A long-running study in the United Kingdom has shown for the first time that combined therapy with two of the main drugs for bipolar disorder is better than using either drug alone.

The study, conducted over the past two years with 330 patients from four nations, showed that combining the anti-seizure drug valproate (Depakote) with the mood stabilizer lithium kept people healthier than using either medication alone.

The results were announced last week at the Eighth International Conference on Bipolar Disorder, which was organized by UPMC and the University of Pittsburgh and held at the David L. Lawrence Convention Center.

Dr. John Geddes, an Oxford University psychiatrist who is the lead investigator of the trial, said the results were particularly striking when the two-drug combination was compared with valproate. The combination was 40 percent better at preventing relapses than valproate alone.

That is important, he said, because the anti-seizure drug has been rising rapidly in popularity as a treatment for bipolar disease, which is characterized by mood swings between depression and mania.

Many psychiatrists today tend to start out giving bipolar patients one drug, and then if they don't respond, adding in a second one. But in the case of valproate, he said, "if a patient is thinking about having valproate therapy, it's probably better to offer lithium as well."

The combined therapy also had an 18 percent advantage over lithium alone, he said, but that was less statistically significant.

Dr. David Kupfer, the psychiatry department chairman at the University of Pittsburgh and one of the conference's organizers, agreed with Dr. Geddes' conclusion.

If a patient were being treated with lithium alone, Dr. Kupfer said he would not add valproate unless the patient didn't respond to the lithium, because the advantage of the combined therapy over lithium by itself is not as strong.

More than 3 million people in the United States are afflicted with bipolar disease, and an estimated 10 to 20 percent of them end up committing suicide.

The raw numbers from the Oxford study show just how tough the disease is.

Even the combination therapy group had a relapse rate of 54 percent. The group taking valproate alone had a relapse rate of 69 percent, Dr. Geddes said.

Because lithium was nearly as effective as the combination therapy, he added, it seems to have clear advantages over valproate -- yet prescription rates have been heading the other way.

One study showed that from 1990-2000, prescriptions for valproate climbed from 11 percent to 30 percent of bipolar patients, while those for lithium, a much older drug, fell from 50 percent to 30 percent. Dr. Geddes said.

One reason for that, he said, is that lithium can be dangerous if patients get too much of it, so doctors have to do regular blood tests to monitor patients' lithium levels.

Another reason for the shift, he said, was that pharmaceutical companies can't make much money with lithium, but can capitalize on marketing a secondary use for valproate, which was originally designed to stop epileptic seizures.

Too often today, Dr. Kupfer added, the drugs used by doctors are the ones promoted by drug companies rather than the ones shown to be effective in scientific studies, "and it's our job in terms of thinking about evidence-based medicine to do something about that."

Even though lithium has been around for decades, researchers are still trying to unravel exactly how it works, Dr. Geddes said.

Once they have determined better how it changes brain activity and chemistry, it may be possible to design new drugs that target the same part of the brain but have fewer side effects.