Statins are typically initiated at a time in the lifespan when we become aware of our own memory decline. This makes non-blinded and observational studies hopelessly confounded, not to mention anecdotal reports. The initial statin mega trials did not prospectively gather baseline data on cognitive status and didn’t rigorously assess for short term changes in cognition, such as memory loss.
So is there any robust evidence? A meta-analysis published in Mayo Clinic Proceedings last November is the best current synopsis. The authors went to great lengths to ensure that their findings approach as close to the truth as is possible: they only included randomised controlled trials that used validated objective measures of cognition as end points. Studies were only included after rigorous assessment for bias, thus reducing the number of studies down from 41 to just 3! With only 300 participants, these well-conducted trials were all consistent, finding no significant hazard from statin therapy (see figure)
It is disappointing that our advice to patients must be based upon only 300 patients. Nevertheless, the rigorous study designs and the trend towards benefit make it unlikely that a significant effect is being missed for the population of patients as a whole. Of course, the rare individual patient may still have an idiosyncratic reaction to medication, so every report of memory loss should be considered thoughtfully.
The same study also assessed the effects of statins upon dementia. It confirmed what is generally accepted, that statins reduce dementia. Five studies found a favourable effect, three found no effect upon the subsequent development of dementia. Taken together, the relative risk reduction was 29%, absolute risk reduction was 2% and the number needed to treat for 6.2 years was 50.
So the current evidence is most compatible with statins having no significant short-term effects on memory and a beneficial role in preventing dementia over the long term.
Swiger KJ, Manalac RJ, Blumenthal RS, Blaha MJ, Martin SS
Mayo Clinic Proceedings November 2013;88(11):1213-1221