Introduction

Retinopathy is prevalent in between10 and 30% of diabetes subjects(Klein et al. 1984b; Eggertsen et al.1993; Henricsson et al. 1996a; Kohneret al. 1998; van Leiden et al. 2002)already at the time of diabetes diagno-sis. This may indicate a 4–7 yearsdelay in the diagnosis of diabetes(Harris et al. 1992), based on theassumption that diabetic retinopathycan not exist before the start of thediabetes disease itself. Shortening thediagnostic delay can make activitiesagainst risk factors for different diabe-tes complications such, as for instanceretinopathy and progression of reti-nopathy, possible at an earlier stageand at its best before clinically evidentcomplications have occurred. Thenumber of diabetes patients withvisual impairment could then bereduced.

Population-based studies indicate,however, that retinopathy in peoplewithout diabetes is not uncommon, isassociated with hypertension and maypredict cardiovascular events and insome cases subsequent risk of diabetes(Yu et al. 1998; Wong et al. 2001,2005, 2006; Liew et al. 2009). Thiskind of retinopathy is usually includedin and labelled as ‘diabetic retinopa-thy’ when prevalence figures are given

The prevalence of retinopathy insubjects with and without type 2diabetes mellitus

Eydis Olafsdottir,1,2,3 Dan K. G. Andersson,4 Inger Dedorsson3

and Einar Stefansson1,2

1Department of Ophthalmology, The National University Hospital, Reykjavik,Iceland2University of Iceland, Reykjavik, Iceland3Department of Ophthalmology, Orebro University Hospital, Orebro, Sweden4Department of Public Health and Caring Sciences, Family Medicine and Clinical

Epidemiology Section, Uppsala University, Uppsala, Sweden

ABSTRACT.

Purpose: To evaluate the prevalence of and risk factors for, retinopathy in a geo-

graphically defined population with type 2 diabetes mellitus compared with a con-

trol group of subjects without diabetes, matched by age, sex and residence in order

to find the retinopathy attributable to type 2 diabetes.

Methods: The study populations are, on one hand, a prevalence cohort of subjects

with type 2 diabetes resident in the community of Laxa, Sweden, and on the other a

control group, matched by age, gender and residence with those with a diagnosis of

type 2 diabetes mellitus. Retinopathy was graded from fundus photographs using a

modification of the Early Treatment Retinopathy Study (ETDRS) adaptation of the

modified Airlie House classification of diabetic retinopathy (DR).

Results: Any retinopathy was found in 34.6% in the type 2 diabetes cohort and in

8.8% in the control group without diabetes. Among the diabetic patients, any reti-

nopathy was significantly associated with duration of diabetes (p = 0.0001),

HbA1c (p = 0.0056), systolic blood pressure (p = 0.0091) and lower serum cho-

lesterol (p = 0.0197) in multivariate logistic regression analyses. Having retinopa-

thy in the control group was associated only with systolic blood pressure

(p = 0.0014) in logistic regression analysis.

Conclusions: The prevalence of retinopathy among patients with type 2 diabetes in

Laxa, Sweden, was similar or somewhat lower compared with other studies in the

Nordic countries. The prevalence of retinopathy in a control group without diabetes

equalled numbers from population studies worldwide. Our study indicates that the

retinopathy that can be attributed to hyperglycaemia in the diabetic state is less

common than is usually accounted for. A considerable fraction of retinopathy in

subjects with diabetes may instead be due to other factors such as hypertension

and should thus be treated correspondingly.

Key words: diabetic retinopathy – duration – hyperglycaemia – hypertension – nondiabetic reti-

nopathy – Type 2 diabetes

Acta Ophthalmol.ª 2013 The Authors

Acta Ophthalmologica ª 2013 Acta Ophthalmologica Scandinavica Foundation. Published by Blackwell Publishing

Ltd.

doi: 10.1111/aos.12095

Acta Ophthalmologica 2013

1

for patients with diabetes, due to thefact that they look alike. As the non-diabetic and the diabetic retinopathymost likely have different causes andalso demand different strategies forprevention and treatment, it is ofinterest to determine their respectivemagnitude and their risk factors.

In this study, we have estimated theoccurrence of retinopathy in a preva-lence cohort with type 2 diabetes, andin an age- and gender-matched con-trol group without diabetes. Our firstaim was to find the prevalence of reti-nopathy that could be attributed tothe diabetes state per se, that is, thedifferences in prevalence between sub-jects with and subjects without diabe-tes. A second aim was to find factorsassociated with retinopathy amongthe studied subjects.

Methods

The study was undertaken in Laxa, arural municipality in the central partsof southern Sweden. The diabetes pop-ulation and the control populationhave been described elsewhere (Olafs-dottir et al. 2007, 2012). In short, thediabetes population included all sub-jects in Laxa with type 2 diabetes mell-itus, as of 1 January 1997. The controlgroup was found through the nationalpopulation register wherefore each dia-betes patient, one nondiabetic subject,living in Laxa, of the same sex andclosest in age to the diabetic patientwas sampled. The nondiabetic statusof these subjects, at the time of the eyeexamination, was secured by means ofmeasurements of fasting whole bloodglucose (FBG) and HbA1c. If indoubt, oral glucose tolerance tests wereperformed. Repeat glucose measure-ments were performed during a 5-yearfollow-up period.

Fundus photographs were takenafter dilatation of the pupil with my-driacyl. Five 45� fields including a ste-reo pair of the macula were takenwith a Topcon fundus camera, TRC-50 VT. Two 30� photographs werealso taken of the optic disc. An expe-rienced ophthalmologist (ID) gradedthe fundus photographs. Diabetic reti-nopathy was graded using a modifica-tion of the ETDRS adaptation of themodified Airlie House classification ofDR (Klein et al. 1994a, 1998). Eyeswere graded according to the follow-ing criteria:

0. No DR (levels 10–13)1. Non-proliferative DR: Mild (levels

14–20)2. Moderate (levels 31–43)3. Severe (levels 47–53)4. Proliferative DR (levels 60–85)

Any DR was defined as levels 14–85.Diabetic macular oedema was definedas the presence of hard exudatesand ⁄or retinal thickening within onedisc diameter of the centre of the mac-ula. Macular oedema was graded aspresent or not present. When datafrom both eyes were used, all the gra-dings of a patient were based on theworst eye. If only one eye could beevaluated, this was considered to bethe worst eye. The ophthalmologistgrading the photographs was notaware of the patient’s identity or clini-cal situation.

Systolic and diastolic blood pressureand heart rate were measured in alying position after five minutes ofrest. The mean of three measurementswas registered. Height was measuredwithout shoes and weight with lightclothing on. Body mass index (BMI)was calculated as weight in kg dividedby height in metre, squared (kg ⁄m2).All participants of the study wereasked to answer a questionnaire.Among the questions asked weresmoking habits. Blood samples,including FBG, HbA1c, cholesterol,triglycerides, HDL-cholesterol, creati-nine, uric acid and gamma glutamyltransferase, were taken and analysedat the laboratory of the nearby localcounty hospital of Karlskoga. Enzy-matic methods were used for choles-terol, triglycerides, HDL-cholesteroland uric acid. Creatinine was analysedusing the Jaffe method and gammaglutamyl transferase according toSzasz-Persijn. These analyses weremade on a Hitachi 911 instrument.HbA1c was analysed with a low-pres-sure chromatography method onGlycomat MTC.

The research ethics committee atthe Orebro County Council gave thestudy its approval. All examinationswere carried out from July 1996 toJune 1998.

Table 1 shows the characteristics ofthe type 2 diabetic and the controlparticipants. Two hundred and sev-enty-six type 2 diabetic patients, all ofwhite Caucasian origin, were enrolledin the study. They included all known

type 2 diabetes patients in Laxa, rep-resenting a crude 3.9% prevalence oftype 2 diabetes. One patient with type2 diabetes declined participation inthe eye examination. Four patientscould not participate in taking fundusphotographs because of high age com-bined with severe disease such asdementia, terminal malignant diseaseor sequel to cerebrovascular disease.In another eight subjects photographswere taken, but could not be gradedbecause of concomitant eye disease.Hence, 263 patient fundus photo-graphs could be evaluated, in 20patients only of one eye.

Originally 275 control subjects weresampled. Of these 36 subjects declinedparticipation, had no photographstaken or the photographs could notbe graded for the same reasons as forthe diabetes subjects, leaving 251 con-trol subjects whose fundus photo-graphs could be evaluated (Table 1).In eight subjects, only one eye couldbe graded.

Statistical analyses

The data were analysed using the sas

software release 9.2 (SAS Institute,Cary. NC, USA). Potential variablesassociated with any retinopathy wereidentified in bivariate logistic regres-sion analyses. Then, multivariatelogistic regression analyses, providingodds ratios (OR), their 95% confi-dence intervals (CI), p-values andWald‘s chi-square estimates, were per-formed with any retinopathy asdependent variable and variablessignificant in bivariate analyses as inde-pendent variables. All tests were two-tailed. p-values <0.05 were accepted asindicating statistical significance.

Results

Of the 263 patients with type 2 diabe-tes, 172 had no retinopathy and 91(34.6%) had any retinopathy. Of thelatter 62 (23.6%) had mild retinopa-thy, 20 (7.6%) moderate retinopathy,5 (1.9%) severe retinopathy and fourpatients (1.5%) had proliferative reti-nopathy. Seven patients (2.7%) hadmacular oedema out of which twohad moderate, three had severe non-proliferative retinopathy and two hadproliferative retinopathy. The preva-lence of diabetic retinopathy increasedfrom 11.8% (2 ⁄ 17) in those with a

Acta Ophthalmologica 2013

2

diabetes duration less than 1 year to68.2% (15 ⁄ 22) in those with a diabetesduration of 20 years or more (Fig. 1).One patient was blind with best visualacuity £0.1, due to proliferative reti-nopathy.

Twenty-two of the 251 control sub-jects (8.8%) had any retinopathy. Ofthese, 20 had mild and two had mod-erate retinopathy. During a 5-year fol-low-up time, two subjects in thecontrol group with retinopathy devel-oped type 2 diabetes. They were diag-nosed 35 and 43 months after the eyeexaminations. Both died shortly (3and 2 months) after the diabetes diag-nosis was confirmed.

Table 1 also shows associations,from bivariate analyses, between thepresence of any retinopathy and vari-ables registered at the cross-sectionalexaminations. Among the diabetic sub-jects, any retinopathy was significantlypositively associated with diabetesduration (p < 0.0001), FBG andHbA1c (p = 0.0100, p = 0.0013), sys-tolic and diastolic blood pressure (p =0.0018, p = 0.0232) and inversely tototal cholesterol (p = 0.098).

When these variables were treated ina series of multivariate logistic analy-

ses, duration of diabetes (p = 0.0001),HbA1c (p = 0.0056), systolic bloodpressure (p = 0.0091) and lower serum

cholesterol (p = 0.0197) remained sta-tistically significant for any retinopathy(Table 2).

Table 1. Characteristics of the study participants.

No diabetes Diabetes

No retinopathy

n = 229

Any retinopathy

n = 22p-value (any

retinopathy ⁄no retinopathy

No retinopathy

n = 172

Any retinopathy

n = 91p-value (any

retinopathy ⁄no retinopathy

Mean

or % SD

Mean

or % SD

Mean

or % SD

Mean

or % SD

Age (year) 69.0 12.6 75.0 10.17 0.0368 68.0 12.24 70.3 11.98 0.1530

Sex (female ⁄male) 0.86 0.50 1.0 0.51 0.7390 0.95 0.50 0.66 0.49 0.1516

Diabetes duration (year) NA NA NA NA NA 7.4 5.98 12.5 7.97 <0.0001

Current smoker (%) 12.4 0.33 15.8 0.37 0.6751 15.1 0.36 13.9 0.35 0.8103

Fasting whole blood

glucose (mm)

4.7 0.57 4.5 0.42 0.2607 7.6 2.66 8.6 3.08 0.0100

HbA1c (%) 4.5 0.58 4.4 0.56 0.8377 6.2 1.43 6.9 1.43 0.0013

Total cholesterol (mm) 6.0 1.02 6.1 1.23 0.6965 5.9 1.08 5.6 1.09 0.0498

HDL-cholesterol (mm) 1.4 0.39 1.4 0.39 0.6879 1.3 0.39 1.3 0.45 0.7995

Triglycerides (mm) 1.6 0.84 1.9 1.42 0.1804 2.1 1.15 2.1 1.45 0.8020

Creatinine (lm) 94.6 18.76 93.8 19.85 0.8597 94.3 22.67 102.7 46.49 0.0774

Uric acid (lm) 339.3 93.42 338.2 70.62 0.9590 364.3 104.26 369.3 104.10 0.7162

Gamma glutamyl

transferase (lm)

0.6 0.61 0.7 0.62 0.5337 0.9 1.22 0.8 0.72 0.3736

Height (cm) 167.3 9.03 166.3 9.55 0.6058 166.9 8.91 168.1 9.05 0.2839

Weight (kg) 73.9 12.19 69.8 12.98 0.1661 80.4 17.10 83.8 17.51 0.1384

Body mass index (kg ⁄m2) 26.4 3.66 25.3 4.60 0.2408 28.8 5.32 29.5 5.35 0.3396

Systolic blood

pressure (mm Hg)

144.5 22.24 162.0 23.35 0.0015 150.0 20.28 158.6 21.03 0.0018

Diastolic blood

pressure (mm Hg)

81.6 9.89 86.6 10.50 0.0282 82.7 8.47 85.2 8.49 0.00232

Heart rate (beats

per minute)

68.2 10.51 67.8 11.14 0.8703 72.1 11.94 72.5 14.26 0.8061

Bold values are statistically significant.

Classes of retinopathy by duration of type 2 diabetes

0

10

20

30

40

50

60

70

80

0–1 2–4 5–9 10–14 15–19 ≥20

Per

cen

t w

ith

ret

inop

ath

y

Duration of diabetes (years)

Proliferative DRP

Severe DRP

Moderate DRP

Mild DRP

Fig. 1. Classes of retinopathy by duration of type 2 diabetes.

Acta Ophthalmologica 2013

3

Among the control subjects, havingretinopathy was significantly positivelyrelated to systolic blood pressure(p = 0.0015), diastolic blood pressure(p = 0.0282) and to age (p = 0.0368).In a multivariate analysis (Table 2),only systolic blood pressure remainedstatistically significant (p = 0.0014, OR1.030, CI 1.012–1.040, v2 = 10.26).

Discussion

We found the prevalence of any reti-nopathy in our diabetic population tobe 34.6%, which is similar to studies inother Nordic countries. Henricssonreported a 36% prevalence of any reti-nopathy in Helsingborg, Sweden (Hen-ricsson et al. 1996a,b), and Kalmdocumented 40% in Gothenburg, Swe-den (Kalm et al. 1989). In Iceland, theprevalence of any retinopathy was 40%(Kristinsson et al. 1994) and in Arhus,Denmark 32% (Hove et al. 2004). Theprevalence found in Wisconsin, USAwas somewhat higher than in the Nor-dic studies, 54% (Klein et al. 1984a).In a study in Kisa, Sweden, the preva-lence of any retinopathy was 27% inall type 2 diabetic patients aged70 years and under (Falkenberg &Finnstrom 1994). A reason for thisslightly lower prevalence may be alower patient age and shorter diabetesduration. The mean age of the Kisapatients and duration of diabetes were60 ± 9 and 7.5 ± 5.5 years versus69 ± 12 and 9 ± 7 years in our study.

In contrast to the findings of Hans-son-Lundblad et al. (2002), we foundno association between patient ageand severity level of retinopathy ormacular oedema. However, as onlyeleven patients in our study had severeor proliferative retinopathy or macu-lar oedema, our results should beinterpreted cautiously.

In accordance with previous studies,duration of diabetes was significantlyassociated with any retinopathy andincreasing grade of retinopathy (Fig. 1).In our study, the frequency of any dia-betic retinopathy increased to 68%when the duration of diabetes was morethan 20 years. This is similar to what isseen in the other Nordic studies.

Glycemic control and blood pres-sure were also significantly associatedwith retinopathy, likewise in accor-dance with previous studies (UK Pro-spective Diabetes Study (UKPDS)Group. (1998a); UK Prospective Dia-betes Study Group., 1998b). Intensify-ing glucose-lowering therapies as wellas blood pressure control can be ofvalue in reducing the risk for retinop-athy and other microvascular compli-cations (UK Prospective DiabetesStudy (UKPDS) Group. (1998a); UKProspective Diabetes Study Group.,1998b). We found an inverse associa-tion between serum cholesterol anddiabetic retinopathy in contrast tosome other studies (Diabetes in Amer-ica 1995). Our finding could not beexplained by any drug treatment fordyslipidemia. Although the associationwas attenuated with increasing sever-ity of diabetic retinopathy, it remainedsignificant and we have so far no rea-sonable explanation for it.

We also found that 8.8% of an age-and gender-matched control group toour type 2 diabetic population hadretinopathy and that it was signifi-cantly associated with systolic bloodpressure, but not to any glucose con-trol measure or to age. To our knowl-edge, this is the first study whereretinopathy has been graded in a con-trol group to a geographically defineddiabetes population, but our resultsare similar to what has been found inother studies in populations without

diabetes, where retinal photographshave been used to grade retinopathy.In the Reykjavik Eye Study, the prev-alence of retinopathy in a cohort ofelderly Icelanders with no diabeteswas 10.2% (Gunnlaugsdottir et al.2012). In the Blue Mountains EyeStudy, the prevalence of retinopathyin subjects with no diabetes was 9.8%,in the Beaver Dam Eye Study 7.8%,in the Cardiovascular Health Study8.3% and in the Hoorn study 9%(Klein et al. 1994b; Yu et al. 1998;van Leiden et al. 2002; Wong et al.2002a, 2003). In the Icelandic study,associated factors included age andmicroalbuminuria. In all the otherstudies, retinopathy was associatedwith high blood pressure and in somestudies to other factors as well, suchas age, BMI, serum cholesterol andtriglyceride levels. We saw an associa-tion with diastolic blood pressure andage in bivariate analyses, but not withBMI or serum lipids possibly becauseour study is smaller than the others.

In diabetic patients, the retinopathypresent at the time of diabetes diagnosishas routinely been seen as diabetic reti-nopathy. However, other pathophysio-logical processes than hyperglycaemia,like hypertension, might explain some ofthe retinopathy observed in diabetespatients. If so, the true prevalence of ret-inopathy that can be attributed to diabe-tes, both at the time of its diagnosis andduring its course, may be overestimatedby 8–9%.

As the current diabetes, diagnosis,first proposed by the National DiabetesData Group in 1979 and later by WHOin 1980, 1985 and in 1999 (WHO-Study-Group, 1980, 1985; Report of aWHO consultation. 1999), is based onan evaluation of the risk of future mi-croangiopathic complications, mainlyretinopathy, due to elevated blood

Table 2. Multivariate regression analyses of the effects of significant characteristics of both the control population and type 2 diabetes population

on having any retinopathy.

No diabetes Diabetes

OR 95% CI p-value v2 OR 95% CI p-value v2

Age 0.3979 0.7147 0.5293 0.3958

Diabetes duration (years) NA NA NA NA 1.090 1.043–1.139 0.0001 14.8641

Fasting whole blood glucose (mm) 0.1966 1.6672 0.4157 0.6624

HbA1c (%) 0.8671 0.0280 1.322 1.085–1.611 0.0056 7.6681

Total cholesterol (mm) 0.6438 0.2138 0.718 0.543–0.948 0.0197 5.4370

Systolic blood pressure (mm Hg) 1.030 1.012–1.049 0.0014 10.2622 1.019 1.005–1.033 0.0091 6.7970

Diastolic blood pressure (mm Hg) 0.4590 0.5484 0.0834 2.9966

Bold values are statistically significant.

Acta Ophthalmologica 2013

4

glucose levels at a 2-h oral glucose tol-erance test (OGTT), our findings mayalso question the validity of the 2-hglucose level as the gold standard forthe diabetes diagnosis and promptfuture research on how to separate dia-betic from other types of retinopathy.

From the UKPDS, we know that forpersons having type 2 diabetes mellitus,it is important to control blood glucoseas well as blood pressure in order tominimize the risk of developing sight-threatening retinopathy (Stratton et al.2001). Isolated retinal micro-aneurysmsand haemorrhages are the earliest signsof retinopathy in people with diabetesmellitus, and their presence has a highpredictive value for worsening retinopa-thy (Kohner et al. 1999). The signifi-cance of retinopathy and the naturalcourse of this phenomenon in personsnot having diabetes is, however, lesswell known, although new studies signalits importance and connection withstroke and coronary heart disease(Wong et al. 2001, 2002b, 2005).

In conclusion, our study shows a34.6% prevalence of retinopathy inour type 2 diabetic prevalence cohort,which is similar to many other Nordicstudies. The retinopathy attributed tothe diabetes disease itself may, how-ever, be overestimated by 8.8% as thiswas the prevalence of retinopathy thatwas found in the control group with-out diabetes.

AcknowledgementsThis study was supported by the Orebro County

Council, Orebro, Sweden, the Research foundation

of the National University Hospital, Reykjavik,

Iceland and the Helga Jonsdottir and Sigurlidi

Kristjansson Memorial Fund, Reykjavik, Iceland.

ReferencesDiabetes in America (1995): Diabetes in America, 2nd

edn. National Institute of Health, National Institute

of Diabetes and Digestive and Kidney Diseases.

Bethesda, MD: National Diabetes Information

Clearinghouse, NIH Publication no 95-1468, 320.

Eggertsen R, Kalm H & Blohme G (1993): The value

of screening for retinopathy and microalbuminuria

in patients with type 2 diabetes in primary health

care. Scand J Prim Health Care 11: 135–140.

Falkenberg M & Finnstrom K (1994): Associations

with retinopathy in type 2 diabetes: a population-

based study in a Swedish rural area. Diabet Med

11: 843–849.

Gunnlaugsdottir E, Halldorsdottir S, Klein R et al.

(2012): Etinopathy in old persons with and without

diabetes mellitus: the Age, Gene ⁄ Environment Sus-

ceptibility–Reykjavik Study (AGES-R). Diabetolo-

gia 55: 671–680.

Hansson-Lundblad C, Holm K, Agardh CD &

Agardh E (2002): A small number of older type 2

diabetic patients end up visually impaired despite

regular photographic screening and laser treatment

for diabetic retinopathy. Acta Ophthalmol Scand

80: 310–315.

Harris MI, Klein R, Welborn TA & Knuiman MW

(1992): Onset of NIDDM occurs at least 4-7 yr

before clinical diagnosis. Diabetes Care 15: 815–819.

Henricsson M, Nilsson A, Groop L, Heijl A & Janzon

L (1996a): Prevalence of diabetic retinopathy in

relation to age at onset of the diabetes, treatment,

duration and glycemic control. Acta Ophthalmol

Scand 74: 523–527.

Henricsson M, Tyrberg M, Heijl A & Janzon L

(1996b): Incidence of blindness and visual impair-

ment in diabetic patients participating in an oph-

thalmological control and screening programme.

Acta Ophthalmol Scand 74: 533–538.

Hove MN, Kristensen JK, Lauritzen T & Bek T

(2004): The prevalence of retinopathy in an unse-

lected population of type 2 diabetes patients from

Arhus County, Denmark. Acta Ophthalmol Scand

82: 443–448.

Kalm H, Egertsen R & Blohme G (1989): Non-stereo

fundus photography as a screening procedure for

diabetic retinopathy among patients with type II dia-

betes. Compared with 60D enhanced slit-lamp exam-

ination. Acta Ophthalmol (Copenh) 67: 546–553.

Klein R, Klein BE & Moss SE (1984a): Visual impair-

ment in diabetes. Ophthalmology 91: 1–9.

Klein R, Klein BE, Moss SE, Davis MD & DeMets

DL (1984b): The Wisconsin epidemiologic study of

diabetic retinopathy. III. Prevalence and risk of

diabetic retinopathy when age at diagnosis is 30 or

more years. Arch Ophthalmol 102: 527–532.

Klein R, Klein BE, Moss SE & Cruickshanks KJ

(1994a): The Wisconsin Epidemiologic Study of

diabetic retinopathy. XIV. Ten-year incidence and

progression of diabetic retinopathy. Arch Ophthal-

mol 112: 1217–1228.

Klein R, Klein BE, Moss SE & Wang Q (1994b):

Hypertension and retinopathy, arteriolar narrow-

ing, and arteriovenous nicking in a population.

Arch Ophthalmol 112: 92–98.

Klein R, Klein BE, Moss SE & Cruickshanks KJ

(1998): The Wisconsin Epidemiologic Study of Dia-

betic Retinopathy: XVII. The 14-year incidence and

progression of diabetic retinopathy and associated

risk factors in type 1 diabetes. Ophthalmology 105:

1801–1815.

Kohner EM, Aldington SJ, Stratton IM, Manley SE,

Holman RR, Matthews DR & Turner RC (1998):

United Kingdom Prospective Diabetes Study, 30:

diabetic retinopathy at diagnosis of non-insulin-

dependent diabetes mellitus and associated risk fac-

tors. Arch Ophthalmol 116: 297–303.

Kohner EM, Stratton IM, Aldington SJ, Turner RC

& Matthews DR (1999): Microaneurysms in the

development of diabetic retinopathy (UKPDS 42).

UK Prospective Diabetes Study Group. Diabetolo-

gia 42: 1107–1112.

Kristinsson JK, Stefansson E, Jonasson F, Gislason I

& Bjornsson S (1994): Screening for eye disease in

type 2 diabetes mellitus. Acta Ophthalmol

(Copenh) 72: 341–346.

van Leiden HA, Dekker JM, Moll AC, Nijpels G,

Heine RJ, Bouter LM, Stehouwer CD & Polak BC

(2002): Blood pressure, lipids, and obesity are asso-

ciated with retinopathy: the hoorn study. Diabetes

Care 25: 1320–1325.

Liew G, Wong TY, Mitchell P, Cheung N & Wang JJ

(2009): Retinopathy predicts coronary heart disease

mortality. Heart 95: 391–394.

Olafsdottir E, Andersson DK & Stefansson E (2007):

Visual acuity in a population with regular screening

for type 2 diabetes mellitus and eye disease. Acta

Ophthalmol Scand 85: 40–45.

Olafsdottir E, Andersson DKG & Stefansson E

(2012): The prevalence of cataract in a population

with and without type 2 diabetes mellitus. Acta

Ophthalmol Scand 90: 334–340.

Report of a WHO Consultation. (1999): Definition,

diagnosis and classification of diabetes mellitus and

its complications. part 1: diagnosis and classifica-

tion of diabetes mellitus. Geneva: World Health

Organization.

Stratton IM, Kohner EM, Aldington SJ, Turner RC,

Holman RR, Manley SE & Matthews DR (2001):

UKPDS 50: risk factors for incidence and progres-

sion of retinopathy in Type II diabetes over 6 years

from diagnosis. Diabetologia 44: 156–163.

UK Prospective Diabetes Study (UKPDS) Group.

(1998a): Intensive blood-glucose control with sul-

phonylureas or insulin compared with conventional

treatment and risk of complications in patients with

type 2 diabetes (UKPDS 33). Lancet 352: 837–853.

UK Prospective Diabetes Study Group. (1998b): Tight

blood pressure control and risk of macrovascular

and microvascular complications in type 2 diabetes:

UKPDS 38. BMJ 317: 703–713.

WHO-Study-Group (1980): Diabetes mellitus: techni-

cal report series 646. Geneva: World Health Orga-

nization.

WHO-Study-Group (1985): Diabetes mellitus: techni-

cal report series 727. Geneva: World Health Orga-

nization.

Wong TY, Klein R, Couper DJ, Cooper LS, Shahar

E, Hubbard LD, Wofford MR & Sharrett AR

(2001): Retinal microvascular abnormalities and

incident stroke: the Atherosclerosis Risk in Com-

munities Study. Lancet 358: 1134–1140.

Wong TY, Hubbard LD, Klein R et al. (2002a): Reti-

nal microvascular abnormalities and blood pressure

in older people: the Cardiovascular Health Study.

Br J Ophthalmol 86: 1007–1013.

Wong TY, Klein R, Sharrett AR, Couper DJ, Klein

BE, Liao DP, Hubbard LD & Mosley TH (2002b):

Cerebral white matter lesions, retinopathy, and

incident clinical stroke. JAMA 288: 67–74.

Wong TY, Klein R, Sharrett AR et al. (2003): The

prevalence and risk factors of retinal microvascular

abnormalities in older persons: the Cardiovascular

Health Study. Ophthalmology 110: 658–666.

Wong TY, Rosamond W, Chang PP, Couper DJ,

Sharrett AR, Hubbard LD, Folsom AR & Klein R

(2005): Retinopathy and risk of congestive heart

failure. JAMA 293: 63–69.

Wong TY, Mohamed Q, Klein R & Couper DJ (2006):

Do retinopathy signs in non-diabetic individuals pre-

dict the subsequent risk of diabetes? Br J Ophthal-

mol 90: 301–303.

Yu T, Mitchell P, Berry G, Li W & Wang JJ (1998):

Retinopathy in older persons without diabetes and

its relationship to hypertension. Arch Ophthalmol

116: 83–89.

Received on June 29th, 2012.

Accepted on December 8th, 2012.

Correspondence:

Eydis Olafsdottir, MD

Department of Ophthalmology

Orebro University Hospital

701-85 Orebro, Sweden

Tel: + 46 19 6021000

Fax: + 46 19 101632

Email: eydiso@simnet.is

Acta Ophthalmologica 2013

5