Week Two - Diet Assessment methods Flashcards
common approach in large epidemiologic studies:
FFQ and 24 hour recall ( to help calibrate FFQ)
a source of measurement error in which measurements consistently depart from true value in same direction; affects sample mean as well as percentiles and can result in incorrect estimates and conclusions
bias (systematic error)
measurable substance in an organism whose presence is indicative of some phenomenon such as disease, infection, or enviro exposure
biomarker
process of using equations to rescale data obtained from more-biased, less-accurate instrument based on info obtained from less-biased, more-accurate instrument
calibration
study that compares two groups of ppl: those with disease or condition under study, and a very similar group of ppl who don’‘t have disease/condition
case control study
study that includes measurements on a group of individuals at a single interval in time
cross-sectional study
reporting error that is diff in nature or magnitude between two groups
differential response bias (differential error)
the explanatory variable, also referred to as independent variable in regression model–often a behaviour such as diet
exposure
the target variable, also referred to as the dependent variable in regression model; often a health outcome like the occurrence of a specific disease
outcome
what is reactivity?
change in behaviour due to awareness that behaviour is being measured–may bias result
type of biomarker directly related to intake and not subject to homeostasis or substantial inter-individual diffs in metabolism
recovery biomarker (ie. doubly labelled water, urinary nitrogen)
probability that a test correctly rejects null hypothesis when alternative hypothesis is true
statistical power
deviation from accurate report related to a variety of factors that affect an individual as they respond to verbal/written question
response bias
stat method for correcting estimated regression coefficients for bias due to measurement error in one or more continuous covariates. Used to adjust relative risk estimates for measurement error in studies of the association of diet and health outcomes
regression calibration
the Primer made for clinic or research?
research
random vs systematic?
unpredictable (day to day fluctuations) and can be adjusted for using repeat measures and stat modelling; measurement consistently depart from true value in same direction (eg. underreport sweets), cannottbe adjusted for without availability of reference measure that contains no or error or only random error
systematic error in food diary?
subject burden–>lower quality data, record consumption less often so be like a recall, reactivity
food records good for:
describe diet intake and examine association between diet and other variables, support individuals to make recommended diet changes
error in 24 hour recall?
day to day variation (random), underestimate energy intake,
why is 24 hour recall considered least biased of self report instruments?
no reactivity bias
systematic error of FFQ?
not capture entire diet, difficulty of recall task
FFQ are popular in ___ because ___
large prospective epidemiological studies/retrospective case control; ease of administration, low cost, asks about diet retrospectively, gets less frequently consumed foods
common use of screeners?
large cross-sectional studies when interview time/resources limited for a few diet factors
4 research objectives?
describe diet intake, examine association between diet and dependent variable, examine association between independent variable and diet as dependent variable, and examine effect of intervention
best method for describing population’s diet intake?
24-hr recall (most accurate)
best method for diet as independent on dependent variable (prospective and cross-sectional)
multi 24-hr recalls and FFQ
best method for diet as dependent variable
similar for when diet is independent, but caution that differential response bias can be significant (independent variable is associated with other variables that affect dietary reporting error)
best method for evaluating intervention effect on diet
biomarkers (if not feasible, then 24-hr recall)
___ + ___ = estimate of intake of nutrients or food components of interest
food consumption data ; nutrient composition of food
why care about measuring diet?
assess/monitor food and nutrient intake, formulate and evaluate gov health and ag policy, using data for commercial purposes, conduct epidemiological research, improve human health!
purpose for measuring diet intake:
compare average nutrient intakes of diff groups, rank individuals within group, estimate individuals usual intake
SDA automated multiple pass method 5 step approach:
1) quick list, purpose: collect list of foods and beverages consumed previous day
2) probe for foods forgotten
3) review/record time and eating occasion for each food/bev item
4) cycle through in detail, inquire about detailed description of food, amount, condiments, cooking methods, brands, other relevant info, review
5) final probe
strengths of 24 hour recall?
short, inexpensive, easy to administer, can provide detailed info on types of food consumed, low respondent burden, used estimate nutrient intake of groups, multi recalls can be used to estimate nutrient intake in individuals, more objective than diet history, doesn’t alter usual diet, useful in clinical
limitations of 24 hour recall?
seldom representative of usual intake, underreporting, relies on memory, omissions of dressings/sauces/beverages, data entry labour intensive
two types of food records
weighed, estimated
strengths of food records?
not depend on memory, provide detailed intake data, provide data about eating habits, multi day data more representative of usual intake, valid up to 5 days
limitations of food records?
lots of work, response burden lead to low response rates, requires literacy, takes more time, reactivity, analysis is labour intensive and expensive, training required to increase reliability of measurements
FFQ formats:
simple non-quantitative, semiquantitative (add portion sizes), quantitative (serving size relative to standard serving size)
common FFQ questionnaires:
willed questionaire, block HHHQ, Diet History Questionnaire, NHANES
what are the advantages of targeted FFQ “screeners” to normal FFQ?
do not require assess total diet, don’t always require quantitative accuracy, shorter (lower respondent burden)
FFQ strengths?
can be self administered, machine readable, modest demand on respondents, relatively inexpensive for large sample, more representative of usual intake than few days of records, based on large population data, method of choice for research
FFQ limitations?
not represent usual foods/portion sizes chosen by respondents, intake data compromised when multiple foods groups within single listings, depends on ability of subject to describe diet, requires memory, can be burdensome
used to assess an individual’s usual diet intake over extended period of time
diet history
traditional diet history:
- collect general info about health habits
- usual eating habits (24 hr recall)
- cross check on data in step 2
- complete 3 day food record
examples of food related general health info:
number of meals per day (frequency of eating), dislikes, general appetite, barrier/symptoms to intake, use of nutrition supplements
examples of lifestyle related general health info:
sleep/rest, work, physical activity, smoking, drugs, alcohol, finances
strengths of diet hx:
assesses usual nutrient intake, can detect seasonal changes, data on all nutrients can be obtained, can correlate well with biochem measures
limitations of diet hx
lengthy interview, highly trained interviewers, expensive and difficult to code, may overestimate nutrient intake, requires cooperative respondent with ability recall usual diet
strengths of digital imaging?
good validity, takes less time than 24-hr recall or food records, respondent burden lower, acceptability to subjects, eating habits less affected by recording
limitations of digital imaging:
large initial expense invovlved, periodic revalidation’s, unable distinguish visually similar foods or document prep methods, subject to technical probs
