After latest studies, is HDL still the 'good' cholesterol?
ORLANDO -- For years, HDL cholesterol was known as "the good cholesterol." A low level was associated with increased cardiovascular risk. More HDL cholesterol was thought to be cardioprotective, perhaps even capable of offsetting at least some of the risk conferred by a high LDL cholesterol level.
But it turns out that when it comes to HDL cholesterol, more isn't always better.
A new study utilizing the increasingly popular "big data" analytic approach indicates that the relationship between HDL cholesterol level and mortality isn't linear. Instead, it's U-shaped, with both low and high HDL cholesterol levels being associated with significantly increased mortality risk, Dr. Dennis T. Ko reported at the American Heart Association scientific sessions.
This is the latest in a string of bad news regarding the former "good cholesterol." Large, multicenter, randomized trials of niacin and cholesterol ester transfer protein (CETP) inhibitors aimed at boosting HDL cholesterol levels in patients with low HDL cholesterol as a means of reducing cardiovascular risk succeeded in raising HDL cholesterol, but with no impact on major cardiovascular endpoints, noted Dr. Ko of the Institute for Clinical Evaluative Sciences and the University of Toronto.
He presented a population-based study of 631,762 Ontario residents who were free of prior cardiovascular disease and at least 40 years old in 2008. This study group, known as the CANHEART cohort, was formed by combining 17 large regional databases.
The advantage of working with such a large study population is that it provides new and statistically powerful insights into the impact of the full range of HDL cholesterol values, not just in terms of cardiovascular events but the full spectrum of disease, Dr. Ko explained.
Up until now, the conventional knowledge about HDL cholesterol has been based largely upon relatively small observational studies, such as the Framingham Heart Study; most of those studies didn't look at noncardiovascular events.
During a mean 4.9 years of follow-up of the CANHEART cohort, 9,339 deaths occurred in men and 8,613 in women.
In an analysis adjusted for age, non-HDL cholesterol levels, cardiac risk factors, sex, comorbid conditions, and income, HDL cholesterol levels below the reference range of 41-50 mg/dL in women and 51-60 mg/dL in men were associated with increased risks of mortality from all three causes: cardiovascular, cancer, and other. The lower the HDL cholesterol level, the greater the risks.
As HDL cholesterol groupings moved decile by decile above the reference ranges, there was no protective effect seen against cardiovascular or noncardiovascular deaths.
Instead, the risk of death due to causes other than heart disease or cancer took a turn upward as HDL cholesterol levels approached the outer end of the bell curve, achieving significance in men with an HDL cholesterol of 71-80 mg/dL and peaking in those with a level greater than 90 mg/dL.
In women, the U-shaped curve was shallower, with an increased mortality risk--again, as in men, restricted to causes other than cancer or heart disease--becoming statistically significant only in women with an HDL cholesterol level greater than 90 mg/dL, Dr. Ko continued.
It's worth noting that men with an HDL cholesterol level of 81 mg/dL or above also showed a trend for increased risks of both cardiovascular and cancer deaths, although this didn't reach statistical significance.
Patients at the low end of the HDL cholesterol spectrum had an increased prevalence of unhealthy lifestyle, COPD and other comorbid conditions, cardiac risk factors, and low income.
In contrast, those with high HDL cholesterol levels were more likely to have a body mass index below 25 kg/[m.sup.2], engage in 30 minutes or more of brisk walking or other moderate exercise daily, and consume five or more servings of fruits and vegetables daily. So they were, overall, healthier.
On the other hand, they were also more likely to be heavy alcohol users, as defined by consuming five or more drinks per occasion at least once monthly during the year prior to study enrollment. Alcohol, like physical exercise, is known to boost HDL cholesterol levels.
These CANHEART data and other evidence warrant a reappraisal of HDL cholesterol as a cardiovascular risk/protective factor, according to Dr. Ko.
"HDL is unlikely to represent a cardiovascular-specific risk factor, given similarities in its association with noncardiovascular outcomes," he observed.
Discussant Jacques Genest concurred.
"Maybe HDL cholesterol mass is the wrong biomarker for HDL function," opined Dr. Genest, professor of medicine at McGill University in Montreal.
He noted that a causal relationship between HDL cholesterol and cardiovascular risk has been cast into doubt not only by the negative randomized trials of niacin and the CETP inhibitors and the U-shaped mortality curve described by Dr. Ko, but also by randomized Mendelian genetic studies suggesting that genes causing HDL deficiency aren't linked to increased cardiovascular risk.
It's likely true that what's important is not HDL cholesterol levels as measured in conventional lipid panels, but rather HDL function.
HDL particles have many beneficial effects: antioxidative, anti-inflammatory, vasodilatory, antiapoptotic, and antithrombotic.
A better biomarker--one that reflects these functional benefits may be cholesterol efflux capacity, as was shown in a recent study by investigators in the Dallas Heart Study (N EnglJ Med. 2014; Dec 18;371:2383-93), according to Dr. Genest.
Session cochair Dr. Christie Ballantyne was particularly interested in the 2.8% of CANHEART participants with an HDL cholesterol level greater than 90 mg/dL and their associated increased mortality from causes other than cancer and cardiovascular disease.
"I've seen similar data once before, in a Russian cohort. Cold weather and alcohol--I wonder if that's a factor here. The alcohol exposure we saw in the Russian cohort with the 90 HDL levels was really rather striking. And it wasn't cardiovascular or cancer mortality that they faced, it was other mortality. So, I wonder if it's not an alcohol-related thing, especially in places where you have long, cold winters," commented Dr. Ballantyne, professor of medicine and of molecular and human genetics at Baylor College of Medicine and director of the Center for Cardiovascular Disease Prevention at the Methodist Debakey Heart Center in Houston.
Dr. Genest was skeptical.
"I think clinicians would argue that seeing an HDL of 90 would be very unusual in a male even if he drinks like an American at a football game," the Canadian quipped. "It's probable that a genetic predisposition plus heavy alcohol is involved."
Dr. Ko said he and his coinvestigators are scrutinizing the "other deaths" in the very-high HDL cholesterol subgroup, looking for an increase in deaths due to liver cirrhosis, trauma, and other obvious alcohol-related causes. "We haven't really found a specific pattern," according to the physician.
Dr. Ballantyne was undeterred. "It may end up being that the person who drinks heavily has some general health issues where trouble occurs because of alcohol in combination with medications rather than a specifically alcohol-related death," he advised.
Key clinical point: Doubt has emerged about the validity of HDL cholesterol as a straightforward cardiovascular risk factor.
Major finding: The relationship between HDL cholesterol and mortality isn't linear, it's U-shaped, with increased risk seen at both low and high levels.
Data source: This registry study included 631,762 Ontario adults free of cardiovascular disease at baseline and followed for a mean of 4.9 years.
Disclosures: The study is sponsored by the Canadian Institutes of Health Research. The presenter reported having no financial conflicts of interest.
Please note: Illustration(s) are not available due to copyright restrictions.
|Printer friendly Cite/link Email Feedback|
|Publication:||Internal Medicine News|
|Date:||Jan 1, 2016|
|Previous Article:||ABIM delays MOC's practice assessment piece till 2019: ABIM will provide at least a full year's notice when the requirements are reinstated.|
|Next Article:||COPD, sudden cardiac death risk linked.|