Refractive Errors in an Elderly Chinese Population in Taiwan: The Shihpai Eye Study

Abstract

purpose. Few epidemiologic data are available on refractive status in elderly Asians. The purpose of the study was to determine prevalence and risk factors associated with refractive errors in a metropolitan elderly Chinese population in Taiwan.

methods. A population-based survey was conducted in the Shihpai district of Taipei, Taiwan. A total of 2045 residents aged 65 years or more were randomly selected and invited to complete a comprehensive questionnaire and undertake a detailed ocular examination, including best corrected visual acuity and measurements of refractive error, using autorefraction. Of the subjects, 1361 (66.6%) participated in the ocular examination. Spherical equivalent (SE) was calculated in diopters (D), and data from right eyes were reported.

results. The age- and sex-adjusted prevalence rates were determined for myopia (SE < −0.5 D, 19.4%; SE < −1.0 D, 14.5%), high myopia (SE < −6.0 D, 2.4%), hyperopia (SE > +0.5 D, 59.0%; SE > +1.0 D, 44.2%), astigmatism (cylinder < −0.5 D, 74.0%; cylinder < −1.0 D, 45.3%), and anisometropia (SE difference between right and left eyes > 0.5 D, 45.2%; SE difference > 1.0 D, 21.8%). The prevalence of myopia, astigmatism, and anisometropia significantly increased with age (all P < 0.01). The prevalence of hyperopia tended to decrease with age. There was no gender difference in prevalence rates in any type of refractive error, except that women had a higher rate of hyperopia (SE > +1.0 D) than men (P = 0.004). Multivariate regression analysis showed that myopia was weakly associated with higher educational level. The severity of lens nuclear opacity was positively associated with the rates of myopia and negatively associated with the rates of hyperopia.

conclusions. The prevalence of myopia in this elderly Chinese population is not much higher than in similarly aged elderly white populations, compared with a much greater difference in prevalence among younger Chinese versus white people. This suggests that changing environmental factors may account for the increased prevalence of myopia in younger cohorts of Chinese.

Refractive error is the cause of a significant proportion of visual impairment and blindness, ^{1 2 3} and the correction of refractive error by spectacles, contact lenses, or refractive surgery can entail a large socioeconomic cost. About half the citizens of developed countries wear glasses or contact lenses. ^{4 5} In the United States, an estimated $12.8 billion was spent to correct refractive error in 1990. ⁴ The public health costs of refractive error are also considerable, such as the costs imposed by the morbidity resulting from myopia, including cataract, glaucoma, retinal detachment, and myopic maculopathy. ⁶  

The prevalence of refractive error varies according to population characteristics, such as age and ethnic group. In recent years, some population-based studies on refractive error have been conducted in Asia, both in developing countries, such as Indonesia ⁷ and India, ^{8 9 10} and in developed countries, such as Taiwan ¹¹ and Singapore. ¹² All these studies were conducted among either school-aged children ^{7 8 9 10 11} or young and middle-aged adults. ^{7 8 12} There is little known about the prevalence of refractive error in the Asian elderly population. 

As in Western communities, the aging population is rapidly growing in Asia. In Taiwan, life expectancy is 72.6 years for men and 78.3 years for women, and 8.6% of the population was aged more than 65 years in 2000. ¹³ With increasing longevity worldwide, age-related ocular diseases are becoming a priority in eye care services. Some population-based surveys of refractive error in the United States have selected those aged 40 years and older, with few persons more than 70 years of age. ^{14 15} To allow for more precise estimates of vision status and prevalence of ocular diseases in the older age group, the Salisbury Eye Evaluation Study examined a random sample of individuals aged more than 65 years, ^{16 17} but their refractive status has not yet been reported. 

Valid data on epidemiology of refractive error in elderly Asians are not readily available. To plan and provide optimal eye health services for this growing segment of the population, it is imperative to obtain accurate information on the refractive status of the elderly. The purpose of this study was to estimate the prevalence of refractive errors, such as myopia, hyperopia, astigmatism, and anisometropia, in an elderly Chinese population in Taiwan and to determine the risk factors associated with the refractive errors. 

Methods

Study Population

This study was a part of the population-based Shihpai Eye Study. Detailed methodology of the Shihpai Eye Study has been reported elsewhere. ^{18 19} In brief, this survey of vision and eye diseases was conducted among noninstitutionalized residents 65 years of age or more in Shihpai, Taipei, Taiwan. The Shihpai community is located in the Peitou district of Taipei. The Peitou district had a population of approximately 247,100 at the end of 1999, ²⁰ making it the second largest district in Taipei. This area was chosen because the population is relatively stable—that is, there was a 0.3% annual population increase between 1994 and 1999. ²⁰ The study was performed between July 1999 and December 2000. 

The government household registration system was used to identify residents aged 65 years and more in Shihpai. This registration system is designed to collate and supply demographic information and to recognize officially personal status and relations. This system provides highly accurate and complete population statistics, which allow the availability of reliable demographic data to assist scholars in academic research. 

Our goal was to recruit a random sample of approximately 2000 residents 65 years of age or older with complete baseline information. According to the official household registration figures for 1999, the total number of residents 65 years of age or older in Shihpai was 4750. Excluding vacant households (658 persons), residents who died before they could be contacted (48 persons), and inpatient, paralyzed, and disabled residents (298 persons), 3746 were eligible for the study. Of these eligible subjects, 2045 were randomly selected and were invited to participate in the survey. 

Procedures

The study protocol was approved by the institutional review board, and the study was conducted according to the tenets of the Declaration of Helsinki of the World Medical Association regarding scientific research on human subjects. 

Before the ocular examination, trained interviewers contacted the participants and administrated a structured questionnaire in the home. The questionnaire was used to obtain baseline information on demographic data, personal medical history, and lifestyle from each subject. Those who had been interviewed were invited to the study hospital for a detailed ocular examination, including autorefractometry, keratometry, best corrected visual acuity assessment, noncontact tonometry, slit lamp examination with cataract grading, fundus examination, and photography. 

All participants underwent initial objective autorefraction (model RK-8100; Topcon, Tokyo, Japan). The result of autorefraction was used as a starting point for a subsequent subjective refraction. The visual acuity was measured in each eye initially without refractive error correction or with distance glasses if worn, using Snellen charts at a distance of 6 m. If the presenting visual acuity was less than 6/6, the examination was repeated with subjective refraction. If visual acuity was 6/6 or better with current distance correction, this correction was measured using a lensmeter (CL-100; Topcon), and the readout was recorded as the refractive error. Because of the age of the study population, cycloplegia was not used. 

Lens opacity was graded using slit lamp biomicroscopy (model BQ900; Haag-Streit, Bern, Switzerland) with the modified Lens Opacity Classification System (LOCS) III. ²¹ Based on this system, nuclear opacity was graded on a scale from 1 to 6, with 6 showing the highest. A physical examination, including blood pressure measurements, was also conducted. Hypertension was defined if the subject was currently using anti-hypertension medication, the measured systolic blood pressure was more than 160 mm Hg, or the measured diastolic blood pressure was more than 95 mm Hg. All personnel were trained and all procedures were standardized before the survey was conducted. Informed consent was obtained from each subject before enrollment in the study. 

Definitions

Spherical equivalent (SE) was used for calculations of refractive error. The SE is derived by adding the spherical component of a refraction to half of the cylindrical component. For the purpose of comparison with other reports, refractive error was defined on two levels. Myopia was defined as an SE of less than −0.5 or −1.0 D. Hyperopia was defined as an SE of greater than +0.5 or +1.0 D. Astigmatism was analyzed in minus cylinders and was categorized as follows: cylinder less than −0.5 or −1.0 D. Anisometropia was defined as a difference in SE of greater than 0.5 or 1.0 D between right and left eyes. In addition, high myopia was defined as an SE of less than −6.0 D. ²² Because there was a high correlation between the fellow eyes (ρ = 0.82, P < 0.001), and because the results based on right eyes and left eyes were similar, data from right eyes only were reported, except in the analysis of anisometropia. Subjects who had pseudophakia or aphakia on ocular examination were excluded from the analysis. Those with ocular conditions that could interfere with accurate refraction, such as corneal opacity or visually impairing opaque media, were also excluded. 

Statistical Analysis

The prevalence of myopia, hyperopia, astigmatism, and anisometropia in subjects with different characteristics was expressed in percentages of the study population with 95% confidence interval (CI). The crude prevalence rates were further age and sex adjusted according to the 1999 Taiwan population, ²⁰ to obtain a more accurate estimate of the actual prevalence of refractive error. 

The association of refractive errors with age, sex, educational level, hypertension, self-reported diabetes, cigarette smoking, alcohol intake, and lens nuclear opacity was assessed. The χ² test was used for univariate analysis. Further, adjusted odds ratios were obtained by using multivariate logistic regression models, allowing for control of the mutually confounding effect of these potential risk factors. All data analyses were performed with a commercial statistical software package (Stata; Stata Corp., College Station, TX). 

Results

Among the 2045 randomly selected subjects, 2038 (99.7%) subjects cooperated with the household interview and completed the questionnaire. Seven (0.3%) people could not be contacted during three visits for the household interview. After finishing the questionnaire, 1361 (66.6%) subjects came to the study hospital and participated in the ocular examination. Six hundred seventy-seven (33.1%) subjects responded to the questionnaire but did not receive the ocular examination, because they refused or had no spare time available to go to the hospital. Table 1 shows the comparison of selected characteristics for subjects who participated in the ocular examination and those who did not. In general, participants (72.2 ± 5.1 years) were younger than nonparticipants (74.3 ± 6.0 years), more likely to be male, and had a higher level of education and a higher frequency of current smoking and alcohol drinking (all P < 0.001). 

Of the 1361 participants, 174 (12.8%) had undergone cataract surgery in their right eyes (169 pseudophakic and 5 aphakic). In an additional 79 (5.8%) subjects, refractive error could not be appropriately measured because of dense cataract, corneal or media opacity, bulbar atrophy, or other reasons. After these subjects were excluded, analyses were based on 1108 right eyes. Included subjects (71.7 ± 4.7 years) were younger than those who were excluded (74.6 ± 5.5 years; P = 0.005), were more likely to be male (P = 0.005), had a higher level of education (P = 0.020), and were less likely to have hypertension (P = 0.010) and self-reported diabetes (P = 0.001). For the analysis of anisometropia, an additional 75 (5.5%) subjects were excluded, because the refractive error in the left eye could not be appropriately measured. 

The distribution of refractive error in SE is shown in Figure 1 . The mean SE was 0.47 D with an SD of 2.47 D. The distribution shows a skew toward myopia. Only 239 (21.6%) of the persons were emmetropic (SE between −0.5 D and +0.5 D). In 91% of the study population, the SE was between −3.0 D and +3.0 D. Figure 2 shows the cylinder distribution in minus power. Of the study population, 95% had an astigmatism less than 3.0 D, and only 3.8% had 0 D of astigmatism. Analyzing the refractive data using the power vector method ²³ showed that the mean primary astigmatism (J₀) was −0.41 ± 0.55 D, and the mean oblique astigmatism (J₄₅) was 0.09 ± 0.37 D in this elderly population (the other component, M, is equal to the SE as defined earlier). 

Table 2 shows the crude and adjusted prevalence rates of refractive errors with the two alternate definitions. Overall, 203 (18.3%) persons had myopia (SE < −0.5 D), 666 (60.1%) had hyperopia (SE > +0.5 D), and 813 (73.4%) had astigmatism (cylinder < −0.5 D). The prevalence rates of different refractive errors by age group are presented in Figure 3 . Both of the two alternate definitions of refractive error had similar trends with age. The rates of myopia were higher in the older age groups, with a peak in persons aged 75 to 79 years. The older groups had higher rates of astigmatism and anisometropia, but lower rates of hyperopia. 

Table 3 shows the prevalence rates of myopia (SE < −0.5 D) and high myopia (SE < −6.0 D) by different characteristics. The prevalence of myopia increased significantly (P < 0.001) with age, from 12.8% at 65 to 69 years of age to 26.5% at 75 to 79 years of age. The prevalence of myopia increased significantly (P < 0.001) with the severity of lens nuclear opacity. Subjects with nuclear opacity LOCS scores of 5 or 6 had the highest rates of myopia (56.7%), whereas those with nuclear opacity LOCS scores of 1 or 2 had the lowest rates (14.1%). The prevalence of high myopia was 2.3% (25 persons). There was no significant gender difference in the prevalence of myopia or high myopia. There were no significant trends in the rates of myopia by other factors. When the association of myopia with the potential risk factors were analyzed using the more stringent definition (SE < −1.0 D), the results were similar. 

The prevalence rates of hyperopia (SE > +0.5 D) are presented in Table 3 . The highest rates of hyperopia were in the group aged 65 to 69 years (62.9%). The rates decreased with increasing age, but the trend was not significant (P = 0.185). Compared with myopia, hyperopia had a reversed pattern of association with nuclear opacity. Subjects with no or mild nuclear opacity (LOCS scores of 1 or 2) had the highest rates of hyperopia (64.3% for hyperopia > +0.5 D), and those with dense nuclear opacity (LOCS scores of 5 or 6) had the lowest rates (36.7%). Using the more stringent definition of hyperopia (SE > +1.0 D), the results of the univariate analysis were similar, except that women had a significant higher rate of hyperopia (> +1.0 D) than did men (50.1% vs. 41.4%, P = 0.004). 

The prevalence of astigmatism was generally high in this elderly population (Table 4) . The rates (cylinder < −0.5 D) increased from 67.8% in the group aged 65 to 69 years to 84.9% in the group aged 80 years or more. The rates of astigmatism (cylinder < −0.5 D) did not vary significantly with nuclear opacity LOCS scores or other factors. Results were similar when the definition of cylinder < −1.0 D was used, except a higher prevalence of astigmatism was weakly associated with non–alcohol drinkers (P = 0.042, data not shown). The most common type of astigmatism in the study population was against-the-rule astigmatism (693 persons, 62.5%), followed by oblique astigmatism (289 persons, 26.1%) and with-the-rule type (126 persons, 11.4%). The frequency pattern was similar across all age groups. 

The prevalence rates of anisometropia (SE difference between right and left eyes > 0.5 D) by different characteristics are shown in Table 4 . The overall prevalence of anisometropia was 42.6%. The rates were significantly (P < 0.001) higher in the older groups and did not vary with gender. In the subjects with a difference in nuclear opacity, an LOCS score of 1 or more between right and left eyes, the rate of anisometropia was 54.4%, which was significantly higher than the rate of 41.5% in those without a difference in LOCS scores between eyes (P = 0.017). 

With multivariate regression analysis, various models were constructed and selectively presented in Table 5 for myopia (SE < −0.5 D), hyperopia (SE > +0.5 D), astigmatism (cylinder < −0.5 D), and anisometropia (SE difference > 0.5 D). After adjusting for other confounding factors (gender, educational level, and nuclear opacity), subjects aged 75 to 79 years had a 2.1-fold higher risk of myopia and subjects aged 70 to 74 years a 1.6-fold higher risk, compared with subjects aged 65 to 69 years. The group with the highest educational level had a twofold higher risk of myopia than did the illiterate group. The positive association of nuclear opacity with myopia and the negative association with hyperopia remained significant in the multivariate analysis. For astigmatism, after controlling for gender and educational level, older subjects had higher risk (1.4- to 2.5-fold) of astigmatism than the younger group (aged 65–69 years). After adjustment for gender, older subjects had higher risk (1.4- to 3.2-fold) of anisometropia than subjects aged 65 to 69 years. When the analyses were performed using the more stringent definition (1.0 D) of refractive error, the results did not change markedly. 

In the multivariate regression models for high myopia (data not presented), after adjustment for gender, age, and educational level, there was a significantly (P = 0.001) higher risk of myopia in subjects with nuclear opacity LOCS scores of 5 or 6 than those with scores of 1 or 2 (odds ratio = 9.9, 95% CI: 2.7–36.7). 

Discussion

To our knowledge, this study provides the first population-based data on the prevalence and distribution of refractive errors in elderly Asians. Myopia is particularly prevalent in East Asia, especially among the Chinese. ²⁴ Most studies in Asia have been conducted in relatively young populations ^{7 9 10 11} or have included only a small proportion of the elderly. ^{8 12} The present study provides an opportunity to compare the prevalence of myopia and other refractive errors with other ethnic populations in similarly aged elderly groups. 

Figure 4 presents the prevalence of myopia and hyperopia in selected population-based surveys. For the purpose of comparison, only data from similarly aged elderly groups are shown. ^{12 14 15 25 26 27} In the present study, the prevalence of myopia (SE < −0.5 D) was 19.4%, which is lower than that reported in similar age groups in Tanjong Pagar, Singapore ¹² and in Barbados, West Indies. ²⁵ By contrast, the prevalence of myopia in our Chinese population is slightly higher than that in similarly aged elderly white populations: 14.7% in the Beaver Dam Eye Study, ¹⁵ 11.1% in the Blue Mountains Eye Study, ²⁷ and 17.9% in the white group in the Baltimore Eye Survey. ¹⁴ Despite the lower prevalence of myopia in blacks than in whites and Asians in previous studies, ^{14 24} the prevalence of myopia is fairly high in the elderly black group in the Barbados Eye Study (Fig. 4) . Because lens nuclear opacity is significantly associated with myopia ^{12 25 26} and subjects who had undergone cataract surgery were usually excluded from the analysis in the studies of refractive error, the lower frequency (3%) of cataract surgery and the higher prevalence (41%) of lens opacity in the black participants in Barbados, ²⁸ compared with other ethnic populations, may account for the high prevalence of myopia. 

The present study supports the theory of cohort effects in the age-related trends in prevalence of myopia. First, the prevalence in this elderly Chinese population is much lower than that previously reported in younger generations in Taiwan ¹¹ and other Asian countries. ^{7 29 30} The prevalence of myopia (defined as SE < −0.25 D) among schoolchildren aged 16 to 18 years in Taiwan was as high as 84%, ¹¹ which is much higher than a recalculated prevalence of 22% in the present study, using the same definition. It is unlikely that such a great difference could be explained by an intrinsic decrease in the amount of an individual’s myopia as part of the aging process. ³¹ Changing environmental factors, most notably increased near-work activity and stringent educational studies in the past two decades, ¹¹ may account in part for the increasing prevalence of myopia among younger generations of Chinese people. Although our data suggest that there is only a weak association between myopia and educational level, this may reflect lower access to education or less near-work demands in the past for this elderly population. Second, the prevalence of myopia in this elderly Chinese population is not much higher than in similarly aged elderly white populations (Fig. 4) , ^{14 15 27} compared with a much greater difference in myopia rates among younger Chinese versus similar white age groups. ^{6 24} The attenuated ethnic difference in the prevalence of myopia in the elderly again supports a greater contribution of recent environmental factors explaining the higher prevalence of myopia in younger generations of Chinese people. 

Changes in refractive error with age are noteworthy (Fig. 4) . Data from cross-sectional studies have shown that after 40 years of age, older persons tend to have lower rates of myopia and higher rates of hyperopia, than do younger persons. ^{12 14 15 26 27} This trend is referred to as the hyperopic shift. ^{32 33 34} However, after 60 or 70 years of age, as depicted in Figure 4 , the hyperopic shift seems less prominent. ^{14 15 26 27} Some studies (such as the Barbados Eye Study ²⁵ and the Tanjong Pagar Survey ¹² ) and ours even show a reverse trend—that is, an increasing prevalence of myopia and a decreasing prevalence of hyperopia with advancing age in the elderly groups. It has been suggested that axial lengths and vitreous chamber depths are the most important predictors of refraction in adults. ^{34 35} Shorter axial lengths and vitreous chamber depths in older than in younger adults may explain the observed hyperopic shift. In the elderly, lens nuclear opacity becomes an additional significant predictor of refractive error. ^{12 25 26 33 36} This is consistent with our findings that the degree of nuclear opacity was positively associated with the prevalence of myopia and inversely associated with the prevalence of hyperopia. Changes in the refractive index of the lens substantially influence the shift of refraction. Thus, denser nuclear opacity in the elderly may drive the refractive error in the minus direction, which makes the hyperopic shift less evident. This is supported by data from the longitudinal Beaver Dam Eye Study, ³⁷ which showed that after a 10-year period, younger adults became more hyperopic, whereas older adults and elderly people became more myopic, and much of this may have been related to increasing nuclear opacity. 

Adult Chinese appear to have a greater prevalence of high myopia than whites. ¹² The prevalence was 2.3% in our elderly Chinese population (SE < −6.0 D), and 4.7% in Singapore Chinese aged 60 years and more (SE < −5.0 D). ¹² By contrast, the prevalence was only 0.87% in whites aged 60 years and more in the Baltimore Eye Survey. ¹⁴ The prevalence of high myopia is an important concern, because the consequences of high myopia, such as macular degeneration, glaucoma, and cataract ^{6 36} may contribute significantly to visual impairment. In a previous study, we found that high myopia macular degeneration contributes to 25% of visual impairment in adult Chinese. ³⁸  

In our study, women had a higher prevalence of hyperopia than men, a finding similar to those in other reports. ^{8 12 25 27} This may be because women’s eyes have a shorter axial length and shallower anterior chamber depth than those of men, ³⁴ and hence a higher probability of being hyperopic. Women’s eyes tended to have steeper corneas than those of men in the present study (data not shown), which also has been shown by others. ^{34 39} Nevertheless, this cannot fully offset the effect of the shorter axial lengths on hyperopia. Our subsequent analysis shows that after adjustment for corneal curvature, women still have a 1.5-fold (95% CI: 1.2–1.9) higher risk of hyperopia than men. 

There are few population-based data available on the prevalence of astigmatism in the elderly. In the present study, almost 75% of the subjects had astigmatism (cylinder < −0.5 D). The prevalence of astigmatism in the Chinese population in the Tanjong Pagar Survey was 51% for the 60- to 69-year age group and 68% for the 70- to 79-year age group. ¹² The prevalence in other ethnic populations is relatively lower. The Baltimore Eye Survey found rates of astigmatism below 49% and 39% in whites and blacks, respectively, in those aged 60 and more. ¹⁴ Chinese appear to have a higher prevalence of astigmatism than other populations, but this must be confirmed. 

Anisometropia (SE difference > 1.0 D) was present in 21.8% of the participants in the present study. This is close to the 26% rate among those aged 60 and more in the Tanjong Pagar Survey. ¹² The prevalence reported from the Baltimore Eye Survey in similarly aged elderly groups was lower, ranging from 5.5% to 18.1% (with variation by gender and race). ¹⁴ Similar to these and other studies, ^{12 14 40} we observed patterns of higher rates of anisometropia in the older age groups. 

We found no significant difference in refractive error between people with and without diabetes. The finding is consistent with the Beaver Dam Eye Study. ¹⁵ As expected, other factors, such as hypertension, smoking, and alcohol intake, did not significantly affect the prevalence of refractive error. 

Population-based studies in the elderly are usually limited by a low response rate. Elderly people usually have a higher frequency of morbidity and disability that hinders travel and motivation to participate. The response rate in the present study was 66.6%. This compares favorably with 65% in the Salisbury Eye Evaluation Study, ^{16 17} which targeted the same age group (≥65 years) and was better than in other studies of this age group that include lengthy clinical examinations—for example, the Cardiovascular Health Study had a 55% response rate. ⁴¹ However, there are still limitations to the present study. First, we acknowledge that nonparticipants tended more often to be older and female (Table 1) , and this may bias our estimates. Because the prevalence of myopia, astigmatism, and anisometropia increases with age, underrepresentation of older people may result in underestimation of the crude prevalence. For hyperopia, we found that older age and female gender affected the prevalence in opposite directions. Thus, we cannot know exactly whether nonparticipants are more or less likely than participants to have hyperopia. Second, we did not collect data on income level and other socioeconomic factors, such as occupation, which might be important potential confounders of the association with refractive error. 

In summary, this study provides epidemiologic data on refractive errors in an elderly Chinese population in Taiwan. This elderly population had a much lower prevalence of myopia than the younger generations of Chinese in Taiwan. However, the prevalence was not much higher than in similarly aged white populations, compared with a greater difference between younger Chinese and white people. This suggests that changing environmental factors—most notably, increasing education and near-work demands—account for the increased prevalence of myopia in recent birth cohorts of Chinese. In contrast, the shift toward myopia in the elderly was mainly related to increasing lens nuclear opacity. 

 Presented at the annual meeting of the Association for Research in Vision and Ophthalmology, Fort Lauderdale, Florida, May 2002.

 Supported by Grants VGH 89-404 and VGH 92-351 from Taipei Veterans General Hospital.

 Submitted for publication February 17, 2003; revised June 15, 2003; accepted July 2, 2003.

 Disclosure: C.-Y. Cheng, None; W.-M. Hsu, None; J.-H. Liu, None; S.-Y. Tsai, None; P. Chou, None

 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.

 

Selected Characteristics for Participants and Nonparticipants in the Shihpai Eye Study

	Participants (%) (n = 1361)	Nonparticipants (%)* (n = 677)
Gender, †
Male	60.4	48.7
Female	39.6	51.3
Age (y), †
65–69	35.3	29.4
70–74	36.2	28.9
75–79	19.7	24.7
80+	8.8	17.0
Educational level, †
Illiterate or no education	16.0	22.5
Elementary school	31.4	34.2
Junior high school	15.4	14.2
Senior high school	17.9	15.9
College and above	19.3	13.2
Marital status
Married	73.7	74.0
Single	1.6	1.6
Widowed	22.0	23.5
Divorced/separated	2.7	0.9
Smoking, †
Nonsmoker	63.6	73.3
Ex-smoker	18.8	11.8
Current smoker	17.6	14.9
Alcohol drinking, †
Nondrinker	82.8	90.4
Ex-drinker	5.5	1.8
Current drinker	11.7	7.8
History
Hypertension	45.4	43.7
Diabetes	15.2	16.0
Cataract	49.6	50.2
Glaucoma	4.9	5.8
Retinopathy	1.9	2.1
Ever seen an ophthalmologist	29.7	27.5

Figure 1.

 

Distribution of SE refractive error in the elderly Chinese in the Shihpai Eye Study (n = 1108 eyes).

Figure 2.

 

Distribution of cylinder in the elderly Chinese in the Shihpai Eye Study (n = 1108 eyes).

Table 2.

 

Crude and Adjusted Prevalence Rates of Refractive Errors in an Elderly Chinese Population in Taiwan

	n	Prevalence Rate (%)
				Crude	Adjusted* (95% CI)
Myopia
SE <−0.5 D	203	18.3	19.4 (16.7–22.1)
SE <−1.0 D	151	13.6	14.5 (12.1–16.9)
SE <−6.0 D	25	2.3	2.4 (1.4–3.4)
Hyperopia
SE > +0.5 D	666	60.1	59.0 (55.8–62.1)
SE >+1.0 D	495	44.7	44.2 (41.1–47.3)
Astigmatism
Cylinder <−0.5 D	813	73.4	74.0 (71.3–76.8)
Cylinder <−1.0 D	478	43.1	45.3 (42.1–48.5)
Anisometropia
SE difference > 0.5 D	440	42.6	45.2 (41.9–48.6)
SE difference > 1.0 D	205	19.9	21.8 (18.8–24.7)

Figure 3.

 

Prevalence rates of refractive errors in the elderly Chinese in the Shihpai Eye Study, by age group and by two alternate definitions—0.5 D (left) and 1.0 D (right). For anisometropia, the number of subjects in the 65- to 69-, 70- to 74-, 75- to 79-, and 80+-year age groups was 414, 396, 162, and 61, respectively.

Table 3.

 

Prevalence Rates of Myopia, High Myopia, and Hyperopia by Potential Risk Factors in an Elderly Chinese Population in Shihpai, Taiwan

	Total	Myopia (SE < −0.5 D)			High Myopia (SE < −6.0 D)			Hyperopia (SE > +0.5 D)
				n	% (95% CI)	n	% (95% CI)	n	% (95% CI)
Gender
Male	689	128	18.6 (15.7–21.7)	14	2.0 (1.1–3.4)	400	58.1 (54.3–61.8)
Female	419	75	17.9 (14.3–21.9)	11	2.6 (1.3–4.6)	266	63.5 (58.7–68.1)
			P = 0.777		P = 0.519		P = 0.073
Age (y)
65–69	429	55	12.8 (9.8–16.4)	8	1.9 (0.8–3.6)	270	62.9 (58.2–67.5)
70–74	417	81	19.4 (15.7–23.6)	8	1.9 (0.8–3.7)	251	60.2 (55.3–64.9)
75–79	189	50	26.5 (20.3–33.3)	8	4.2 (1.8–8.2)	108	57.1 (49.8–64.3)
80+	73	17	23.3 (14.2–34.6)	1	1.4 (0.0–7.4)	37	50.7 (38.7–62.6)
			P < 0.001		P = 0.249		P = 0.185
Educational level
Illiterate or no education	168	29	17.3 (11.9–23.8)	4	2.4 (0.7–6.0)	104	61.9 (54.1–69.3)
Elementary school	333	57	17.1 (13.2–21.6)	5	1.5 (0.5–3.5)	200	60.1 (54.6–65.4)
Junior high school	175	30	17.1 (11.9–23.6)	9	5.1 (2.4–9.5)	114	65.1 (57.6–72.2)
Senior high school	209	35	16.7 (12.0–22.5)	2	1.0 (0.1–3.4)	123	58.9 (51.9–65.6)
College and above	222	51	23.0 (17.6–29.1)	5	2.3 (0.7–5.2)	126	56.3 (49.5–62.9)
			P = 0.383		P = 0.059		P = 0.470
Hypertension
No	637	115	18.1 (15.1–21.3)	14	2.2 (1.2–3.7)	384	60.3 (56.4–64.1)
Yes	471	88	18.7 (15.3–22.5)	11	2.3 (1.2–4.1)	282	59.9 (55.3–64.3)
			P = 0.789		P = 0.879		P = 0.890
Self-reported diabetes
No	932	174	18.7 (16.2–21.3)	23	2.5 (1.6–3.7)	554	59.4 (56.2–62.6)
Yes	151	25	16.6 (11.0–23.5)	2	1.3 (0.2–4.7)	101	66.9 (58.5–74.3)
			P = 0.534		P = 0.385		P = 0.083
Smoking
Nonsmoker	704	122	17.3 (14.6–20.4)	14	2.0 (1.1–3.3)	426	60.5 (56.8–64.1)
Ex-smoker	204	36	17.6 (12.7–23.6)	5	2.5 (0.8–5.6)	125	61.3 (54.2–68.0)
Current smoker	199	45	22.6 (17.0–29.1)	6	3.0 (1.1–6.4)	115	57.8 (50.6–64.7)
			P = 0.226		P = 0.676		P = 0.738
Alcohol drinking
Nondrinker	912	171	18.8 (16.3–21.4)	23	2.5 (1.6–3.8)	540	59.2 (56.0–62.4)
Ex-drinker	59	9	15.3 (7.2–27.0)	1	1.7 (0.0–9.1)	39	66.1 (52.6–77.9)
Current drinker	136	23	16.9 (11.0–24.3)	1	0.7 (0.0–4.0)	87	64.0 (55.2–72.0)
			P = 0.718		P = 0.406		P = 0.361
Nuclear opacity (LOCS III score)
1–2	611	86	14.1 (11.4–17.1)	11	1.8 (0.9–3.2)	393	64.3 (60.4–68.1)
3–4	460	99	21.5 (17.9–25.6)	10	2.2 (1.0–4.0)	258	56.1 (51.4–60.7)
5–6	30	17	56.7 (37.4–74.5)	4	13.3 (3.8–30.7)	11	36.7 (19.9–56.1)
			P < 0.001		P < 0.001		P = 0.001

Table 4.

 

Prevalence Rates of Astigmatism and Anisometropia by Potential Risk Factors in an Elderly Chinese Population in Shihpai, Taiwan

	Astigmatism (Cylinder < −0.5 D)					Anisometropia (SE Difference > 0.5 D)
		Total	n	% (95% CI)	Total	n	% (95% CI)
Gender
Male	689	510	74.0 (70.6–77.3)	644	268	41.6 (37.8–45.5)
Female	419	303	72.3 (67.8–76.5)	389	172	44.2 (39.2–49.3)
			P = 0.533			P = 0.413
Age (y)
65–69	429	291	67.8 (63.2–72.2)	414	150	36.2 (31.6–41.1)
70–74	417	313	75.1 (70.6–79.1)	396	173	43.7 (38.7–48.7)
75–79	189	147	77.8 (71.2–83.5)	162	78	48.2 (40.2–56.1)
80+	73	62	84.9 (74.6–92.2)	61	39	63.9 (50.6–75.8)
			P = 0.003			P < 0.001
Educational level
Illiterate or no education	168	134	79.8 (72.9–85.6)	154	75	48.7 (40.6–56.9)
Elementary school	333	241	72.4 (67.2–77.1)	309	130	42.1 (36.5–47.8)
Junior high school	175	126	72.0 (64.7–78.5)	167	74	44.3 (36.6–52.2)
Senior high school	209	157	75.1 (68.7–80.8)	193	74	38.3 (31.5–45.6)
College and above	222	154	69.4 (62.9–75.3)	209	86	41.2 (34.4–48.1)
			P = 0.200			P = 0.381
Hypertension
No	637	461	72.4 (68.7–75.8)	594	252	42.4 (38.4–46.5)
Yes	471	352	74.7 (70.6–78.6)	439	188	42.8 (38.1–47.6)
			P = 0.379			P = 0.898
Self-reported diabetes
No	932	684	73.4 (70.4–76.2)	872	376	43.1 (39.8–46.5)
Yes	151	115	76.2 (68.6–82.7)	136	54	39.7 (31.4–48.4)
			P = 0.473			P = 0.454
Smoking
Nonsmoker	704	513	72.9 (69.4–76.1)	660	271	41.1 (37.3–44.9)
Ex-smoker	204	152	74.5 (68.0–80.3)	184	78	42.4 (35.2–49.9)
Current smoker	199	147	73.9 (67.2–79.8)	188	91	48.4 (41.1–55.8)
			P = 0.882			P = 0.199
Alcohol drinking
Nondrinker	912	674	73.9 (70.9–76.7)	850	365	42.9 (39.6–46.3)
Ex-drinker	59	40	67.8 (54.3–79.3)	54	22	40.7 (27.6–55.0)
Current drinker	136	98	72.0 (63.7–79.4)	128	53	41.4 (32.8–50.4)
			P = 0.552			P = 0.909
Nuclear opacity (LOCS III score)
1–2	611	438	71.7 (67.9–75.2)	—	—	—
3–4	460	348	75.7 (71.5–79.5)	—	—	—
5–6	30	22	73.3 (54.1–87.7)	—	—	—
			P = 0.347

Table 5.

 

Multivariate Logistic Regression Models Assessing the Adjusted Odds Ratio (95% CI) of Refractive Errors

	Myopia (SE < −0.5 D)	Hyperopia (SE > +0.5 D)	Astigmatism (Cylinder < −0.5 D)	Anisometropia (SE difference > 0.5 D)
Gender
Male	1.0	1.0	1.0	1.0
Female	1.1 (0.8–1.6)	1.3 (1.0–1.7)*	0.8 (0.6–1.1)	1.1 (0.9–1.5)
Age (y)
65–59	1.0	1.0	1.0	1.0
70–74	1.6 (1.1–2.4)*	0.9 (0.7–1.2)	1.4 (1.0–1.9)*	1.4 (1.0–1.8)*
75–79	2.1 (1.3–3.3)*	0.9 (0.6–1.2)	1.6 (1.1–2.4)*	1.6 (1.1–2.4)*
80+	1.6 (0.8–3.1)	0.8 (0.5–1.3)	2.5 (1.3–5.0)*	3.2 (1.8–5.6), †
Educational level
Illiterate or no education	1.0		1.0
Elementary school	1.2 (0.7–2.1)		0.7 (0.5–1.1)
Junior high school	1.3 (0.7–2.3)		0.6 (0.4–1.1)
Senior high school	1.3 (0.7–2.3)		0.7 (0.4–1.2)
College and above	2.0 (1.1–3.5)*		0.6 (0.3–0.9)*
Nuclear opacity (LOCS III score)
1–2	1.0	1.0
3–4	1.6 (1.1–2.2)*	0.7 (0.6–0.9)*
5–6	8.1 (3.7–17.7), †	0.3 (0.2–0.7)*

Figure 4.

 

Prevalence of myopia and hyperopia by age group in selected population-based surveys that include data from similarly aged elderly groups. Numbers represent the overall prevalence. *Age groups are 65 to 74, 75 to 84, and 85+ years, respectively, for the Beaver Dam Eye Study and the Blue Mountains Eye Study. Age groups have been collapsed into 65 to 69, 70 to 79, and 80+ years, respectively, for the Shihpai Eye Study for comparison. **Data on hyperopia are unavailable.