UNITE DE RECHERCHES STATISTIQUES (I.N.S.E.R.M. U. 21), VILLEJUIF CHAIRE DE DIABETOLOGIE, HOTEL-DIEU, PARIS CLINIQUE OPHTALMOLOGIQUE, HOTEL-DIEU, PARIS DIABETIC R E T I N O P A T H Y , DURATION OF DIABETES AND RISK FACTORS OF ATHEROSCLEROTIC CARDIOVASCULAR DISEASE DII)IEt JoB

EVELINE ESCHWkGE

CLAUDINE GUYOT-ARGENTON

GEOR¢lZS TCHOBROUTSKY

JEAN-PIERREAUBRY

MAURICEDI~ROT

An important aspect of the epidemiological study of diabetic retinopathy consists in tracing factors linked with its evolution. Numerous studies have been carried out on chronological factors: age at the clinical onset of diabetes, present age, known duration of diabetes ~ Because of the high risk of atherosclerotic cardiovascular disease in diabetics, the measurement of blood pressure, body weight and serum lipids all form part of the routine check for diabetes. However, their relation to the development of diabetic retinopathy has been studied relatively little and apparently never comprehensively *' ~' ,2. ,~. ~6 The present work analyses, in a group of insulin-dependent diabetics, the relationship between the degree of evolution of diabetic retinopathy and weight, blood pressure and serum lipid levels, as well as the relationship between all these factors and the chronological variables. The subjects chosen form a reliable group for the study of these rdationships, as, contrary to maturity onset diabetes, the existence of the disease is independent of the weight factor and is a certainty. MATERIAL AaND METHODS The study was carried out on 145 adult, insulin-dependent diabetics, aged less than 60, out-patients or in-patients of the Department of Diabetology of the Hgtel-Dieu, Paris, between January 1970 and January 1971. Their weight, height, supine blood pressure and ser~_ cholesterol and lipid leveIs7. ,s were noted. Fifty-three of them (group G1) form part of a longitudinaI study of diabetic retinopathy*. For these, fluorescein angiography of the eye fundus was taken. This fluorography *Essai th&apeufique concernant la pr6vention des r6tinopatl~es chez les diab&iques insutino-ddpendants. A.F.E.D.I.A. (ASSAN, C.~NIVE'r, CAT'~LtNEAU, CH~ME,'C~S, DEBRY, D~ROT,DEVIL,ESCHWq~GE,LEBOUC,LLrBETZKI,PEQUIGNOT,RAMBERT,RATtmRY,ROSSELIN, SC~,eA~TZ, TCHOBROUTSKY,THERWer,V.~LL~ON). Key-words: Blood pressure," Cholesterol; Diabetics retinop~thy; Duration o/ diabetes; Weight. Received: December 20, 1974. Acta diabe¢, lat. t2, 1, 1975.

DIABETIC

RETINOPATHY

A N D RISK F A C T O R S

OF A T H E R O S C L E R O T I C

CARDIOVASCULAR

DISEASE

is centred systematically on the macula. It explores a circular surface totalling about 5 papillary diameters including the papilla. The fluorographic aspects are described quantitativeiy giving the numbers of microaneurysms per quadrant, the retinal haemorrhages and new vessel formation. Only the number of microaneurysms, considered the most specific lesions of the disease, is used in this study. The inclusion of haemorrhages, ex-udates and new- vessels, which are neither numerous nor normally distributed, does not contribute further information or alter the results in any way. In this group, the evolution of the retinal lesions ranged from stage I to stage I I I of ALAE~TS and SLOSSE's dassification 2. Stages 0, IV and V are by definition not included in the longitudinal study. The other 92 patients (group G2) were in-patients. Their data were obtaincd by retrospective examination of the records of all insulin-dependent diabetics hospitalised during the period when group G1 subjects were recruted. The only records excluded from the study were those of patients hospitalised for severe diabetic ketosis or ketoacidosis, because these have marked disturbing effect on humoral parameters, blood pressure and weight. All the parameters used in this work form part of the routine check upon admission. In particular, the data for the eye fundus of the in-patients are those of the clinical routine examination and not of fluorescein angiography which explains the qualitative classification of the results. The patients fell into 4 classes: Class A : normal eye fundus Class B : isolated abnormalities of the perimacular small b!ood vessels C'lass C : mieroaneurysms and/or haemorrhages and/or exudates Class D : presence of new vessels. A few patients of class C e,xhibit non-specific lesions, haemorrhages or exudates without mieroaneurysms. It was considered that in such a group of insulin-dependent subjects, who on average are young and not suffering from high blood pressure, these lesions are very likely to be of diabetic origin. In group G1 as in group G2, the lesions described are those of the more severely involved eye. All the data were set down on questionnaires and worked out by computer at the Centre de CaIcul de I'I.N.S.E.R.M. RESULTS

Description o~ the groups studied (tabs 1-3) The study is based on diabetic subjects who became insulin-dependent at the onset or soon after, and w h o were young at the onset of diabetes. The duration of clinical diabetes is shorter in group G2 than in group G1. This corresponds to the presence, within group G2, of an important subgroup of patients with a normal eye fundus. This sub-group does not exist in group G1 by reason of the selection criteria operating for the sub3ects included in this group. Generally, subjects in both groups showed little overweight: 3 subjects only were more than 2 0 % overweight, their blood pressure was not high and their serum lipid levels were within the limits considered normal.

Relationship between diabetic retinopat~y, age at the onset o/ diabe~es~ presen~ age and duration o~ diabetes (tab~ 4) i n both groups, the degree of evolution of diabetic retinopathy was closely correlated to the duration of diabetes. O n the other band, it was not related to age at onset of diabetes nor, in group G1, to present age. Statistical analysis shows that the significant link in group G 2 between severity of retinopathy and present age reflects solely the fact that older subjects have diabetes of longer standing.

D. 30B, E. ESCHWEGE,G. TCHOt3ROUTSKY,C. GUYOT-ARGENTON,I. P. AUBRY,M. DI~ROT a) Retinal lesions and sex G1(%)

ocular lesions

G2(%)

0

A

52

B

26

15

C

55

21

D

i9

12

59

63

G12 SEM) (mean -t-

(mean +G22 SEM)

sex (% men)

b) Other parameters studied

t present age (years)

40±

3

34±

age at onset of diabetes (years)

27±

3

27 +__ 2.5

dttration of diabetes (years)

t3+

2

6.5 +_ 1.5

duration of insulin treatment (years}

12_+ 2

body weight (kg)

63+

3

3

59 ~- 3

height (cm)

165 + 3

t66±

systolic blood pressure (mmHg)

135 ± 6

125=h 4

diasmlic blood pressure (mmHg)

80± 3

75±

2 2

serum choIesterol (mg/I00 mi)

220 ± 11

230 __+10

total serum lipids (rag/00 ml)

600 __+36

630 -- 30

Table 1 - Description of groups G1 (53 patients) and G2 (92 patients). The division of group G1 subjects into classes A, B, C, D was made according to the rules described for ~oup G2 (see text) but using the data of fiuorescein angiography. Relationship between diabetic retinolathy and weight, blood pressure, serum cholesterol, and lipid levels (tab. 5) I n both groups, the degree of evolution of diabetic retinopathy increased with weight and blood pressure. I n group G I it also increased with the cholesterol level. This relationship was not found in group G2. N o relationship was found between retinal disease and total lipid levels. Statistical analysis showed that b o d y weight, blood pressure and cholesterol level were each intrinsically linked to the development of retinal lesions, that is each of these variables remained linked to the state of the eye when the effect

DIABETIC ILETINOPATHY AND RISK FACTORS OF A T H E R O S C L E R O T I C

CARDIOVASCULAR DISEASE

% of subjects (weight in proportion to height) groups

above 20% above over 20% 10% weight ideal weight above ideal ideal weight

< ideal weight

= ideal weight

GI

24

39

31

6

0

G2

54

20

16

6.5

3.5

Percentage difference from ideal weight in both groups, according to the tables of the Metropolitan Life Insurance Company, supplied by Documenta Geigy.

Table 2 -

of the other two was eliminated by the use of partial correlation. Similarly, the retinal state remained linked to weight, blood pressure and cholesterol level, when the effect of the duration of diabetes was taken into account. DISCUSSION Numerous studies 5 have shown the frequency and severity of diabetic retinopathy to increase with two factors: duration of the disease and age of the patient at the time of diagnosis. The relationship between severity of a) Diastolic blood pressure % of subjects with diastolic blood pressure (ram Hg) groups -~ 75

8045

90-95

100-105

> 105

G1

24

42

22

12

0

G2

44

37

12

7

0

b) Systolic blood ~ressure % of subjects with systolic blood pressure (ram Hg) groups ~< 120

125-135

G1

26

36

G2

44

31

Table 3 -

Blood pressure in both groups.

140~150

155-.165

> 170

24

8

6

17

2

6

1

D. JOB, E. ESCHWEGE,G. TCHOBROUTSKY,C. GUYOT-ARGENTON,J. V. AUBRY,M. DEROT a) Group GI: corrections age

number of microaneurysms degree of significance

at onset of diabetes

present age

duration of diabetes

r = --0.11

r = 0.18

r = 0.47

n.s.

n.s.

p < 0.00t

b) Group G2: comparison of averages class

age ar onset of diabetes (years)

A

26

29

3

B

32

37

5

C

26

39

13

D

24

40

16

p < 0.0i

p < 0.001

degree of significance

n.8,

present age (years) duration (years) of diabetes

Table 4 - Relationship between diabetic retinopathy and chronological factors. (r = correlation coefficient; n.s. = non significant).

retinopathy and duration of diabetes was confirmed by the present study. However, in the group examined here, the severity of retinal lesions seemed unrelated to age at onset of diabetes. This apparent discrepancy from data found in the literature could be explained by the fact that the influence of age at onset of diabetes was established by comparkqg groups of subjects who were not only of different ages at the time of diagnosis but also whose diabetes was not equally severe, the older subjects at the time of diagnosis of the disease being less frequently insulin-dependent ~. In these conditions it is difficult to separate the influence of age at clinical onset of diabetes from the influence of the undetected duration of the disease, which is a~ O" " 4 the lon~,er the less severe diabetes . Age at onset of diabetes was found to have no influence, This was in a group of subjects who were a!t insulin-dependent. In their case the duration of the undetected evolution of the disease is presumably short and of limited range, I t would therefore seem that the relationship usually found between age at onset of diabetes and delay in the appearance of diabetic retinopathy reflects the influence ok the undetected duration of the disease rather than the intrinsic influence of the patient's age at the time of diagnosis of diabetes. 5

DIABETIC RETINOPATHY AND RISK FACTORS OF ATttEROSCLEROTIC CARDIOVASCULAR D I S E A S E

a) Group GI: correlations corpulence index

systolic blood pressu re

diastolic blood pressure

Serum

8erlmll

cholesterol

~pids

number of microaneurysms

r = 0.39

r = 0.26

r = 0.42

r = 0.4t

r = 0.20

degree of significance

p < 0,01

p < 0.10

p