Efficacy of atropine for myopia control in children: A meta-analysis of randomized controlled trials

ABSTRACT

Objective:

To conduct a meta-analysis of randomized controlled trials (RCTs) on the effects of atropine eye drop in slowing myopia progression.

Methods:

A systematic search of relevant articles was done through a computerized search on PubMed, Medline, Cochrane Library, and Google Scholar on June 16, 2022. A supplementary search was done on ClinicalTrials.gov on the same date. After thorough search and analysis, seven relevant RCTs, double-masked with atropine eye drop as intervention arm and placebo as control arm, were selected for meta-analysis. Jadad scoring was used to evaluate the quality of RCTs. The outcome measurements included in the present meta-analysis are mean changes in the spherical equivalent (SE) of myopic error, and mean changes in axial length (AL) during the study period.

Result:

Pooled summary effect size, calculated by random effect model, for SE of myopia progression was 1.08 with 95% confidence interval (CI) (0.31–1.86) which was statistically significant (P-value = 0.006). Pooled summary effect size, calculated by random effect model, for axial length was − 0.89 with 95% CI (−1.48 to − 0.30) which was statistically significant (P-value = 0.003).

Conclusion:

In summary, atropine was demonstrated to be effective in controlling myopia progression in children. Both outcome measures, mean SE changes and mean AL elongation responded to atropine intervention compared to placebo.

Introduction

Myopia, shortsightedness, has assumed epidemic proportion during the past half-century. A higher myopia prevalence has been reported from East and Southeast Asia compared to Western and European nations. Singapore, China, Hong Kong, and Taiwan are the regions, where the problem is being reported more commonly.[1] The World Health Organization (WHO) enlisted myopia among 10 priority diseases in VISION 2020 campaign.[2] It aims to eliminate avoidable blindness in the world.

A projection estimate predicts the worldwide prevalence of myopia to go up to nearly 5 billion and high myopia up to 1 billion by 2050.[3] Our comprehensive eye care services should prepare to accommodate it. The coronavirus disease 2019 (COVID-19) pandemic has forced the world’s population to lead a sedentary life. Home confinement has led to an unusual amount of time spent indoors. Online classes, increased burden of near work, and long hours of digital screen usage have placed children at a higher risk of myopia development and progression.[4] An axial length (AL) of 26 mm or greater and refractive errors of −6 diopters (D) or greater are reported to have significant association with an increased lifetime risk of visual impairment.[5] A younger age of onset of myopia is a greater risk factor for development of high myopia.[6,7] High myopes are more susceptible to develop macular degeneration, retinal detachment, and glaucoma, and, therefore, subsequent substantial visual loss.[8,9,10] A treatment that would halt or at least decelerate myopia’s progression rate is the need of the hour. Several interventions have been attempted. However, the most ideal approach for efficacious prevention and treatment of myopia needs to be established with sufficient safety and clinical acceptability.

Atropine is a nonselective, muscarinic antagonist. This alkaloid is produced by Atropa belladonna. Its traditional usage in ophthalmic practice is as a mydriatic and cycloplegic drug.[11] In recent years, it has been put to clinical trials as a potential anti-myopia drug.[12,13] The exact mechanism of the suppressive effect of atropine on myopia is still unknown. Fischer et al.[14] concluded that cholinergic signaling in the retina is not required for the atropine-mediated suppression of myopia.

Although involvement of cholinergic amacrines was suggested by earlier reports,[12,13,15,16] Carr et al.[17] hypothesized that muscarinic acetylcholine receptor (mAChR) antagonists could inhibit myopia via a2A-adrenoceptors. Barathi et al.[18] found involvement of gamma-aminobutyric-acid-mediated (GABAergic) signaling. A significant increase in the release of retinal dopamine by atropine instillation was suggested by other studies.[19] It implicates nitric oxide (NO) downstream of dopamine in myopia signaling.[20,21]

However, atropine usage for myopia control remains an off-label treatment. Its approval is still pending by the FDA.

The effects of different concentrations of atropine on myopia development and its long-term effects on visual function in children have been evaluated by several clinical trials. There are still some uncertainties and controversies. Thus, well-conducted evidence-based analysis is needed to evaluate the efficacy of atropine on ameliorating the progression of myopia. In this meta-analysis, with a statistical method of high confidence to achieve scientific result, we aimed to evaluate the overall efficacy of different doses of atropine with more updated RCTs, with only placebo as the control arm.

Material and Methods

Design

The Preferred Reporting Items for a Systematic Review and Meta-Analyses (PRISMA) was used as a guideline.[22]

Inclusion criteria

Comparative studies, that is, RCTs were included according to the following criteria:

• (1) A randomized double-blinded, placebo-controlled clinical trial.

• (2) A human study based on effect of topical atropine on myopia progression in school-aged children (between 4 and 16 years)

• (3) Diagnosis of myopia with spherical equivalent refraction (SER) between −0.50 D and −6 D as measured by cycloplegic autorefraction, and astigmatism ≤2.5 D.

• (4) Concentrations of atropine considered are as follows: 0.01%, 0.025%, 0.05%, 0.5% and 1.0%.

• (5) Atropine as the intervention arm and placebo as the control arm in RCTs

• (6) Measure of outcome in terms of the annual change of SE of myopia progression and in terms of the annual change of AL

• (7) Quality assessment with Jadad score ≥2.[23]

Exclusion criteria

• 1) Duplicate publications and data

• 2) Review articles

• 3) If original data could not be retrieved even after contacting the author

• 4) Experimental study on animals

• 5) If control arm in RCT is not placebo

• 6) If mean spherical error change from baseline to end of study period, indicator of myopia progression and mean AL change measured from baseline to end of study period were not provided in the study

• 7) The studies are not published in English and also its translation is unavailable.

Search strategy

A PRISMA flowchart is depicted for search strategy in Figure 1. Two investigators (SK and RA) independently searched databases PubMed, Medline, Cochrane Library, and Google Scholar on June 16, 2022 with no language restriction. Studies published in other languages were included if their English translation was available. ClinicalTrials.gov was also searched on the same date. The keywords ((“Myopia” OR “Refractive error”) AND “Children” AND “Atropine” AND “Vision” AND “Progression”) were searched. Mendeley reference manager and desktop applications were used for data compilation and duplicate removal. Two independent assessors resolved any discrepancies by consensus. The objective of the study was to look for studies evaluating the efficacy of atropine eye drop in different concentrations on myopia progression. After preliminary screening, 526 related studies were obtained. One hundred fifteen duplicate studies were removed. After excluding 364 unrelated studies, 47 full-text publications were assessed for eligibility. Systematic reviews and meta-analysis were thoroughly searched for relevant RCTs, to meet our objective. After excluding 40 studies for various reasons, 7 studies were eventually included.

Figure 1.

PRISMA flowchart showcasing the search strategy for meta-analysis

Outcome

Efficacy endpoint was myopia progression, defined as mean spherical error change from baseline to end of study period after continuous administration of the study drug. Another efficacy endpoint was mean axial length change measured from baseline to end of study period.

Data extraction

Two investigators (SK and RA) extracted data from the articles in standard file and third independent investigator (ST) validated data extraction. Data collected from each paper are shown in Table 1. They are as follows: the study subject’s characteristics (number of groups and number of participants in each group); the subject’s characteristics (subject type, age, range); experimental treatment (type of treatment, active ingredients, dose, frequency of dose and duration of treatment); and results (mainly quantitative scores of different cognitive tests expressed as mean and SD between baseline and the last follow-up assessment).

Table 1.

Basic characteristics of studies included for meta-analysis

Study	Year	Design	Country/Area	Follow-up (Months)	Included Atropine dose%	Age (Range) in Years	Intervention Group	Control group	Total No. of Patients Completing Study (Atropine/Control group)
Chua et al.[24]	2006	RCT	Singapore	24	1%	6-12	Atropine 1%	Placebo	166/190
Yi et al.[25]	2015	RCT	China	12	1%	7-12	Atropine 1%	Placebo	64/68
Wang et al.[26]	2017	RCT	China	12	0.5%	5-10	Atropine 0.5%	Placebo	54/55
Yam et al.[27]	2018	RCT	Hong Kong	12	0.05% 0.025% 0.01%	4-12	Atropine 0.05% 0.025% 0.01%	Placebo	102 91 97 93
Wei et al.[28]	2020	RCT	China	12	0.01%	6-12	Atropine 0.01%	Placebo	76/83
Saxena et al.[29]	2021	RCT	India	12	0.01%	6-14	Atropine 0.01%	Placebo	47/45
Hieda et al.[30]	2021	RCT	Japan	24	0.01%	6-12	Atropine 0.01%	Placebo	77/81

RCT, randomized controlled trial

If in studies, summary statistics were reported as mean and standard error, the standard error was transformed into SD using formula SD = σ*√n.

Data synthesis and statistical analysis

All analysis were done using RStudio. Meta-analysis to estimate the overall treatment effect of atropine in myopia relative to placebo was performed. Pooled standardized mean difference of atropine (last evaluation at end of follow-up minus baseline data) was computed using fixed effect model and random effects model.

When there was heterogeneity in effect size across all studies, Q-statistics was used to examine this heterogeneity, which follows Chi-square distribution and I²-statistics explains the degree of heterogeneity in effect size across all the studies. Heterogeneity in meta-analysis means effect sizes varies from study to study, therefore identifying this effect size and quantifying heterogeneity is important point to be considered. On the basis of these two measures of heterogeneity (Q and I²), the appropriate model (fixed effect model and random effects model) is chosen to generate pooled effect size. If the degree of heterogeneity in effect size was significantly high (i.e. I² > 30%) random effect model is used; else fixed effect model is used. As Q-statistic and P value will provide only the presence/absence of heterogeneity but not degree of heterogeneity, the magnitude of heterogeneity in effect size across selected studies will be assessed by I² statistics. The I² statistic describes the percentage of variation across studies that are due to real heterogeneity rather than chance alone. The I² statistic calculated along with 95% confidence interval while conducting any meta-analysis to explore the degree of heterogeneity in effect size across various selected studies.

Forest plot was drawn to display the result of individual included studies. Their 95% confidence interval (CI) and pooled effect size with 95% CI are also displayed at the bottom of the graph. Publication bias assessment was done by funnel plot. The graphical method, to check the Publication bias of studies was also constructed.

Qualitative assessment

The assessment of methodological quality of controlled trials was done by Jadad score.[23] Three key methodological features of clinical trial, that is, randomization, double masking, and description of withdrawals and dropouts were used to score the studies.

Other points to down score these trials were used if generating the sequence of randomization and method of double blinding was inappropriately described. One point was added for a “yes” answer to each of the first three items and, also for appropriate description of randomization sequence generation and double blinding. One point is subtracted for a “yes” answer to either of the last two items, for an overall score from 0–5.

Results

Characteristics of the study

A total 526 relevant studies were identified during literature search of atropine for treatment of myopia. Out of 411 records after duplicate removal, 7 RCTs could be included for meta-analysis [Figure 1].

Among the 7 studies, meta-analysis was performed based on mean spherical equivalent (SE) change and mean AL elongation. The 7 studies included in meta-analysis have representation from 1389 subjects, out of which 803 were treated with atropine and 586 with placebo.

For mean SE change, one study, by Yam et al. in 2018,[27] used three doses of drugs; therefore the total number of studies used for meta-analysis was nine. For mean AL elongation, one study, by Wang et al. in 2017,[26] did not report the mean difference of score, and one study, by Yam et al. in 2018,[27] used three doses of drugs; therefore the total number of studies in meta-analysis was eight.

Mean spherical equivalent change of myopia progression

Meta-analysis was performed on nine studies with 1575 number of observations. Mean SE changes are depicted in Table 2. Heterogeneity across nine studies in effect size was statistically significant (Q-value = 143.67 and P value < 0.0001), but degree of heterogeneity was I² = 94.4% (95% CI 91.4%–96.4%). Therefore random effects model was used to calculate standardized mean difference for this indicator. The forest plot of this meta-analysis for mean SE change is shown in Figure 2.

Table 2.

Mean spherical equivalent change in each study

Study	Atropine Group				Control Group				Difference Between Groups
	Concentrations	No. of Patients	Mean of Change (D)	SD	Control	No. of Patients	Mean of Change (D)	SD
Chua et al. 2006[24]	1%	166	−0.28	0.92	Placebo	190	−1.20	0.69	−0.92
Yi et al. 2015[25]	1%	64	0.32	0.22	Placebo	68	−0.85	0.31	−1.17
Wang et al. 2017[26]	0.5%	54	−0.8	1.49	Placebo	55	−2.0	1.51	1.2
Yam et al. 2018[27]	0.05%	102	−0.27	0.61	Placebo	93	−0.81	0.53	0.54
	0.025%	91	−0.46	0.45					0.35
	0.01%	97	−0.59	0.61					0.22
Wei et al. 2020[28]	0.01%	76	−0.49	0.42	Placebo	83	−0.76	0.50	0.26
Saxena et al. 2021[29]	0.01%	47	−0.16	0.40	Placebo	45	−0.35	0.40	0.19
Hieda et al. 2021[30]	0.01%	77	−1.26	0.4	Placebo	81	−1.48	0.41	0.22

D, diopter; SD, standard deviation

Figure 2.

Forest plot showing the result of meta-analysis for indicator mean spherical equivalent of myopia progression

All nine studies showed positive standardized mean difference (SMD). Their 95% CI are as follows: 1.14 (0.92, 1.36); 4.31 (3.68, 4.93); 0.79 (0.40, 1.18); 0.94 (0.64, 1.23); 0.71 (0.41, 1.01); 0.38 (0.10, 0.67); 0.58 (0.26, 0.90); 0.48 (0.07, 0.09); 0.54 (0.22, 0.86). The pooled effect size was 1.08 (95% CI 0.31–1.85) which was statistically significant (P = 0.006). Chua et al.[24] was assigned the highest weight (11.3%) and Yi et al.[25] was assigned lowest weight (10.6%).

Funnel plot [Figure 3] shows two study out of inverted funnel; therefore, publication bias was present.

Figure 3.

Graphic method (Funnel plot) to check the presence of publication bias

The progression of myopia was defined as the change in SER relative to baseline. For this scale, positive value indicated myopia improvement and negative value indicated myopia progression.

Axial length elongation

Meta-analysis was performed on eight studies with 1466 number of observations. Mean AL changes are depicted in Table 3. Heterogeneity across eight studies in effect size was statistically significant (Q-value = 108.85 and P value < 0.0001), but degree of heterogeneity was I² = 93.6% (95% CI 89.6%–96.0%). Therefore random effects model was used to calculate standardized mean difference for this indicator. The forest plot of this meta-analysis for myopia progression is shown in Figure 4.

Table 3.

Mean axial length changes in each study

Study	Atropine Group				Control Group			Difference Between Groups
	No. of Patients	Concentrations	Mean of Change (mm)	SD	No. of Patients	Mean of Change (mm)	SD
Chua et al. 2006[24]	166	1%	−0.02	0.35	190	0.38	0.38	0.40
Yi et al. 2015[25]	64	1%	−0.03	0.07	68	0.32	0.15	0.35
Yam et al. 2018[27]	102	0.05%	0.20	0.25	93	0.41	0.22	−0.21
	91	0.025%	0.29	0.20				−0.12
	97	0.01%	0.36	0.29				−0.05
Wei et al. 2020[28]	76	0.01%	0.32	0.19	83	0.41	0.19	0.09
Saxena et al. 2021[29]	47	0.01%	0.22	0.20	45	0.28	0.28	0.06
Hieda et al. 2021[30]	77	0.01%	0.63	0.17	81	0.77	0.18	−0.14

SD, standard deviation

Figure 4.

Forest plot showing the result of meta-analysis for indicator axial length elongation

All eight studies showing negative SMD with their 95% CI are as follows − 1.09 (−1.31, 0.87); −2.94 (−3.44, −2.45); −0.89 (−1.18, −0.59); −0.57 (−0.86, −0.27); −0.20 (−0.49, 0.08); −0.47 (−0.79, −0.16); −0.25 (−0.66, 0.16); and −0.80 (−1.12, −0.47). The pooled effect size was −0.89 (95% CI −1.48 to −0.30) which was statistically significant (P = 0.003). Chua et al.[24] was assigned the highest weight (12.8%) and Yi et al.[25] was assigned lowest weight (11.9%).

Funnel plot [Figure 5] shows three studies out of inverted funnel; therefore, publication bias was present.

Figure 5.

Graphic method (Funnel plot) to check the presence of publication bias

Changes in AL was defined as mean baseline value subtracted by mean end point value. For this scale, negative value indicated myopia improvement and positive value indicated myopia progression.

Jadad score

Jadad score was calculated using seven studies included after systematic review. Five columns were used for scoring purpose independently. The five columns (domains) on which scoring was done have headings as follows: Described as Randomization, Method of Randomization, Described as Double Blinded, Allocation Concealment and Description of Withdrawals and Dropouts.[23]

A total of six studies (Chua et al.,[24] Wang et al.,[26] Yam et al.,[27] Wei et al.,[28] Saxena et al.,[29] and Hieda et al.[30]) received highest score (5), whereas one study (Yi et al.[25]) received the lowest score (4). Table 4 displays the result of Jadad score in detail.

Table 4.

Jadad scoring of studies included after systematic review

Study	Described as Randomized	Method of Randomization	Described as Double Blinded	Allocation Concealment	Description of Withdrawals and Dropouts	Jada Scoring
Chua et al. 2006[24]	Yes	Adequate	Yes	Appropriate	Adequate	5
Yi et al. 2015[25]	Yes	Described but not appropriately	No	Not appropriate	Adequate	2
Wang et al. 2017[26]	Yes	Adequate	Yes	Appropriate	Adequate	5
Yam et al. 2018[27]	Yes	Adequate	Yes	Appropriate	Adequate	5
Wei et al. 2020[28]	Yes	Adequate	Yes	Appropriate	Adequate	5
Saxena et al. 2021[29]	Yes	Adequate	Yes	Appropriate	Adequate	5
Hieda et al. 2021[30]	Yes	Adequate	Yes	Appropriate	Adequate	5

Discussion

Key findings

Our meta-analysis performed on SE indicator for myopia progression shows improvement of myopia in atropine-treated group. This result was statistically significant. The funnel plot constructed was showing presence of publication bias in our current meta-analysis.

Our meta-analysis performed on AL elongation as indicator shows improvement in atropine-treated group. This result was statistically significant. The funnel plot constructed was showing presence of publication bias in our current meta-analysis.

Uniqueness of our meta-analysis

To the best of our knowledge, this is the first meta-analysis of RCTs evaluating atropine efficacy on myopia control with only placebo as control arm.

Comparative outcomes of different studies

The first trial by Chua et al.,[24] Atropine for the treatment of Myopia 1 (ATOM1) study, demonstrated that the progression of myopia is retarded by approximately 76% over the 2-year treatment period by use of 1% atropine eye drop.

However, the follow-up study by Tong L et al.,[31] showed a strong rebound effect, a 300% increase in myopia progression rate compared to placebo during the first 12 months after the cessation of atropine. It eliminated the benefits by approximately 60% gained during the 2-year treatment period. The side effects, photophobia, reduced accommodation amplitude, and blurred vision are caused by use of 1% atropine.

The follow-up trial, ATOM2, demonstrated the effects of 0.5%, 0.1%, and 0.01% atropine on the progression of myopia in children. It showed that the progression of myopia is retarded by 75% in 0.5% atropine-treated group, while progression is reduced by 68% and 59% in 0.1% and 0.01% atropine-treated group, respectively.[32]

A 218% rebound increase in the progression rate compared to placebo in 0.5%-atropine-treated group was reported after cessation of treatment. A 170% increase in the progression rate of myopia in 0.1%-atropine-treated group during the first 12 months after cessation of the drug was reported,[33] whereas there was an approximate drop of 30% in progression rate of myopia in 0.01%-atropine-treated group was noted. Therefore, it was suggested that 0.01% concentration is a good option for myopia control.[33]

The recent 5-year follow-up study demonstrated that the cessation of treatment with higher initial atropine dose leads to greater myopia progression. Therefore, treatment with 0.01% atropine is advocated for the best long-term outcome.[34]

Low-concentration atropine for myopia control (LAMP) study by Yam et al.[27] suggested a dose-dependent suppressive effect of low-dose atropine on myopia progression. This study showed that progression of myopia is retarded by 27% over a 1-year period in 0.01% atropine group, while 43% and 67% retardation are reported in 0.025%- and0.05%-atropine-treated group, respectively. All these concentrations were well tolerated by children, causing minimal side effects. Therefore, low-dose atropine to control myopia progression in children is suggested.

Ongoing clinical trials

Ongoing clinical trial, CHAMP-UK-study, “Low-dose (0.01%) atropine eye-drops to reduce progression of myopia in children: a multicenter placebo-controlled randomized trial in the UK” will be the first randomized trial in UK population reporting outcomes of low-dose atropine eye-drops in myopic children.[35]

The other ongoing clinical trial, MOSAIC study, “Myopia Outcome Study of Atropine in Children (MOSAIC): an investigator-led, double-masked, placebo-controlled, randomized clinical trial protocol”, is the first RCT to explore the efficacy, mechanisms of action, and safety of unpreserved 0.01% atropine in a predominantly White population.[36]

Another large scale ongoing clinical trial, (WA-ATOM) study, “Western Australia Atropine for the Treatment of Myopia (WA-ATOM) study: Rationale, methodology and participant baseline characteristics”, will explore the efficacy, safety, tolerability and long-term effects of low-dose atropine eyedrops in myopia control in Australian children, including those of non-Asian background.[37]

Once the results of these trials[35,36,37] are published, the efficacy of atropine in controlling myopia progression in different ethnic groups worldwide would be clear.

Critical assessment of included RCTs

Chua et al.[24] included low and moderate myopia group, and only uniocular treatment with 1% atropine in the study. Magnitude of efficacy (0.79 D) of 1% atropine seen in 1 year was slightly larger than previously published studies. Again, long-term uniocular treatment with atropine could lead to anisometropia and aniseikonia, whereas bilateral treatment with atropine would have introduced a confounding factor in the form of bifocal or progressive additional glasses, needed to correct blurred near vision.

Yi et al.[25] studied effect of 1% atropine on early myopia. Though myopia improved in most cases with atropine intervention, there was no response in few cases. Heterogenous drug response, environmental factors, disease severity, lifestyle, and genetics are some factors responsible for variable response to treatment.

Wang et al.[26] included participants only from Han ethnicity with low myopia. Therefore, results from this trial may not be generalized to other Chinese ethnicities.

Yam et al.[27] proved concentration-dependent response with low concentrations of atropine. Concentration-dependent response had been earlier reported with higher concentrations of atropine: 0.5%, 0.25%, and 0.1%,[38] whereas, ATOM 2 study[32] reported a small difference in efficacy of 0.01% atropine compared with 0.1% and 0.5%. It raises a question on the existence of concentration-dependent response. Yam et al. proved a clear concentration-dependent response, with 0.05% atropine better than 0.025% and 0.01%. It would open up the possibility of increasing the frequency of atropine eye drop in controlling myopia progression.

Wei et al.[28] reported a loss to follow-up of approximately 30%. They concluded that younger age at baseline was associated with the risk of progressive myopia based on an unadjusted regression analysis. It was noted that relative reduction of 34.2% in myopia progression in their study was smaller than the study conducted in the Europe.[39] It was hypothesized that less pigmented eyes, on average, show more sensitivity to cycloplegic agents.

Saxena et al.[29] reported favorable effects of 0.01% atropine on controlling myopic SE progression while there was minimal effect on AL elongation.

Hieda et al.[30] reported moderate reduction in myopia progression with 0.01% atropine. It was emphasized that it takes as long as 2 years to achieve beneficial effects of therapy. It was suggested that environment, country, and lifestyle could influence the final outcome. A strong interaction was observed on treatment-by-subgroup, that is, uncorrected near visual acuity (VA) and mesopic pupil diameter in SE, and the enrolled AL and mesopic pupil diameter in AL. This opens up a possibility of atropine therapy on special target population of myopic children. They also observed an increased photopic pupil size in the early phase and no change in the mesopic pupil size in the atropine group. This raises a question whether pupil dilation is one of adverse effects or not and whether it is involved in inhibition of myopia progression.

Side effects of atropine

Ocular use of atropine is not commonly associated with systemic side effects; for example, dry mouth, headache, face flush, constipation, difficulty in micturition, increased blood pressure, and disturbances related to the central nervous system.[40] Photophobia, blurriness of near vision, impaired amplitude of accommodation, and local allergic reactions are the commonly seen ocular side effects. Photophobia is the most common ocular side effect. Yen et al.[41] reported photophobia in 100% of subjects and also the major reason for more than 50% dropout after receiving 1% atropine in their study. Shih et al.[38] reported photophobia in only 22% of subjects in 0.5%-atropine-treated group while it was reported by 7% of subjects in 0.25%-atropine-treated group. No significant photophobia was reported in the 0.1%-atropine-treated group.

ATOM 2 study reported allergic conjunctivitis in 4.1% of subjects in the atropine 0.1% and 0.5% groups. Negligible effect on pupil size and accommodation, and no effect on near vision, and uncommonly reported photophobia were seen in children who were treated with 0.01% atropine. Only 7% of children in this group opted for photochromatic lenses.[32]

These data indicate that incidence of photophobia is positively correlated with the concentration of atropine.

1% or 0.5% of atropine concentration have been reported to be very effective in retarding myopia progression. But the high side effects of photophobia (in up to 100%), and higher dropout rate (16%–58%) are associated with these high concentrations of atropine.[13,41] Potential long-term systemic or ocular side effects and rebound effect after atropine cessation, particularly with higher concentration of atropine are of concern.

Several recent publications from Asia have reported efficacy of 0.01% atropine in controlling myopia progression but with lower rates of side effects.[40]

Side effects of atropine in RCTs included in our meta-analysis are as follows:

Chua et al.[24] reported no serious adverse effects with 1% atropine. Allergic or hypersensitivity reactions or discomfort in 4.5% of subjects, glare in 1.5% of subjects, and blurred near vision in 1% of subjects were the reason behind withdrawal of treatment. There was no deterioration in best-corrected visual acuity.

Yi et al.[25] reported photophobia due to mydriasis as main adverse effect in 1%-atropine-treated group. Anti-UV sunglasses were prescribed particularly during outdoor activities. Only 7 cases of near blurring required near add glasses.

Wang et al.[26] reported no serious adverse effects in 0.5% atropine group in their study.

Yam et al.[27] reported that the accommodation loss was comparable in 0.01% atropine group and in placebo group. Vision and quality of life of children on 0.05%, 0.025%, and 0.01% atropine treatment were comparable to those in placebo group. There were no serious adverse effects related to atropine.

Wei et al.[28] reported 4.5% of subjects with photophobia in the 0.01% atropine group. 0.9% of subjects in the control group reported the same. Four children complained of allergic conjunctivitis; 1 of them was in the control group. Near-blurred vision was reported by none. No serious adverse effects were seen in the atropine group.

Saxena et al.[29] conducted I-ATOM study in India. They reported stable mesopic and photopic pupil size, accommodation amplitude, distance, and near vision on follow-ups in 0.01% atropine group. No significant change was noted between atropine and placebo groups. None of the patients reported photophobia or blurring of vision. Therapy was not discontinued in any subject for atropine-related side effects.

Hieda et al.[30] reported 1 case with suspected hemiplegic alteration migraine due to photophobia and one more case with light sensitivity in 0.01% atropine group. They had to discontinue due to adverse events.

Strengths of our study

Literature search strategy was rigorous. Research question was supported by clear eligibility criteria. Each step-in review was done by multiple reviewers to ensure accuracy. Preferred Reporting Items of Systematic Review and Meta Analysis during preparation of manuscript is followed. Meta-analysis was conducted adhering to guidelines of Cochrane Handbook of Systematic Review and Meta-analysis.

Limitations of our study

Studies selected for meta-analysis in our study are less in number. Meta-regression was not suitable in current study as the number of studies should be minimum 10 to get robust result by this technique. Instead of comparing single concentration of atropine eye drops with placebo, varying concentrations of atropine have been considered in RCTs included. Treatment period in included RCTs are of shorter duration, that is, 12 months and 24 months, while the period of myopia progression and ocular axial growth extend beyond 2 years in children commonly.

Conclusion

In conclusion, the present meta-analysis, which was performed with 7 RCTs suggested that administration of atropine for a duration of 12 months (5 studies) or 24 months (2 studies) was effective to control myopia progression in children. Both outcome measures, mean spherical equivalent changes and mean axial length changes responded to atropine intervention compared to placebo.

Future remarks

Future clinical trials and meta-analysis should focus on uncertainty regarding initiation and cessation of therapy, duration and optimum concentration of drug-related protocols in different ethnic groups, and in different rates of myopia progression.

Atropine therapy in non-responders or poor responders, and its effectiveness in late onset myopia also needs to be addressed.

List of Abbreviations

Abbreviation	Definition
AL	Axial length
CI	Confidence interval
D	Diopter
FDA	Food and Drug Administration
mAChR	Muscarinic acetylcholine receptor
RCTs	Randomized controlled trials
SD	Standard deviation
SE	Spherical equivalent

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.