GlaxoSmithKline P.L.C.'s diabetes drug Avandia does not increase the risk of heart-related hospitalizations and deaths, according to a company-funded study released yesterday.

Critics said flaws in the study design rendered the results meaningless, and the authors of the paper, published in the British journal Lancet, acknowledged that the data on heart-attack risks were "inconclusive."

Avandia generated $2.2 billion in 2008 sales, but those numbers have been falling since 2007, when Cleveland Clinic cardiologist Steve Nissen published research saying that Avandia increased the risk of heart attacks by about 40 percent. Some of Glaxo's own research has shown increased heart-attack risk.

In the study released yesterday, researchers randomly assigned about 4,400 people to take either Avandia with an older diabetes drug or a combination of two other older drugs.

The researchers found no difference between the two groups in the rate of cardiovascular problems that led to hospitalization or deaths over an average study period of about five years.

Avandia "carries no increased risk of overall cardiovascular death or hospitalization compared to the most commonly used diabetes medicines, metformin and sulfonylurea, which have been used for decades," lead researcher Philip D. Home of the University of Newcastle said in a news release.

But critics, who have included Nissen and federal government researchers, have criticized the study, known as Record.

About 40 percent of patients stopped taking Avandia before the end of the study.

"It is impossible to assess the safety of a drug when patients are not actually taking it," Nissen said yesterday.

GlaxoSmithKline spokeswoman Mary Anne Rhyne said most patients who stopped taking Avandia did so near the end of the trial, allowing researchers to draw conclusions.

Patients taking Avandia also were more likely to take a statin, which lowers cholesterol and could have affected the heart-attack results, Nissen said.

Another design problem was that the researchers knew which patients were taking which drugs, which can bias results. GlaxoSmithKline said it was not able to keep this information from researchers and patients, a method known as double-blinding, because some patients needed insulin, which is delivered by injection.

In commentary accompanying the study, Ravi Retnakaran and Bernard Zinman, scientists at Toronto's Mount Sinai Hospital, said: "Definitive conclusions about the relation between rosiglitazone and cardiovascular disease remain elusive, owing to study limitations."

Zinman has consulted for GlaxoSmithKline on Avandia in the past, but Retnakaran reported no conflicts.

Problems with the design of Record were one reason the FDA told Glaxo to begin another study, known as Tide, but the results of that will not be known for several years.

The Record study did confirm a higher rate of heart failure - already a well-known risk of Avandia and similar drugs. The study's conclusions were based on hopitalization and death rates for all types of heart problems, not just heart failure.

Contact staff writer Miriam Hill

at 215-854-5520 or hillmb@phillynews.com.