Not in a human environment would be a challenge, but IMO somewhat double. But not in a human environment, where we _do_ have historical weight data? Forget it.

As to the control mice: my hunch would be that 'let the control animals eat at will' has changed over time. Maybe the diameter of sugar water tubes has increased slightly?

Certainly, our notion of what constitutes overweight has changed. If you look at photos of kids in the 1950s, almost all of them look extremely skinny. We may project that on lab animals, and size the 'eat at will' containers accordingly ("hm, the control mice get too fat. Let's make it a tiny bit harder for them too get their food")

Also, chances are that lab animals live at better controlled temperatures. Possibly, they used to eat less in hot summers because the temperature in their cages was higher in summer?

Of course, that's speculation, but it is the simplest explanation I can think of.

I have data here on the weight of cats trapped on Kerguelen island since 1999, and I'm no statistician but I see no trend at all (data should also exist on the feral rabbits and mice, but I don't think I have it).

Isle Royale would be a good candidate for this kind of study. It's a small island in the middle of Lake Superior that has its own isolated wolf population. They take a lot of data from the wolf and moose populations, and it wouldn't be surprising if they had some way to track/estimate weight.

I know the research facilities there were the target of severe budget cuts several years ago. It's a shame, because this is an example of where other, unrelated research would benefit from the basic research of an environment mostly untouched by humans.

http://en.wikipedia.org/wiki/Isle_Royale