Cancer survival rates are rising - but specialists are unsure whether people are really living longer or merely being diagnosed earlier.

Figures published yesterday by the Office for National Statistics show five year breast cancer survival improved by 6 per cent among women diagnosed in 1991-93 compared with those diagnosed in 1986-90.

For men with prostate cancer, survival rose 7 per cent in the same period, and for either sex survival of colon and rectal cancer increased by 1 to 4 per cent, but lung cancer survival declined by 1 per cent.

The figures suggest survival rates are accelerating. A study published in the British Medical Journal yesterday, based on earlier figures from 1981-85 and 1986-90, showed overall cancer survival rates had improved by 3.3 per cent.

Survival from cancer is partly affected by treatment and partly by the point at which the disease is diagnosed. The earlier a cancer is found, the longer a patient will appear to have lived with it even if treatment has no effect.

In the case of prostate and breast cancer, the increase in screening has made a major contribution to the apparent improvement in survival. For breast cancer, the largest improvement in survival, of 7 to 10 per cent, was among women aged 50 to 69, compared with 1 to 4 per cent increases for older and younger women. Women aged 50-64 are routinely screened.

For prostate cancer, the increase was 7 to 9 per cent in men aged 50 to 79, compared with 1 to 3 per cent for older and younger men. This is the first improvement since the 1980s and almost certainly reflects the growth of PSA testing (a blood test for prostate-specific antigen) which may indicate the presence of a tumour.

The Office for National Statistics said: "Earlier diagnosis from mammography [breast screening] or PSA testing will increase survival time even if death is not delayed by treatment." But the fact that the increase in one-year and five-year survival rates are different suggests that better treatment is helping too, it said.

Dr John Toy, the medical director of the Imperial Cancer Research Fund, said: "In breast cancer, one can be much more optimistic that treatment has postponed death. With prostate cancer you have to be much more guarded because there has been no improvement in treatment in the way there has been in breast cancer."

The main advance in breast cancer treatment has been the wider use of tamoxifen, the hormonal drug, after surgery. Improvements in survival from bowel cancer are also likely to be real because screening is almost non-existent and there have been advances in surgical techniques and chemotherapy.

But, in prostate cancer, there is no new treatment or surgical technique. "There are too few good quality surgeons," Dr Toy said. "If a surgeon is not able to remove the tumour completely there is an increased risk of metastases [cancer spread]. The quality of the surgery will also have an impact on side effects such as impotence and incontinence."

Some specialists fear cancer screening increases anxiety among patients without increasing survival. "Someone with prostate cancer who is going to die in 10 years will have a better quality of life if it is not diagnosed for nine years," Dr Toy said. "Sometimes ignorance is bliss."