The influence of hepatic lipase (HL) and lipoprotein lipase (LPL) activity on the low density lipoprotein (LDL) subclass pattern was studied in a population of males with coronary heart disease and without severe hypercholesterolemia. LDL subclass patterns, lipases and plasma lipoproteins were determined in 326 patients. In part of the study population, fasting insulin and glucose levels were also determined. The LDL subclass pattern was determined by gradient gel electrophoresis (GGE) and classified according to Austin et al. (J. Am. Med. Assoc. 260 (1988) 1917 (predominantly large LDL = A-pattern, predominantly small LDL = B-pattern). An LDL subclass A-pattern was exhibited by 199 subjects; 108 exhibited a B-pattern. In 19 subjects no distinctive A- or B-pattern was present (A/B-pattern). Hepatic and lipoprotein activities differed significantly between patients with the A- or B-pattern. The median hepatic lipase activity was lower (384 vs. 417 mU/ml, P = 0.006), and the lipoprotein lipase activity higher (122 vs. 101 mU/ml, P = 0.001) in the A-pattern subjects than in the B-pattern subjects. In subjects with the A/B pattern the lipase activities were intermediate between the values in the A- and B-pattern subjects (HL 408 +/- 87 mU/ml, LPL 115 +/- 55 mU/ml). Plasma triglyceride, very low density lipoprotein (VLDL)-triglyceride, intermediate density lipoprotein (IDL)-triglyceride and LDL-triglyceride were higher in the patients with a B-pattern (+84%, +171%, +10% and +16%, respectively). Total plasma cholesterol was not different between A- and B-pattern subjects. VLDL- and IDL-cholesterol were higher in the B-pattern group (+174% and +66%, respectively), while LDL- and HDL-cholesterol were higher in the A-pattern group (+2 and +24%, respectively). In univariate analysis HL, LPL, plasma (and VLDL) triglyceride, HDL-cholesterol and IDL-cholesterol were each significantly associated with the LDL subclass pattern. In multivariate analysis plasma triglyceride (or VLDL-triglyceride) and HDL-cholesterol appeared to be independently associated with the LDL subclass pattern. No additional discriminative value of HL or LPL was found. Similar results were obtained if the patients with or without beta blocker were evaluated separately. An estimate of insulin resistance (EIR), calculated from plasma insulin and glucose in part of the study population (n = 145), was significantly higher in the subjects with a B-pattern than in those with an A-pattern (3.12 vs. 2.00, P < 0.003). EIR correlated positively with plasma triglyceride (P < 0.0001), but not with HL or LPL.(ABSTRACT TRUNCATED AT 400 WORDS)