Dr Maree Farrow
Often we see headlines in the media that might tell us a certain behaviour will increase our risk of dementia while another behaviour might reduce our risk of dementia. Those headlines, unfortunately, can be misleading. They can leave us believing that we could eliminate our risk of dementia completely simply by eating more chocolate or drinking more coffee. But what do those headlines really mean?
Let’s talk about how researchers actually find out how a particular behaviour affects our risk of developing dementia. Let’s say we wanted to find out how eating cheese is related to the risk of dementia. Now it’s not, as far as we know. There’s been no studies looking at this, so please don’t worry about eating cheese. This is a hypothetical example that we’re going to use here.
One of the best ways to look at this question would be to conduct what we call a prospective cohort study. In this type of study, we’d recruit a random sample of people and we’d ask them how much cheese they eat. Then we’d follow them up over a number of years to determine who gets dementia, and whether how much cheese they eat was related to their risk of developing dementia. Then we would use statistics to determine whether eating cheese had a significant effect on your risk of developing dementia and also how big that effect might be.
In our hypothetical research study, we’re going to start by recruiting two thousand people who are aged over 65. We’re going to ask them how much cheese they eat and we’re going to split them into two groups. We’re going to take the people who eat more than 100 grams of cheese per week and then the people who eat less than that amount per week. So in this hypothetical study, let’s say we were able to split the group in half. So we have one thousand people who eat more than 100 grams of cheese per week. We’re going to call them the high cheese group. Then we have another one thousand people who eat less than 100 grams of cheese per week, and we’re going to call them the low cheese group.
To do this study well, we’d need to match our two groups on things like age and gender, the other things that they eat in their diet, and any other factors that might affect dementia risk, so that how much cheese they ate was the only thing that was different about the two groups. We would also test their cognitive function at the beginning of the study, to make sure that they didn’t have dementia at the start. After this baseline testing, we’ll follow our two thousand participants for the next ten years, and we’ll meet with them every two years to test their cognitive function, and in that way we can determine if anyone develops dementia over that time.
So in our hypothetical study, let’s say we found that over the ten years 120 people developed dementia. Now 80 of those were from the high cheese group, so that represents 8% of the people in that group who did develop dementia over those ten years.
In the low cheese group, however, there were only 40 people who developed dementia, so that represents 4% of the low cheese group.
And we might start to think about whether eating cheese is such a good idea. But what do these numbers really mean? There’s two different ways we can look at the findings - depending on whether we’re interested in the effects of eating a lot of cheese or the effects of eating not much cheese.
另外我们可以开始考虑吃奶酪是否是一个好主意。但这些数字真正的含义是什么呢？我们可以有两种不同的方式来看结果 - 取决于我们是否对吃大量的奶酪或吃不太多的奶酪的影响感兴趣。
Let’s first look at how eating high amounts of cheese affects dementia risk in our hypothetical study. From our finding that 8% of people in the high cheese group developed dementia compared to 4% in the low cheese group, we can conclude that eating high amounts of cheese might double the risk of developing dementia. We could also say that eating high amounts of cheese increases the risk of dementia by 100%. Then, as researchers, we would calculate what we call a risk ratio, and in this case the risk ratio is simply eight divided by four, which is two.
A risk ratio of two means exactly the same thing as increasing the risk of dementia by 100% or doubling the risk through eating high amounts of cheese.
We would then use statistics to determine if our risk ratio is significant. A statistically significant risk ratio means that we can be reasonably certain there’s a real relationship between eating cheese and the risk of developing dementia, rather than it just being due to chance.
To investigate the effects of eating small amounts of cheese, we would look at our results the other way around. From our results that 4% of people in the low cheese group developed dementia, compared to 8% in the high cheese group, we could conclude that eating small amounts of cheese halves the risk of developing dementia, or that the risk is reduced by 50%. And the risk ratio in this case would be four divided by eight or 0.5. So again having a risk ratio of 0.5 is exactly the same as halving the risk or reducing the risk by 50%. And once again we would need to use statistics to determine if that relationship was statistically significant.
Doubling your risk of dementia sounds pretty scary, while halving your risk of dementia sounds like a good idea. So would you give up eating cheese based on these findings? There are some very important points to remember when looking at research findings like this.
Firstly, although the relative risk of developing dementia was doubled by eating high amounts of cheese in our hypothetical study, the absolute risk changed by only a small amount from 4% to 8%. The other 92% of people in the high cheese group did not develop dementia, despite the fact that they consumed those higher amounts of cheese.
Secondly, a finding like this does not mean that eating cheese causes dementia. What we found was an association between the two things, but this does not represent a causal link.
And finally, a study like this can’t tell us everything about the relationship between eating cheese and developing dementia. What about younger people? If they ate lots of cheese would that affect their risk of developing dementia later in life? What about the type of cheese you eat? Is parmesan cheese better or worse than gorgonzola? And what about the amount of cheese you eat? If 100 grams of cheese per week increases the risk, would 200 grams of cheese per week increase the risk even further?
What this finding means is that older people who eat more of any kind of cheese might have a slightly higher chance of developing dementia. Don’t forget though, this is a hypothetical question. So if I did eat lots of cheese, based on these findings, I might think about cutting back to reduce my risk of dementia, but I would do so knowing that it wouldn’t stop me getting dementia, and I’d also know that it was only one of many choices I could make to potentially reduce my risk of dementia.
Now what would the media make of these findings? We might see headlines such as “Don’t eat cheese if you don’t want dementia”. But we need to look at the research and the real findings behind these headlines to understand how relevant they are to us.
Here’s an example of some real newspaper headlines from a few years ago. Firstly, “Coffee – just the shot to fend off Alzheimer’s.” And “Caffeine addicts rejoice – a cup a day may keep Alzheimer’s at bay.” You might expect the research was conducted in a similar way to our hypothetical cheese study. A human prospective cohort study investigating the link between coffee consumption and the risk of developing Alzheimer’s Disease.
“咖啡 - 只是防止阿尔茨海默病的镜头”和“咖啡因成瘾者喜出望外 - 一天一杯可能会远离阿尔茨海默病”，这是几年前某些真正报纸头条的一个例子。首先，你可能希望以类似前面我们假设的奶酪研究的方式进行一项人类前瞻性队列研究来调查喝咖啡和发展阿尔茨海默病的风险之间的联系。
However, here’s the title of the article published by the researchers in the Journal of Neuroinflammation in 2008. “Caffeine blocks disruption of the blood brain barrier in a rabbit model of Alzheimer’s Disease”. The participants in this particular study were rabbits fed a high cholesterol diet.
The researchers tested whether chronic ingestion of caffeine in their rabbits could protect them against changes in the blood brain barrier brought about by their high cholesterol diet. High cholesterol diets and high cholesterol levels, as well as disruptions in the blood brain barrier, have been associated with Alzheimer’s Disease, and also there have been studies suggesting that coffee consumption might be protective against Alzheimer’s Disease. So this is a reasonable research question.
The researchers found that in their rabbits, caffeine consumption blocked the usual changes in the blood brain barrier brought about by their cholesterol-rich diet. And the researchers concluded that caffeine, and drugs similar to caffeine, might be effective in the treatment of Alzheimer’s Disease.
Now it’s rather a long stretch to go from that research to suggesting that humans could fend off Alzheimer’s Disease by drinking more coffee.
The participants in the study were rabbits, not humans. They didn’t have Alzheimer’s Disease. They were fed a cholesterol-rich diet as a potential model for Alzheimer’s Disease. They didn’t drink coffee, they were given pure caffeine in their drinking water. Also, the researchers suggested that caffeine could potentially be used as a treatment for Alzheimer’s Disease, not as something that would prevent Alzheimer’s Disease.
So be wary about making decisions based on media headlines. They may not accurately reflect the research that they’re based on. And also remember the difference between relative risk and absolute risk. A doubling in the relative risk of developing dementia might be something we should take notice of. But it may mean just a small increase in the absolute risk of developing dementia.