Prevalence and Risk Factors of Undercorrected Refractive Errors among Singaporean Malay Adults: The Singapore Malay Eye Study

Abstract

purpose. To describe the prevalence and the risk factors of undercorrected refractive error in an adult urban Malay population.

methods. This population-based, cross-sectional study was conducted in Singapore in 3280 Malay adults, aged 40 to 80 years. All individuals were examined at a centralized clinic and underwent standardized interviews and assessment of refractive errors and presenting and best corrected visual acuities. Distance presenting visual acuity was monocularly measured by using a logarithm of the minimum angle of resolution (logMAR) number chart at a distance of 4 m, with the participants wearing their “walk-in” optical corrections (spectacles or contact lenses), if any. Refraction was determined by subjective refraction by trained, certified study optometrists. Best corrected visual acuity was monocularly assessed and recorded in logMAR scores using the same test protocol as was used for presenting visual acuity. Undercorrected refractive error was defined as an improvement of at least 0.2 logMAR (2 lines equivalent) in the best corrected visual acuity compared with the presenting visual acuity in the better eye.

results. The mean age of the subjects included in our study was 58 ± 11 years, and 52% of the subjects were women. The prevalence rate of undercorrected refractive error among Singaporean Malay adults in our study (n = 3115) was 20.4% (age-standardized prevalence rate, 18.3%). More of the women had undercorrected refractive error than the men (21.8% vs. 18.8%, P = 0.04). Undercorrected refractive error was also more common in subjects older than 50 years than in subjects aged 40 to 49 years (22.6% vs. 14.3%, P < 0.001). Non-spectacle wearers were more likely to have undercorrected refractive errors than were spectacle wearers (24.4% vs. 14.4%, P < 0.001). Persons with primary school education or less were 1.89 times (P = 0.03) more likely to have undercorrected refractive errors than those with post–secondary school education or higher. In contrast, persons with a history of eye disease were 0.74 times (P = 0.003) less likely to have undercorrected refractive errors. The proportion of undercorrected refractive error among the Singaporean Malay adults with refractive errors was higher than that of the Singaporean Chinese adults with refractive errors.

conclusions. Undercorrected refractive error is a significant cause of correctable visual impairment among Singaporean Malay adults, affecting one in five persons.

Uncorrected and undercorrected refractive errors are the leading causes of visual impairment worldwide. ¹ Uncorrected refractive error refers to individuals with refractive errors who do not use any form of optical correction (i.e., spectacles or contact lenses) to correct their refractive errors. Undercorrected refractive errors include individuals with uncorrected refractive errors and individuals with refractive errors that are undercorrected by their prescriptions. 

A total of 153 million people are estimated to be visually impaired by uncorrected refractive error. ¹ It is estimated that 59% to 83% of adults with visual impairment in Australia and the United States have undercorrected refractive error. ^{2 3 4 5} However, given that myopia and astigmatism are more common in urban areas of East Asia like Hong Kong, Taiwan, and Singapore, ^{6 7 8 9 10} it is expected that the greatest number of individuals with undercorrected refractive error would be in East Asia. ¹¹ Despite the potential magnitude of this problem in urban areas of East Asia, there are few population-based studies on the prevalence and risk factors of undercorrected refractive error. In Taiwan, the prevalence of undercorrected refractive error among elderly Taiwanese Chinese was estimated at 9.6%. ¹² In the Tanjong Pagar study in Singapore, the prevalence of undercorrected refractive error was 17.3% among Singaporean Chinese persons aged 40 to 79 years. ¹³ These studies involved only Chinese persons and to our knowledge, no data on undercorrected refractive error are available on people with other ethnic backgrounds who live in the urban areas of East Asia. To the best of our knowledge, there are no population-based studies of the prevalence and risk factors of undercorrected refractive error that have been conducted in an adult Malay population living in an urban area. 

In Singapore, while the government provides health subsidies for patients with eye diseases, optometric services are not part of the government health subsidy plan. Individuals with refractive errors have to use their own funds to obtain optical corrections. They may choose to purchase cheaper off-the-shelf spectacles or to visit a private optometrist for a refraction and prescription of an appropriate pair of spectacles that is more expensive. Lack of funds may affect the prevalence rate of undercorrected refractive errors in the Singaporean adult population and may also result in differences in the prevalence rate of undercorrected refractive errors among the various ethnic populations in Singapore. 

The purpose of this study was to document the prevalence and risk factors of undercorrected refractive error in an adult urban Malay population in Singapore. 

Materials and Methods

Study Population

The study population is from the Singapore Malay Eye Study, which was a population-based survey of 3280 Malay adults aged 40 to 79 years living in Singapore that was conducted from August 2004 to June 2006. ^{14 15 16 17 18 19}  

The methodology for the Singapore Malay Eye Study, which adhered to the guidelines of the Declaration of Helsinki, has been described in detail elsewhere. ¹⁴ Briefly, the sampling frame consisted of all Malays aged 40 to 79 living in designated study areas in southwestern Singapore. From a list of 16,069 names provided by the Ministry of Home Affairs, the government agency that supplied the information, age-stratified random sampling was used to select 5,600 names (1,400 people from each decade: 40–49, 50–59, 60–69, and 70–79 years). The target sample size for this study was 3150 persons. Of the 4168 eligible individuals, 3280 (78.7%) took part in our study. 

All subjects underwent a clinic-based examination by trained optometrists and ophthalmologists that included presenting and best corrected visual acuity assessments, refraction measurement, and a full ophthalmic assessment. A detailed interviewer-administered questionnaire was used to collect relevant sociodemographic and medical information (including educational level, marital status, occupation, housing, income, and the presence and details of any eye or systemic diseases), by trained interviewers fluent in Malay. 

Ocular Examination

All subjects were instructed to bring their optical corrections (i.e., spectacles or contact lenses). During the evaluation in the clinic, all subjects were asked to put on their optical corrections, even if they were not wearing them when they entered the clinic. Individuals who had optical corrections but did not bring them to the clinic were categorized as non-spectacle or non–contact lens wearers as it was likely that they did not use their optical corrections habitually if they did not bring their optical corrections with them to the clinic. 

Distance presenting visual acuity was monocularly measured with a logarithm of the minimum angle of resolution (logMAR) number chart (Lighthouse International, New York, NY) at a distance of 4 m, with the participants wearing their “walk-in” optical correction (spectacles or contact lenses), if any. A number chart was used for participants who were illiterate in the Latin alphabet. If no numbers were read at 4 m, the participant was moved to 3, 2, or 1 m, consecutively. If no numbers were identified on the chart, visual acuity was assessed as counting fingers, hand movements, perception of light, or no perception of light. 

Refraction was determined by subjective refraction by trained and certified study optometrists. A trial frame was placed and adjusted on the participant’s face. Autorefraction readings (Auto Ref-Keratometer RK-5; Canon, Tokyo, Japan) were used as the starting point, and refinement of sphere, cylinder, and axis were performed until the best visual acuity was obtained. Best corrected visual acuity was monocularly assessed and recorded in logMAR scores using the same test protocol as that used for presenting visual acuity. Cycloplegic refraction was not performed. 

Definitions

Refractive error was expressed as spherical equivalent (SE; defined as sphere+half negative cylinder). Myopia was defined as an SE of less than −0.50 D, hyperopia as an SE of more than +0.50 D, and astigmatism as cylinder greater than −0.50 D. ¹³ Since the spherical equivalent refraction in the right and left eyes correlated highly (r = 0.84, P < 0.001), the refractive results of the right eyes were used to determine the type of refractive errors. The results of analysis with refractive error defined using left eye data were essentially similar (data not shown). 

Undercorrected refractive error was defined as an improvement of at least 0.2 logMAR (equivalent to 2 lines) in best corrected visual acuity compared with presenting visual acuity in the eye with the better presenting visual acuity. ¹³  

Statistical Analysis

Subjects with bilateral pseudophakia were excluded from the study. In those with unilateral pseudophakia, only the nonpseudophakic eyes were analyzed for refractive error. 

Population prevalence rates of undercorrected refractive error were calculated by direct age and sex standardization to the 2000 population census of Singapore Malays. ²⁰ Proportions were compared by using the χ² test and the means compared by using the t-test, if parametric assumptions were fulfilled. Multiple logistic regression models were developed with undercorrection of refractive error as the dichotomous dependent variable and relevant predictors as covariates. All probabilities quoted were two-sided and were considered statistically significant at P < 0.05. Commercially available statistical software was used (SPSS for Windows, ver. 15; SPSS, Chicago, IL). 

Results

The Singapore Malay Eye Study (SiMES) enrolled 3280 subjects, representing a 78.7% response rate. ¹⁹ Data from 3115 subjects were included in our study and analyzed. Of the 888 nonparticipants, 52.4% were women (compared with 52% of the participants of our study). The mean age of the subjects included in our study was 58 ± 11years. However, the proportion of older people was higher among the nonparticipants (40–49 years [17.4%], 50–59 years [27.5%], 60–69 years [26.2%], and 70–79 years [28.9%]) compared with the participants (40–49 years [24.9%], 50–59 years [29.2%], 60–69 years [23.8%], and 70–79 years [22.1%]). ¹⁹ Of the nonparticipants, 55.1% were aged between 50 and 79 years compared with 45.9% in the participant group, whereas 24.9% of the participants were aged between 40 and 49 years compared with 17.4% in the nonparticipant group. 

The prevalence rate of undercorrected refractive error in our study population is summarized in Table 1 . Of 3115 subjects in our study, 634 (20.4%) had undercorrected refractive error. The calculated age-standardized prevalence rate of undercorrected refractive error is 18.3% (95% confidence interval [CI] 16.7–20.0). More women (21.8%) have undercorrected refractive error than do men (18.8%; P = 0.04). Undercorrected refractive error is also more common in subjects older than 50 years compared with subjects aged 40 to 49 years (P < 0.001, for all the adults). 

Table 2shows that the vision of 14.4% of subjects who wore spectacles or contact lenses is undercorrected (i.e., a gained of two or more lines). Only 10 subjects wore contact lenses. Among those who did not wear spectacles or contact lenses, 24.4% required correction and had obviously completely uncorrected refraction. The proportion with four or more lines gained was 6.7% for all the adults: 3.9% for spectacles or contact lens wearers, and 8.6% for non-spectacle or non–contact lens wearers. Women (P = 0.002) and non-spectacle or non–contact lens wearers (P < 0.001) are more likely to gain more lines of visual acuity than are men and spectacle or contact lens wearers. 

The proportion of undercorrected refractive error was 28.9% in the adults with myopia only, 28.7% for those with hyperopia only, 20.8% in those with astigmatism only, 32.3% in those with myopia and astigmatism, and 32.2% in those with hyperopia and astigmatism (Fig. 1) . The age- and sex-adjusted odds ratio of undercorrected refractive error was higher in the adults with combined astigmatism and myopia or hyperopia compared with the adults with only myopia, hyperopia, or astigmatism (Table 3) . 

Table 4summarizes the multiple logistic regression model of the predictors of undercorrected refractive error. After adjustment for age, educational level, and history of eye disease, the female adult Malays were 1.22 times (P = 0.034) more likely to have undercorrected refractive errors than were the male adult Malays. The adult Malays with primary school education or less are 1.89 times (P = 0.03) more likely to have undercorrected refractive errors than were those with post–secondary school education or higher. Further analysis of sex and educational level revealed that undercorrected refractive error was associated with lower educational levels in the female Singaporean Malay adults, but not in the males (Table 5) . The women with primary or lower education (OR 6.03; 95% CI, 1.87–19.50) and with secondary education (OR 4.49; 95% CI, 1.35–14.91) had higher odds of having undercorrected refractive error than did the women with post–secondary education or higher (P = 0.005). Such associations were not seen in the men (OR 1.25; 95% CI, 0.73–2.15 for primary or lower education and OR 1.51; 95% CI 0.86–2.68 for secondary education). Analysis of the sex–economic relationship with undercorrected refractive error in our population showed that unemployed, retired or homemaker men are more likely to have undercorrected refractive errors compared with employed men (P = 0.045). However, there was no effect of sex when we compared the risk of undercorrected refractive error with economic activity (P = 0.377). 

Subjects with a history of eye disease were 0.74 times (P = 0.003) less likely to have undercorrected refractive errors. The risks of undercorrected refractive error were not significantly related to marital status, type of housing, whether the individual lives alone, individual monthly income, ability to read, or ability to write. The risk of undercorrected refractive error was also not significantly associated with the presence of cataracts, hypertension, or diabetes. 

Discussion

In our study, one in five Singaporean Malay adults aged 40 to 80 years had undercorrected refractive error. This finding expands the data of our previous analysis that showed undercorrected refractive error to be the main cause of both unilateral and bilateral low vision in this population. ¹⁹  

The prevalence rate of undercorrected refractive error in our study and among Singaporean Chinese adults ¹³ was higher than the estimated prevalence rate of undercorrection among other published studies (Table 6) , including Australians aged 49 years and older (10.2% in the Blue Mountains study), ⁴ Latinos aged 40 years and older in the United States (15.1% in the Los Angeles Latino Eye Study), ²¹ and Taiwanese Chinese persons aged 65 years and older (9.55% in the Shihpai Eye Study). ¹² However, we should be cautious when directly comparing results between studies because of different inclusion criteria, definitions used, and sampling methods in the various studies. Most studies define undercorrected refractive errors as individuals with a presenting visual acuity of less than 20/40 in the better eye which improved by 2 lines after refraction. The Victoria Visual Impairment Project (VVIP) has a less stringent definition and defined undercorrected refractive error as individuals with presenting visual acuities worse than 20/20 minus 2 letters which improved by five letters or more (1 line or more of visual acuity) in the better eye after refraction. ²² However, the prevalence of undercorrected refractive error in the VVIP study is still lower than among Singaporean Malays and Singaporean Chinese. The Blue Mountain study assessed subjects older than 49 years, whereas most other studies included at subjects aged 40 years and older. ⁴ An older population should have a higher prevalence of undercorrected refractive error, as increasing age is a risk factor for undercorrected refractive error. Despite this difference, the prevalence of undercorrected refractive error is still higher among Singaporean Malays and Singaporean Chinese compared with Caucasians in the Blue Mountain study. Thus, undercorrected refractive error appears to be a more significant problem among Singaporean Malays and Singaporean Chinese than among Caucasians, Latinos and Mexican Americans. Direct comparison of our study results with those of the Taiwanese Chinese in the Shihpai Eye study was not possible, as the Shihpai Eye study included only Taiwanese Chinese adults aged 65 years and older. ¹²  

Direct comparison of our study with the Tanjong Pagar study which was conducted in Singapore among Chinese persons is possible, however, ¹³ because the definitions, sampling strategies, age ranges of the populations and methodology are similar. After age standardization, the prevalence of undercorrected refractive error in Singapore Malay adults is similar to that in Singaporean Chinese adults (age-standardized prevalence rate was 17.3% in the Tanjong Pagar study). ¹³ Although the prevalence rate of undercorrected refractive error between the Singaporean Chinese and Malay adults are similar, Singaporean Chinese adults have a higher rate of refractive error. ^{18 23} The age-standardized rates of myopia, hyperopia, and astigmatism among Singaporean Chinese adults are 38.7%, 28.4%, and 37.8%, respectively, compared to 30.7%, 27.4%, and 33.3%, respectively, among Singaporean Malay adults. ^{18 23} These data show that the percentage of those with identified need for correction or refractive errors who have undercorrected visual acuity among Singaporean Malay adults is higher than among Singaporean Chinese adults. This suggests that the need for correction is not being identified as well in Singaporean Malays as it is in Singaporean Chinese, and/or that there are higher barriers to correction among Singaporean Malay adults. Figure 2shows that the percentage of undercorrected refractive error among Singaporean Malay adults was higher than in Singaporean Chinese adults for almost all refractive error types. The proportion of Singaporean Chinese adults with higher educational levels was higher compared with the proportion of Singaporean Malay adults. ²⁴ Thus, we postulate that differences in socioeconomic levels and knowledge of refractive errors as a consequence of educational levels may account for the higher percentage of people with an identified need for correction or refractive errors that were not corrected among Singaporean Malay adults. However, further studies may be needed to confirm this postulation. 

Older age, lower education levels and non–spectacle wearers are risk factors of undercorrected refractive error in other published studies (Table 6) . To the best of our knowledge, our study is the only one that shows that the individual’s sex is a risk factor for undercorrected refractive error. Our study identified Singaporean Malay adult women with lower educational level to have the highest risk of undercorrected refractive error in the population. Differences in employment status do not appear to be the cause of the higher risk of undercorrection among Singaporean Malay adult women (Table 5) . Further studies may be needed to determine why the rate of undercorrected refractive error is higher in this group of individuals. 

Our study showed interethnic differences in undercorrection among adults with refractive errors. The possible reasons include interethnic differences in health awareness and access to health care, interethnic differences in socioeconomic status, and interethnic cultural differences. The prior reports of low percentage of undercorrection among those with refractive error in Singapore Chinese are not applicable to the entire population. ¹³ The percentage is higher in Malays, and the problem of undercorrected refractive error among Singaporean Malays with identified need for correction is substantial. 

We believe that the knowledge and beliefs of undercorrected refractive error may be an important factor as to why a person with undercorrected refractive error remains undercorrected. In a separate subset study of the SiMES population (n = 503) on the knowledge and beliefs of refractive errors, 26.6% (n =387) of subjects interviewed do not know that they have refractive errors. ¹⁷ Subjects with undercorrected refractive errors were also more likely to have a lack of knowledge and awareness of refractive error. Thus, lack of awareness, rather than specific physical barriers such as distance from the clinic, may be the primary cause of undercorrected refractive error in Singaporean Malays. A similar study among Singaporean Chinese adults would be useful and may shed more light on why a higher proportion of Singaporean Malay adults with identified need for correction remain undercorrected. 

Subjects with undercorrected refractive error are more likely to report difficulties with their activities of daily living. ^{25 26} Undercorrected refractive error is a significant problem among Singaporean Malay adults that can result in the impairment of their vision. ¹⁹ Lamoureux et al. ¹⁵ showed that severe visual impairment in the worse eye significantly increased the risk of falling (60%; OR 1.6; 95% CI 1.1–2.3) among Singaporean Malay adults. They also showed that visual impairment is significantly associated with poor vision-specific functioning among Singaporean Malay adults. ¹⁶ Undercorrected refractive errors can easily be corrected by wearing the appropriate spectacles. Our study shows that Singaporean Malay adults with lower educational levels are at higher risk of undercorrected refractive error. We postulate that a lower educational level may result in a lower socioeconomic status and reduced knowledge of refractive errors. Both factors may result in a higher prevalence of undercorrected refractive error. Our subset analysis study showed that the lack of knowledge and awareness in this same group of individuals is an important reason that visual acuity in these adults remains undercorrected. ¹⁷ Thus, we recommend that health education programs to educate Singaporean Malay adults on undercorrected refractive error and the importance of wearing appropriate spectacles be promulgated. Such health education programs should be tailored to the Singaporean Malay community. For example, the health promotion messages should be written in Malay and disseminated on the Malay radio and television channels. Engaging prominent Malay personalities to deliver the health promotion messages may also be useful. These programs should be targeted at the high risk groups—Singaporean Malay adults with refractive errors who are older, women, and non–spectacle wearers and those with lower education levels. 

The health education programs should also advise all adults with refractive errors to visit the eye care professional at regular intervals to update their spectacle prescription. In addition, adults with low vision should be advised to seek immediate ophthalmic care to determine the cause and subsequently treat any ocular disorders including undercorrected refractive errors. As optometric service in Singapore is not subsidized by the government, the cost of purchasing an appropriate pair of spectacles from private optometrist may be an important factor in the higher burden of undercorrected refractive error among the Singaporean Malay population. This lack of funding may be a reason why a higher proportion of Singaporean Malay adults with identified need for correction remain without it compared with Singaporean Chinese adults. The availability of inexpensive, subsidized spectacles may help increase the uptake and compliance with full optical correction. 

Our study has several strengths. We are able to assess the prevalence of undercorrected refractive error in a large sample of ethnic Malays in a population-based setting with a high response rate (79% response). We are also able use our data to estimate the prevalence rate of undercorrected refractive error among the general Malay population as our study involved a substantial landmass of Singapore (110.4 km² of Singapore’s total size of 699.4 km², estimated to be approximately 15.8%), included various types of residential areas, and covered a large proportion of the populated area in Singapore. Our study followed standardized protocols based on the Blue Mountains Eye Study which enable us to compare our result with other similar population-based studies. 

However, our study has several limitations. The proportion of older people is higher among the nonparticipants. Thus, it is likely that the actual prevalence rate of undercorrected refractive error in our study was underestimated. Another limitation is that the Malay population in our study is not representative of most of the Malay population in South East Asia, as Singaporean Malays are highly urbanized compared with a large proportion of Malays in South East Asia who live in rural communities. Thus, we are not able to extrapolate our results for all Malays in South East Asia. Our study used manifest noncycloplegic refraction instead of cycloplegic refraction as the method of measurement of refractive error. However, the Tehran Eye Study has shown that manifest noncycloplegic refraction tends to overestimate the rate of myopia and underestimate the rate of hyperopia compared with cycloplegic refraction, especially in the younger age groups and for hyperopia. ²⁷ Noncycloplegic refraction may result in some inaccuracy in determining the refractive status of our subjects. Another limitation is that we did not ask all the undercorrected subjects in our study why they were not using an appropriately prescribed optical correction. Determining why they are not using a prescribed correction may shed more light as to why they remain undercorrected. 

In conclusion, in our study one in five Singaporean Malay adults aged 40 to 80 years had undercorrected refractive error. This is a significant cause of correctable visual impairment among Singaporean Malay adults. ¹⁹ The proportion of undercorrected refractive error among Singaporean Malay adults with refractive errors is higher than in Singaporean Chinese adults with refractive errors. Increasing age, being female, non–spectacle use and lower educational level are significant risk factors for undercorrection among Singaporean Malay adults. Health education programs to increase the awareness of undercorrected refractive error targeted at Singaporean Malay adults, especially those with risk factors, may reduce the prevalence of undercorrected refractive error in this population. 

 Supported by Grant 0796/2003 from the National Medical Research Council (NMRC), Singapore; Grant 501/1/25-5 from the Biomedical Research Council (BMRC), Singapore; the Singapore Prospective Study Program; and the Singapore Tissue Network, A*STAR.

 Submitted for publication September 3, 2008; revised December 15, 2008, and January 25 and February 6, 2009; accepted May 5, 2009.

 Disclosure: M. Rosman, None; T.Y. Wong, None; W.-T. Tay, None; L. Tong, None; S.-M. Saw, 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.

 

Prevalence Rates of Undercorrected Refractive Error

	All							Spectacle or Contact Lens Wearers							Non-spectacle or Non–contact Lens Wearers
	N	n	Prevalence Rate % (95% CI)	P (Trend)	N	n	Prevalence Rate % (95% CI)	P (Trend)	N	n	Prevalence Rate % (95% CI)	P (Trend)
Total	3115	634	20.4 (18.9–21.8)		1250	180	14.4 (12.5–16.3)		1862	454	24.4 (22.4–26.3)
All (y)				<0.001				0.005				<0.001
40–49	810	116	14.3 (11.9–16.7)		211	15	7.1 (3.6–10.6)		599	101	16.9 (13.9–19.9)
50–59	945	201	21.3 (18.7–23.9)		393	54	13.7 (10.3–17.1)		551	147	26.7 (23.0–30.4)
60–69	728	173	23.8 (20.7–26.9)		340	64	18.8 (14.7–23.0)		388	109	28.1 (23.6–32.6)
70–80	632	144	22.8 (19.5–26.1)		306	47	15.4 (11.3–19.4)		324	97	29.9 (25.0–34.9)
3 Missing*
Age standardized			18.3 (16.7–20.0)				11.3 (9.3–13.9)				22.2 (20.1–24.6)
Men (y)				0.002				0.003				0.001
40–49	378	49	13.0 (9.6–16.3)		61	4	6.6 (0.3–12.8)		317	45	14.2 (10.4–18.0)
50–59	424	83	19.6 (15.8–23.4)		154	11	7.1 (3.1–11.2)		269	72	26.8 (21.5–32.1)
60–69	357	76	21.3 (17.0–25.5)		150	28	18.7 (12.4–24.9)		207	48	23.2 (17.4–28.9)
70–80	337	73	21.7 (17.3–26.1)		175	29	16.6 (11.1–22.1)		162	44	27.2 (20.3–34.0)
All men	1496	281	18.8 (16.8–20.8)		540	72	13.3 (10.5–16.2)		955	209	21.9 (19.3–24.5)
1 Missing*
Women (y)				0.001				0.12				<0.001
40–49	432	67	15.5 (12.1–18.9)		150	11	7.3 (3.2–11.5)		282	56	19.9 (15.2–24.5)
50–59	521	118	22.6 (19.1–26.2)		239	43	18.0 (13.1–22.9)		282	75	26.6 (21.4–31.8)
60–69	371	97	26.1 (21.7–30.6)		190	36	18.9 (13.4–24.5)		181	61	33.7 (26.8–40.6)
70–80	295	71	24.1 (19.2–28.9)		131	18	13.7 (7.8–19.6)		162	53	32.7 (25.5–39.9)
All women	1619	353	21.8 (19.8–23.8)		710	108	15.2 (12.6–17.9)		907	245	27.0 (24.1–29.9)
2 Missing*
P (sex)	0.036					0.349					0.010

*  Missing information on spectacle wear (1 man, 2 women).

Table 2.

 

Distribution of the Number of Lines Gained on the LogMAR Chart after Refraction

Number of Lines Gained	Total	Spectacle or Contact Lens Wearers*	Non-spectacle or Non–contact Lens Wearers
All	3115	1250	1862
0	1786 (57.2)	830 (66.4)	955 (51.3)
1 or more	1329 (42.8)	420 (33.6)	907 (48.7)
2 or more	634 (20.4)	180 (14.4)	454 (24.4)
3 or more	332 (10.7)	81 (6.5)	251 (13.5)
4 or more	209 (6.7)	49 (3.9)	160 (8.6)
Men	1496	540	955
0	882 (59.0)	368 (68.1)	514 (53.8)
1 or more	614 (41.0)	172 (31.9)	441 (46.2)
2 or more	281 (18.8)	72 (13.3)	209 (21.9)
3 or more	130 (8.7)	31 (5.7)	99 (10.4)
4 or more	83 (5.5)	20 (3.7)	63 (6.6)
Women	1619	710	907
0	904 (55.8)	462 (65.1)	441 (48.6)
1 or more	715 (44.2)	248 (34.9)	466 (51.4)
2 or more	353 (21.8)	108 (15.2)	245 (27.0)
3 or more	202 (12.5)	50 (7.0)	152 (16.8)
4 or more	126 (7.8)	29 (4.1)	97 (10.7)

 Data are expressed as the number (percentage of the total group).

*  Only 10 subjects wore contact lenses.

Figure 1.

 

Refractive status of subjects with undercorrected refractive error versus subjects who do not have undercorrected refractive error.

Table 3.

 

Age- and Sex-Adjusted Odds Ratio of Undercorrection versus Type of Refractive Error

	At Risk (n)	Undercorrected n (%)	Age–Sex OR (95% CI)	P
Refractive Error				0.001
Myopia only	277	80 (28.9)	1.0 (Reference)
Hyperopia only	662	190 (28.7)	0.99 (0.72–1.36)
Astigmatism only	365	76 (20.8)	0.64 (0.44–0.93)
Myopia and astigmatism	399	137 (34.3)	1.29 (0.92–1.81)
Hyperopia and astigmatism	366	118 (32.2)	1.16 (0.81–1.65)

Table 4.

 

Logistic Regression Model of the Predictors of Undercorrected Refractive Error

Characteristics	N	n (%)	Age–Sex OR (95% CI)*	P	Multivariate OR (95% CI), †	P
Age (y)	3115	634 (20.4)	1.02 (1.01–1.03)	<0.001	1.02 (1.01–1.03)	<0.001
Sex				0.016		0.034
Male	1496	281 (18.8)	1.0		1.0
Female	1619	353 (21.8)	1.24 (1.04–1.48)		1.22 (1.02–1.45)
Education				0.020		0.030
Primary or lower	2311	505 (21.9)	1.97 (1.22–3.16)		1.89 (1.18–3.05)
Secondary	491	108 (18.3)	1.89 (1.15–3.11)		1.87 (1.13–3.07)
Post secondary or higher	208	21 (10.1)	1.0		1.0
Income				0.061		0.093
<S$1000	1709	373 (21.8)	1.43 (0.99–2.06)		1.37 (0.95–1.98)
S$1000–S$2000	649	99 (15.3)	1.09 (0.75–1.59)		1.87 (0.72–1.54)
>S$2000	346	47 (13.6)	1.0		1.0
History of eye disease				0.002		0.003
Yes	890	157 (17.6)	0.73 (0.59–0.99)		0.74 (0.60–0.90)
No	2220	475 (21.4)	1.0		1.0

*  Adjusted for age and sex.

†  Adjusted for age, sex, education, income, and history of eye disease.

Table 5.

 

Relationship of Education and Economic Activity and Undercorrected Refractive Error, by Sex

Characteristics	N	n (%)	Multivariate OR (95% CI)*	P	P Interaction, †
Education
Women				0.005
Primary or lower	1278	305 (23.9)	6.03 (1.87–19.50)
Secondary	269	45 (16.7)	4.49 (1.35–14.91)
Post secondary or higher	72	3 (4.2)	1.0		0.014
Men				0.300
Primary or lower	1033	200 (19.4)	1.25 (0.73–2.15)
Secondary	322	63 (19.6)	1.51 (0.86–2.68)
Post secondary or higher	136	18 (13.2)	1.0
Economic activity
Women				0.511
Employed	424	76 (17.9)	1.0
Retired or homemaker	1104	259 (23.5)	1.16 (0.85–1.58)
Unemployed	89	18 (20.2)	0.94 (0.52–1.70)		0.377
Men				0.045
Employed	954	152 (15.9)	1.0
Retired or Homemaker	417	96 (23.0)	1.30 (0.90–1.88)
Unemployed	123	33 (26.8)	1.75 (1.11–2.75)

*  Adjusted for age

†  Sex–education interaction term or sex–economy activity interaction term.

Table 6.

 

Comparison of Prevalence and Risk Factors of Undercorrected Refractive Error in Selected Population-Based Studies

Study (Year of Study)	Ethnicity	n	Age (y)	Undercorrection in Study Population (%)	Undercorrection among Participants with Refractive Error (%)	Estimated Undercorrection in General Population (%)	Definitions	Multivariate Risk Factors
LALES 21 (2000–3)	Latinos	6129	≥40	15.1%	NA	NA	(Best corrected VA − presenting VA) ≥2 lines improvement in better eye	Older age, lack of health insurance, lower education, lower BMI, and being unemployed
Proyecto VER 3 (1997–9)	Mexican Americans	4774	≥40	6.3%*	NA	NA	Presenting visual acuity <20/40 in the better eye (Best corrected VA − presenting VA) ≥2 lines improvement in better eye	Older age, less than high school education, low index of acculturation, no health insurance coverage in the past year, and not seeing an eye-care provider in the past 2 years
BMES 4 (1992–4)	Caucasians	3654	≥49	10.2%	53.9% of hypermetropic subjects	NA	Presenting visual acuity <20/40 (Best corrected VA − presenting VA) improvement of ≥10 letters (2 lines on the logMAR chart) in the better eye	Increasing age, living alone, occupations of trade and laborer, receipt of a government pension, hyperopia and duration from the last eye examination
					12.7% of myopic subjects
VVIP 22 (1992–6)	Caucasians	4735	≥40	10%	7.5% of hypermetropic subjects	NA	Presenting visual acuity <20/20 minus two letters (Best corrected VA − presenting VA) ≥1 lines improvement in better eye	Increasing age, main language spoken at home, lower education level, longer time since last eye examination, not wearing any distance correction lenses, type of refractive error and presence of ocular disease condition
					11.1% of myopic subjects
Shihpai Eye Study 12 (1999–2000)	Taiwanese Chinese	1361	≥65	9.55% (8.2%, †)	11.3% of hypermetropic subjects	NA	Presenting visual acuity <20/40 in the better eye (Best corrected VA − Presenting VA) ≥2 lines improvement in better eye	Older age, nonemmetropic eye, not wearing distance eyeglasses during examination and lower level of education
					34.4% of myopic subjects, ‡
Tanjong Pagar Survey 13 (1997–8)	Singaporean Chinese	1152	40–79	21.7%	NA	17.3%	Presenting visual acuity <20/40 in the better eye (Best corrected VA − Presenting VA) ≥2 lines improvement in better eye	Older age, fewer years of education, not wearing spectacles, cataracts
SiMES (Current)	Singaporean Malays	3115	40–80	20.4%	28.7% of hypermetropic subjects	18.3%	Presenting visual acuity <20/40 in the better eye (Best corrected VA − Presenting VA) ≥2 lines improvement in better eye	Older age, female sex, lower education level
					28.9% of myopic subjects

 LALES, Los Angeles Latino Eye Study; Proyecto VER, Proyecto Vision and Eye Research (US Hispanic study); VIP, Victoria Visual Impairment Project; BMES, Blue Mountains Eye Study; BDES; SiMES, Singapore Malay Eye Study.

*  Percentage of uncorrected refractive error was calculated from the paper.

†  Percentage of subjects with correctable visual impairment who had improvement of at least 2 lines after refractive correction.

‡  Percentage was calculated from the paper.

Figure 2.

 

Comparison of the percentage of undercorrected refractive error among Singaporean Chinese adults (TPGS) and Singaporean Malay adults (SiMES).