Variations in Ocular Biometry in an Adult Chinese Population in Singapore: The Tanjong Pagar Survey

Abstract

purpose. To describe the variation in ocular biometry in adult Chinese individuals in Singapore.

methods. This study was a population-based, cross-sectional survey of adult Chinese persons aged 40 to 81 years residing in Tanjong Pagar district, Singapore. Axial ocular dimensions, including axial length (AL), anterior chamber depth (ACD), lens thickness (LT), and vitreous chamber depth (VCD) were measured using an A-scan ultrasound device. Corneal curvature (CC) and noncycloplegic refraction were measured with an autorefractor, with refraction further refined subjectively. Lens nuclear opacity (NO) was graded clinically using the modified Lens Opacity Classification System III (LOCS III) score.

results. A total of 1717 subjects were eligible for the survey, of whom 1232 (71.8%) participated. Biometric and refraction data were available for 1004 (58.5%) phakic subjects. The AL, ACD, LT, VCD, CC, and LOCS III scores were 23.23 ± 1.17 mm, 2.90 ± 0.44 mm, 4.75 ± 0.47 mm, 15.58 ± 1.11 mm, 7.65 ± 0.27 mm, and 3.2 ± 0.9 (mean ± SD), respectively. On average, people aged 40 to 49 years, when compared with those 70 to 81 years, had longer ALs (mean difference, +0.58 mm), deeper ACDs (+0.52 mm), longer VCDs (+0.72 mm), but thinner lenses (−0.70 mm) and less severe NO (−1.7 LOCS III score). CCs did not vary significantly with age. After controlling for age, women had shorter ALs and VCDs, shallower ACDs, but thicker lenses and steeper CCs than men. The variation in noncycloplegic refraction with age was nonlinear. Among people aged 40 to 59 years, a higher prevalence of hyperopia was seen in older compared with younger persons (on average, a difference of +1.3 D for every 10-year difference in age, P < 0.001), explained principally by shorter AL (and VCD) in older persons. Among those 60 to 81 years, this pattern was not obvious (a difference of −0.03 D for every 10-year difference in age, P = 0.12), as NO became an additional determinant of refraction, with greater degrees of NO in older person’s driving refraction in the “minus” direction.

conclusions. Ocular dimensions vary with age and gender in adult Chinese persons in Singapore. The variation in noncycloplegic refraction in people 40 years and older may be explained by differences in axial lengths (principally vitreous chamber depths) between older and younger persons and, from 60 years onwards, differences in lens nuclear opacification as well.

The variation in refraction in adults has been described in several large population-based, cross-sectional studies. In general, these studies have shown that after the age of 40 years, there was a cross-sectional trend in refraction in the “plus” direction, with older persons having higher prevalence of hyperopia and correspondingly lower prevalence of myopia, compared with younger persons. ^{1 2 3 4 5} After 70 to 80 years, this pattern was less distinct, ^{1 2 3 4 5} with some studies indicating a trend in the“ minus” direction, with further increases in age. ^{1 2} The explanation for these observations is not clear and appears to be related to either a cohort effect (differences in rates of myopia and hyperopia between younger and older generations) ^{6 7 8} or to actual changes in refraction with increasing age. ⁹  

Biometric data (e.g., axial length, corneal curvature, lenticular power) may help explain the variation in refraction observed. In children, adolescents and young adults, for example, biometric studies have shown that differences in axial lengths (principally vitreous chamber depths) account for most of the variation in refraction in the young. ^{10 11 12 13 14 15 16 17 18 19} Likewise, differences in axial lengths have been hypothesized to explain the variation in refraction observed in older adults. ⁶ However, biometric data are lacking from current population-based studies to support this hypothesis. ^{1 2 3 4 5}  

In a recent survey in adult Chinese persons 40 years and older residing in Singapore, we observed a similar cross-sectional pattern of refraction as in other studies. ^{1 2 3 4 5 20} We now describe the variation in ocular dimensions in this population and evaluate the association between individual components and refraction. 

Methods

Singapore is an urban city-state with a population of 3.2 million, 78% of whom are ethnic Chinese. ²¹ Because a large number of Chinese residents in our study age group are first-generation immigrants from the southern provinces of Fujian and Guandong in China, some generalizability of our data to the population in China may be possible. 

The study was part of a population-based survey of ocular disorders among adult Chinese living in Singapore. Detailed population selection and methodology have been previously reported. ^{20 22} In brief, the 1996 Singapore electoral register in the district of Tanjong Pagar was used as the sampling frame in this study. Tanjong Pagar is located in the center of Singapore and was chosen because the population demographics of the Chinese residents are representative of the rest of Singapore. The electoral register listed 15,082 names of Chinese persons between the ages of 40 to 79 years residing in the district. Two thousand (13.3%) names were selected using a disproportionate (with more weights given to the older age groups), stratified, clustered, random sampling method. These persons were invited for a comprehensive eye examination at the study center. After this, an abbreviated domiciliary examination on nonrespondents was conducted. This study was approved by the ethics committee of Singapore National Eye Center and carried out in accordance with the tenets of the World Medical Association’s Declaration of Helsinki. Written informed consent was obtained, and the study was conducted between October 1997 and August 1998. 

Procedures

Measurements of axial length (AL), anterior chamber depth (ACD), lens thickness (LT) and vitreous chamber depth (VCD) were obtained using a 10-MHz A-mode ultrasound device (Storz Compuscan; Storz, St. Louis, MO). The hard-tipped, corneal contact ultrasound probe was mounted on a tonometer (Haag-Streit, Bern, Switzerland) set to the individual’s intraocular pressure. The mean of 16 separate readings was recorded, together with the SD of each parameter. An SD for axial length measurement of less than or equal to 0.13 mm was required. If the SD was greater, the reading was repeated up to another two times. If it was not possible to achieve an SD within these limits, the data were still accepted but were not analyzed in this report. 

Corneal curvature was assessed using a handheld autorefractor/keratometer (Retinomax K-plus; Nikon, Tokyo, Japan). The device recorded up to eight separate estimates of corneal curvature along two meridia, each 90° apart. A mean value along each meridian was recorded, and the mean corneal curvature (CC) was calculated as the average of the greater and lesser curvature. Lens nuclear opacity (NO) was graded at a slitlamp using the modified Lens Opacity Classification System III (LOCS III) score. ²³ Under this system, nuclear opacity was classified in increasing severity as grades 1 to 6. 

Noncycloplegic refraction was performed as follows. First, objective refraction was assessed with the same handheld autorefractor used to measure corneal curvature (Retinomax K-plus; Nikon). A single optometrist then performed a subjective refinement of the refractive correction with a phoropter, using the results of the objective refraction as the starting point. 

Definitions and Analysis

Biometric data for right and left eyes were analyzed separately. Because the results between the two eyes were similar, only data on the right eye are presented in this study. Noncycloplegic refraction data were converted to spherical equivalents (SE) and were based on subjective refraction when participants had both subjective and objective refraction and on objective refraction when only this information was available. Preliminary analysis indicated high overall agreement between objective and subjective refraction in SE. ²⁰  

The data analysis was conducted as follows. First, the variation in individual biometric components was analyzed across the entire age spectrum (40–81 years). Next, the participants were divided into two age groups (40–59 and 60–81 years), and the variation in biometric components within each group was analyzed separately. This was done because our previous study showed refraction was nonlinearly associated with age. ²⁰ Before 60 years, there was a trend toward“ plus” SE in older compared with younger persons, but after 60 years, this pattern was less distinct (see Results and Fig. 7 ). Within each group, linear regressions were performed to assess the effect of age and gender (independent variables) on the variations in individual biometric components and refraction (dependent variables). Linear regression models were then constructed to evaluate the independent effects of different biometric components (independent variables) on refraction (dependent variable). Standardized regression coefficients in these models were used to determine the relative role/importance of each biometric component on refraction. ²⁴ Statistical analyses of the data were carried out using SPSS (SPSS Inc., Chicago, IL). 

Results

Among the 2000 names selected from the sampling frame, 46 persons had died and 235 had moved to addresses outside the Tanjong Pagar district before or during the study period. Two persons were considered unfit to ever undergo examination, leaving 1717 persons who were invited for this study. The total number of persons who participated in this study was 1232 (71.8%), and they were examined in either the research clinic or in their homes. Only the 1090 persons (63.5%) who were examined at the research clinic were considered eligible for this report, because those in the home setting did not receive a biometric examination. Of the 1090 persons, 80 had prior cataract extraction in their right eye, and 4 had corneal scarring, dense cataracts, or other media opacities that prevented reliable biometry or refraction in their right eye. Another 2 persons had no biometry or refraction data available, leaving 1004 (58.5%) for this analysis. Table 1 shows the demographic characteristics of the 1004 subjects, compared with the 228 excluded from this analysis. In general, subjects included in our analysis (our study population) were younger; more likely to be professionals and office workers, production operators, or salespeople; lived in better housing; and had higher education levels and individual income. 

Table 2 shows the mean biometric components and noncycloplegic refraction in the study population. Gender variation was present among a number of biometric parameters. After controlling for age, women had significantly shorter ALs, explained principally by shorter VCDs and shallower ACDs. Women also had significantly steeper CCs and thicker lens. 

The variation of different biometric components and noncycloplegic refraction with age is shown in Table 3 and depicted in Figures 1 2 3 4 5 6 7 . On average, persons aged 40 to 49 years, when compared with those 70 to 81 years, had longer ALs (mean difference, 0.58 mm), longer VCDs (mean difference, 0.72 mm), and deeper ACDs (mean difference, 0.52 mm). However, persons aged 70 to 81 years had thicker lenses (mean difference, −0.70 mm) and greater NOs (mean difference, −1.7 LOCS III score). There was no significant variation in corneal CC between young and old. From 40 to 60 years, older persons had more “plus” SE refraction than younger persons (Fig. 7) , but from 60 years onward, the difference in refraction between younger and older persons was minimal. 

Linear regressions were performed to quantify the variation in individual biometric components and noncycloplegic refraction with age, controlling for gender (Table 4) . Because the data were cross-sectional, the results should be interpreted as follows: Among those 40 to 59 years, older persons had, on average, 0.31 mm (95% CI: 0.12, 0.50) shorter ALs, compared with persons 10 years younger, after adjusting for gender. Older persons also had shallower ACDs (on average, 0.25 mm shallower than persons 10 years younger), shorter VCDs (0.41 mm shorter), but thicker lenses (0.34 mm thicker) and higher LOCS III scores (0.5 higher). Older persons were more hyperopic by an average of +1.4 D for every 10 years difference in age. CC was similar in younger and older persons. 

Among persons aged 60 to 81 years, AL, VCD, and SE did not differ significantly between older and younger persons. However, older persons had shallower ACD (on average, 0.17 mm shallower than persons 10 years younger), thicker lenses (0.19 mm thicker), and greater lens nuclear opacity (0.6 greater LOCS III score) compared with younger persons. 

The relationship between different ocular dimensions and noncycloplegic refraction is shown in Table 5 . As expected, persons with a more “minus” SE were younger, had longer ALs, deeper ACDs, and longer VCD than subjects with “plus” SE. However, subjects with “plus” SE had thicker lens and greater severity of nuclear opacity. There was no significant variation in CCs between different refractive states. 

Linear regression models were constructed to evaluate the independent effects of biometric components on noncycloplegic refraction (final models presented in Table 6 ). Models with axial length (1 and 3) were analyzed separately from models with vitreous chamber depth, anterior chamber depth, and lens thickness (models 2 and 4). The relative “effect” of each biometric component was estimated by the absolute magnitude of the standardized regression coefficient (SRC). ²⁴ The data could be interpreted as follows. In persons aged 40 to 59 years, AL (model 1) and VCD (model 2) were the most “important” relative predictors of refraction. In model 1, an increase in 1 mm of AL was associated independently with a decrease in 2.02 D in SE, after controlling for corneal curvature and lens nuclear opacity. In terms of SRC, an increase in 1 SRC of axial length was associated independently with a decrease in 0.91 D in SE (the highest relative predictor). Similarly, in model 2, an increase in 1 mm of VCD was associated independently with a decrease in 2.28 D in SE, after controlling for other biometric components. In terms of SRC, an increase in 1 SRC of VCD was associated independently with a decrease in 0.93 D in SE (again the highest relative predictor). In both models, an increase in 1 SRC of CC was associated with a 0.53 D increase in SE (the next highest predictor). However, NO was not a significantly predictor of SE in this age group (SRC = 0.031 in model 1, and SRC = 0.034 in model 2). 

In persons aged 60 to 81 years, AL (model 3) and VCD (model 4) were again the most “important” relative predictors of SE, followed by CC. However, in both models, NO became an additional significant predictor of the SE. An increase in 1 SRC of NO was now associated with 0.36 and 0.33 reduction in SE in models 3 and 4, respectively. 

Discussion

Our study provides cross-sectional, population-based data on the variation in ocular biometric components in adult Chinese older than 40 years living in Singapore. First, we found that older persons had shorter ALs, shallower ACDs, shorter VCDs, but greater LTs and NOs than those of younger persons. Variation in CCs was minimal between younger and older people. Second, after controlling for age, we found that women had shorter ALs, shallower ACDs, shorter VCDs, but thicker lenses and steeper corneas than men. NO, as measured clinically, was not significantly different between men and women. Third, our biometric data are useful in explaining observed patterns in refraction reported in previous studies. ^{1 2 3 4 5}  

Existing biometric data are based mostly on studies in children, adolescents, and young adults ^{11 12 13 14 16 17 18} or on selected populations (such as medical students ¹⁵ or clinical microscopists ¹⁹ ). In adults after 40 years, comparable population-based data are less readily available. ^{1 2 3 4 5} Among Eskimos in Alaska, shorter ALs in older compared with younger persons were observed by van Rens and colleagues. ²⁵ In a nonrandom survey of 220 adult Chinese older than 40 years living in Hong Kong, older persons and women were found to have shorter ALs and VCDs, compared with younger persons and men, respectively. ²⁶ Other non–population-based studies have noted shorter ALs in older people, ^{27 28} relative stability of CCs with age, ^{29 30} and shorter ALs and CCs in women. ^{31 32 33} The explanation for the variation in ocular dimensions with gender is not clear but is likely to reflect differences in both genetic and environmental factors between men and women. ^{6 7 8 32}  

More interesting, and difficult to explain, is the variation in axial ocular dimensions with age. Inferences on age-related (longitudinal) trends can be problematic, based on cross-sectional data. This is illustrated by comparing alternative explanations for the variation in AL versus LT in our study. Shorter ALs in older compared with younger persons could be related to either a cohort effect (i.e., with older generations have shorter ALs because of poorer nutrition, general health, or other unknown factors than younger generations ⁶ ) or to an actual reduction in AL with increasing age. Data on which of the two hypotheses is more likely are inconsistent. Although some studies indicate that an age-related reduction in a persons AL is possible, ^{27 28} other studies have also shown that this parameter remains fairly constant in adolescents. ^{11 12 13 29 34} Our data do not provide sufficient information to distinguish whether an age-related change or a cohort effect is more likely. On the other hand, a cohort effect is an unlikely explanation for variation in LT between older compared with younger persons. The crystalline lens has been shown to continue growing throughout life. ³⁵ Previous research has demonstrated that the LT increases at a rate of approximately 0.02 mm/y. ³⁶ Our cross-sectional data were consistent with this observation. We found a difference in lens thickness of 0.035 mm/y in subjects aged 40 to 59 years and 0.019 mm/y in subjects aged 60 to 81 years. Therefore, the variation in lens thickness appears to be an age-related longitudinal change rather than a cohort effect. 

Our biometric data are useful in explaining the variation in refraction observed in previous studies ^{1 2 3 4 5} and in ours. ²⁰ Specifically, these studies have shown that in adults after age 40 years, older persons tend to have higher rates of hyperopia (and lower rates of myopia), compared with younger persons (which will be referred to as the “hyperopic shift” in this discussion). ^{1 2 3 4 5} After 60 years, the hyperopic shift appears to be less prominent, ^{1 2 3 4 5} and some studies have shown a pattern in the opposite “minus” direction with age, with increasing prevalence of myopia seen in elderly persons. ^{1 2} The variation in our study is depicted in Figure 7 . Refractive status in children, adolescents, and younger adults are explained mainly by variation in ALs and VCDs. ^{10 11 12 13 14 15 16 17 18 19} Likewise, we found AL and VCD to be the most“ important” relative predictors of refraction in our adult population (Table 6 , models 1–4). Shorter ALs and VCDs in older compared with younger persons appeared to explain the hyperopic shift in our population. After 60 years of age, we showed that NO became an additional significant predictor of refraction (Table 6 , models 3 and 4). Greater nuclear opacification in older persons appeared to drive the refraction in the “minus” direction, explaining why the hyperopic shift was less prominent in the group aged 60 to 81 years. This was consistent with data from the Visual Impairment Project in Melbourne ³ and in the Beaver Dam Eye Study, ⁹ which showed that increased nuclear opacity of the lens in older persons was associated with a more “minus” refraction status. 

However, we should point out that as cycloplegia was not used in our study, the effect of accommodation on refraction in our population was not known. In younger adults, this could be significant, and the hyperopic shift in Figure 7 could also be explained by the loss of accommodation in older compared with younger persons. Manual subjective refraction techniques in our study would theoretically reduce the effects of accommodation (compared with objective autorefraction), and in participants more than 50 years old, accommodative ability was expected to be minimal. 

The principal strength of our study was the population-based random sampling strategy, avoiding the bias seen in studies of biometry in specific, highly selected patient groups. ^{14 15 16 17 19} In addition, as the prevalence of refractive errors in our population was high, any variation in the biometric components was potentially accentuated. The problems with inferences based on cross-sectional data and noncycloplegic refraction have already been noted. Another potential problem was selection bias. For example, shorter axial dimensions in older persons and women could be explained by selective exclusion of older persons and women with longer axial dimension, due either to higher cataract extraction rates in these people (biometric data on pseudophakic and aphakic participants were not analyzed), higher mortality, or other unknown reasons for nonparticipation. 

In conclusion, we observed age and gender variation in ocular biometry in adult Chinese persons more than 40 years old in Singapore, with older people and women generally having shorter axial ocular dimensions. Vitreous chamber depth was the most “important” determinant of refraction in adults more than 40 years old. Shorter vitreous chamber depths in older compared with younger persons appear to explain the hyperopic shift seen cross-sectionally in previous population-based studies. After 60 years of age, lens nuclear opacity was a significant additional determinant of refraction, with greater lens opacities in older persons driving the refraction in the“ minus” direction, possibly explaining why the hyperopic shift in older persons was less prominent. 

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

 Submitted for publication June 20, 2000; revised August 9, 2000; accepted September 12, 2000.

 Commercial relationships policy: N.

 

Table 1.

 

Comparison of Subjects Included (Study Population) and Excluded from Analyses

	Included (n = 1004)	Excluded (n = 228)
Age (y)
40–49	260 (25.9)	16 (7.0)
50–59	281 (28.0)	25 (11.0)
60–69	270 (26.9)	73 (32.0)
70–81	193 (19.2)	114 (50.0)
Sex
Females	547 (54.5)	128 (56.1)
Education*
No education	243 (24.2)	83 (36.4)
Primary	402 (40.0)	96 (42.1)
Secondary	277 (27.6)	33 (14.5)
Tertiary	77 (7.7)	6 (2.6)
Occupation*
Professionals and office workers	200 (19.9)	24 (10.5)
Sales persons	170 (16.9)	24 (10.5)
Production operators	339 (33.8)	50 (21.9)
Laborers and cleaners	36 (3.6)	44 (19.3)
Homemakers	224 (22.3)	68 (29.8)
Unemployed	31 (3.3)	11 (4.8)
Housing*
1–2 room flats	171 (17.0)	64 (28.1)
3 room flats	540 (53.8)	118 (51.8)
4–5 room flats	262 (26.1)	37 (16.2)
Executive flats	17 (1.7)	2 (0.9)
Private housing	5 (0.5)	0 (0.0)
Individual monthly income*
Less than $1000	590 (58.8)	18.9 (82.9)
$1000–2000	184 (18.3)	9 (3.9)
$2000–3000	64 (6.4)	1 (0.4)
More than $3000	51 (5.1)	0 (0.0)
Retired	100 (10.0)	20 (8.8)
Smoking*
Yes	189 (18.8)	36 (15.8)

 Values are number of subjects, with percentage in parentheses.

*  Percentages may not add up to 100 because of missing data.

Table 2.

 

Ocular Biometry and Noncycloplegic Refraction, by Gender, Adult Chinese Residents in Singapore

	All (n = 1004)*	Males (n = 457)*	Females (n = 547)*	P , †
AL (mm)	23.23 ± 1.17	23.54 ± 1.10	22.98 ± 1.16	<0.001
ACD (mm)	2.90 ± 0.44	2.99 ± 0.45	2.81 ± 0.42	<0.001
LT (mm)	4.75 ± 0.47	4.73 ± 0.47	4.78 ± 0.47	0.003
VCD (mm)	15.58 ± 1.11	15.82 ± 1.08	15.39 ± 1.09	<0.001
CC (mm)	7.65 ± 0.27	7.73 ± 0.29	7.59 ± 0.24	<0.001
NO (LOCS III), ‡	3.2 ± 0.9	3.2 ± 0.9	3.2 ± 0.9	0.29
SE (D)	−0.49 ± 2.69	−0.40 ± 2.41	−0.56 ± 2.89	0.43

*  Values are means ± SD.

†  P values are based on age-adjusted difference in means between male and female subjects.

‡  LOCS III scores range from 1 to 6.

Table 3.

 

Ocular Biometry and Noncycloplegic Refraction, by Age and Gender, Adult Chinese Residents in Singapore

	Male Subjects							Female Subjects							All
		40–49 y (n = 120)	50–59 y (n = 106)	60–69 y (n = 142)	70–81 y (n = 89)	40–49 y (n = 140)	50–59 y (n = 175)	60–69 y (n = 128)	70–81 y (n = 104)	40–49 y (n = 260)	50–59 y (n = 281)	60–69 y (n = 270)	70–81 y (n = 193)
AL (mm)	23.80 ± 1.20	23.54 ± 1.19	23.37 ± 1.13	23.38 ± 0.85	23.40 ± 1.37	23.01 ± 1.16	22.73 ± 1.03	22.66 ± 0.75	23.58 ± 1.31	23.21 ± 1.20	23.08 ± 1.09	23.00 ± 0.88
ACD (mm)	3.25 ± 0.38	3.00 ± 0.42	2.92 ± 0.43	2.74 ± 0.40	3.08 ± 0.37	2.83 ± 0.39	2.70 ± 0.38	2.55 ± 0.35	3.16 ± 0.38	2.89 ± 0.41	2.82 ± 0.42	2.64 ± 0.38
LT (mm)	4.36 ± 0.35	4.71 ± 0.41	4.87 ± 0.41	5.05 ± 0.45	4.44 ± 0.33	4.77 ± 0.43	4.92 ± 0.45	5.14 ± 0.38	4.40 ± 0.34	4.75 ± 0.42	4.89 ± 0.43	5.10 ± 0.42
VCD (mm)	16.19 ± 1.15	15.83 ± 1.14	15.62 ± 1.00	15.60 ± 0.85	15.88 ± 1.25	15.41 ± 1.06	15.11 ± 0.96	14.98 ± 0.73	16.02 ± 1.21	15.57 ± 1.11	15.37 ± 1.01	15.28 ± 0.85
CC (mm)	7.71 ± 0.26	7.68 ± 0.42	7.71 ± 0.29	7.78 ± 0.33	7.56 ± 0.22	7.62 ± 0.37	7.57 ± 0.24	7.57 ± 0.27	7.63 ± 0.25	7.67 ± 0.24	7.64 ± 0.32	7.66 ± 0.32
NO (LOCS III)*	2.4 ± 0.5	2.9 ± 0.6	3.5 ± 0.7	4.3 ± 0.8	2.4 ± 0.5	3.0 ± 0.5	3.5 ± 0.7	4.2 ± 0.8	2.4 ± 0.5	2.9 ± 0.6	3.5 ± 0.7	4.1 ± 0.8
SE (D)	−1.39 ± 2.58	−0.24 ± 2.53	0.02 ± 2.05	0.08 ± 2.22	−2.08 ± 3.30	−0.10 ± 2.31	0.34 ± 2.55	−0.42 ± 2.84	−1.76 ± 3.00	−0.15 ± 2.40	0.18 ± 2.30	−0.19 ± 2.58

 Values are means ± SD.

*  LOCS III scores range from 1 to 6.

Figure 1.

 

Mean axial lengths (mm) in adult Chinese Singaporeans, by age. Vertical bar, 95% confidence interval for the mean.

Figure 2.

 

Mean anterior chamber depths (mm) in adult Chinese Singaporeans, by age. Vertical bar, 95% confidence interval for the mean.

Figure 3.

 

Mean lens thickness (mm) in adult Chinese Singaporeans, by age. Vertical bar, 95% confidence interval for the mean.

Figure 4.

 

Mean vitreous chamber depths (mm) in adult Chinese Singaporeans, by age. Vertical bar, 95% confidence interval for the mean.

Figure 5.

 

Mean corneal curvatures (mm) in adult Chinese Singaporeans, by age. Vertical bar, 95% confidence interval for the mean.

Figure 6.

 

Mean lens nuclear opacity (LOCS III score) in adult Chinese Singaporeans, by age. Vertical bar, 95% confidence interval for the mean.

Figure 7.

 

Mean noncycloplegic refraction (spherical equivalent diopters) in adult Chinese Singaporeans, by age. Vertical bar, 95% confidence interval for the mean.

Table 4.

 

Linear Regression Models of Ocular Biometry and Noncycloplegic Refraction, per Year Difference in Age, Adjusted for Gender in Adult Chinese Residents in Singapore

Biometric Components and Noncycloplegic Refraction	Biometric Components and Noncycloplegic Refraction, per Year Difference in Age*
		Age 40–59 y			Age 60–81 y
		Regression Coefficient, †	P	Regression Coefficient, †	P
AL (mm)	−0.031 (−0.05, −0.012)	0.001	−0.003 (−0.018, 0.013)	0.777
ACD (mm)	−0.025 (−0.031, −0.019)	0.000	−0.017 (−0.023, −0.010)	0.000
LT (mm)	0.034 (0.029, 0.040)	0.000	0.019 (0.012, 0.025)	0.000
VCD (mm)	−0.041 (−0.059, −0.024)	0.000	−0.003 (−0.018, 0.012)	0.688
CC (mm)	0.001 (−0.003, 0.005)	0.565	0.005 (0.000, 0.009)	0.031
NO (LOCS III), ‡	0.05 (0.04, 0.06)	0.000	0.06 (0.05, 0.07)	0.000
SE (D)	0.14 (0.095, 0.18)	0.000	−0.030 (−0.068, 0.008)	0.122

*  In each regression model, biometric components and noncycloplegic refraction were the dependent variables and age and gender were the independent variable. For example, in the first model, among subjects aged 40–59 years, after adjusting for gender, a person 1 year older would be expected to have an axial length 0.031 mm shorter.

†  Values in parentheses are 95% confidence interval.

‡  LOCS III scores range from 1 to 6.

Table 5.

 

Ocular Biometry, by Noncycloplegic Refraction in Adult Chinese Residents in Singapore

	Noncycloplegic Refraction (in Spherical Equivalent Diopters)
		<−9.0 (n = 15)	−9.0 to −7.01 (n = 20)	−7.0 to −5.01 (n = 35)	−5.0 to −3.01 (n = 64)	−3.0 to −1.01 (n = 141)	−1.0 to −0.01 (n = 252)	0.0 to +1.00 (n = 223)	+1.01 to +3.0 (n = 228)	>+3.0 (n = 26)
Age (y)	51.07 ± 9.40	53.45 ± 11.23	54.09 ± 10.95	55.42 ± 11.61	58.08 ± 12.01	56.27 ± 11.51	58.04 ± 10.02	62.61 ± 8.52	64.46 ± 7.53
AL (mm)	26.34 ± 1.18	25.25 ± 1.31	25.09 ± 1.28	24.11 ± 1.16	23.54 ± 0.99	23.22 ± 0.82	22.96 ± 0.88	22.63 ± 0.85	22.08 ± 0.87
ACD (mm)	3.14 ± 0.46	3.14 ± 0.54	3.16 ± 0.40	3.04 ± 0.44	2.98 ± 0.49	2.92 ± 0.43	2.89 ± 0.41	2.73 ± 0.39	2.65 ± 0.38
LT (mm)	4.62 ± 0.24	4.64 ± 0.65	4.60 ± 0.50	4.72 ± 0.48	4.68 ± 0.47	4.76 ± 0.52	4.70 ± 0.46	4.92 ± 0.38	4.87 ± 0.41
VCD (mm)	18.58 ± 1.11	17.47 ± 1.20	17.34 ± 1.30	16.35 ± 1.10	15.89 ± 0.89	15.54 ± 0.82	15.39 ± 0.78	14.97 ± 0.81	14.55 ± 0.80
CC (mm)	7.58 ± 0.24	7.43 ± 0.34	7.67 ± 0.30	7.54 ± 0.31	7.59 ± 0.37	7.64 ± 0.27	7.66 ± 0.25	7.70 ± 0.37	7.75 ± 0.22
NO (LOCS III)*	3.0 ± 1.1	3.2 ± 1.1	3.2 ± 1.0	3.2 ± 1.1	3.4 ± 1.1	3.2 ± 0.9	3.0 ± 0.7	3.3 ± 0.7	3.5 ± 0.9

 Values are means ± SD.

*  LOCS III scores range from 1 to 6.

Table 6.

 

Linear Regression Models of Noncycloplegic Refraction (in Spherical Equivalents Diopters), per Unit Difference in Biometric Components in Adult Chinese Residents in Singapore

Biometric Components	Noncycloplegic Refraction (in Spherical Equivalents Diopters) per Unit Difference in Biometric Components*
		Age 40–59 y					Age 60–81 y
		Unstandardized Regression Coefficient, †	Standardized Regression Coefficient	P	Unstandardized Regression Coefficient, †	Standardized Regression Coefficient	P
	Model 1 (R2 = 0.75)					Model 3 (R2 = 0.48)
AL (mm)	−2.02 (−2.13, −1.92)	−0.91	0.000	−1.45 (−1.61, −1.28)	−0.68	0.000
CC (mm)	5.78 (5.26, 6.29)	0.53	0.000	3.51 (2.91, 4.10)	0.46	0.000
NO (LOCS III)	0.15 (−0.06, 0.35)	0.03	0.168	−0.98 (−1.17, −0.80)	−0.36	0.000
	Model 2 (R2 = 0.74)					Model 4 (R2 = 0.52)
VCD (mm)	−2.28 (−2.40, −2.16)	−0.933	0.000	−1.68 (−1.87, −1.49)	−0.755	0.000
CC (mm)	6.14 (5.64, 6.64)	0.529	0.000	3.95 (3.35, 4.56)	0.528	0.000
LT (mm)	−1.53 (−1.9, −1.16)	−0.230	0.000	−1.29 (−1.74, −0.84)	−0.258	0.000
ACD (mm)	−0.63 (−1.01, −0.25)	−0.094	0.003	−0.51 (−0.98, −0.04)	−0.096	0.036
NO (LOCS III), ‡	0.17 (−0.03, 0.38)	0.034	0.148	−0.88 (−1.07, −0.69)	−0.333	0.000

*  In each regression model, noncycloplegic refraction was the dependent variable with each biometric component as independent variables. For example, in model 1, among subjects aged 40–59 years, a person with an axial length 1 mm longer would be expected to have a 2.02 D more negative noncycloplegic refraction in spherical equivalents.

†  Values in parentheses are 95% confidence interval.

‡  LOCS III scores range from 1 to 6.