Distribution of height by age and educational level.
Distribution of axial length (right eyes) by height and age (A), sex (B), and education (C) in the Beaver Dam Eye Study.
Lee KE, Klein BEK, Klein R, Quandt Z, Wong TY. Association of Age, Stature, and Education With Ocular Dimensions in an Older White Population. Arch Ophthalmol. 2009;127(1):88-93. doi:10.1001/archophthalmol.2008.521
To describe ocular biometry relationships in older white adults.
Ocular dimensions were measured with partial coherence laser interferometry in 1968 persons (aged 58-100 years, 59% female) seen at the fourth examination of the Beaver Dam Eye Study. Generalized estimating equations–modeled associations of age, sex, height, and education with ocular dimensions: axial length, corneal curvature radius, and anterior chamber depth.
The mean axial length was 23.69 mm; mean corneal curvature radius was 7.70 mm; and mean anterior chamber depth was 3.11 mm. Participants younger than 65 years had larger eyes (longer axial length, greater corneal curvature radius, and deeper anterior chamber depth) than persons aged 75 years or older. Mean axial length was 23.86 mm, 23.66 mm, and 23.55 mm in people aged 64 years and younger, 65 to 74 years, and 75 years or older, respectively. Generally, larger eyes were observed in men (vs women) and in taller (>178 vs ≤158 cm) and more educated (>16 vs <12 years) persons. Adjustment for height accounted for all sex differences. Age differences in axial length were attenuated (P = .06) after adjustment for both height and education.
In this older white population, age and sex variations in ocular dimensions are partially explained by differences in stature and education.
Biometric measurements of the globe, such as axial length (AL), corneal curvature radius (CR), and anterior chamber depth (ACD), have been found to differ by sex and age from childhood through adulthood.1- 5 These measures also differ by stature5- 9 as well as education level.9- 11 Age, sex, stature, and education are all interrelated (eg, older adults are more likely to be shorter and less likely to be highly educated). Axial length in particular is an important determinant of refraction,4,12,13 which has shown to have strong relationships with age and education.2,13- 19
Although there have been studies that have assessed the interrelationships of ocular biometry measures with age, sex, stature, and education, most of these studies have been confined to Asian and Hispanic populations.1,2 Some of these studies have found that AL is shorter in older people and women,2,6 while others have reported relationships of height and education with various biometry measures.5- 7,20- 22 However, there is very little understanding about the complex interrelationships of age, sex, height, and education with ocular biometry measurements in older white adults. Further understanding of such relationships may provide insight into observed trends and patterns of myopia.13- 19
In the Beaver Dam Eye Study, a population-based cohort study of age-related eye diseases, we had the opportunity to add measurements of ocular dimensions at the fourth examination. In this primary report on these measures, we describe the relationships of ocular biometry components with age, sex, height, and education in this white adult population.
Ocular dimensions were measured during the fourth examination of the Beaver Dam Eye Study cohort. The study began in 1988 with a private census of the population of Beaver Dam, Wisconsin. All individuals (N = 5924) between the ages of 43 and 84 years were identified and invited for a baseline examination from 1988 through 1990. Follow-up examinations were performed every 5 years after, with the fourth examination occurring from May 2003 to May 2005. At each examination, participation rates were over 80%, with the primary reason for nonparticipation being death. At each examination, living nonparticipants were older, less educated, had poorer vision, had higher blood pressure, and smoked more than participants.23- 26 Informed consent was obtained at each examination with institutional review board approval. Tenets of the Declaration of Helsinki were followed. Similar protocols were followed during each examination. Using the National Institutes of Health classification of race, we identified 99% of the population as white.
Ocular biometry measures were added for the fourth examination, and those data are used in this article. The relevant portions of this examination, summarized herein, include standardized noncycloplegic refraction using an automated refractor (Humphrey, San Leandro, California) with further modification if resulting visual acuity was 20/40 or worse27; ocular biometry using partial coherence laser interferometry (IOL Master; Carl Zeiss, Jena, Germany) for AL, ACD, and CR; measurements of height and weight; questions about education; and assessment of lens status using standardized lens photography and grading.28
Weight was measured to the nearest quarter pound and converted to kilograms by multiplying by 0.4536. Height was measured to the nearest quarter inch and converted to centimeters by multiplying by 2.54. If we were unable to measure height or weight (3% of population), self-reported measurements were used. Education was assessed with the question: “What was the highest year of school or college you completed?” During slitlamp examination, the examiner assessed lens status.
Ocular biometry was measured following manufacturer's recommendations in 1976 participants (83%) before pupillary dilation (both eyes in 1962 participants, right eye only in 6, left eye only in 8). Reasons for not measuring ocular biometry (right eye) included the following: examinations conducted off site (n = 243), physical limitations (n = 41), not enough time (n = 56), and other/unspecified/refusal/inability (n = 67). The average of the 2 corneal curvature meridians was used for analysis of CR. Anterior chamber depth measurements were not included in the analyses in persons without a lens or with an intraocular lens (n = 269). Comparison of characteristics in those included for various analyses is presented in Table 1. In general, those included in analyses were younger, taller, and, after age adjustment, less likely to have diabetes, cataract, or age-related macular degeneration. There were no differences in education or refraction (spherical equivalent).
For all analyses, SAS, version 9.1 (SAS Institute Inc, Cary, North Carolina), was used. Normality of each ocular biometry measure (AL, CR, and ACD) was assessed. Mean ocular dimensions were calculated for each eye separately and are reported for the right eye, as analyses for left eyes were similar (data not shown). However, both eyes were included in analysis of variance models with the generalized estimating equations approach to adjust for correlation between the eyes. All adjustments were done for the continuous version of the variables. Models were assessed using appropriate contrast statements.
The associations of age and sex with the ocular dimensions and refractions are presented in Table 2. Men and younger persons had longer AL, greater (flatter) CR, and deeper ACD than women and older individuals, respectively. Persons younger than 65 years of age had a mean AL of 23.86 mm, a mean CR of 7.72 mm, and a mean ACD of 3.19 mm. Persons who were aged 75 years or older had a mean AL of 23.55 mm, a mean CR of 7.68 mm, and a mean ACD of 2.99 mm.
Both height and education decreased with increasing age. Persons younger than 65 years had a mean height of 167.1 cm (standard deviation [SD], 9.9 cm) and had completed a mean of 13.6 years (SD, 2.6 years) of school. On average, persons aged 75 years or older were 162.0 cm (SD, 9.6 cm) tall and had completed 12.1 years (SD, 2.9 years) of school. The age trends were similar within sex, and women were shorter (mean [SD], 158.6 [6.6] vs 172.9 [7.2] cm) and had completed less schooling (mean [SD], 12.5 [2.3] vs 13.1 [3.2] years) than men. Within each age category, persons with more education were taller (Figure 1). The same trends were observed within each sex (data not shown).
Ocular dimensions varied with height and education and were similar in men and women (Table 3). Taller individuals generally had larger eyes. In all participants, mean AL was 23.36 mm in those shorter than 158 cm and 24.20 mm in those taller than 178 cm (β = .31, P < .001 per 10 cm). Similarly, CR increased from 7.60 to 7.84 mm (β = .08, P < .001), and ACD increased from 3.04 to 3.21 mm (β = .05, P < .001). Persons with more education had longer AL (β = .08, P < .001 per year), greater (flatter) CR (β = .009, P < .001), and deeper ACD (β = .02, P < .001). When comparing CR by categories of education, only those with 16 or more years of education had significantly higher CR (vs those with 1-11 years of education). Additional adjustment for height attenuated the associations, but they remained statistically significant (not shown).
To further explore whether height differences may explain associations of age, sex, and education with the biometry measures, we plotted AL by height for age (Figure 2A), sex (Figure 2B), and education (Figure 2C). For persons of the same height, those younger than 65 years and those with more education appeared to have longer AL. Similar figures for CR and ACD (not shown) also suggested that among those with the same height, younger persons had deeper ACD and those with 16 or more years of education had greater (flatter) CR and deeper ACD.
In models adjusting for height and education (Table 4), associations of age and sex with the ocular components were attenuated. Importantly, the association of age with AL was nonsignificant after adjustment for both height and education (P = .06). The association of age with CR was no longer significant after adjustment for height. Education did not add information beyond height for CR. When both height and education were added to the model for ACD, the age association remained significant. The associations of sex with AL, ACD, and CR were no longer significant after adjustment for height.
In this study, we report on the distribution of ocular biometry measures in an older white population. We found that men and younger, taller, and more educated people had generally longer AL and larger eyes. We found that height explained most of the variations in the measures between men and women as well as the variation in CR between younger and older people. Adjustment for education in addition to height explained some of the variation in AL between younger and older people.
As in previous studies in Asian and Hispanic populations, we found that younger persons had longer AL, greater (flatter) CR, and deeper ACD than older persons. In a separate analysis in which we used age groups similar to those in other studies,1,2 this white population had, on average, longer AL and greater CR than Asian and Hispanic populations (K.E.L., unpublished data, 2007). In women aged 60 to 69 years, the mean AL was 23.7 mm in the Beaver Dam Eye Study, 22.7 mm in the Tanjong Pagar Survey,1 23.2 mm in Mongolia,29 and 23.1 mm in the Los Angeles Latino Eye Study.2 These differences may be influenced by differing methods used to obtain biometry measures (the IOL Master compared with ultrasonography). Additionally, our population was slightly older and taller, had more years of education, and had a spherical equivalent between that of the Asian and Hispanic populations.1,2
The age and sex differences observed for AL, CR, and ACD may be associated with older persons and women having smaller statures and thus smaller eyes. In our population, height decreased by 1.3 cm for every 5-year increase in age, and women were on average 14 cm shorter than men. In addition, shorter persons had significantly shorter AL, shorter CR, and shallower ACD, which is consistent with observations in Asian populations.5,6 Height was not related to the AL:CR ratio (K.E.L., unpublished data, 2007), supporting the notion that AL and CR remained in similar proportions for taller and shorter persons. Adjustment for height attenuated the associations of age and sex with the ocular biometry measures. The sex differences for all measures were no longer significant after height adjustment. However, despite some attenuation, older individuals still had shorter AL and shallower ACD, but there were no significant differences in CR after adjustment for height. The AL:CR ratio was greater in younger persons (K.E.L., unpublished data, 2007) even after adjustment for height, suggesting that these persons have proportionately longer AL, given the CR, than older persons.
The association between education and AL has not been widely studied.30 It may reflect excessive AL growth associated with near work (use-abuse theory),13 as supported by data from studies of medical students or microscopists that show an increase in AL with adult-onset myopia.10,11,31 It may also reflect larger stature resulting from a higher socioeconomic status and better nutrition associated with more-educated groups. Indeed, we found that in persons of the same age, those with more education tended to be taller. We also observed that in persons of the same height, those with more education had longer AL. We found that more education was significantly associated with longer AL, greater (flatter) CR, and deeper ACD. With the exception of CR, these associations remained despite adjustment for age and height (data not shown). Thus, height differences associated with education do not explain longer AL in persons with more education.
The major limitation in this and other reported population-based studies is their cross-sectional design, which does not allow evaluation of change in these measures with age. To investigate a real effect of aging as opposed to a cohort effect, longitudinal data are needed. While such data are lacking for AL, data on myopia suggest that the aging process may explain age-related declines in myopia observed in many cross-sectional studies.32 It is possible that some of these results may be affected by differential survival, in which survivors may be taller and more educated. However, we did not observe large differences in baseline height or education level in participants in this analysis compared with all participants in the baseline examination phase.
An important observation in our study was that adjustment for both education and height attenuated the association between age and AL. Younger persons had more education and were taller. In persons with a similar height, those with 16 or more years of education had longer AL. Our finding that age variation in AL, a key determinant of refraction, is largely explained by height and education may partially explain the cohort effect of more myopic refractions at younger ages.
Correspondence: Kristine E. Lee, MS, Department of Ophthalmology and Visual Sciences, University of Wisconsin–Madison, 610 N Walnut St, 405 WARF Bldg, Madison, WI 53726-2336 (firstname.lastname@example.org).
Submitted for Publication: January 23, 2008; final revision received June 12, 2008; accepted August 19, 2008.
Financial Disclosure: None reported.
Funding/Support: This study was supported by grant EY06594 from the National Institutes of Health (Drs R. Klein and B. E. K. Klein) and in part by the Research to Prevent Blindness (Drs R. Klein and B. E. K. Klein, Senior Scientific Investigator Awards).