Strengths and limitations
The ALSPAC study provides rich life course data for many participants across a geographic region, offering sufficient power to investigate rare exposures, though some adjusted analyses may have limited power. However, the lack of representation from ethnic minorities and less affluent families may restrict the generalisability of findings.8
Self-reported data such as smoking or alcohol intake are susceptible to misreporting, potentially biasing associations towards null values. Observer variability and measurement errors may affect physical measures, despite training and protocols.
Retinal photography, while widely used in clinical practice, faces challenges in optic disc measurement due to landmark identification and magnification error, mitigated by using dimensionless ratios like CDAR. Retinal photography and OCT have been demonstrated to yield comparable results with relative measures, though absolute measures may be around 10% smaller using photography.12
Monoscopic software usage may underestimate CDAR compared to stereoscopic methods, reducing cupping prevalence estimations.11 Although graders tended to underestimate parameters, their reasonable agreement questions the necessity of a reference standard with inherent errors. A priori analysis protocols aimed to minimize bias. Residual confounding from unmeasured exposures remains possible.
Interpretation
The average 12-year-old child typically exhibits a CDAR of around 0.2, regardless of gender, similar to findings in Australian six-year-olds.13 While a cup-to-disc ratio > 0.2 has been reported to occur in fewer than 1% of children our study shows a prevalence of CDAR > 0.3 in 11% of participants, suggesting a potential need to reassess the clinical implications of CDAR thresholds.14
Similarly, the mean cup-to-fovea over disc diameter (CF/DD) in our cohort was 2.48, with 6% of children exhibiting CF/DD > 3.0, questioning the appropriateness of this threshold for diagnosing optic nerve hypoplasia.2,3
Regarding interocular differences, only 0.5% of children in our study showed asymmetry > 0.2, highlighting how rare pathological asymmetry is according to proposed thresholds of > 0.25.15 Further exploration of asymmetry's clinical significance in childhood is warranted to establish appropriate abnormal interocular difference levels.
A small head circumference below the 9th percentile was associated with a small optic disc. It is unlikely that this is a proportional phenomenon given that CF/DD is a ratio, and large optic discs were not seen in those with large head circumference. One study reported smaller disc diameter in Australian children with small head circumference, which was associated with larger cup-to-disc ratio.16 Small head circumference has been reported to be associated with neurodevelopmental abnormalities such as attention deficit, which, in turn, has been associated with subtle morphological changes in the optic nerve.17,18
Being underweight at age 11 was associated with disc cupping independent of maternal BMI, aligning with adult studies associating low BMI with smaller neuro-retinal rim area and larger vertical and area cup-to-disc ratios.19,20 Nutritional deficiencies may mediate this association, given the optic nerve's vulnerability to deficiencies like B vitamins, folic acid, and proteins containing sulphur-containing amino acids.21 In later life, glaucoma development has been linked with low vitamin A and vegetable fat intake.22
Maternal smoking is a significant factor associated with small optic discs, likely due to its impact on foetal development and neuro-ophthalmic complications. Potential mechanisms include reduced placental blood flow through nicotine-induced vasoconstriction; foetal hypoxia from carbon monoxide binding to haemoglobin; vascular neogenesis; and endothelial function disturbance.23 Additionally, direct toxic, ischaemic, or hypoxic effects on cell proliferation or migration during critical periods may contribute.23
Studies of premature infants reported high mean CDAR and high prevalence of cupped discs, possibly influenced by factors precipitating preterm birth or immaturity-associated morbidity.24 Immature apoptotic pruning of supernumerary fibres and extra-uterine environmental influences like oxygen delivery and carbon dioxide tension may cause excessive elimination of axons.24 Persistence of cupping in older preterm children suggests limited catch up growth. Disentangling effects of gestational age from factors like low birth weight, ROP, cerebral injury, and early life events is challenging. Cupped appearance in white matter injury may stem from axonal interruption via retrograde trans-synaptic degeneration, akin to optic nerve hypoplasia, with the differing appearance due to the timing of the insult.24,25
Some exposures hypothesized to cause the extreme phenotype of optic nerve hypoplasia, like maternal age and parity, showed a lack of evidence of association.
Implications and future work
This study contributes to a field primarily composed of small-scale investigations. We reveal that cupped and small optic discs may be more prevalent in children than previously believed. Therefore, it is crucial to scrutinize the clinical characteristics of children presenting with what is typically considered a cupped disc to ascertain whether these features are being overly referred, pathologized, or monitored.
The long-term prognosis of preterm cupped discs remains uncertain. While a larger cup-to-disc ratio in adulthood is linked to a higher risk of glaucomatous optic neuropathy and associations have been reported between optic disc morphology and systemic neuro- and cardiovascular development in adults26, the significance of cupped discs in preterm children as a form of optic nerve hypoplasia remains unclear. Further research is needed to elucidate the relationship between visual acuity and disc size.
Additionally, the association between small discs and head circumference raises questions about the need for further assessment in infants with smaller heads but not microcephaly. Work is needed to understand the relationship between timing and coordination of craniofacial growth and the morphology of optic fidelity.
Finally, addressing maternal smoking, a preventable risk factor, is essential to combat health disparities among socioeconomic groups.