Ugh. I find it upsetting how poorly most reporters understand the science they cover (Saletan normally being a counter-example). Neuroscience in particular has this problem.

It should come as no surprise to anyone that eating activates areas of the brain associated with our reward system. Food sex and safety are kinda the reason that a reward system evolved in the first place (i.e. maintained itself after first arising). Certain chemicals happen to override parts of the reward system and can cause some people to act in a way detrimental to their survival. Correlations of fMRI activity are thus pretty much meaningless.

Yep, it does seem a bit obvious. The problem is, there are people who seriously advocate the regulation of food as if it were a drug, despite the fact that we haven't really been all that fantastic at regulating drugs up to this point.

As far as I'm concerned, unless they find convincing evidence that the Colonel with his wee beady eyes has been spiking his chicken with a chemical that causes you to crave it fortnightly, then they need to slow the overregulation train down a bit.

blueshift:

Ugh. I find it upsetting how poorly most reporters understand the science they cover (Saletan normally being a counter-example). Neuroscience in particular has this problem.

It should come as no surprise to anyone that eating activates areas of the brain associated with our reward system. Food sex and safety are kinda the reason that a reward system evolved in the first place (i.e. maintained itself after first arising). Certain chemicals happen to override parts of the reward system and can cause some people to act in a way detrimental to their survival. Correlations of fMRI activity are thus pretty much meaningless.

Correlations of fMRI activity are only as good as the task's design. These studies were designed poorly as numerous interpretations can be made from the data.

The regions identified in this study are part of the reward system in the brain. However, this activity pattern has also been implicated in task control and attention (see dosenbach et al, neuron, 2006). Therefore, an alternative explanation might be that this pattern of activity reflects the fact that the subjects are performing a task.
Unfortunately, addiction is not studied in the context of attentional mechanisms, so nearly all addiction studies are heavily confounded.

Thank you Shusaku. Rereading my last sentence I see that I was unclear. I meant that finding a correlation in this instance was meaningless. fMRI is obviously a great tool when used in a well designed and controlled study. The problem is that fMRI data looks impressive no matter what, and a mediocre paper with some marinably valuable results can get picked up by people that don't understand the confounds. The message to the public easily becomes a conflation of implicated systems.