Teaching case. A preschool child with constant unilateral esotropia. Refraction after cycloplegia is almost symmetrical, the media are clear, the disc and macula are unremarkable. Best-corrected acuity is worse in the eye that does not hold fixation. If you think "the eye is lazy", you want to stimulate that eye with devices. If you think "the cortex has learned to ignore one channel", it becomes clear why you touch the fellow eye at all.
Amblyopia, as formulated in the AAO Amblyopia PPP 2022 (Updated 2024), is a unilateral or, less commonly, bilateral reduction in BCVA that usually occurs in an otherwise normal eye. It is a developmental disorder of the central nervous system: abnormal processing of visual images leads to reduced acuity. Less commonly, amblyopia exists alongside a structural abnormality of the eye or the pathway — and then the reduction cannot be attributed entirely to that abnormality. Such eyes can have deficits of contrast sensitivity and accommodation. The fellow eye is often not "perfectly normal" either: subtle functional deficits are described in it (the PPP cites works designated in the document as 4 and 5).
This definition does two practical things.
First: the diagnosis is not made until there is an amblyogenic factor — strabismus, unequal or high refractive error, media opacity, ptosis, a structural obstruction — and a corresponding acuity deficit. The PPP states directly: amblyopia in the absence of strabismus, unequal refractive error, media opacity or a structural abnormality is rare. If there is no factor, you look for another diagnosis rather than "idiopathic amblyopia" as a convenient shelf.
Second: the retina can be anatomically quiet, and this does not refute the diagnosis. On the contrary, a quiet fundus with a persistent BCVA deficit and an obvious factor is the typical picture. The reverse is also true and is missed more often: with disc hypoplasia, myelinated nerve fibres, macular traction after retinopathy of prematurity, or uveitis, part of the reduction may be treatable amblyopia — either through the mechanism of form deprivation or because of coexisting strabismus and refractive error (PPP). A structural label does not exempt you from attempting to treat the share that is still plastic.
Why it is the cortex. The classic biological argument comes not from the clinic but from the laboratory. Wiesel and Hubel showed in the cat in 1963 and 1970 that early monocular deprivation (closure of one eyelid) sharply changes the ocular dominance of cells in the primary visual cortex: after early deprivation almost no cells remain that can be driven through the deprived eye, whereas normally most cells respond to either of the two eyes. Binocular deprivation of equivalent duration gives a different picture: the cells respond less well, but are not "rewritten" in favour of one eye as brutally. This is not a human RCT and not grounds for prescribing occlusion to a cat. It is a mechanism that explains why a unilateral cataract in an infant is on average more severe than a bilateral one of comparable density: on top of the direct degradation of the image comes interocular competition for cortical representation. The PPP states the same for humans: unilateral obstruction of the visual axis produces a greater loss than bilateral deprivation of a similar degree.
Hence the working vocabulary, which is not spelled out again in the course.
- BCVA — best-corrected visual acuity: the acuity measured with spectacle (or contact lens) correction, not "as is".
- Fellow eye — the better-seeing eye in unilateral amblyopia. Not automatically "healthy": it may have subtle deficits, and it can be harmed by excessive occlusion.
- Amblyogenic factor — a condition that makes the input from one or both eyes unsuitable for normal development: defocus, non-fusable input in strabismus, form deprivation.
- Interocular competition — the cortical mechanism by which the sharper / fixating channel suppresses the less sharp one. It explains why unilateral causes are usually more severe than symmetrical ones.
Prevalence, so as not to wave around a round "one in ten". Population-based studies in children aged 30–71 months, summarised in the PPP, give a range of 0.7–2.6%; older school samples are usually higher, 1.0–5.5%, and the figure depends on age, race/ethnicity, method and the definition of amblyopia itself. Most cases are unilateral; the proportion of bilateral cases varies across studies from about 5% to 41% — again because of the acuity threshold, not because of "a different disease". These are observational prevalence estimates, not the risk for a particular child in clinic.
Factors that the PPP associates with a higher frequency of amblyopia and its causes: prematurity, small for gestational age, developmental delay, amblyopia in a first-degree relative. Maternal smoking and alcohol/drug use during pregnancy are associated in some studies with an increased risk of amblyopia or strabismus; other population-based studies do not find this association. An observational association here remains an association.
The lesson in one sentence. Amblyopia is the cortex not having learned to use one of the channels, not the eye "not wanting to look". Treatment therefore always answers the question: what input is the cortex receiving now — and what input can we make usable.