1. In order to determine whether rheological changes occur in neuropat
hic diabetic patients in the absence of smoking, proteinuria, retinopa
thy or other factors thought to influence haemorheology, three groups
of subjects were studied; 24 non-diabetic control subjects (C), 24 non
-neuropathic (D) and 24 neuropathic (N) diabetic patients, The groups
were matched for age, sex, type and duration of diabetes, No patient o
r control was a current smoker, No patient had clinically detectable r
etinopathy or microalbuminuria, Neuropathy was defined as a peroneal c
onduction velocity < 40 ms(-1). All subjects were tested resting semi-
recumbent after a light breakfast. 2. There were no significant differ
ences in rheological or microvascular parameters between uncomplicated
diabetic patients and non-diabetic control subjects, although peronea
l nerve motor conduction velocity was significantly reduced in otherwi
se uncomplicated diabetic patients [C 51.7 +/- 6.0 ms(-1) (mean +/- SD
) versus D 45.1 +/- 5.2 ms(-1) (P < 0.05 C versus D)]. 3. Transcutaneo
us oxygen and laser Doppler flux measured at 44 degrees C were higher
in control subjects than in neuropathic diabetic patients [C 76 +/- 16
mmHg versus D 71 +/- 10 mmHg versus N 63 +/- 9 mmHg, and C 72 +/- 40
flow units versus D 64 +/- 41 flow units versus N 50 +/- 26 flow units
respectively (both P not significant C versus D, P < 0.05 N versus C)
. 4. Erythrocyte aggregation, plasma viscosity and plasma fibrinogen w
ere all significantly higher in the neuropathic diabetic patients comp
ared with nondiabetic control subjects (all P < 0.05 N versus C), Eryt
hrocyte filtration was not significantly different between groups but
was lower in diabetic patients, Whole-blood viscosity (corrected to 45
% haematocrit) was significantly higher at both high (100 s(-1)) and l
ow (1 s(-1)) shear rates in neuropathic diabetic patients than in non-
diabetic control subjects (both P not significant C versus D, P < 0.05
N versus C). There were no significant differences in whole-blood vis
cosity at a shear rate of 0.01 s(-1). 5. In summary, there were no sig
nificant differences in rheological or microvascular parameters betwee
n matched groups of uncomplicated diabetic patients and control subjec
ts, but erythrocyte aggregation, fibrinogen and plasma and corrected w
hole-blood viscosity were all significantly different in neuropathic d
iabetic patients compared with control subjects, as were assessments o
f microvascular flow, These results suggest that haemorheological chan
ges are associated with disturbances of microvascular flow and diabeti
c peripheral neuropathy in the absence of other diabetic complications
.