The exclusion of non-cycloplegic studies was necessary

The gap

Evidence profile

Sourced from the limitations section and limitations of the source papers, classified as general, spanning 3 journals.

Supporting evidence — 3 representative gaps

• Neuro-ophthalmic review on pediatric myopia: Advancing from refraction to a brain-centric model of axial growth (2026) · World Journal of Clinical Pediatrics · doi

  The exclusion of non-cycloplegic studies was necessary, as non-cycloplegic refraction in children tends to overestimate the degree of myopia due to active accommodation. The mean age of participants varied across the included studies, and additional analyses were performed to account for differences in age distribution. The review was limited to studies published between 2002 and 2022.

  generallimitations section
  Keywords: exclusion non-cycloplegic studies was necessary refraction children tends
• Refractive outcomes following anti-VEGF, vitrectomy, cryotherapy, and laser photocoagulation for retinopathy of prematurity: a systematic review and meta-analysis (2026) · Frontiers in Medicine · doi

  The strengths of this study include its comprehensive scope, incorporation of both randomized controlled trials and real- world studies, and the use of subgroup and heterogeneity analyses. However, limitations remain, including heterogeneity in refractive assessment methods. Study design heterogeneity is another important limitation because most included studies were retrospective cohorts, with only a small proportion of RCTs. Dierences in study design may aect patient selection, treatment criteria, follow-up schedules, and outcome ascertainment, which can introduce systematic bias. Confounding by indication cannot be excluded because treatment selection is influenced by infant maturity and disease severity, including gestational age, birth weight, and ROP zone and stage. Many included studies reported unadjusted outcomes, and uniform adjustment across studies was not feasible. Therefore, observed dierences in refractive outcomes may partly reflect baseline dierences rather than treatment eects. Furthermore, the absence of detailed biometric data, such as axial length, in many studies may have limited the ability to fully explain refractive trends. Substantial heterogeneity was observed across several pooled analyses. This heterogeneity likely reflects dierences in follow- up duration, age at refractive assessment, refractive measurement techniques, and variations in treatment protocols across centers and time periods. Baseline disease severity and infant maturity also varied across studies and were not uniformly reported, which may further contribute to between-study variability. Although prespecified subgroup analyses were performed where data were available, more detailed exploration of heterogeneity, such as sensitivity analyses by specific anti-VEGF agent, follow-up strata, or baseline maturity, was limited by incomplete reporting and the lack of consistently adjusted estimates across studies. Therefore, pooled estimates should be interpreted with caution, and the direction of eects may be more informative than the exact magnitude in some comparisons. This systematic review and meta-analysis suggests that anti- VEGF therapy results in significantly better refractive outcomes in children treated for ROP, with lower degrees of myopia and reduced prevalence of high myopia compared to laser photocoagulation, cryotherapy, and vitrectomy. Among the available treatment options, cryotherapy yielded the highest myopic burden, while vitrectomy–though essential for retinal detachment–also conferred considerable refractive disadvantage. Laser therapy, while eective for peripheral ablation, was associated with substantial myopia, particularly in posterior disease. The relatively mild refractive profile observed after anti-VEGF therapy supports its use as a preferred modality, especially for Zone I and aggressive posterior ROP; however, long-term follow-up remains critical for early detection of refractive changes and reactivation of neovascularization. Moving forward, research should focus on standardizing refractive measurement protocols, integrating biometric outcomes such as axial length and anterior chamber depth, monitoring long-term visual and functional outcomes, and comparing newer anti-VEGF agents and dosing strategies.

  generallimitationsevidence 5/5
  Keywords: refractive heterogeneity treatment outcomes across analyses erences follow anti vegf maturity disease observed baseline therapy
• Global Patterns of Myopia, Age, Sex, and Vision Loss: A Comparative Analysis of US and South Korean National Surveys (2026) · Ophthalmology Science · doi

  This study has several limitations. Both NHANES and KNHANES used noncycloplegic refraction. Although the lack of cycloplegia may introduce some misclassification of refractive error in younger participants, this is unlikely to materially affect the main findings. In the Beaver Dam Offspring Study, the mean difference between cycloplegic and noncycloplegic refraction in 40 to 49 year-old myopes was 0.3 D and >60 years of age was just 0.03 D. 42 Recent work in various populations has also demonstrated that the difference between noncycloplegic and cycloplegic refraction is lowest among myopes, and particularly older myopes, with those aged ≥60 years having a difference of 0.07 D. 43,44 The cross-sectional design precludes direct causal inference and limits assessment of longitudinal changes in visual impairment risk. As with the Kaplan—Meier curves presented in Tideman et al, 16 the Kaplan—Meier curves in Figure 2 should be interpreted as estimates of the cumulative probability of visual impairment at a given age, that is prevalence, rather than incidence because the underlying data are cross-sectional. In addition, the direct cause of visual impairment was not assessed; hence, the results represent the associations be- tween myopia, sex, age, and visual impairment of all causes. Axial length was not measured for either the NHANES or KNHANES and as such the relationship between axial length and vision impairment could not be determined. It has, however, been shown that most myopia is axial in na- ture and in this case, SER is likely a reasonable surrogate for axial length. 16 Other unmeasured environmental, genetic, or phenotypic factors that may modify susceptibility to visual impairment in the presence of myopia may have contributed to the observed regional differences. Choroidal thickness, which was not measured in the NHANES and KNHANES surveys, has been identified as an independent risk factor for myopia-related pathology, 45 but little data exists regarding regional differences in this parameter. considerations may Alternatively, methodological contribute, including differences in the distribution of high myopia between surveys. The greater representation of individuals with very high myopia in KNHANES may accentuate effects at the upper end of the refractive error spectrum. Despite this variation, the existence of a consistent, dose—response relationship between myopia and visual impairment in both surveys supports the conclu- sion that myopia-related risk of visual impairment is a robust and generalizable concept across populations. Differences in survey methodology, examination protocols, and population characteristics between the 2 countries may have influenced comparisons. Visual acuity assessment differed between NHANES and KNHANES, which may in- fluence absolute estimates of visual impairment and should be considered when interpreting intercountry comparisons. The NHANES used a letter-based chart integrated into an autor- efractor, whereas KNHANES employed a logMAR chart with numeric optotypes. Differences in optotype design, chart characteristics, and testing conditions are known to produce systematic variation in measured acuity. Such methodological differences could contribute to observed intercountry Moore et al • Myopia and Vision Loss by Age and Sex differences in vision impairment, particularly at higher levels of myopia. Accordingly, intercountry comparisons should be interpreted cautiously. However, these differences are unlikely to explain the consistent within-country patterns observed across refractive error severity, age, or sex. Finally, unmea- sured confounding factors such as socioeconomic status, ac- cess to eye care, or comorbid health conditions may also have contributed to observed differences.

  generallimitationsevidence 5/5
  Keywords: impairment myopia visual differences knhanes nhanes axial observed noncycloplegic refraction refractive error difference myopes risk

Questions about this gap