Nutritional Epidemiology What is it? How is it important?

1
Nutritional EpidemiologyWhat is it?How is it
important?

• Reference Nutrition Epidemiology
• by Walter Willet

2
Importance of Nutrition problems in disease
prevention

• Small relative risk (1.5 3)
• large population attributable risk
• (RR1-RR0)/RR1
• High fat diet
• (2 1) / 2 1/2
• 75 x 1/2 38
• Radiation
• (100-1) /100 0.99
• 0.001 x 0.99 0.00099 0.001

3

• What is Nutritional epidemiology
• Concept
• diet influences occurrence of diseases
• Relatively new
• the basic method used for gt 200 years to identify
  essential nutrients
• ?Example?
• Observations that fresh fruits and vegetables
• could cure scurvy by Lind
  in 1753.

4

• Nutrition problems in the past
• Typical deficiency syndromes
• Protein energy malnutrition
• Iron deficiency anemia
• Goiter
• High frequency among those with very low intake
• Short latent periods
• Can be reversed within days or weeks

5

• Contemporary nutritional epidemiology
• Major diseases of Western civilization
• Heart disease
• Cancer
• Osteoporosis
• Cataracts
• Stroke
• Diabetes
• Congenital malformations

6
Why is it hard to study contemporary
nutrition-related disease?

• Characteristics
• Multiple causes
• diet, genetic, occupational, psychosocial, and
  infectious factors levels of physical activity
  behavioral characteristics
• Long latent periods
• cumulative exposure over many years, or
  relatively short exposure occurring many years
  before diagnosis
• Occur with relatively low frequency
• despite a substantial cumulative lifetime risk
• Conditions not readily reversible
• May result from excessive and / or insufficient
  intake of dietary factors

7

• The complex nature of diet has posed an unusually
  difficult challenge to nutritional epidemiology
• Diet represents an unusually complex set of
  exposures that are strongly intercorrelated,
  cannot be characterized as present or absent
• Continuous variables often with a rather limited
  range of variation
• Individuals rarely make clear changes in their
  diet at identifiable points in time typically
  eating patterns evolve over periods of years
• Individuals are generally not aware of the
  content of the foods that they eat ? consumption
  of nutrients is usually determined indirectly
  based on the reported use of foods or on the
  level of biochemical measurements

8

• Limitation in nutritional epidemiology
  research
• Lack of practical methods to measure diet for
  large of subjects
• Dietary assessment methods must be
• reasonably accurate
• relatively inexpensive
• Diets of persons within one country are too
  homogeneous to detect relationships with disease

9

• ?Example?(Shekelle et al., 1981) the standard
  deviation (SD) for dietary cholesterol is 68
  mg/1000kcal, and the SD for serum cholesterol is
  54 mg/dl. From metabolic ward studies of Mattson
  et al.(1972), a 10 mg/1000kcal change in dietary
  cholesterol causes 1.2 mg/dl change in serum
  cholesterol thus the expected SD of serum
  cholesterol variation due to dietary cholesterol
  variation is 8.2 mg/dl. The theorectically
  expected correlation between cholesterol intake
  and serum cholesterol is therefore
• Expected SD due to diet 8.2
• Total SD for serum cholesterol 54
• - the SD for dietary cholesterol in the example
  is overstated because it included measurement
  error as well as true variation
• - some factors are associated with both reduced
  cholesterol intake and higher serum cholesterol,
  such as low levels of physical activity and
  knowledge of hypercholesterolemia
• ? The true relationship between diet and serum
  levels is distorted toward an inverse association
  in cross-sectional studies

10

• ?Example?(Shekelle et al., 1981) the standard
  deviation (SD) for dietary cholesterol is 68
  mg/1000kcal, and the SD for serum cholesterol is
  54 mg/dl. From metabolic ward studies of Mattson
  et al.(1972), a 10 mg/1000kcal change in dietary
  cholesterol causes 1.2 mg/dl change in serum
  cholesterol thus the expected SD of serum
  cholesterol variation due to dietary cholesterol
  variation is 8.2 mg/dl. The theorectically
  expected correlation between cholesterol intake
  and serum cholesterol is therefore
• Expected SD due to diet 8.2
• Total SD for serum cholesterol 54
• - correlation between dietary lipid intake and
  serum cholesterol was only 0.08 because of other
  determinants of serum cholesterol
• - some factors are associated with both reduced
  cholesterol intake and higher serum cholesterol
  (e.g. low levels of physical activity and
  knowledge of hypercholesterolemia)
• ? The true relationship between diet and serum
  levels is distorted toward an inverse association
  in cross-sectional studies

11
(No Transcript)
12
(No Transcript)
13
EPIDEMIOLOGIC APPROACHES to DIET and DISEASE

• Sources of the concepts, hypotheses, and
  techniques of nutritional epidemiology
• - biochemistry
• - cell culture methods
• - experiments in laboratory animals
• - metabolic and biochemical studies among human
  subjects
• Findings from in vitro studies and animal
  experiments cannot be extrapolated directly to
  humans
• The basic science areas provide critical
  direction for information that can aid in the
  interpretation of the epidemiologic findings
• New methods for measuring genetic and
  environmental exposures that can be applied in
  epidemiologic studies

14

• Correlation Studies
• Comparisons of disease rates in populations
  with the population per capita consumption of
  specific dietary factors
• Usually the dietary information is based on
  disappearance data
• food produced and imported
• minus the food exported, fed to animals, or..
• ?Example?
• Correlation between meat intake and incidence of
  colon cancer is 0.85 for men and 0.89 for women

15
(No Transcript)
16
(No Transcript)
17

• Correlation Studies
• Strengths of international correlational studies
• - contrasts in dietary intake are typically
  very large
• - the average of diets for persons residing in
  a country are likely to be more stable over-time
  than are the diets of individual persons within
  the country
• - the cancer rates on which international
  studies are based are usually derived from
  relatively large populations and are subject to
  only small random errors

18

• Correlation Studies
• Problems of correlational studies
• -- Many potential counfounders
• genetic predisposition, other dietary factors
  (total energy intake), other environmental or
  lifestyle practices
• ?Example?
• Countries with a low incidence of colon cancer
  tend to be economically underdeveloped
• Limited by the use of food disappearance data
• indirectly related to intake
• variable quality
• Cannot be independently reproduced

19

• Correlation Studies
• Ecologic studies have unquestionably been useful,
  but are not sufficient to provide conclusions
  regarding the relationships between dietary
  factors and disease and may sometimes be
  completely misleading

20

• Special Exposure Groups
• a population that consume unusual diets
  (religious or ethnic)
• Strengths
• - Same strengths as ecologic studies
• - Often live in the same general environment as
  the comparison group ? reduce chance of
  alternative explanations
• Limitations
• - Subject to many of the same limitations as
  ecologic studies
• - Many factors (dietary and nondietary) are
  likely to distinguish these special groups from
  the comparison population

21
(No Transcript)
22

• Migrant Studies and Secular Trends
• Strengths
• - Tease out genetic factors
• - Useful to examine the latency or relevant
  time of exposure
• - Changes in the rates of a disease within a
  population over time provide evidence that
  nongenetic factors play an important role
• Environmental factors are primary causes of
  certain diseases
• Genetic factors may still influence who becomes
  affected given an adverse environment

23

• Case-Control and Cohort Studies
• Case-control studies
• information about previous diet cases and
  control
• Cohort studies
• information dietdisease-free subjects who are
  then followed
• Strengths
• - confounding effects of other factors can be
  controlled
• in the design (matching or restriction)
• in the analysis (multivariate methods)
• - dietary information can be obtained for the
  individuals

24

• Case-Control and Cohort Studies
• Case-control studies
• - More efficiently and rapidly than cohort
  studies because the number of subjects is
  typically far smaller and no follow up is
  necessary
• Problems of case-control studies
• - potential methodologic bias limited range of
  variation in diet, inevitable error in measuring
  intake ? relative risks in most studies of diet
  and disease are likely to be modest (0.5-2.0)
• - these relatively risks are usually based on
  small differences in means for cases and controls
  of only about 5 ? a systematic error of even 3
  or 4 can seriously distort such a relationship
• ? It would not be surprising if case-control
  studies of dietary factors lead to inconsistent
  findings

25

• Case-Control and Cohort Studies
• Problems of case-control studies (cont)
• - Selection of an appropriate control group
• (1) use patients with another disease
  (assuming that the exposure under study is
  unrelated to the condition of this control group)
• (2) use a sample of persons from the general
  population ? low participation rates diets of
  those who participate may differ substantially
  from those who do not
• - potential methodologic bias may be
  particularly troublesome due to the inherent
  biologic complexity of nutrient-nutrient
  interactions (effect of one nutrient may depend
  on the level of another), which may result in
  apparently inconsistent findings
• ?Example?highly consistent positive
  associations between total energy intake and risk
  of colon cancer have been seen in case-control
  studies. In prospective studies, however, either
  no or inverse associations have been found.

26

• Case-Control and Cohort Studies
• Strengths of prospective cohort studies
• - avoid most of the potential sources of
  methodologic bias associated with case-control
  studies
• - dietary information is collected before the
  diagnosis of disease, illness cannot affect the
  recall of diet
• - distributions of dietary factors in the study
  population may be affected by selective
  participation in the cohort low participation
  rates at enrollment, however, will not distort
  the relationships between dietary factors and
  disease
• - provide the opportunity to obtain repeated
  assessments of diet over time and to examine the
  effects of diet on a wide variety of disease,
  including total mortality, simultaneously

27

• Case-Control and Cohort Studies
• Limitations of prospective cohort studies
• - losses to follow-up that vary by level of
  dietary factors can result in distorted
  associations
• - necessary to enroll tens of thousands of
  subjects ? use of structured, self-administered
  questionnaires has made studies of this size
  possible
• - for diseases of relatively low frequency,
  even very large cohorts will not accumulate a
  sufficient number of cases within a reasonable
  amount of time

28

• Controlled Trials
• - Most rigorous evaluation of a dietary
  hypothesis optimally conducted as a randomized,
  doubled-blind experiment
• Strengths
• - minimize the possibility of confounding by
  randomization
• - it may be possible to create a larger
  contrast between the groups being compared by use
  of an active intervention
• - particularly practical for evaluating
  hypothese that minor components of the diet
  (trace elements or vitamins) can prevent disease
  as these nutrients can be formulated into pills
  or capsules
• - can provide unique information on the latent
  periods for change in an exposure and change in
  disease, which is usually difficult in
  observational studies

29

• Controlled Trials
• Limitations
• - time between change in the level of a dietary
  factor and any expected change in the incidence
  of disease is typically uncertain
• ? any lack of difference between treatment
  groups may be due to insufficient duration
• - compliance with treatment diet likely to
  decrease during an extended trial, particularly
  if treatment involves a real change in food
  intake, or the control group may well adopt the
  dietary behavior of the treatment group if the
  treatment diet is thought to be benefical
• - participants enrolled tend to be highly
  selected on the basis of health consciousness and
  motivation ? subjects at highest potential risk
  on the basis of their dietary intake (thus
  susceptible to intervention) are seriously
  underrepresented

30

• Controlled Trials
• Limitations
• ?Example?
• Low folic acid intake is thought to be a risk
  factor for lung cancer.
• A trial of folic acid supplementation is
  conducted among a health conscious population.
• No effect might be observed.
• Because the study population was already
  receiving the maximal benefit of folic acid
  through its usual diet.
• ? Measure dietary intake of folic acid
  before starting the trial
• exclude those with high intakes either before
  randomization or in subanalyses.
• - usually produce an imprecise measure of the
  effect of exposure due to marginally adequate
  sample sizes
• - ethical considerations that require stopping
  soon after a statistically significant effect is
  seen
• - sometimes impossible to conduct due to
  practical or ethical reasons (e.g. smoking and
  lung cancer, alcohol use and human breast cancer
  risk)

31
DASH Feeding Schedule
32
J Am Diet Assoc 1999 8 Suppl.
33
J Am Diet Assoc 1999 8 Suppl.
34
INTERPREDATION of EPIDEMIOLOGIC DATA

• Usually it is concerned whether an association
  represents a true cause-effect relationship when
  it is observed
• Criteria for causality (Hill, 1965) may not be
  applicable in nutritional epidemiologic studies
• - strength of association not likely to be
  strong in nutritional epidemiology
• - consistency of a finding in various studies
  and populations null findings should sometimes
  be expected in nutritional epidemiology
• - presence of a dose-response gradient likely
  to be nonlinear, may be of almost any shape (Fig
  1-2)

35
INTERPREDATION of EPIDEMIOLOGIC DATA

• Criteria for causality (Hill, 1965) may not be
  applicable in nutritional epidemiologic studies
  (cont)
• - appropriate temporal relationship
• - biologic plausibility post hoc biologic
  explanations should be viewed cautiously because
  they can be developed for most observations
• - coherence with existing data pathology of
  most cancers and many other chronic diseases is
  poorly understood ? lack of a well-defined
  mechanism should not be construed as evidence
  against causality

36
INTERPREDATION of EPIDEMIOLOGIC DATA

• Knowledge that an association exists is not
  sufficient to make public or personal decisions
• - such actions require some knowledge of the
  shape and quantitative aspects of the
  dose-response relationship
• - knowledge of the approximate latent period
  between alteration in diet and change in disease
  incidence would be important
• Interpretation of Null Associations
• Possibilities of null associations
• - variation in diet is insufficient
• - variation may exist for the study population,
  but only within a flat portion of the total
  dose-response relationship

37
INTERPREDATION of EPIDEMIOLOGIC DATA

• Interpretation of Null Associations
• Possibilities of null associations (cont)
• - method of measuring dietary intake is not
  sufficiently precise to measure differences that
  truly exist
• - low statistical power due to inadequate
  number of diseased and nondiseased subjects
• - temporal relationship between the measure
  exposure and the occurrence of disease did not
  emcompass the true latent period
• - some unmeasured third variable was related to
  exposure and disease in opposite directions
  (negative confounding variables)
• - methodologic sources of bias

38
INTERPREDATION of EPIDEMIOLOGIC DATA

• Interpretation of Null Associations
• Description of the conditions/limitations of the
  null finding
• - demonstrate that true variation in diet
  exists within the study population and that the
  method of measuring diet provides useful
  discrimination among subjects
• - confidence intervals provide a sense of the
  range of values that are still consistent with
  the data should be adjusted for measurement
  error
• - include a priori power calculations in
  reports
• - range of latent periods encompassed by the
  study
• - dietary and nondietary correlates of the
  primary exposure that have been evaluated as
  potential confounding variables
• - aspects have or have not been evaluated

39
INTERPREDATION of EPIDEMIOLOGIC DATA

• Interpretation of Null Associations
• ?Example?Vitamin C intake determined by a
  detailed quantitative method was 40 mg/day for
  the 10th percentile and 200 mg/day for the 90th
  percentile. During a 5-year follow-up period the
  observed relative risk was 1.0 with a 95
  confidence interval of 0.8-1.3 after adjusting
  for exposure measurement error for a difference
  of 50 mg/day of vitamin C intake, which
  corresponds to a 50 increase for the average
  subject. Finally, adjustment for parental history
  of colon cancer and intakes of dietary fiber and
  calcium did not alter the findings
• ? (1) effects of very low and very high
  vitamin C diets are not being evaluated
• (2) a 10 (but not 30) reduction in risk
  by a 50 increase in vitamin C intake later in
  life may still be possible

40
INTERPREDATION of EPIDEMIOLOGIC DATA

• Multivariate Relationships of Diet and Disease
• Types and amounts of food eaten may be related to
  mportant nondietary determinants of disease (e.g.
  age, smoking, exercise, and occupation), which
  may both distort or confound and modify
  relationships with diet
• Intakes of specific nutrients tend to be
  intercorrelated so that associations with one
  nutrient may be confounded by other aspects of
  the diet
• Intake of one nutrient may modify the absorption,
  metabolism, or requirement for another nutrient,
  thus creating a biologic interaction
• ? Necessity of employing multivariate techniques
  (both stratified analyses and statistical models)
  to adjust for potentially confounding variables
  and examine interactions

41
INTERPREDATION of EPIDEMIOLOGIC DATA

• Multivariate Relationships of Diet and Disease
• Use of multivariate methods requires a careful
  consideration of the precise question that is
  being posed and whether potential covariates are
  true confounders as opposed to effects of the
  primary exposure
• ?Example?inclusion of blood pressure, glucose
  tolerance, serum lipid levels, and a body fat
  measure in a multivariate model would result in a
  misleading conclusion that obesity has little
  relationship with coronary heart disease.
• Application of multivariate methods in
  nutritional epidemiology necessitates maximal use
  of existing knowledge regarding the effects of
  dietary factors to avoid similar problems in the
  future

42
(No Transcript)
43
(No Transcript)