The first thing I always do with study headlines like this is find a link to the study itself. For all we know, the article we first read may simply be a regurgitation of some pre-written press release that was sent out to media outlets for the express purpose of achieving maximum exposure, rather than actual journalism where somebody actively seeks out the story. Best to go beyond the headline and go straight to the source, where possible.

In this case, the study itself is published here.

I had a read of it a couple of weeks ago. To my relatively unqualified eyes, the raw data seems fine in terms of how much of which nutrients are being consumed in different parts of the world, even allowing for the inherent weaknesses in 24 hour diet recall as a method of data collection ("please tick everything that you ate yesterday from this list").

Where I would start to challenge the findings is when that raw data is compared to the recommended levels of those nutrients. Those recommended levels were set by calculating the amounts associated with the lowest risk of death, the key word being "associated". For example, eating red meat may be associated with a small increase of risk in developing cancer, but since smokers also tend to eat more meat than non-smokers, any direct association with meat and cancer can't really be proven without first addressing the confounder. So, to claim that we eat too much red meat on a basis as weak as this, is itself pretty hollow.

Any recommended levels of anything should really only be defined through data from higher-quality experimental studies, i.e. trials. Observational data proves very little on its own, and in this case, we're talking about observational data based on other observational data, which is even worse. So to me, there really is nothing meaningful to see here... but we see it anyway, because people know that catchy headlines sell and details don't really concern the majority of audiences these days.

If you haven't read The Big Fat Surprise by Nina Teicholz, I'd recommend it. It's interesting - in a depressing kind of way - to understand the lengths to which some people will go to get their message out, regardless of its accuracy to the truth.