Skip to main content

Archived Comments for: The effect of body mass index on global brain volume in middle-aged adults: a cross sectional study

Back to article

  1. Inadequate adjustment for body size

    Tim Cole, UCL Institute of Child Health, London, UK

    9 January 2006

    The study demonstrates a significant association between brain volume and BMI which the authors tentatively conclude to be causal. Their conclusion is dubious for two reasons:

    1. The association is cross-sectional, and association does not imply causation. The sample is non-random and may be unrepresentative of the population, and the modestly significant result may be a type 1 error.

    2. The regression analysis did not adjust for body size, only age. Head circumference (and hence brain size) is known to be strongly correlated with height in children and young adults, yet this association is not normally interpreted as meaning that taller people are more intelligent.

    It is quite possible that the fatter subjects here were also shorter, which would imply smaller heads and smaller brains but no associated reduction in intelligence. So the association may be meaningless in functional terms.

    The authors should be more cautious in their conclusions.

    Competing interests

    None

Advertisement