Best headline on a press release from last week? It has to be this from the University of Bristol: "Cannabis causes cognitive chaos in users".

In my day we called it getting stoned – but who am I to stand between a scientist and his research grant? Still, it is always comforting to have one's beliefs reinforced. Two other studies seemed to do the same. One, by a team at Oxford University, found taking the contraceptive pill for 10 years almost halved the risk of ovarian cancer. Good news, no? But as another expert from Newcastle University pointed out, the Pill is known to slightly increase the risk of breast cancer, which is much more common. Thus, among 100,00 women taking the Pill for 10 years there will be 12 fewer ovarian cancers – but 50 extra breast cancers. Result: cognitive chaos for patients.

The second study confirmed another long-held belief – that aspirin protects against cancer. The evidence has been growing for decades, but this by Oxford University was the first randomised controlled trial, medicine's gold standard, showing that for people at high risk of bowel cancer a couple of the little white pills daily cut their risk by over 60 per cent.

Impressive. But the trial was done in people with a genetic pre-disposition to the disease, who number one in 1,000 of the population. For the rest of us, the calculation is more difficult. Aspirin protects against heart attacks and stroke, as well as various cancers, hence its "wonder-drug" status. Professor Sir John Burn, author of the study in The Lancet, said he takes it himself.

But I remain impressed by a conversation with an expert who pointed out that internal bleeding caused by low-dose aspirin – one of the two commonest causes of adverse drug reactions – accounted, in all, for one in 15 hospital admissions. Who was that? Professor Sir Alasdair Breckenridge, chairman of the Medicines and Healthcare Products Regulatory Agency which oversees the safety of medicines in the UK.

Those who hope for cognitive clarity on drugs are whistling for the moon.