To identify high-density lipoprotein (HDL) subfractions associated with longevity in men.DESIGN:
Fifty-three-year prospective follow-up of Gofman's Livermore Cohort between 1954 and 2008.SETTING:
Lawrence Livermore National Laboratory.PARTICIPANTS:
One thousand one hundred forty-four men who consented to the study, had analytic ultracentrifuge measurements of lipoprotein subfractions at baseline, and were old enough at baseline to have survived to age 85 during follow-up.MEASUREMENTS:
Survival was determined according to participant contact, Social Security Death Index, and National Death Index.RESULTS:
Three hundred ninety men survived to 85 years old (34.1%). Survivors were less likely than nonsurvivors to be in the lowest HDL3 (% (standard error) 18.5% (2.0%) vs 27.3% (1.6%), P < .001) and HDL2 (22.1% (2.1%) vs 27.7% (1.6%), P = .04) quartiles. Logistic regression analyses showed that the lowest HDL3 quartile significantly predicted shorter longevity (P = .002), whereas the linear increases per mg/dL of HDL3 did not (P = .38), suggesting a risk threshold proximal to the 25th percentile. Men who were above the 25th HDL3 percentile had 70% greater odds of surviving until age 85 than those below this level, which persisted when adjusted for HDL2, very low-density lipoprotein (LDL), and standard risk factors. Proportional hazard analyses of survival before age 85 showed that being in the lowest HDL3 quartile increased age-adjusted cancer risk by 39% (P = .05) and noncancer risk by 23% (P = .04) when adjusted for other risk factors. Survivors also smoked less (mean ± SD 0.31 ± 0.48 vs 0.57 ± 0.56 packs/d, P < .001), had lower systolic (118.36 ± 11.08 vs 122.81 ± 13.55 mmHg, P < .001) and diastolic (70.61 ± 8.59 vs 73.14 ± 9.22 mmHg, P < .001) blood pressures and lower LDL mass (359.55 ± 80.42 vs 374.37 ± 86.10 mg/dL, P = .009) and total cholesterol concentrations (229.51 ± 43.21 vs 235.89 ± 45.40 mg/dL, P = .04) than nonsurvivors.CONCLUSION:
Low HDL3 reduces the odds of extended survival in men, independent of HDL2, other lipoproteins, and standard risk factors.