Never has there been a time like the present when so many different sources are being tapped for information about what foods are best to eat, from a health or any other point of view. Compiling even a list of providers of advice about food available on the internet poses a nearly impossible task. Health, food and nutrition bloggers have gained the trust of a third of millennials, according to a survey undertaken by the International Food Information Council, itself self-described as “a non-partisan, non-lobbying Washington-based non-profit group that communicates science-based information on food issues to those providing information to consumers.” The immensity of doing that reflects the extent to which anything called “science” is lacking from most advice given about food.

It was Kris Sollid, the Council’s director of nutrition communications, who in addressing the International Sweetener Symposium warned that more and more consumers are seeking and trusting information about food from unreliable sources. Trusting bloggers is a particular problem for the millennial generation, he said, noting that “bloggers aren’t always the most credible sources.”

He attributed this trust to lack of knowledge derived from mis-reading labels. He noted how consumer attitudes toward G.M.O. labeling depend on whether genetic manipulation is stated in the question or not. It is only when included in asking about G.M.O. on labels that 42% of interviewed millennials favored requiring such information. Even as these and other health-related issues attract attention, Mr. Sollid said that taste remains the primary deciding factor in food purchases for 84% of people surveyed. It is followed by price at 71%, health and fitness 64%, convenience 52% and sustainability 41%.

Examining those influences and realizing how they are being affected, even lessened, by concerns about the health effects of what is eaten prompt an urgent call for science-based guidance. Too much of what is being offered by bloggers has no basis in science, a situation that is promoted by questioning the findings behind the Dietary Guidelines for Americans issued by two departments of the federal government.

It doesn’t take much study to lead to the conclusion that it is almost impossible to measure accurately the health-related qualities of food from studies that follow eating patterns of a large number of people. Too many experts engaged in this sort of work are highly skeptical of the way that outcomes are largely determined ahead of the study being undertaken. “I can get you any result you want in any observational data set,” declares one authority. How then may such studies be used, and how difficult is it to come up with an approach that is not subject to manipulation?

That search possibly leads to other fields where science has the overwhelming influence on outcomes. One place worthy of a careful look is the research being undertaken by the National Cancer Institute to discover the causes of cancer. The only causes on which all agree are smoking and specific toxins. Yet, it might be tempting to look at genetic, environmental and behavioral factors among specific populations. The hope is that findings resulting from such research may be used in suggesting clinical practice guidelines for physicians. Many different approaches have been used to identify potential causes of cancer, and a similar method makes sense for gaining a scientific basis for what people eat to assure good health. In the cancer search, a variety of epidemiological approaches are being tried as well as examining trends in cancer incidence and mortality.

In much of this work, dietary experiences are included for analysis to relate certain exposures to increased or decreased risk of cancer. It is acknowledged that large studies may be required to have the statistical power required to establish an association. This is expensive work, but well worth it in trying to identify the causes of cancer and certainly well worth doing in seeking scientific findings on the relationship of food to health.