Assessment of factors affecting anterior chamber depth from data obtained from health checkup participants in Japan

Abstract

Measurement of anterior chamber depth (ACD), an important marker for the screening of primary angle-closure glaucoma, requires biometry, which is not readily used. This study assessed the relationship between ACD and health check-up data findings from participants with good corrected visual acuity in Japan. Participants underwent ophthalmic, anthropometric, and hematological assessments. The mean ACD of all 3060 participants was 3.33 ± 0.34 mm [2.22–4.72 mm]. Multivariable linear regression analysis was performed to determine factors that were significantly correlated with ACD, and logistic regression analysis was performed to predict ACD < 2.70 mm. Multivariable linear regression analysis showed that age, sex, intraocular pressure, spherical equivalent refractive error (SER), height, and fasting blood sugar levels significantly correlated with ACD (P < 0.05). Logistic regression analysis showed that age, sex, and SER were the best predictors of ACD < 2.70 mm. The area under receiving operator characteristic curves of ‘age and SER’ and ‘age, SER, and sex’ were 0.821 and 0.835, respectively, with no significant difference (P = 0.122). In conclusion, ACD correlates with several parameters, and age and SER may be particularly important for predicting ACD in participants undergoing health checkups.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-024-82096-1.

Introduction

Primary angle-closure glaucoma (PACG) is the leading cause of blindness in the Asian population^1,2and can occur even in individuals with good visual acuity (VA). Among the ocular risk factors for PACG, a shallow anterior chamber is the most consistent^3–5. Anterior chamber depth (ACD) is important for PACG screening, as many studies support the notion that ACD can be treated as a surrogate marker for PACG^6–8. However, the measurement of ACD requires biometry, which is not readily available in daily ophthalmology practice. Therefore, demographic and anthropometric risk factors are more practical for population screening for PACG.

Many factors, including sex^3,5,9,10, height^11–15, age^{4,5,9,11,16,17}, intraocular pressure (IOP)^18,19, blood sugar status^20,21, and obesity^13,22,23are correlated with ACD, although correlation strength varies by the race reported in each study. This could be because changes in the anterior segment structures include a variety of factors such as vacuolation of the iris pigment epithelium^24,25, thickening of the basement membrane of the ciliary processes^26,27, thickening of the cornea^28,29, and cataract formation^30,31. Detecting the parameters that affect the anterior ocular segment biometry for each race or region will help in early intervention and effective treatment to reduce the risk of vision loss.

Old age, female sex, and short height are significantly associated with shallow ACD³², however, they cannot be used solely to predict ACD; the area under the curves (AUCs) of receiving operator characteristic (ROC) curves for predicting an ACD < 2.70 mm (used as one of the thresholds of PACG³²) with age, sex, and height were less than 0.7. There may be other combinations of parameters that can predict ACD more robustly.

Determining factors that correlate with a shorter ACD in those with good corrected VA at health checkups is important in the selection of participants for thorough follow-up examinations. This study aimed to determine the relationship between ACD and ocular examination findings, hematologic tests, and anthropometric data in health checkup participants from mainland Japan.

Methods

Study design

This was a retrospective single-center study, which adhered to the tenets of the Declaration of Helsinki. The study protocol was approved and requirement for informed consent was waived by the Institutional Review Board of Nagoya University Graduate School of Medicine (Nagoya, Aichi, Japan) (Approval No. 2017 − 0283) owing to the retrospective nature of the study. An opt-out method was implemented to obtain informed consent from participants. All participant data were anonymized before analysis.

Participants

Participants underwent health and eye screening between October 2022 and September 2023 at the Aichi Health Promotion Foundation (Aichi, Japan). All participants underwent ophthalmic parameter and anthropometric assessments as well as hematological tests. The inclusion criteria were (1) participants aged 20–99 years at the time of enrollment and (2) participants with corrected VA of 20/20 or greater. The exclusion criterion was a history of ocular surgery, including cataract, vitreous, and corneal surgery, as obtained from a medical questionnaire (See Figure S1 for details on the questionnaire administered). Based on the aforementioned criteria, one eye from each participant was included in the study. In cases where both eyes met the inclusion criteria, the right eye was considered the study eye.

Examination protocols

All participants underwent complete ophthalmologic assessments on a single day, during which anthropometric and hematologic tests were performed and medical histories were obtained. VA was measured using an automatic vision tester (CA-1000, Tomey, Nagoya, Aichi, Japan). The refractive error was measured using an auto Ref/keratometer (Canon, RK-F2, Tokyo, Japan). The ACD was measured as the distance between the anterior corneal surface and anterior lens surface by partial coherence interferometry (Tomey, OA-2000, Nagoya, Aichi, Japan). The IOP was measured using a non-contact tonometer (Canon, TX-20P, Tokyo, Japan). Fasting blood samples were collected by venipuncture, and serum was used for the assays. Routine biochemical analyses were performed in the laboratory of our facility. Anthropometric measurements (height and weight) were obtained during screening and used to calculate body mass index (BMI) (kg/m²). All data were evaluated by two ophthalmologists (T.I. and K.Y.).

Statistical analyses

For statistical analyses, decimal values of VA were converted to logarithm of the minimum angle of resolution (logMAR) units.

The spherical equivalent refractive error (SER) was calculated using the Eq. (1):

1

Since the exclusion of a history of intraocular surgery, including cataracts, was questionnaire-based, outliers for ACD were excluded using the Smirnov–Grubbs test. The Kolmogorov-Smirnov test was conducted to determine the distribution of the data. Continuous data are presented as mean ± standard deviation (SD) and categorical data as number (%).

Univariable linear regression analysis was performed separately for each variable. ACD was used as the response variable, and the explanatory variables included age, sex, IOP, SER, height, weight, BMI, systolic and diastolic blood pressure, average blood pressure, heart rate, and hematologic test results (up to 17 hematologic parameters). To build a multivariable model, a stepwise selection method using Akaike information criterion (AIC) was adopted with an entry P value of < 0.05 after Holm–Bonferroni correction. Standardized partial regression coefficients (β) were calculated for the independent variables after unifying the units of the variables by standardizing all variables to a mean of 0 and variance of 1. The unstandardized partial regression coefficient (B) and 95% confidence intervals (95% CI) were calculated to present the results of the final model with backward stepwise selection. The variance inflation factor (VIF) was calculated to estimate the severity of multicollinearity. In addition, logistic regression analysis with backward stepwise model selection using AIC was performed to predict ACD < 2.70 mm. The efficacy of predicting ACD < 2.70 mm was analyzed using ROC curves, and the AUCs were calculated. For AUC values, a value of 1.00 is considered a perfect, 0.90–0.99 an excellent, 0.80–0.89 a good, 0.70–0.79 a fair, 0.51–0.69 a poor, and 0.50 a worthless test³³. DeLong’s test was used for comparing AUCs. Significance was set at a P value < 0.05 for all statistical analyses. Considering that participants whose fasting glucose values were < 110 mg/dL were classified as normal in Japan³⁴, fasting blood sugar (FBS) was also analyzed as a categorical variable with < 110 mg/dL set as normal and ≥ 110 mg/dL as abnormal.

Analyses were performed with scikit-learn version 0.24.0 (https://scikit-learn.org/stable/install.html) based on Python version 3.6.7 (https://www.python.org/downloads/release/python-367/) and lme4 version 1.1–30 based on R software version 4.2.2. (https://www.r-project.org/).

Results

Demographic information

A total of 4679 individuals took part in the health and eye check-up screening. Of these, 3276 eyes met all the inclusion criteria and did not meet any exclusion criteria. Of the 3276 individuals with these eyes, 3072 did not have any missing data and were used for further statistical analyses. The Smirnov–Grubbs test for ACD identified 12 individuals as outliers and excluded them (ACDs = 5.78, 5.75, 5.32, 5.23, 5.21, 5.17, 5.13, 4.98, 4.88, 4.77, and 4.76 mm). The mean ACD ± standard deviation (SD) was 3.33 ± 0.34 mm (range: 2.22–4.72 mm), and the distribution of ACD was normal (P = 0.21 by Kolmogorov–Smirnov test). The mean ACDs in women and men were 3.25 ± 0.35 mm and 3.37 ± 0.32 mm, respectively (P < 0.001). The main demographic information from the participants used in the statistical analyses is presented in Table 1.

Table 1.

Demographic information.

Anthropometric data
Number of eyes	3060
Age (years)	53.6 ± 10.2 [23, 84]
Sex (Male/Female)	2121 (69%)/939 (31%)
Height (cm)	166.9 ± 8.3 [138.7, 191.6]
Weight (kg)	65.9 ± 13.2 [26.2, 128.6]
BMI (kg/m2)	23.5 ± 3.8 [12.1, 43.2]
sBP (mmHg)	121.5 ± 16.6 [75, 205]
dBP (mmHg)	76.0 ± 12.0 [35, 133]
average BP (mmHg)	91.2 ± 12.8 [53.7, 151]
HR (/min)	68.9 ± 10.3 [34, 133]
Ocular examination data
VA (logMAR units)	−0.078 ± 0.070 [−0.18, 0.0]
IOP (mmHg)	13.6 ± 2.8 [8.0, 27.0]
SER (D)	−2.5 ± 2.9 [−14.25, + 4.0]
ACD (mm)	3.33 ± 0.34 [2.22, 4.72]
Hematologic data
WBC (103/mL)	5.24 ± 1.43 [2.1, 14.4]
RBC (104/mL)	469 ± 44 [321, 668]
Hb (g/dL)	14.3 ± 1.32 [6.9, 15.2]
Ht (%)	43.3 ± 3.5 [23.8, 55.1]
Plt (104/µL)	23.7 ± 5.0 [6.5, 50]
TG (mg/dL)	112 ± 85 [15, 1438]
HDL (mg/dL)	61.6 ± 15.3 [29, 153]
LDL (mg/dL)	127 ± 29.2 [28, 265]
Total cholesterol (mg/dL)	200 ± 32 [96, 377]
FBS (mg/dL)	102 ± 18 [57, 280]
HbA1c (%)	5.7 ± 0.60 [4.4, 11.2]
Cre (mg/dL)	0.81 ± 0.35 [0.32, 17.51]
eGFR (mL/min/1.73m2)	74.8 ± 13.7 [2.6, 189.7]
AST (U/L)	22.4 ± 11.1 [6, 261]
ALT (U/L)	24.0 ± 19.5 [4, 28]
γGTP (U/L)	38.6 ± 44.1 [5, 42]
ALP (U/L)	67.8 ± 19.0 [13, 229]

BMI, body mass index; sBP, systolic blood pressure; dBP, diastolic blood pressure; HR, heart rate; VA, visual acuity; logMAR, logarithm of the minimum angle of resolution; IOP, intraocular pressure; SER, spherical equivalent refractive error; D, diopter; ACD, anterior chamber depth; WBC, white blood cell; RBC, red blood cell; Hb, hemoglobin; Ht, hematocrit; Plt, platelet; TG, triglyceride; HDL, high-density lipoprotein cholesterol; LDL, low-density lipoprotein cholesterol; FBS, fasting blood sugar; HbA1c, hemoglobin A1c; Alb, albumin; Cre, creatinine; eGFR, estimated glomerular filtration rate; AST, aspartate aminotransferase; ALT, alanine aminotransferase;γGTP, γ-glutamyl transpeptidase; ALP, alkaline phosphatase.

Data are mean ± standard deviation [range].

Associations between ACD and other health checkup factors

First, the significance of the correlations between ACD and 28 health check-up measurements including ophthalmic, anthropometric, and hematologic findings were determined. The results of the univariable linear regression analysis are presented in Table 2. The factors that significantly correlated with ACD (P value < 0.05) after the Holm–Bonferroni correction were age, sex, IOP, SER, height, weight, BMI, systolic blood pressure, red blood cell, hemoglobin, hematocrit, triglyceride, high-density lipoprotein cholesterol (HDL), total cholesterol, FBS, hemoglobin A1c, estimated glomerular filtration rate, and alanine aminotransferase.

Table 2.

Univariable linear regression analyses between ACD and explanatory variables.

	Regression coefficient (β)	95% CI [0.025, 0.975]	P value	P value corrected using the Holm–Bonferroni method
Age (years)	−1.1 × 10−2	−1.2 × 10−2, −1.0 × 10−2	< 0.001*	< 0.001*
Sex (Male/Female)	−1.2	−1.4, −8.9 × 10−2	< 0.001*	< 0.001*
IOP (mmHg)	1.2 × 10−2	8.0 × 10−3, 1.6 × 10−2	< 0.001*	< 0.001*
SER (D)	−4.5 × 10−2	−4.9 × 10−2, −4.1 × 10−2	< 0.001*	< 0.001*
Height (cm)	8.6 × 10−3	7.1 × 10−3, 1.0 × 10−2	< 0.001*	< 0.001*
Weight (kg)	3.9 × 10−3	3.0 × 10−3, 4.8 × 10−3	< 0.001*	< 0.001*
BMI (kg/m2)	5.6 × 10−3	2.4 × 10−3, 8.8 × 10−2	< 0.001*	0.008*
sBP (mmHg)	−1.2 × 10−3	−2.0 × 10−3, −5.5 × 10−4	< 0.001*	0.007*
dBP (mmHg)	5.1 × 10−4	−7.5 × 10−4, 1.3 × 10−3	0.623	0.623
average BP (mmHg)	−5.7 × 10−4	−1.5 × 10−3, 3.7 × 10−4	0.235	n.s.
HR (/min)	6.5 × 10−4	−5.2 × 10−4, 1.8 × 10−3	0.276	n.s.
WBC (103/mL)	7.9 × 10−4	−5.0 × 10−5, 1.6 × 10−3	0.065	0.392
RBC (104/mL)	1.2 × 10−3	9.2 × 10−4, 1.5 × 10−3	< 0.001*	< 0.001*
Hb (g/dL)	2.9 × 10−2	2.0 × 10−2, 3.8 × 10−2	< 0.001*	< 0.001*
Ht (%)	9.4 × 10−3	6.0 × 10−3, 1.3 × 10−2	< 0.001*	< 0.001*
Plt (104/µL)	3.2 × 10−3	8.9 × 10−4, 5.7 × 10−3	0.007	0.071
TG (mg/dL)	2.1 × 10−4	7.2 × 10−5, 3.5 × 10−4	0.003	0.034*
HDL (mg/dL)	−3.1 × 10−3	−3.9 × 10−3, −2.4 × 10−3	< 0.001*	< 0.001*
LDL (mg/dL)	1.9 × 10−4	−2.3 × 10−4, 6.0 × 10−4	0.377	0.754
Total cholesterol (mg/dL)	−6.5 × 10−4	−1.0 × 10−3, −2.8 × 10−4	< 0.001*	0.008*
FBS (mg/dL)	−1.3 × 10−3	−2.0 × 10−3, −6.4 × 10−4	< 0.001*	0.002*
HbA1c (%)	−6.5 × 10−2	−8.5 × 10−2, −4.5 × 10−2	< 0.001*	< 0.001*
Cre (mg/dL)	3.8 × 10−2	3.4 × 10−3, 7.3 × 10−2	0.031	0.250
eGFR (mL/min/1.73m2)	3.0 × 10−3	2.2 × 10−3, 3.9 × 10−3	< 0.001*	< 0.001*
AST (U/L)	5.5 × 10−4	−5.3 × 10−4, 1.6 × 10−3	0.322	0.966
ALT (U/L)	1.2 × 10−3	6.3 × 10−4, 1.9 × 10−3	< 0.001*	< 0.001*
γGTP (U/L)	3.2 × 10−4	5.0 × 10−5, 6.0 × 10−4	0.020	0.183
ALP (U/L)	−6.7 × 10−4	−1.3 × 10−3, −3.8 × 10−5	0.038	0.265

CI, confidence interval; IOP, Intraocular pressure; SER, spherical equivalent refractive error; D, diopter; ACD, anterior chamber depth; BMI, body mass index; sBP, systolic blood pressure; dBP, diastolic blood pressure; HR, heart rate; WBC, white blood cell; RBC, red blood cell; Hb, hemoglobin; Ht, hematocrit; Plt, platelet; TG, triglyceride; HDL, high density lipoprotein cholesterol; LDL, low density lipoprotein cholesterol; FBS, fasting blood sugar; HbA1c, Hemoglobin A1c; Alb, albumin; Cre, creatinine; eGFR, estimated glomerular filtration rate; AST, aspartate aminotransferase; ALT, alanine aminotransferase;γGTP, γ-glutamyl transpeptidase; ALP, alkaline phosphatase; n.s., not significant.

* P value < 0.05.

Adjusted P values for factors that were not significant in univariable models are not shown as they were insignificant after Holm–Bonferroni adjustment as well.

Explanatory variables (P value < 0.05) after Holm–Bonferroni correction were then tested using multivariable linear regression analysis; considering multicollinearity, weight, hemoglobin, hematocrit, and triglycerides were not used. Backward stepwise model selection using AIC showed that multivariable linear regression analysis with explanatory variables of age, sex, IOP, SER, height, BMI, HDL, and FBS were the best explanatory factors for variations in ACD (R² = 0.24). Age (P < 0.001), sex (P < 0.001), IOP (P = 0.012), SER (P < 0.001), height (P = 0.012), and FBS (P < 0.001) significantly correlated with ACD (Table 3).

Table 3.

Multivariable linear regression analysis between ACD and explanatory variables.

	Standardized partial regression coefficient (β)	Partial regression coefficient (B)	95% CI [0.025, 0.975]	P value	VIF
Age (years)	−0.19	−7.8 × 10−3	−9.0 × 10−3, −6.6 × 10−3	< 0.001*	1.3
Sex (Male:0, Female:1)	−3.8 × 10−2	−9.5 × 10−2	−0.13, −6.0 × 10−2	< 0.001*	2.4
IOP (mmHg)	3.8 × 10−2	5.0 × 10−3	1.1 × 10−3, 8.9 × 10−3	0.012*	1.1
SER (D)	−0.26	−3.6 × 10−2	−4.0 × 10−2, −3.2 × 10−2	< 0.001*	1.1
Height (cm)	5.1 × 10−2	2.4 × 10−3	5.3 × 10−4, 4.3 × 10−3	0.012*	2.2
BMI (kg/m2)	3.9 × 10−2	3.2 × 10−2	2.9 × 10−2, 3.5 × 10−2	0.063	1.4
HDL (mg/dL)	−3.6 × 10−2	−6.1 × 10−4	−1.4 × 10−3, 2.2 × 10−4	0.090	1.5
FBS (mg/dL)	−0.13	−1.4 × 10−3	−2.1 × 10−3, −7.4 × 10−4	< 0.001*	1.2

CI, confidence interval; VIF, variance inflation factor; IOP, intraocular pressure; SER, spherical equivalent refractive error; D, diopter; BMI, Body mass index; HDL, high density lipoprotein cholesterol; FBS, fasting blood sugar.

* P value < 0.05.

Because smaller ACD has a higher correlation with acute cases of PACG and ACD < 2.70 mm has been used as one of the thresholds of PACG³², the predictability of this threshold was also analyzed. Logistic regression analysis adapted to backward stepwise model selection using the AIC revealed that age, sex, and SER were the best predictors of ACD < 2.70 mm (Table 4). For these three items that showed statistically significant differences (age, sex, and SER), ROC curves of seven different combinations of parameters for predicting ACD < 2.70 mm were plotted (Fig. 1). The AUCs of ROC curves of ‘age and SER’ and ‘age, sex and, SER’ were 0.821 and 0.835, respectively (Table 5), with no significant difference (P = 0.122); those of ‘sex and SER’ and ‘age, sex, and SER’ were 0.817 and 0.835, respectively, and were significantly different (P = 0.047). To include FBS, which was not statistically significant in the logistic regression analysis (Table 4), ROC curves of 15 different combinations of parameters for predicting ACD < 2.70 mm were also plotted and the superiority of the detection ability by the combination of ‘age and SER’ was consistent (Figure S2).

Table 4.

Logistic regression analysis between ACD and explanatory variables.

	Odds ratio	95% CI [0.025, 0.975]	P value	VIF
Age (years)	0.948	0.925, 0.972	< 0.001*	1.18
Sex (Male:0, Female:1)	0.339	0.213, 0.538	< 0.001*	1.10
SER (D)	0.656	0.579, 0.745	< 0.001*	1.14
FBS (mg/dL)	0.990	0.979, 1.000	0.084	1.11

The odds ratio was calculated using a binary logistic regression model, in which a deeper anterior chamber depth (ACD ≥ 2.70 mm) was defined as 1 and a shallower chamber depth (ACD < 2.70 mm) was defined as 0.

CI, confidence interval; VIF, Variance Inflation Factor; SER, spherical equivalent refractive error; D, diopter; FBS, fasting blood sugar; ACD, anterior chamber depth.

* P value < 0.05.

Fig. 1.

ROC curves of 7 different combinations of parameters for predicting ACD < 2.70 mm. ROC curves of the different combinations of parameters for predicting ACD < 2.70 mm are plotted. ROC, receiver operating characteristic; SER, spherical equivalent refractive error; ACD, anterior chamber depth.

Table 5.

The AUC of ROC curves of seven different combinations of parameters.

Combination	AUC	95% CI [0.025, 0.975]	P value
Age	0.726	0.673, 0.778	< 0.001*
Sex	0.605	0.552, 0.658	< 0.001*
SER	0.805	0.764, 0.847	0.008*
Age and sex	0.758	0.708, 0.807	< 0.001*
Age and SER	0.821	0.782, 0.861	0.122
Sex and SER	0.817	0.775, 0.859	0.047*
Age, sex, and SER	0.835	0.796, 0.874	-

AUC, area under the curve; CI, confidence interval; SER, spherical equivalent refractive error.

DeLong’s test was used for comparing AUCs based on the ROC curve of ‘age, sex and, SER.’.

* P value < 0.05.

In the analysis using FBS as a categorical variable, with < 110 mg/dL as normal and ≥ 110 mg/dL as abnormal, backward stepwise model selection using AIC also showed that multivariable linear regression analysis with explanatory variables of age, sex, IOP, SER, height, BMI, HDL, and FBS were the best explainers for the variations of ACD (R² = 0.24). Age (P < 0.001), sex (P < 0.001), IOP (P = 0.022), SER (P < 0.001), height (P = 0.015), and FBS (P = 0.010) significantly correlated with ACD (Table S1). Logistic regression analysis using backward stepwise model selection with AIC and FBS as categorical variables showed that age, sex, and SER were the best predictors of ACD < 2.70 mm (Table S2).

Discussion

This study is the first to measure ACDs in over 3000 people from mainland Japan. Multivariable regression analysis showed that age, sex, IOP, SER, and FBS significantly correlated with ACD. The results suggest that age and SER alone may be useful as parameters to determine the indicator for a threshold of PACG (ACD < 2.70 mm in healthy participants with good corrected VA).

Compared with a previous report³² determining factors that correlated with ACD, the AUC of the ROC curve for predicting ACD < 2.70 mm with age and sex was higher in this study (0.687 vs. 0.758). This may be due to population differences; the previous study had a higher proportion of women than the present study (56.7% vs. 30.7%) as well as a wider SD and ACD range (2.96 ± 0.45 mm, range 1.86–4.45 mm vs. 3.33 ± 0.34 mm, range 2.22–4.72 mm).

Female sex is traditionally viewed to be a risk factor for PACG^2,4,8. Although female sex was a factor in determining ACD in the present analysis, the spherical equivalent (especially tending towards hyperopia) might be a more important factor in predicting ACD < 2.70 mm that may result in PACG, considering the AUCs of ROC curves for combinations of ‘age, sex, and SER,’ ‘age and SER,’ and ‘sex and SER’. These results are in agreement with those of earlier studies, which showed that a deeper anterior chamber is present in middle-aged and elderly individuals with myopia^16,35–38.

Although FBS was not a statistically significant predictor of ACD < 2.70 mm in the logistic regression analysis (P = 0.084), a statistically significant negative correlation was observed between FBS and ACD in the multivariable linear regression analysis (P< 0.001). A previous report showed that ACD was significantly narrower in participants with diabetic mellitus after controlling for age and sex²¹. It was suggested that the difference might be due to enhanced oxidative stress and/or reduction in antioxidant status^20,39–41, which may lead to thickening of the anteroposterior lens diameter, reduction in anterior chamber volume, and narrowing of the angle. The present results support these findings and suggest that FBS can be used as a screening tool for narrower ACD to rule out angle closure, although differences in retinal findings, including fundus color, were not examined.

The positive correlation between IOP and ACD (higher IOP = deeper ACD) in multivariable regression analysis is controversial. This is contrary to what is found in PACG with high IOP, where shallow ACD is a known risk factor. Previous reports have also shown no statistically significant correlation was observed between IOP and ACD^19,42. This suggests that the elevation in IOP seen in normal individuals is due to factors other than ACD; future studies that include factors that may influence IOP, such as central corneal thickness and axial length, are needed to confirm the correlation between IOP and ACD.

This study had some limitations. First, linear regression analysis cannot capture nonlinear associations. Additionally, no causal relationship between ACD and the findings of ocular examinations, hematological test values, or anthropometric data could be ascertained because this analysis was cross-sectional and no longitudinal analysis was performed. The study population was racially homogeneous, consisting only of Japanese participants, and the cohort consisted of participants who resided in a single city; therefore, the results may not be generalizable. In addition, medical history was obtained through a questionnaire, but history of prior therapeutic interventions (including medications) and minor surgeries could not be completely identified. Particularly, the absence of exclusion criteria based on medications may have influenced the results, although the sample size may have mitigated some of these effects.

In conclusion, ACD was found to correlate with several hematological findings such as FBS and anthropometric parameters such as age, sex, IOP, and SER. Age and SER may be particularly important in clinically predicting ACD < 2.70 mm in health checkup participants with good corrected VA in mainland Japan.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Author contributions

K. Y. and T. I. wrote the manuscript, acquired the data, and prepared the figures and tables. K. Y., T. I., A.S., T. K., T. M., J. O., R. T., Y. K., H. T., S. U., Y. I., and K. M. N. analyzed and interpreted the data. All the authors have read and approved the final version of the manuscript.

Data availability

The datasets analyzed for this article are included within the article and supplementary materials.

Declarations

Competing interests

The authors declare no competing interests.