FSIS Logo Food Safety and Inspection Service
United States Department of Agriculture
Washington, D.C. 20250-3700
Speeches

Office of the Under Secretary for Food Safety
U.S. Department of Agriculture

Nutrition, Food Safety, And Risk Assessment--
A Policy-Maker's Viewpoint

Remarks prepared for delivery by Dr. Catherine Woteki, Under Secretary for Food Safety, before Purdue University, West Lafayette, IN, June 18, 1998. 

It’s a pleasure to be here today to talk about nutrition, food safety and risk assessment. As a government food safety official, I am often asked to make policy decisions that could affect the public health. The nature of such decisions varies greatly, ranging from what level of dioxin in poultry poses a hazard, to what is an appropriate performance standard for Salmonella in raw meat and poultry products.

Oftentimes, these decisions must be made without the benefit of complete information. Public health decisions cannot always wait for more scientific data. Of course, this certainly is not a new situation for public health officials. To illustrate this point, I like to tell a story about John Snow, a founding member of the London Epidemiological Society. In 1849, a large waterborne outbreak of cholera occurred in London, and Snow set out to find the source. By studying the distribution of cases, he was able to conclude that the problem was coming from a well in one part of town. To control the epidemic, he simply took the handle off of the pump. It wasn't until 34 years later--in 1883, that the cholera vibrio was identified. But he was able to make the association between disease and the sources--and end the epidemic--without knowing all of the details.

While it is possible to make good public health decisions based on incomplete data, as John Snow did, this certainly is not the preferred mode of operation. We will never reach a point where we have all of the information we need. But we must strive toward not only gathering the best information available, but toward finding ways to organize that information in a manner that helps us to make sound public policy decisions.

Thus, I would like to focus on nutrition, food safety, and risk assessment from a policy-maker's viewpoint. What type of information do policy-makers such as myself need to make sound decisions on food safety and nutrition issues?

Statistics play a very important role in helping us to make such decisions. There are a variety of new, and not so new, methods that have arisen over the last few decades to help decision-makers like myself organize scientific information about risk. Meta-analysis, where the results of many studies are combined to arrive at a conclusion, is one such method. Risk analysis is another such method.

Today, I want to spend my time talking about this second statistical method--risk analysis. I will illustrate its value in helping policy makers make decisions with two examples--one in the nutrition area, and one in the food safety area.

Risk Analysis

Let me begin, however, by first describing the three components of risk analysis--risk assessment, risk management, and risk communication.

Risk assessment is a structured process for determining the risks associated with any type of hazard--biological, chemical, or physical--in a food. It has as its objective a characterization of the nature and likelihood of harm resulting from human exposure to agents in the environment. The characterization of risk typically contains both qualitative and quantitative information and is associated with a certain degree of scientific uncertainty.

Risk management is the process of weighing policy alternatives in light of the results of the risk assessment and selecting and implementing appropriate control options. The goal of risk management is to protect the public health by controlling risks as effectively as possible through the selection and implementation of appropriate measures.

Risk communication is the exchange of information and opinion on risk among risk assessors, risk managers, and other interested parties, including the general public.

This risk analysis approach is a very helpful way for someone like me to deal with very controversial and complex situations that involve public health. Risk assessment helps to organize scientific information and to characterize the nature and likelihood of harm to the public. It helps to identify and define uncertainties, identify data gaps, and identify what assumptions must be made. Thus, risk assessment provides the decision-maker with a level of confidence in that decision.

There are four very distinct steps in the risk assessment process. The definitions for these steps differ slightly depending on whether you are talking about nutritional risk assessment or food safety risk assessment.

The first step is hazard identification, which involves the collection, organization, and evaluation of all information pertaining to a pathogen or a nutrient. Second is dose-response assessment, which determines the relationship between a nutrient or a pathogen and any adverse effects. Third is exposure assessment, which for nutrition involves evaluating the distribution of total daily nutrient intakes among members of a healthy population. For food safety, it involves determining how much of a pathogen might be ingested in a serving of food. The fourth, and last, step is risk characterization, which involves evaluating the risk.

Now I would like to talk about two specific applications of risk assessment that I believe will be of special interest to this audience.

The Development of Tolerable Upper Intake Levels

The first example is a project I was involved with several years ago to revise the Recommended Dietary Allowances (RDA's). As part of this revision, the Food and Nutrition Board was concerned about how to define a tolerable upper level of intake for nutrients. This question is of great importance to nutritionists. The increasing use of dietary supplements, in combination with the long-standing practice of food fortification, provide opportunities for individuals and groups of individuals to ingest levels of nutrients that might put them at risk.

We know from case reports, and from a limited number of studies, that there are adverse health effects associated with high levels of intakes over sustained periods of time. In some cases, we understand the mechanism--and in some, we don't. In some cases, we know there are highly sensitive sub-populations--for others, we don't. A good example of a sensitive sub-population are those people who carry the gene for hemochromatosis, and thus have a predisposition to iron-storage disease.

The task facing the Food and Nutrition Board was how do you take information about a hazard, the dose-response, and intake levels, and characterize the risk associated with high levels of intake. Some had recommended developing a mathematical model to answer this question, but the Food and Nutrition Board decided to use risk assessment.

In applying the risk assessment framework to the development of Upper Limits (UL) for nutrients, the Board had to account for a number of factors.

First, the risk assessment model had to consider the variability in the sensitivity of individuals to adverse effects. Physiological changes and common conditions associated with growth and maturation that occur during an individual’s lifespan may influence sensitivity to nutrient toxicity. Therefore, to the extent possible, Upper Limits had to be developed for each separate age or life-stage group. The Board noted, however, the need to exclude subpopulations with extreme and distinct vulnerabilities due to genetic predisposition or other considerations, which otherwise would lead to Upper Limits that are significantly lower than are needed to protect most people.

Bioavailability also can affect the nature and severity of toxicity. It is well established, for instance, that certain nutrient interactions and the nutritional status of an individual can alter bioavailability. Bioavailability was considered, however, only after more conventional adverse responses were evaluated and a tentative Upper Limit was derived.

The first step in the process of developing the Upper Limits was hazard identification. Human studies and animal studies were the primary types of data used for identifying nutrient hazards. Six key issues addressed in the data evaluation were: (1) evidence of adverse effects in humans, (2) causality, (3) relevance of experimental data, (4) mechanisms of toxic action, (5) quality and completeness of the data base, and (6) identification of distinct and highly sensitive populations.

The second step in the process was dose-response assessment. The Board selected the most appropriate or critical data sets for deriving the Upper Limit based on set criteria, and then identified the "no-observed-adverse-effect level" (NOAEL) or, if that was not available, the "lowest-observed adverse-effect level" (LOAEL). These figures were then adjusted based on the degree of uncertainty to arrive at the Upper Limit. The larger the uncertainty, the smaller the Upper Limit, which is consistent with the ultimate goal of the risk assessment—to provide an estimate of a level of intake that will protect a healthy population. The committee recommended tolerable upper intake levels for B vitamins and choline.

That is the risk assessment model that was used to identify upper levels of nutrient intake. The details of how the committee applied this approach to the B vitamins can be found in the recently available report, Dietary Reference Intakes—Thiamin, Riboflavin, Niacin, Vitamin B6, Folate, Vitamin B12, Pantothenic Acid, Biotin, and Choline.

Risk Assessment for Salmonella Enteritidis

The second example I will use to illustrate the value of risk assessment is in the food safety arena. A group of USDA, FDA, CDC, and university scientists recently completed a risk assessment on Salmonella Enteritidis in eggs and egg products. This is our first quantitative, farm-to-table microbial risk assessment, and we expect it to serve as a prototype for future risk assessments.

We began the SE risk assessment in response to an increasing number of human illnesses attributed to the consumption of contaminated eggs. Data from the Centers for Disease Control and Prevention indicate that SE is one of the most commonly reported causes of bacterial foodborne illness in the United States and has been increasing since 1976.

This increase has occurred despite a variety of initiatives we, and the industry, have had underway to address the problem. In fact, data from the risk assessment indicate that consumption of contaminated eggs results in an average of 661,633 human illnesses per year.

The risk assessment has several modules, reflecting each stage of the farm-to-table continuum. The first module—shell egg production, simulates the daily SE-positive egg infection frequency for U.S. commercial flocks. The second module—shell egg processing and distribution—follows the shell eggs from collection on the farm through processing, transportation, and storage. The third module—egg products processing and distribution—tracks the change in numbers of SE organisms in egg processing plants from receiving through pasteurization. The fourth module—preparation and consumption—describes exposure from the consumption of eggs and egg-containing foods that are contaminated with SE. The fifth module—public health outcomes—links exposure to eggs and egg products containing SE with the public health outcomes of morbidity and mortality.

With this risk assessment, we now have a farm-to-table model that we can use to determine the effects of specific interventions on the incidence of illness. In fact, as part of the risk assessment, the team evaluated a number of possible interventions on the expected number of human illnesses. They included shell egg cooling, diverting eggs from flocks with a high prevalence of SE-positive hens to breaker plants for pasteurization, and reducing the prevalence of SE-positive flocks. The potential uses for such a model in assessing risk reduction strategies are evident.

Now that we have completed this risk assessment, the results will be available to risk managers as they develop a comprehensive strategy to address the problem of SE in eggs and egg products. We recently published a notice in the Federal Register announcing our intention to initiate a comprehensive and coordinated process of addressing the SE problem in shell eggs. The risk assessment will be very useful as we proceed with this process of risk management analysis.

Strengths and Limitations of Risk Assessment

It is clear that risk assessment is a valuable tool is helping us make public health decisions in both the nutrition and food safety areas. As I said earlier, risk assessment helps to organize information about risk in a helpful manner. It also characterizes the nature and likelihood of harm to the public. It helps to define the uncertainties and provides some level of comfort with the inferences that are made. And it points out data gaps that can help us to prioritize research needs.

But there are some weaknesses to the risk assessment process as well. Risk assessment has its roots in toxicology and carcinogenicity studies, and its application to other disciplines poses significant challenges. For food safety, one challenge relates to the fact that, unlike chemical, environmental, or toxicological contaminants, bacteria can multiply and produce toxins as conditions change as food moves through the farm-to-table continuum. For nutritional risk assessments, one must consider that nutrients are not substances to be avoided, like pathogens or toxic chemicals, but are essential for human well-being and often for life, although they may share with other chemicals the capability of producing adverse effects at excessive exposures.

In addition to the difficulties in applying risk assessment to pathogens and nutrients, risk assessment also is subject to two types of uncertainties-- those related to data, and those associated with any assumptions that are required when directly applicable data are not available.

It would be helpful for statisticians and nutritionists—at this conference and others like it—to establish a dialogue on these inherent weaknesses and other questions for which answers are needed. A publication entitled Risk Assessment in the Federal Government: Managing the Process, which was published by the National Research Council in 1983, lists a number of such questions, and I would like to review some of these with you. Although the book is now 15 years old, these questions are as germane today as they were then. They are organized according to the steps in the risk assessment process.

In the area of hazard identification, these questions were raised:

In the area of dose-response assessment:

In the area of exposure assessment,

In the area of risk characterization,

Closing

I hope I have left you with enough to think about. I’m sure that you will have answers to these questions in a few hours. But seriously, I do look forward to our making progress on these and other difficult questions that are related to conducting sound risk assessments. Having the answers will make it a lot easier for people like me to make those tough decisions.

Banner

For Further Information Contact:
FSIS Food Safety Education and Communications Staff
Public Outreach and Communications
Phone: (202) 720-9352
Fax: (202) 720-9063

Speeches Menu | FSIS Home Page | USDA Home Page