Improving Diet/Disease Studies by Using Biomarkers

Roxanne Nelson, RN, BSN

March 18, 2020

Almost on a daily basis, conflicting results from the latest diet/disease study stir up a media frenzy ― this food is linked to cancer, then it isn't.

Most diet/disease studies are based on self-reports and are notoriously unreliable. One way to improve accuracy would be to implement the use of biomarkers in nutritional research, as highlighted in an opinion piece published in the Annals of Internal Medicine.

The use of biomarkers measured in urine, blood, or other biospecimens could strengthen assessments of diet, says Ross Prentice, PhD, member of the Cancer Prevention Program in the Clinical Research Division, Fred Hutchinson Cancer Center, Seattle, Washington.

"These biomarkers may be able to be applied directly in disease association analyses, or may be used to calibrate self-report assessments to reduce systematic and random measurement error influences," he writes.

Important but "Low-Glory Research"

Whether it is feasible to use biomarkers for dietary research is an important question, but there are practical difficulties, said a researcher who was approached for comment.

"There are not many biomarkers that are validated, and these approaches take time and the studies are expensive to conduct," noted Nigel Brockton, PhD, vice president of research at the American Institute for Cancer Research (AICR).

Unfortunately, Brockton pointed out, there is only so much money to go around to fund research. "This is low-glory research," he said. "People want research to give them answers, and there is a lot of competition for funding — developing tools isn't sexy, and these are the studies that will often not get funded.

"They want to look at the causality related to diet but not the tools you need to get that answer," he continued. "It's kind of like being the village blacksmith — he keeps the village moving, but it's not very glamorous."

Biomarkers Would Improve Accuracy

In an interview with Medscape Medical News, Prentice explained that most dietary studies tend to rely on data that are self-reported by study participants. "There a long history studying diet and disease, going back at least 50 years," he said. "It seemed sensible at the beginning, but since then, we have been surprised at how difficult it is to obtain reliable dietary intake and dietary patterns."

Self-reporting is often unreliable, he noted. "Most people report inaccurately even if they're not aware of it, with the most inaccurate being calories or total energy intake."

Most of the studies that have evaluated the relationship of either nutrient intake or specific food intake to disease incidence have been observational. There have been few randomized controlled trials, because these tend to be expensive, take a long time to conduct, and are logistically complicated.

Prentice also points out in his article that systematic reviews of specific recommended dietary behaviors often conclude that the evidence "for chronic disease benefits is of low certainty and that any benefits are probably small." Recent examples of these are systematic reviews of sodium, eggs, and red/processed meat. Findings from these studies are often vigorously debated, and opinions tend to differ regarding the reliability of the epidemiologic data that led to the dietary recommendations.

"Biomarkers can also help differentiate if the association with a chronic disease is indeed the diet or some other individual characteristic," Prentice said. "It can be difficult to tease out causation in observational studies."

For example, is it the diet that is lowering the risk for a certain cancer, or is it that the person is a nonsmoker, has a healthy body weight, and is physically active?

"Even if you could measure diet exactly and all the inputs that you would like to add in, it would be a difficult statistical challenge to sort out which foods or dietary patterns are related to the chronic diseases that appear to be related to diet," he said.

Biomarkers have the potential to reduce the biases associated with self-reported diet, but they tend to relate to limited periods relative to the person's lifespan and risk for disease.

Assessing Biomarkers

In a study published last year in the American Journal of Clinical Nutrition, Prentice and colleagues looked to see whether the use of intake biomarkers could yield more accurate measures of the impact of diet on health.

Their goal was to assess the association of serum-based carotenoid and tocopherol intake biomarkers with the incidence of major cardiovascular diseases, cancers, and diabetes in a subset of 5488 women who were participants of the Women's Health Initiative. They also wanted to accurately estimate short-term micronutrient consumption and intake biomarkers for four micronutrients (α- and β-carotene, lutein plus zeaxanthin [L+Z], and α-tocopherol). The intake biomarkers were obtained by combining serum concentrations with participant characteristics.

Results showed that there were somewhat lower risks for specific cardiovascular outcomes, breast cancer, and diabetes associated with a higher intake of α- and β-carotene. Higher L+Z intake was associated with a lower risk for diabetes, and a higher intake of α-tocopherol was associated with higher risks for certain cardiovascular outcomes. These patterns remained after persons who used dietary supplements were excluded.

The study demonstrated that suitable biomarkers can be calculated from blood specimens obtained in large cohorts and can be used directly in disease-association analyses without the need of self-reported dietary information.

Prentice and his group are now studying small molecules that circulate in body fluids in an effort to identify new biomarkers. They are working with metabolomics expert Daniel Raftery, PhD, also from Fred Hutchinson and the University of Washington.

"I think biomarkers are the logical way forward," Prentice said, "but the problem is that we don't have biomarkers for many aspects of diet."

I think biomarkers are the logical way forward. Dr Ross Prentice

Prentice added, "We are limited by our weak research tools to elucidate dietary data, and there should be a bigger effort in the community to strengthen research enterprise to identify intake biomarkers for other components of diet."

Biomarkers Are Important

Brockton agrees that biomarkers are important and that objective measures are a step forward.

Current studies are hampered by reliance on individuals' recall, especially 24-hour recall. "With that, you only get one snapshot, and it's what the person happened to eat that particular day, and it still comes down to self-report," he said.

There has been a lot of criticism of dietary studies during the past 5 years, "some deserved and some not," Brockton noted. "But unless there is an investment in tools we can use, so we can get more data, observational studies are the only ones that are likely to be done."

Randomized controlled trials are generally not ethical or feasible. "You can't really put people on a diet that may be harmful," he emphasized, "and people will refuse to be randomized to a diet they won't eat."

Another problem is cost. "Observational trials are less expensive than a randomized trial because you can look at multiple endpoints as well, whereas in a randomized trial, you are testing a specific hypothesis, so you are limited," Brockton said. "It's a different approach, but another factor is the time scale. With intervention studies, an issue is compliance — how do you get people to comply and not contaminate the results over years or decades? Which is what may be needed with a dietary study."

The media frenzy that often occurs with new diet/disease studies that contradict one another leads to people not understanding, not knowing, or not believing what's good or what's bad, he added. This confusion is likely to continue as studies roll out.

"Even if it's a biomarker study or a randomized controlled trial, one study is never going to be the definitive answer except in very rare cases," Brockton commented.

Prentice has disclosed no relevant financial relationships. Brockton is affiliated with the AICR.

Ann Intern Med. Published online March 3, 2020. Abstract

For more from Medscape Oncology, follow us on Twitter: @MedscapeOnc.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.