Dieters know the expression, “Watch what you eat.”

But apparently, on a global scale, it applies to everyone.

Turns out what we eat actually can be measured in emissions as well as pounds.

An engineer at the University of Michigan’s Center for Sustainable Systems is one of several researchers who published a paper in “Environmental Research Letters” that looks at how much variability there is in the greenhouse gas emissions of American diets.

In fact, 46% of the total emissions from food came from the diets of just one-fifth of the population. According to the study, mostly that’s because those people eat a lot more meat than the others, especially methane-spewing beef. Animal protein jacks up emissions for the top consumers, while the ecologically minded folks at the bottom consume more plant proteins.

Suddenly, the “Eat More Kale” push here in Vermont has some new legs.

“I don’t think any of us really had a strong sense of how distributed the greenhouse gas emissions would be,” Martin Heller, the researcher, said in media reports. “That was perhaps the most striking result.”

Heller and one of his co-authors had done previous work on how U.S. diets contribute to climate change, but no one has ever really looked into the variability between different diets before.

In part, that may be because the work is just so complicated. To come up with estimates, the group had to link several large databases of information, including the Food Commodities Intake Database, which is run by the Environmental Protection Agency.

The researchers came up with what is called “a life cycle analysis,” which involves understanding the agricultural processes required at every step of a vegetable’s life before it leaves a farm. That includes not only how much fuel the tractor uses in the field, but how much energy goes into making the fertilizer and pesticides.

Unsurprisingly, fertilizer is often one of the bigger contributors, the report found. It’s an energy-intensive process that requires making ammonia and to make ammonia, we often use natural gas, so the life cycle analysis for a tomato, for example, has to include the natural gas drilling required to make the fertilizer that went into the soil to grow the fruit. And then it has to include the emissions created by that fertilized soil. Ammonia increases the amount of nitrogen in soil, and that excess nitrogen has the unfortunate side effect of promoting nitrous oxide production. (That nitrous oxide has 300 times the global warming potential of carbon dioxide, and it’s unavoidably created whenever you fertilize soil.)

When you look at meat, it’s an entirely different equation when you take in feed production and, simply put, gassy animals.

Heller and his team did not even add into the equation the factor of emissions and transporting the food products from farm to supermarket or food-processing facility.

“The whole local food movement has really emphasized the impact of food miles,” Heller was quoted as saying in Popular Science. “But most of the research points out that that’s not really a huge part of the total. What goes on at the farm is a much bigger piece.”

Here in Vermont, we like to think we are making a dent in climate change by supporting farmers’ markets and farm-to-table operations.

But the researchers suggest those choices don’t necessarily mean fewer emissions.

“Often times those are the things that people’s imaginations jump to, because those are the kinds of choices they’re presented with in the store. Those are small differences, though, and they’re relatively small compared to the large differences we see between food types,” he told interviewers.

If added up by cost per unit (meat, vegetable or fruit), as consumers we would probably be stunned by the economics of “actual costs.” More intimidating would be realizing just how much we might be contributing to climate change with our daily diets.

And, of course, there is no way — nor should there be — to make policy to try to change what we eat. That’s why we have choices.

But you do wonder how we can get 20% of Americans to change what they eat.

Here in Vermont, whether it’s through local distributors, farmers’ markets, health care networks or community action and other agencies that regularly work with the poor, we are pretty good about encouraging healthy eating.

Clearly we can be doing more — starting with our pantries. Consider it food for thought.