Prevalence of and associations with asteroid hyalosis 331

Correspondence andreprint requests to:Dr. Paul MitchellDept. of OphthalmologyCentre for Vision ResearchUniversity of SydneyWestmead HospitalHawkesbury RdWestmead, NSWAustralia, 2145.Tel.: +61 2 9845 7960Fax: +61 2 9845 6117E-mail:paulmi@westgate.wh.usyd.edu.au

Acknowledgements:This study was supported by theAustralian National Health &Medical Research Council (Grant #974159) and the WestmeadMillennium and Save SightInstitutes, University of Sydney.

Abstractpurpose To assess the prevalence of and associations with asteroidhyalosis in an older population.

methods The Blue Mountains Eye Study was a cross-sectional studyof an older community (aged 49–97 years). Subjects included werethose attending the baseline (n = 3654) and 5-year examinations (n =2335) of this cohort. Asteroid hyalosis was diagnosed clinically by thepresence of cream-white spherical bodies within the vitreous or fromgrading of stereo retinal photographs of both eyes. Logistic regressionassessed age-sex adjusted associations with relevant characteristics,including diabetes and cardiovascular variables.

results Asteroid hyalosis was found in 36 subjects (1.0%), 95% con-fidence interval (CI) 0.7% to 1.3%, and was bilateral in three affectedsubjects (8.3%). An age-related increase in prevalence was observed,increasing from 0% of persons aged less than 55 years to 2.1% ofpersons aged 75 years or older. The prevalence of this sign was signif-icantly higher in men (1.4%) than in women (0.6%), the age-adjustedodds ratio (OR) was 2.54 (CI 1.25–5.16). No statistically significantassociations were found between asteroid hyalosis and a history ofheart disease, gout, current smoking, the highest level of alcohol con-sumption or with presence of diabetes (diagnosed from history orfasting blood glucose tests).

conclusions Asteroid hyalosis was detected in 1% of participantsin this Australian older population. No significant associations werefound, apart from age and male gender. Our study provides similar age-specific prevalence data to a recent report from the Beaver Dam EyeStudy for asteroid hyalosis.

Ophthalmic Epidemiology0928-6586/03/US$ 16.00

Ophthalmic Epidemiology– 2003, Vol. 10, No. 5,pp. 331–335© Swets & Zeitlinger 2003

Accepted 15 April 2003

Asteroid hyalosis in an older population:the Blue Mountains Eye Study

Paul MitchellMaria Y. WangJie Jin Wang

Department of Ophthalmology, University of Sydney, Australia

Original article

Oph

thal

mic

Epi

dem

iol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y M

ichi

gan

Uni

vers

ity o

n 10

/31/

14Fo

r pe

rson

al u

se o

nly.

Key words Asteroid hyalosis; cross-sectional study; Blue Moun-tains Eye Study; disease prevalence; elderly population

Introduction Dispersed through the vitreous as pallid sphericalparticles resembling cosmic dust, asteroid hyalosis seldom causes anyvisual symptoms. However, since its recognition in the nineteenthcentury, a number of reports have suggested that this sign may be asso-ciated with systemic conditions, including diabetes,1 hyperlipidemia,2,3

hypertension1 and hyperopia.1 Data from the population-based U.S.Beaver Dam Eye Study, however, could not confirm these systemic vas-cular or refractive associations.4

In this brief report, we aimed to describe the prevalence of asteroidhyalosis in a representative older Australian population and to inves-tigate systemic and ocular associations, both previously reported andunreported, with the presence of this sign in either eye.

Methods The baseline survey of the Blue Mountains Eye Study(BMES I) examined 3654 persons aged 49 years or older, residing in a2-postal-code area of the Blue Mountains region, west of Sydney, Aus-tralia, during the period 1992–1994, as described in a previous report.5

Five years after the initial examinations, 2335 (75.0% of survivors)attended follow-up assessments (BMES II), which were conductedduring the period 1997–1999.6

A detailed medical history was taken, including history of cardiovas-cular diseases, gout, smoking and alcohol consumption. Height, weightand seated blood pressure were measured and body-mass index wascalculated. Participants were also asked to return for fasting blood tests,including glucose and lipids; over 90% complied.

At both examinations, participants had a comprehensive eye exami-nation, before and after pupil dilatation. The assessment includedstereoscopic 30° retinal photographs, using a Zeiss FF3 fundus camera(Carl Zeiss, Oberkochen, Germany) of multiple fields in both eyes.5

Signs of asteroid hyalosis were graded either from the presence oftypical cream-white spherical bodies seen at the slit-lamp or on retinalexamination or from grading of the stereo retinal photographs of eacheye. Cataract, glaucoma and diabetic retinopathy were graded usingstandard criteria, as described previously.7–9

We performed logistic regression analyses, using the Statistics Analy-sis System (Version 6.08, SAS Institute Inc. Cary, NC, USA), to assessage-sex adjusted associations with asteroid hyalosis.

Results Of 3654 BMES I participants, 3583 had gradable fundusphotographs. Asteroid hyalosis was diagnosed in 36 subjects, including18 right and 21 left eyes. This sign was graded as present in both eyesof three participants (8.3% of those affected). Among the 36 affectedcases, 29 had signs of asteroid hyalosis observed during BMES I, anda further 7 cases had asteroid hyalosis that was only evident in pho-tographs taken during the BMES II examinations. These two groupswere then amalgamated and the age at examination when this condi-tion was detected was calculated.

332 P. Mitchell et al.

Oph

thal

mic

Epi

dem

iol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y M

ichi

gan

Uni

vers

ity o

n 10

/31/

14Fo

r pe

rson

al u

se o

nly.

Fig. 1. Prevalence of asteroidhyalosis (%) by age group and sex.

The overall population prevalence was 1.0%, 95% confidence interval (CI), 0.7% to 1.3%, and increased significantly with age (p for trend <0.001). The prevalence was 0% in persons aged <55 years,0.4% in persons aged 55–64 years, 1.1% in persons aged 65–74 yearsand 2.1% among those aged 75 years or older. The prevalence doubledfor each decade of increasing age after age 55 years, odds ratio (OR)2.05 (CI 1.46, 2.89). A significant gender difference was also observed,with asteroid hyalosis being significantly more frequent in men (1.4%)than in women (0.6%). This difference was present at all ages, oddsratio (OR) 2.54 (CI 1.25, 5.16), after adjusting for age, as shown inFigure 1.

Table 1 presents the age-specific prevalence of asteroid hyalosistogether with age-sex adjusted odds ratios for associations between anumber of variables and asteroid hyalosis. Increased odds were foundfor associations between asteroid hyalosis and a history of heartdisease, gout, current smoking and the highest level of alcohol con-sumption, but none of these associations were statistically significant.No significant associations were present for any of the other systemicor ocular variables examined, including diabetes, diagnosed from thehistory or from a finding of elevated fasting blood glucose in personswithout a history of diabetes.

Discussion This study has confirmed asteroid hyalosis as a rela-tively uncommon, mostly unilateral, age-related vitreous pathologyoccurring with much greater frequency in men than in women. In thispopulation, we were not able to detect any case before age 55 yearsbut documented that the prevalence of this condition increased withage from 55 years on. The associations with age and sex were previ-ously reported by the U.S. Beaver Dam Eye Study.4

Most previous reports of the frequency of asteroid hyalosis have beenof clinic-based series1,10 or case-control studies.11 The Beaver Dam EyeStudy is the only other population-based study to report the prevalenceof this well-recognised sign.4

Our prevalence findings are in very close agreement with those fromthe Beaver Dam Eye Study,4 which reported a prevalence of 1.2%

Prevalence of and associations with asteroid hyalosis 333

0

0.5

1

1.5

2

2.5

3

<55 55-64 65-74 75+

age group

%

Women

Men

All

Oph

thal

mic

Epi

dem

iol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y M

ichi

gan

Uni

vers

ity o

n 10

/31/

14Fo

r pe

rson

al u

se o

nly.

334 P. Mitchell et al.

table 1. Associations with asteroid hyalosis.

Characteristic Prevalence Age- and sex-adjusted odds ratio

number at risk %(95% confidence interval)

age at diagnosis (per decade) 2.10 (1.48–2.98)<55 477 055–64 1158 0.4365–74 1198 1.0975+ 750 2.13

sexfemale 2024 0.59 1.0male 1559 1.41 2.54 (1.25–5.16)

body mass index (kg/m2)<20 205 0.98 0.98 (0.21–4.51)20–25 1339 0.90 1.026–29 1377 1.16 1.47 (0.68–3.14)≥30 603 0.50 0.76 (0.21–2.75)

blood pressure statusnormotensive 1931 1.04 1.0controlled 655 1.07 0.98 (0.41–2.35)uncontrolled 489 1.02 0.77 (0.28–2.08)untreated 496 0.40 0.29 (0.07–1.25)

systolic blood pressure 0.98 (0.96–1.00)diastolic blood pressure 0.97 (0.94–1.00)history of angina or myocardial infarction

no 2989 0.77 1.0yes 573 1.75 1.67 (0.78–3.56)

history of goutno 2914 0.82 1.0yes 390 1.79 1.50 (0.63–3.60)

history of diabetes or fasting glucose ≥7.0mmol/ Lno 3307 0.94 1.0yes 276 1.09 0.99 (0.30–3.26)

smokingnever smoked 1802 0.67 1.0ex-smoker 1243 1.29 1.52 (0.69–3.36)current smoker 535 1.12 2.07 (0.75–5.74)

pack-years (per 10) 1.05 (0.97–1.14)alcohol consumption (drinks/day)

none 1.0<1 0.80 (0.33–1.96)1–3 1.02 (0.42–2.48)≥4 1.56 (0.33–7.50)

serum cholesterol 1.23 (0.89–1.69)serum triglycerides 1.18 (0.95–1.46)serum HDL cholesterol 0.95 (0.39–2.29)diabetic retinopathy

no 3501 0.94 1.0yes 82 1.22 1.22 (0.16–9.08)

glaucomano 3479 0.95 1.0yes 104 0.96 0.60 (0.08–4.53)

intraocular pressure 1.05 (0.94–1.17)any cataract

no 2253 0.89 1.0yes 1209 1.16 0.59 (0.27–1.27)

Oph

thal

mic

Epi

dem

iol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y M

ichi

gan

Uni

vers

ity o

n 10

/31/

14Fo

r pe

rson

al u

se o

nly.

Prevalence of and associations with asteroid hyalosis 335

References1 Bergren RL, Brown GC, Duker JS.

Prevalence and association ofasteroid hyalosis with systemicdiseases. Am J Ophthalmol. 1991;111(3):289–293.

2 Smith JL. Asteroid hyalitis.Incidence of diabetes mellitus andhypercholesterolemia. J Am MedAssoc. 1958;168:891–893.

3 Bard LA. Asteroid hyalitis.Relationship to diabetes andhypercholesterolemia. Am JOphthalmol. 1964;58:239–242.

4 Moss SE, Klein R, Klein BEK.Asteroid hyalosis in a population:the Beaver Dam eye study.Am J Ophthalmol. 2001;132(1):70–75.

5 Mitchell P, Smith W, Attebo K,Wang JJ. Prevalence of age-relatedmaculopathy in Australia. The Blue Mountains Eye Study.Ophthalmology. 1995;102(10):1450–1460.

6 Mitchell P, Wang JJ, Foran S, SmithW. Five-year incidence of age-related maculopathy lesions: TheBlue Mountains Eye Study.

Ophthalmology. 2002;109(6):1092–1097.

7 Mitchell P, Cumming RG, Attebo K,Panchapakesan J. Prevalence ofcataract in Australia: the BlueMountains Eye Study.Ophthalmology. 1997;104(4):581–588.

8 Mitchell P, Smith W, Attebo K,Healey PR. Prevalence of open-angle glaucoma in Australia. TheBlue Mountains Eye Study.Ophthalmology. 1996;103(10):1661–1669.

9 Mitchell P, Smith W, Wang JJ, AtteboK. Prevalence of diabeticretinopathy in an older community.The Blue Mountains Eye Study.Ophthalmology. 1998;105(3):406–411.

10 Potter JW, Newcomb R. Prevalenceof asteroid bodies in a VAoptometry clinic. J Am OptomAssoc. 1980;51(1):19–25.

11 Luxenberg M, Sime D.Relationship of asteroid hyalosis todiabetes mellitus and plasma lipidlevels. Am J Ophthalmol. 1969;67(3):406–413.

(95% CI 0.9% to 1.5%). Our overall prevalence rate of 1.0% is withinthe confidence limits of the Beaver Dam prevalence and our confidenceintervals of 0.7% to 1.3% overlap those found in the Beaver DamStudy.The age-specific prevalence rates were also very similar betweenthe two studies.

Our data also provide no evidence to support earlier observations ofa significant association between asteroid hyalosis and diabetes,1 as alsoreported by the Beaver Dam Eye Study.4 We could not, however,confirm the Beaver Dam Eye Study findings that suggested associa-tions between asteroid hyalosis and higher body mass index, a historyof gout and of higher levels of alcohol consumption.

Important limitations of this study include the relatively smallnumber of identified cases of asteroid hyalosis and the possibility ofunder-diagnosis for mildly affected cases. This small number limitationis particularly relevant in relation to the assessment of associations withsystemic diseases such as diabetes. Nevertheless, the prevalence dataindicate a very close concordance with findings from a similar popula-tion-based survey. The systemic origin of this comparatively trivial andasymptomatic ocular condition, however, remains obscure.

Oph

thal

mic

Epi

dem

iol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y M

ichi

gan

Uni

vers

ity o

n 10

/31/

14Fo

r pe

rson

al u

se o

nly.

Oph

thal

mic

Epi

dem

iol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y M

ichi

gan

Uni

vers

ity o

n 10

/31/

14Fo

r pe

rson

al u

se o

nly.