Long-term Impact of COVID-19–Related Behavioral Changes on Myopia Progression in Children: A Four-Year Cohort Study

Abstract

Purpose

To evaluate the effect of behavioral changes during the COVID-19 pandemic on myopia progression in children and to determine whether these changes modified the efficacy of low-concentration atropine.

Methods

This retrospective longitudinal study included children aged 7 to 15 years who underwent annual examinations from 2019 to 2023. A subset of participants completed a behavioral questionnaire in 2021. Annual spherical equivalent measurements from pre-COVID-19 (2019), during the COVID-19 (2020), and post-COVID-19 (2021–2023) periods were analyzed using linear mixed-effects models. Changes in mobile device use and outdoor activity were assessed with repeated measures ANOVA. Subgroup analyses were performed based on the degree of increase in device use and the duration of remote education.

Results

Seventy-seven children (140 eyes) were included in the refractive analysis, of whom 37 (69 eyes) completed the behavioral questionnaire and were included in behavioral analyses. Myopia progression during COVID-19 increased compared with the pre-COVID-19 period. A greater increase in mobile device use was significantly associated with faster progression between 2021 and 2022 (p = 0.005). Longer duration of remote education showed a similar association (p = 0.035). Changes in outdoor activity were not clearly associated with refractive outcomes. Low-dose atropine did not show a meaningful treatment effect in this cohort.

Conclusions

Myopia progression accelerated during the pandemic and was strongly associated with increased screen time and prolonged remote education. These behavioral factors had a greater influence on progression than low-dose atropine, underscoring the importance of addressing visual habits in pediatric myopia management.

Myopia has emerged as a public health concern due to its rapidly increasing prevalence in recent decades [1,2]. In particular, early-onset myopia in children is particularly worrisome, as it could be associated with a higher risk of developing high myopia and serious vision-threatening complications, such as retinal detachment, myopic maculopathy, and glaucoma [3–6]. Therefore, identifying and managing the factors associated with myopia progression is crucial for preventing potential vision loss.

Various lifestyle factors contribute to myopia progression. Increased outdoor activity is associated with a reduced risk of myopia onset in children; however, its effect on slowing progression in children already diagnosed with myopia is limited [7]. In contrast, excessive near work—especially digital screen use—is associated with a higher risk of myopia [8,9]. However, studying the impact of these behavioral factors on myopia progression poses challenges, primarily because it is difficult to alter children’s daily habits and quantify their effects over time.

The COVID-19 pandemic provided a unique context for investigating these factors, as it resulted in sudden and widespread changes in children’s lifestyles. Lockdowns and remote learning significantly increased screen time while reducing outdoor activities, raising concerns about accelerated myopia progression. Recent studies have documented a marked increase in myopia during the pandemic [10–12]. A prospective cross-sectional study observed a significant myopic shift during the pandemic following home confinement [10]. A comprehensive systematic review and meta-analysis of 33 studies further substantiated these global trends, demonstrating an overall increase in myopia prevalence, a decrease in spherical equivalent (SE), and an increase in axial length (AL) during the pandemic [12].

Various interventions have been developed to slow myopia progression, with low-concentration atropine eye drops (LAMP) attracting considerable attention due to their relative safety and efficacy [13,14]. However, the pandemic’s impact on the effectiveness of LAMP remains inadequately investigated [15]. Understanding this relationship is crucial for optimizing treatment strategies in environments characterized by extended near work and reduced outdoor exposure.

In this study, we examined longitudinal changes in refractive error among myopic children across the pre-, during-, and post-COVID-19 periods, incorporating 3 years of follow-up data (2019–2023). We also investigated the relationship between behavioral factors—such as screen time, outdoor activity, and remote education duration— and myopia progression. Additionally, we assessed the impact of LAMP treatment under these behavioral conditions.

Materials and Methods

Ethics statement

Written consent was obtained from individuals who consented to complete the questionnaire. The requirement for written consent did not apply to participants in the retrospective chart review who did not complete the questionnaire. To ensure confidentiality, all personal health information was strictly anonymized and de-identified before analysis. This study adhered to the ethical principles of the Declaration of Helsinki and received approval from the Institutional Review Board of National Health Insurance Service Ilsan Hospital (No. 2022-09-002).

Study design and population

This longitudinal study included myopic children who received regular ophthalmic follow-up at National Health Insurance Service Ilsan Hospital from November 2018 to January 2024. Medical records of children aged 7 to 15 years who had myopic refraction of at least −0.5 diopters (D) in both eyes and astigmatism of less than 2.5 D were reviewed. In 2021, 1 year following the onset of the COVID-19 pandemic, parents were requested to fill out a written questionnaire.

Inclusion and exclusion criteria

Inclusion criteria were as follows: (1) participants with a baseline SE of ≤−0.50 D; (2) those with astigmatism of <2.5 D; (3) those that showed up to three follow-up visits approximately 1 year apart; and (4) those that were initiated on low-concentration atropine eye drops pre-COVID-19 or did not use low-concentration atropine eye drops during the study. Exclusion criteria included participants with a history of orthokeratology or multifocal contact lens wear initiated after the onset of the COVID-19 pandemic, participants who began using atropine eye drops of any concentration post-COVID-19 pandemic, and those with ocular pathologies such as amblyopia, strabismus, or a history of ocular surgery.

Atropine treatment protocol

A standardized treatment protocol was established for subjects prescribed LAMP. Initially, 0.02% atropine eye drops were used. If myopia advanced by >0.5 D over 6 months, the concentration was escalated to 0.05% under the guidance of the supervising physician (HYK), considering the patient’s age and baseline refractive status. The eye drops were formulated by diluting 1% Isopto Atropine (Alcon) with 0.9% saline and were administered nightly. For participants who discontinued LAMP during the follow-up period, only the duration of actual atropine use was included in the analysis so that the analyzed periods reflect active treatment exposure. In cases where the atropine concentration was escalated, participants were analyzed according to their initial starting concentration to maintain consistency in treatment group classification.

Definition of the COVID-19 study periods

The government implemented remote education and restricted outdoor activities during the COVID-19 epidemic in 2020. Consequently, the study periods were defined as follows: January through December 2019 constituted the pre-COVID-19 period; January to December 2020 was designated as the COVID-19 period; and the post-COVID-19 period was further segmented by year for in-depth analysis, spanning from January 2021 to December 2023. Each patient underwent at least one annual eye checkup throughout this time.

Myopia assessment and data collection

Refractive errors were measured using a Huvitz HRK-7000A autorefractor-keratometer (Coburn Technologies), and the SE was determined. For those prescribed LAMP, AL measurements were performed using an IOL Master 600 (Carl Zeiss Meditec Inc.). Myopia progression was quantified as the annual change in spherical equivalent refraction (ΔSER) (D/yr), with ΔSER1 representing the change from 2019 to 2020 (during the COVID-19 pandemic), ΔSER2 from 2020 to 2021, ΔSER3 from 2021 to 2022, and ΔSER4 from 2022 to 2023. AL elongation was assessed over the same time intervals. Parents/caregivers also provided data on digital device usage, outdoor activity, and remote education duration through questionnaires.

Statistical analysis

Statistical analyses were conducted using R software ver. 4.5.0 (R Foundation for Statistical Computing). Continuous variables are presented as mean ± standard deviation, while categorical data are shown as frequencies and percentages. Demographic variables were compared using t-tests and chi-square tests. To assess annual ΔSER and annual axial length elongation (ΔALR), we employed linear mixed models with random intercepts for each subject to account for repeated measures. Post hoc pairwise comparisons were conducted using Tukey method to identify specific time periods with significant differences.

We analyzed changes in screen time and outdoor activity across the three periods using one-way repeated measures ANOVA followed by post hoc Tukey tests. Participants were also stratified into subgroups based on median splits of remote education duration (≤12 months vs. >12 months), screen time increase (≤three-fold vs. >three-fold), and outdoor activity decrease (≤0.5-fold vs. >0.5-fold). Differences in myopia progression between subgroups were evaluated using linear mixed models.

To identify factors associated with myopia progression during the post-COVID-19 period (ΔSER3), univariable and multivariable linear mixed-effects regression analyses were conducted. The rationale for using mixed models rather than treating fellow eyes as independent observations was supported by intraclass correlation coefficients (ICC) (2, 1) calculated from our dataset. The ICC for baseline SE was 0.911 (95% confidence interval, 0.860–0.950), and ICCs for annual progression rates (ΔSER1–ΔSER4) ranged from 0.395 to 0.566 (all p < 0.001), confirming meaningful inter-eye correlation. Predictor variables included age, sex, baseline SE, remote education duration, screen time increase, and outdoor activity. Variables with p < 0.1 in univariable analysis were incorporated into multivariable models. A two-sided p-value < 0.05 was considered statistically significant.

Results

A total of 140 eyes from 77 children (39 girls and 38 boys) were analyzed, with 41 eyes receiving LAMP intervention. Of the 77 patients included in the final analysis, 37 patients (69 eyes) completed the behavioral questionnaire and were included in behavioral analyses; the remaining 40 patients (71 eyes) were included only in refractive analyses (Fig. 1). No significant differences were found in age (11.8 ± 2.0 years vs. 12.4 ± 2.0 years, p = 0.172), sex distribution (p = 0.729), or baseline SE (−3.41 ± 2.29 D vs. −3.51 ± 2.04 D, p = 0.783). The mean age of the participants was 12.1 ± 2.0 years, and the baseline SE was −3.4 ± 2.2 D. There were no significant differences in age or sex distribution between the LAMP and non-LAMP groups (Table 1).

Fig. 1.

Flowchart of patient selection. A total of 93 patients (186 eyes) were assessed for eligibility. One patient (7 eyes) was excluded due to implausible refractive progression (annual change in SE > +1.0 D). Nine patients (14 eyes) were further excluded for insufficient myopia (baseline SE ≥ −0.50 D). Ten patients (25 eyes) in the LAMP group were excluded due to atropine initiation after COVID-19 onset or a first-year progression rate of less than −0.75 D/yr suggesting suboptimal treatment response. Ultimately, 77 patients (140 eyes) were included in the final analysis: 53 patients (99 eyes) in the LAMP (−) group and 24 patients (41 eyes) in the LAMP (+) group. Of these, 37 patients (69 eyes) completed the behavioral questionnaire and were included in behavioral analyses, while 40 patients (71 eyes) who did not complete the questionnaire were included in refractive analyses only. No significant differences in age, sex, or baseline SE were found between questionnaire responder and nonresponder (all p > 0.05). SE = spherical equivalent refraction; D = diopters; LAMP = low-concentration atropine for myopia progression.

Table 1.

Baseline characteristics and annual myopia progression in children with and without LAMP treatment

Characteristic	LAMP (−) group*	LAMP (+) group†	p-value
Age (yr)	12.2 ± 2.2	11.9 ± 1.5	0.487
Sex			>0.999
Female	27 (50.9)	12 (50.0)
Male	26 (49.1)	12 (50.0)
Baseline SE (D)	−2.78 ± 1.57	−5.11 ± 2.51	<0.001‡
ΔSER1 (D/yr)	−0.47 ± 0.51	−0.24 ± 0.32	0.035
ΔSER2 (D/yr)	−0.49 ± 0.50	−0.40 ± 0.52	0.349
ΔSER3 (D/yr)	−0.55 ± 0.55	−0.61 ± 0.56	0.574
ΔSER4 (D/yr)	−0.33 ± 0.40	−0.30 ± 0.44	0.825

Values are presented as mean ± standard deviation or number (%). Comparisons of age and sex were based on 77 patients, not eyes. Age was compared using an independent t-test, and sex distribution was compared using the chi-square test. Comparisons of baseline SE and annual SE changes (ΔSER1 to ΔSER4) were based on eye-level data and analyzed using a linear mixed model to account for repeated measures. ΔSER1 representing the change from 2019 to 2020; ΔSER2, 2020–2021; ΔSER3, 2021–2022; and ΔSER4, 2022–2023. The LAMP (+) group had significantly greater baseline myopia and slower progression during ΔSER1 (p = 0.035), with no significant differences observed in subsequent intervals. Eye-level comparisons were analyzed using linear mixed models with random intercepts for subject ID to account for inter-eye correlation (see Statistical Methods).

LAMP = low-concentration atropine for myopia progression; SE = spherical equivalent; D = diopters; ΔSER = change in spherical equivalent refraction.

^*LAMP (−) group: 53 patients, 99 eyes;

^†LAMP (+) group: 24 patients, 41 eyes;

^‡Statistically significant (p < 0.05).

Myopia progression was assessed by calculating the ΔSER. The mean ΔSER was −0.40 ± 0.47 D for ΔSER1 (2019–2020), −0.47 ± 0.51 D for ΔSER2 (2020–2021), −0.56 ± 0.55 D for ΔSER3 (2021–2022), and −0.32 ± 0.41 D for ΔSER4 (2022–2023) (p < 0.001). Post hoc analysis indicated that myopic progression was most pronounced between 2021 and 2022 (ΔSER3) compared to the pre-COVID-19 period (ΔSER1, p = 0.003) (Fig. 2). Additionally, ΔSER4 was significantly greater (less myopic change) than ΔSER3 (p < 0.001).

Fig. 2.

Annual changes in spherical equivalent refraction (ΔSER) in the control group from 2019 to 2023. The line graph shows the mean annual ΔSER across four time intervals: ΔSER1 (2019–2020), ΔSER2 (2020–2021), ΔSER3 (2021–2022), and ΔSER4 (2022–2023). Values are presented as mean ± standard deviation. The rate of myopia progression peaked during ΔSER3, which was significantly higher than ΔSER1 (p = 0.003). No statistically significant differences were observed among the other intervals. ns = not significant; D = diopters. Statistically significant (*p < 0.05).

Based on questionnaire data, the mean duration of remote education was found to be 12.9 ± 6.6 months. Mobile device usage increased significantly during the COVID-19 pandemic. The average daily usage was 1.64 ± 1.03 hours before the pandemic, which increased to 4.69 ± 2.09 hours during the pandemic, and then partially decreased to 2.41 ± 1.30 hours post-COVID-19 pandemic, with an overall significant difference across the time points (p < 0.001) (Fig. 3). Post hoc analysis revealed that screen time during the pandemic was significantly higher compared to both the pre-COVID-19 pandemic ( p < 0.001) and post-COVID-19 pandemic periods (p < 0.001), while the difference between pre- and post-COVID-19 pandemic screen times approached but did not reach statistical significance (p = 0.090).

Fig. 3.

Changes in daily screen time and outdoor activity before, during, and after the COVID-19 pandemic. Mean daily hours spent on screen (left) and outdoor activity (right) are shown for three periods: pre-COVID-19, COVID-19, and post-COVID-19. During the COVID-19 pandemic, screen time increased significantly while outdoor activity decreased significantly compared to the pre-COVID-19 period (all p < 0.001). Post-COVID-19, screen time decreased (p < 0.001), while outdoor activity increased but did not fully return to pre-COVID-19 levels (p = 0.012). Error bars represent standard deviation. Statistical comparisons were performed using one-way ANOVA with post hoc Tukey multiple comparison tests. Statistically significant (*p < 0.05, ***p < 0.001).

Outdoor activity showed the opposite trend. Average time spent outdoors was 2.35 ± 1.37 hr/day before the pandemic, which decreased to 1.01 ± 0.77 hr/day during the pandemic, and partially recovered to 1.86 ± 1.49 hr/day post-COVID-19 pandemic, with a significant difference across time points (p < 0.001) (Fig. 3). Post hoc comparisons demonstrated a significant reduction in outdoor activity during the pandemic compared to the pre-COVID-19 pandemic period (p < 0.001), and a significant increase post-COVID-19 pandemic compared to during the pandemic (p = 0.012). However, the difference between pre-and post-COVID-19 pandemic outdoor activity was not statistically significant (p = 0.220).

Table 1 shows the baseline SE and ΔSER1 to ΔSER4 for both the LAMP and non-LAMP groups. The LAMP group exhibited a significantly myopic baseline SE compared to the non-LAMP group ( p < 0.001). The LAMP group showed a smaller myopic change before the COVID-19 pandemic (ΔSER1, p = 0.035). However, no significant difference in ΔSER was found between the two groups since the onset of the COVID-19 pandemic (Fig. 4A). In the LAMP group, changes in AL were analyzed and revealed that AL elongation between 2021 and 2022 (ΔALR3) was significantly greater than that between 2021 and the COVID-19 pandemic (ΔALR2, p = 0.032) (Supplementary Table 1).

Fig. 4.

Annual myopia progression (ΔSER) stratified by LAMP use, remote education duration, and mobile device usage. (A) Annual ΔSER for children with and without LAMP treatment. The LAMP (+) group showed significantly reduced myopia progression compared to the LAMP (−) group during ΔSER1 (p = 0.035), with no significant differences observed in subsequent years. (B) ΔSER stratified by duration of remote education: ≤12 months vs. >12 months. Children with >12 months of remote learning exhibited significantly greater myopia progression during ΔSER3 (p = 0.023). (C) ΔSER stratified by increase in mobile device usage: ≤3-fold vs. >3-fold increase after COVID-19 onset. Greater screen time (>3-fold) was associated with significantly greater myopia progression during ΔSER3 (p = 0.005). ΔSER1 representing the change from 2019 to 2020; ΔSER2, 2020–2021; ΔSER3, 2021–2022; and ΔSER4, 2022–2023. ns = not significant; D = diopters; LAMP = low-concentration atropine for myopia progression; ΔSER = change in spherical equivalent refraction. Statistically significant (*p < 0.05, **p < 0.01, ***p < 0.001).

A longer duration of remote education was associated with greater myopia progression, particularly during the post-COVID-19 period. Children engaged in remote education for more than 12 months exhibited significantly greater myopic progression during the 2021–2022 interval (ΔSER3) compared to those with shorter remote education duration (p = 0.033) (Fig. 4B). Similarly, children whose screen time increased more than three-fold during the COVID-19 pandemic relative to pre-COVID-19 pandemic levels showed significantly greater myopic progression during the same interval (ΔSER3, p = 0.005) (Fig. 4C). In contrast, the degree of reduction in outdoor activity was not significantly associated with myopia progression during the post-COVID-19 period (Supplementary Fig. 1).

To identify factors associated with myopia progression during the post-COVID-19 period (ΔSER3), we conducted univariable and multivariable linear regression analyses. In the univariable analysis, younger age, higher baseline myopia (more negative SE), and a longer duration of remote education were significantly associated with increased myopic progression. However, in the multivariable model, only remote education duration emerged as a significant predictor of ΔSER3, suggesting that prolonged remote learning was independently associated with accelerated myopia progression (Table 2).

Table 2.

Univariable and multivariable linear mixed model results for myopia progression (ΔSER3)

Variable	Univariable estimate (β)	Std.Err	p-value	Multivariable estimate (β)	Std.Err	p-value
LAMP (yes vs. no)	−0.067	0.119	0.574
Age (yr)	0.073	0.026	0.007*	0.025	0.038	0.514
Sex (male vs. female)	−0.135	0.108	0.215
Remote education (mon)	−0.033	0.011	0.006*	−0.032	0.011	0.007*
Screen time increases (folds)	−0.050	0.032	0.122
Outdoor activity decreases (folds)	0.244	0.277	0.385
Baseline SE (D)	−0.058	0.023	0.015*	−0.032	0.031	0.304

Estimates represent the effect of each variable on annual change in spherical equivalent refraction (D/yr). Random intercepts for subject IDs were included in all models.

ΔSER3 = annual change in spherical equivalent refraction between 2021 and 2022; Std.Err = standard error; LAMP = low-concentration atropine for myopia progression; SE = spherical equivalent; D = diopters.

^*Statistically significant (p < 0.05).

Discussion

This study examined the impact of behavioral changes during the COVID-19 pandemic on myopia progression in children. We specifically focused on the following parameters—use of mobile devices, decreased outdoor activity, and remote education—to determine their contributions to worsening myopia. Additionally, we compared myopia progression between children treated with LAMP and those without, assessing whether the pandemic altered treatment outcomes. Our findings indicate a significant acceleration in myopia progression associated with the pandemic, particularly evident during the 2021–2022 period, rather than immediately following the onset of the pandemic. We also found that increased mobile device use and remote education were linked to myopia progression, whereas decreased outdoor activity was not.

The COVID-19 pandemic led to significant behavioral changes, notably a decline in outdoor activities and an increase in digital screen time [16–18]. According to the questionnaire results, the average screen time increased from 1.6 to 4.7 hr/day during the pandemic, while outdoor activity decreased from 2.4 to 1.0 hr/day. These two factors emerged as key contributors to myopic progression. Previous studies corroborate these changes; one group in Hong Kong reported a drop in outdoor activity from 1.27 to 0.41 hr/day and a doubling of screen time from 2.45 to 6.89 hr/ day, correlating with a myopic shift of −0.50 D in SE and 0.29 mm in AL over 8 months [17]. A large cross-sectional study in China similarly found a significant myopic shift (approximately −0.3 D) and a 1.4 to 3.0 times increase in myopia prevalence among younger children aged 6 to 8 years in 2020 compared to previous years [10]. Furthermore, a meta-analysis encompassing 33 studies worldwide established a general rise in myopia prevalence (pooled odds ratio of 1.11), a decrease in SE of −0.61 D, and an increase in AL of 0.42 mm, associated with an average increase of 6.25 hr/day in screen time and a decrease of 1.52 hr/day in outdoor activity [12].

A key finding of our study is the delayed manifestation of myopia progression, with the most significant refractive change occurring in 2021–2022 (ΔSER3), rather than in the immediate post-COVID-19 period (2020–2021, ΔSER2). Previous studies have reported myopic changes after the pandemic; however, most focused on short-term follow-ups, with few reporting a long-term follow-up [17,19–25]. This temporal lag suggests that the cumulative effects of increased near work and reduced outdoor exposure during the pandemic may require time to translate into observable refractive changes. Moreover, the continuation of high-risk behaviors after the pandemic—such as excessive screen time and reduced outdoor activity—may have contributed to this delayed but pronounced myopic progression.

Several biological mechanisms may underlie the delayed manifestation of myopia progression observed in our cohort. First, cumulative changes in retinal dopamine signaling which is stimulated by bright light exposure and suppressed by prolonged near work, modulate ocular growth inhibition over time, suggesting that environmental shifts may not immediately translate into refractive changes but require sustained exposure to alter growth signals [26]. Second, alterations in choroidal blood flow and thickness in response to changed outdoor activity and light exposure have been implicated in regulating axial elongation, with structural adaptation occurring gradually rather than acutely [27]. Finally, molecular-level mechanisms such as scleral endoplasmic reticulum stress and extracellular matrix remodeling have been shown to control AL growth in animal models, indicating that protracted biomechanical and cellular responses to altered visual environments may contribute to delayed refractive adaptation [28].

Our results also highlight that increased screen time and remote education may diminish the efficacy of LAMP. Before the pandemic, the LAMP group demonstrated significantly less myopia progression compared to the non-LAMP group (ΔSER1). However, during and after the pandemic (ΔSER2 to ΔSER4), the progression rate in the LAMP group increased to levels comparable with the non-LAMP group. This observation aligns with findings from the Western Australia atropine for myopia study, which noted a diminished effect of 0.01% atropine during lockdown [29]. These findings suggest a possibility that behavioral factors, particularly associated with increased near work, may influence the apparent effectiveness of LAMP under certain environmental conditions. However, given the observational nature of this study, further prospective investigations should be necessary to clarify whether modification in atropine dosing or the addition of adjunctive interventions is beneficial.

Behavioral data provided additional insights into potential modifiable risk factors. In our study, children demonstrating a more than three-fold increase in screen time or engaging in remote learning for over 12 months exhibited significantly greater myopia progression during 2021–2022. One study indicated that the increased digital screen exposure contributed to myopic progression during the COVID-19 pandemic [24]. Another investigation among Chinese children observed a high odds ratio of myopic progression associated with online learning duration and digital screen use [22]. Similarly, a population-based prospective study also showed that reading and writing time were related to the changes in SE and AL [17]. Our results corroborate previous findings of a positive relationship between increasing screen time and myopic progression, highlighting the need for public health strategies aimed at mitigating the visual impacts of increased near work.

In contrast, changes in outdoor activity did not demonstrate a significant association with myopia progression. Although previous studies have established outdoor exposure as a protective factor against myopia onset, its impact on progression—especially among already myopic children—appears more limited and less consistent. One study examining risk factors for rapid myopia progression during the COVID-19 lockdown reported that decreased sun exposure (<1 hr/day) was significantly associated with progression [30]. Another study also reported that spending at least 2 hours outdoors had a protective effect on myopia progression during the pandemic [19]. However, a study conducted in China failed to demonstrate significant effects of outdoor activities on myopia progression [22]. These contradictory results imply that the crucial aspect may not be the time spent outdoors, but rather the activities undertaken outdoors, such as sun exposure. Additionally, variations due to race or cultural background could affect the results.

We also analyzed the AL changes in the LAMP group, which showed progressive elongation post-COVID-19 pandemic. Notably, AL elongation between 2021 and 2022 (ΔALR3) was significantly greater than that between 2020 and 2021 (ΔALR2), paralleling refractive changes. One previous study showed a smaller AL change between 2020 and 2021 (0.01 mm) compared to that between 2019 and 2020 (0.03 mm), suggesting variability in AL response during different phases of the pandemic [31]. Therefore, further research is warranted to explore the factors associated with AL changes. Additionally, while younger age and higher baseline myopia were associated with greater progression in univariable analyses, these associations did not reach significance in multivariable models due to the limited sample size.

Despite the strengths of this study, including its longitudinal design and integration of behavioral data, several limitations should be mentioned. First, AL measurements were only available for the LAMP group, limiting direct comparison of structural progression between groups. Because AL elongation is considered one of the most important objective indicators of myopia progression, the absence of AL data in non-LAMP group restricts interpretation of structural changes across our study’s entire cohort. Second, behavioral data such as screen time and outdoor activity were collected via retrospective questionnaires, which may be susceptible to recall bias and parental reporting inaccuracies; whereas the duration of remote education is typically anchored to structured schedules and therefore may be more reliably reported. Accordingly, findings based on individual sub-questionnaire items should be interpreted with caution. Third, the sample size was relatively small, particularly after stratification into subgroups based on behavioral factors. It may potentially limit the statistical power to detect subtle differences or interactions, especially in multivariable models. Furthermore, because participants who discontinued LAMP contributed data only during the period of actual use and those with dose escalation were analyzed according to their initial starting concentration, the estimated treatment effect may not fully reflect true exposure intensity. This methodological approach may therefore limit the interpretability of the observed LAMP efficacy. Lastly, although we observed associations between behavioral factors and myopia progression, causality cannot be established due to the observational nature of this study. Future prospective studies with objective activity monitoring and larger samples would strengthen and validate these findings.

In conclusion, our findings indicate that the COVID-19 pandemic accelerated myopia progression in children. We identified a critical window (2021–2022) during which myopia progression was most pronounced compared to the periods during or immediately after the pandemic. Increased screen time and prolonged remote learning were significantly associated with greater refractive changes. Additionally, the efficacy of LAMP treatment may also have been compromised under these behavioral conditions. These results suggest the need for targeted interventions, including screen time management and adjusted atropine dosing, to mitigate myopia progression in the post-COVID-19 pandemic era.