Your child's diopter prescription tells you how blurry their distance vision is. Their axial length in millimetres tells you how much their eye has grown — and that's the number most closely linked to long-term eye health risk.
A myopia prescription — expressed in negative diopters (e.g., −2.50D) — measures how much optical power a lens needs to bring a patient's distant vision to clear focus. It is a functional measurement of visual blur, not a direct measurement of eye structure. Several factors influence the measured refraction including whether cycloplegia was used, the time of day, recent near work, and individual variation in corneal curvature and lens power.
Because refraction can fluctuate between visits even when the eye is growing, prescription alone is an unreliable marker of whether myopia is "stable." A child's prescription could remain unchanged for 12 months while their axial length continues to grow — especially if concurrent changes in corneal curvature or lens power partially compensate for the axial elongation.
This is why clinicians sometimes tell parents "the prescription hasn't changed" while the child's eye is still growing longer. The prescription is a snapshot of functional vision; axial length is a direct, irreversible measure of structural change.
Axial length is the physical distance from the front of the cornea to the retinal pigment epithelium, measured in millimetres using optical biometry (devices such as the Lenstar or IOLMaster) or ultrasound A-scan. It does not require cycloplegia. It is highly repeatable (within ~0.01–0.02mm across measurements) and is the standard outcome measure in all modern myopia control RCTs.
Why does axial length matter more than diopters for risk? The structural complications of myopia — myopic maculopathy, retinal detachment, glaucoma, and cataract — arise from the mechanical stretching of ocular tissues as the eye elongates, not from the diopter number on a prescription. A large-scale multicentre analysis found that axial length is more closely correlated with myopic complications than refractive error in diopters. In high myopia, the odds ratios for serious conditions are striking: myopic macular degeneration (OR 845 for ≥−6.00D vs low myopia), retinal detachment (OR 12.62), open-angle glaucoma (OR 2.92).
The IMI 2025 Digest cites Singapore longitudinal data finding that a significantly greater risk of myopic retinopathy was associated with axial length ≥26.5mm, myopia over 8.00D, and age over 19 years. This threshold-based risk framing (by AL, not diopters) is central to why the IMI recommends tracking axial length rather than refraction as the primary outcome of myopia management.
One child might have a −4.00D prescription with AL of 24.5mm; another might have a −3.00D prescription with AL of 26.2mm. The second child has a lower prescription but a longer eye — and is therefore at greater long-term risk of structural complications. Without axial length measurement, the first child would appear more concerning and the second would be undermanaged. This is the "mismatch" scenario described in clinical literature and is one of the strongest arguments for routine AL measurement.
As a rough approximation, each 1.00D of myopia corresponds to approximately 0.30–0.35mm of axial elongation in school-aged children — but this relationship is not fixed. It varies with age, individual corneal curvature, lens power changes during childhood (the lens thins and loses power as the eye grows, partially compensating for axial growth), and level of myopia.
In the COMET study, 1D of myopia was associated with 0.50mm of axial length in progressing myopes — a larger ratio than in non-progressors. This means that the diopter-to-millimetre ratio itself is a clinical signal: rapidly progressing myopes tend to show larger axial changes per diopter of refractive change.
For monitoring purposes, the IMI 2025 Digest confirms that axial length measurement is more accurate than refraction at detecting lower levels of myopia progression — meaning it can catch early progression that a refraction-only follow-up would miss.