The researchers divided 77 men and women with diabetes into two groups and measured the liver fat of 44 of the participants.

Why only 44 of the 77 participants? I called Dr. Stewart at the annual meeting to find out.

"When we started the study, we didn't plan to study hepatic fat," he replied. "But after we started, some of my colleagues persuaded me to include it."

Dr. Stewart says the team's study is the first to specifically demonstrate the beneficial role that exercise plays in controlling hepatic fat levels in people with diabetes.

The researchers put half of the study participants through a moderate program of sustained aerobic exercise consisting of three weekly 45-minute sessions. The participants could bicycle, run on a treadmill, or take brisk walks. In addition, they lifted stacked weights for about 20 minutes, also three times a week -- and not at a heavy-duty pace. They asked the other half of the participants to avoid any formal aerobic fitness or gym classes.

Special magnetic resonance imaging scans showed much lower levels of liver fat in the active group, which remained the same in the non-exercising group. The exercising group had 5.6 percent liver fat after six months. The non-exercising group had 8.5 percent.

Until Dr. Stewart and his team does more studies, we can't tell how significant this is. He says his team's next steps will be to analyze the long-term effect of moderate exercise on diabetes.

We already knew that exercise makes you feel and look better. It takes glucose out of your blood to use for energy. It helps prevent heart disease, depression, and even some forms of cancer. If you do enough, it will help you to lose weight.

But sadly, the link between exercise and fatty liver came too late for Catherine. It is timely for everyone else who has type 2 diabetes.