Wear-Related Visible Surface Alteration and Pseudomonas aeruginosa Adhesion in Orthokeratology Lenses

Abstract

Purpose

To compare grayscale-derived visible surface characteristics and Pseudomonas aeruginosa adhesion between unworn orthokeratology lenses and lenses worn for approximately 12 months, and to explore associations between grayscale-derived parameters and bacterial adhesion.

Methods

Sixty orthokeratology lenses were analyzed, including 20 unworn lenses and 40 lenses worn for approximately 12 months. Each worn lens was collected from a different wearer during routine replacement. Standardized slit-lamp images were analyzed with ImageJ to obtain mean gray value and gray-level standard deviation within the central 6.0-mm optical zone. In vitro P. aeruginosa adhesion was quantified as log10-transformed colony-forming units. Group differences were assessed using two-way ANOVA followed by exploratory simple-effect comparisons with Holm adjustment, and correlations between grayscale-derived parameters and bacterial adhesion were explored among worn lenses.

Results

Clinical wear was associated with significantly higher mean gray value, gray-level standard deviation, and P. aeruginosa adhesion compared with unworn lenses (all P < 0.001). Exploratory simple-effect comparisons identified higher mean gray value and bacterial adhesion in worn type A than worn type Z lenses after Holm correction, whereas the gray-level standard deviation difference did not remain significant after correction. None of the wear × lens-system interaction terms in the primary ANOVA was statistically significant. Grayscale-derived parameters were only weakly correlated with bacterial adhesion, and these associations were not significant after correction for multiple testing.

Conclusion

Orthokeratology lenses worn for approximately 12 months showed greater grayscale-derived visible surface alteration and higher P. aeruginosa adhesion than unworn lenses. These laboratory findings do not establish an increased clinical infection risk or an optimal replacement interval and require prospective clinical validation.

Introduction

Orthokeratology is widely used as a non-surgical approach for myopia control in children and adolescents, with clinical studies, reviews, and meta-analyses supporting its effectiveness in slowing axial elongation.^1–5 Because orthokeratology requires repeated overnight lens wear, ongoing follow-up is an essential part of clinical management, including assessment of refractive effect, lens fitting, ocular surface response, patient compliance, and lens condition.^5,6 Within this follow-up framework, microbial safety remains an important consideration, particularly because lens surface changes during long-term wear may influence deposit retention and bacterial adhesion.^7–13

The safety of long-term orthokeratology wear depends not only on fitting accuracy and patient compliance, but also on the condition of the lens surface after repeated use.^5,6,14–18 During nightly wear and daily handling, orthokeratology lenses are exposed to tear proteins, lipids, care solutions, manual cleaning, and mechanical interaction with the ocular surface, all of which may contribute to progressive surface alteration.^14–18 These changes may present as visible deposits, micro-scratches, altered surface uniformity, and reduced optical quality.^14,15,17,18 From a clinical follow-up perspective, such changes are relevant because they may affect lens performance, comfort, tear-lens interaction, and microbial retention.

Pseudomonas aeruginosa is one of the major pathogens associated with contact lens-related microbial keratitis because of its adhesive capacity, corneal invasiveness, and biofilm-forming potential.^{12,13,19–22} Its established relevance to contact lens-related infection and its ability to adhere to orthokeratology lens materials make it a clinically relevant model organism for laboratory adhesion studies. Previous studies have shown that P. aeruginosa can adhere to orthokeratology lens materials and that lens deposition and its removal influence bacterial adhesion to orthokeratology lens surfaces.^16,20 Although in vitro bacterial adhesion does not directly represent clinical infection risk, it provides a controlled laboratory outcome for comparing bacterial retention on worn and unworn lens surfaces.

In routine practice, lens surface condition is mainly assessed by visual inspection, slit-lamp examination, replacement history, and clinical judgment. High-resolution methods such as atomic force microscopy and scanning electron microscopy can provide detailed surface characterization, but they are not practical for routine clinical monitoring or larger sample analysis.^17,18 Previous work has demonstrated that image-based methods can quantify deposit-related cloudiness and deposit thickness on worn orthokeratology lenses,¹⁵ but mean gray value and gray-level standard deviation have not been established as validated measures of structural wear, surface roughness, or material degradation. Accordingly, the present study used these grayscale-derived variables only as standardized, exploratory descriptors of visible optical-zone surface alteration. Image processing was performed in ImageJ.²³

Commercial orthokeratology lens systems differ as complete products in reverse-geometry design, material properties, oxygen permeability, fenestration configuration, tear-exchange characteristics, and surface treatment. The two systems examined here were selected because both were used in routine clinical practice at the study center and represented distinct commercial system-level designs. Clinical observations of differences in lens appearance after prolonged use provided the rationale for exploring post-wear surface characteristics, but the study was not designed to attribute any observed difference to a single material or design feature.

The specific knowledge gap addressed was whether lenses collected after a clinically relevant long-wear interval show concurrent changes in standardized visible surface descriptors and in vitro bacterial adhesion. Approximately 12 months was selected because it corresponded to the routine replacement interval recommended for these lenses in clinical practice, allowing collection at a common replacement time point rather than implying an optimal replacement threshold. P. aeruginosa was selected because it is a major pathogen in contact lens-related microbial keratitis and has established relevance to adhesion studies involving orthokeratology lenses.^{12,13,19–21} Accordingly, this study compared unworn lenses with lenses collected at routine replacement after approximately 12 months of wear, compared two commercial lens systems at the system level, and explored whether central optical-zone grayscale parameters were associated with bacterial adhesion.

Methods

Study Design and Lens Samples

This laboratory-based observational study was conducted from August 1, 2025, to February 28, 2026, as the first non-interventional component of a broader prospectively registered research project on orthokeratology lens structural design and ocular surface microecology (Chinese Clinical Trial Registry, ChiCTR2500106186; registered July 17, 2025). The registration covers the overall research program, including subsequent prospective components involving tear and conjunctival sampling, ocular surface microbiome analysis, inflammatory protein profiling, and lens-design intervention studies. The present manuscript reports only the laboratory-based observational component conducted before those prospective intervention phases. No participant was prospectively assigned to a lens-design intervention in the work reported here, and the registered primary outcomes related to the microbiome, inflammatory protein, and intervention components were not evaluated in this manuscript.

The present study evaluated lens surface status and bacterial adhesion using a cross-sectional comparison of unworn and clinically worn lenses rather than longitudinal measurements of the same lenses. Sixty lenses were included: 10 unworn and 20 worn type A lenses, and 10 unworn and 20 worn type Z lenses. Each worn lens came from a different wearer, with one unilateral lens included per wearer to ensure statistical independence. No formal a priori sample-size or power calculation was performed for this exploratory laboratory-based component. The sample size reflected the eligible lenses available during the predefined collection period, informed by preliminary experiments showing marked differences between unworn and worn lenses. Orthokeratology lenses were custom manufactured to order according to individual lens parameters rather than produced in conventional batches. Within each commercial system, unworn comparator lenses were obtained from the same brand/manufacturer and matched lens specifications to the worn lenses to minimize differences attributable to lens geometry.

Before inclusion, all worn lenses were examined by slit-lamp biomicroscopy. Lenses with obvious accidental mechanical damage, such as large-area scratches caused by lens dropping or localized lens breakage, were excluded. Routine wear-related visible surface changes were not an exclusion criterion. Because this exclusion was based on clinical slit-lamp judgment rather than a quantitative damage scale, the study may not represent the most severely damaged lenses encountered in practice.

For the worn-lens groups, wearer-related clinical information was recorded, including age, sex, spherical equivalent refraction, mean simulated keratometry, and duration of lens wear. Wear duration across the 40 worn lenses averaged 1.047 ± 0.112 years and ranged from 0.833 to 1.395 years (approximately 10.0–16.7 months). Type A and type Z groups had mean wear durations of 1.036 ± 0.095 and 1.057 ± 0.127 years, respectively.

All wearers were prescribed a standardized lens care regimen consisting of daily cleaning and disinfection with multipurpose solution and weekly enzymatic protein removal (Menicon, Nagoya, Japan). Care and wear information was documented during routine follow-up at approximately 3-month intervals, including care-solution use, manual rubbing, protein-removal frequency, non-wear days, sleep duration, and use of low-concentration atropine when applicable. Wear frequency of at least 6 nights per week was confirmed from these follow-up records. After collection, all worn and unworn lenses underwent the same standardized protein-removal procedure before image acquisition, and the same lenses were subsequently used for the bacterial adhesion assay.

Standardized Image Acquisition and ImageJ Analysis

Before imaging, each lens underwent standardized protein removal treatment using Menicon AB solution and was rinsed with sterile deionized water. Lens images were acquired using a digital slit-lamp microscope (SL-D4; Topcon, Tokyo, Japan). The illumination beam was positioned at approximately 45°, with the binocular observation path approximately 90° to the incident illumination. The slit was opened to its maximum width, magnification was fixed at 16×, a black light-absorbing background was used, and focus was centered on the lens optical zone. The same acquisition configuration was maintained across lenses; a numeric illumination-intensity setting was not retained in the laboratory record.

All images were imported into ImageJ software and converted to 8-bit grayscale, yielding dimensionless pixel intensity values from 0 (black) to 255 (white). Image scale was calibrated using the known 10.6-mm overall lens diameter so that the predefined 6.0-mm circular region of interest (ROI) represented the same physical area in every image. The ROI was centered on the optical zone; reverse-curve and peripheral zones were excluded to reduce region-specific deposit accumulation and illumination artifacts. Mean gray value and gray-level standard deviation were measured directly within the ROI without threshold-based segmentation. This approach was adapted from prior image-based quantitative assessment of deposits on worn orthokeratology lenses,15 while the present study used the parameters as standardized descriptors of visible post-wear surface appearance rather than validated measures of structural degradation.

Two grayscale-derived parameters were extracted from each ROI: mean gray value and gray-level standard deviation. Mean gray value described overall pixel intensity within the optical zone, whereas gray-level standard deviation described local grayscale heterogeneity. These measures were treated as practical image descriptors rather than direct measures of surface roughness, physicochemical degradation, or microbiological risk. Three independently acquired images were analyzed for each lens, and the average of the three measurements was used for statistical analysis. Image analysis was performed by a masked examiner who was unaware of lens group allocation.

P. aeruginosa Adhesion Assay

A standardized P. aeruginosa strain (ATCC 27853) was used for the in vitro bacterial adhesion assay. Bacteria were recovered from frozen glycerol stock, cultured overnight in Luria-Bertani broth at 37°C with shaking, and adjusted spectrophotometrically to a 0.5 McFarland standard (approximately 1.5×10⁸ CFU/mL). The use of a single reference strain provided experimental standardization but was not intended to represent the full phenotypic diversity of clinical P. aeruginosa isolates.

Each lens was placed concave side upward in a 24-well plate containing 1 mL bacterial suspension and incubated statically at 37°C for 6 h to allow initial bacterial adhesion. Following incubation, lenses were rinsed three times with sterile phosphate-buffered saline to remove non-adherent bacteria and transferred into tubes containing 1 mL sterile phosphate-buffered saline. Adherent bacteria were detached by ultrasonication at 100 W for 5 min followed by vortex mixing for 1 min. This condition was selected through preliminary optimization using the same lens–bacteria system to obtain reproducible recovery of culturable adherent bacteria for CFU enumeration. Samples were kept on ice during sonication to minimize temperature elevation.

Serial 10-fold dilutions (10⁻¹–10⁻⁴) were prepared, and 20 μL aliquots were plated onto Luria-Bertani agar plates. After incubation at 37°C for 12 h, plates containing 30–300 colonies were counted. Bacterial load was calculated as follows:

All colony counting was performed by an investigator blinded to lens group allocation. Bacterial counts were log10-transformed prior to statistical analysis.

Statistical Analysis

Statistical analyses were performed using R software (version 4.2.0) and JASP software (version 0.16). All tests were two-sided, and P < 0.05 was considered statistically significant.

Baseline comparability between worn type A and worn type Z lens groups was assessed using Welch independent-samples t tests for continuous variables with approximately symmetric distributions and no major outliers, and Fisher’s exact test for categorical variables. Distributional form was examined using histograms, Q-Q plots, and Shapiro–Wilk tests. Because wear duration in the type Z group showed evidence of non-normality, a Wilcoxon rank-sum sensitivity analysis was additionally performed and yielded the same inference regarding between-group comparability. Standardized mean differences were calculated, with absolute values <0.3 considered to indicate acceptable between-group balance.

For each outcome variable (mean gray value, gray-level standard deviation, and log10CFU), two-way analysis of variance was performed with lens wear (unworn vs worn) and lens system (type A vs type Z) as fixed factors. Main effects, interaction effects, F values, P values, and partial eta-squared values were reported. Residual distributions and variance homogeneity were examined using Q-Q plots, Shapiro–Wilk tests, and Levene tests. Departures from normality and/or homogeneity were observed for the grayscale outcomes; therefore, the inferential emphasis was placed on the large wear-related effects, effect sizes, confidence intervals, and the consistency of direction across outcomes, while system-specific findings were interpreted cautiously. The log10 transformation of CFU improved variance homogeneity.

Following the factorial ANOVA, six type A-versus-type Z comparisons were examined as exploratory simple-effect analyses: one comparison within the worn group and one within the unworn group for each of the three outcomes. Because these comparisons were not specified in a formal a priori statistical analysis plan, they were not treated as confirmatory planned contrasts. Multiplicity across the six comparisons was controlled using the Holm procedure. Estimated differences, 95% confidence intervals, unadjusted P values, and Holm-adjusted P values were reported and interpreted in the context of the corresponding wear × lens-system interaction.

Repeatability of grayscale measurements was assessed using an average-measure intraclass correlation coefficient based on a two-way mixed-effects, absolute-agreement model, with 95% confidence intervals estimated by bootstrap resampling.

Among worn lenses, Spearman rank correlation analysis was used to explore associations between grayscale-derived surface parameters and bacterial adhesion. Because grayscale analysis was restricted to the central optical-zone ROI whereas bacterial adhesion was quantified from the whole lens, these correlation analyses were interpreted as exploratory assessments of association rather than validation of grayscale parameters as direct surrogates for bacterial adhesion. The Benjamini-Hochberg procedure was applied to control the false discovery rate across the three prespecified correlation analyses.

Results

Lens System Characteristics

The structural and material characteristics of the two commercial orthokeratology lens systems are summarized in Table 1. Both systems had the same optical zone diameter and overall diameter, but differed in design type, material, oxygen permeability, fenestration configuration, and plasma treatment.

Table 1.

Structural and Material Characteristics of the Orthokeratology Lens Systems Evaluated in this Study

Parameter	Type A Lens	Type Z Lens
Commercial product	Alpha Ortho-K	Menicon Z Night Ortho-K
Manufacturer	Alpha Corporation, Japan	Menicon Corporation, Netherlands
Design type	Four-zone, five-curve reverse-geometry design	Three-zone reverse-geometry design with a tangential zone
Material	Boston EM	Menicon Z
Oxygen permeability, Dk*	104	163
Optical zone diameter	6.0 mm	6.0 mm
Overall diameter	10.6 mm	10.6 mm
Fenestrations	None	Three fenestrations in the reverse zone
Plasma treatment	No	Yes

Notes: *Dk values are reported according to manufacturer-provided information.

Baseline Comparability of Worn-Lens Groups

A total of 40 worn orthokeratology lenses (20 type A and 20 type Z), obtained from 40 different wearers, were included in the baseline comparability analysis. Wearer-related baseline characteristics are summarized in Table 2. No statistically significant between-group differences were observed in age, spherical equivalent refraction, mean simulated keratometry, wear duration, or sex distribution. Mean wear duration was 1.036 ± 0.095 years for type A and 1.057 ± 0.127 years for type Z lenses. Across all worn lenses, the duration ranged from 0.833 to 1.395 years (approximately 10.0–16.7 months). Standardized mean differences for all continuous variables were <0.3, indicating broadly similar baseline characteristics between the two worn-lens groups.

Table 2.

Baseline Characteristics of Worn-Lens Groups

Variable	Type A (n = 20)	Type Z (n = 20)	Test	P value	SMD
Age, years	11.709 ± 1.199	11.872 ± 1.531	Welch t test	0.710	0.118
Spherical equivalent, D	−2.200 ± 0.963	−2.069 ± 1.094	Welch t test	0.689	0.127
Mean Sim-K, D	44.264 ± 1.541	44.483 ± 1.461	Welch t test	0.646	0.146
Duration of lens use, years	1.036 ± 0.095	1.057 ± 0.127	Welch t test	0.558	0.187
Sex, male/female	7/13	8/12	Fisher exact test	1.000	-

Notes: Values are presented as mean ± standard deviation unless otherwise indicated.

Abbreviations: SMD, standardized mean difference; Sim-K, simulated keratometry; D, diopter.

Descriptive Statistics

Descriptive statistics for grayscale-derived surface parameters and bacterial adhesion across the four experimental groups are summarized in Table 3. Representative grayscale slit-lamp images and bacterial culture plates are shown in Figure 1, and quantitative group distributions are shown in Figure 2. Unworn type A and type Z lenses showed low mean gray values and gray-level standard deviations, consistent with minimal visible surface alteration before wear. In contrast, worn lenses of both types showed higher grayscale values, greater grayscale heterogeneity, and higher P. aeruginosa adhesion. These wear-related differences were consistent with the visual appearance of representative lens images and the corresponding bacterial culture plates.

Table 3.

Descriptive Statistics of Outcome Variables in Each Group

Lens System	Lens Wear	n	Mean Gray value	Gray-Level Standard Deviation	log10CFU
Type A	Unworn	10	0.243 ± 0.223	1.710 ± 1.213	6.586 ± 0.229
			0.175 [0.112, 0.269]	1.611 [0.749, 2.278]	6.614 [6.543, 6.747]
	Worn	20	54.387 ± 24.993	19.519 ± 7.148	7.011 ± 0.220
			51.794 [41.820, 69.920]	20.011 [13.885, 24.679]	7.038 [6.926, 7.166]
Type Z	Unworn	10	0.345 ± 0.542	1.681 ± 1.427	6.606 ± 0.205
			0.060 [0.025, 0.490]	1.292 [0.572, 2.223]	6.669 [6.568, 6.718]
	Worn	20	39.816 ± 12.321	15.544 ± 6.489	6.800 ± 0.267
			36.446 [33.339, 43.272]	13.506 [11.484, 16.520]	6.801 [6.669, 7.028]

Notes: Values are presented as mean ± standard deviation and median [Q1, Q3].

Abbreviation: CFU, Colony-forming units.

Figure 1.

Representative grayscale slit-lamp images and corresponding P. aeruginosa culture plates of orthokeratology lenses. (A–D) show representative grayscale slit-lamp images of unworn type A, worn type A, unworn type Z, and worn type Z lenses, respectively. (E–H) show representative bacterial culture plates from the corresponding groups following the in vitro bacterial adhesion assay.

Figure 2.

Quantitative comparison of grayscale-derived surface parameters and bacterial adhesion across orthokeratology lens groups. Box-and-whisker plots showing (A) mean gray value, (B) gray-level standard deviation, and (C) log10-transformed P. aeruginosa colony-forming units (log10CFU) in unworn and worn type A and type Z orthokeratology lenses.

Among worn lenses, type A lenses had numerically higher mean gray value, gray-level standard deviation, and log10CFU than type Z lenses (Table 3 and Figure 2). These descriptive differences were examined further in exploratory simple-effect analyses and were interpreted in the context of the nonsignificant wear × lens-system interactions.

Main Effects of Lens Wear and Lens System

Two-way analysis of variance showed significant main effects of lens wear on all three outcome measures (Table 4; Figure 2). Specifically, lens wear was associated with mean gray value (F = 110.87, P < 0.001, partial η² = 0.664), gray-level standard deviation (F = 103.90, P < 0.001, partial η² = 0.650), and bacterial adhesion quantified as log10CFU (F = 22.89, P < 0.001, partial η² = 0.290). The estimated marginal worn-versus-unworn differences were 46.86 for mean gray value (95% CI, 37.95 to 55.76), 15.84 for gray-level standard deviation (95% CI, 12.72 to 18.95), and 0.309 for log10CFU (95% CI, 0.180 to 0.439).

Table 4.

Two-Way ANOVA for Effects of Lens Wear and Lens System on Visible Surface Alteration and Bacterial Adhesion

Outcome	Effect	df	F value	P value	Partial η2
Mean gray value	Lens wear	1,56	110.87	< 0.001	0.664
	Lens system	1,56	5.337	0.0254	0.0878
	Wear × Lens system	1,56	2.7269	0.1047	0.046
Gray-level standard deviation	Lens wear	1,56	103.90	< 0.001	0.650
	Lens system	1,56	3.30	0.075	0.056
	Wear × Lens system	1,56	1.61	0.209	0.028
log10CFU	Lens wear	1,56	22.89	< 0.001	0.290
	Lens system	1,56	4.83	0.032	0.079
	Wear × Lens system	1,56	3.19	0.079	0.054

Lens system also showed significant main effects on mean gray value and log10CFU, but not on gray-level standard deviation. Specifically, lens system was associated with mean gray value (P = 0.0254) and log10CFU (P = 0.032), whereas its effect on gray-level standard deviation did not reach statistical significance (P = 0.075).

No statistically significant interaction was observed between lens wear and lens system for mean gray value, gray-level standard deviation, or log10CFU (all P >0.05). Although the interaction term for log10CFU approached significance (P = 0.079), it did not meet the predefined threshold for statistical significance.

Exploratory Simple-Effect Comparisons within Wear Conditions

Exploratory simple-effect comparisons between type A and type Z lenses within each wear-condition subgroup are summarized in Table 5 and illustrated in Figure 2. Holm adjustment was applied across all six type A-versus-type Z comparisons. These subgroup analyses were used to describe system-level differences within a given wear condition and were not interpreted as evidence of differential response to wear in the absence of a statistically significant wear × lens-system interaction in the primary ANOVA.

Table 5.

Exploratory Simple-Effect Comparisons Between Lens Systems within Each Wear-Condition Subgroup

Outcome	Contrast	Lens Wear	Estimate	95% CI	Unadjusted P	Holm-Adjusted P
Mean gray value	Type A − Type Z	Worn	14.57	4.29 to 24.85	0.0063	0.0378
Mean gray value	Type A − Type Z	Unworn	−0.102	−14.65 to 14.44	0.989	1.000
Gray-level standard deviation	Type A − Type Z	Worn	3.98	0.38 to 7.57	0.0308	0.1231
Gray-level standard deviation	Type A − Type Z	Unworn	0.03	−5.05 to 5.11	0.991	1.000
log10CFU	Type A − Type Z	Worn	0.211	0.061 to 0.361	0.0065	0.0378
log10CFU	Type A − Type Z	Unworn	−0.020	−0.232 to 0.191	0.850	1.000

Notes: Comparisons were examined following the factorial ANOVA. Multiplicity across all six type A-versus-type Z comparisons was controlled using the Holm procedure. Estimates represent type A minus type Z.

Abbreviations: CFU, colony-forming units; CI, confidence interval.

Among worn lenses, the estimated type A-type Z difference was 14.57 for mean gray value (95% CI, 4.29 to 24.85; unadjusted P = 0.0063; Holm-adjusted P = 0.0378), 3.98 for gray-level standard deviation (95% CI, 0.38 to 7.57; unadjusted P = 0.0308; Holm-adjusted P = 0.1231), and 0.211 for log10CFU (95% CI, 0.061 to 0.361; unadjusted P = 0.0065; Holm-adjusted P = 0.0378). Thus, after multiplicity correction, the worn-group differences remained statistically significant for mean gray value and bacterial adhesion, but not for gray-level standard deviation.

Among unworn lenses, the corresponding type A-type Z differences were −0.102 for mean gray value (95% CI, −14.65 to 14.44; unadjusted P = 0.989; Holm-adjusted P = 1.000), 0.03 for gray-level standard deviation (95% CI, −5.05 to 5.11; unadjusted P = 0.991; Holm-adjusted P = 1.000), and −0.020 for log10CFU (95% CI, −0.232 to 0.191; unadjusted P = 0.850; Holm-adjusted P = 1.000).

Thus, the exploratory subgroup analyses identified selected system-level differences among worn lenses but not among unworn lenses. However, because none of the primary wear × lens-system interaction terms was statistically significant, these comparisons do not establish that type A and type Z lenses respond differently to clinical wear.

Associations Between Grayscale-Derived Parameters and Bacterial Adhesion

Spearman correlation analyses among worn lenses are summarized in Table 6.

Table 6.

Spearman Correlations Among Grayscale-Derived Surface Parameters and Bacterial Adhesion in Worn Lenses

Variable x	Variable y	ρ	95% CI	P value	Adjusted P value
Mean gray value	log10CFU	0.282	−0.009 to 0.535	0.078	0.118
Gray-level standard deviation	log10CFU	0.196	−0.146 to 0.518	0.226	0.226
Mean gray value	Gray-level standard deviation	0.741	0.579 to 0.839	<0.001	<0.001

Notes: Adjusted P values were calculated using the Benjamini-Hochberg procedure across the three prespecified correlation analyses reported in this table.

Abbreviations: CFU, colony-forming units; CI, confidence interval.

Mean gray value showed a weak positive correlation with bacterial adhesion (ρ = 0.282, 95% CI −0.009 to 0.535), but this association was not statistically significant after false discovery rate correction (adjusted P = 0.118). Gray-level standard deviation also showed a weak positive correlation with bacterial adhesion (ρ = 0.196, 95% CI −0.146 to 0.518), which was not statistically significant after correction (adjusted P = 0.226).

In contrast, mean gray value was strongly correlated with gray-level standard deviation (ρ = 0.741, 95% CI 0.579 to 0.839; adjusted P < 0.001).

Overall, the two grayscale-derived parameters captured related aspects of visible surface status, whereas their associations with whole-lens bacterial adhesion were weak in the present dataset.

Discussion

This laboratory-based observational study compared unworn orthokeratology lenses with lenses worn for approximately one year. Worn lenses showed greater grayscale-derived visible surface alteration and higher P. aeruginosa adhesion under controlled laboratory conditions. Associations between central-zone grayscale parameters and whole-lens bacterial adhesion were weak. The principal finding is therefore a wear-status difference across both image-based and microbiological laboratory outcomes, rather than evidence for a clinical infection risk or a specific replacement interval.

The wear-related differences were consistent across image-based and microbiological outcomes. Worn lenses had higher mean gray value, greater gray-level heterogeneity, and higher P. aeruginosa adhesion than unworn lenses. Previous studies have shown tear-film deposition and changes in lens surface properties after orthokeratology wear.^14,15 Repeated exposure to tear components, care solutions, manual rubbing, and ocular surface contact may contribute to deposits and microscopic surface changes that alter bacterial attachment.^{14–16,19–21} The present findings extend this laboratory evidence by showing concurrent wear-status differences after a similar long-wear interval, but they do not establish when those changes developed or whether they affect clinical outcomes.

Although lens system was associated with selected outcomes in the primary ANOVA, the main effect of wear status showed larger effect sizes. More importantly, no statistically significant wear × lens-system interaction was observed for any outcome. The exploratory within-wear-condition comparisons should therefore not be interpreted as evidence that type A and type Z lenses respond differently to wear. They are secondary system-level observations that may reflect the combined influence of multiple product characteristics and require confirmation in studies designed for direct system-by-wear comparisons.

The exploratory subgroup analyses showed selected type A-type Z differences among worn lenses but not among unworn lenses after multiplicity correction. Because the two products differ simultaneously in material, oxygen permeability, reverse-geometry design, fenestration configuration, and surface treatment, no individual feature can be identified as the cause of these differences. More importantly, none of the wear × lens-system interactions in the primary ANOVA was statistically significant. The subgroup findings should therefore be regarded as secondary system-level observations rather than evidence that one system is intrinsically more susceptible to wear-related change.

Mean gray value and gray-level standard deviation are indirect image descriptors. Higher values were associated with worn status, but these measures have not been validated against objective measures of surface morphology, nanoscale roughness, material degradation, or deposit composition. Prior orthokeratology research has shown that image-based methods can quantify deposit-related cloudiness and deposit thickness, but those methods used different metrics. Accordingly, the present grayscale variables are best interpreted as standardized descriptors of visible optical-zone alteration rather than direct measures of physical degradation. Their weak associations with bacterial adhesion further indicate that visible grayscale appearance should not be treated as a surrogate for microbiological risk.

The measurement design may partly explain the weak grayscale-adhesion relationship. Grayscale analysis was limited to the central 6.0-mm optical zone after standardized cleaning, whereas the adhesion assay quantified bacteria recovered from the whole lens. Reverse-curve and adjacent peripheral regions have distinct geometry and tear-exchange characteristics and may accumulate deposits or localized surface changes during long-term wear. Because these regions were not included in the grayscale ROI, central optical-zone grayscale measures may not fully represent whole-lens surface status, potentially attenuating their association with whole-lens bacterial recovery. The present grayscale method was developed for standardized ex vivo imaging of removed lenses and was not designed for direct on-eye measurement. Any on-eye application would require a separate acquisition and validation framework because the optical environment differs in vivo.

Mean gray value and gray-level standard deviation were strongly correlated, indicating that the two image-derived parameters captured related aspects of the standardized optical-zone image. This internal association does not constitute validation against physical surface properties. Grayscale analysis may nevertheless provide a simple, non-destructive supplementary method for standardized laboratory comparison of visible lens surface status, while methods such as atomic force microscopy, scanning electron microscopy, contact-angle measurement, or chemical deposit analysis are required for direct characterization of surface morphology, roughness, wettability, or composition.

The clinical implications should therefore remain limited. Greater recoverable P. aeruginosa in the in vitro assay does not demonstrate greater microbial contamination during wear, ocular surface inflammation, microbial keratitis, or any other patient-level complication. Similarly, comparing unworn lenses with lenses worn for approximately 12 months cannot identify an optimal replacement interval or show that replacement at a particular time improves safety. The findings are best regarded as preliminary laboratory evidence that may motivate prospective longitudinal studies linking lens-surface measurements with ocular-surface and infection-related outcomes.

This study has several limitations. The sample size was modest and the unworn subgroups contained only 10 lenses per system, limiting precision and statistical power, particularly for interaction and between-system estimates. No formal a priori sample-size calculation was performed. The cross-sectional comparison of independent unworn and approximately 12-month-worn lenses cannot define the temporal progression of surface change or an optimal replacement interval. Excluding lenses with obvious accidental large-area scratching or breakage may have reduced representation of the most severely damaged lenses. The grayscale outcomes showed departures from ideal ANOVA distributional assumptions, and the image-derived measures were not validated against atomic force microscopy, scanning electron microscopy, contact-angle measurement, or direct surface roughness assessment. The bacterial adhesion assay used a single laboratory strain (P. aeruginosa ATCC 27853), so the findings may not generalize to other clinical strains or to in vivo microbial keratitis risk. Although the detachment conditions were selected through preliminary protocol optimization, bacterial recovery efficiency and viability were not formally quantified in a separate validation experiment. In addition, the in vitro assay cannot reproduce tear turnover, blinking, mucin interactions, microbial competition, routine care-solution exposure, or host defense on the ocular surface. Future studies should include staged or longitudinal wear-duration cohorts, larger samples, design-controlled comparisons, multimodal surface assessment, multiple clinically relevant bacterial strains, and patient-level ocular-surface outcomes.

Conclusion

Orthokeratology lenses worn for approximately 12 months showed greater grayscale-derived visible surface alteration and higher P. aeruginosa adhesion than unworn lenses. Grayscale-derived parameters were only weakly related to bacterial adhesion. The system-level subgroup differences do not demonstrate different wear responses because the wear × lens-system interactions were not statistically significant. These laboratory findings do not establish clinical infection risk or an optimal lens replacement interval and require prospective longitudinal validation.

Funding Statement

This work was supported by the Fundamental Research Funds for the Central Universities, China (No. DUT25YG270), and the 2025 Dalian Municipal Guiding Plan Project for Life and Health (No. 2025ZDJH01PT076). The funders had no role in the study design; collection, analysis, or interpretation of the data; preparation of the manuscript; or the decision to submit the manuscript for publication.

AI-Assisted Writing Disclosure

During the preparation of this manuscript, the authors used DeepSeek V4 solely for English language polishing because the authors are not native English speakers. No AI-assisted tool was used to generate scientific content, data, analyses, interpretations, or conclusions. After using this tool, the authors carefully reviewed and edited the manuscript and take full responsibility for the content of the submitted work.

Clinical Trial Registration

This manuscript reports the non-interventional laboratory-based observational component of a broader prospectively registered research project (Chinese Clinical Trial Registry, ChiCTR2500106186; registered July 17, 2025). The broader registration includes subsequent prospective components involving ocular surface microbiome analysis, inflammatory protein profiling, and lens-design interventions. No participant was prospectively assigned to an intervention in the work reported in this manuscript, and the registered prospective primary outcomes are not reported here.

Abbreviations

ANOVA, analysis of variance; CFU, colony-forming units; Dk, oxygen permeability; ROI, region of interest; Sim-K, simulated keratometry; SMD, standardized mean difference.

Data Sharing Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Ethics Approval and Informed Consent

The study protocol was approved by the Institutional Review Board of The Third People’s Hospital of Dalian, Affiliated to Dalian Medical University (Approval No. 2025-155-001; approval date: June 27, 2025), and was conducted in accordance with the tenets of the Declaration of Helsinki. For worn-lens samples and associated clinical data, written informed consent was obtained before study initiation. Because all wearers contributing worn lenses were minors, written informed consent was obtained from their legal guardians, and age-appropriate assent was obtained from the wearers when applicable. Unworn manufacturer-supplied lenses did not involve human participants, biological samples, or identifiable personal data. No animal experiments were conducted in this study.

Consent for Publication

The manuscript contains no identifiable personal information, images, or recordings.

Author Contributions

All authors made a significant contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work.

Disclosure

Dr Chunxiao Yan and Dr Lijun Zhang report support for the manuscript from provision of study materials, during the conduct of the study.The authors report no conflicts of interest in this work.