BOSTON — High-density lipoprotein cholesterol may not be as beneficial as once believed, according to the results of a large data analysis.
Among women, high levels of HDL cholesterol were associated with a near 40% increased risk for unprovoked venous thromboembolic events (VTE), compared with women with lower levels of the so-called “good cholesterol,” according to Dr. Knut H. Borch from the Center for Atherothrombotic Research, University of Trosmø in Norway.
“We found no association between HDL cholesterol levels and risk of provoked VTE in men or in women. However, high HDL cholesterol was associated with increased risk of unprovoked VTE in women, but not in men,” Dr. Borch explained at a meeting of the International Society on Thrombosis and Haemostasis.
The investigators looked for a potential protective effect of HDL on VTE by examining data from a prospective population-based study on all men and women age 25 and older in the town of Tromsø.
A total of 26,676 adult men and women were included in the study, after exclusion of some participants for missing HDL data, or for other reasons. The investigators looked for all first lifetime events of VTE during follow-up from the date of enrollment (1994-1995) through Sept. 1, 2007.
Cases were identified by a computerized index of discharge diagnoses, autopsy registry, and radiology procedure registry. VTE was confirmed either by diagnostic procedure or autopsy, and the data collected included diagnosis of deep vein thromboembolism or pulmonary embolism, signs and symptoms consistent with VTE and VTE treatment.
The investigators further classified VTE as being provoked or unprovoked. Provoked VTEs were those associated with major surgery or an acute medical condition within 8 weeks of the event, cancer at the time of the event, prolonged immobilization, or other known risk factors. Unprovoked VTEs were those occurring in the absence of any of the known provoking factors.
The median follow-up was 12.5 years, and a total of 288,572 person-years.
During that time, there were 458 VTEs in the cohort, 191 (42%) of which were identified as provoked, and of this group, 64% were DVT. This translated into an incidence of 1.6 VTEs per 1,000 person-years. There were no between-gender differences in provoking factors or type of VTE.
In a multivariate analysis, the researchers found no significant differences between the lowest and highest HDL quartiles for either women (hazard ratio .90, .73-1.11) or men (HR 1.06, .83-1.35).
In terms of unprovoked VTE events, however, there was a significantly increased risk for VTE among women with the highest versus the lowest HDL quartiles (HR 1.39, 1.10-1.75). There was no significant difference in unprovoked VTE events among men in the highest versus lowest HDL quartiles (HR 1.15, .87-1.53).
Dr. Borch said that the strengths of the study include its prospective design, large number of participants, high compliance rate (77%), and exclusive hospital care by the University Hospital of Northern Norway. Limitations included HDL blood draws performed at different times of day with patients in a nonfasting state.
“However, HDL is not known to have a substantial diurnal variation or to be severely affected by meals,” Dr. Borch said.
He also noted that a common problem of cohort studies is that risk factors for individual patients may change over time, especially when there is a long follow-up period. Additionally, the researchers had no information about treatment during follow-up, and, therefore, could not control for it.
The findings are consistent with those of two previous cohort studies examining a possible link between HDL cholesterol levels and VTE, and raise doubts about potential antithrombotic properties of HDL particles (Blood 2008;112:2675-80; J. Thromb. Haemost. 2009;7:588-96).
Dr. Borch said that further studies are needed to determine whether the increased risk of unprovoked VTE in women is a direct result of HDL particles or of some unrecognized confounder.
Dr. Borch reported that neither he nor his colleagues had any financial disclosures relevant to the study.