The recent study was a "good study but only a small study" Sacks said. The second study is "designed too see if we can prevent heart attacks and get a definitive answer," he said.

The NIH classifies cholesterol levels of healthy individuals into three groups: "high" with cholesterol levels above 240, "border-line" with levels around 200, and "desirable" with levels less than 200. The classifications change for patients with coronary disease.

Cholesterol levels above 170 are labeled sub-optimal and levels below 170 are optimal.

Despite the unexpected result, Dr. Basil M. Rifkind, senior scientific advisor to the division of vascular disease at HIH, said Sacks' study would not alter the NIH recommendations.

"Dr. Sacks' study raises the issue that we need more information before we change guidelines," Rifkind said, "It is not justified to alter cholesterol recommendations and treatment based on this study, since other studies in this field suggest evidence to the contrary."

"We still need to know a lot more about patients with low cholesterol and coronary disease," Rifkind said.

Rifkind summarized Sacks' study as a "well done study, a small study with not all the evidence relating to this study pointing to Sacks' results."