"We were so surprised with the findings about immunity that we repeated the tests over and over again as more patients entered the trial," said lead researcher Dr Barbara Andersen from Ohio State University in Columbus.

Writing in the current edition of the Journal of Clinical Oncology, the team explains that little research has been undertaken into the possible benefits of psychological interventions for cancer survivors.

To investigate further, they studied 227 women who had undergone breast cancer surgery and were awaiting radiation therapy, hormonal therapy or chemotherapy. Information on the participants' psychological, behavioural, medical and treatment status was gathered during a structured interview and blood samples were taken.

The women were then randomly assigned to receive usual care (control group) or attend 18 counselling sessions over 4 months, each lasting 1.5 hours. Counselling sessions included standard psychosocial strategies to reduce stress, improve mood, alter health behaviours and aid adherence to breast cancer treatment and care.

The researchers found, as predicted, that women who attended counselling had significantly lower anxiety and more perceived support than those in the control group. In addition, they had improved their dietary and smoking habits, whereas women given usual care made no significant changes to their diet and actually increased their cigarette use.

The researchers also examined the participants' levels of natural killer cells and T-cells, which help identify and destroy infected or mutated cells that can lead to cancer and other diseases. Low levels of such cells would indicate a suppressed immune system, putting the patient's health at risk.

Importantly, women given usual care showed a significant decline in T-cell proliferation. However, the patients who received counselling had a sustained or improved T-cell proliferation.

With patient monitoring continuing, Dr Andersen added: "Ours is the first study to use an experimental design to discover what, if any, relationship exists between psychological intervention and risk of recurrence.

"Statistically, we just don't have enough data yet to determine that, but, so far, we know that the intervention is beneficial and the findings are robust and that gives us greater confidence we have a strong test regarding impact on recurrence."