This study examined the effect of myopic defocus on visual acuity (VA) over time, with attention being paid to the first point at which blur adaptation had a significant and measurable effect on defocused VA. Visual acuity was sampled at a higher rate than previous studies in order to assess the time course of blur adaptation processes in myopic and emmetropic observers.
Participants were 24 normally-sighted observers (12 emmetropes and 12 myopes, median age: 22.5 years). All ametropic participants wore their full refractive correction throughout the experiment. 1 D and 3 D of myopic defocus were introduced in two separate, randomised sessions. Visual acuity was measured using Test Chart 2000 at 2 min intervals over a 30 min session whilst looking through defocus lenses. Recovery clear VA was also measured every 2 min for a further 20 min.
Defocused VA was found to improve significantly within 4 min after the introduction of defocus for both 1 D (P < 0.0001) and 3 D conditions (P < 0.0001). The improvements reached a plateau shortly after, with no significant further improvements in defocused VA after 6 min. There were no significant differences found in the temporal blur adaptation profiles between emmetropes and myopes (P = 0.267). Data were fitted with an exponential decay function; the lowest R2 value for this fit was 0.95.
Blur adaptation has a clinically significant and measurable effect on VA within 4 min of exposure to defocus. This finding indicates that the visual system instigates the neural compensatory mechanisms shortly after the appearance of defocus. Our results relate particularly to real-life vision of uncorrected myopes or myopes who remove their correction for part of the day.
Blur adaptation is a phenomenon which allows the visual system to compensate for optical defocus, and ultimately results in an improvement in defocused visual acuity (VA) without altering ocular refraction.[1, 2] Previously, it has been hypothesized that recalibration of spatial frequency processing channels by increasing the gain of high frequency selective channels and decreasing the gain of low frequency selective channels serves to improve resolution.[1, 3] Mankowska et al reported a new finding of blur adaptation occurring in the parafovea and proposed that neural calibration of spatial frequency channels is spread across a range of frequencies.
Some studies have revealed differences in response to defocus due to refractive status, namely that myopes demonstrate a greater degree of adaptation compared to emmetropes. Elevated blur detection thresholds after adaptation to defocus resulted in an improvement in low-contrast grating VA in myopic individuals more so than for any other refractive group.[5, 6] This may be due to increased tolerance or decreased sensitivity to blur compared to emmetropes.[7-10] In addition, a larger depth of focus may also contribute to this finding. Thorn et al postulated that emmetropes are generally more susceptible to change in VA and contrast sensitivity with the introduction of defocus. In contrast, measures of blur adaptation using high-contrast VA often result in similar performance between emmetropes and myopes.
The phenomenon of blur adaptation is widely known to cause a significant improvement in VA under defocused conditions after at least 30 min of exposure to a constant defocus environment. Previous studies have been conducted with various adaptation periods from 30 min to 3 h, and differing defocus levels from 1 D to 3 D of myopic defocus or myopes' uncorrected refractive error.[1, 2, 4, 6, 13-15] Findings from these studies showed improvements in VA under defocused conditions ranging from 0.04 to 0.27 logMAR. Until now, no study has examined the time course of the effects of blur adaptation at a sufficiently high temporal sampling rate to determine the onset and duration of these adaptive processes.
Previously, small changes in the perception of blur have been found to occur within a few minutes of exposure to defocus. They demonstrated significantly altered perception of blur and judgments of best focus produced by vivid adaptation after-effects. Participants were presented with physically blurred or sharpened images, and a staircase procedure was implemented to judge best focus of the image after an adaptation period of 3 min. From this, it can be postulated that the visual system responds quickly to changes in the visual environment and begins counteracting defocus at a much earlier time than 30 min. Rosenfield et al commented on findings that showed the robustness of the blur adaptation phenomenon, and provided evidence that its effects are extended over a certain period of time. They found that once adaptation was established, minimal loss of adaptation occurred, i.e. VA was maintained for a significant period of time after the adaptation period. This study used participants' habitual refractive error (1.50 D–3.00 D of myopia) which shows the effect of adaptation in real life. Another study reports that intervening periods of clear vision does not result in a change in the VA achieved after adaptation has been established. Once again, this indicates the robust nature of blur adaptation.
In the experiments reported here, we wished to ascertain how defocused VA changes with blur adaptation immediately after defocus is introduced, and in particular when blur adaptation first causes a significant improvement in defocused VA, i.e. how quickly does the visual system begin to respond and counteract the defocus, and when does this start to have a prominent effect on vision. We will investigate different defocus levels in addition to refractive group differences.
A total of 24 participants were recruited for the study; 12 emmetropes (median age: 22 years, range: 19–24 years) and 12 myopes (median age: 25.5 years, range: 18–41 years). Emmetropes were classified as having a spherical refraction of between +0.75 DS and −0.75 DS (mean spherical error: +0.10 ± 0.36 D). Myopes were defined as having a negative spherical refraction of at least −0.75DS (mean spherical error: −3.44 ± 2.32 D, range: −0.75–−8.25 DS). Participants had no more than 1.00D astigmatism in either refractive group. Mean astigmatic error was −0.27 DC and −0.40 DC for emmetropes and myopes, respectively. All participants had best corrected VA of 0.00 logMAR or better. Refractive errors were corrected using full aperture trial lenses.
All participants were free from any binocular vision abnormality and ocular disease, and were required to have undergone an eye examination in the last two years and possess up to date spectacles (if required). All prescriptions were rechecked and participants with more than 0.50DS change in refraction were excluded from the experiment. Ametropic participants had worn their correction for at least two hours before starting the experiment to ensure that no prior blur adaptation had taken place. Contact lenses were not worn on the day of the experiment to avoid problems such as dry eye and reduced vision. Informed consent was obtained from all participants and the study was conducted in accordance with the Declaration of Helsinki. Institutional ethical approval was also obtained.
Visual acuity was measured in the right eye only. LogMAR letters on a Bailey-Lovie design chart were viewed on the Thomson Test Chart 2000 system (www.thomson-software-solutions.com) on a standard LCD computer screen at a distance of 3 m (NEC LCD Accusync 72VM – model no. ASLCD72VM-BK-1, mean luminance; 239.8 cd m−2). Test Chart 2000 settings were adjusted for use at 3 m. Letters were randomised for each VA measurement to reduce potential learning effects and allow an accurate reading to be taken.
The experiment was completed across two sessions of approximately one hour each. Two different blur levels were used separately for each session (+1.00 D and +3.00 D). The sessions were separated by one week to allow blur adaptation effects to dissipate from the first session. Randomisation of blur levels for each session ensured no investigator bias.
Subjective refraction was conducted on all participants using standard procedures (Jackson crossed cylinder, and modified Humphriss binocular balancing), to the Test Chart 2000 screen at 3 m working distance. All participants wore this optimum refractive correction in a trial frame, with additional spherical convex lenses in front of both eyes to induce +1.00 or +3.00 D of blur. Visual acuity measurements were made on the right eye only (left eye was occluded temporarily during measurements). Participants were instructed to look at the smallest line of letters they could see through the blurring lenses for the duration of the experiment. Visual acuity measurements through blurring lenses were taken for every 2 min for a period of 30 min and ‘0 min’ corresponds to initial defocused VA measurement prior to any adaptation taking place. The blurring lenses were then removed to measure clear VA every two minutes for a further 20 min.
To ensure our results were reliable, participants were encouraged to read as far down the chart until at least half of the letters were named incorrectly on a line. Letter-by-letter scoring was implemented as this is known to improve the repeatability of logMAR acuity scores.
In a subset of 12 participants (six emmetropes and six myopes), a control trial was conducted where VA measurements were made every 2 min for a period of 30 min with the optimum refraction in place. The rationale here was to control for the influence of time and repeated measurements on VA.
Analysis of the data was conducted to investigate the first point at which blur adaptation causes a significant and measurable improvement in defocused VA, how defocused VA changed over the course of adaptation and lastly, if there were any differences due to refractive status. A two-way repeated measures anova was applied to the data using spss Ver.19 software (www-01.ibm.com/software/analytics/spss).
The introduction of defocus immediately reduced baseline clear VA. Myopes, as expected, had consistently better initial defocused VA (i.e. immediately after the introduction of defocus) prior to any adaptation being undertaken. However, statistically, this difference was insignificant for both defocus levels (P > 0.05). Blur adaptation resulted in a significant improvement in defocused VA over time for both 1 D and 3 D defocus levels (F(5.30,116.65) = 47.58, P < 0.0001). Overall, 30 min of defocused viewing produced significant adaptation effects (P < 0.0001) for both defocus levels. This is shown as averaged data for emmetropes and myopes in Table 1, and as individual data in Table 2.
Table 1. Mean (±1 S.D.) change in defocused visual acuity (VA) from start to finish of the adaptation period (30 min) for both defocus levels. Negative logMAR VA scores indicate an improvement in VA
Mean improvement in VA (logMAR)
−0.17 ± 0.11
−0.20 ± 0.09
−0.19 ± 0.10
−0.16 ± 0.10
−0.23 ± 0.14
−0.19 ± 0.12
Table 2. Individual values for change in defocused visual acuity (VA in logMAR) following 30 min adaptation to 1 D or 3 D blur. Negative logMAR VA scores indicate an improvement in VA
Change in logMAR VA; baseline to 30 min
1 D blur adaptation
3 D blur adaptation
For 1 D defocus, the first point at which blur adaptation caused a significant and measurable effect (i.e. improvement) in defocused VA was at 4 min into the adaptation period (P < 0.0001). Although subsequent measurements over the course of the 30 min adaptation period indicated an on-going improvement in defocused VA, none of these adjacent measures were significantly different from the previous one. This suggests that a plateau has been reached. For example, from 4 to 6 min adaptation, the improvement in defocused VA was not significant (P = 1.00), and so on. For 3 D defocus, significant improvement in defocused VA was also found at 4 min of adaptation (P < 0.0001). Similarly, a plateau was also reached at 6 min adaptation to 3 D defocus (P = 1.00).
Refractive status did not have a statistically significant effect on the time course of blur adaptation (F(1,22) = 1.30, P = 0.27) for either defocus levels. These data are shown in Figures 1 and 2 for 1 D and 3 D respectively. Data were fitted with an exponential decay function; the lowest R2 value for this fit was 0.95.
After a 30 min adaptation period, defocus lenses were removed and recovery clear VA was measured. There were no significant changes in recovery clear VA over time (F(5.63,123.93) = 0.81, P = 0.56), i.e. post-adaptation clear VA are similar to initial clear VA, and remained similar over the 20 min recovery period. Refractive status did not have any influence on post-adaptation clear VA. Data for the recovery periods following 1 D and 3 D blur are presented in Figure 3.
Figure 4 presents the results for the control experiment, where VA was measured every 2 min for 30 min under clear conditions. There was no significant change in VA during the 30 minute clear control period, indicating an absence of a learning effect regarding the visual task.
Effect of time on blur adaptation
This study provides new information about the time course of the blur adaptation process. Blur adaptation causes significant improvements in defocused VA within 4 min of exposure to a defocused environment, followed by a plateau of the effect at 6 min. This is represented well by an exponential decay function. The improvements in defocused VA for 1 D and 3 D blur levels were clinically significant in our study, and are in line with previous findings.[1, 6, 13] Cufflin et al measured logMAR VA under defocused conditions every 10 minutes using 1 D and 3 D defocus. Change in defocused VA was significant compared to baseline, and equal for both defocus levels. In contrast to our study, improvements in visual acuity became significant after 30 min of adaptation. Cufflin et al reported no significant improvements in VA in the optimally focused control group which indicates that blur adaptation is not a learning effect. Thus, we can be certain that blur adaptation has a real and measurable effect on VA.
From our findings, it is apparent that the neural compensatory processes[1, 3] that serve to improve high contrast VA as part of a continuous recalibration mechanism occur rapidly. Our results are in agreement with previous reports of improvements in VA using simulated and optically-induced astigmatic defocus after only 10 min of adaptation. Another study reports increased supra-threshold contrast sensitivity after adaptation to +3 D defocus for 10 min. Although these findings strongly suggest that blur adaptation effects are present shortly after the introduction of defocus, direct comparison with previous work is difficult, due to a lack of high-contrast VA measurements at sufficiently small temporal intervals. There have been reports suggesting that an element of individual variability is present during adaptation and that personality influences tolerance of defocus and perception of image quality, i.e. individuals that exhibit greater magnitudes of adaptation may be more tolerant to defocus.
It would be interesting to examine whether blur adaptation is dose-dependent, i.e. is blur adaptation still effective when periods of defocused vision are interleaved with clear viewing. It may be the case that short periods of clear vision prevent the blur adaptation process from taking place. Alternatively, blur exposure may be cumulative, in which case as the dose of blur builds up, so does the adaptive effect. That may have clinical ramifications, for example the small periods of blur that observers may be exposed to as a result of nearwork-induced transient myopia. This may also relate directly to myopes who remove their correction for part of the day. A protocol of periods of incremental blur exposure could be used to investigate the magnitude and time course of blur adaptation effects.
Effect of ametropia on time course of adaptation
Our study did not find a significant difference in the time course of blur adaptation between emmetropes and myopes. Due to variability in VA measurements with defocus, it is possible that any refractive group differences could have been masked. However, it is worth noting that myopes had better initial defocused VA prior to adaptation. Long term exposure to defocus may be associated with increased subjective tolerance to defocus, and therefore uncorrected myopes are more proficient at interpreting blurred signals. It can be postulated that a myopic individual without vision correction may be in a constant state of blur adaptation, and may therefore have better unaided vision than initially expected compared to the degree of myopia present.
LogMAR chart repeatability
High-contrast logMAR VA charts have been used extensively to measure defocused VA prior to and following a period of adaptation.[1, 2, 14] LogMAR charts are beneficial as they control the visual task and crowding effects for all acuity levels. The presence of defocus results in an increased variability of VA measurements, i.e. increased test-retest variability of VA measurements under myopic defocus and reduced accuracy of the endpoint of high-contrast VA measurements.
The findings of this study show how rapidly the perception of defocus can change in response to alteration of the visual environment. Blur adaptation has a significant and measurable effect on VA within 4 min of exposure to defocus. This finding indicates that the visual system instigates the neural compensatory mechanisms shortly after the appearance of defocus. Our results relate particularly to real-life vision of uncorrected myopes or myopes who remove their correction for part of the day.
Kiren Khan is supported by a Bradford School of Optometry and Vision Science departmental studentship. Aleksandra Mankowska is supported by a Doctoral Training Account grant from the Engineering and Physical Sciences Research Council.