Pa. Grootenhuis et al., DISSIMILAR ASSOCIATION OF CONVENTIONAL IMMUNE-REACTIVE VERSUS SPECIFIC INSULIN WITH CARDIOVASCULAR RISK-FACTORS - A CONSEQUENCE OF PROINSULINAEMIA, Diabetes research and clinical practice, 40(2), 1998, pp. 81-90
In this study involving 365 non-diabetic elderly Caucasians, we examin
ed the relationship of immune-specific insulin (ISI), total immune-rea
ctive insulin (IRI), proinsulin (PI) and proinsulin-insulin ratio (PI:
ISI) to serum high-density lipoprotein cholesterol (HDL-C), triglyceri
de (TG); systolic blood pressure (SBP) and diastolic blood pressure (D
BP), mean blood pressure (MBP) and pulse pressure. In a multiple regre
ssion analysis, adjusted for age, sex and obesity, a 1.3-fold stronger
inverse association with HDL-C levels was found for IRI than for ISI,
with a 1.6-fold better fit of the regression equation. The positive a
ssociation of insulin with TG was 1.6-fold stronger for IRI compared t
o ISI, with a 2.5-fold better fit. In contrast, the positive associati
on of IRI with the various blood pressure parameters was 1.5-1.9-fold
weaker than for I:SI, with a 2.1-3.8-fold worse fit. Both PI:ISI ratio
and PI were independently associated with TG levels, but not with HDL
-C. The PI:ISI ratio, but not PI, was associated with blood pressure,
but dependent on glycaemia. In conclusion, compared to ISI, IRI overes
timates the association of insulin with serum lipids and underestimate
s the association of insulin with blood pressure. The use of non-speci
fic insulin assays may explain the inconsistencies in the findings of
previous epidemiological studies. (C) 1998 Elsevier Science Ireland Lt
d. All rights reserved.