‘Dementia rates “higher near busy roads”’ was a headline from the BBC News website last week which I was invited to comment on.  The article concerned an 11-year Canadian study which had monitored around 2.2 million people to investigate the risks associated with living close to a busy road and incidence of dementia. This study found that for people living within 50 metres of a busy road, the risk of dementia increased by 7%.
A shocking statistic when you consider the fact that dementia affects almost 50 million people worldwide and there are over 7.5 million new cases every year. So, should we all think about picking up sticks and moving to the countryside in search of a dementia free life? Well, here is an example of where risk isn’t all that it seems.
The BBC isn’t wrong that living close to a busy road may cause an increase in the risk of dementia, but by presenting only the relative risk, they are misleadingly overemphasising its effect. What was worrying regarding the BBC reporting of this study was that they continued to run the story under the headline ‘Dementia rates “higher near busy roads”’ throughout the afternoon and evening on their news programmes despite my explanation of the statistics underpinning the study in my interview with the Radio Four programme, ‘The World at One’, in the early afternoon.
The BBC Trust Review of Statistics published in August of last year found that BBC journalists often do not go ‘beyond the headlines’ and there was a tendency to report statistics straight from a press release without looking further into it. The reporting of this story was an obvious example of a headline prevailing at the cost of the nuances of the issue.
What do I mean by a relative risk? A relative risk is one where it is just presented as an increase or a decrease compared with something else: you have a 7% increase in the risk of dementia living within 50 metres of a busy road compared with more than 300 metres away. It is important to note that relative risks don’t tell you anything about what the underlying risks actually are. Absolute risks, on the other hand, tell us what the actual chances are of something happening. The higher the risk, the more likely it is.
There were around 250,000 diagnoses of dementia in the Canadian study, so around 11.4% of those studied.* If the risk of dementia is 11.4%, that means that for every 500 people in the population not living close to a busy road, we would expect 57 of them to get dementia anyway. What is a 7% increase of 57? This is only an extra 4 in every 500 people, so for 500 people living within 50 metres of a busy road, we would expect 61 of them to get dementia. Not so shocking, especially when you factor in the potential benefits of living near a busy road. It should also be noted that there are a number of other risk factors that have a bigger effect on the chances of being diagnosed with dementia.
The supplementary material from the study in fact states that smoking increases the risk of dementia by 30%, obesity 64% and physical activity reduces it by 14%. So what was a headline story on BBC News, presenting seemingly shocking statistics on the risk of being diagnosed with dementia, turns out to be something we don’t really need to worry about.
Let’s throw in the fact that, as any statistician knows, correlation doesn’t mean causation and also consider that what Canadians define as a busy road may not be same as what constitutes ‘busy’ worldwide. Additionally, the study didn’t account for family history of dementia, which is a well-known risk factor for the disease.
This isn’t to say that this study wasn’t a well-executed study with a large sample size which has provided a useful insight into the risks associated with being diagnosed with dementia, but this particular article from the BBC is an example of how only presenting relative risks can generate public hysteria and provide misleading information on which the general public is encouraged to make important decisions about their everyday lives.
*note the average lifetime risk for a 65 year old is 13%, but this study considered 55-85 year olds, so 11.4% may be reasonable(5)
 Chen H, Kwong JC, Copes R, Tu K, Villeneuve PJ, Donkelaar A, Hystad P, Martin RV, Murray BJ, Jessiman B, Wilton AS, Kopp A, Burnett RT. Living near major roads and the incidence of dementia, Parkinson’s disease, and multiple sclerosis: a population-based cohort study. The Lancet, 2017. In Press.