Refractive Errors, Axial Ocular Dimensions, and Age-Related Cataracts: The Tanjong Pagar Survey

Abstract

purpose. To describe the relationship of refractive errors and axial ocular dimensions and age-related cataract.

methods. Population-based, cross-sectional survey of ocular diseases among Chinese men and women aged 40 to 81 years (n = 1232) living in the Tanjong Pagar district in Singapore. As part of the examination, refraction and corneal curvature were determined with an autorefractor, with refraction further refined subjectively. Ocular dimensions, including axial length, anterior chamber depth, lens thickness, and vitreous chamber depth, were measured with an A-mode ultrasound device. Lens opacity was graded clinically according to the Lens Opacity Classification System (LOCS) III system. Refraction, biometry, and cataract data on right (n = 989) and left (n = 995) eyes were analyzed separately.

results. In analyses controlling for age, gender, education, diabetes, and cigarette smoking, nuclear cataract was associated with myopia (−1.35 D vs. −0.11 D, P < 0.001, comparing right eyes with and without nuclear cataract), but not with any specific biometric component. Cortical cataract was associated with thinner lenses (4.67 mm vs. 4.79 mm, P = 0.001, comparing right eyes with and without cortical cataract), but not with refraction and other biometric components. Posterior subcapsular cataract was associated with myopia (−1.80 D vs. −0.39 D, P < 0.001, comparing right eyes with and without posterior subcapsular cataract), deeper anterior chamber (3.00 mm vs. 2.89 mm, P = 0.02), thinner lens (4.62 mm vs. 4.77 mm, P = 0.001), and longer vitreous chamber (15.78 mm vs. 15.57 mm, P = 0.09), but not with overall axial length and corneal curvature. Adjustment for vitreous chamber depth attenuated the association between posterior subcapsular cataract and myopia by 65.5%, but did not substantially change the association between nuclear cataract and myopia.

conclusions. These population-based data support the associations between nuclear and posterior subcapsular cataracts and myopia reported in previous studies. Posterior subcapsular cataract is also associated with deeper anterior chamber, thinner lens, and longer vitreous chamber, with vitreous chamber depth explaining most of the association between posterior subcapsular cataract and myopia.

The relationship between refractive errors and age-related cataract is not clear. ¹ Several large population-based studies have demonstrated a cross-sectional association between myopia and nuclear cataract in adults aged more than 40 years. ^{2 3 4 5} This association is widely believed to reflect increasing nuclear sclerosis of the lens with age, leading to a myopic shift in refraction. ⁶ In the Blue Mountains Eye Study, myopic refraction was also related to posterior subcapsular cataract, even when controlling for severity of nuclear sclerosis, indirectly suggesting that axial myopia may be related to posterior subcapsular cataract. ⁷ However, prospective data from the Beaver Dam Eye Study did not identify an association between myopia and 5-year incidence or progression of nuclear, cortical, or posterior subcapsular cataracts. ⁸ Instead, weak associations were found between hyperopia and nuclear and cortical cataracts. Results in other studies in smaller, selected populations have also been inconsistent. ^{9 10 11 12 13 14 15 16}  

However, these associations are difficult to interpret, because the final refractive state of an eye is dependent on the interaction between individual ocular biometric components (i.e., axial dimensions, corneal curvature, and lens power). ⁶ Thus, an alternative approach is to determine whether a particular type of cataract (e.g., posterior subcapsular) is related to a specific ocular component (e.g., vitreous chamber depth), rather than to the refractive status of that eye. 

The purpose of this present analysis was to describe the relationship between refractive errors, axial ocular dimensions, and age-related cataracts and specifically to determine whether the refractive associations of these cataract are axial (i.e., related to axial length or vitreous chamber depth) in nature. 

Methods

Study Population

The Tanjong Pagar Survey was a population-based, cross-sectional study of ocular disorders among adult Chinese persons in Singapore. The study was conducted between October 1997 and August 1998, with its population selection and methodology previously reported in detail. ^{5 17 18} In brief, the 1996 Singapore electoral register in the district of Tanjong Pagar was used as the sampling frame in this study. Because electoral registration is a legal requirement in Singapore, the register provides a complete record of all Singapore citizens aged 21 years and older. The electoral register listed 15,082 names of Chinese aged between 40 and 79 years residing in the district. Two thousand (13.3%) names were initially selected by a stratified, clustered, random sampling method, with more weights given to the older age groups. Among the 2000 names selected, 46 died and 235 moved to addresses outside the district before the study period and 2 people were excluded on grounds of ill health, leaving 1717 subjects considered eligible to participate in the study. These persons were invited for a comprehensive eye examination at the study clinic, after which an abbreviated home examination was conducted on nonrespondents. 

The total number of subjects examined was 1232 (71.8%), but only the 1090 (63.5%) subjects examined at the study clinic had an ocular biometry examination. Of these, 61 persons had cataract extraction, or data were missing for lens, refraction, or biometry in both eyes, leaving for this analysis 1029 persons with data in either eye (989 right eyes, and 995 left eyes). Refraction, biometry and cataract data on right and left eyes were analyzed separately. Table 1 shows the characteristics of participants included (n = 1029) compared with those excluded (n = 203). In general, subjects included were younger, had higher education levels and individual monthly income, lived in better housing, were less likely to have diabetes and hypertension, and were less likely to be cigarette smokers. 

Study Procedures

All examination procedures followed a written standardized protocol described elsewhere. ^{5 17 18} Only relevant portions are summarized in this report. The study was approved by the Ethics Committee of Singapore National Eye Center and was performed in accordance with the tenets of the World Medical Association’s Declaration of Helsinki. Visual acuity was first determined with the distance spectacle correction (if any) at initial examination, using the logarithm of minimum angle of resolution (logMAR) chart under standard lighting conditions at 4 m. ¹⁹ Refraction and corneal curvature were assessed with a handheld autorefractor-keratometer. For refraction, a single optometrist further performed a subjective refinement, the best corrected visual acuity was determined, and both the derived refraction and the visual acuity were recorded. ⁵ The autorefractor-keratometer reported eight separate estimates of corneal curvature radius along two meridians, each 90° apart. A mean value along each meridian was recorded, and the mean corneal curvature radius was calculated as the average of the greater and lesser radii of the curvature. Measurements of ocular dimensions, including axial length, anterior chamber depth, lens thickness, and vitreous chamber depth, were obtained with a 10-MHz A-mode ultrasound device. ¹⁸ The hard-tipped, corneal contact ultrasound probe was mounted on a tonometer set to the individual’s intraocular pressure, and the mean of 16 separate readings was recorded. 

Cataract was determined clinically using the Lens Opacity Classification System (LOCS) III system. ²⁰ The procedure for assessment of cataract in the present study has been described. ²¹ After dilation of pupils with tropicamide 1% and phenylephrine hydrochloride 2.5% eye drops (application repeated twice if necessary), the participant was examined at a slit lamp by the study ophthalmologist (PJF), and the presence and severity of a specific lens opacity were compared and documented according to LOCS III standard photographs. 

Trained study personnel, masked to the participants’ refraction, biometry, and cataract status, ascertained information on demographics, education, medical history, and other variables from a standardized interview. 

Definitions

Data on refraction were converted to spherical equivalent diopters and were based on subjective refraction when participants had both subjective and objective refraction, and on objective refraction when only this information was available. ⁵ In this study, refraction was categorized as follows: high myopia as a spherical equivalent of less than −6.00 D, moderate myopia between −5.99 and −3.00 D, mild myopia between −2.99 and −0.51 D, emmetropia between −0.50 and +0.50 D, and hyperopia as greater than +0.50 D. Biometry data were divided into quintiles for analysis, with the 1st quintile representing the lowest 20% of the population for that parameter. 

LOCS III includes an assessment of nuclear opalescence (NO), nuclear color (NC), cortical cataract (C), and posterior subcapsular cataract (P). ²⁰ For analysis, a LOCS III score of 4.0 or more for NO or 4.0 or more for NC was defined as significant nuclear cataract, a score of 2.0 or more for C as significant cortical cataract, and a score of 2.0 or more for P as significant posterior subcapsular cataract. ²¹ Definitions were based on similar criteria published elsewhere. ²²  

Age was the current age at the time of examination. Education was ascertained by the question, “What was your highest education level?” and categorized into four groups: no formal education, primary (6 years or less), secondary (7–10 years), and tertiary (11 years or more, including university education). Housing type was recorded to one of four groups: one- or two-room government flats, three-room government flats, four- or five-room government flats, and executive government flats or private housing. Individual monthly income were ascertained in Singapore dollars (approximate exchange rate of Sing$1.7 = US$1) and categorized as $1000 or less, $1001 to $2000, $2001 to $3000, more than $3000, and not currently working (retired). Diabetes and hypertension were ascertained by asking, “Have you been told by a doctor that you have diabetes (hypertension)?” followed by further questions on treatment. Diabetes was classified as diabetes treated with oral diabetic medications and/or insulin injection, diabetes treated by diet only, and no diabetes. Hypertension was classified as present (yes) or not (no). A history of current cigarette smoking was ascertained by asking, “Do you smoke regularly (at least once a week)?” and categorized as yes or no. 

Statistical Analysis

Data on right and left eyes were analyzed separately. This approach is statistically valid, easy to interpret, and does not result in substantial loss of power when the correlation between eyes for the parameters concerned are high (e.g., correlation between eyes for spherical equivalent refraction and nuclear cataract are 0.83 and 0.90, respectively). ²³ Because the results in the left eyes were similar, most of the data presented are based on the right eye. 

We calculated the mean refraction, axial biometric components, and corneal curvature radius, in the presence versus the absence of nuclear, cortical, and posterior subcapsular cataracts, using analysis of covariance to adjust first for age and gender and then further for education, diabetes, and cigarette smoking. The latter variables have been found to be associated with refractive error (education) and cataract (diabetes and smoking). Multiple logistic regression was used to determine the effects of categories of refraction or quintiles of specific biometric components on the odds of each type of cataract, adjusting similarly for age, gender, education, diabetes, and cigarette smoking. Finally, axial biometric components (e.g., axial length, vitreous chamber depths) were entered into analysis of covariance models to determine their effects on the difference in mean refraction between eyes with and without cataract. The relative effect (%) of these components was defined as [(Difference in means in the reference model - Difference in means in models with the specific biometric components added)/Difference in means in the reference model]. The reference model adjusted for age, gender, education, diabetes, cigarette smoking, and corneal curvature radius (corneal curvature is correlated strongly with refraction). Analyses of all data were performed on computer (SPSS, ver. 9.0; SPSS Science Inc., Chicago, IL). 

Results

Among the 989 people with data on the right eye, 338 (34.2%) had significant nuclear cataract, 324 (29.1%) significant cortical cataract, and 98 (9.9%) significant posterior subcapsular cataract. The mean refraction was −0.51 ± 2.67 D (SD), distributed as follows: 48 (4.9%) had high myopia, 85 (8.6%) moderate myopia, 207 (20.9%) mild myopia, 301 (30.4%) emmetropia, and 348 (35.2%) hyperopia. 

Table 2 shows mean refractions, axial biometric components, and corneal curvature radius in the right eye, in the presence and absence of nuclear, cortical, and posterior subcapsular cataracts. When analysis was made controlling for age, gender, education, diabetes, and cigarette smoking, nuclear cataract was associated with myopic refraction (−1.35 D vs. −0.11 D, P < 0.001, comparing right eyes with and without nuclear cataract), but not with any specific biometric component. Cortical cataract was associated with thinner lens (4.67 mm vs. 4.79 mm, P = 0.001), but not with refraction and other biometric components. Posterior subcapsular cataract was associated with myopic refraction (−1.80 D vs. −0.39 D, P < 0.001), deeper anterior chamber (3.00 mm vs. 2.89 mm, P = 0.02), thinner lenses (4.62 mm vs. 4.77 mm, P = 0.001), and longer vitreous chamber (15.78 mm vs. 15.57 mm, P = 0.09), but not with axial length and corneal curvature. 

Table 3 shows the distribution and multivariate-adjusted odds ratio of specific cataracts in the right eye by categories of refraction and quintiles of ocular biometry and corneal curvature, with the lowest category or quintile serving as the reference. Similarly, nuclear cataract was associated with a myopic refraction (see odds ratios in Table 3 ; test of trend; P < 0.001) but not with axial biometric components or corneal curvature. Cortical cataract was not related to refractive errors, but was associated with thinner lenses (P = 0.003) and flatter corneas (P = 0.02). Posterior subcapsular cataract correlated with myopic refraction (P < 0.001), deeper anterior chamber (P = 0.008), thinner lens (P < 0.001), and a tendency toward a longer vitreous chamber (P = 0.07). 

Logistic regression models were also constructed, with refraction, biometry, and corneal curvature entered as continuous variables. In general, the pattern of associations was similar (data not shown). 

To determine the nature of the associations of myopia with nuclear and posterior subcapsular cataracts, axial biometric parameters (e.g., anterior chamber depths, vitreous chamber depths) were entered in a stepwise fashion in analysis of covariance models (Table 4) . The difference in mean refraction between eyes, with and without nuclear (or posterior subcapsular) cataract, was compared between models with a biometric component entered (models 2–8) versus the reference model without biometric components (model 1). The relative effect of a specific component (e.g., vitreous chamber depth) estimates the amount of attenuation in the association between myopia and cataract. For example, an interpretation of relative effect of vitreous chamber depth in model 4 is as follows: Vitreous chamber depth explained 16.4% of the mean difference in refraction between eyes with and without nuclear cataract, but explained 65.5% of the mean difference in refraction between eyes with and without posterior subcapsular cataract. In general, adding axial length (or its individual components) attenuated the difference in mean refraction between eyes with and without nuclear cataract by a maximum of 17.2%. In contrast, vitreous chamber depth accounted for more than 60% (model 4), whereas nuclear opacity accounted for only 21.6% (model 6) of the difference in mean refraction between eyes with and without posterior subcapsular cataract. Simultaneous adjustment for both vitreous chamber depth and nuclear opacity significantly attenuated the association between myopia and posterior subcapsular cataract (model 7, mean difference between eyes with and without posterior subcapsular cataract −0.23 D, P = 0.20). 

Analyses were repeated with alternative cutoffs points used for cataract definition. We found the pattern of associations to be similar (data not shown). For example, a myopic refraction was associated with nuclear cataract by using definitions of LOCS III score of 3.0 or more for NO or 3.0 or more for NC 3 (OR of hyperopia [reference], emmetropia and mild, moderate and high myopia were 1.0, 1.3, 1.6, 1.9, and 2.4, test of trend; P < 0.001) or LOCS III score of 5.0 or more for NO or 5.0 or more for NC 3 (OR: 1.0, 1.2, 5.9, 18.8, and 21.0, test of trend; P < 0.001). 

In general, analysis of left eye data (n = 995) showed a similar pattern of associations. Nuclear and posterior subcapsular cataracts correlated strongly with myopic refraction (OR of hyperopia [reference], emmetropia and mild, moderate, and high myopia were 1.0, 1.4, 2.2, 2.8, and 7.9 for nuclear cataract and 1.0, 1.9, 1.5, 2.3, and 3.9 for posterior subcapsular cataract; test of trend; P < 0.001 for both) but cortical cataract was not related to refractive errors (OR: 1.0, 1.1, 1.1, 0.4, and 0.9, P = 0.11). 

Discussion

Refractive errors and age-related cataract are frequent ocular conditions in adults worldwide. Approximately a third of the population aged more than 40 years has myopia, ^{2 3 4 5} whereas age-related cataract is the most common cause of visual impairment in the elderly. ²⁴ The relationship between refractive errors and age-related cataracts is not clear, despite several clinic-based, ^{9 10 11 12 13 14 15 16} and population-based investigations. ^{7 8} Although these inconsistencies may be the result of variations in populations, methodology, and definitions (e.g., a history of spectacle use as an indicator of myopia ¹⁶ ), this may also be because refraction is an imperfect summary measure of different biometric components of the eye. 

Our study provides cross-sectional, population-based data, not only on the association between refraction and cataracts but, as important, on the association between axial ocular dimensions and age-related cataracts. After controlling for age, gender, and cataract risk factors (diabetes, smoking, and education), nuclear and posterior subcapsular cataracts, but not cortical cataract, were associated with a myopic refraction. Nuclear cataract was not further associated with any specific biometric component, and axial biometric components do not appear to explain its refractive association (models 2–5, Table 4 ). In contrast, posterior subcapsular cataract was associated with deeper anterior chamber, thinner lens, and a longer vitreous chamber. Variation in vitreous chamber depth explained more than 60% of the difference in mean refraction between eyes with and without posterior subcapsular cataract (model 4, Table 4 ). In contrast, variation in nuclear sclerosis accounted for only 20% of the difference in mean refraction between eyes with and without posterior subcapsular cataract (model 6, Table 4 ). This pattern suggests that the refractive association of posterior subcapsular cataract is axial. Finally, although not related to refractive errors, cortical cataract was associated with thinner lenses. The significance of these findings will be described in turn. 

The cross-sectional association between nuclear cataract and myopia has been demonstrated in several population-based studies among adults of different ethnicities. ^{2 3 4 5} This association has been hypothesized by the investigators to reflect increasing nuclear sclerosis of the lens with age, causing a myopic shift in refraction (i.e., index myopia). Consistent with this hypothesis, the Beaver Dam Eye Study showed no prospective relationship between a myopic refraction and 5-year risk of development of nuclear cataract. ⁸ In addition, in the same cohort, eyes with severe nuclear sclerosis at baseline were more likely to have a myopic change in refraction after 10 years, compared with a hyperopic change in eyes with only mild nuclear sclerosis. ²⁵ Our findings that nuclear cataract was associated with myopia but not with axial ocular dimensions and that neither axial length nor vitreous chamber depth explains the refractive association of nuclear cataract provide evidence to support the index-myopia hypothesis regarding nuclear cataract. 

The relationship between myopia and posterior subcapsular cataract is controversial. As for nuclear cataract, cross-sectional associations between myopia and posterior subcapsular cataracts have been observed in similar populations. ^{5 7 26} Unlike nuclear cataract, however, posterior subcapsular cataract does not appreciably affect refraction, and it has therefore been suggested that this relationship may be causal (i.e., myopia may be a risk factor for development of posterior subcapsular cataract). This premise is supported by findings from the Blue Mountains Eye Study, in which a myopic refraction and early-onset myopia (defined as a history of wearing spectacles for distance before age of 20 years) were related to increased odds of posterior subcapsular cataract, despite adjusting for nuclear sclerosis. ⁷ Our study now suggests that posterior subcapsular cataract is also related to a deeper anterior chamber, thinner lens, and a longer vitreous chamber, and that adjusting for these components, in particular vitreous chamber depth, attenuates the association of posterior subcapsular cataract with myopic substantially. Anterior chamber depth, lens thickness, and vitreous chamber depth correlated highly and, because our data are cross sectional, it is not possible to distinguish which of these factors were more important or occurred first. For example, the initial hypothesis could be that eyes with longer vitreous chambers are more likely to have development of posterior subcapsular cataract. ⁷ These eyes would in turn have thinner lenses, resulting from a compensatory mechanism to maintain emmetropia. ^{27 28} However, an alternative explanation is that lenses with posterior subcapsular cataract are thinner to begin with, ^{29 30 31 32} either because of a decreased rate of lens fiber formation ²⁹ or because of leakage of lens protein. ^{33 34} Thinner lenses in turn would be associated with relatively longer anterior and vitreous chambers, thus explaining our findings. Future longitudinal data with biometric measurements would help clarify the picture. 

Finally, our study supports the findings of previous studies that cortical cataract is not related to refractive errors, either cross sectionally ^{7 26} or longitudinally. ⁸ Our finding that cortical cataract was associated with a thinner lens has been noted in Beaver Dam ³⁵ and other non-population-based studies. ^{29 30} As for posterior subcapsular cataract, the underlying biological explanation remains unknown. ^{30 33 34}  

Limitations of this study should be considered. First, as noted already, in a cross-sectional study, we cannot infer a temporal relationship for any of these factors. Second, selection biases must be considered, in that some people were excluded because they did not have a biometric examination or had had severe cataract and had undergone cataract surgery. For example, the myopic associations of cataract could be explained if people with hyperopia were more likely to have cataract surgery and therefore to be excluded from analyses. However, data from the Beaver Dam Eye Study indicate the reverse appears to more likely, because eyes with myopia were more likely to undergo cataract surgery. ⁸ Third, our inability to adjust for other unmeasured potential confounders (e.g., ultraviolet light exposure) may have masked some associations and accentuated others. These are not expected to be significant, as the multivariate ORs were little changed from the age- and sex-adjusted OR, when we further controlled for smoking, diabetes, and education (see Table 2 ). Fourth, the relatively lower prevalence of posterior subcapsular cataract may have reduced the study’s power to identify significant relationships. Finally, our findings may not be applicable to Western populations with a lower prevalence of refractive errors ^{2 3 4} and eyes that are biometrically different from Chinese eyes. ³⁶  

In summary, our study shows that myopia is associated with nuclear and posterior subcapsular, but not cortical, cataract in adult Chinese people aged 40 to 81 years. Nuclear cataract was not associated with any specific biometric component and is therefore probably explicable on the grounds of increasing refractive index of the lens (index myopia). In contrast, eyes with posterior subcapsular cataract were more likely to have deeper anterior chambers, thinner lenses, and longer vitreous chambers. Controlling for vitreous chamber depth attenuated most of the association of myopia with posterior subcapsular cataract, suggesting that the refractive association of this form of cataract is axial. 

 Supported by the National Medical Research Council, Singapore and the British Council for the Prevention of Blindness, London, United Kingdom.

 Submitted for publication May 31, 2002; revised October 7, 2002; accepted October 15, 2002.

 Commercial relationships policy: N.

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

Table 1.

 

 

Comparison of Subjects Included and Excluded from Right Eye Analyses

Table 1.

 

Comparison of Subjects Included and Excluded from Right Eye Analyses

		Included (n = 1029)			Excluded (n = 203)			P
				n	%	n	%
Age (y)	40–49	259	25.2	17	8.4	<0.001
	50–59	229	27.1	27	13.5
	60–69	282	27.4	61	30.3
	70–81	209	20.3	98	48.3
Gender	Male	469	45.6	88	43.3	0.56
	Female	560	54.4	115	56.7
Education*	No education	252	24.6	74	38.1	<0.001
	Primary (1–6 y)	417	42.8	81	41.8
	Secondary (7–10 y)	276	27.0	34	17.5
	Tertiary (11 y or more)	78	6.5	5	2.6
Individual monthly income*	Less than $1000	613	59.9	166	84.3	<0.001
	$1000–2000	182	17.8	11	5.6
	$2000–3000	64	6.3	1	0.5
	More than $3000	51	5.0	0	—
	Retired	103	10.1	17	8.6
Housing type*	1–2 room flats	175	17.2	60	30.5	<0.001
	3 room flats	557	54.7	101	51.3
	4–5 room flats	265	26.0	34	17.3
	Executive flats or private housing	23	2.2	2	1.0
Diabetes	Yes	108	10.5	32	16.1	0.06
	No	919	89.5	166	83.8
Hypertension	Yes	290	28.3	69	34.8	0.03
	No	737	71.8	129	65.2
Cigarette smoking	Yes	191	18.6	29	14.6	0.06
	No	836	81.4	169	85.3

Table 2.

 

 

Adjusted Means of Refraction and Ocular Biometric Components by Cataract Status in the Right Eye

Table 2.

 

Adjusted Means of Refraction and Ocular Biometric Components by Cataract Status in the Right Eye

		Nuclear Cataract					Cortical Cataract					Posterior Subcapsular Cataract
				Present (n = 338)	Absent (n = 650)	P	Present (n = 288)	Absent (n = 701)	P	Present (n = 98)	Absent (n = 890)	P
Refraction (D)	Age and gender*	−1.29 ± 0.17	−0.10 ± 0.11	<0.001	−0.59 ± 0.16	−0.48 ± 0.10	0.59	−1.78 ± 0.27	−0.37 ± 0.09	<0.001
	Multivariate, †	−1.35 ± 0.17	−0.11 ± 0.11	<0.001	−0.69 ± 0.17	−0.46 ± 0.10	0.28	−1.80 ± 0.27	−0.39 ± 0.09	<0.001
Axial length, (mm)	Age and gender*	23.23 ± 0.07	23.24 ± 0.05	0.90	23.20 ± 0.07	23.26 ± 0.04	0.56	23.38 ± 0.12	23.23 ± 0.04	0.22
	Multivariate, †	23.25 ± 0.07	23.24 ± 0.05	0.98	23.25 ± 0.07	23.25 ± 0.04	0.97	23.29 ± 0.12	23.23 ± 0.04	0.23
Anterior chamber depth (mm)	Age and gender*	2.88 ± 0.07	2.91 ± 0.02	0.39	2.88 ± 0.03	2.90 ± 0.02	0.39	3.01 ± 0.04	2.89 ± 0.01	0.007
	Multivariate, †	2.88 ± 0.03	2.91 ± 0.02	0.42	2.89 ± 0.03	2.91 ± 0.02	0.66	3.00 ± 0.04	2.89 ± 0.01	0.02
Lens thickness (mm)	Age and gender*	4.80 ± 0.07	4.74 ± 0.02	0.13	4.69 ± 0.03	4.79 ± 0.02	0.003	4.64 ± 0.04	4.77 ± 0.01	0.003
	Multivariate, †	4.79 ± 0.03	4.74 ± 0.02	0.17	4.67 ± 0.03	4.79 ± 0.02	0.001	4.62 ± 0.05	4.77 ± 0.001	0.001
Vitreous chamber depth (mm)	Age and gender*	15.56 ± 0.05	15.60 ± 0.05	0.72	15.64 ± 0.07	15.57 ± 0.04	0.42	15.75 ± 0.11	15.57 ± 0.04	0.13
	Multivariate, †	15.58 ± 0.07	15.60 ± 0.05	0.86	15.68 ± 0.07	15.56 ± 0.04	0.14	15.78 ± 0.12	15.57 ± 0.04	0.09
Corneal radius (mm)	Age and gender*	7.64 ± 0.02	7.67 ± 0.01	0.19	7.63 ± 0.02	7.67 ± 0.01	0.07	7.65 ± 0.03	7.66 ± 0.01	0.88
	Multivariate, †	7.64 ± 0.02	7.67 ± 0.01	0.18	7.63 ± 0.02	7.67 ± 0.01	0.03	7.65 ± 0.03	7.66 ± 0.01	0.89

Table 3.

 

 

Multivariate-Adjusted Odds Ratio of Nuclear, Cortical, and Posterior Subcapsular Cataract, by Refraction and Ocular Biometric Components in the Right Eye

Table 3.

 

Multivariate-Adjusted Odds Ratio of Nuclear, Cortical, and Posterior Subcapsular Cataract, by Refraction and Ocular Biometric Components in the Right Eye

	Categories (Range)	Eyes at Risk (n)	Nuclear Cataract					Cortical Cataract					Posterior Subcapsular Cataract
						n	%	OR (95% CI)	n	%	OR (95% CI)	n	%	OR (95% CI)
Refraction (D)	Hyperopia	348	126	36.2	1.0	117	33.6	1.0	25	7.2	1.0
	Emmetropia	301	85	28.2	1.1 (0.7, 1.8)	82	27.2	1.1 (0.7, 1.6)	30	10.0	1.9 (1.0, 3.4)
	Low myopia	207	84	40.6	2.6 (1.5, 4.3)	6	29.5	1.1 (0.7, 1.8)	22	10.6	1.7 (0.9, 3.3)
	Moderate myopia	84	26	31.0	2.2 (1.0, 4.8)	17	20.0	0.8 (0.4, 1.6)	15	17.6	4.7 (2.2, 10.3)
	High myopia	48	17	35.4	7.1 (2.8, 18.4)	11	22.9	1.7 (0.7, 3.9)	6	12.5	4.2 (1.4, 11.9)
					P < 0.001, †			P = 0.56, †			P < 0.001, †
Axial length (mm)	1st (20.6–22.3)	189	72	38.1	1.0	69	36.5	1.0	12	6.3	1.0
	2nd (22.4–22.8)	189	80	42.3	1.4 (0.8, 2.4)	60	31.7	0.8 (0.5, 1.3)	21	11.1	2.0 (0.9, 4.4)
	3rd (22.9–23.4)	191	66	34.6	0.9 (0.5, 1.6)	54	28.3	0.7 (0.4, 1.2)	23	12.1	2.3 (1.0, 5.2)
	4th (23.5–24.0)	188	55	29.3	0.9 (0.5, 1.7)	49	26.1	0.7 (0.4, 1.1)	21	11.2	2.5 (1.1, 5.6)
	5th (24.1–28.6)	189	49	25.9	1.0 (0.5, 1.9)	40	21.1	0.7 (0.4, 1.3)	13	6.8	1.6 (0.6, 4.0)
					P = 0.59, †			P = 0.19, †			P = 0.24, †
Anterior chamber depth (mm)	1st (1.7–2.4)	190	101	53.2	1.0	92	48.4	1.0	22	11.6	1.0
	2nd (2.5–2.7)	191	80	41.9	0.8 (0.5, 1.4)	59	30.9	0.5 (0.3, 0.9)	18	9.4	1.0 (0.5, 2.1)
	3rd (2.8–3.0)	183	62	33.9	1.0 (0.6, 1.8)	40	21.9	0.5 (0.3, 0.9)	17	9.3	1.3 (0.8, 3.3)
	4th (3.1–3.3)	194	41	21.1	0.7 (0.4, 1.2)	49	25.3	0.9 (0.5, 1.5)	12	6.2	1.4 (0.6, 3.0)
	5th (3.4–4.3)	187	37	19.8	0.9 (0.5, 1.7)	32	17.0	0.7 (0.4, 1.2)	21	11.2	3.1 (1.4, 6.9)
					P = 0.47, †			P = 0.42, †			P = 0.008, †
Lens thickness (mm)	1st (3.4–4.2)	186	25	13.4	1.0	34	18.2	1.0	17	9.1	1.0
	2nd (4.3–4.5)	194	31	16.0	0.8 (0.4, 1.7)	37	19.1	0.8 (0.4, 1.4)	13	6.7	0.5 (0.2, 1.2)
	3rd (4.6–4.8)	183	59	32.2	0.9 (0.5, 1.7)	57	31.1	0.7 (0.4, 1.2)	16	6.7	0.3 (0.1, 0.7)
	4th (4.9–5.1)	189	88	46.6	1.3 (0.7, 2.5)	65	34.4	0.5 (0.3, 0.9)	21	11.1	0.4 (0.2, 0.8)
	5th (5.2–6.9)	192	117	60.9	1.4 (0.7, 2.6)	78	40.6	0.4 (0.2, 0.8)	22	11.5	0.2 (0.1, 0.5)
					P = 0.11, †			P = 0.003, †			P < 0.001, †
Vitreous chamber depth (mm)	1st (12.8–14.6)	187	77	41.2	1.0	62	33.2	1.0	13	7.0	1.0
	2nd (14.7–15.2)	189	84	44.4	1.2 (0.7, 2.1)	64	33.9	1.1 (0.7, 1.8)	20	10.6	1.8 (0.8, 4.1)
	3rd (15.3–15.7)	190	61	32.1	0.8 (0.5, 1.4)	56	29.5	1.0 (0.6, 1.7)	24	12.6	2.9 (1.3, 6.4)
	4th (15.8–16.3)	191	53	27.7	0.8 (0.4, 1.4)	53	27.7	1.1 (0.6, 1.8)	19	10.0	2.3 (1.0, 5.3)
	5th (16.4–21.1)	187	44	23.7	1.0 (0.5, 1.8)	36	19.3	1.0 (0.6, 1.8)	13	7.0	2.1 (0.9, 5.2)
					P = 0.47, †			P = 0.87, †			P = 0.07, †
Corneal curvature radius (mm)	1st (6.8–7.4)	201	68	33.8	1.0	70	34.8	1.0	19	9.5	1.0
	2nd (7.5–7.6)	197	68	34.5	1.3 (0.8, 2.3)	53	26.9	0.7 (0.4, 1.1)	19	9.6	1.2 (0.6, 2.6)
	3rd (7.6–7.7)	197	74	37.6	1.6 (0.9, 2.9)	58	29.4	0.7 (0.4, 1.2)	22	11.6	1.4 (0.7, 2.9)
	4th (7.7–7.8)	195	66	33.8	1.0 (0.5, 1.8)	55	28.2	0.6 (0.4, 1.0)	20	10.3	1.1 (0.5, 2.2)
	5th (7.9–9.6)	193	61	31.6	0.8 (0.5, 1.5)	52	26.8	0.5 (0.3, 0.9)	18	9.3	1.1 (0.5, 2.3)
					P = 0.32, †			P = 0.02, †			P = 0.99, †

Table 4.

 

 

Difference in Mean Refraction between Eyes with and without Cataract, Adjusted for Axial Ocular Biometric Components in the Right Eye

Table 4.

 

Difference in Mean Refraction between Eyes with and without Cataract, Adjusted for Axial Ocular Biometric Components in the Right Eye

Models (Additional Parameters)*	Mean Refraction (D), by Nuclear Cataract Status									Mean Refraction (D), by Posterior Subcapsular Cataract Status
		Present	Absent	Difference in Means	P†	Relative Effect* (%)	Present	Absent	Difference in Means	P†	Relative Effect* (%)
1 (Reference)	−1.31	−0.09	−1.22	<0.001	Reference	−1.76	−0.37	−1.39	<0.001	Reference
2 (ACD)	−1.24	−0.10	−1.14	<0.001	6.6	−1.25	−0.41	−0.84	0.002	39.6
3 (LT)	−1.22	−0.11	−1.11	<0.001	9.0	−1.33	−0.40	−0.93	0.001	33.0
4 (VCD)	−1.16	−0.14	−1.02	<0.001	16.4	−0.92	−0.44	−0.48	0.01	65.5
5 (AL)	−1.15	−0.14	−1.01	<0.001	17.2	−1.04	−0.43	−0.61	0.001	56.1
6 (NO)	—	—	—	—	—	−1.49	−0.40	−1.09	<0.001	21.6
7 (VCD, NO)	—	—	—	—	—	−0.70	−0.46	−0.24	0.20	82.7
8 (AL, NO)	—	—	—	—	—	−0.80	−0.45	−0.35	0.06	74.8