Ocular Surface Outcomes with Different 0.05% Atropine Dosing Frequencies in Children with Myopia: A Secondary Analysis of a Randomized Trial

Abstract

Introduction

This study aimed to compare the effects of three dosing frequencies of 0.05% atropine on the ocular surface in children with myopia.

Methods

In this sub-analysis of a randomized controlled trial, children with myopia (aged 6–14 years) were randomized to 0.05% atropine once daily (Qd), twice weekly (Biw), or once weekly (Qw). Objective parameters, including noninvasive tear breakup time (NIBUT), tear meniscus height (TMH), lipid layer thickness (LLT), incomplete blink rate (IBR), and bulbar conjunctival hyperemia index (BCHI), were assessed at baseline, 2 weeks, and 12 months. Subjective symptoms were evaluated using the Ocular Surface Disease Index (OSDI).

Results

Among 130 children (260 eyes) completing objective assessments (Qd: n = 43; Biw: n = 49; Qw: n = 38), a representative subset of 78 completed the OSDI questionnaire. Dosing frequency had no differential effect on any outcome (time-by-group interaction: all P > 0.05). Across all groups, first and average NIBUT increased significantly from baseline through 12 months (P < 0.001), with a shift toward less severe dry eye grades (P < 0.001). IBR decreased, indicating improved blink quality (P < 0.001) and BCHI decreased by 12 months (P = 0.001). LLT increased transiently at 2 weeks (P = 0.003), while TMH remained stable (P = 0.712). OSDI scores remained stable, with no significant differences among regimens (P = 0.193).

Conclusions

During the 12-month follow-up period, ocular surface parameters and subjective symptoms remained comparable among the three 0.05% atropine dosing regimens, with no evidence of clinically significant deterioration.

Trial registration

ChiCTR2100043506.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1007/s40123-026-01467-1.

Key Summary Points

Why carry out this study?

Although 0.05% atropine provides an effective balance between myopia control and tolerability, concerns remain regarding its potential long-term effects on tear film stability and the ocular surface, particularly when combined with treatments such as orthokeratology.

This study evaluated whether once-daily, twice-weekly, and once-weekly dosing of 0.05% atropine differed in their effects on ocular surface parameters and symptoms in children with myopia over 12 months.

What was learned from the study?

Dosing frequency did not significantly alter objective or subjective outcomes (all P > 0.05). Over 12 months across all treatment groups, noninvasive tear breakup time improved (P < 0.001), incomplete blink rate decreased (P < 0.001), and Ocular Surface Disease Index (OSDI) scores remained stable (P = 0.193).

Once-daily, twice-weekly, and once-weekly 0.05% atropine showed comparable ocular surface tolerability, with no detectable deterioration in the measured parameters or symptoms. Because no placebo or untreated control group was included, the observed improvements in tear film stability cannot be attributed to atropine treatment.

Introduction

Low-concentration atropine is widely regarded as an effective pharmaceutical method for controlling myopia in children [1, 2]. However, the requirement for long-term daily dosing has raised concerns about its safety profile. Beyond frequent visual side effects such as photophobia and attenuated near vision, atropine’s long-term impact on the ocular surface microenvironment remains a major concern.

As a nonselective muscarinic antagonist, atropine competitively blocks lacrimal gland receptors, inhibiting tear secretion and disrupting tear film integrity [3]. Maintaining a steady ocular surface is especially important as multimodal techniques, such as combining atropine with orthokeratology or contact lenses, become more widespread [4, 5]. As a result, reduced ocular surface integrity could compromise the ocular surface tolerability and long-term viability of these combined therapies.

While 0.01% atropine exhibits a favorable safety profile with minimal ocular surface impact [6, 7], 0.05% atropine has emerged as a potentially optimal concentration, offering a better balance between efficacy and visual side effects [8, 9]. However, recent data indicate that 0.05% atropine may induce greater transient tear film instability in young adults compared with lower concentrations [10]. It remains unclear whether similar ocular surface disturbances occur in children, and whether altering the dosing frequency could mitigate these potential adverse effects. To address this concern, we conducted a randomized clinical trial to evaluate the effects of different dosing frequencies of 0.05% atropine on the ocular surface. We hypothesized that reducing the instillation frequency could preserve ocular surface homeostasis without requiring a reduction in atropine concentration. While the refractive efficacy and visual function outcomes of this cohort have been reported elsewhere [11, 12], to date, no longitudinal study has systematically compared the ocular surface effects of different frequencies of 0.05% atropine in children. The present study aims to evaluate both objective ocular surface parameters, including lipid layer thickness, tear volume, and stability, and subjective ocular symptoms via the Ocular Surface Disease Index (OSDI), in children with myopia assigned to receive 0.05% atropine once daily, twice weekly, or once weekly over a 12-month period.

Methods

This study was conducted as a pre-specified secondary analysis of ocular surface health data derived from the randomized controlled trial “Clinical observation and study of 0.05% atropine inhibiting myopia” [11, 12] (Registration No: ChiCTR2100043506). The trial protocol was approved by the Ethics Committee of Changsha Aier Eye Hospital (ID: KYPJ021) and adhered to the tenets of the Declaration of Helsinki. Written informed consent was obtained from the parents or legal guardians of all participants for study participation and the use of their clinical data.

Participants

Children with myopia aged 6–14 years were enrolled between March 2021 and August 2023. The key inclusion criteria were: cycloplegic spherical equivalent (SE) refraction of −1.00 D or more myopic (≤  −1.00 D) in both eyes. Exclusion criteria included: history of other myopia control interventions (e.g., orthokeratology, atropine of other concentrations); coexisting ocular pathologies (e.g., amblyopia, strabismus); clinically significant dry eye disease; or known allergy to atropine or its excipients.

For the specific purpose of this longitudinal ocular surface analysis, participants were included in the final statistical analysis only if they met the following additional criteria: (1) completion of all scheduled visits (baseline, 2 weeks, and 12 months); and (2) availability of high-quality, readable images from both LipiView II and Keratograph 5M at every time point. Participants with incomplete follow-up or imaging data that could not be reliably analyzed were excluded from the objective complete case analysis. Some images could not be analyzed because of difficulties with cooperation during pediatric image acquisition. A total of 130 participants met the criteria for the complete case analysis and were included in the analysis of objective ocular surface signs (Supplementary Figure S1). Among them, 78 (60.0%) completed the subjective questionnaire at all time points and constituted the Questionnaire Subset.

Intervention

Eligible participants were randomly assigned to one of three intervention groups using a computer-generated sequence: the once-daily group (Qd), instilling eye drops every night; the twice-weekly group (Biw), instilling eye drops on two fixed nights per week; and the once-weekly group (Qw), instilling eye drops on one fixed night per week. Preservative-free 0.05% atropine sulfate eye drops (containing sodium chloride and dilute hydrochloric acid for pH adjustment) were prepared by the Changsha Aier Eye Hospital. All participants wore single-vision spectacles.

Ocular Surface Parameter Measurements

The LipiView II Ocular Surface Interferometer (Johnson & Johnson Vision, Jacksonville, FL, USA) was used to measure the average lipid layer thickness (LLT) and to record blink patterns for calculating the incomplete blink rate (IBR). The Keratograph 5M (Oculus GmbH, Wetzlar, Germany) was employed to assess tear meniscus height (TMH), first noninvasive breakup time (NIBUT-first), average noninvasive breakup time (NIBUT-avg), and bulbar conjunctival hyperemia index (BCHI). Dry eye severity was graded based on NIBUT-avg values: grade 0 (normal): ≥ 14 s; grade 1 (mild): 7–13 s; grade 2 (moderate): < 7 s. Because the treatment groups followed different dosing schedules, the interval between the most recent atropine instillation and the ocular surface assessment was not standardized across groups. Subjective ocular symptoms were evaluated using the Ocular Surface Disease Index (OSDI) questionnaire in the questionnaire subset at baseline, 2 weeks, and 12 months. The OSDI is a validated 12-item questionnaire that assesses symptoms of ocular discomfort, functional limitations, and environmental triggers. The total score ranges from 0 to 100, with higher scores indicating more severe symptoms. Children completed the questionnaire with parental assistance when necessary, and trained clinical examiners provided clarification to ensure that each item was understood before responses were recorded. All ocular surface imaging and subjective grading were performed by trained clinical examiners who were masked to the participants’ treatment group assignments. For this secondary analysis, NIBUT-avg was defined as the primary ocular surface endpoint. Secondary outcomes included NIBUT-first, LLT, IBR, TMH, BCHI, and dry eye severity grade. OSDI score was analyzed as an exploratory subjective outcome because it was available only in the questionnaire subset.

Statistical Analysis

Statistical analyses were performed using IBM SPSS Statistics (Version 25.0, IBM Corp., Armonk, NY, USA). Continuous variables are presented as mean ± standard deviation, and categorical variables as frequencies and percentages.

The baseline demographic and clinical parameters were compared across groups using one-way ANOVA for continuous variables and Chi-square tests for categorical data. To ensure that the questionnaire subset was representative, a nonresponder analysis was performed, comparing participants who completed the survey to those who did not, using independent sample t-tests and Chi-square tests. To account for within-subject inter-eye correlations, generalized estimating equations (GEE) were used on objective ocular surface metrics across the entire cohort. These models used treatment group, time (baseline, 2 weeks, and 12 months), and group-by-time interactions as predictors. The ordinal outcome of dry eye severity was analyzed using a GEE model with an ordinal logistic link function. Subjective OSDI scores (questionnaire subset) were analyzed using repeated-measures analysis of covariance (ANCOVA). To account for any discovered baseline imbalances, baseline age was used as a covariate to analyze the main effects of group and time, as well as their interaction. Post hoc pairwise comparisons were made using Bonferroni adjustments. A two-sided P value < 0.05 indicated statistical significance.

Results

Study Cohort and Baseline Characteristics

From the initial cohort of 205 participants, a total of 130 participants (260 eyes) met the strict inclusion criteria for the objective ocular surface analysis (Full Cohort). These participants attended all scheduled visits and provided high-quality imaging data. The final analytical cohort consisted of 43 participants in the Qd group, 49 in the Biw group, and 38 in the Qw group. Within this full cohort, 78 participants (60.0%) completed the OSDI questionnaire at all time points (Questionnaire Subset). To ensure that this subset was not biased, a nonresponder analysis was performed. Comparison between the questionnaire subset (n = 78) and the noncompleters (n = 52) revealed no significant differences in baseline age, sex, or objective ocular surface parameters (all P > 0.05, Supplementary Table 1). This confirms that the subset used for subjective analysis was representative of the full study population. Baseline demographic and clinical characteristics of the full cohort were well-balanced across the three treatment groups, with no significant differences in age, sex, or any baseline ocular surface measure (all P > 0.05, Table 1). Adverse events, treatment discontinuations, and adherence have been reported in our previous trial publications [12]. These data are referenced here to provide clinical context, whereas the present secondary analysis focuses specifically on ocular surface tolerability.

Table 1.

Baseline demographics and ocular surface characteristics of the full cohort (n = 130 participants, 260 eyes)

Characteristica	Qd group	Biw group	Qw group	Pb
n (eyes)	86	98	76
Age (years)	10.58 (1.83)	10.22 (1.91)	10.03 (1.87)	0.40
Sex, n (%)				> 0.99
Male	18 (42%)	21 (43%)	16 (42%)
Female	25 (58%)	28 (57%)	22 (58%)
LLT, nm	60.44 (24.67)	64.03 (21.21)	62.64 (22.18)	0.71
IBR	0.70 (0.31)	0.61 (0.36)	0.63 (0.33)	0.33
TMH (mm)	0.21 (0.06)	0.21 (0.05)	0.21 (0.04)	0.79
NIBUT-first (s)	8.82 (5.77)	10.56 (7.22)	10.68 (6.78)	0.16
NIBUT-avg (s)	11.90 (5.98)	13.21 (7.07)	13.62 (6.29)	0.34
Dry eye gradec, n (%)				0.43
Grade 0	31 (36.0)	48 (49.0)	39 (51.3)
Grade 1	33 (38.4)	24 (24.5)	20 (26.3)
Grade 2	22 (25.6)	26 (26.5)	17 (22.4)
BCHI	0.58 (0.27)	0.62 (0.31)	0.61 (0.27)	0.78

Qd once nightly, Biw twice weekly, Qw once weekly, LLT lipid layer thickness, IBR incomplete blink rate, TMH tear meniscus height, NIBUT-first first noninvasive breakup time, NIBUT-avg average noninvasive breakup time, BCHI bulbar conjunctival hyperemia index

^aUnless otherwise indicated, data are expressed as mean (SD)

^bP values: age was analyzed using one-way ANOVA; sex and dry eye grade were analyzed using the Chi-square test; all other ocular surface parameters were analyzed using generalized estimating equations (GEE) to account for inter-eye correlations

^cDry eye severity grading was based on NIBUT-avg thresholds: grade 0 ≥ 14 s; grade 1 = 7–13 s; grade 2 < 7 s

Average Lipid Layer Thickness

Average LLT exhibited a significant main effect of time (P = 0.004). Post hoc pairwise comparisons indicated that LLT increased significantly from baseline to 2 weeks (P = 0.003). However, LLT at 12 months did not differ significantly from either baseline or the 2-week visit (all P > 0.05). No significant main effect of group or time-by-group interaction was detected (all P > 0.05), suggesting that the initial increase in lipid layer thickness was transient and levels returned to baseline by the end of the study (Fig. 1A, Table 2).

Fig.1.

Longitudinal changes in ocular surface parameters following 12-month atropine treatment with different dosing regimens. A Average lipid layer thickness (LLT). B Incomplete blink rate (IBR). C Tear meniscus height (TMH). D Bulbar conjunctival hyperemia index (BCHI). Data are presented as estimated marginal means ± standard error of the mean (SEM; error bars). P-values indicate the main effect of time (Time), group (Group), and their interaction (Int) derived from GEE models. *P < 0.05, **P < 0.01, ***P < 0.001 versus baseline. Qd once nightly, Biw twice weekly, Qw once weekly

Table 2.

Longitudinal changes in objective ocular surface parameters in the full cohort (n = 130 participants, 260 eyes)

Parameter	Group	Baseline EMMa (95% CI)	2 Weeks EMM (95% CI)	12 Months EMM (95% CI)	P Valueb (time)	P Valueb (group)	P Valueb (interaction)
LLT (nm)	Qd	60.86 (54.16, 67.56)	68.03 (62.35, 73.71)	65.83 (58.99, 72.66)	0.004	0.43	0.93
	Biw	64.37 (59.29, 69.44)	71.37 (65.58, 77.15)	70.05 (64.66, 75.44)
	Qw	63.24 (56.90, 69.59)	67.28 (59.53, 75.02)	65.36 (58.95, 71.78)
IBR	Qd	0.70 (0.62, 0.78)	0.53 (0.45, 0.62)	0.57 (0.48, 0.65)	< .001	0.45	0.17
	Biw	0.62 (0.53, 0.71)	0.62 (0.53, 0.70)	0.53 (0.44, 0.61)
	Qw	0.63 (0.54, 0.72)	0.55 (0.46, 0.64)	0.45 (0.35, 0.55)
TMH (mm)	Qd	0.21 (0.20, 0.23)	0.21 (0.19, 0.22)	0.21 (0.19, 0.23)	0.71	0.29	0.42
	Biw	0.21 (0.20, 0.22)	0.22 (0.20, 0.23)	0.21 (0.19, 0.24)
	Qw	0.20 (0.19, 0.22)	0.20 (0.18, 0.22)	0.19 (0.17, 0.21)
NIBUT-first (s)	Qd	8.83 (7.43, 10.23)	12.41 (10.44, 14.38)	13.84 (11.97, 15.72)	< .001	0.94	0.37
	Biw	10.63 (8.93, 12.33)	11.91 (10.23, 13.59)	13.25 (11.55, 14.96)
	Qw	10.70 (8.97, 12.42)	11.55 (9.47, 13.63)	13.74 (11.61, 15.86)
NIBUT-avg (s)	Qd	11.91 (10.25, 13.57)	14.43 (12.48, 16.38)	15.85 (14.21, 17.50)	< .001	0.64	0.73
	Biw	13.27 (11.49, 15.05)	15.43 (13.88, 16.97)	15.73 (14.18, 17.28)
	Qw	13.59 (11.87, 15.32)	14.54 (12.54, 16.55)	16.29 (14.43, 18.16)
BCHI	Qd	0.75 (0.46, 1.03)	0.74 (0.62, 0.86)	0.61 (0.36, 0.86)	< .001	0.36	0.87
	Biw	0.67 (0.44, 0.89)	0.65 (0.56, 0.74)	0.60 (0.44, 0.76)
	Qw	0.61 (0.48, 0.74)	0.61 (0.50, 0.72)	0.48 (0.39, 0.58)

Qd once nightly, Biw twice weekly, Qw once weekly, LLT lipid layer thickness, IBR incomplete blink rate, TMH tear meniscus height, NIBUT-first first noninvasive breakup time, NIBUT-avg average noninvasive breakup time, BCHI bulbar conjunctival hyperemia index

^aValues represent estimated marginal means (95% CI)

^bP values derived from generalized estimating equations (GEE) for main effects (time, group) and interaction

Incomplete Blink Rate

The IBR decreased significantly during the trial (main effect of time, P < 0.001), indicating an improvement in blink quality. IBR decreased from baseline to 2 weeks (β = 0.085, P = 0.012) and 12 months (β = 0.138, P < 0.001), although there was no significant difference between the two visits (P > 0.05).There was no significant main effect of group or time-by-group interaction (both P > 0.05), and the improvement was consistent across all treatment groups (Fig. 1B, Table 2).

Tear Meniscus Height

TMH levels remained steady throughout the 12-month research period. There were no significant main effects of time or treatment group, nor were there any significant time-by-group interactions (all P > 0.05). Pairwise comparisons revealed no significant differences between any time points, indicating that atropine treatment did not significantly affect TMH (Fig. 1C, Table 2).

Bulbar Conjunctival Hyperemia Index

BCHI demonstrated a significant time-dependent reduction (P < 0.001). Post hoc analysis showed that BCHI was significantly lower at 12 months compared with baseline (β = 0.112, P = 0.001), while the change from baseline to 2 weeks was not statistically significant (P > 0.05). This gradual reduction in conjunctival hyperemia was observed consistently across all dosing regimens  (Fig. 1D, Table 2).

Tear Film Stability and Dry Eye Grade

Both NIBUT-first and NIBUT-avg showed a significant main effect of time (both P < 0.001), with no significant group or interaction effects. NIBUT-first and NIBUT-avg increased significantly from baseline to 2 weeks and remained elevated at 12 months (P < 0.05). No significant changes were found between the 2-week and 12-month time points (Fig. 2A, B, Table 2). Consistent with NIBUT measurements, the distribution of dry eye severity grades shifted significantly toward milder categories over time (P < 0.001). This positive shift was observed across all treatment groups, with a significant main effect of time (P < 0.001), but no significant effect of group or time-by-group interaction (all P > 0.05) (Fig. 2C, Table 3).

Fig.2.

Longitudinal changes in tear film stability and dry eye severity following 12-month atropine treatment with different dosing regimens. A First noninvasive tear breakup time (NIBUT-first). B Average noninvasive tear breakup time (NIBUT-avg). C Stacked bar chart showing the distribution of eyes across dry eye severity grades (grade 0, 1, and 2) at baseline (BL), 2 weeks (2W), and 12 months (12M) for each treatment group. Data are presented as estimated marginal means ± standard error of the mean (SEM; error bars). P-values indicate the main effect of time (Time), group (Group), and their interaction (Int) derived from GEE models. *P < 0.05, **P < 0.01, ***P < 0.001 versus Baseline. Qd once nightly, Biw twice weekly, Qw once weekly

Table 3.

Longitudinal changes in dry eye severity grade distribution in the full cohort (n = 130 participants, 260 eyes)

Dosing group	Dry eye gradea	Baselineb n (%)	2 Weeksb n (%)	12 Monthsb n (%)
Qd (n = 86)	Grade 0 (Normal)	31 (36.0)	42 (48.8)	56 (65.1)
	Grade 1 (Mild)	33 (38.4)	28 (32.6)	23 (26.7)
	Grade 2 (Moderate)	22 (25.6)	16 (18.6)	7 (8.1)
Biw (n = 98)	Grade 0 (Normal)	48 (49.0)	61 (62.2)	61 (62.2)
	Grade 1 (Mild)	24 (24.5)	22 (22.4)	28 (28.6)
	Grade 2 (Moderate)	26 (26.5)	15 (15.3)	9 (9.2)
Qw (n = 76)	Grade 0 (Normal)	39 (51.3)	39 (51.3)	47 (61.8)
	Grade 1 (Mild)	20 (26.3)	24 (31.6)	23 (30.3)
	Grade 2 (Moderate)	17 (22.4)	13 (17.1)	6 (7.9)
Overall P valuec	Time: < .001	Group Effect: 0.25	Time × Group: 0.78

Qd once nightly, Biw twice weekly, Qw once weekly

^aGrading criteria (average noninvasive breakup time,NIBUT-avg): grade 0 ≥ 14 s; grade 1 = 7–13 s; grade 2 < 7 s

^bData are presented as number of eyes (percentage of eyes within the group at that time point). Percentages may not total 100 owing to rounding

^cP values were derived from a generalized estimating equations (GEE) model

Subjective Ocular Symptoms (OSDI Sub-analysis)

Analysis of subjective symptoms was restricted to the questionnaire subset (n = 78). Although a baseline imbalance in age was identified in this subset (P = 0.043), a repeated-measures ANCOVA adjusting for baseline age revealed no significant main effect of treatment group (F = 1.683, P = 0.193) or time-by-group interaction (F = 0.728, P = 0.560). The adjusted mean OSDI scores were comparable across the Qd, Biw, and Qw groups at both 2 weeks (P = 0.779) and 12 months (P = 0.391), indicating that higher dosing frequency did not induce greater subjective discomfort (Table 4).

Table 4.

Longitudinal changes in subjective ocular symptoms (OSDI Score) in the questionnaire subset (n = 78)

Parameter	Group	Baseline EMMa (95% CI)	2 Weeks EMM (95% CI)	12 Months EMM (95% CI)	P valueb (time)	P valueb (group)	P valueb (interaction)
OSDI	Qd (n = 28)	4.01 (1.82–6.21)	4.35 (2.58–6.12)	4.20 (1.62–6.78)	0.398	0.193	0.560
	Biw (n = 25)	6.61 (4.34–8.88)	5.07 (3.25–6.90)	7.29 (4.62–9.95)
	Qw (n = 25)	4.13 (1.83–6.42)	5.13 (3.29–6.98)	6.14 (3.44–8.83)

Qd once nightly, Biw twice weekly, Qw once weekly, CI confidence interval, EMM estimated marginal means, OSDI Ocular Surface Disease Index

^aValues are presented as estimated marginal means (95% CI)

^bP values were derived from repeated-measures analysis of covariance (ANCOVA), adjusted for baseline age

Discussion

This study evaluated the ocular surface tolerability of 0.05% atropine at different dosing frequencies over 12 months. Although anticholinergic medications may affect tear function, the present findings showed no detectable deterioration in the measured ocular surface parameters during the 12-month follow-up. No significant differences were observed among the three dosing regimens in any measured ocular surface parameters or OSDI score. NIBUT and IBR improved over time, and BCHI decreased by 12 months across all groups. These longitudinal changes should be interpreted as observational because no placebo or untreated control group was included. The recent MOSAIC trial showed that nightly 0.05% atropine was associated with a higher frequency of blurred near vision and photophobia, with no increase in treatment discontinuation [9]. The vehicle-controlled STAR trial further showed that photophobia and blurred vision were more common with atropine, particularly at 0.03%, whereas corneal staining remained clinically comparable among groups over 24 months [13]. Our study complements these findings by evaluating ocular surface tolerability across different dosing frequencies of 0.05% atropine.

Lipid Layer Thickness and Incomplete Blink Rate

We observed a transient increase in LLT at 2 weeks, which returned to baseline by 12 months. Despite methodological variations across studies, this functional stability aligns with the structural integrity reported in previous studies: Li et al. found no meibomian gland loss with 0.01% atropine over 12 months [7], and Luo et al. observed stable meiboscores with 0.05% atropine over the short term [10]. By contrast, Li et al. reported different dynamics in their 0.02% group, noting a fluctuation in meibomian gland parameters (a decrease in visibility score at 3 months followed by a rebound at 6 and 12 months) [7]. The authors hypothesized that this later increase might be owing to a complex compensatory mechanism of the meibomian glands [7]. In our study, long-term use of 0.05% atropine did not lead to such fluctuations; instead, the lipid layer remained stable throughout the 12-month period. Concurrently, IBR decreased significantly, indicating a greater proportion of complete blinks. Incomplete blinking is associated with evaporative dry eye and meibomian gland dysfunction, whereas complete blinking facilitates meibum expression and distribution across the ocular surface. These concurrent findings suggest a possible association between blink completeness and lipid layer stability. However, the observational design of this study precludes causal inference regarding the contribution of blink behavior to the observed LLT pattern.

Tear Film Stability and Dry Eye Grade

The most significant finding in our study was the sustained improvement in NIBUT (both first and average) and the corresponding shift toward milder dry eye grades across all groups. This contradicts the traditional pharmacological expectation that atropine, as a competitive muscarinic antagonist, would inhibit lacrimal secretion and induce dry eye. Previous animal studies have shown that 1% atropine can induce dry eye models by blocking neural reflex loops [14]. However, our study revealed that 0.05% atropine did not replicate this deleterious effect in children. Clinical reports on the effects of low concentration atropine on tear film stability are inconsistent. Cheng et al. [6] and Li et al. [15] reported that NIBUT values showed no significant change in children and adult patients with the use of 0.01% atropine. Similarly, Li et al. observed no significant changes in either NIBUT-first or NIBUT-avg within the 0.01% and 0.02% concentration groups throughout the 1-year course of eye drop treatment [7]. Adding to this, Cho et al. found that dynamic tear film stability in atropine-treated children was comparable to healthy controls, with the notable observation that participants treated for more than 2 years presented significantly better stability than those treated for less than 2 years [16]. This aligns with our finding of sustained improvement over time.

By contrast, our results diverge from the adult data reported by Luo et al., where 0.05% atropine caused a transient reduction in NIBUT [10]. The discrepancy between our findings (improvement) and Luo et al.’s (transient worsening) may be attributed to both the age-related decline in tear fluid production and quality and the more rapid recovery of tear film stability in children as highlighted by Cho et al. [10, 16]. Recent computational evidence suggests that atropine related mechanisms in myopia and dry eye disease may overlap through inflammatory signaling and extracellular matrix remodeling [17]. These findings generate testable hypotheses that require validation using tear inflammatory biomarkers and tear osmolarity measurements. These proposed pathways were not evaluated in the present study. The potential effects of formulation characteristics also remain uncertain. The study formulation contained 0.9% sodium chloride and no preservatives. However, because no vehicle only control group was included and tear atropine concentration, tear osmolarity, and inflammatory biomarkers were not measured, the contribution of the formulation to the observed outcomes remains uncertain.

Tear Meniscus Height (TMH)

Our findings demonstrated that TMH maintained stability throughout the 12-month study period, unaffected by dosing frequency. This is consistent with the findings of Cheng et al. and Li et al. (0.01% group) in children [6, 7]. Similarly, Luo et al. (0.05% group) also found no significant change in TMH in adults [10]. However, it contrasts with the findings of Li et al. (0.02% group), who reported a significant reduction in TMH in their 0.02% atropine group at 12 months [7]. The authors attributed this decline primarily to environmental and behavioral factors, such as the dry winter climate and increased academic pressure associated with extended near-work [7]. In our study, despite using a higher concentration of 0.05%, we did not observe such a decline over the same 12-month duration. Furthermore, Cho et al. extended this observation, reporting that children treated with 0.01%–0.05% atropine exhibited TMH levels comparable to healthy controls, with no significant alterations even after treatment durations exceeding 2 years [16].

Theoretically, as a muscarinic acetylcholine receptor antagonist, atropine could inhibit glandular secretion [16, 18]. Cai et al. demonstrated that the transcutaneous application of atropine gel could effectively reduce tear flow and control epiphora in patients with submandibular gland transplantation [18]. However, a key pharmacokinetic distinction exists between the two delivery methods. As Cho et al. highlighted, the transcutaneous gel used by Cai et al. contained permeation enhancers to maintain efficacy for 3–5 h [16]. By contrast, the small volume of topical eye drops is subject to rapid nasolacrimal drainage [16]. Nevertheless, TMH is an indirect indicator of the aqueous tear compartment rather than a direct measure of lacrimal gland function. Therefore, the stable TMH observed in this study should not be interpreted as evidence that 0.05% atropine preserves lacrimal gland function or has no effect on tear secretion.

The current study demonstrated no significant influence of atropine administration frequency on children’s OSDI scores. There were no between-group differences at baseline, 2 weeks, or 12 months, and no modifications from baseline were found at any time point. These data imply that atropine used once a night or less does not enhance subjective ocular discomfort.

Bulbar Conjunctival Hyperemia Index (BCHI)

Our findings are consistent with those of Luo et al., who found no significant change in bulbar redness after 2 weeks dose of 0.05% atropine [10]. In terms of long-term effects, Cheng et al. reported no significant difference in bulbar redness with 0.01% atropine after 6 months [6]. Cho et al. confirmed this tolerability profile, observing that bulbar redness scores in children treated with atropine were not significantly different from those of normal children and were unaffected by treatment duration, even in participants treated for more than 2 years [16]. However, by extending the observation period to 12 months in our study, we discovered a substantial time-dependent drop in the BCHI value compared with the baseline. Although bulbar conjunctival hyperemia is commonly associated with ocular surface inflammation or irritation, the cause and clinical significance of the observed reduction in BCHI remain uncertain. Because no placebo group, vehicle only control group, or untreated control group was included and inflammatory biomarkers were not measured, this reduction cannot be attributed to the preservative free formulation, improved tear film stability, altered blink behavior, or reduced ocular surface inflammation.

Strengths and Limitations

This study supports the comparable 12-month ocular surface tolerability of once-daily 0.05% atropine and reduced-frequency regimens. By integrating objective imaging parameters with subjective symptom scores, we provide evidence that increased dosing frequency was not associated with worse measured ocular surface outcomes. Because this was a secondary analysis of a parent trial powered for refractive efficacy rather than ocular surface outcomes, the ocular surface findings should be interpreted as exploratory. However, the absence of a placebo or untreated control group limits causal interpretation. Therefore, the observed longitudinal changes in NIBUT, IBR, LLT, and BCHI cannot be attributed to 0.05% atropine or its formulation. A major limitation is the potential selection bias associated with the complete-case analysis. Of the 205 randomized participants, only 130 were included in the final analysis of objective ocular surface parameters. The remaining 75 participants were not included because of incomplete follow-up or missing or nonanalyzable imaging data at one or more visits. Some images could not be reliably analyzed because of cooperation difficulties during pediatric image acquisition, including frequent blinking or unstable fixation. However, frequent blinking may also indicate ocular surface discomfort or dry eye. Consequently, children with poorer ocular surface status, lower treatment adherence, or reduced tolerance of treatment or study examinations may have been disproportionately excluded from the complete-case analysis. Although baseline age and sex did not differ significantly between the included and excluded participants, residual selection bias and informative missingness cannot be excluded and should be considered when interpreting the findings. The discrepancy in sample sizes between the full cohort (n = 130) for objective signs and the questionnaire subset (n = 78) for subjective symptoms remains a primary limitation. Although the comparison between included and excluded participants showed no significant differences in baseline age or sex, selection bias cannot be fully excluded. In addition, because the OSDI was originally developed for adults, its use for symptom assessment in children should be considered exploratory. The relatively small sample size and 12-month follow-up period may limit the generalizability of the results. Seasonal, environmental, and behavioral factors were not fully controlled. Additionally, the study did not assess tear osmolarity or inflammatory cytokines. Recent longitudinal data showed that nightly 0.125% atropine was associated with a modest increase in intraocular pressure relative to controls, without detectable differences in retinal nerve fiber layer or ganglion cell complex thickness [19]. Although these findings cannot be directly extrapolated to 0.05% atropine, they underscore that ocular surface tolerability does not constitute a comprehensive assessment of ocular safety. Future studies should evaluate these characteristics to clarify the mechanisms underlying these clinical findings.

Conclusions

During the 12-month follow-up, once-daily, twice-weekly, and once-weekly 0.05% atropine regimens showed no significant differences in measured ocular surface parameters or OSDI scores in children with myopia. No detectable deterioration was observed in NIBUT, TMH, LLT, IBR, BCHI, or subjective symptoms. However, because no placebo or untreated control group was included, the observed time-related changes in NIBUT and IBR cannot be attributed to atropine treatment. These findings are limited to ocular surface tolerability and subjective symptoms and should not be interpreted as a comprehensive assessment of the overall safety profile of atropine.

Supplementary Information

Below is the link to the electronic supplementary material.

Author Contribution

Xiaoning Li and Zhikuan Yang designed the study; Minsong Xue, QingLin Xu, Haoran Wu, Yao Tang, Zhiwei Luo, WenYu Gao, Yuexiang Deng, and Lijun Hou performed the study; Minsong Xue, QingLin Xu, Haoran Wu, and Yao Tang analyzed data; Minsong Xue, QingLin Xu, Haoran Wu, Yao Tang, Zhiwei Luo, WenYu Gao, Yuexiang Deng, Xiaoning Li, and Zhikuan Yang drafted the manuscript; Xiaoning Li and Zhikuan Yang reviewed the manuscript. All authors read and approved the final version of the manuscript.

Funding

This research was supported by the Natural Science Foundation of Hunan Province (grant no. 2023JJ70036), the Clinic Research Foundation of Aier Eye Hospital Group (grant no. AIM2301D06), and the Key Technologies for Myopia Prevention, Control, Diagnosis and Treatment in Children and Adolescents (grant no. 2019SK2051). The Rapid Service Fee was funded by the funding sources listed above.

Data Availability

The datasets used during the current study are available from the corresponding author upon reasonable request.

Declarations

Conflict of Interest

Minsong Xue, QingLin Xu, Haoran Wu, Yao Tang, Zhiwei Luo, WenYu Gao, Yuexiang Deng, Lijun Hou, Xiaoning Li, and Zhikuan Yang have nothing to disclose.

Ethical Approval

The trial protocol was approved by the Ethics Committee of Changsha Aier Eye Hospital (ID: KYPJ021) and adhered to the tenets of the Declaration of Helsinki.