1. Introduction
Myopia is a major and growing global public health problem, particularly in East and Southeast Asia, where prevalence in young adults exceeds 80% and 20–30% may develop high myopia [1,2,3,4]. Childhood myopia progression is primarily driven by excessive axial elongation and is associated with an increased lifetime risk of irreversible complications such as retinal detachment (RD), glaucoma, and myopic macular degeneration (MMD). Effective strategies to slow myopia progression during childhood are therefore essential to reduce long-term visual morbidity.
Current myopia management includes pharmacological treatment with low-dose atropine (LDA) [5,6,7,8,9] and optical interventions designed to induce peripheral myopic defocus, such as highly aspherical lenslet target (HALT) [10,11] and defocus incorporated multiple segments (DIMS) [12,13] spectacle lenses. Both approaches have demonstrated substantial stand-alone efficacy in reducing spherical equivalent progression and axial elongation. However, treatment response is variable and a subset of children continues to progress despite LDA monotherapy, highlighting the need for alternative or adjunctive treatment strategies.
Combination therapy targeting complementary mechanisms has emerged as a rational approach to enhance myopia control. In our previous 12-month study (Sim et al., 2025 [14]), adding HALT lenses to children inadequately controlled on LDA reduced myopia progression from a mean of −0.60 D or 0.24 mm to −0.15 D or 0.14 mm over 12 months, without significant adverse effects. The present study extends these observations to 24 months, evaluating the longer-term efficacy and safety of combination therapy and examining the effect of subsequent atropine dose adjustment, with a comparison to findings from randomized controlled trials and other combination studies.
2. Materials and Methods
A prospective interventional cohort study was conducted at the Singapore National Eye Centre. Patients were enrolled from March 2022 to July 2023. Inclusion criteria were children aged 6–11 years with ≥0.50 D progression over 6 months while on LDA (atropine 0.01% once nightly, atropine 0.01% twice daily or atropine 0.025% once nightly).
Children had to be compliant with atropine treatment at recruitment and agreed to be compliant throughout the study period.
Exclusion criteria included ocular diseases other than myopia, strabismus, amblyopia, or prior contact lens therapy. Informed consent and assent were taken from the parent and patient respectively. The study was conducted in accordance with the tenets of the Declaration of Helsinki, and ethics approval was obtained from the SingHealth Institutional Review Board (IRB).
Participants continued their established LDA regimen and were prescribed Essilor Stellest (HALT) lenses. Cycloplegic refraction and axial length were measured at baseline and at 6, 12, and 24 months. Cycloplegic autorefraction was also measured at least 30 min after administration of 2 drops of cyclopentolate 1% (Alcon, Singapore) given 10 min apart at baseline, using a table-mounted autorefractor (CKR2, Topcon Corporation, Tokyo, Japan).
Five readings, each less than 0.25 D apart, were averaged. AL was measured using the IOLMaster 500 (Carl Zeiss AG, Oberkochen, Germany). Five readings, each less than 0.10 mm apart, were averaged.
Treatment was maintained unchanged for the first 12 months but, subsequently, atropine dose/frequency could be modified based on clinical discretion (i.e., decreased, maintained, or increased). In general, atropine dose/frequency could be tapered/decreased when patients were older with stable myopia and increased when their myopia progressed. Primary outcomes were changes in spherical equivalent (SER) and axial length (AL) over 24 months.
Participants were divided into 3 groups—where atropine treatment was decreased (i.e., tapered or stopped, Group A), continued unchanged (Group B), or increased (Group C). Differences in baseline age, gender, race, baseline SER and AL and changes in SER and AL 6 months prior, in first year, second year and over 2 years were compared between groups. Paired and independent t-tests were used to compare changes between time periods.
Univariate analysis, with the change in SER and AL over 24 months and change between 12 and 24 months (dependent variables), was used to evaluate associations with age, sex, and SER and AL changes, using cluster analysis to account for inter-eye correlation,. Multivariable regression models were constructed to identify independent predictors after adjustments for age, gender, race, and treatment changes. Statistical significance was defined as p < 0.05.
All analyses were conducted using STATA/IC version 11.2 (StataCorp, College Station, TX, USA) and IBM SPSS Statistics for Windows, version 29.0.2.0 (IBM Corp., Armonk, NY, USA).
3. Results
Fifty children were enrolled in the original study. Two children (four eyes) without the full 24-month follow-up were excluded from this study. In the remaining 48 children (96 eyes), baseline age was 9.0 ± 1.1 years, with more children being ethnic Chinese (85%) and female (65%).
Across all eyes, first-year progression (−0.07 ± 0.45 D; 0.13 ± 0.14 mm) was much less than the prior 6 months of atropine alone (−0.61 ± 0.39 D; 0.24 ± 0.12 mm).
In the second year, 23% of eyes had atropine dose either tapered (n = 18 eyes) or stopped (n = 4 eyes) (group A), 67% maintained (group B), and 10% increased (group C). Children in Group A were significantly younger than those in group B (Table 1). Otherwise, baseline SER and AL were very similar between groups.
Six children (12 eyes) switched from HALT lenses to alternative myopia-control lenses in the second year (two from Group A and four from group B). The reasons for this change included increased myopia (n = 3), cost (n = 2) and wear and tear of the lenses (n = 1).
There was no difference in SER and AL progression over the second year between groups, with SER changes of −0.31 ± 0.50 D, −0.24 ± 0.40 D and −0.30 ± 0.32 D (p = 0.495–0.937) and AL changes of 0.13 ± 0.25 mm, 0.17 ± 0.15 mm and 0.21 ± 0.10 mm (p = 0.256–0.541) in Groups A, B and C, respectively (Table 1). There was also no significant difference in the change in SER and AL between groups. Second-year progression (−0.26 ± 0.41 D; 0.17 ± 0.17 mm) was significantly greater than in the first-year for SER (p = 0.003) but not AL (p = 0.133) (Table 1).
However, this progression was still much lower than 6 months prior to the start of combination treatment (Figure 1).
Overall, 71 eyes (73.9%) progressed by ≤0.5 D, while 14 eyes (14.5%) progressed by ≥1.0 D over 2 years. There were 66 eyes (71.7%) that progressed ≤0.4 mm, while 15 (16.3%) progressed by ≥0.6 mm over 2 years.
Univariate analysis suggested that the greater change in SE and AL over 24 months was associated with the greater change in SER and AL over the first year, with the change in AL also associated with younger age (Table A2). These changes were not associated with sex, race, baseline SE or AL, the change in SE or AL 6 months prior, or the treatment group (Table A2). The change in SER in the second year was mainly associated with the change in AL in the 6 m prior to combination and in the first year.
The change in AL over the second year was associated with age and the changes in SER and AL over the first 12 months.
Multivariate models were created, taking care not to include the changes in SER and AL, which were correlated, into the same model. The change in SER and AL over 24 months (dependent variable) was analyzed with models including gender, age, treatment group and the change in AL over the first 12 months (Model 1) and the change in SER over the first 12 months (Model 2). The analysis suggested that the change in SER and AL over 24 months was associated with a greater change in both SER and AL over the first 12 months, with the change in AL over 24 months also being associated with younger age (Table 2).
The change in SER and AL in the second year was analyzed in models including gender, age, the treatment group and the change in AL over the first 12 months (Model 3) and SER over the first 12 months (Model 4). The greater change in SER in the second year was associated with the greater change in AL over the first year while the greater change in AL in the second year was both associated with the greater change in AL and SER over the first year, with age no longer being relevant (Table 2). There were no complaints of glare or near or peripheral blur in children after starting the combination treatment.
4. Discussion
In this prospective cohort, 50 children who were still progressing by ≥0.5 D on low-dose atropine (0.01% or 0.025%) were started on highly aspherical lenslet target (HALT) lenses. In a previous publication, we reported the 12-month follow-up findings which demonstrated that the addition of HALT lenses resulted in a reduced progression from −0.60 D/+0.24 mm in the preceding 6 months to only −0.15 D/+0.14 mm over 12 months. In the second year, clinicians were allowed to alter the treatment according to their clinical judgment.
In this paper, we report the findings in the 48 children who were still under follow-up at 24 months. Treatment remained unchanged in 67%, reduced in 23% and increased in 10%. Compared to the first year of treatment, there was significantly more change in SER in the second year (−0.26 D vs. −0.07 D, p = 0.003) but no significant change in AL (0.17 mm vs.
0.13 mm, p = 0.133). However, the overall change over 24 months was still much less than that noted 6 months prior to when treatment was started. This suggested that despite alterations in the atropine dose, there was a sustained benefit of the combination treatment over 24 months.
In the second year, changes in the atropine dose were made according to clinical judgment, introducing confounding by indication. Given the small group sizes, the absence of statistically significant differences between groups should not be taken as evidence of similar treatment effects.
Children enrolled in this study represent a special clinical group who were still progressing despite treatment with low-dose atropine (0.01% or 0.025%). The mean progression over the preceding 6 months was comparable to, or even greater than, the placebo or single-vision lens groups in many myopia-control trials and substantially greater than that reported for children treated with similar doses in other studies (Table 3). These children may therefore be either rapid progressors in whom the atropine dose was insufficient or children who were destined to be non or poor responders to atropine therapy.
As such, this cohort differs from those enrolled in most atropine or spectacle lens trials, where participants were treatment-naïve and may have consisted of many who would respond well to atropine therapy. Differences in age, baseline myopia, the presence of any previous myopia treatment and study design between studies may also make direct comparisons difficult.
Following the introduction of HALT lenses, first-year AL elongation (+0.13 mm) was lower than that reported in the randomized HALT trial by Bao et al. [10] and comparable to that reported by Sankaridurg et al. [11], while SER progression (−0.07 D) was smaller than both studies (Table 3).
When compared with studies combining atropine with HALT or DIMS lenses, first-year axial elongation in our cohort (0.13 mm) was within the range reported by most studies (0.05–0.16 mm) [23,24,25,26,28,29] and better than several others (0.21–0.28 mm) [22,27]. Refractive progression was also among the lowest reported, being most comparable to Guemes-Villahoz et al. (ASPECT) [22], Zhao et al. [23] and Lee et al. [24] studies (Table 3).
These comparisons should be interpreted cautiously given the differences in age, ethnicity, baseline myopia, atropine concentration or formulation and study design. Overall, however, there is growing evidence that combined optical–pharmacology treatments may be more effective than single treatments. These findings support a practical treatment-escalation strategy, where an optical intervention can be added to atropine, or vice versa, should myopia control be deemed inadequate.
There are few studies reporting the 24-month outcomes of atropine–spectacle lens combination therapy (Table 3). In this study, although progression in the second year was more than in the first year, it was still much lower than that observed before the combination treatment. The change in treatment in some children may have altered the trajectory of myopia progression, but overall, progression over 24 months compared favorably with the published studies of atropine or HALT monotherapy.
This suggests that treatment could be modified according to children’s response over the first year. Interestingly, multivariate analysis suggests that the greater progression of SER and AL over 24 months and in the second year were most strongly associated with the greater change in AL over the first 12 months and to lesser degrees with the change in SER over the first 12 months and younger age (Table 2). This was likely a reflection of how AL change was more rapid in younger children [5,6,15,30], and highlights the importance of monitoring both SER and AL.
4.1. Clinical Implications
This study addresses the clinical problem that arises when children continue to progress while on low doses of atropine. One common cause of a poor response is poor compliance. If compliance is good, however, options include increasing atropine dose or frequency, although this may result in more blur and glare.
Alternatively, as noted in this study, it is also possible to add an optical treatment (such as the HALT lenses). Combination therapy may also be particularly useful in younger children and those with rapid myopia progression.
That said, 14% of children may still progress > 1.0 D and 20% may progress > 0.5 mm over 24 months. Children who are also already progressing more quickly on treatment are likely to be at greater risk. The options are then, besides altering atropine dose, to better manage environmental triggers (e.g., lack of outdoor time) or swap to an alternate optical treatment (spectacles or contact lenses) or new light-based therapies.
4.2. Strengths and Weaknesses
The strengths of this study include its prospective design, high retention rate, standardized measurements, and extended 24-month follow-up. Importantly, it provides real-world data on the effect of treatment modification, an area that has received little attention in the existing myopia-control literature.
Several limitations should be acknowledged. There was no control or monotherapy group for the comparison and the effect measured may be that of HALT alone, rather than an additive effect of combination. Over the second year, the treatment in some children was changed, introducing possible selection bias.
Those children progressing faster in the first year may have been started on a higher dose of atropine, which would blunt the overall progression in the second year. Similarly, rebound when atropine was tapered or stopped may have had the opposite effect. Compliance with atropine and spectacle lens wear was also not objectively monitored and the levels of compliance cannot be confirmed with certainty.
The inclusion of both eyes from each participant in the analysis may also introduce a confounding effect.
5. Conclusions
In this cohort of children who had previously progressed on LDA, myopia progression remained relatively limited over 24 months after combination treatment was introduced. Without a concurrent atropine-only group, the additional effect of HALT and the durability of any treatment effect cannot be determined. Longer follow-ups and larger randomized trials with monotherapy and combination arms are needed to identify which children may benefit from combination treatment, its optimal duration, and how treatment can be safely tapered or stopped.