A. Jager et al., Prognostic implications of retinopathy and a high plasma von Willebrand factor concentration in type 2 diabetic subjects with microalbuminuria, NEPH DIAL T, 16(3), 2001, pp. 529-536
Background. Microalbuminuria in subjects with type 2 diabetes may be hetero
geneous with respect to clinical features, renal histology, and prognosis.
There may be at least two types of microalbuminuria in diabetes, namely wit
h and without generalized endothelial dysfunction. We investigated whether,
among microalbuminuric subjects with type 2 diabetes, the presence of gene
ralized endothelial dysfunction, as indicated by the presence of retinopath
y or a high plasma von Willebrand factor (vWf) level, has prognostic implic
ations.
Methods. In 173 type 2 diabetic subjects of a population-based cohort, we a
ssessed the urinary albumin-to-creatinine ratio, the plasma vWf level, and
the presence of retinopathy. The main outcome was cardiovascular mortality.
Results. The absolute difference in 7 years' cardiovascular mortality betwe
en microalbuminuric (albumin-to-creatinine ratio 2.0-30.0 mg/mmol) and norm
oalbuminuric subjects was higher in the presence as compared to the absence
of retinopathy (55.6 vs 11.1%). The age- and sex-adjusted relative risk (9
5% confidence interval) of cardiovascular mortality, as compared to normoal
buminuric subjects without retinopathy, was 1.1 (0.1-9.2) for normoalbuminu
ric subjects with retinopathy, 1.8 (0.5-6.7) for microalbuminuric subjects
without retinopathy, and 9.8 (3.1-30.9) for microalbuminuric subjects with
retinopathy. The absolute difference in risk of 7 years' cardiovascular mor
tality between microalbuminuric and normoalbuminuric subjects was higher in
the presence as compared to the absence of a high (>1.89 IU/ml) vWf level
(49.8 vs 16.4%). The age- and sex-adjusted relative risk of cardiovascular
mortality, as compared to normoalbuminuric subjects without a high vWf leve
l, was 1.5 (0.4-5.5) for normoalbuminuric subjects with a high vWf level, 2
.6 (0.7-9.6) for microalbuminuric subjects without a high vWf level, and 12
.0 (2.9-49.5) for microalbuminuric subjects with a high vWf level. These di
fferences in risk of cardiovascular mortality did not change materially aft
er further adjustment for known duration of diabetes, hypertension, creatin
ine clearance, level of glycated haemoglobin and high-density lipoprotein c
holesterol, and presence of cardiovascular disease. Analysis of all-cause i
nstead of cardiovascular mortality showed a similar difference in risk of m
ortality between microalbuminuric subjects with or without retinopathy or a
high vWf level.
Conclusion. Among type 2 diabetic subjects with microalbuminuria, the prese
nce of retinopathy or a high plasma vWf level affects the risk of cardiovas
cular death. Although larger studies are necessary, these findings support
the concept that microalbuminuria in type 2 diabetes can occur in the absen
ce or the presence of generalized endothelial dysfunction, and that the lat
ter is a much more 'malignant' condition than the former.