A couple of weeks ago, the American Diabetes Association had their annual conference. A bunch of news reports found their way into my inbox and left me chomping at the bit with stuff I wanted to rant about. This one is about two large epidemiological studies in Canada, both showing that highly walkable neighbourhoods are associated with fewer new cases of diabetes and less ‘obesity’.
My first thought was ‘Well, duh!’ So many possibly confounding factors are associated with low-walkable neighbourhoods – socioeconomic status being the obvious one. Low SES is one of the strongest known risk factors for diabetes (and obesity). Poverty, crime, possibly fewer educational opportunities, high unemployment and so on, are not conducive to a low-stress lifestyle, and chronic stress is recognised as a major determinant of metabolic health.
One of the studies’ authors is quoted as saying:
“This suggests that changes in zoning, urban planning and design that promote walking and other forms of active transportation may help to curb the ongoing rise in obesity and diabetes. … Further research is needed to understand the full impact such interventions will have.”
It does? OK, I thought, before jumping to conclusions based on a news report, I should go back to the source – maybe they controlled for all that stuff. Maybe their findings and discussion are more nuanced and insightful than represented in this soundbite. Unfortunately, the ADA does not make the abstracts of oral presentations at their meetings freely available. Big surprise. But a quick search found this paper published earlier this year.
The researchers split Toronto into 5 levels of walkability. They defined walkability as a combination of population density, housing density, and having stuff like shops and schools within walking distance. And yes, they found that people who lived in more walkable neighbourhoods had lower average BMI, and lower risk of diabetes. In fact, people in the least walkable neighbourhoods had 1/3 higher risk of each of these than people in the most walkable. But here’s the weird thing. The most walkable neighbourhoods weren’t the wealthy suburbs, but the ‘poorer’ downtown, high-density neighbourhood.
Now traditionally, people with lower SES live in denser areas and higher SES is more associated with sprawling suburbs. The demographics seem to support this: 50% more people in the most walkable neighbourhoods lived below the low-income cut-off (23% vs 15%), and earned an average of $17,000 per year less (76K vs 59K) than the people in the most walkable neighbourhoods. Not only did they walk more, but they cycled more too (and took public transport more often), meaning they were generally more active in going about their daily lives. This doesn’t speak to leisure pursuits though.
This makes the study results a little surprising. It means that the richer people had more obesity and more diabetes, which flies in the face of everything we know. A 2008 report from the York University School of Health Policy and Management in Toronto found that the strongest correlation between diabetes and risk factors was with poverty and marginalisation.
Even within the lowest income group, moving from under $15,000 to the next level up, £15,000 to $29,999, the risk of diabetes in nearly doubles. Accounting for education, BMI and physical activity didn’t make much of a dent in that figure.
So what’s going on with this new study then? It’s hard to know exactly, but here are some thoughts. The most walkable neighbourhoods did have the lowest average income, but this was still C$59,232, so nothing to be sneezed at. The authors give no indication of the spread of incomes, and do not adjust their diabetes figures for socio-economic status. I’ll pause here for people who understand epidemiology to gasp in shock. So at best, I’d say, that given a moderately comfortable standard of living, more physical activity is associated with lower rates of diabetes. This is not news. Physical activity is generally good at improving most markers of health – it’s just rubbish for weight loss.
So yes, improving the walkability of neighbourhoods with higher average incomes may well help increase physical activity in that sociodemographic, but should it really be our first priority?