Diet Disease Relationships

0.0(0)
Studied by 0 people
call kaiCall Kai
learnLearn
examPractice Test
spaced repetitionSpaced Repetition
heart puzzleMatch
flashcardsFlashcards
GameKnowt Play
Card Sorting

1/45

encourage image

There's no tags or description

Looks like no tags are added yet.

46 Terms

1
New cards

What is noise when it comes to nutrition research?

vibration or fluctuation in the data that interferes with our ability to perceive the signal


2
New cards

What is signal when it comes to nutrition research?

relationship between a dietary exposure and a health outcome


3
New cards

What are three assessment methods for dietary data?

  1. food records

  2. 24 hr recalls

  3. food frequency questionnaires


4
New cards

What are some characteristics associated with food records.


5
New cards

What are some advantages of food records (6)?


6
New cards

What are some disadvantages of food records (7)?


7
New cards

What are some characteristics of 24-hr dietary recalls?


8
New cards

What are some advantages of 24-hr recalls? (6)


9
New cards

What are some disadvantages of 24-hr recalls? (3)


10
New cards

Characteristics associated with food frequency questionnaires (FFQ).


11
New cards

Advantages of FFQs (7)


12
New cards

Disadvantages of FFQs. (5)


13
New cards

What are the trade-offs in terms of the different dietary assessment methods?

24 hr more common bc they are the in-between

14
New cards

What are new technologies in terms of dietary assessment?

  1. food record apps

  2. automated cameras

  3. incorporation of AI (could help predict food ID and portion size from image, but requires population specific training datasets)


15
New cards

What are three different types of measurement errors that can occur in dietary data assessment?

  1. omission (leaving something out)

  2. estimation of quantities

  3. bias


16
New cards

In what cases can we see omission in dietary data?

  • missed foods or details in food records

  • forgotten foods in 24 hr or FFQ

  • addition of habitual or desirable foods also possible


17
New cards

In what cases can we see estimation of quantities errors in dietary data?

  • portion sizes in 24 hr or FFQ

  • frequencies in FFQ


18
New cards

In what cases can we see measurement bias in dietary data?

  • altered eating patterns in food records

  • misreporting in 24 hr or FFQ


19
New cards

What types of noise is present within diet data (3)?

  1. systemic error/bias

  2. random error

  3. correlations among aspects of diet or between diet and lifestyle


20
New cards

What is systemic error?

consistent under or over estimation of intakes


21
New cards

What are three different types of biases in dietary assessment?

  1. social desirability bias

  2. recall bias

  3. flat-slope syndrome


22
New cards

What occurs as a result of social desirability bias?

overestimate of “good foods” and underestimate of “bad foods”

23
New cards

What occurs as a result of recall bias?

  • cases in a case/control study recall their food intake differently than controls bc of the knowledge of their disease status

  • related but different from recall error (errors of omission or addition due to misremembering)


24
New cards

What occurs as a result of flat slop syndrome?

overestimate foods consumed rarely, underestimate foods consumed frequently

lighter green = truth

darker green = what was reported

25
New cards

Do dietary assessments on average under or overestimate intake?

dietary assessments on average tend to underestimate intakes

  • results of biomarker validation studies show that 0 to > 30% underestimate energy intake

  • vias varies by study population , respondent characteristics


26
New cards

Describe the relationship of RCTs with bias. (2)


27
New cards

Describe the relationship of observational studies with bias. (2)


28
New cards

What is the difference between systemic and random error?

  • scattered around the mean/truth, doesn’t move the mean.

  • random error is a mis-estimation of intakes scattered evenly above and below true values

measurement error, variation in nutrient contents of foods, day-to-day variability in diet can contribute to random error

29
New cards

In dietary assessments, why is there variation in nutrient content of foods?


30
New cards

How does day-to-day variability in a diet contribute to random error?

  • can have different average consumptions amongst people but on day to day there can be overlap amongst the individuals depending on the day


31
New cards

What are other noisy issues?

diet data is noisy, we need to strengthen the signal

32
New cards

What are the three components that go into intake variance?


33
New cards

What is the problem with random error if the mean is still the same? (3)

  • the larger variance leads to wider confidence intervals and less statistical power to detect a difference between groups

  • odds ratio and RR estimates will often be biased towards the null

  • regression coefficients will often be biased towards the null (will be stretched and lead to less association as the conclusion)

overall biased towards the null


34
New cards

What happens to OR when we have systemic misclassification?


35
New cards

What happens to OR when you have both systemic bias and random error?

synergistic effect

  • estimate is even farther from the truth


36
New cards

What are two things you need to ask in order to evaluate dietary assessment tools?


37
New cards

What are three ways to validate nutrient intake estimates?

there are still sources of error that can reside in these measures

38
New cards

What are two examples of nutrient intake biomarkers ?

  1. skin carotenoids via “veggie meter” as a measure of fruit and vegetable intake

  2. doubly labeled water estimate of energy expenditure


39
New cards

What are some sources of error that could occur in biomarker measurements?


40
New cards

What are some sources of error that could occur in FFQs?


41
New cards

What are some sources of error that could occur in repeated weighed food records?


42
New cards

What are some reasons as to why nutrient biomarkers do not correlated strongly with dietary data?


43
New cards

How do you reduce random error in 24 hr recalls?

repeat 24 hr recalls over multiple days

in general repeat measures to estimate each indvidual’s usual intake

44
New cards

When looking at group level intake, how do we improve precision in dietary data?


45
New cards

When looking at individual level intake, how do we improve precision in dietary data?


46
New cards

What are two ways to strengthen the signal?

  1. statistical modeling methods that account for day today variability using repeat assessments among sub-sample

  2. study design and analyses that account for covariation