If you're filling out a form and you're asked to enter your weight or height, it can be tempting to fudge the numbers. New research suggests that Australians often understate their weight and overestimate their height -- but we might be getting slightly better at it.
Picture from The Biggest Loser
A new research paper from the Australian Bureau of Statistics looks at Australian data relating to body mass index (BMI), a widely-used measurement for tracking weight trends in large populations. BMI is calculated by taking an individual's mass in kilograms and dividing it by the square of their height in metres.
While widely adopted, the use of BMI is not universally accepted, with some researchers arguing that it fails to properly take into account variations such as athletes with high muscle mass. However, the Australian research focuses on a different question: just how accurately do Australians know their own height and weight?
ABS researchers Tim Ayre, Jason Wong and Anil Kumar looked at two studies, the 1995 National Nutrition Survey (NNS) 1995 and the 2007-2008 National Health Survey (NHS). Both studies included a mixture of self-reported height and weight figures and accurately measured numbers, which meant that the researchers could determine how people's estimates varied, as well as applying regressions to see if self-reported data could be adjusted to more accurately reflect trends. In simple terms, this is what they found:
The results confirm that, on average, respondents overestimate their height and underestimate their weight and BMI. However, the magnitude of the misreporting was significantly smaller in the NHS 2007–08 than in NNS 1995. Adjusting self-reported BMI is found to provide significantly more accurate estimates of the distribution of BMI than using self-reported BMI directly.
Both trends aren't entirely surprising; we're a society that generally argues that obesity is bad but that height is somehow useful (the latter assumption gets messy when you board a plane). What's interesting is that the rate of error is smaller in the more recent data, though we'd need a lot more studies to confirm whether we really have become more aware or more honest about our weight.
Other variables also played a role:
The analysis shows evidence that females tend to have greater reporting errors than males. The size of the reporting errors were also found to depend on the reported values of height, weight and BMI, as well as various demographic and socio-economic variables, and other health-related risk factors.
The big lesson? If you ask someone how tall they are, it's probably still safer to knock off a few centimetres from the figure. (I'll stick to believing it's rude to ask someone their weight unless you're a medical professional or they're about to board a light aircraft.)