[Lipoprotein (a) in a population-based study: more significant in Turkish women than men?]

Serum lipoprotein(a) [Lp(a)] concentrations, determined in 665 persons in the 2003/04 survey of the Turkish Adult Risk Factor Study, were investigated in regard to distribution, determinants and relationship to cardiovascular risk factors, metabolic syndrome (MS) and coronary heart disease (CHD).

Diagnosis of MS was based on Adult Treatment Panel III criteria, that of CHD on the presence of clinical findings and Minnesota coding of resting electrocardiograms. Metabolic syndrome was observed in 44%, CHD in 14% of the study sample. Behring nephelometry was used for Lp(a) values measurements which were log-transformed for analyses because of skewing.

Geometric mean values of Lp(a) in 286 men and 379 women, aged 55.5 +/-12.0 years, were 9.46+/-2.90 mg/dL and 10.46+/-3.00 mg/dL (p>0.2), respectively. Apart from a slight correlation with age, Lp(a) exhibited significant positive correlations with apolipoproteins A-I and B, low density lipoprotein-cholesterol (LDL-C) (r =0.15), total cholesterol, high density lipoprotein-cholesterol (HDL-C), systolic blood pressure and log C-reactive protein, and inverse ones with thyroid stimulating hormone (r =-0.25) in men, and log gamma glutamyltransferase in women. Further 10 variables were not significantly correlated in either gender. In linear regression analyses for independent covariates of Lp(a), positive associations were noted with serum total cholesterol and systolic blood pressure, and inverse ones with waist circumference, triglycerides and (only in women) with gamma glutamyltransferase. Logistic regression analyses revealed in men no association with either MS or CHD likelihood. Among women, age-adjusted Lp(a) was associated inversely at a borderline significance with MS, as did levels of Lp(a) >30 mg/dl vs. the remaining sample, controlled for age and MS, display an odds ratio (OR) of 1.62 for prevalent CHD (p=0.20). An OR of 1.92 (p<0.19) was noted in all adults for the coexistence of Lp(a) >30 mg/dl and LDL-C >150 mg/dl, after controlling for age, MS, smoking status and LDL-C categories.

Lipoprotein(a), the variance of which is known to be overwhelmingly due to the apo(a) isoforms, proved to have a significant inverse independent association with a measure of abdominal obesity. Lipoprotein(a) levels appeared not to be associated with risk for MS or CHD among men. In women, however, high Lp(a) levels were accompanied with an environment less prone to MS, and - without attaining significance -- tended to be associated with CHD likelihood, independent of age and MS. Further studies are warranted in this area.