Statistics from Altmetric.com
Increased chances of finding an effective optical method of arresting myopia development
In The Marriage of Heaven and Hell, William Blake says that “If the doors of perception were cleansed everything would appear to man as it is: infinite.” In vision, of course, there is a simple connection between optical infinity and perceived visual clarity, at least for distance vision in emmetropes. Contact lens practitioners and refractive surgeons have taken things one step further. By exploiting the brain’s ability to perceptually suppress central vision in one eye when the two eyes are receiving disparate stimuli, they have found that it is often possible to correct presbyopic ametropes using a distance correction for the dominant eye, and a near correction for the non-dominant eye. In this “monovision” situation, patients thus have to suppress the central vision in their non-dominant eye for distance tasks, and in their dominant eye for near tasks.
In essence, monovision is a form of deliberately introduced anisometropia, and it is this property that John Phillips has exploited in a highly original study, in this issue of BJO (p 1196), that provides new insight into the role of blur in regulating eye growth and refractive development in children. In this small clinical trial, children received a full myopic correction for their dominant eye and an undercorrection of up to +2.00 D for their non-dominant eye (as discussed below, the undercorrection led to the vision in the children’s non-dominant eyes being continually blurred). The results were striking: the rate of myopia progression in the undercorrected eye was found to be approximately 50% of that in the fully corrected eye. Furthermore, the reduced rate of myopic progression was attributed to a reduced rate of vitreous chamber elongation, consistent with a slowing of the primary structural change responsible for causing myopia.
Using the chicken as a model, Schaeffel and co-workers1 first showed that refractive state could be altered in response to retinal blur imposed by wearing a spectacle lens over one or both eyes. Remarkably, the change in refractive error Schaeffel et al found depended on the sign of the defocus: negative spectacle lenses, which would have diverged the light entering the eye and shifted the image plane behind the retina (leading to “hyperopic defocus”) induced myopia. Meanwhile, positive spectacle lenses, which would have converged the light entering the eye, shifting the image plane in front of the retina (producing “myopic defocus”) induced hyperopia. In each case, the refractive change was mostly the result of an altered rate of vitreous chamber elongation. Chicks becoming myopic showed an acceleration in their normal rate of axial eye growth. Chicks developing hyperopia showed a slowing or cessation of the normal rate. Similarly elegant experiments have demonstrated that refractive plasticity of this kind, which represents an active form of emmetropisation, is a feature of the early development of many other species, including marmosets2 and monkeys.3
So it seems highly likely that human infants also use blur cues to guide refractive development during emmetropisation. What is less clear is whether blur contributes to the development of myopia in school aged children, as well. A decade ago, Gwiazda and colleagues4 hypothesised that the lag of accommodation at near could explain the intriguing link between near work and myopia. Children with an accommodative lag at near would experience hyperopic blur at the retina that, according to the animal models, would lead to myopia development. However, conclusive proof for this theory is lacking, and more recent results from a longitudinal study suggest that there is no difference in accommodative lag before the onset of myopia development, between those children who remain emmetropic and those who become myopic.5
Furthermore, experiments in which animals have been allowed short periods (minutes) of unrestricted vision, in between long periods of hyperopic defocus, show that these intervals of sharp focus quickly counteract the tendency towards myopia. This might mean that children undergoing even prolonged periods of near work induced blur are protected from myopia by distance viewing in between times. Also, Chung and colleagues6 undercorrected both eyes of a group of 47 young myopes, by approximately +0.75 D. After 2 years, instead of the myopic blur these children having been exposed to making them less myopic, Chung et al found that the children had become even more myopic than a control group who had been fully corrected. So perhaps humans are unusual, in not being able to emmetropise in response to defocus? Or perhaps by the time they reach school age, the plasticity of the human emmetropisation system has decreased below a clinically significant level?
Even before the study by Phillips, however, there was indirect evidence that exposure to blur could influence refractive development in school age children. In the COMET study, a randomised, multicentre, double masked clinical trial to evaluate whether progressive addition lenses (PALs) slowed myopia development in children, PALs were indeed found to slow myopia progression significantly, though only by about 14% over 3 years. While it is possible that reduced accommodation itself was responsible for this slowed rate of progression, it seems more likely to have been the reduction in hyperopic defocus brought about by the extra plus power at near lessening the stimulus for myopia development, since in the subgroup of children who had larger lags of accommodation, PALs slowed myopia progression to a greater extent (by 21% in general, and by 37% in children who also had esophoria at near).7
How does the monovision study carried out by Phillips add to our knowledge? Crucially, unlike presbyopic monovision wearers, the children in the Phillips study were found to posture their accommodation according to their dominant eye, for both distance and near. This meant that the non-dominant eye would have experienced myopic blur in both viewing conditions, and hence throughout the period of spectacle wear. However, the amount of accommodation these children exerted would presumably have been similar to that when wearing a full distance correction. Therefore, the new study strongly suggests that it was the continuous myopic blur experienced by the undercorrected eye that signalled its slower rate of myopic progression.
Critics might argue that the study by Phillips is too small to get overly excited about; after all only 13 children were involved in the trial. Yet the “within subject” design afforded by the monovision approach (compared to the “between subject” design necessitated by myopia control trials such as the COMET study) meant that despite its small size, the study was adequately powered, as the results confirm. However, the lack of a control group does leave open room for a slight doubt: because we do not know what the baseline rate of myopia progression is in this particular group of children, it is conceivable that the fully corrected eyes had shown an accelerated rate of myopia progression, rather than the undercorrected eyes having shown a slower rate of progression.
The new study immediately brings to mind two key questions. Firstly, is it possible to exploit this monovision paradigm clinically to slow myopia progression? And, secondly, why did the undercorrected eyes still progress towards myopia, instead of halting their progression completely?
In answer to the first question, it is important to emphasise that the protocol used by Phillips induced anisometropia, because only the non-dominant eye was exposed to myopic blur. Therefore, clinically, some kind of periodic reversal of the treatment regimen would be required, so that the dominant eye also spent time exposed to myopic blur while the non-dominant eye received clear vision. A crucial issue is whether children will tolerate the undercorrection of their dominant eyes in the reversed monovision situation. If they do, then the prospects look promising, but much work will still be required to optimise the treatment. For instance, would greater amounts of blur give a stronger effect, or would this tip the balance and induce form deprivation myopia? What would be the optimal treatment duration—for example, would it be better to switch the undercorrection on alternate days, alternate weeks, or alternate months, perhaps? Also, a more rigorous study design, including masked observers and a fully corrected control group will be required to prove conclusively that monovision is effective. It should noted, as well, that if a 50% reduction in myopia progression turns out to be the best that can be achieved, then taking into account the fact that the undercorrection would only be present for 50% of the time, then the overall progression rate might be slowed by only 25%, which would be disappointing.
Why should the undercorrected eyes have become more myopic, despite the myopic blur they were exposed to? At this age, the eye is normally elongating at approximately 0.2 mm per year,8 so there may be an endogenous, developmental push towards axial elongation. However, in animal models, visual cues are generally able to dominate this innate propensity, to effectively halt axial elongation completely. Table 1 of Phillips’s paper shows that there was some intersubject variation in response, and an impression that full time monovision wearers may have derived a slightly greater benefit than part-time wearers. If so, then at least part of the explanation may be that during periods when the monovision correction was not worn, the non-dominant eyes experienced hyperopic defocus (either because of children reverting to a conventional spectacle correction, or doing near work while uncorrected). Alternatively, there may be additional environmental risk factors for myopia progression9 to which these children were exposed, or it could be that once myopia progression has begun, that it is somehow self perpetuating (there is a precedent for this in marmosets10).
In conclusion, by providing evidence that further implicates blur in driving the progression of myopia in school age children and, most importantly, in providing arguably the strongest evidence yet that myopia progression can be slowed by imposing myopic blur at the retina, this study makes the chances of finding an effective optical method of arresting myopia development look significantly brighter.
Note in Proof
Increased chances of finding an effective optical method of arresting myopia development
Competing interests: none declared