Researchers reported delayed increases in refractive and corneal astigmatism among Hong Kong schoolchildren following COVID-19 restrictions, according to a population-based cross-sectional study.
Unlike previously reported myopia increases that emerged during pandemic restrictions, astigmatism rates remained statistically unchanged in 2020 and increased beginning in 2021, after restrictions were gradually lifted. Researchers suggested the delayed trajectory may reflect the need for cumulative corneal exposure over time, rather than an acute response to restrictions.
The study included 21,655 children aged 6 to 8 years in the Hong Kong Children Eye Study between 2015 and 2023. Children underwent comprehensive ocular examinations at 2 academic medical centers in Hong Kong, including cycloplegic autorefraction and optical biometry. No child was analyzed more than once.
Researchers defined refractive astigmatism as cylindrical refraction of at least 1.0 diopter and corneal astigmatism as an absolute difference of at least 1.0 diopter between the flattest and steepest keratometry readings. Refractive astigmatism includes both corneal and internal components, and researchers separately evaluated the corneal contribution using keratometry measurements. Analyses used right-eye data because of high intereye correlation.
Refractive astigmatism prevalence was 23% during 2015 to 2019, approximately 25% in 2020, 30% in 2021, and 35% in 2022 to 2023. The 2020 prevalence increase was not statistically significant compared with baseline.
Corneal astigmatism prevalence showed a similar delayed pattern. Prevalence was 60% during 2015 to 2019 and remained statistically unchanged at 59% in 2020, before increasing to 67% in 2021 and 65% in 2022 to 2023.
Mean refractive astigmatism magnitude remained stable at approximately 0.70 diopters through 2020 before increasing to 0.81 diopters in 2021 and 0.87 diopters in 2022 to 2023. Mean corneal astigmatism magnitude likewise remained unchanged at 1.24 diopters through 2020 before increasing to 1.34 diopters in 2021 and 1.35 diopters in 2022 to 2023.
In adjusted models comparing 2020 to 2023 with 2015 to 2019, the pandemic period was associated with 1.20 times the odds of refractive astigmatism and 1.26 times the odds of corneal astigmatism. Models adjusted for age, sex, parental astigmatism, spherical refractive value, and family income.
The pandemic period was also associated with adjusted increases of 0.04 diopters for refractive astigmatism and 0.05 diopters for corneal astigmatism. Unadjusted mean differences were larger at 0.10 diopters and 0.08 diopters, respectively, suggesting the smaller adjusted figures reflected the pandemic-associated contribution independent of concurrent myopia increases.
The study distinguished refractive astigmatism from its corneal component, which researchers said may help explain the delayed changes observed following the pandemic. Although overall mean keratometry remained stable across study years, the difference between the steepest and flattest corneal curvatures widened, suggesting localized corneal shape changes rather than generalized corneal steepening.
Most children continued to have with-the-rule astigmatism, meaning the steeper corneal curvature was primarily located in the vertical meridian. Researchers hypothesized that prolonged downward gaze during digital device use and near work may have contributed to these corneal changes over time.
Outdoor time declined from 1.43 hours per day before the pandemic to 1.16 hours per day during 2020 to 2023, while near work increased from 3.33 to 4.91 hours per day.
Researchers noted that the delayed increase in astigmatism compared with myopia may have implications for postpandemic pediatric eye surveillance, although the study did not evaluate screening strategies or long-term clinical outcomes.
The study had several limitations. Most participants were Han Chinese children living in a high-density urban environment, which may limit generalizability to other populations. Researchers also noted that P values were not adjusted for multiple comparisons. Recruitment declined substantially in 2022 to 2023 compared with earlier years, which invited commentators said may have affected how representative the later cohorts were of the broader population.
Parental astigmatism history was self-reported and subject to recall bias, although researchers reported that sensitivity analyses excluding parental astigmatism from regression models did not materially alter the findings.
In an invited commentary, Jonathan R. Morse, MPH, of the University of Michigan School of Medicine, and colleagues wrote that the study highlighted both the value and challenges of population-based pediatric ophthalmology research. The commentators noted that reduced recruitment during the pandemic, particularly in 2022 and 2023, raised concerns about whether enrolled children fully represented the underlying target population.
Co-study author Ian Morgan, PhD, reported receiving support from Eyerising International outside the submitted work. Co-study author Jason C. Yam, MD, reported a pending patent related to low-concentration atropine for delaying myopia onset and disclosed roles as a nonexecutive director and shareholder of OCUS Innovation in Hong Kong outside the submitted work. No other conflicts of interest were reported. The study received support from multiple Hong Kong research grants and institutional programs. The invited commentators reported no conflicts of interest.
Source: JAMA Ophthalmology
