The origins of the increase in the prevalence of myopia in many parts of the world remain elusive. Although various factors have been proposed to account for this change, near work and time spent on outdoor activity are so far the only environmental factors that have been definitely linked to myopia development and progression.1–5
One factor that could contribute to refractive change is the eye movement associated with the performance of visual tasks. Two important studies imply that myopia progression is not necessarily caused by high accommodative demands. Adams and McBrien6 studied a group of microscopists and Simensen and Thorud7 analyzed myopia development and progression in textile workers. The microscopists6 looked at images that were placed nominally at infinity and the textile workers7 inspected cloth for weaving errors at a distance of approximately 50 cm (see Goldschmidt8 for illustration). The results showed that, despite the low accommodative demands, the microscopists and textile workers tended to develop myopia and to progress at a rate higher than that usually reported in the literature.9–11 These results therefore suggest that special types of posture or eye movements could contribute to myopization. Eye movements in reading combined with lid pressure have been shown to cause corneal distortion, which could lead to myopization,1,12–14 although limited experimental studies by our group have so far failed to find any major differences in eye movements and posture in emmetropes and myopes when carrying out specific tasks.15,16
In the present study, we studied saccades because saccadic patterns might be characteristic for refractive error groups such as myopes and emmetropes. Saccades are rapid eye movements that bring points of interest onto the fovea (e.g., Robinson17). Under normal conditions, saccades may occur up to several times per second in humans, predominantly along the horizontal meridian. They allow visual sampling of the environment, and retinal stability is maintained through microsaccades.18 We studied saccades because of their importance in many visual tasks. In addition, our hypothesis was that the stresses associated with their high frequency and velocity might produce stronger distorting effects on the globe than those of slower eye movements.
We note that any study comparing existing myopes with emmetropes cannot differentiate between whether any differences in saccades (or any other characteristic) are the cause of the myopia or are simply its effect. It might be, for example, that once the myopia had developed, the longer and heavier eyeball, associated perhaps with subtle differences in the attachment of the extraocular muscles or greater lid pressures, affects saccades, possibly slowing them slightly. Nevertheless, demonstration of differences in the patterns of saccades between different refractive groups would form at least a suggestive first step toward exploring the significance of such effects in relation to refractive development.19
One earlier study20 compared the saccadic eye movements of a group of myopes with those of emmetropes. The authors found differences which they felt were primarily associated with the wear of different types of correction by the myopes. Peak saccadic velocity was determined for amplitudes 7.5, 15, 22.5, and 30 degrees using electro-oculography electrodes. The myopic subjects were corrected by glasses or contact lenses. Their strongest finding was that myopes with higher prescriptions (myopia >6 D) seemed to have slower peak saccadic velocities than emmetropes, particularly when corrected with contact lenses: myopes with corrections of less than 6 D did not differ from emmetropes. From the optical point of view, the result with high myopes is surprising, in that spectacle magnification considerations would lead to the expectation that spectacle-corrected myopes would have to make a smaller saccade and hence that their peak eye velocity would be lower than contact lens–corrected myopes, whose spectacle magnification effects are much smaller.
In the light of these suggestive but slightly perplexing results, which do not seem to have been followed up, the present study aimed to reexamine the characteristics of saccadic eye movements in myopes and emmetropes. In particular, because we hypothesized that a longer eye might move more slowly, owing to its greater mass and reduced orbital space, than a shorter eye, we wished to determine whether saccadic performance depended systematically on the degree of myopia or axial length.
Twenty-eight participants (15 men and 13 women) were recruited. The age of the subjects ranged between 19 and 39 years (mean [SD], 27.0 [4.7] years). All subjects were free of any ocular pathologic abnormality and could achieve a visual acuity of 6/6 or better when corrected. Best-sphere corrections were in the range of −7.13 D to +0.50 D (mean, −1.55 ± 2.11 D). The group included 14 myopes (spherical equivalent power between −7.13 D and −0.88 D; mean, −3.10 ± 1.99 D) and 14 emmetropes (spherical equivalent power between −0.50 D and +0.50 D; mean, 0.01 ± 0.38 D). In all cases, astigmatism was less than 1.50 D. The mean ages of the myopes and emmetropes were 27.6 (5.1) and 26.5 (4.3) years, respectively. Myopic subjects were corrected with their habitually worn, spherical, soft contact lenses because spectacles induced additional reflections that did not allow precise eye tracking. The study followed the tenets of the Declaration of Helsinki, and written informed consent was obtained from all participants after the nature of the study and possible consequences of the study had been explained. The project protocol was approved by the Senate Committee on the Ethics of Research on Human Beings of the University of Manchester.
Stimuli and Procedures
Eye movements were recorded while the participants alternated their binocular fixation between the centers of two crosses that were horizontally separated by 20 degrees and were on the same vertical level. Head movements were stabilized using a chin rest. The widths of the limbs of the crosses subtended 0.02 degrees, the total lengths of the horizontal and vertical limbs being about 2.6 degrees. The crosses were shown alternately for 2 seconds each on a monitor at a distance of 55 cm from the corneal plane of the subject and were presented in a consecutive and therefore predictable order. Participants were instructed to refixate immediately when they were aware that the target had changed, so that the horizontal eye movement record approximated to a square wave of a 4-second period and 10-degree amplitude to each side of the median line. Each target was presented eight times, which resulted in 15 eye movements between the two targets.
Note from Figure 1 that the nature of the eye turn made in each fixation differed between the two eyes and that it depended slightly on the pupillary distance (PD). With a PD of 65 mm, the left eye had a left turn &thetas;1 = 6.5 degrees when fixating target A and a right turn &thetas;2 = 12.9 degrees when fixating target B. The eye therefore had to turn through 19.4 degrees, when altering between the two targets. Because the range of PDs in our subjects was such that differences in the necessary rotation angles were minor, we did not correct for this effect but assumed in the calibration process that 20 degrees of eye rotation was required for each fixational movement.
An Eye Link II (SR Research Ltd., Mississauga, Canada) was used in conjunction with Motion Monitor software (Innovative Sports Training Inc., Chicago, IL) to monitor the saccades. Eye movements were recorded binocularly at 500 Hz in horizontal and vertical planes using two video-based cameras mounted on a helmet. The total weight of the helmet and associated head-mounted equipment was 420 g. The cameras were set for pupil tracking mode. The data sets were calibrated using a MATLAB script. After the subjects had been familiarized with the task, two complete data sets of movements were recorded: the first set being used for calibration; the second, for analysis. Checks showed no significant differences between the calibrations produced by the two data sets.
A LenStar LS 900 (Haag Streit, Koeniz, Switzerland) was used to measure the axial length of the right eye. The device has been found to be repeatable and valid.21–25 Five readings were taken for each participant, and the internal software calculated the mean of five readings automatically. This mean was used for further analysis.
To determine the refractive error, a subjective refraction was performed on all subjects to an accuracy of ±0.25 DS and ±0.25 DC. The cylindrical component was found, if existent, using a cross-cylinder. To refine the spherical component at the end of the routine, the duochrome test was used. All subjective refractions were performed by one of the authors (A.H.). For further analysis, the spherocylindrical result was converted into a spherical equivalent.
Figure 2 represents a typical eye movement recording. For some fixational movements, there is undershooting or overshooting of the primary saccade, which is then followed by a small corrective movement. In the right-eye case illustrated, undershoots are more common during abduction (rightward fixation change) than adduction (leftward eye movement).
Only the horizontal data from right eyes were analyzed. A customized MATLAB code (The MathWorks Inc., Natick, MA) was used to calculate the following parameters for each participant:
- (i) Mean durations of rightward and leftward main saccades as well as all saccades (right and leftwards) combined
- (ii) Mean amplitudes of rightward and leftward main saccades as well as all saccades combined
- (iii) Mean peak velocities of rightward and leftward main saccades as well as all saccades combined
- (iv) The percentages of undershoots, overshoots, and exact fixations among the main saccades
The starting point of the main saccade was selected manually and was taken as the instant at which the eye movement started, the end point as the instant when either a secondary eye movement with markedly different temporal characteristic commenced (i.e., a corrective movement) or exact fixation was directly established at the end of the main saccade.
All statistical analysis was performed using SPSS 16.0 (SPSS Inc., Chicago, IL). Between-groups analysis of variance was used to compare differences between myopes and emmetropes. Mixed analysis of variance was used to check for interactions between repeats (in terms of rightward and leftward) and refractive error groups. Two-tailed Pearson correlations were applied for comparisons in reference to refractive error and axial length. For multiple comparisons, the conservative Bonferroni adjustment was used.
Axial Length and Refractive Error
These are shown for the two refractive groups in Figure 3. Although in the myopic group there is the expected tendency for more myopic eyes to have longer axial lengths (Pearson product-moment correlation for myopes and emmetropes combined: r = −0.80, p < 0.001), the data for the emmetropes emphasize that a substantial range of axial lengths may be associated with the same value of refractive error.
Characteristics of Main Saccades
Table 1 shows the main characteristics of saccades in myopes and emmetropes. As can be seen from Table 1, any differences between the two refractive groups were minor and failed to reach statistical significance (taken initially as p < 0.05), particularly if multiple testing was taken into account (p < 0.006 required when Bonferroni correction for multiple comparisons was included).
For the three parameters (peak velocity, amplitude, and duration of saccades), no significant differences were found between abduction and adduction movement when using a mixed analysis of variance, where rightward and leftward movements were the dependent variables and the refractive error group was the independent variable (duration: F1,25 = 2.18, p = 0.15; amplitude: F1,25 = 0.02, p = 0.89; peak velocity: F1,25 = 0.03, p = 0.87).
Proportions of Corrective Movements
The percentages of undershoots, overshoots, and exact fixations among the main saccades made by the two refractive groups are shown in Table 2. Emmetropes had relatively more undershoots than myopes, as would be expected on the basis of their lower main saccade amplitudes (Table 1). However, no statistically significant differences were found between myopes and emmetropes for any of the categories in Table 2 (between-groups ANOVA, p > 0.05).
Correlations Between Saccadic Parameters, Axial Lengths, and Refractions
When emmetropes and myopes were analyzed together, correlations between refractive error or axial length and saccadic parameters such as duration, amplitude, and maximum velocity were not significant (p > 0.05). Analyzing myopes and emmetropes separately gave no significant correlations for emmetropes; but in myopes, the correlations of peak velocity with refractive error and axial length were significant at the p = 0.05 level (Pearson product-moment correlation: refractive error, r = −0.65, p = 0.01; axial length, r = 0.55, p = 0.04; Fig. 4). Peak velocity increased with axial length and the degree of myopia. However, because we explored a substantial number of possible correlations (n = 18), Bonferroni correction would render these correlations not significant (p < 0.003 required). Further, axial length and refractive error in myopes are themselves correlated (Fig. 3). Nonetheless, Figure 4A raises the possibility that some differences in peak velocity may occur between the two groups.
The correlations for duration and amplitude separated for myopes and emmetropes were nonsignificant (p > 0.05).
The present study evaluated differences in saccadic eye movements between groups of myopes and emmetropes using a predictive saccade paradigm. When considering the whole study population, no significant differences were found between the two refractive groups, nor were any of the saccadic parameters correlated with axial length, which ranged quite widely, from 22.2 to 27.5 mm, or mean spherical refractive error (range, −7.13 D to +0.50 D). Thus, the results offered no support for the hypothesis that the generally greater length of the myopic eye or any other difference affected saccadic velocities or, indirectly, that saccadic differences might be involved in myopization. When possible correlations between the saccadic parameters, axial length, and refraction were studied separately for the myopic and emmetropic groups, the only result of marginal significance was that, in myopes, peak velocity correlated with refractive error and axial length. When Bonferroni correction due to multiple comparisons is considered, these results are nonsignificant. In the literature, there is disagreement between the importance of applying corrections for multiple comparisons. For example, Perneger26 and Rothman27 argue that corrections for multiple comparisons are not necessary, whereas Bender and Lange28 defend the importance of using adjustments. Overall, the results give no support for the influence of predictive saccadic eye movements on myopization.
The data, obtained for 20-degree saccades, do not directly support the results of Müller et al.,20 although their task was similar to that in our experimental setup. They suggested that contact lens–corrected myopes had saccades of significantly lower peak velocity than emmetropes, for a saccade amplitude of 22.5 degrees (Student t test on their data, p = 0.024 for all myopes; not significant for myopes < 6 D [p > 0.05] and significant for myopes > 6 D [p = 0.001]). We note, however, that Müller et al. took p = 0.05 as their level of significance despite multiple testing (36 comparisons). If a Bonferroni correction had been applied, it would seem that their result was significant only for the higher myopes. Our subjects included only 5 myopes with corrections >6 D, so it may be that there are differences for very high levels of myopia, which we could not detect through a lack of suitable subjects. The peak velocities recorded by Müller et al. using their electro-oculography technique were, however, substantially lower than ours: around 280 deg/s for a 22.5-degree saccade compared with our value of about 460 deg/s for a 20-degree saccade. Therefore, some doubts remain about the validity of their findings owing to the use of the electro-oculography device. A possible reason might be the recording frequency of 30 Hz, whereas we recorded the saccades at a frequency of 500 Hz. In our study, the use of contact lenses in the myopes could, in principle, have had an impact on the results of the measurements by affecting, for example, lid-eye interactions. However, the soft lens were characterized by minimal thickness, and the absence of significant differences between the saccades of the two refractive groups suggests that they had no detectable influence on the experimental results.
In general, the characteristics of the saccades listed in Table 1 are similar to those found elsewhere in the literature for saccades of similar amplitude. We found a greater prevalence of hypometria (undershooting) than hypermetria (overshooting).29,30 The mean undershoot of about 0.8 degrees is similar to that found by Collewijn et al.30 and Bötzel et al.31 under similar experimental conditions. Henson32 reasoned that undershooting keeps the visual target at the same side of the fovea, which might lead to greater precision and shorter latency for the corrective saccade. The peak velocities in the present study (about 460 deg/s) are also in the range of earlier findings for 20-degree saccades. For example, Boghen et al.33 found a peak velocity of 375 deg/s and Baloh et al.34 reported a peak velocity of 420 deg/s, while Bahill et al.35 found a much higher peak velocity of 657 deg/s.
As noted earlier, the present study does not support the hypothesis that the characteristics of saccades between two fixed points are markedly different in existing myopes as compared to emmetropes, at least for the range of myopia and axial lengths studied (Table 1). This makes it unlikely that saccadic velocity, as such, could play any role in myopization. Our study involved a fixed head position and repetitive saccades. It still remains possible that the mixture of head movements, saccades, and other eye movements used when carrying out a more complex visual task by emmetropes who were at risk of developing myopia might differ from that of those emmetropes whose refractions remained stable. However, we have, as yet, found it difficult to demonstrate differences in head and eye movements between emmetropes and existing myopes when carrying out simple reading and writing tasks.15,16
The present investigation had a number of limitations. It studied regular binocular saccades between targets separated by 20 degrees, placed symmetrically about the midline. The predictable and repetitive nature of our task allowed participants to easily plan ahead and quickly learn to correct any errors, potentially reducing any differences between the groups. Differences in saccadic characteristics might have emerged had a more challenging experimental paradigm been used, involving, for example, using nonpredictable targets presented at different amplitudes, positions, and times. It is also possible that if the saccadic task was continued over longer periods, as in real-life situations, greater differences might emerge. As discussed by Tatler et al.36 salience-based fixation targets do not reflect natural conditions. In a previous work,15 we tried to analyze more natural conditions, such as reading. There we did not find differences in eye movements between myopes and emmetropes. However, for future work, the analysis of saccades made under more natural conditions should be considered.
It could also be argued that, although saccadic characteristics might be similar, the myopic or potentially myopic eyeball is more susceptible to the external stresses associated with eye movement, thus causing small changes in axial length and refraction to occur after lengthy sequences of movement. Such possibilities deserve further investigation. In addition, we only used 20-degree saccades, and it remains possible that differences would have emerged using smaller saccades owing to the momentum of a larger eye.
A further, more fundamental, limitation of this study is that the data were obtained on adults rather than children who are likely to progress at a much faster rate. However, there is evidence in the literature that myopia progression is not only dependent on age. For instance, works from Adams and McBrien6 and Simensen and Thorud7 indicate that myopia progression can onset in adults as well.
A final limitation might be the use of a head-mounted eye tracker and chin rest. These involve artificial conditions that constrain head movements and hence disturb an individual’s usual mixture of head and eye movements when changing fixation. A future study on young progressing myopes with a helmet-free technique for tracking eye movements would be useful for studying eye movements under more natural conditions.
To summarize, our present results show that saccadic eye movements of the same amplitude are similar in emmetropes and existing myopes. Hence, they offer no support for the hypothesis that differences in saccadic characteristics are implicated in myopization.
Received: January 31, 2012; accepted May 10, 2013.
1. Collins MJ, Buehren T, Bece A, Voetz SC. Corneal optics after reading, microscopy and computer work. Acta Ophthalmol Scand 2006; 84: 216–24.
2. Wu PC, Tsai CL, Hu CH, Yang YH. Effects of outdoor activities on myopia
among rural school children in Taiwan. Ophthalmic Epidemiol 2010; 17: 338–42.
3. Rose KA, Morgan IG, Smith W, Burlutsky G, Mitchell P, Saw SM. Myopia
, lifestyle, and schooling in students of Chinese ethnicity in Singapore and Sydney. Arch Ophthalmol 2008; 126: 527–30.
4. Rose KA, Morgan IG, Ip J, Kifley A, Huynh S, Smith W, Mitchell P. Outdoor activity reduces the prevalence of myopia
in children. Ophthalmology 2008; 115: 1279–85.
5. Mutti DO, Mitchell GL, Moeschberger ML, Jones LA, Zadnik K. Parental myopia
, near work, school achievement, and children’s refractive error
. Invest Ophthalmol Vis Sci 2002; 43: 3633–40.
6. Adams DW, McBrien NA. Prevalence of myopia
and myopic progression in a population of clinical microscopists. Optom Vis Sci 1992; 69: 467–73.
7. Simensen B, Thorud LO. Adult-onset myopia
and occupation. Acta Ophthalmol (Copenh) 1994; 72: 469–71.
8. Goldschmidt E. The mystery of myopia
. Acta Ophthalmol Scand 2003; 81: 431–6.
9. Jacobsen N, Jensen H, Goldschmidt E. Does the level of physical activity in university students influence development and progression of myopia
?—A 2-year prospective cohort study. Invest Ophthalmol Vis Sci 2008; 49: 1322–7.
10. Saw SM, Gazzard G, Au Eong KG, Tan DT. Myopia
: attempts to arrest progression. Br J Ophthalmol 2002; 86: 1306–11.
11. Walline JJ, Jones LA, Sinnott L, Manny RE, Gaume A, Rah MJ, Chitkara M, Lyons S. A randomized trial of the effect of soft contact lenses on myopia
progression in children. Invest Ophthalmol Vis Sci 2008; 49: 4702–6.
12. Collins MJ, Buehren T, Trevor T, Statham M, Hansen J, Cavanagh DA. Factors influencing lid pressure on the cornea. Eye Contact Lens 2006; 32: 168–73.
13. Buehren T, Collins MJ, Carney L. Corneal aberrations and reading. Optom Vis Sci 2003; 80: 159–66.
14. Buehren T, Collins MJ, Carney LG. Near work induced wavefront aberrations in myopia
. Vision Res 2005; 45: 1297–312.
15. Hartwig A, Gowen E, Charman WN, Radhakrishnan H. Working distance and eye and head movements during near work in myopes and non-myopes. Clin Exp Optom 2011; 94: 536–44.
16. Hartwig A, Gowen E, Charman WN, Radhakrishnan H. Analysis of head position used by myopes and emmetropes when performing a near-vision reading task. Vision Res 2011; 51: 1712–7.
17. Robinson DA. The mechanics of human saccadic eye movement. J Physiol 1964; 174: 245–64.
18. Kowler E. Eye movements
: the past 25 years. Vision Res 2011; 51: 1457–83.
19. Foulsham T, Teszka R, Kingstone A. Saccade control in natural images is shaped by the information visible at fixation: evidence from asymmetric gaze-contingent windows. Atten Percept Psychophys 2011; 73: 266–83.
20. Müller C, Stoll W, Schmal F. The effect of optical devices and repeated trials on the velocity of saccadic eye movements
. Acta Otolaryngol 2003; 123: 471–6.
21. Holzer MP, Mamusa M, Auffarth GU. Accuracy of a new partial coherence interferometry analyser for biometric measurements. Br J Ophthalmol 2009; 93: 807–10.
22. Buckhurst PJ, Wolffsohn JS, Shah S, Naroo SA, Davies LN, Berrow EJ. A new optical low coherence reflectometry device for ocular biometry in cataract patients. Br J Ophthalmol 2009; 93: 949–53.
23. Cruysberg LP, Doors M, Verbakel F, Berendschot TT, De Brabander J, Nuijts RM. Evaluation of the LenStar LS 900 non-contact biometer. Br J Ophthalmol 2010; 94: 106–10.
24. Rohrer K, Frueh BE, Walti R, Clemetson IA, Tappeiner C, Goldblum D. Comparison and evaluation of ocular biometry using a new noncontact optical low-coherence reflectometer. Ophthalmology 2009; 116: 2087–92.
25. O’Donnell C, Hartwig A, Radhakrishnan H. Correlations between refractive error
and biometric parameters in human eyes using the LenStar 900. Cont Lens Anterior Eye 2011; 34: 26–31.
26. Perneger TV. What’s wrong with Bonferroni adjustments. BMJ 1998; 316: 1236–8.
27. Rothman KJ. No adjustments are needed for multiple comparisons. Epidemiology 1990; 1: 43–6.
28. Bender R, Lange S. Adjusting for multiple testing—when and how? J Clin Epidemiol 2001; 54: 343–9.
29. Weber RB, Daroff RB. The metrics of horizontal saccadic eye movements
in normal humans. Vision Res 1971; 11: 921–8.
30. Collewijn H, Erkelens CJ, Steinman RM. Binocular co-ordination of human horizontal saccadic eye movements
. J Physiol 1988; 404: 157–82.
31. Botzel K, Rottach K, Buttner U. Normal and pathological saccadic dysmetria. Brain 1993; 116 (Pt 2): 337–53.
32. Henson DB. Corrective saccades
: effects of altering visual feedback. Vision Res 1978; 18: 63–7.
33. Boghen D, Troost BT, Daroff RB, Dell’Osso LF, Birkett JE. Velocity characteristics of normal human saccades
. Invest Ophthalmol 1974; 13: 619–23.
34. Baloh RW, Sills AW, Kumley WE, Honrubia V. Quantitative measurement of saccade amplitude, duration, and velocity. Neurology 1975; 25: 1065–70.
35. Bahill AT, Brockenbrough A, Troost BT. Variability and development of a normative data base for saccadic eye movements
. Invest Ophthalmol Vis Sci 1981; 21: 116–25.
36. Tatler BW, Hayhoe MM, Land MF, Ballard DH. Eye guidance in natural vision: reinterpreting salience. J Vis 2011; 11: 5.