Recent clinical research has shown that the degree of myopia is significantly less following laser therapy when compared with cryotherapy for severe retinopathy of prematurity (ROP).1² This important finding is confirmed by Laws et al in this issue of the BJO (p12). While the associations of prematurity, ROP, and myopia are well known, they are not precisely defined and mechanisms are even less well understood—an ideal opportunity for us to delve briefly into the evidence.

Myopia is probably the normal refractive state in infants before full term^3 4 with the eye becoming more hypermetropic in early infancy. Compared with the eye of the full term baby the features of this myopia are shorter axial length, flatter anterior chamber, and more spherical lens.4 The term myopia of prematurity is not applied to this physiological and temporary type of myopia.

Over three decades ago Fledelius studied a cohort of preterm babies and observed a disproportionate number of myopes; he found that this refractive state persisted even to 18 years of age.5 This type of myopia, myopia of prematurity (MOP), has an early onset and compared with full term and juvenile onset myopes the MOP eye exhibits a relatively highly curved cornea, shallow anterior chamber, and thick lens. Axial lengths are shorter than expected for the dioptric value.6 The hallmark of MOP is arrested development of ocular anterior segment. With refreshing honesty Fledelius, in 1996,7 stated that owing to a paucity of neonatal data in his early study it was not categorically known whether the MOP in this cohort was, or was not, associated with previous ROP. It is now confirmed that MOP without previous ROP does occur,7-11and at higher frequencies than in the full term population, and with the characteristics described above.12

Few reports on the refractive outcome contain detailed information of the neonatal period. Laws et al 13 studied ROP stage and refractive outcome at 6 months’ corrected age and found that while there was a trend for increasing myopia with ROP presence and severity, this only reached significance with stage 3. Several authors have reported that there is a dramatic jump in the prevalence of myopia when stage 3 ROP is reached.791114 To cite one study⁹ the incidence of myopia by ROP stage was as follows: none 13%; mild ROP <20%, stage 3 ROP >44%. Of particular interest is the high incidence of myopia in both cryotherapy treated and non-treated eyes with severe ROP.81516 This is quite distinct from the outcome after laser therapy as shown by Lawset al and others.12 Acknowledging the large data spread, these differences are not trivial, with refractions at 1 year after laser of −0.50 and −0.37 dioptres (right and left eyes), compared with −5.25 and −6.00 (right and left eyes) following cryotherapy (Laws et al, this issue).

Many theories have been put forward to explain how myopia develops in premature babies. These include bone deficiency, temperature, light, visual deprivation, and retinal dysfunction. Pohlandt17 speculated that MOP could be attributed to postnatal bone mineral deficiency, an idea which unsurprisingly generated a sharp critique.18 Secondly, preterm neonates experience a temperature deficit¹⁹ at a time when corneal growth is especially active. While we can speculate that this deficit might impede corneal growth, it is most unlikely that it fully explains MOP. Furthermore, while this temperature deficit affects all such babies myopia is not an invariable sequela of preterm birth. This last point also mitigates against light exposure as a myopiagenic factor. While light is known to influence eye growth in chickens,20and preterm neonates are generally exposed to high levels of non-cycled lighting,21 its role for human eye growth is unknown. Visual deprivation is, in our opinion, an unlikely candidate as a cause of either MOP or ROP induced myopia, not least because normal vision in preterm neonates is so low that it is relatively insensitive to blur and deprivation. Thus, macular haemorrhage (even unilateral) in full term neonates does not adversely influence visual²² or refractive development.23 Macular haemorrhage permits peripheral retinal function, but it is interesting to note that more generalised deprivation such as a dense vitreous haemorrhage does cause myopia, but this has been reported only in older babies and children, and only if it persists for months.24 It has been postulated that even mild acute ROP renders the posterior retina dysfunctional possibly by retarding photoreceptor maturation and migration²⁵ and ‘so alters eye growth signals’. While we cannot agree with or refute this general statement, it does not marry with the data on ROP stage and refractive development.

ROP induced myopia cannot be fully explained by increased axial length as it is also associated with evidence of arrested development of the anterior segment: microcornea, steep corneal curvature, thickened lens.26-29 This points to a mechanical restriction of ocular growth. Supportive evidence comes from the non-linear refractive development associated with ROP. As the prevalence of myopia rises sharply as stage 3 is achieved, so does the prevalence of anisometropia, and astigmatism.13 The last exhibits a greater than normal spread of axis—with a tendency for the axis to rotate according to the location of ROP residua.13It could be argued that the mechanical effect is exerted by the ROP lesion which is located in that portion of the globe where maximum growth occurs in late fetal and early postnatal life. Restricted growth in this region would be expected to inhibit growth of both the anterior sclera and the anterior segment.

What can account for the differential refractive outcome of cryotherapy and laser? Trans-scleral cryotherapy is more tissue destructive³⁰ compared with laser.31 Cryo applications are large and confluent. Laser lesions are smaller, discrete, and, as they are spaced by lesion-sized gaps,32it could be argued that this is less likely to impede ocular growth.

To summarise, three types of myopia are associated with premature birth: (1) physiological and temporary myopia (nature); (2) myopia without ROP (MOP; nurture); and (3) myopia induced by severe ROP (disease). That laser results in less myopia than cryotherapy is clinically important. However, it is also important to appreciate that both cryotherapy and laser offer significant benefit to the eye at risk of blindness due to severe ROP. Clearly, there is vital work to do not only with regard to refractive development, but also to determine the visual outcome of these babies who are nurtured in an abnormal environment and may suffer a range of severe visual pathway complications. So, returning to the title of this editorial, it is hopefully now apparent that nature, nurture, and disease all contribute to myopia associated with prematurity—we need to know more.