BACKGROUND: Blood lipoprotein(a) Lp(a) concentrations are an important risk factor for atherosclerosis. The basis for this atherogenic property of Lp(a) and the factors that in¯uence its cross-population levels, however, remain poorly understood. OBJECTIVES: To investigate the relationship between serum Lp(a) and metabolic and anthropometric parameters in a healthy Kuwaiti population. DESIGN: Cross-sectional study. SUBJECTS: 177 (72 male, 105 female) randomly recruited healthy Kuwait Arabs aged 17 ± 60 y MEASUREMENTS: Metabolic parameters in serum: Lp(a), apo(a) phenotypes, lipids and lipoproteins, glucose and urate. Anthropometric parameters: body mass index (BMI) and waist:hip-ratio (WHR). RESULTS: The distribution of Lp(a) concentrations was positively skewed (median 153 mgal, range 0 ± 1086). Women had higher concentrations ± (194, 0 ± 1086) than men (117, 0 ± 779), P 0.069. Lp(a) and insulin concentrations were signi®cantly higher when the men and women were obese. In all subjects, there were signi®cant correlations between Lp(a) and BMI (r 0.23), total cholesterol (TC) (r 0. Multivariate analyses con®rmed BMI and low-density lipoprotein (LDL) as the signi®cant determinants of serum Lp(a).On apo (a) phenotyping, 114 (67%), 51 (30%) and 6 (4%) had single, double and null phenotypes respectively. The isoforms and their corresponding kringle IV repeat numbers were: F (14 repeats in 3%, mean Lp(a) 497 mgal); S1 (19 repeats in 14%, mean 245 mgal); S2 (23 repeats in 16%, mean 264 mgal); S3 (27 repeats in 35%, mean 236 mgal); and S4 (35 repeats in 28%, mean 235 mgal). DISCUSSION AND CONCLUSION: The results from the Kuwaiti population studied suggest that: (1) serum Lp(a) concentrations and distribution are similar to the pattern in Caucasians and Asians but not African ± Americans or Africans; (2) serum Lp(a) is variably in¯uenced by BMI and LDL ± the impact of either factor differs between the sexes; (3) there is a high frequency of the single-banded phenotype; (4) contrary to reports in some Caucasian and Asian populations, there is no simple relationship between kringle IV repeat numbers and plasma Lp(a) concentrations.