The global sweet tooth seems pretty much insatiable. Sugar consumption—whether of cane sugar or high fructose corn syrup—has been steadily on the rise for decades. In the low fat diet craze of the 1990s, sugar was seen as a relatively benign replacement for some of the fats in our foods. More recently, however, worries have shifted, with sugar taking the place of saturated fat as the ingredient that tops our suspicions for being the greatest contributor to poor health and obesity.
The problem for scientists is that it’s tremendously hard to figure out even some of the most basic information about what we eat.
“Foods perceived as undesirable tend to be underreported,” said Brenda Davy, a nutritionist at Virginia Tech. “And there’s a lot of literature that links added sugars and sugar-sweetened beverages with adverse health outcomes.”
It’s why Davy paired up with geochemist Hope Jahren from the University of Hawaii. Jahren’s work uses subtle variations in atoms like carbon and nitrogen to determine the diets of ancient animals. Several years ago, Jahren began to think that she could apply these same techniques to tease apart our modern diets. The result of the collaboration between Davy and Jahren is a blood test that can accurately measure how much sugar is in a person’s diet.
“It’s a low cost way to determine sugar intake,” says Josh Bostric, a researcher in Jahren’s lab. “It can also tell us how much sugar is too much, which can ultimately shape policy.”
Not only do the researchers hope to use this test in large-scale public health nutrition studies, they also hope that one day, doctors can use this test as a quick and easy measure of how much sugar their patients are eating, and whether people might benefit from cutting back.
Scientists estimate that Americans now consume around 16 percent of their daily calories from added sugars, such as sodas, juice, cookies, and other sweets. But this number is, at best, an estimate, Davy explains, because asking people to report what they eat is fraught with errors.
For one, we forget what we eat, even in the recent past. Another difficulty is estimating portion size, which studies have shown we are notoriously bad at. Lastly, people might not want to admit to other people just how much soda they drink, and so they conceal or reduce the amount of these foods they eat when asked, whether consciously or not. What researchers need is an objective measure of how much sugar we actually eat that doesn’t depend on our fickle memories or desire to look good in front of others. The answer, it turns out, can be found in the very plants that satisfy our sweet tooth.
Like nearly all of the tiny atoms that make up our world, carbon, nitrogen, and oxygen have a unique number of proteins in their nucleus. Some of these atoms, however, have different numbers of neutrons. While this doesn’t change how these elements behave chemically, it does give them slightly different weights. Carbon, for example, has six protons and, in more than 99 percent of all its atoms, also has six neutrons, which is called carbon-12. A few carbon atoms, though, have six proteins and seven neutrons, or carbon-13. Unlike carbon-14, which has six protons and eight neutrons, carbon-12 and -13 are stable and not radioactive.
When plants use the carbon dioxide from the atmosphere to make sugar during photosynthesis, they use molecules with either carbon isotope. However, due to the small differences in the weight of the different carbon isotopes, different plants create sugar molecules with different concentrations of carbon-13 based on which of two methods of photosynthesis they use. When we eat these plants—or eat the animals that eat these plants—these different ratios of carbon isotopes are reflected in our tissues. Sophisticated equipment can separate the different isotopes based on differences in their weight, which lets researchers determine how much of each carbon isotope is in a tissue sample.
“You really are what you eat,” Bostic said.
In the American diet, the two overwhelming sources of added sugar are sugarcane and corn syrup, both of which have a higher proportion of carbon-13 than other sugary or carbohydrate-rich foods, such as fruits and wheat. Other studies in animals told Jahren and Davy that humans who ate diets high in added sugar would have more carbon-13 in their blood, hair, urine, and other tissues. They first tested the idea on a small group of undergraduates, and found that they could separate the students who reported consuming the most added sugars from those who ate the least based on the amount of carbon-13 in their blood.
Other scientists soon began conducting similar studies. Sarah Nash, currently an epidemiology fellow at the National Cancer Institute, used stable isotopes to measure added sugars in a group of Yup’ik people in Alaska. Because their diet is so heavy in seafood, Nash and colleagues had to measure levels of both carbon and nitrogen isotopes to measure added sugars.
“The carbon isotope ratio is distinctly high in corn- and sugar cane-based sweeteners, which is why the carbon isotope ratio has proven a useful indicator of sugar intake in several study populations, but this ratio is also elevated in animal proteins,” Nash said.
In the US, most livestock has a corn-based diet, which elevates the carbon isotope ratio just like the consumption of sugar. Fish and marine mammals have a similar effect, although for different reasons. The end result is that measuring carbon isotopes alone may not be sufficient for certain populations, and the tests may have to be adjusted based on local diet.
Earlier this spring, Davy and Jahren validated their idea in an underserved rural population in southwestern Virginia that typically drinks a lot of sugar-sweetened beverages, as part of a larger, NIH-funded study looking to decrease added sugar consumption in these individuals. They gathered a sample of 216 adults who reported consuming at least 200 calories in sugar-sweetened beverages each day. The researchers compared stable carbon isotope ratios in a fingerstick blood sample (the kind diabetics use to test their blood sugar) with records of all the food and beverages each participant consumed over three different 24 hour periods.
A study published in the Journal of Nutrition showed that, unlike in the Yup’ik people, those consuming a more standard American diet did not need the addition of nitrogen isotopes to measure added sugar consumption. Measuring just carbon isotopes was sufficient. A second study published in Public Health Nutrition revealed that individuals with high, medium, and low levels of sugar-sweetened beverage consumption could readily be distinguished by their carbon isotope signatures.
Besides showing that the method is effective, the results also show promise for use in the doctor’s office. Currently, physicians can measure blood glucose, which is a snapshot for the amount of sugar in your bloodstream at any given moment, and they can also measure hemoglobin A1c levels. While these tests are unquestionably useful, they don’t tell physicians how much sugar a person is actually eating.
“The problem with those tests is that they’re only going to be elevated in folks who have problems with glucose tolerance, like pre-diabetes or diabetes. If you have folks who are metabolically very healthy, then they could drink sugar-sweetened beverages and it may not have an obvious impact on their blood glucose or hemoglobin A1c levels,” Davy said.
The carbon isotope tests don’t depend on how someone metabolizes sugar, so it’s an independent measure of just how large their sweet tooth really is.
The tests aren’t ready for clinical use just yet. For one, they’re not quite sensitive enough yet to be useful outside a research setting. For another, the equipment needed to measure the isotope ratios is expensive (at least $100,000), although the price is falling rapidly. Soon enough, though, your basic lab tests from the doctor’s office may have an extra line, telling you all about how much added sugar you eat.