Eleven years of orthokeratology contact lens wear for slowing myopia progression in children

Abstract

SIGNIFICANCE:

The myopia control effect of orthokeratology accrues over time, with 11 years of lens wear providing a cumulative absolute reduction in axial elongation of −0.69 mm in comparison with spectacle lens wear. Steeper corneas are likely to benefit from enhanced myopia control efficacy.

PURPOSE:

To compare axial length growth between a group of orthokeratology contact lens wearers and a control group of distance single-vision lens wearers over an 11-year period.

METHODS:

White European subjects 6 to 12 years old with myopia −0.75 to −4.00DS and astigmatism ≤1.00DC were prospectively allocated orthokeratology or distance single-vision spectacle correction for 2 years. Axial length measurements (Zeiss, IOLMaster) were taken at 6-month intervals during the initial 2 years of the study. Subjects were contacted approximately 5 and 9 years later (i.e., 7 and 11 years after the beginning of the study, respectively) and axial length measurements were repeated.

RESULTS:

Thirty-one orthokeratology and 30 control subjects were initially recruited, but only 10 orthokeratology and 15 control subjects attended the 11-year visit. In comparison with the control group, the change in axial length for the orthokeratology group was reduced by 0.04, 0.10, 0.14, 0.22, 0.45, and 0.69 mm after 0.5, 1, 1.5, 2, 7, and 11 years of lens wear, respectively. Significant differences between groups were found in mean unadjusted changes in axial length at the 1-, 1.5-, and 2-year time points (unpaired t-test, p < 0.05), whereas standard contrasts revealed statistical differences between groups in the estimated marginal means of the change in axial length at the 7- and 11-year time points (p < 0.05).

CONCLUSIONS:

Eleven years of orthokeratology lens wear provided a substantial slowing in the axial elongation of the eye, with a treatment effect of up to 0.69 mm after 11 years of lens wear in comparison with single-vision lens wear.

Myopia is widely recognized as a significant risk factor for a range of sight-threatening ocular complications.¹ As such, it is particularly worrying that the prevalence of myopia has increased substantially in recent decades, affecting approximately 30% of the world’s population, with estimates indicating that it will continue to increase to affect around 50% of the world’s population by 2050.² Furthermore, myopia incurs substantial healthcare expenditure³ and loss of productivity.⁴ The growing prevalence of myopia together with its healthcare and economic implications has resulted in increased interest in methods to slow its progression, leading to several therapies, including topical atropine, specially designed spectacles, dual-focus contact lenses, multifocal soft contact lenses, and overnight orthokeratology, that have demonstrated clinically significant effects in slowing myopia progression.⁵ Among these options, there is a substantial body of scientific evidence supporting the effectiveness of orthokeratology lens wear in slowing the axial elongation of the eye, with systematic reviews and meta-analyses reporting that orthokeratology lens wear reduces axial elongation in pediatric patients by approximately 0.28 mm compared with control groups of single-vision lens wearers over a 2-year period.^5,6 However, most evidence supporting its effectiveness is limited to studies in which lenses were worn for up to 2 years only, with only a few studies reporting on its efficacy in slowing myopia progression over longer treatment periods.^7–11 Although axial elongation is the preferred metric to assess the efficacy of myopia control interventions,¹² most studies evaluating orthokeratology lens wear over periods longer than 2 years relied on retrospective designs and estimated changes in refractive corneal power. As such, the findings from these studies provide limited insight into the long-term efficacy of orthokeratology in slowing myopia progression.^7,9–11 One prospective study, however, evaluated changes in axial length in orthokeratology versus single-vision spectacle lens wearers over a 5-year period.⁸ The latter study found orthokeratology lenses to significantly slow the axial elongation of the eye in comparison with spectacle lens wear during the first 3 years of treatment, but not during the 4th and 5th years of lens wear. Since orthokeratology contact lenses are frequently used to slow myopia progression, they may be worn over many years—potentially throughout childhood and adolescence, when the risk of myopia progression is highest,^13,14 and possibly into adulthood for refractive correction. Thus, the purpose of this study, as the primary outcome measure, is to assess differences in axial length growth between orthokeratology and control single-vision lens wearers over an 11-year period.

METHODS

This study was part of a larger, nonrandomized, clinical trial designed to assess different aspects of orthokeratology lens wear specifically prescribed for the control of myopia progression in children, with the methods of special interest to this study being described in detail elsewhere.^15,16 In brief, White European subjects 6 to 12 years old with myopia −0.75 to −4.00DS and astigmatism ≤1.00DC were prospectively allocated orthokeratology (i.e., tisilfocon A, Menicon Z Night, Menicon Co., Ltd, Nagoya, Japan) or single-vision spectacle correction for 2 years; subjects were allowed to freely choose one or the other mode of visual correction at baseline. Axial length measurements (Zeiss, IOL Master 500, Germany) were taken at 6-month intervals during the initial 2 years of the study (ClinicalTrials.gov number: NTC04806763). Spectacles or contact lenses, and contact lens solutions, full ocular examinations, contact lens fittings, and aftercares were provided to all subjects free of charge throughout the initial 2 years of the study. Approximately 5 and 9 years after completion of the study (i.e., approximately 7 and 11 years since the beginning of the study), subjects were contacted by phone and invited to return to the clinic for a free ocular examination in which ocular refractive and biometric parameters were collected, including axial length measurements (ClinicalTrials.gov number: NTC04806711).¹⁶ At all follow-up visits, three separate measurements of axial length were recorded, and a mean was obtained. The study was conducted in accordance with the Tenets of the Declaration of Helsinki and approved by the Institutional Ethical Committee Review Board of Novovision Ophthalmology Clinic (Madrid, Spain). Full informed consent and child assent were obtained in writing from the parents/guardians before the start of all experimental work and data collection. Patient participation in the study could be discontinued at the examiner’s discretion should significant symptoms or slit-lamp findings occur. Subjects were instructed they could withdraw from the study at any time.

Statistical analysis

Subjects’ demographics and baseline data were first checked for normality, and differences between groups were subsequently assessed using unpaired sample t-tests for all variables, except for the male:female ratio, which was tested using a chi-square test. Changes in axial length (relative to baseline) were also first checked for normality and compared between treatment groups (i.e., orthokeratology vs. control) at each of the different time points (i.e., baseline and 0.5, 1, 1.5, 2, 7, and 11 years) using an unpaired t-test. Subsequently, changes in axial length over the 11-year period were further compared between groups using a “modified” or “per-protocol” intent-to-treat approach whereby data from all subjects, regardless of whether some subjects who not attended all study visits or were lost to follow-up at different time points, to prevent data from subjects being removed from the analysis whenever one or more data points were missing (i.e., listwise deletion). Subjects who switched lens wear modalities were excluded from the analysis, but subjects who switched from distance, single-vision spectacles to distance, single-vision soft contact lenses after the initial 2 years of the study were retained as part of the control group. The latter analysis was performed using linear mixed models in which the change in axial length was selected as the dependent variable, group and time as fixed effects, and subjects’ identification number as the random effect; this allows for testing differences between groups, over time and their interaction. Estimated marginal means were calculated for the changes in axial length for the two study groups at each of the different time points, and standard contrasts were used to assess differences between groups at each time point; the latter included adjustments for multiple comparisons using the Holm method. The effects of baseline demographic variables, including age, gender, mean spherical refractive error, axial length, mean central corneal power, corneal shape (i.e., p value), and pupil diameter, on the changes in axial length were successively tested in the model by including them as fixed factors to assess their interactions with group and time. For those baseline demographic variables showing significant interaction, simple linear regressions against the 11-year change in axial length were calculated for both groups separately. The strength of association is summarized using linear regression equations, R² squared values, and p values. Data are expressed as mean ± 1 standard error of the mean unless otherwise stated. Data from the right eye only were used for analysis. Statistical analyses were performed using JASP software (version 0.14.1, University of Amsterdam, Amsterdam, The Netherlands). The level of statistical significance was set at 5%.

RESULTS

Thirty-one orthokeratology and 30 control subjects were initially recruited; 29 orthokeratology and 24 control completed the initial 2 years of the study; 14 orthokeratology and 16 control subjects attended the 7-year visit; 10 orthokeratology and 15 control subjects attended the 11-year visit; but only 10 orthokeratology and 10 control subjects attended all study visits. The 10 orthokeratology and 15 control subjects that attended the final visit of this study were found to have completed 11.27 ± 0.05 and 11.00 ± 0.13 years of follow-up, respectively (Fig. 1); the latter difference was not statistically significant (p > 0.05).

FIGURE 1.

Flowchart of the subjects followed up in the study. SCL = single-vision soft contact lenses; SV = distance single-vision spectacles.

No statistically significant differences were found in any of the baseline demographics, including age, gender, axial length, mean spherical refractive error, mean central corneal curvature, corneal p value, or pupil diameter, between those initially recruited versus those who completed 11 years of follow-up within both the orthokeratology and control groups (all p > 0.05) (Table 1). However, significant differences were observed between retained and dropout orthokeratology subjects at 11 years in baseline age and the male/female ratio (both p < 0.05), but no significant differences were found in baseline demographics between retained and dropout control subjects at 11 years (all p > 0.05).

TABLE 1.

Baseline demographics for the subjects initially recruited and for those who did (retained) and did not (dropout) complete 11 years of follow-up

Study visit	Orthokeratology			Control
	Baseline (N = 31)	11 years		Baseline (N = 30)	11 years
		Retained (N = 10)	Dropout (N = 21)		Retained (N = 15)	Dropout (N = 15)
Age (y)	9.67 ± 0.30	10.50 ± 0.60	9.25 ± 0.30	9.84 ± 0.32	9.80 ± 0.44	9.90 ± 0.48
Male/female ratio	15/16	8/2	7/14	15/15	8/7	7/8
MSE (D)	−2.28 ± 0.21	−2.25 ± 0.31	−2.29 ± 0.27	−2.44 ± 0.25	−2.23 ± 0.29	−2.74 ± 0.45
Axial length (mm)	24.46 ± 0.14	24.53 ± 0.24	24.42 ± 0.18	24.20 ± 0.19	24.17 ± 0.21	24.25 ± 0.35
Mean K (D)	43.27 ± 0.28	43.17 ± 0.42	43.32 ± 0.37	43.77 ± 0.33	43.71 ± 0.47	43.85 ± 0.45
Corneal p value	0.69 ± 0.02	0.71 ± 0.03	0.67 ± 0.02	0.72 ± 0.02	0.72 ± 0.02	0.72 ± 0.02
Pupil diameter (mm)	4.03 ± 0.09	3.90 ± 0.14	4.11 ± 0.11	3.61 ± 0.11	3.46 ± 0.11	3.87 ± 0.19

Variables are expressed as mean ± 1 standard error of the mean.

D = diopters; K = keratometry; MSE = mean spherical refractive error; N = number of subjects.

The orthokeratology group exhibited significantly less axial length growth than the control group at 1-, 1.5-, and 2-year time points, and almost reached statistically significant differences at 7- and 11-year visits (Fig. 2 and Table 2, unpaired t-tests).

FIGURE 2.

Mean unadjusted changes in axial length (mm) from baseline for the orthokeratology (green, solid circles and line) and control (red, solid circles and line) groups. Error bars represent one standard error of the mean.

TABLE 2.

Mean unadjusted changes in axial length (from baseline) for the orthokeratology and control groups at each of the different time points

Visit	Intervention	No. subjects	Axial length change (mm ± SEM)	Difference (mm)	95% confidence Interval	Statistical difference (p value)
0.5 y	Orthokeratology	30	0.12 ± 0.02	0.04	−0.02 to 0.10	p=0.15
	Control	23	0.17 ± 0.02
1 y	Orthokeratology	30	0.22 ± 0.02	0.10	0.03 to 0.17	p=0.005
	Control	23	0.32 ± 0.03
1.5 y	Orthokeratology	28	0.37 ± 0.02	0.14	0.03 to 0.26	p=0.016
	Control	25	0.51 ± 0.05
2 y	Orthokeratology	29	0.47 ± 0.03	0.22	0.08 to 0.37	p=0.003
	Control	24	0.70 ± 0.07
7 y	Orthokeratology	14	0.91 ± 0.17	0.45	0.02 to 0.92	p=0.062
	Control	16	1.36 ± 0.16
11 y	Orthokeratology	10	1.11 ± 0.31	0.69	0.06 to 1.44	p=0.069
	Control	15	1.80 ± 0.21

SEM = standard error of the mean.

Using the linear mixed model, statistically significant differences in axial length changes (from baseline) were found over time, between groups and for the time × group interaction (all p ≤ 0.001). The estimated marginal means of the change in axial length for the two study groups are reported in Table 3. Standard contrasts revealed statistical differences between groups solely at the 7- and 11-year time points. Statistically significant group and group × time interactions were found for age, axial length, and mean central corneal curvature. Younger age at baseline was associated with larger increases in axial length at 11 years in the control group (R² = 0.38, p = 0.02), but no such significant relationship was found for the orthokeratology group (R² = 0.39, p = 0.07) (Fig. 3). A trend for shorter baseline axial length to be associated with larger increases in axial length at 11 years was found in the control (R² = 0.22, p = 0.08), but not in the orthokeratology group (R² = 0.01, p = 0.83) (Fig. 4). Greater mean corneal power (i.e., steeper cornea) was associated with smaller increases in axial length at 11 years in the orthokeratology group (R² = 0.63, p = 0.01), but no significant relationship was found in the control group (R² = 0.01, p = 0.81) (Fig. 5).

TABLE 3.

Estimated marginal means of the change in axial length (from baseline) for the orthokeratology and control groups at each of the different time points

Visit	Intervention	Axial length change (mm ± SEM)	Difference (mm)	95% confidence Interval	Statistical difference (p value)
0.5 y	Orthokeratology	0.12 ± 0.06	0.04	−0.15 to 0.22	0.72
	Control	0.16 ± 0.07
1 y	Orthokeratology	0.22 ± 0.06	0.10	−0.08 to 0.28	0.57
	Control	0.32 ± 0.07
1.5 y	Orthokeratology	0.37 ± 0.06	0.14	−0.04 to 0.33	0.38
	Control	0.51 ± 0.07
2 y	Orthokeratology	0.48 ± 0.06	0.21	0.03 to 0.40	0.086
	Control	0.69 ± 0.07
7 y	Orthokeratology	0.92 ± 0.09	0.41	0.18 to 0.64	0.003
	Control	1.33 ± 0.08
11 y	Orthokeratology	1.13 ± 0.10	0.64	0.39 to 0.90	<0.001
	Control	1.77 ± 0.08

SEM = standard error of the mean.

FIGURE 3.

Simple linear regressions between the change in axial length at 11 years relative to baseline and baseline age for the orthokeratology (green, solid circles and line) and control groups (red, solid circles and line).

FIGURE 4.

Simple linear regressions between the change in axial length at 11 years relative to baseline and baseline axial length for the orthokeratology (green, solid circles and line) and control groups (red, solid circles and line).

FIGURE 5.

Simple linear regressions between the change in axial length at 11 years relative to baseline and baseline mean central corneal power for the orthokeratology (green, solid circles and line) and control groups (red, solid circles and line).

DISCUSSION

This study assessed the efficacy of orthokeratology lens wear in slowing the axial elongation of the eye in comparison with a control group of single-vision lens wearers over an 11-year period in White European myopic subjects, thus providing evidence with regards to the longest, reported period of orthokeratology lens wear in slowing the axial elongation of the eye. Significant differences between groups were found in mean unadjusted changes in axial length at the 1-, 1.5-, and 2-year time points (Table 2), whereas standard contrasts revealed statistically significant differences between groups in the estimated marginal means of the change in axial length at the 7- and 11-year time points (Table 3) relates to differences in the two statistical approaches employed, with the former assessing differences between groups at different time points individually, and the latter adjusting for multiple comparisons across time points. Overall, these results indicate that while a large proportion of the total 11-year treatment effect in slowing myopia progression (i.e., 0.69 mm) were obtained during the 1st (i.e., 0.10 mm) and 2nd (i.e., 0.22 mm) years of orthokeratology treatment, the myopia control effect continues to accrue over time. This cumulative effect leads to progressive reductions in axial elongation with longer periods of orthokeratology lens wear. This initial burst in myopia control efficacy over the first 2 years of lens wear appears to be a common feature among different myopia control interventions,¹² including orthokeratology.¹⁷

Statistically significant group and group × time interactions were found for age, axial length, and mean central corneal curvature. Younger age at baseline was associated with larger increases in axial length in the control but not in the orthokeratology group (Fig. 3). More specifically, inspection of Fig. 3 reveals that orthokeratology lens wear slows the axial elongation of the eye by about 0.5 mm over 11 years of lens wear in comparison with control children of the same age, indicating that the efficacy of orthokeratology lens wear in slowing myopia progression appears to be independent of age, which is in agreement with previous studies.^12,17 Greater mean corneal power (i.e., steeper cornea) was associated with smaller increases in axial length at 11 years in the orthokeratology group, but not in the control group (Fig. 4), which might indicate that a steeper cornea allows enhanced corneal reshaping as steeper corneas tend to respond better to the flattening effect of orthokeratology lenses, potentially leading to more effective increases in spherical aberration and peripheral myopization as potential mechanisms responsible for the myopia control effect behind orthokeratology lens wear.¹⁸ The latter result implies that, despite all orthokeratology lens-wearing subjects achieved successfully refractive correction and optimal levels of visual acuity, individuals with flatter corneas are likely to benefit less from orthokeratology lens wear for slowing myopia progression compared with individuals with steeper corneas.

The 2-year efficacy in slowing the axial elongation of the eye appears to be relatively similar between different myopia control interventions,^5,12 but of interest is understanding the long-term effectiveness of these interventions. To contextualize the findings of this study, comparisons were made with other long-term investigations into the efficacy of myopia control interventions (Table 4). It is important to note that comparisons across studies are inherently challenging because many long-term studies lack conventional control groups. In most clinical trials, control subjects are often invited to switch to the intervention after the trial period, limiting the ability to assess cumulative treatment effects over time. Despite these challenges, the results of the present study align well with the 5-year Japanese study by Hiraoka et al., which assessed axial elongation in children using orthokeratology compared with single-vision spectacle lenses. The observed similarities are particularly notable given the greater axial elongation typically reported in Asian children compared with non-Asian children of the same age.⁸ Furthermore, to address the absence of long-term control group data, the estimated axial elongation for control groups in various studies was calculated using the model proposed by Brennan et al.²² (Table 4). More specifically, the model was used to estimate the axial elongation of control groups from other long-term studies that lacked such data. This was done using two approaches: (1) based on baseline age data from the control group or (2) using axial elongation data from the 1st year from the control group to extrapolate the estimated elongation for a given year, applying a 15% annual slowing rate. While this model provides valuable insights, it is worth noting that estimates based on baseline age tend to underestimate axial elongation, those derived from 1st-year axial elongation data align more closely with the values reported in Japanese children over 5 years and in White European children over 7 and 11 years in this study.

TABLE 4.

Comparison between studies reporting on the long-term efficacy of different interventions for myopia control.

Study (country)	Duration (y)	Age range (y)	No. subjects at baseline: test/control	No. subjects completed test/control (% of baseline)	Study design	Test lens: cumulative axial elongation (mm)	Control lens: cumulative axial elongation (mm)
Hiraoka et al.8 (Japan)	5	8–12	29/30	22 (76)/21 (70)	Non-RCT	Orthokeratology: 0.99 ± 0.47	Spectacles: 1.41 ± 0.68* 1.31† or 1.41‡
Chamberlain et al.19 (Canada, Portugal, Singapore, and UK)	6	8–12	52/56	45 (86)/NA	RCT	Dual focus: 0.49 ± 0.39	Soft lenses: 1.27† or 1.00‡
Lam et al.20 (Hong Kong)	6	8–13	93/90	36 (34)/NA	RCT	DIMS 0.60 ± 0.49	Spectacles: 1.47† or 1.33‡
Zhang et al.21 (China)	5	4–12	109:108: 110/111	33 (30):27 (25): 39 (35)/NA	RCT	Atropine 0.05%: 0.79 ± 0.54 Atropine 0.025%: 1.11 ± 0.46 Atropine 0.01%: 1.24 ± 0.72	Placebo: 1.05† or 1.52‡
This study (Spain)	7	8–12	31/30	14 (45)/16 (53)	Non-RCT	Orthokeratology: 0.91 ± 0.17	Spectacles: 1.36 ± 0.16* 1.19† or 1.45‡
This study (Spain)	11	8–12	31/30	10 (32)/15 (50)	Non-RCT	Orthokeratology: 1.11 ± 0.31	Spectacles: 1.80 ± 0.21* 1.47† or 1.78‡

Reported/calculated from the study.

^*Estimated using a model reported by Brennan et al.²² based on control group data related to either baseline age

^†or axial elongation from the first year‡.

The baseline age and 1st-year axial elongation values used for the latter two estimations were as follows: Hiraoka et al.⁸ (10.0 y, 0.38 mm); Chamberlain et al.¹⁹ (10.1 y, 0.24 mm); Lam et al.²⁰ (10.0 y, 0.32 mm); and Zhang et al.²¹ (8.4 y, 0.41 mm). DIMS = Defocus-incorporated multiple segment (DIMS) spectacle lenses; NA = not applicable; RCT = randomized clinical trial.

Nevertheless, the cumulative efficacy of myopia control interventions, as shown in Table 4, demonstrates consistent trends across different treatment modalities, including orthokeratology, dual-focus soft contact lenses, defocus-incorporated multiple-segment spectacle lenses, and atropine. These results highlight that the efficacy of myopia control interventions accrues over time and emphasize the importance of considering long-term data when evaluating treatment strategies. However, they also underline the complexities of cross-study comparisons owing to variations in methodology, demographics, and the availability of long-term control data.

Like any study following up subjects for over a decade, this study suffers from several limitations, such as the potential bias introduced by subjects’ self-selection to continue wearing orthokeratology or single-vision correction for such a long follow-up period. Another limitation is the relatively low retention rate of orthokeratology and control subjects particularly at the last two study visits. The relatively high dropouts found at the 7- and 11-year visits might be attributed to several reasons including patients no longer using orthokeratology lenses owing to the expenses associated with the treatment or to myopia stabilization, thus no longer requiring myopia control treatment. It is also possible that control subjects with relatively stable myopia lost motivation to attend follow-up visits, thus leaving a larger number of subjects with rapidly progressive myopia within this group. Likewise, it is also possible orthokeratology subjects who continued wearing lenses for the entire 11-year period did so because of increased motivation as a result of successful myopia control efficacy. The latter could have resulted in potential imbalances between groups, potentially affecting the results of this study^12,22; however, of notice is that both study groups were relatively well matched (Table 1). Although Fig. 1 outlines the available data on participants’ treatment modalities and follow-up, detailed information on variability in treatment protocols, prescriber philosophies, or potential interruptions in orthokeratology lens wear after the initial 2 years of the study was not systematically collected and represents a limitation of this study. Another limitation relates to the long-term efficacy being solely tested in one single ethnicity (i.e., White European), although the short-term efficacy of orthokeratology lens wear in slowing myopia has been reported to be independent of ethnicity.¹⁷

CONCLUSIONS

Notwithstanding the above-mentioned limitations, this is the longest study to document the impact of orthokeratology contact lens wear in slowing the axial elongation of the eye in comparison with single-vision lens wear. Despite the initial burst in myopia control efficacy found following the initial 2 years of orthokeratology lens wear, this study demonstrates that the myopia control effect accrues over time leading to cumulative decreases in the axial elongation of the eye with longer periods of orthokeratology lens wear. More specifically, 11 years of orthokeratology lens wear provided a substantial slowing in the axial elongation of the eye, with a mean treatment effect of −0.69 mm in comparison with single-vision lens wear. Thus, these results inform eye care practitioners as to what sort of myopia control efficacy can be expected with long-term orthokeratology lens wear.