Dietary advice complex

Here's the Point 1st Part in Series: Science 'totality' finds no link 2nd Part in Series: Diet's role in cancer unclear 3rd Part in Series: Dietary advice complex

EPIDEMIOLOGY'S role in the regulatory sector is widespread, i.e., epidemiology is used by the Environmental Protection Agency to develop and enforce environmental standards, by the Occupational Safety & Health Administration to assure workplace safety and by the U.S. Department of Health & Human Services and U.S. Department of Agriculture to make recommendations for diet and health, according to a report by Exponent Inc.

-Last in a series-

Findings from epidemiologic studies may provide "a clear, accurate and precise measure" of public health benefit or risk between an exposure or set of exposures and an outcome or set of outcomes, or findings may be inconsistent or unclear and make the issue more perplexing, according to the report, written by Exponent's Health Science Practice.

The report was commissioned by the National Beef Board and National Pork Board in response to an epidemiologic study three years ago by the American Institute of Cancer Research (AICR) and World Cancer Research Fund (WCRF) that declared "a convincing link" between the consumption of red and processed meats and cancer (Feedstuffs, Nov. 5, 2007).

An epidemiologic study concerns the incidence, distribution and control of a disease in a population segment.

Exponent is a scientific study group that performs risk assessments across several environmental and public health issues.

Nine criteria

In epidemiologic parlance, to draw the link AICR/WCRF did requires that nine causal criteria be found between the risk factor, i.e., red and processed meats, and the disease outcome, i.e., cancer, according to Dr. Dominik Alexander, lead author of the report.

These criteria were developed by Sir Austin Bradford Hill in 1965 and have been applied in hundreds of epidemiologic studies since then, Alexander said.

He broke down the extent to which links between red and processed meat consumption and cancer have failed to meet the Hill criteria.

Temporality. For some cancers, such as breast and colorectal, there is a large volume of work evaluating the effect of meat consumption, but for most cancers, there are relatively few studies on the impact of meat consumption, which limits the causal assessment.

Strength of association. A relative risk estimate between a risk factor and an outcome of less than 1.5 generally is considered weak, and a weak assessment is less likely to reflect a causal association than a strong assessment because other additional or alternative explanations are more likely. Therefore, bias and confounding factors may influence judgment in such a way that an observed association is, in fact, not an association at all.

Most associations between red and processed meats and cancer "across the entire body of epidemiologic literature" are considered weak, with relative risks below 1.5, and most are not considered statistically significant.

Consistency. Generally, it's believed that the more consistent the results, the more probable that the observed association is causal. For some cancers, the majority of associations of red and processed meats with cancer are in the positive direction, but associations across all studies are inconsistent.

Dose response. There is no established dose-response relationship between red and processed meat consumption and cancer, and there is no threshold of consumption beyond which cancer risk begins increasing.

Biological plausibility and coherence. Several factors have been postulated for red and processed meats that would link consumption with cancer, but epidemiological investigations of these factors have not produced coherent, or consistent, conclusions.

Specificity. This criterion refers to an exposure that produces a specific outcome or an outcome that occurs due to a specific exposure. For cancers, specificity cannot be met because their causes are multifactorial.

Experimentation. This criterion refers to decreasing or eliminating an exposure of interest to determine if there is an impact in incidence, but this "is not a practical application" in studies of red and processed meat consumption and cancer.

Analogy. This is the extent to which one exposure-outcome relationship is similar to another exposure-outcome relationship. For instance, if a specific dietary factor is thought to cause a specific cancer, does the same factor then cause cancer at another organ site? It is not clear how or whether such an analogy could be applied to epidemiologic studies about red and processed meat consumption and cancer.

Dietary recommendations "are complex decisions" that involve scientific, economic and ethical considerations, Alexander said, and the underlying science surrounding an exposure and disease or health outcome "should be the foundation on which a recommendation is built."

If the underlying science is tenuous or unclear, as in the association of red and processed meat consumption with cancer, then any subsequent dietary recommendation "may not achieve the effect it's intended to elicit," he said.

Well within

The Exponent report concludes with a discussion of the government's "Dietary Guidelines for Americans" and the 2005 "MyPyramid," which will be revised this year, as it is every five years.

The report notes that MyPyramid has five specific food groupings and currently recommends consuming 5-7 oz. per day from the meat group, which also includes eggs, nuts, seeds and soy-based foods such as tofu.

The report then more or less asks what all the controversy is about anyway because Americans consume an average of only 2.3 oz. of red meat per day (Table) -- not only well within MyPyramid recommendations but even those of the AICR/WCRF study, which recommends that Americans consume 18 oz. or less of red meat per week.

The complete Exponent report, "Red Meat & Processed Meat Consumption & Cancer: A Technical Summary of the Epidemiologic Evidence," can be downloaded at www.beefresearch.org in the "Human Nutrition Research" section and at www.porkandhealth.org.