To investigate possible relationships between plasma low density lipop
rotein (LDL) cholesterol and fasting plasma insulin in the elderly, cr
oss-sectional random samples of age cohorts (65, 75, 80 and 85 years,
n = 1188, M/F 38/62%) were studied in the neighbouring cities of Helsi
nki and Vantaa, Finland. Plasma total and high density lipoprotein (HD
L) cholesterol, plasma triglycerides, blood glucose and plasma insulin
were measured after an overnight fast. LDL cholesterol was calculated
using the Friedewald equation. Statistical analyses were performed se
parately in subjects with non-insulin-dependent diabetes mellitus (NID
DM, n = 219) and non-diabetic subjects (n = 969). Comparison of lipid
levels by insulin quartile (I < 7.4 IU/l, II 7.4-10.0, III 10.1-15.0,
IV > 15.0) showed that total and LDL cholesterol decreased in the high
est insulin quartile (P = 0.003). This trend prevailed after adjustmen
ts for age, gender, body mass index, blood glucose and serum triglycer
ides,.and it was significant also in normotriglyceridemic (serum trigl
ycerides < 2.3 mmol/l) subjects. Furthermore, the association between
high insulin and lower cholesterol was seen in normoglycemic (fasting
blood glucose < 6.7 mmol/l) and diabetic subjects. Lower LDL cholester
ol in elderly subjects with higher fasting insulin may reflect poor he
alth or a 'harvesting' effect, but the results may also be due to effe
cts of insulin on LDL catabolism and/or cholesterol absorption.