the Colvard pupillometer and the Zywave for measuring scotopic pupil diameter. J Refract Surg 2004;20:248–52.
28. Loewenfeld IE, Newsome DA. Iris mechanics. I. Influence of pupil size on dynamics of pupillary movements. Am J Ophthalmol 1971;71:347–62.
29. Robl C, Sliesoraityte I, Hillenkamp J, Prahs P, Lohmann CP, Helbig H, Herrmann WA. Repeated pupil size measurements in refractive surgery candidates. J Cataract Refract Surg 2009;35:2099–102.
mesopic pupil size; Colvard pupillometer; refractive state; age; pachymetry