During the third year, participants who consistently wore or switched to wearing HAL exhibited significantly less choroidal thinning than did children who wore SVL. Wearing HAL for 3 years resulted in choroidal thickening only during the first year, thinning in the second year and stabilization in the third year.

Several studies have demonstrated the efficacy of interventions such as orthokeratology lenses [7, 16], multifocal contact lenses [8, 17], spectacle lenses with myopic defocus signals [9, 13], atropine [18] and repeated low-level red light [19, 20] in slowing myopia progression while also leading to choroidal thickening. However, these studies reported only ChT changes for a maximum duration of 2 years, and longer-term findings are lacking. Here, participants who wore HAL for three years exhibited relatively stable ChT during the third year. Our previous study found changes in ChT for participants who wore HAL for 2 years where there was an initial increase in ChT during the first year followed by a decrease in the second year; the extent of thinning was similar to the initial thickening observed in the first year [13]. We speculated that the peak effect of the myopia control lenses on choroidal thickening occurred within the first year, and afterwards, either the lenses no longer caused choroidal thickening, or that the amount of thickening was less than the age-related thinning. This was due to the retina gradually “adapting” to the myopic control signals provided by the lenses, thereby weakening their role in myopia control and losing the continued thickening effect of the choroid. However, compared with children wearing the SVL, which exhibited significant choroidal thinning, children in the HAL group demonstrated no substantial changes in ChT, which indicated that the subtle thickening effect caused by the lenses counteracted the thinning effect resulting from increases in AL and age. This finding suggested that although the impact of lens-induced increases in ChT may diminish over time, it does not completely disappear.

The effect on choroid and myopia control was less pronounced in the SVL-HAL group in the third year than in the HAL group in the first year [14]. Notably, participants in the SVL-HAL group were older during their third year and had greater myopia and greater AL than participants in the HAL group who were during their first year. This difference implied a steeper retinal shape and potentially weakened the effect of peripheral myopic defocus from wearing HAL. However, we did not find any correlations between baseline age, refractive error, or AL and myopia progression or changes in the ChT in the HAL group. Chamberlain et al. also discovered that age and baseline myopia were found to be independent factors influencing myopia control efficacy among children who wore Misight lenses for 3 years [21]. Therefore, further verification is required regarding assumptions about age, baseline myopia, and their impact on myopia control.

The lenslets of SAL are less aspherical than those of the HAL [10], apparently resulting in a decreased thickening effect and a quicker “retinal adaptation” to myopic control signals. The SAL-HAL group exhibited an increase in the ChT only during the first half of the year when they wore SAL. Upon switching to wearing HAL, the changes in ChT in the third year in the SAL-HAL group were less pronounced than those in the SVL-HAL group but more pronounced than those in the HAL group. We speculated that when the participants switched from SAL to HAL, only the additional myopic defocus signals (the difference in signals between two lenses) had an impact. For the SVL-HAL group, all signals from the HAL had an impact; however, no new signals were induced for the HAL group. Therefore, it can be inferred that the effects of SAL-HAL lie between those of the HAL and SVL-HAL groups. Based on this conjecture, we hypothesized that augmenting defocus signals after a period of use, such as 1 year, may be more beneficial for controlling myopia progression.

When comparing changes in the AL and ChT between the HAL, SAL-HAL and SVL-HAL groups over 3 years, the changes in the AL were significantly smaller in the HAL group than in the SVL-HAL group at all visits (all P < 0.001) [14]. However, the changes in the ChT were no longer significantly different among the three groups at the end of the 3-year follow-up. This may be attributed to the greater variability in ChT changes among participants relative to the AL changes. Therefore, a larger change is required for significant differences to emerge. Additionally, while there was a significant correlation between the AL and ChT changes in both the HAL and SAL-HAL groups, this correlation was not observed in the SVL-HAL group. Therefore, choroidal changes cannot fully explain the myopia control effect of HAL.

There were several limitations in this study. First, we did not strictly enforce a consistent examination time for each participant across all visits, instead we allowed them to schedule between 9 a.m. and 4 p.m.; thus, we could not completely eliminate the potential impact of circadian rhythm on the ChT. In a follow-up study with a large sample size of children, it is difficult to require that participants and their parents adhere to a strict examination schedule for all visits. As this was a randomized controlled study, follow-up dates were determined by the protocol’s follow-up window, while specific times were chosen by the parents. We did not intervene in any participant or group timing choices; therefore, we believe that circadian rhythm had only a minimal effect, if any, on the intergroup comparisons. Second, OCT measurements were taken after cycloplegia, which may have affected the ChT [22]. However, as all groups at all visits experienced cycloplegia during the examinations, we assumed that its effects were negligible when comparing intra- and inter-group changes. Previous studies have shown that thickening and thinning first occur in the medium and large blood vessel layers and change the choroidal vascular index [23,24,25,26]. Specific changes in choroidal tissue may be more meaningful for myopia and myopia control. However, the thickness of different choroidal layers and the choroidal vascular index were not measured as they were outside the scope of the study.