Consider the strength of evidence

When I was at medical school in the Nineties, there were certain subjects that were almost universally disliked.

Chief among these was the module on research, which was guaranteed to induce somnolence or mind wandering.

I must admit, I did not pay it adequate attention at the time, but I now recognise its importance. Many of my patients come to me with questions about a product or intervention about which they had heard great things – from boosting brain performance through to reversing or curing dementia.

For anyone to be able to provide meaningful advice, it is critical to know about the studies that underpin the claims. It is helpful for everyone to know a little about evidence to make informed decisions about their health.

The best level of evidence is level (or class) 1 and arises from studies called randomised controlled trials (RCTs) or an analysis of multiple RCTs (a meta-analysis). These studies compare a group of people randomly given an intervention (for example, a medication or treatment) with a group that is not.

The groups should otherwise be as similar as possible. In this way, if the outcome is different for the intervention group, you can conclude that the intervention was responsible – that it caused the change.

When it comes to brain health, there are only a few things that have this level of evidence: blood pressure control, treating poor hearing and physical exercise. The pivotal Finnish Geriatric Intervention Study to Prevent Cognitive Impairment and Disability (FINGER) trial also counts as level 1 evidence and involved multiple interventions concurrently: diet, exercise, cognitive training and control of vascular (blood supply) risk factors.

Level or class 2 evidence is still reasonable but is derived from lesser-quality RCTs and strong so-called cohort studies. In the latter, the intervention may not be randomly allocated to the groups or there may be no specific intervention and people are just recorded as having certain attributes (for example, following a healthy diet) and then followed up to see what happens to them.

If there is a strong correlation (such as those eating a Mediterranean diet are less likely to develop dementia), that might lead you to conclude that one thing causes the other. The issue is that other variables may have caused the change (for example, those eating a Mediterranean diet, pictured, might also exercise more) and so it can be hard to confirm that the one thing being recorded causes the outcome.

Despite these limitations, class/level 2 evidence is still seen as pretty good. For promoting brain health, the following have this level of evidence: Mediterranean diet/MIND diet, cognitive stimulation, smoking cessation, treating sleep apnoea, treating depression, early life education, control of diabetes and lifelong learning.

Level 3 and 4 evidence is considerably less robust and often relies on small or retrospective studies where the researchers look back and see what someone with or without a disease may have been doing or experiencing: for example, finding out that those who drink a little red wine have better memory.

These are signals that the behaviour may have influenced the outcome, but it is hard to conclude.

There are many things that fall into this category for brain health, including moderate alcohol use (hotly debated), omega 3 supplementation, B and D vitamins, curcumin, resveratrol, sauna use, intermittent fasting and more.

Ultimately, we all must make decisions based on incomplete data, but it is good to know the strength of the evidence.

If it is a low-risk intervention, you might decide to do it, even if it has a low level of evidence.

But if not, you really want to be assured that the evidence is robust.

 Kailas Roberts is a psychogeriatrician and author of Mind Your Brain: The Essential Australian Guide to Dementia, available at bookstores and online. Visit yourbraininmind.com or uqp.com.au