One should always read the science press with a skeptical eye: scientists are just like the rest of us in that they sometimes get things wrong and often don't know everything of interest in a field. That latter point is especially true as most fields of research in the life sciences have expanded far beyond the awareness capacity of a single busy professional. So it's very helpful to have some background knowledge yourself, to help sort out the relevant from the potentially flawed.
Take this recent work on calorie restriction for example, which is announced as a potential proof that calorie restriction is not going to be useful in healthy humans:
"Our study questions the paradigm that caloric restriction is universally beneficial," Sohal said. "Contrary to what is widely believed, caloric restriction does not extend (the) life span of all strains of mice."
By measuring the animals' metabolic rate, Sohal and his colleagues came to a deceptively simple conclusion: Caloric restriction is only useful when, as in the case of the obese mice, an animal eats more than it can burn off.
"Your energy expenditure and your energy intake should be in balance," Sohal said. "It's as simple as that. And how do you know that? By gain or loss of weight.
"The whole thing is very commonsensical."
For humans of normal weight, Sohal strongly cautions against caloric restriction. In a 2003 study, he and Forster found that caloric restriction begun in older mice - both in DBA and leaner C57 individuals - actually shortened life span.
Given the very large number of studies showing strong positive effects on health and longevity in individuals of normal weight through calorie restriction, one should immediately be cautious in looking at this sort of claim in a single paper. Over at the Immortality Institute, calorie restriction expert Michael Rae dives in to take it apart in more detail:
These guys have really made some extremely unreasonable leaps of reasoning.
First, yes, it's true that DBA/2 mice are resistant to the effects of CR on lifespan -- that's been shown repeatedly, since at least the early 80s, and not just by Sohal; and (per this study) yes, they apparently have some aesthetically happy genetic perk, that allows them to eat ad libitum without gaining any weight, while other mice gain weight as one would expect.
But so what? First, as Anthony noted, while they have demonstrated these two facts, they haven't done anything to prove that this resistance to weight gain in response to an AL diet is what *causes* their resistance to the age-retarding effects of CR in this strain.
Second, as Matt notes, these are sickly mice
...
in [their prior referenced 2003] study, they violated almost every rule of adult-onset CR in the book: they administered CR as shock therapy instead of stepwise to allow for gradual adaptation, and they didn't boost the %protein in the diet to ensure that they got the full rodent 'RDA' of the AL mice. Weindruch and Walford's big 1982 breakthru', showing that you COULD actually make CR work in adults, was based precisely on not making those mistakes - and it's been repeated numerous times since.
...
the current study doesn't come anywhere close to proving what Sohal is evidently asserting about it.
For those of us without this level of background knowledge, the weight of science test is the best way to go: only start to pay serious attention to a claim when it has a cluster of papers backing it. One paper standing in opposition to dozens or hundreds of others is either wrong or the first sign of a change of viewpoint - for an observer outside the scientific community there's no harm done in dismissing it in either case. If change is underway, there will be more papers and studies to support the new viewpoint in due course.