Apparently, it’s possible to have too much of a good thing.

At least that’s what researchers are suggesting about high-density lipoprotein (HDL) cholesterol, widely known as the “good” type of cholesterol.

Danish researchers studied 116,508 adults in their country’s general population for an average of six years. Mean age of the sample population was 57, and 55 percent of participants were women. A total of 10,678 total deaths occurred during follow-up.

“The association between HDL cholesterol on a continuous scale and all-cause mortality was U-shaped for men and women, as both low and high concentrations were associated with high all-cause mortality,” wrote lead researcher Christian M. Madsen, of the department of health and medical sciences at the University of Copenhagen, and colleagues in the European Heart Journal. “The association between extreme high HDL cholesterol and high mortality was most pronounced for men.”

After adjusting for other factors, HDL cholesterol levels of 73 milligrams per deciliter in men and 93 in women were associated with the lowest risk of all-cause mortality. Compared to those levels, men with HDL of 97 to 115 mg/dL had a 36 percent increased risk for death, and more than double the risk above 116.

Women had death risk increases of 10 and 78 percent, respectively, for HDL levels of 116-134 and 135 or above.

Similar to other studies, low HDL levels were associated with greater mortality. At less than 39 mg/dL, the lower the HDL, the greater the risk for death.

The extreme high HDL group in both men and women were found to be at more than double the risk for cardiovascular mortality when compared to the reference groups, the researchers found. Low concentrations of HDL cholesterol were associated with the greatest risk of cancer-specific mortality.

The authors noted genetic variants could play a role in the link between mortality and extreme high HDL cholesterol concentrations. Another implication of the study is medications designed to substantially raise HDL cholesterol levels, such as cholesteryl ester transfer protein inhibitors, may need to be re-evaluated in clinical use.

“Current guidelines recommend measuring HDL cholesterol for risk estimation and before initiation of lipid-lowering therapy,” they wrote. “Our results indicate that the common belief that, the higher the concentration of HDL cholesterol the better, does not hold for extreme high concentrations, as the relationship between HDL cholesterol and all-cause mortality was not linear over the entire range of HDL concentrations.”

Two characteristics of the study population could limit the impact of the findings, Madsen and colleagues acknowledged. First, the relative low number of participants with high HDL cholesterol levels could limit the statistical power of the results. Also, considering all participants were white and of Danish descent, the findings may not apply to other regions and ethnicities.