Myopic Refractive Error in Adult Latinos: The Los Angeles Latino Eye Study

Abstract

purpose. To estimate gender- and age-specific prevalence of myopia and identify risk indicators for myopia in a population-based sample of Latino adults aged 40 years and older in La Puente, California.

methods. Noncycloplegic automated refraction with supplemental standardized subjective refraction was performed for presenting visual acuity worse than 20/20. Potential risk indicators for myopia were evaluated by questionnaire and clinical examination. The gender- and age-specific prevalence of spherical equivalent myopia in phakic eyes was calculated. Multiple logistic regression identified independent risk indicators for myopia.

results. Refractive error was analyzed for the worse eye of 5927 of 6357 participating Latinos. The overall prevalence of myopia ≤ −1.0 D was 16.8%, and of high myopia ≤ −5.0 D, 2.4%. When participants were stratified by age, myopia initially decreased with age and then increased in the oldest groups in association with nuclear opacification. Risk indicators for myopia were: oldest and youngest age groups, presence of diabetes mellitus, high acculturation, high school or higher education, and birth in the United States. The latter two were also predictors of high myopia.

conclusions. Risk indicators for myopia in adult Latinos include higher education, birth in the United States, high acculturation, and diabetes. The burden of myopia in older Latinos exceeds that in African Americans and non-Hispanic whites in the United States.

The Los Angeles Latino Eye Study (LALES) is a population-based ophthalmologic survey of adult Latinos in the Los Angeles metropolitan area. Latinos are the fastest-growing minority in the United States, projected to constitute more than 24% of the population by 2050, ¹ and so it is important to gather basic information on the prevalence of eye disease in this population. Refractive error (RE) is of special interest because it is an easily correctable cause of visual disability and a significant cause of visual impairment worldwide ^{2 3} ; myopia in particular is important because of associated retinal degeneration and retinal detachment, ⁴ and potentially modifiable environmental risk factors. ⁵ Furthermore, the prevalence of myopia is apparently increasing in younger generations, especially in East Asia, ^{5 6} but also in the United States. ⁷ Although many population-based studies worldwide have examined adult RE, ^{8 9 10 11 12 13 14 15 16 17 18 19} only one has studied U.S. Latinos, ²⁰ and the RE analysis does not address risk indicators. ²¹ The purpose of this article is to describe the burden and distribution of myopia in a Los Angeles Latino population, providing insight into potentially modifiable risk factors. Data on the ocular biometric determinants of RE in LALES are published separately. ²²  

Methods

Study Population

The LALES population consists of residents in six census tracts in La Puente, California, ≥40 years old and of self-identified Latino ethnicity. Eligible residents electing to participate after informed consent reviewing the nature and possible consequences of the study were interviewed at home and examined in the clinic. The study was approved by the Los Angeles County/University of Southern California Medical Center Institutional Review Board and adhered to the tenets of the Declaration of Helsinki for research in human subjects. 

Data Collection

The in-home interview reviewed country of birth, education, marital status, working status, income, acculturation, smoking/alcohol use, healthcare utilization, ocular disease, comorbid medical conditions (self-reported diabetes, arthritis, stroke/brain hemorrhage, hypertension, angina, heart attack, heart failure, asthma, skin cancer, other cancer, back problems, hearing problems, or another major health problems), and health and vision insurance. Acculturation was measured with a nine-item questionnaire modified from Cuellar et al. and others, ^{23 24} that recorded Spanish, English, and preferred language for speech, reading and writing and scored the results on a 5-point scale (5, most acculturated). After the interview, participants received a comprehensive standardized eye examination at the Local Eye Examination Center in La Puente, including visual acuity (VA), refraction, slit lamp examination with LOCS II lens opacity grading, ²⁵ dilated fundus examination, intraocular pressure, visual fields, and ocular biometry. Height, weight, blood pressure, serum glucose, and glycosylated hemoglobin were also assessed. Additional interview and examination details are reported elsewhere. ²⁶  

Refractive Error Measurement

If presenting VA was 20/20 (55 letters) in each eye according to the Early Treatment Diabetic Retinopathy Study protocol at 4 m, refraction was the presenting spectacle correction or plano. Otherwise, VA was remeasured after noncycloplegic autorefraction, followed by standardized subjective refraction for VA <20/20 (at 1 m for patients unable to read letters at 4 m). If refractions yielded identical VA, subjective refraction took precedence. 

Definitions of Outcome Variables

Myopia was defined by the spherical equivalent (SE) refractive error (RE) in the worse eye (phakic eye with larger absolute RE, or phakic eye with RE if there was only one). This definition was used for comparison with Eye Disease Prevalence Research Group (EDPRG) ²¹ data, and risk indicator analyses. Any myopia (≤ −1.0 D SE) and high myopia (≤ −5.0 D SE) prevalence was stratified by age and gender using RE of the worse eye, right eye (for phakic right eyes), or better eye (defined as phakic eye RE, for only one phakic eye, or else RE of the worse eye’s fellow eye). For comparison with Bangladeshi ¹⁷ and Singaporean ¹⁴ populations, myopia was defined as < −0.5 D, and high myopia was < −5.0 D SE. 

Statistical Analyses

Prevalence was calculated as the ratio of the number of individuals with a given level of myopia to the total number of individuals whose myopia status was known, using a given definition (worse eye, better eye, or right eye). Data were stratified by age, gender, and selected variables identified as risk indicators. Myopia risk indicators were explored using univariate and multiple logistic regression. Sociodemographic factors evaluated included age, gender, birth country (United States, other), education level (0–6, 7–11, ≥12 years), marital status (married/with stable partner, never married, widowed, separated/divorced), working status (used, retired, unemployed), annual income (<$20,000, $20,000–$40,000, >$40,000), acculturation (low, ≤1.9, high, >1.9), smoking and alcohol history (none, former, current), healthcare utilization (visit within past year), and health-vision insurance. Clinical factors included hypertension (by history, or LALES evaluation), diabetes (determined by history, by hemoglobin A1c ≥6.5%, or by random blood glucose ≥200 mg/dL), body mass index (normal, ≤24.9, overweight, 24.0 to 29.9; obese >29.9), self-reported comorbid nonocular conditions (0, 1, ≥2), and self-reported history of ocular disease (diabetic retinopathy, cataract, age-related macular degeneration or glaucoma). Univariate logistic regression analyses were first performed for all of these sociodemographic and clinical risk indicators. Variables showing at least marginally significant associations (P < 0.1) in univariate regression were considered candidates for subsequent forward stepwise multiple logistic regression. Odds ratios are reported for the significant independent risk indicators included in the final model. To further control for confounding by nuclear lens opacification (NO) of associations of age and ocular disease with myopia, age/gender-stratified myopia prevalence and risk indicators were also assessed in a cohort from which all participants with NO (LOCS II grade ≥2 either eye) were excluded. All analyses were conducted on computer (SAS software; SAS Institute, Cary, NC) and a significance level of P < 0.05. 

Results

Study Cohort

Of 7789 eligible La Puente residents, 6357 (82%) participated and were examined. Of those who were not examined, a third completed an in-home interview. Compared with individuals who were examined, those who were only interviewed were more often male, U.S.-born, more acculturated, and more educated, having higher incomes, more health insurance, and more previous cataract surgery. ²⁶  

Of the 6357 participants, 5927 (93%) were included in RE analysis based on the worse eye. Of those, 5396 remained in the cohort after exclusion of NO. Ninety-five percent of the participants were of Mexican-American self-reported ancestry (5% Native American), 76% were born outside the United States, and 58% were female. 

Prevalence of Myopia and High Myopia

The overall prevalence of myopia ≤ −1.0 D in the worse eye was 16.8% (males, 16.1%; females, 17.2%). High myopia ≤ −5.0 D was present in 2.4% of individuals (males, 2.1%; females, 2.6%). Age stratification showed an initial slight decrease in myopia prevalence with increasing age, followed by higher prevalence in older groups (Tables 1 2 ; Figs. 1 2 ). Myopia prevalence in the better eye and in the right eye are also presented (Tables 1 2) . The Pearson coefficient of correlation between right and left eye spherical equivalent refractive error was 0.81. 

Comparison to raw data from African Americans and non-Hispanic whites (NHWs) from the Baltimore Eye Survey (BES), ⁹ NHWs from the Beaver Dam Eye Study (BDES), ⁸ Chinese residents of Singapore from the Tanjong Pagar Survey, ¹⁴ and participants in the Bangladesh National Blindness and Low Vision Survey (NBLVS) ¹⁷ is shown in Figures 1 and 2 , using comparable myopia definitions in all instances (raw data for Hispanics from Proyecto VER ^{20 27} were not available). 

The higher myopia prevalence in older Latinos is reduced if subjects with NO are excluded from analysis (Fig. 3) . 

Risk Indicators for Any Myopia and High Myopia

Stratification of myopia prevalence (≤ −1.0 D) by age, gender and either education, acculturation, country of birth, or diabetes (selected for their independent associations with myopia; described later) is shown in Table 3 . These factors increased myopia risk primarily in younger age groups. 

For myopia ≤ −1.0 D, stepwise multiple logistic regression identified the following independent risk indicators: age (lower risk of myopia in 55- to 69-year-olds than in younger or older groups), education, country of birth, acculturation, ocular disease history, and diabetes (Table 4 , showing odds ratios for all risk indicators included in the final model). When the same analysis was performed after excluding participants with NO, ocular disease was no longer significantly associated (Table 4) . The other factors remained significant, whereas females became more likely to be myopic. 

High myopia risk indicators were assessed by comparing subjects with no myopia with those with high myopia (excluding myopia between −1.0 D and −5.0 D). Multiple logistic regression identified education, country of birth, and ocular disease history as significant associations (Table 4) . After exclusion of NO, ocular disease was no longer associated, whereas education and country of birth remained significant, and retired individuals became less likely to be myopic (Table 4) . 

Discussion

The prevalence of myopia (≤−1.0 D in the eye with greater refractive error) in this population of Latinos over 40 years of age ranged between 10% and 20% for men and women, except in the oldest groups. High myopia (≤−5.0 D) was present in less than 3% of the population. 

The strongest risk indicators for myopia were birth in the United States, more education, greater acculturation, diabetes, and (unless participants with NO were excluded) a history of ocular disease; intermediate age groups were at lower risk. After excluding NO, females were at higher risk than males. For high myopia, significant associations included birth in the United States, education level, and (unless NO was excluded) a history of ocular disease; retirement became protective after exclusion of NO. 

When myopia was stratified by age, the initial slight decrease in prevalence with increasing age echoed the aforementioned protective value of intermediate age (discussed further later in the article). The subsequent increase in prevalence at older ages, is reduced when participants with NO were excluded from the analysis. 

LALES myopia prevalence was compared with data from other Latinos and other ethnic-racial groups in the United States reported by the EDPRG. ²¹ We compared our data with those for Hispanics from Proyecto VER ^{20 27} in Arizona, African Americans from the BES, ⁹ and NHWs from the BES ⁹ and the BDES. ⁸ Lacking population-based data on U.S. Asians, we also considered studies from East Asia and the Indian subcontinent: the Tanjong Pagar survey of adult Chinese in Singapore, ¹⁴ and the Bangladeshi NBLVS. ¹⁷  

Myopia ≤ −1.0 D was less prevalent in LALES Latinos than in NHWs from the fifth to the seventh decade, especially at younger ages. High myopia was also less common among the younger Latinos. This may relate to environmental differences such as duration of education, particularly among younger persons: only 39% of LALES participants aged 40 to 54 years had ≥12 years of schooling, compared with 89% of comparable BDES participants ⁸ (among those aged ≥75 years, the gap is smaller, with 45% in BDES and 30% in LALES having ≥12 years). Among young Latinos with ≥12 years’ schooling, myopia prevalence is close to that in BES NHWs, though still lower than in the youngest BDES participants; residual differences after controlling for secondary education may relate to patterns of tertiary education (not studied in LALES). 

By contrast, at older ages, myopia was more prevalent in LALES Latinos than in NHWs. Compared with BES African Americans and Hispanics from Proyecto VER (EDPRG report ²¹ ), LALES participants had similar myopia prevalence at younger ages, but again showed higher prevalence at older ages. Since the higher prevalence of myopia in older persons may be related to NO, differences between population groups may be related to differences in the rates of lens opacification and/or rates of cataract surgery. Because studies differ in lens opacity classification methods, even when the rates of cataract surgery are reported, this possibility cannot be directly evaluated. ^{28 29}  

In noting that a higher myopia prevalence with older age is associated with NO, the LALES data qualitatively resemble findings from the Bangladeshi NBLVS ¹⁷ : in Bangladesh there was a steep increase in myopia prevalence with age, even more pronounced and beginning earlier than in LALES Latinos, and myopia prevalence at older ages was only halved if participants with NO were excluded from the analysis. 

Compared with Chinese adults in Singapore, ¹⁴ myopia is less prevalent in LALES Latinos at all ages, although this population, like LALES, showed an increase in myopia prevalence with age, from the sixth decade onward. A high prevalence of myopia in East Asia is hypothesized to reflect an interaction of environmental factors with a genetically determined predisposition toward myopia. ^{5 30} Evidence for a genetic influence comes from twin studies ^{31 32 33 34} and other heritability studies, ³⁵ and supports the notion that broad genetic differences between racial/ethnic groups could partially account for disparities in myopia prevalence between populations. However, from our current observations, we cannot distinguish the relative contributions of environmental and genetic differences to the observed variation in myopia prevalence between the LALES and Singaporean populations. 

Of the myopia risk indicators in LALES, several are familiar from other studies. Education as a myopia risk factor is described in numerous population-based studies. ^{8 9 10 11 12 13 14 16 17 36 37 38} This correlation may reflect an underlying association between near work (study-related reading/computer work) and myopia, ^{39 40 41 42} with the driving force being retinal defocus resulting from prolonged accommodative effort, either as an after-effect (reading-induced defocus for distance from failure to relax accommodation after near work) ^{43 44} or from inaccurate accommodation during near work. ^{45 46} Alternatively, because even brief periods of myopic defocus protect against development of myopia in animal models, ^{47 48} perhaps a relative lack of distance visual activities rather than near work per se underlies the association between education and myopia. 

In LALES Latinos, U.S. birth and high acculturation increased the risk of myopia, independent of education level. These correlations may represent environmental factors unrelated to education. Alternatively, however, these variables may be surrogate measures for some aspect of education not captured by our years-of-schooling measure. For instance, we did not evaluate tertiary education. We also did not consider the country where schooling occurred: different national education systems may have different schooling-related risks for myopia development, because of differences in curriculum, educational media used (e.g., visual displays ⁴⁰ ), time devoted to indoor-versus-outdoor activities, and academic competitiveness. Birth country and acculturation may capture the impact of country of education. The fact that the influence of birth country or acculturation is most pronounced in younger age groups, as is the influence of education, is compatible with the idea that acculturation and country of birth may be associated indirectly with myopia through education. 

The association of self-reported ocular disease (which included cataract) with myopia and high myopia in LALES can be attributed to NO-related myopic shift, since it is reduced with exclusion of NO. NO is well-known to be associated with myopia in cross-sectional studies, ^{11 12 13 17 19 37 38 49 50} and longitudinal studies show that, although people on average get more hyperopic with age, the subset with NO at baseline show a myopic shift. ^{7 51 52} However, the impact on cross-sectional, age-related prevalence patterns depends on several factors. Populations vary in the prevalence of nonsurgically treated NO, whose association with both age and myopia makes myopia prevalence increase with age, but also in the magnitude of cohort effects (e.g., related to education) that result in higher prevalence of myopia in younger age groups, ⁷ causing an opposing trend with age. High cataract burden without large cohort effects can result in marked age-related increases in myopia prevalence in some countries, ^{11 13 17 19} contrasting with a pattern of decreasing myopia prevalence with age in cross-sectional data from the United States, Western Europe, and Australia. ²¹ The LALES population falls midway between these extremes. 

Diabetes was found to be an independent risk indicator for myopia in LALES, although studies in other populations, ^{11 16 19} including the BDES, ⁸ have not observed this association; indeed, in the BDES, diabetics underwent a hyperopic shift over time, ⁷ and diabetes was unrelated to refractive change in the Blue Mountains Eye Study. ⁵² The discrepant findings may reflect differences among diabetic subsets of different populations. For example, the higher frequency of very elevated hemoglobin A1c among BDES diabetic participants compared with LALES (Varma R, unpublished data, 2003) may be associated with hyperglycemia-induced hypermetropic shifts in RE, ^{53 54} masking any tendency for diabetes to cause myopia (although if hyperglycemia induces myopic RE shifts in diabetics, as in healthy volunteers, ⁵⁵ glycemic control cannot account for the differences observed). Alternatively, it may be that diabetic Latino individuals are more susceptible to some myopia-inducing aspect of the diabetic disease process, just as they are more susceptible to diabetic retinopathy than NHWs, having twice the risk of diabetic retinopathy as NHWs even after adjustment for known risk factors (Varma R, et al. IOVS 2005;46:ARVO E-Abstract 1164). 

The finding that younger age groups were at relatively higher risk of myopia, even in a model that adjusted for education and thereby eliminated potential education-related age cohort effects, implies one of the following: Our model does not entirely capture the effect of education (e.g., there may be confounding by tertiary education), or there are other unidentified cohort effects at work, or the difference in risk reflects true longitudinal age-related change. ⁵⁶ One source of longitudinal change could be decreasing accommodative tone and/or amplitude with aging. Changes in accommodation have been evoked to explain age-related RE distribution shifts in both cross-sectional and longitudinal studies which, like LALES and most adult studies of RE epidemiology, used noncycloplegic measures of RE. ^{21 51} Loss of accommodation could cause not only age-related increases in manifest hypermetropia, but also age-related decreases in apparent myopia related to incomplete relaxation of accommodation during refraction; although such effects are likely to be small in adults over 40 years of age, data from the Tehran Eye Study ¹⁸ suggest that unimpeded accommodation may contribute to small amounts of pseudomyopia even in adulthood. 

After exclusion of participants with NO, two additional myopia risk indicators emerged from our analysis: female gender, for any myopia, and nonretirement, for high myopia. Findings from other population-based studies on the influence of gender have been mixed, but several have reported higher age-adjusted myopia prevalence in women, ^{8 14 18} and one study found, like LALES, higher prevalence in women only after exclusion of NO. ³⁷ With regard to the protective effect of retirement, work status is probably not directly related to high myopia, but it may be indirectly related through an environmental risk factor, such as type of profession, ^{14 16} that was not evaluated in our study; alternatively, the association may have arisen by chance. 

Although LALES has the advantage of being a population-based study employing standardized procedures, our prevalence estimates could be affected by participation biases. Nonparticipants were more likely than participants to be U.S.-born, educated, and acculturated, all of which, based on our analysis, would bias us toward underestimation of prevalence. However, given our high participation rate, any such bias is unlikely to be large. The findings from our study are applicable to Latinos of Mexican-American descent, and may not generalize to all Latino-Americans. Noncycloplegic refraction may have caused some myopia overestimation in younger age groups (see the Discussion section); however, this effect is probably small in adults aged ≥40 years. Furthermore, noncycloplegic refraction more closely reflects actual distance refractive needs than does cycloplegic refraction (especially for manifest hypermetropia). Finally, noncycloplegic refraction makes our findings comparable to most other adult population-based epidemiologic studies of RE. 

In summary, myopia is moderately prevalent among adult Latinos in Los Angeles. Higher rates among older Latinos are associated with nuclear lens opacification in older individuals. Strong demographic indicators of increased risk for both high myopia and any myopia among Latinos are a higher level of education, birth in the United States, and a greater degree of acculturation. Diabetes is also a significant independent risk indicator for myopia in this population. 

Appendix 1

The Los Angeles Latino Eye Study Group, University of Southern California, Los Angeles, CA: Rohit Varma, Sylvia H. Paz, LaVina Abbott, Stanley P. Azen, Lupe Cisneros, Elizabeth Corona, Carolina Cuestas, Denise R. Globe, Sora Hahn, Mei Lai, George Martinez, Susan Preston-Martin, Ronald E. Smith, Mina Torres, Natalia Uribe, Jennifer Wong, Joanne Wu, and Myrna Zuniga. 

Battelle Survey Research Center, St. Louis, MO: Sonia Chico, Lisa John, Michael Preciado, and Karen Tucker. 

Ocular Epidemiology Grading Center, University of Wisconsin, Madison, WI: Ronald Klein. 

⁴  See ^Appendix for members of the Los Angeles Latino Eye Study Group.

 Supported by Grants EY11753 and EY03040, the National Eye Institute and the National Center on Minority Health and Health Disparities, and an unrestricted grant from Research to Prevent Blindness, New York, New York. RV is a Research to Prevent Blindness Sybil B. Harrington Scholar.

 Submitted for publication August 29, 2005; revised January 4, 2006; accepted March 3, 2006.

 Disclosure: K. Tarczy-Hornoch, None; M. Ying-Lai, None; R. Varma, 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.

 

Age- and Gender-Specific Prevalence of Myopia

Age Group	Worse Eye			Better Eye			Right Eye
		n	% Prevalence (95% CI)	n	% Prevalence (95% CI)	n	% Prevalence (95% CI)
Males
40–44	478	16.53 (13.31–20.17)	474	11.18 (8.49–14.37)	474	14.14 (11.13–17.60)
45–49	494	17.81 (14.54–21.48)	494	11.74 (9.04–14.91)	492	15.45 (12.37–18.95)
50–54	459	14.60 (11.49–18.16)	457	10.94 (8.23–14.17)	457	12.69 (9.78–16.10)
55–59	304	12.83 (9.28–17.12)	303	8.58 (5.68–12.32)	300	11.00 (7.69–15.10)
60–64	281	11.74 (8.22–16.10)	281	6.41 (3.84–9.94)	274	8.76 (5.69–12.75)
65–69	193	13.99 (9.43–19.70)	193	9.33 (5.62–14.34)	189	11.64 (7.44–17.09)
70–74	142	20.42 (14.12–28.00)	141	11.35 (6.63–17.77)	140	17.86 (11.90–25.22)
75–79	92	26.09 (17.48–36.29)	92	13.04 (6.93–21.68)	91	19.78 (12.16–29.45)
80+	36	38.89 (23.14–56.54)	36	33.33 (18.56–50.97)	35	34.29 (19.13–52.21)
Total	2479	16.14 (14.71–17.64)	2471	10.64 (9.45–11.93)	2452	13.66 (12.33–15.09)
Females
40–44	698	19.48 (16.61–22.62)	696	14.80 (12.24–17.66)	697	17.93 (15.15–20.99)
45–49	667	19.04 (16.13–22.23)	664	11.45 (9.12–14.12)	666	16.07 (13.36–19.08)
50–54	610	16.89 (14.00–20.10)	609	11.66 (9.22–14.48)	608	13.82 (11.17–16.82)
55–59	458	13.10 (10.15–16.54)	455	7.69 (5.42–10.54)	454	10.57 (7.90–13.77)
60–64	391	14.07 (10.78–17.91)	391	6.39 (4.18–9.29)	388	10.05 (7.25–13.48)
65–69	286	15.73 (11.71–20.48)	286	7.69 (4.88–11.41)	281	13.17 (9.44–17.69)
70–74	175	18.86 (13.35–25.45)	175	9.71 (5.76–15.10)	172	13.95 (9.15–20.05)
75–79	100	20.00 (12.67–29.18)	98	12.24 (6.49–20.41)	96	14.58 (8.21–23.26)
80+	63	23.81 (13.98–36.21)	63	12.7 (5.65–23.50)	59	16.95 (8.44–28.97)
Total	3448	17.23 (15.98–18.53)	3437	10.74 (9.72–11.82)	3421	14.26 (13.11–15.48)

 Prevalence is shown by gender and age group of myopia ≤ −1 D, with refractive error defined according to the worse eye, better eye, or right eye. Better eye/worse eye sample size differences are because the refracted eye in bilaterally phakic participants with refraction missing for one eye was defined as the worse eye.

Table 2.

 

Age- and Gender-Specific Prevalence of High Myopia

Age Group	Worse Eye			Better Eye			Right Eye
		n	% Prevalence (95% CI)	n	% Prevalence (95% CI)	n	% Prevalence (95% CI)
Males
40–44	478	1.67 (0.73–3.27)	474	1.05 (0.34–2.44)	474	1.48 (0.60–3.02)
45–49	494	2.43 (1.26–4.20)	494	1.62 (0.70–3.17)	492	2.24 (1.12–3.97)
50–54	459	2.40 (1.20–4.25)	457	1.31 (0.48–2.84)	457	1.75 (0.76–3.42)
55–59	304	1.64 (0.54–3.80)	303	0.99 (0.20–2.87)	300	1.67 (0.54–3.85)
60–64	281	1.42 (0.39–3.60)	281	1.07 (0.22–3.09)	274	0.73 (0.09–2.61)
65–69	193	2.07 (0.57–5.22)	193	1.55 (0.32–4.48)	189	1.59 (0.33–4.57)
70–74	142	2.11 (0.44–6.05)	141	0.71 (0.02–3.89)	140	2.14 (0.44–6.13)
75–79	92	4.35 (1.20–10.76)	92	2.17 (0.26–7.63)	91	3.30 (0.69–9.33)
80+	36	5.56 (0.68–18.66)	36	2.78 (0.07–14.53)	35	5.71 (0.70–19.16)
Total	2479	2.14 (1.61–2.79)	2471	1.30 (0.89–1.82)	2452	1.79 (1.31–2.40)
Females
40–44	698	3.15 (1.99–4.73)	696	2.44 (1.43–3.88)	697	2.73 (1.65–4.22)
45–49	667	2.40 (1.38–3.87)	664	1.20 (0.52–2.36)	666	1.65 (0.83–2.94)
50–54	610	2.79 (1.63–4.42)	609	2.30 (1.26–3.83)	608	2.47 (1.39–4.04)
55–59	458	1.31 (0.48–2.83)	455	0.66 (0.14–1.91)	454	1.32 (0.49–2.85)
60–64	391	2.56 (1.23–4.65)	391	0.51 (0.06–1.84)	388	2.06 (0.89–4.02)
65–69	286	4.20 (2.19–7.21)	286	0.35 (0.01–1.93)	281	2.85 (1.24–5.53)
70–74	175	2.86 (0.93–6.54)	175	2.29 (0.63–5.75)	172	2.91 (0.95–6.65)
75–79	100	3.00 (0.62–8.52)	98	2.04 (0.25–7.18)	96	2.08 (0.25–7.32)
80+	63	0.00 (0.00–5.69)	63	0.00 (0.00–5.69)	59	0.00 (0.00–5.69)
Total	3448	2.64 (2.13–3.23)	3437	1.48 (1.11–1.95)	3421	2.16 (1.70–2.71)

 Prevalence is shown by gender and age group of high myopia ≤ −5 D, with refractive error defined according to the worse eye, better eye, or right eye.

Figure 1.

 

Age-related prevalence of myopia in males (A, B) and females (C, D). In (A) males and (C) females, myopia was defined as ≤ −1.0 D spherical equivalent refractive error according to the phakic eye with the larger absolute value spherical equivalent refractive error for participants in LALES and for other racial-ethnic groups, in the United States, as reported in the refractive error study of the EDPRG. ²¹ In (B) males and (D) females, myopia was defined as < −0.5 D spherical equivalent refractive error in the right eye for phakic eyes for LALES and for adult Chinese participants in the Tanjong Pagar survey ¹⁴ and the National Blindness and Low Vision Survey of Bangladesh. ¹⁷ Note that the data from Bangladesh for the 70- to 79-year-old age group includes all participants over 70 years of age.

Figure 2.

 

Age-related prevalence of high myopia in males (A, B) and females (C, D). In (A) males and (C) females, myopia was defined as spherical equivalent refractive error of ≤ −5.0 D according to the phakic eye with the larger absolute value spherical equivalent refractive error for participants in LALES and for other racial/ethnic groups in the United States, as reported in the refractive error study of the EDPRG. ²¹ In (B) males and (D) females, myopia was defined as < −5.0 D spherical equivalent refractive error in the right eye for phakic eyes for LALES and for adult Chinese participants in the Tanjong Pagar survey ¹⁴ and the National Blindness and Low Vision Survey of Bangladesh. ¹⁷ Note that the data from Bangladesh for the 70- to 79-year-old age group includes all participants over 70 years of age.

Figure 3.

 

Age-related prevalence of myopia ≤ −1.0 D in the phakic eye with the larger absolute spherical equivalent refractive error in all phakic LALES participants (males and females combined), and those phakic participants without nuclear opacification.

Table 3.

 

Myopia Prevalence Stratified by Education, Acculturation, Country of Birth, and Diabetes

	40–54 y	55–64 y	65–74 y	≥75 y	Total, All Ages
Males
Education
0–11 y	109/831	38/399	42/252	24/89	213/1571
	13.12*	9.52*	16.67	26.97	13.56*
≥12 y	125/595	34/186	14/82	13/38	186/901
	21.01*	18.28*	17.07	34.21	20.64*
Acculturation
Score <1.9	124/936	41/413	37/211	16/50	218/1610
	13.25*	9.93*	17.54	32.00	13.54*
Score ≥1.9	110/492	31/172	19/123	21/77	181/864
	22.36*	18.02*	15.45	27.27	20.95*
Country of birth
Born outside of U.S.	158/1114	46/468	41/227	16/56	261/1865
	14.18*	9.83*	18.06	28.57	13.99*
Born in U.S.	76/314	26/117	15/107	21/71	138/609
	24.20*	22.22*	14.02	29.58	22.66*
Diabetes mellitus
No	178/1181	42/395	36/214	28/89	284/1879
	15.07*	10.63	16.82	31.46	15.11*
Yes	56/250	30/190	20/121	10/39	116/600
	22.40*	15.79	16.53	25.64	19.33*
Females
Education
0–11 y	157/1247	69/615	52/344	33/136	311/2342
	12.59*	11.22*	15.12	24.26	13.28*
≥12 y	208/721	46/231	26/114	2/27	282/1093
	28.85*	19.91*	22.81	7.41	25.80*
Acculturation
Score <1.9	169/1358	63/602	46/290	21/94	299/2344
	12.44*	10.47*	15.86	22.34	12.76*
Score ≥1.9	197/614	52/246	32/170	14/69	295/1099
	32.08*	21.14*	18.82	20.29	26.84*
Country of birth
Born outside of U.S.	243/1589	68/665	48/299	22/99	381/2652
	15.29*	10.23*	16.05	22.22	14.37*
Born in U.S.	123/383	47/183	30/161	13/64	213/791
	32.11*	25.68*	18.63	20.31	26.93*
Diabetes mellitus
No	292/1666	69/578	48/318	26/114	435/2676
	17.53*	11.94*	15.09	22.81	16.26*
Yes	74/309	46/271	30/143	9/49	159/772
	23.95*	16.97*	20.98	18.37	20.60*

 Prevalence is shown of any myopia (≤ −1.0 D), with the number of myopes/total group and percentage of total group. Subjects with missing data are excluded.

*  Significant difference in the proportion of participants with myopia between levels of various risk indicators (P < 0.05, χ²).

Table 4.

 

Independent Risk Indicators for High Myopia ≤ −1.0 D and High Myopia ≤ −5.0 D

Risk Indicator	Myopia ≤ −1.0 D			High Myopia ≤ −5.0 D
		Odds Ratio (95% CI)	Odds Ratio (95% CI) Excluding Nuclear Lens Opacification	Odds Ratio (95% CI)	Odds Ratio (95% CI) Excluding Nuclear Lens Opacification
Acculturation score	[1]	[1]	[N/A]	[N/A]
≤1.9	1	1
>1.9	1.48 (1.17–1.86)*	1.48 (1.16–1.90)
Education (y)	[2]	[2]	[3]	[2]
0–6	1	1	1	1
7–11	0.85 (0.68–1.06)	0.89 (0.69–1.14)	0.81 (0.48–1.35)	1.17 (0.61–2.26)
≥12	1.45 (1.18–1.78)*	1.55 (1.24–1.95)*	1.59 (1.05–2.40)*	2.11 (1.21–3.69)*
History of ocular disease	[3]	[N/A]	[2]	[N/A]
No	1	—	1	—
Yes	1.48 (1.18–1.87)*	—	2.45 (1.64–3.67)	—
Diabetes mellitus	[4]	[3]	[N/A]	[N/A]
No	1	1	—	—
Yes	1.39 (1.17–1.66)*	1.57 (1.29–1.91)*	—	—
Age (y)	[5]	[4]	[N/A]	[N/A]
40–44	1	1	—	—
45–49	1.00 (0.79–1.25)	1.01 (0.80–1.28)	—	—
50–54	0.83 (0.65–1.06)	0.84 (0.65–1.07)	—	—
55–59	0.69 (0.52–0.91)*	0.68 (0.51–0.90)*	—	—
60–64	0.65 (0.48–0.87)*	0.59 (0.43–0.81)*	—	—
65–69	0.70 (0.50–0.98)*	0.45 (0.29–0.69)*	—	—
70–74	0.90 (0.62–1.29)	0.54 (0.33–0.88)*	—	—
75–79	0.84 (0.54–1.31)	0.17 (0.05–0.57)*	—	—
80+	1.52 (0.89–2.58)	0.88 (0.24–3.17)	—	—
Country of birth	[6]	[5]	[1]	[1]
U.S.	1	1	1	1
Outside of US	0.76 (0.60–0.95)*	0.73 (0.57–0.93)*	0.51 (0.35–0.75)*	0.36 (0.23–0.57)*
Gender	[N/A]	[6]	[N/A]	N/A
Male	—	1	—	—
Female	—	1.24 (1.05–1.47)*	—	—
Work status	[N/A]	[N/A]	[N/A]	[3]
Employed	—	—	—	1
Not working	—	—	—	0.79 (0.50–1.26)
Retired	—	—	—	0.32 (0.13–0.81)*

 Odds ratios of risk indicators for myopia ≤ −1.0 D and high myopia ≤ −5.0 D, with and without inclusion of participants with nuclear lens opacification in the cohort analyzed, determined by stepwise multiple logistic regression. Numbers in brackets indicate order of entry of variable in stepwise multiple logistic regression for the given myopia level and cohort. N/A, odds ratio not applicable, because the variable was not a significant independent predictor in the final model for the given myopia level and cohort.

*  Odds ratios with confidence intervals excluding 1.0.

The authors thank all the members of the LALES Group (see ^Appendix) and the LALES External Advisory Committee (Roy Beck, Natalie Kurinij, Leon Ellewein, Helen Hazuda, Eve Higginbotham, Lee Jampol, M. Cristina Leske, Donald Patrick, and James M. Tielsch) and Ronald Klein, Kristine Lee, and John Kempen, for providing data as reported by the EDPRG for the BDES and the BES.