Is peak expiratory flow rate a predictor of complications in diabetes?

The Wisconsin Epidemiologic Study of Diabetic Retinopathy

Barbara E.K. Kleina,*, Scot E. Mossa, Ronald Kleina, Karen J. Cruickshanksa,b

aDepartment of Ophthalmology and Visual Sciences, University of Wisconsin-Madison, 610 North Walnut Street, 460 WARF, Madison, WI 53705-2397, USAbDepartment of Preventive Medicine, University of Wisconsin Medical School, Madison, WI 53705-2397, USA

Received 28 February 2001; received in revised form 11 June 2001; accepted 18 June 2001

Abstract

The objective of this study was to determine whether peak expiratory flow rate is a predictor of complications of diabetes. Peak expiratory

flow rate was measured at the 10-year follow-up (third examination) of a cohort of persons with younger-onset diabetes. The relationships of

progression of diabetic retinopathy by two steps, progression to proliferative retinopathy and of incidences of macular edema, sore or ulcers

on feet or ankles, lower extremity amputation, proteinuria, and cardiovascular disease 4 years after this examination with respect to peak

expiratory flow rate were evaluated. Study procedures including measurements of blood pressure, height and weight, grading of fundus

photographs, peak expiratory flow rate, urinalysis, and medical history were performed according to standard protocols. Peak expiratory flow

rate was not associated in univariate analyses with progression of retinopathy, incidences of proliferative retinopathy, macular edema or lower

extremity amputation, sores or ulcers on feet or ankles, gross proteinuria, or self-reported cardiovascular disease. However, when using

multivariable models to include the effects of other risk factors, peak expiratory flow rate was significantly associated with the combined

incidences of sores or ulcers on feet and ankles, or lower extremity amputations (OR= 0.61, 95% CI 0.42–0.88). These data suggest that

peak expiratory flow rate is a predictor of subsequent complications in the lower extremities in those with long duration of younger-onset

diabetes. Evaluating this association in an incipient cohort would illuminate whether the relationship we found is likely to be causal. D 2001

Elsevier Science Inc. All rights reserved.

Keywords: Complication of diabetes; Peak expiratory flow rate

1. Introduction

Pulmonary and respiratory functions have been found to

be compromised in persons with Type 1 diabetes. Bell et al.

(1988) found decreased lung volume in persons with Type 1

diabetes. Schuyler, Niewoehner, Inkley, and Kohn (1976)

found increased lung elastic recoil in male subjects with

juvenile onset diabetes. Thickening of basal laminae of the

alveoli (Vracko, Thoring, & Huang, 1979) and diabetic

microangiopathy in the capillaries of the alveolar septa have

been reported (Walton, Byrd, Fields, Ossorio, & Roy, 1994).

These may affect diffusion of carbon monoxide (Sandler,

Bunn, & Stewart, 1986; Walton et al., 1994). Moreover,

transcutaneous oxygen pressure has been found to be

diminished in patients with Type 1 diabetes (Breuer, Breuer,

& Berger, 1988; Walton et al., 1994). While a large and

varied battery of pulmonary and respiratory function tests

(Breuer et al., 1988; Primhak, Whincup, Tsanakas, &

Milner, 1987; Ramirez et al., 1991) have been commonly

used in clinic based studies of people with diabetes, peak

expiratory flow rate is easily measured and can be incorpo-

rated in field studies (Dahlquist, Eisen, Wegman, & Kriebel,

1993). Moreover, peak expiratory flow rate has been shown

to be associated with mortality (Cook et al., 1991).

The development and progression of complications of

diabetes have been thought to be due to relative hypoxia in

the affected organs and tissues (Cameron & Cotter, 1999;

Dyck & Giannini, 1996; Linsenmeier et al., 1998; Nakhos-

tine, Nadeau, & Lamontagne, 1997; Veves et al., 1996).

Population-based studies rarely collect information about

hypoxia on the tissue or cellular level. However, it is

possible that pulmonary and respiratory dysfunction influ-

ence oxygenation of tissues. Thus, it may be informative to

investigate the association of pulmonary function to vas-

1056-8727/01/$ – see front matter D 2001 Elsevier Science Inc. All rights reserved.

PII: S1056 -8727 (01 )00170 -2

* Corresponding author. Tel.: +1-608-263-0276; fax: +1-608-263-

0279.

Journal of Diabetes and Its Complications 15 (2001) 301–306

cular complications of diabetes. We present the experience

derived from the Wisconsin Epidemiologic Study of Dia-

betic Retinopathy (WESDR) of the association of peak

expiratory flow rate and the development of these endpoints.

2. Subjects and methods

2.1. Population

The population has been described in detail in previous

reports (Klein, Klein, Moss, et al., 1984; Klein, Klein, Moss,

& Cruickshanks, 1994; Klein, Klein, Moss, Davis, &

DeMets, 1984a, 1984b; Klein, Klein, Moss, Davis, &

DeMets, 1989a, 1989b). In brief, a sample of 2990 persons

was selected for the baseline examination from a total of

10,135 diabetic patients identified as receiving primary care

in an 11-county area in southern Wisconsin from July 1,

1979 through June 30, 1980. This sample was composed of

two groups. The first group consisted of all patients with a

diagnosis of diabetes before 30 years of age who took

insulin (1210 patients), the ‘‘younger-onset’’ group. The

second group consisted of a probability sample of the 5431

patients with a diagnosis of diabetes at 30 years of age or

older. Data from this ‘‘older-onset’’ group is omitted

because of the high mortality in this group in the later years

of follow-up.

Of the 1210 eligible younger-onset persons, 996 partici-

pated in the baseline examination in 1980–1986 (Examina-

tion 1) and 795 participated in the 10-year follow-up

(Examination 3). Of these, 651 were seen four years later

(Examination 4). Peak expiratory flow fate was first meas-

ured at the 10-year follow-up (examination 3). This serves

as the ‘‘baseline’’ for this paper and follow-up (incidence)

refers to Examination 4, an interval of 4 years.

2.2. Procedures

The examinations were performed in a mobile van in or

near the city where the participants resided. Pertinent parts

of the ocular and physical examinations included measuring

blood pressures (Hypertension Detection and Follow-up

Program Cooperative Group, 1976), dilating the pupils

and taking stereoscopic color fundus photographs of seven

standard fields (Diabetic Retinopathy Study Research

Group, 1981), determining random blood glucose and

glycosylated hemoglobin levels from a finger-prick capil-

lary blood sample (at baseline) and from venous blood (at

the follow-up examinations) (Quick-Step Fast Hemoglobin

Test System, Isolab, Akron, OH) (Moss, Klein, Klein,

Spennetta, & Shrago, 1988; Quick-Step Fast Hemoglobin

Test System, 1981), and obtaining a urine specimen for a

semiquantitative measurement of urine protein. Measure-

ment of peak expiratory flow was added at the 10-year

follow-up (Examination 3) (Wright, 1978) using the mini-

Wright peak flow meter (Clement Clarke, Columbus, OH

43219). The test was administered by the study examiners

who had been trained in the protocol, which is described by

the manufacturer (mini-Wright Peak Flow Meter, Clement

Clarke). Subjects were standing for all measurements. Three

measurements were made and the highest value was used in

these analyses. The meters were calibrated at eight flow

rates using a shop vacuum to generate flow (Cook, Evans,

Scherr, et al., 1989).

A standard structured interview was conducted, includ-

ing questions about age, age at diagnosis of diabetes,

cigarette smoking, cardiovascular disease, sores and ulcers

of the feet and ankles, and lower extremity amputations at

each study examination. At the 10-year examination, the

Respiratory Symptom Questionnaire recommended by the

American Thoracic Society was added (Ferris, 1978).

2.3. Grading protocol

To determine the incidence and progression of retinop-

athy in each eye, the fundus photographs were graded in a

masked fashion using the Early Treatment Diabetic Retin-

opathy Study (ETDRS) adaptation of the modified Airlie

House classification scheme (Early Treatment Diabetic

Retinopathy Study Research Group, 1991; Klein, Davis,

Segal, et al., 1984).

3. Definitions

At Level 10, no retinopathy was present; Levels 21, 31,

37, 43, 47, and 53 included nonproliferative diabetic retin-

opathy of increasing severity, and Levels 60, 61, 65, 71, 75,

and 85 included proliferative retinopathy of increasing

severity. In the analyses, eyes with proliferative retinopathy

at Levels 60–85 were grouped together as Level 60 + . This

included eyes that were treated with panretinal photocoagu-

lation during the follow-up interval. Eyes that could not be

graded for retinopathy levels because of opacities in the

media or enucleation not related to diabetic retinopathy were

classified as ‘‘cannot grade.’’

In determining retinopathy levels for a participant, the eye

with the higher level was given greater weight. Participants

in a given level were divided into two groups: those with this

level in each eye and those with a lesser level in one eye. For

example, the level for a participant with Level 37 retinopathy

in each eye is specified by the notation ‘‘Level 37/37.’’ This

scheme provided a 15-step scale (10/10, 21/ < 21, 21/21, 31/

< 31, 31/31, 37/ < 37, 37/37, 43/ < 43, 43/43, 43/ < 47, 47/47,

53/ < 53, 53/53, 60+/ < 60+, and 60+/60+) when all levels of

proliferative retinopathy were grouped as one level. For

purposes of classification, if retinopathy severity could not

be graded in an eye, this eye was considered to have the same

score as the participant’s other eye.

The incidence of any retinopathy was estimated from all

persons who had no retinopathy at the 1990–1992 exam-

ination (Level 10/10) and who participated 4 years later.

B.E.K. Klein et al. / Journal of Diabetes and Its Complications 15 (2001) 301–306302

Any retinopathy was defined as a retinopathy severity Level

21/ < 21 or worse at follow-up. Progression to proliferative

retinopathy was estimated for all persons who were clas-

sified as having no or nonproliferative retinopathy severity

(Level 53/53 or better at the 1990–1992 examination) and

who participated 4 years later. Proliferative retinopathy was

defined as a retinopathy severity Level 60+/ < 60+ or worse

at follow-up.

For persons with nonproliferative or no retinopathy,

progression was defined as an increase in the retinopathy

severity by two or more steps (e.g., 10/10 to 21/21 or worse

or 37/ < 37 to 43/ < 43 or worse). Progression was not

examined in persons who had proliferative retinopathy at

the 10-year examination.

Macular edema was defined as thickening of the retina

with or without partial loss of transparency within one disc

diameter (DD) from the center of the macula (Early Treat-

ment Diabetic Retinopathy Study Research Group, 1985) or

the presence of focal photocoagulation scars in the macular

area (associated with a history of development of macular

edema) as documented by stereoscopic fundus photographs.

Clinically significant macular edema was based on the

detailed gradings and was defined as the presence of any

one of the following: thickening of the retina located 500 mmor less from the center of the macula (Early Treatment

Diabetic Retinopathy Study Research Group, 1985), hard

exudates with thickening of the adjacent retina 500 mm or less

from the center of the macular, or a zone of retinal thickening

one disc area or larger in size, located one disc diameter or

less from the center of the macula. If macular edema could not

be graded in an eye, the individual was assigned the score of

the other eye. The incidence of macular edema was estimated

from data for all persons who had no macular edema and had

not been previously treated with photocoagulation at the

10-year examination and who participated in the follow-up

examination 4 years later.

Cardiovascular disease status was determined as fol-

lows: a person was classified as having a history of

cardiovascular disease if he/she reported having been told

by a physician that he/she had angina, or had a heart

attack or stroke. The history of angina, heart attack, and

stroke was verified by a physician. Incidence of cardio-

vascular disease was presence at the 14-year examination

among participants without it at the 10-year examination.

Urine protein was considered present at a concentration of

0.3 g/l or greater. Incidence of urine protein and incidence

of lower extremity amputation were defined in a manner

similar to incidence of cardiovascular disease. We also

defined the incidence of sores or ulcers on the feet or

ankles or lower extremity amputation as the occurrence of

either event among participants without both at the

10-year examination.

Current age was defined as the age at the time of the

examination in 1990–1992. Age at diagnosis of diabetes

was defined as the age at the time the diagnosis was first

recorded by a physician on the patient’s chart or in a

hospital record. The duration of diabetes was that period

between the age at diagnosis and the age at the 1990–

1992 examination.

The means of both systolic and diastolic blood pressures

were the averages of the last two of three measurements

(Hypertension Detection and Follow-up Program Coopera-

tive Group, 1976).

Cigarette smoking status was determined as follows: a

person was classified as having never smoked if he/she had

smoked fewer than 100 cigarettes in his/her lifetime, as

being an ex-smoker if he/she smoked more than this number

of cigarettes in his/her lifetime but had stopped smoking

before the examination, or as currently smoking if he/she

had not stopped. For purposes of analysis, two dichotomous

variables were defined: one to compare persons who had

previously smoked with those who had never smoked and

one to compare persons who currently smoked with those

who had never smoked. Because lower extremity amputa-

tions were infrequent (N = 13) and, because they may be a

result of sores or ulcers on the feet or ankles, we combined

information for incidences of these two categories of out-

come in the analytic models.

4. Statistical analysis

To determine the relative risks and 95% confidence

intervals for the endpoints, the peak expiratory flow rate

Table 1

Values at baseline of those with diabetes of younger onset participating at baseline and at 10- and 14-year follow-up

Baseline 10-year 14-year

Characteristic N Mean S.D. N Mean S.D. N Mean S.D.

Age, year 996 29.3 13.3 784 27.1 11.8 642 26.7 11.1

Duration, year 996 14.7 10.6 784 13.0 9.4 642 12.5 8.9

Glycosylated hemoglobin, % 950 10.8 2.1 746 10.8 2.1 610 10.6 2.0

N % N % N %

Sex, % male 996 51.4 784 49.4 642 49.5

Hypertension, % present 988 21.9 779 15.7 640 14.5

Smoking status, % current 996 24.6 784 22.1 642 20.9

Dead, n 0 150 214

SD = standard deviation

B.E.K. Klein et al. / Journal of Diabetes and Its Complications 15 (2001) 301–306 303

at the 10-year examination was divided into quartile ranges.

Tests for trends in rates were performed by the Mantel–

Haenszel procedure (Mantel, 1963). Multivariate analyses

for predicting the endpoints were performed by logistic

regression (Hosmer & Lemeshow, 1989). Power estimates

were obtained from nQuery Advisor and were based on the

c2 test of proportions (Elashoff, 1999).

5. Results

A comparison of persons who participated in the first

WESDR examination and in the third and fourth examina-

tions reveals that those participants who contributed data to

the current analyses were younger at baseline, more likely

to be nonsmokers and to have had lower blood pressure

(Table 1). Most who did not return for the later follow-up

examinations had died.

We first investigated the univariate relationships of

peak expiratory flow rate to incidence of endpoints 4 years

later. There were too few people who did not have diabetic

retinopathy (n = 25) at the 1990–1992 examination to

evaluate incidence of any retinopathy 4 years later. The

sample that was eligible for progression was more robust

(n = 372); there was no evidence of a significant relation-

ship to peak expiratory flow rate (Table 2). Similarly,

there was no evidence of a significant association of any

of the individual endpoints with peak expiratory flow rate

(Table 2).

However, these analyses do not consider other variables

which may confound these relationships. Therefore, in order

to investigate whether a relationship of peak expiratory flow

rate to these outcome variables might be influenced by other

characteristics previously shown to be associated with these

outcomes, we performed multiple logistic regression ana-

lyses including those additional characteristics (Table 3).

This analysis indicated a statistically significant association

between peak expiratory flow rate and incidence of lower

extremity amputations and sores or ulcers of the feet or

ankles. Respiratory symptoms did not add significantly to

the multivariable models.

An analysis corrected for height was hampered by the

paucity of data from a similar outside population. We

attempted an evaluation by developing an expected value

of peak expiratory flow rate based on the regression

described by Nunn and Gregg (1989). The regressions

were based on healthy nonsmoking volunteer subjects.

Using values for our population calculated from their

equations resulted in odds ratios that differed little from

those we found.

Table 2

Incidencea of complications of diabetes by quartile of peak expiratory flow rate

Peak expiratory flow rate (l/min)

Endpoint 200–430 440–490 500–580 590–760 P valueb

Two-step progression of diabetic retinopathy N 81 92 94 105

% 43.2 41.3 44.7 50.5 .25

Progression to proliferative diabetic retinopathy N 81 92 94 105

% 11.1 10.9 12.8 7.6 .51

Incidence of macular edema N 90 96 102 105

% 8.9 3.1 8.8 10.5 .36

Incidence of sores and ulcers on the feet or ankles or lower extremity amputation N 119 116 124 121

% 10.1 8.6 12.1 5.0 .31

Incidence of proteinuria N 87 96 100 93

% 6.9 15.6 9.0 6.5 .52

Incidence of cardiovascular disease N 133 141 158 142

% 3.8 7.1 9.5 5.6 .40

a Incidence refers to the development of the endpoint at the 14-year follow-up (4 years after the measurement of peak expiratory flow rate).b Based on the Mantel–Haenszel test of trend.

Table 3

Odds ratio and 95% confidence interval for an endpoint for a 100-unit increment in peak expiratory flow rate controlling for other risk factors

Endpoint Odds ratio 95% confidence interval P value

Two-step progressiona 1.08 0.84, 1.41 .54

Progression to proliferative diabetic retinopathyb 0.85 0.50, 1.47 .56

Incidence of macular edemab 1.43 0.82, 2.50 .21

Incidence of sores or ulcers on the feet and ankles or lower extremity amputationc 0.61 0.42, 0.88 .008

Incidence of proteinuriaa 0.73 0.48, 1.11 .14

Incidence of cardiovascular disease historyd 1.29 0.90, 1.84 .17

a Controlling for sex and glycosylated hemoglobin.b Controlling for sex, glycosylated hemoglobin, and baseline retinopathy level.c Controlling for sex, glycosylated hemoglogin, and smoking status.d Controlling for age, systolic blood pressure, total cholesterol, HDL cholesterol, and proteinuria.

B.E.K. Klein et al. / Journal of Diabetes and Its Complications 15 (2001) 301–306304

6. Discussion

Our analyses suggest that peak expiratory flow rate was

inversely associated with progression to proliferative retin-

opathy, incidence of lower extremity amputation or of sores

or ulcers on the feet or ankles, and incidence of proteinuria,

although only the relationship to the lower extremity lesions

was significant.

Diminished pulmonary and respiratory functions may be

associated with tissue oxygenation and other metabolic

functions at the cellular level. These could be causally

related to the complications that we have evaluated. Dia-

betes is associated with hyalinization of alveolar basement

membranes, microvascular changes (Vracko et al., 1979),

and abnormal lung elasticity (Schuyler et al., 1976). There

are various measures of pulmonary and respiratory function

that are used to detect which mechanism is more likely to be

compromised. Peak expiratory flow rate is only one such

measure and it is thought to best reflect large airway

functions, while our subjects may be experiencing pulmon-

ary/respiratory dysfunction on many levels. Because it is not

feasible to perform more extensive testing, we are limited in

our ability to determine the presence and severity of other

pulmonary/respiratory functions that may have been influ-

enced by diabetes in our population.

The prevalence and incidence of sores and ulcers on the

feet or ankles and lower extremity amputations (Moss,

Klein, & Klein, 1992) in this cohort have been shown to

be related to many characteristics, some of which are

related directly to diabetes, others not. The findings from

the current investigation add another factor. Peak expir-

atory flow rate is at least a risk indicator, if not a cause of

these problems.

A possible explanation for the lack of significant asso-

ciations between poorer peak expiratory flow rate and other

vascular complications, aside from the possibility that no

causal relationship exists, is that these complications are

more strongly associated with the other risk factors and with

local factors than is true for the lower extremity lesions.

While peak expiratory flow rate has been linked to mortality

from cardiovascular disease (Cook et al., 1991), there are

little published data describing its association to measures

of tissue oxygenation which may be a direct result of

diabetic microangiopathy.

The underlying approach we have taken is the longit-

udinal one which is appropriate when searching for poten-

tially causal associations. However, it is likely that

diminished peak expiratory flow rate is another ‘‘complica-

tion’’ of diabetes and one which may precede the other

complications. It may or may not be causally related to the

development or progression of the other lesions. We are

unable to sort these questions out in our observational

epidemiologic investigation.

Another consideration in evaluating our findings is that

the average duration of diabetes of our cohort when we did

our initial measurement of peak expiratory flow rate was 23

years. The group had already experienced significant incid-

ence and progression of diabetic retinopathy (Klein et al.,

1989a, 1989b), macular edema (Klein, Moss, Klein, Davis,

& DeMets, 1989), proteinuria (Klein, Klein, & Moss, 1991),

amputations (Moss et al., 1992), end stage renal disease

(Klein, Klein, & Moss, 1996), and death (Moss, Klein, &

Klein, 1991). Thus, those at highest risk for many of the

endpoints of interest, even if they were related to pulmonary

function, were excluded from the analyses by virtue of

having experienced the event at or before we measured

peak expiratory flow rate and therefore were not eligible to

be incident cases.

We note that peak expiratory flow rate is dependent upon

subject effort (Clement & Van de Woestijne, 1971; Sobel &

Emergil, 1964). Effort may be a function of physical fitness

and therefore peak expiratory flow rate may, in a sense, be a

psycho-physical test. However, it is associated with age,

smoking, and respiratory symptoms suggesting that it is

reflecting pathologic or possibly degenerative states (Cook

et al, 1989).

Lastly, we have limited power to detect some of these

relationships. For example, the power to find a halving of

the incidence of proteinuria between the lowest and highest

quartiles of peak expiratory flow rate is 10–15%.

In light of research findings of others showing a

relationship of peak expiratory flow rate to cardiovascular

disease death (Cook et al., 1991) and to findings in this

cohort indicating the relationship of glycosylated hemo-

globin to peak expiratory flow rate, studies of the

relationship of microvascular endpoints to pulmonary

function early in the course of disease are of interest

and may be best addressed by studying an incipient

cohort. They may suggest that the importance of glyce-

mia in complications may, in part, be mediated by

pulmonary function.

Acknowledgments

This research is supported by National Institute of Health

grant EY12198 (R. Klein, BEK Klein).

References

Bell, D., Collier, A., Matthews, D. M., Cooksey, E. J., McHardy, G. J., &

Clarke, B. F. (1988). Are reduced lung volumes in IDDM due to defect

in connective tissue? Diabetes, 37, 829–831.

Breuer, H. W., Breuer, J., & Berger, M. (1988). Transcutaneous oxygen

pressure measurements in Type I diabetic patients for early detection of

functional diabetic microangiopathy. European Journal of Clinical In-

vestigation, 18, 454–459.

Cameron, N. E., & Cotter, M. A. (1999). Effects of antioxidants on nerve

and vascular dysfunction in experimental diabetes. Diabetes Research

and Clinical Practice, 45, 137–146.

Clement, J., & Van de Woestijne, K. P. (1971). Variability of maximum

expiratory flow-volume curves and effort independency. Journal of

Applied Physiology, 31, 55–62.

B.E.K. Klein et al. / Journal of Diabetes and Its Complications 15 (2001) 301–306 305

Cook, N. R., Evans, D. A., Scherr, P. A., Speizer, F. E., Vedal, S., Branch,

L. G., Huntley, J. C., Hennekens, C. H., & Taylor, J. O. (1989). Peak

expiratory rate in an elderly population. American Journal of Epidemi-

ology, 130, 66–78.

Cook, N. R., Evans, D. A., Scherr, P. A., Speizer, F. E., Taylor, J. O., &

Hennekens, C. H. (1991). Peak expiratory flow rate and 5-year mortal-

ity in an elderly population. American Journal of Epidemiology, 133,

784–794.

Dahlquist, M., Eisen, E. A., Wegman, D. H., & Kriebel, D. (1993). Re-

productivity of peak flow measurements. Occupational Medicine, State

of the Art Reviews, Vol. 8, No. 2, Philadelphia, Hanley and Belfus,

April – June.

Diabetic Retinopathy Study Research Group. (1981). Report 7: a modifi-

cation of the Airlie House classification of diabetic retinopathy. Inves-

tigative Ophthalmology and Visual Science, 21, 210–226.

Dyck, P. J., & Giannini, C. (1996). Pathologic alterations in the diabetic

neuropathies of humans: a review. Journal of Neuropathology and Ex-

perimental Neurology, 55, 1181–1193.

Early Treatment Diabetic Retinopathy Study Research Group. (1985).

Photocoagulation for diabetic macular edema. Early treatment dia-

betic retinopathy study report 1. Archives of Ophthalmology, 103,

1796–1806.

Early Treatment Diabetic Retinopathy Study Research Group. (1991).

Grading diabeti retinopathy from stereoscopic color fundus photo-

graphs—an extension of the modified Airlie House classification.

ETDRS Report No. 10. Ophthalmology, 98, 786–806.

Elashoff, J. D. (1999). nQuery Advisor, Version 3.0 User’s Guide. Los

Angeles, CA.

Ferris, B. G. (1978). Epidemiology standardization project (American Thora-

cic Society). American Review of Respiratory Disease, 118, 1–120.

Hosmer, D. W., & Lemeshow, S. (1989). Applied logistic regression. New

York: Wiley.

Hypertension Detection and Follow-up Program Cooperative Group.

(1976). The hypertensive detection and follow-up program. Preventive

Medicine, 5, 207–215.

Klein, B. E. K., Davis, M. D., Segal, P., Long, J. A., Harris, W. A., Haug,

G., Magli, Y., & Syrjala, S. (1984). Diabetic retinopathy. Assessment of

severity and progression. Ophthalmology, 91, 10–17.

Klein, R., Klein, B. E. K., Moss, S. E., et al (1984). Prevalence of diabetes

mellitus in southern Wisconsin. American Journal of Epidemiology,

119, 54–61.

Klein, R., Klein, B. E. K., & Moss, S. E. (1991). The incidence of gross

proteinuria in people with insulin-dependent diabetes mellitus. Archives

of Internal Medicine, 151, 1344–1348.

Klein, R., Klein, B. E. K., & Moss, S. E. (1996). Relation of glycemic

control to diabetic microvascular complications in diabetes mellitus.

Annals of Internal Medicine, 124, 90–96.

Klein, R., Klein, B. E. K., Moss, S. E., & Cruickshanks, K. J. (1994). The

Wisconsin epidemiologic study of diabetic retinopathy: XIV. Ten-year

incidence and progression of diabetic retinopathy. Archives of Ophthal-

mology, 112, 1217–1228.

Klein, R., Klein, B. E. K., Moss, S. E., Davis, M. D., & DeMets, D. L.

(1984a). The Wisconsin epidemiologic study of diabetic retinopathy: II.

Prevalence and risk of diabetic retinopathy when age at diagnosis is less

than 30 years. Archives of Ophthalmology, 102, 520–526.

Klein, R., Klein, B. E. K., Moss, S. E., Davis, M. D., & DeMets, D. L.

(1984b). The Wisconsin epidemiologic study of diabetic retinopathy:

III. Prevalence and risk of diabetic retinopathy when age at diagnosis is

30 or more years. Archives of Ophthalmology, 102, 527–532.

Klein, R., Klein, B. E. K., Moss, S. E., Davis, M. D., & DeMets, D. L.

(1989a). The Wisconsin epidemiologic study of diabetic retinopathy:

IX. Four-year incidence and progression of diabetic retinopathy when

age at diagnosis is less than 30 years. Archives of Ophthalmology, 107,

237–243.

Klein, R., Klein, B. E. K., Moss, S. E., Davis, M. D., & DeMets, D. L.

(1989b). The Wisconsin epidemiologic study of diabetic retinopathy:

X. Four-year incidence and progression of diabetic retinopathy when

age at diagnosis is 30 years or more. Archives of Ophthalmology, 107,

244–249.

Klein, R., Moss, S. E., Klein, B. E. K., Davis, M. D., & DeMets, D.

L. (1989). The Wisconsin epidemiologic study of diabetic retino-

pathy: XI. The incidence of macular edema. Ophthalmology, 96,

1501–1510.

Linsenmeier, R. A., Braun, R. D., McRipley, M. A., Padnick, L. B., Ahmed,

D. L., Hatchell, D. L., McLeod, D. S., & Lutty, G. A. (1998). Retinal

hypoxia in long-term diabetic cats. Investigative Ophthalmology and

Visual Science, 39, 1647–1657.

Mantel, N. (1963). Chi-square tests with one degree of freedom: extensions

of the Mantel–Haenszel procedure. Journal of the American Statisti-

cian Association, 58, 690–700.

Moss, S. E., Klein, R., & Klein, B. E. K. (1991). Cause-specific mortality in

a population-based study of diabetes. American Journal of Public

Health, 81, 1158–1162.

Moss, S. E., Klein, R., & Klein, B. E. K. (1992). The prevalence and

incidence of lower extremity amputation in a diabetic population. Ar-

chives of Internal Medicine, 152, 610–616.

Moss, S. E., Klein, R., Klein, B. E. K., Spennetta, T. L., & Shrago, E. S.

(1988). Methodologic considerations in measuring glycosylated hemo-

globin in epidemiologic studies. Journal of Clinical Epidemiology, 41,

645–649.

Nakhostine, N., Nadeau, R., & Lamontagne, D. (1997). Altered hypoxia-

induced coronary vasodilatation in diabetic rabbit heart. Canadian Jour-

nal of Physiology and Pharmacology, 75, 1267–1272.

Nunn, A. J., & Gregg, J. (1989). New regression equation for predict-

ing peak expiratory flow in adults. British Medical Journal, 298,

1068–1070.

Primhak, R. A., Whincup, G., Tsanakas, J. N., & Milner, R. D. (1987).

Reduced vital capacity in insulin-dependent diabetes. Diabetes, 36,

324–326.

Quick-Step Fast Hemoglobin Test System (1981). Akron, OH: Isolab.

Ramirez, L. C., Dal Nogare, A., Hsia, C., Arauz, C., Butt, I., Strowig, S. M.,

Schnurr-Breen, L., & Raskin, P. (1991). Relationship between diabetes

control and pulmonary function in insulin-dependent diabetes mellitus.

American Journal of Medicine, 91, 371–376.

Sandler, M., Bunn, A. E., & Stewart, R. I. (1986). Pulmonary function in

young insulin-dependent diabetic subjects. Chest, 90, 670–675.

Schuyler, M. R., Niewoehner, D. E., Inkley, S. R., & Kohn, R. (1976).

Abnormal lung elasticity in juvenile diabetes mellitus. American Review

of Respiratory Disease, 113 (1), 37–41.

Sobel, B. J., & Emergil, C. (1964). Subject effort and the expiratory flow

rate. American Review of Respiratory Disease, 89, 402–408.

Veves, A., Donaghue, V. M., Sarnow, M. R., Giurini, J. M., Campbell, D.

W., & LoGerfo, F. W. (1996). The impact of reversal of hypoxia by

revascularization on the peripheral nerve function of diabetic patients.

Diabetologia, 39, 344–348.

Vracko, R., Thorning, D., & Huang, T. W. (1979). Basal lamina of alveo-

lar epithelium and capillaries: quantitative changes with aging and in

diabetes mellitus. American Review of Respiratory Disease, 120 (5),

973–983.

Walton, S., Byrd, R. R. Jr., Fields, C. L., Ossorio, M. A., & Roy, T. M.

(1994). Abnormal pulmonary function and juvenile onset diabetes mel-

litus. Journal of the Kent Medical Association, 92, 101–104.

Wright, B. M. (1978). A miniature Wright peak flow meter. British Medical

Journal, 2, 1627–1628.

B.E.K. Klein et al. / Journal of Diabetes and Its Complications 15 (2001) 301–306306