A pooled analysis of the association of isolated low levels of high-density lipoprotein cholesterol with cardiovascular mortality in Japan

• Published in: European journal of epidemiology 2017-07, Vol.32 (7), p.547-557
• Main Authors: Hirata, Takumi, Sugiyama, Daisuke, Nagasawa, Shin-ya, Murakami, Yoshitaka, Saitoh, Shigeyuki, Okayama, Akira, Iso, Hiroyasu, Irie, Fujiko, Sairenchi, Toshimi, Miyamoto, Yoshihiro, Yamada, Michiko, Ishikawa, Shizukiyo, Miura, Katsuyuki, Ueshima, Hirotsugu, Okamura, Tomonori
• Format: Article
• Language: English
• Subjects: Abnormalities; Adult; Aged; Aged, 80 and over; Biomarkers - blood; Cardiology; Cardiovascular disease; Cardiovascular diseases; Cardiovascular Diseases - blood; Cardiovascular Diseases - mortality; Cholesterol; Cholesterol, HDL - blood; Cohort Studies; Confidence intervals; Coronary artery disease; Coronary Disease - mortality; Epidemiology; Female; Follow-Up Studies; Hazards; Health risk assessment; Heart; Heart diseases; Hemorrhage; High density lipoprotein; Humans; Infectious Diseases; Ischemia; Japan - epidemiology; Low density lipoprotein; Male; Medicine; Medicine & Public Health; Men; META-ANALYSIS; Middle Aged; Oncology; Proportional Hazards Models; Public Health; Risk Factors; Stroke; Stroke - mortality
• ISSN: 0393-2990; 1573-7284; 1573-7284
• DOI: 10.1007/s10654-016-0203-1

Low levels of serum high-density lipoprotein cholesterol (HDL-C) have been shown to be associated with increased risk of coronary heart disease (CHD). However, because this is usually observed in the context of other lipid abnormalities, it is not known whether isolated low serum HDL-C levels are an independent risk factor for CHD. We performed a large pooled analysis in Japan using data from nine cohorts with 41,206 participants aged 40-89 years who were free of cardiovascular disease at baseline. We divided participants into three groups: isolated low HDL-C, non-isolated low HDL-C, and normal HDL-C. Cohort-stratified Cox proportional hazards models were used to estimate multivariate-adjusted hazard ratios (HRs) for death due to CHD, ischemie stroke, and intracranial cerebral hemorrhage; during a 12.9-year follow-up, we observed 355, 286, and 138 deaths, respectively, in these groups. Non-isolated low HDL-C was significantly associated with increased risk of CHD compared with normal HDL-C (HR 1.37, 95 % confidence interval (CI) 1.04-1.80); however, isolated low HDL-C was not. Although isolated low HDL-C was significantly associated with decreased risk of CHD (HR 0.51, 95 % CI 0.29-0.89) in women, it was significantly associated with increased risk of intracranial cerebral hemorrhage in all participants (HR 1.62, 95 % CI 1.04-2.53) and in men (HR 2.00, 95 % CI 1.04-3.83). In conclusion, isolated low HDL-C levels are not associated with increased risk of CHD in Japan. CHD risk may, therefore, be more strongly affected by serum total cholesterol levels in this population.