Download Regulatory issues regarding Se-enriched foods: Kathleen Ellwood

Nutrition Related Regulatory
Issues
Kathleen Ellwood, Ph.D.
Director, Division of Nutrition Programs
and Labeling
Office of Nutritional Products, Labeling,
and Dietary Supplements
Topics To Be Covered
• Fortification Policy
• Biotechnology Labeling Guidance
• Health Claims
Fortification Policy
• Established in 1980
• A uniform set of principles as a model for the
rational addition of nutrients to foods
• Lists the 20 vitamins/minerals and ptn for which
there were established U.S. RDAs in 1980;
appropriate fortification levels based on a nutrient
density
Fortification Policy – con’t.
Addition of nutrients to foods can:
• be an effective way to maintain and improve
overall nutritional quality of the food supply
• create nutrient imbalances in the food supply
• contribute to deceptive or misleading claims for
certain foods
Fortification Policy – con’t.
FDA does not consider it appropriate to fortify:
• fresh produce;
• meat, poultry, or fish;
• sugars; or
• “snack foods such as candies or carbonated
beverages.”
Biotechnology Labeling
Guidance
Voluntary Labeling
Indicating Whether Foods Have or Have Not
Been Developed Using Bioengineering
Labeling of
Bioengineered Foods
• Food must be labeled when there is a significant
change in the food:
• nutritional;
• compositional;
• change in conditions of use; or
• when an allergenic component has been
introduced in a food where it does not naturally
occur.
FDA Actions
Held three public meetings in 1999
Conducted focus group studies
Developed draft guidance in
January, 2001
Biotechnology Draft Guidance
• Does not require special labeling of all
bioengineered foods based on production method
• Restates agency’s policy regarding when special
labeling is required for bioengineered foods
• On web at:
http://www.cfsan.fda.gov/~dms/biolabgu.html
What It Takes
To Get A Health Claim
Evidence-Based Ranking System
The review of the evidence for
SSA health claims and
qualified health claims is the
same review.
Reviewing the Evidence
1999 and 2003






Define substance/disease
relationship
Identify relevant studies
Classify studies
Rate studies for quality
Rate for strength of body of
evidence: quantity, consistency,
relevance
Report “rank”
Substance/Disease Relationship
Causal relationship between a substance and a
disease or health related condition for the
general U.S. population or subpopulation:
• Substance – food or food component
• Disease – damage to an organ, structure or
system of the body such that it does not
function properly (e.g. CHD), or a state of health
leading to dysfunctioning (e.g. hypertension)
SSA and Qualified Health Claims
Health Related Condition:
• A condition so equivalent to the disease as to be
indistinguishable from the disease (e.g.
hypertension), or
• A validated modifiable risk factor for disease
(e.g. LDL cholesterol)
56 Fed Reg 2478 at 2481, Jan 6, 1993.
Relationships Between
Diet and Disease
Or Health Related Conditions
Substance
Validated
Modifiable
Risk Factors
Disease/
Health Related
Condition
Health Claim
21 USC 343(r)(5)(D)
Disease
Symptoms
and Signs
Drugs
21 USC 321 (g)(1)(B)
Diseases and Validated
Modifiable Risk Factors
CHD
Total/LDL cholesterol
Cancer
Colorectal/polyps
Diabetes
Blood sugar levels
Osteoporosis
Bone mineral density
Dementia
Mild cognitive impairment
Identify Relevant Studies
• Studies involving healthy humans
• Studies that measure validated, modifiable risk
factors
• Studies that measure reduced risk of disease or
health related condition
Classify Studies
• Intervention studies – “ Gold standard”
-- Measured modifiable risk factor, i.e. LDLcholesterol
soy protein and CHD
walnuts and CHD
-- Measured incidence of onset of disease
omega-3 fatty acid and CHD
fiber-containing grain products, fruit and
vegetables and cancer
Classify Studies – con’t.
• Observational studies – less reliable
-- measurement of substance
-- confounders of disease (other nutrients)
• Prospective cohort
• Case-control
• Cross-sectional – least reliable
Rate Studies for Quality
• Study design
• Data collection
• Inclusion and exclusion of study subjects
• Statistical methods/analysis
+ (good)
 (moderate)
 (poor)
Rate for Strength of the Evidence
• Type of studies
• Quality
• Quantity of studies and sample size
• Consistency of findings
• Relevance to general population or target
subgroup
Report “Rank”
Significant Scientific Agreement
Strong, high quality, relevant and consistent body
of evidence. Conclusions not likely to be
reversed with new data (1999 and 2003)
High Level of Comfort = A
Report “Rank”
Qualified Health Claims
• Moderate/good level of comfort (B)
• Low level of comfort (C)
• Extremely low level of comfort (D)
Qualifying Language
Intended to qualify the level of credible scientific
evidence for a substance-disease (risk-reduction)
relationship in the general population.
But Not Intended To Qualify The…
• Extrapolation of treatment/mitigation data (drugs) to
risk reduction (nutrients).
• Lack of a validated, modifiable risk factor which
serves as the link between a substance and disease
risk reduction.
• Fact that evidence on disease risk reduction is only
observed in animals.
Selenium and Cancer
Selenium and Cancer
Intervention Studies
• Clark et al., 1996 (Nutritional Prevention of Cancer Trial) –
U.S. No effect on the primary end-point (non-melanoma
skin cancer), reduced risk of total, prostate, lung and
colorectal cancer
• Duffield-Lillico et al., 2002;
• Reid et al., 2002 – post-hoc analysis, 7.9 yr follow-up,
reduced risk of total and prostate cancer; no longer with
lung and colorectal; especially for males with low plasma
selenium levels
• Yu et al., 1991 – reduced rate of liver cancer in
malnourished Chinese
• Blot et al., 1993 (Linxian Trial) – lower stomach cancer
mortality in malnourished Chinese
Selenium and Cancer
Intervention Studies – con’t.
Multi-nutrient supplement provided - Not able to
evaluate the effect of Selenium per se
• Blot et al., 1993, 1995 - China
• Li et al., 1993 - China
• Prasad et al., 1995 - India
• Bonelli et al., 1998 - Italy
Selenium and Cancer
Observational Studies
36 Studies
16 prospective cohort studies,
3 retrospective case-control studies,
17 ecological and cross-sectional
½ supported an association for total cancers
½ did not support an association for total cancers
Observational studies showed the greatest benefit with
prostate cancer and no association of Se intake and
breast cancer.
Qualified Health Claim
•
Nutritional Prevention of Cancer Trial, showed a
benefit in males with low plasma selenium levels
with post-hoc analysis; only study applicable to
U.S. population
•
Yu et al., 1991, China, indicated Se per se
decreased liver cancer
•
Blot et al., 1993, 1995, China, Se containing multinutrient decreased stomach cancer
FDA determined evidence to be limited and not
conclusive.
What’s Needed?
• Well controlled clinical trials
• Substance well defined
• Well defined and measured surrogate endpoints
that demonstrate risk reduction
• If validated endpoints lacking, then trials need to
be of long enough duration to measure incidence
of disease
Food Labeling Information
www.cfsan.fda.gov/label.html