Week 1 – Study designs
Study designs used in nutrition research:
- Experimental
o Intervention/ trials
This is the best research design because only this method, can reveal a causal relation between
a particular exposure and outcome.
- Observational
With observational studies, the researchers don’t change anything, but just observe the dietary
habits of a large number of people and examine whether there is relationship with health or
disease. People are thus in their natural context.
Observational studies cannot lead to conclusions. They can rather conclude that certain diet
components are associated with health/diseases.
o Cohort study (prospective)
o Case-control study
o Cross sectional study
o Ecological (population-based) study
Dietary Assessment: If in an observational study we want to make a link between dietary
intake and risk of disease, we need to have solid info on the composition of the diet of the
people in the study.
Self-reported Dietary Assessment: when participants report diet information themselves.
done with:
- Food Frequency Questionnaire (FFQ) (different for different countries)
- Dietary recall/ diary: recall all the foods participants ate for e.g. 24h or a longer
period.
o Advantage: can be quite accurate if done well
o Disadvantage: People tend to change their food consumption
Expensive and time consuming
People lose motivation and become sloppy
From diet to nutrient intake food composition tables
Disadvantage: many food products are not listed, which forces researchers to select an
alternative food product to calculate nutrient intake.
Main advantage of these methods: fairly efficient and can be used to collect the nutrient intake
data of a lot of people.
Main disadvantage: people tend to be less than 100% truthful
, Cross-sectional study:
dietary intake and the outcome are assessed at same time.
Reverse causality: e.g. people with a high BMI are more likely to switch to diet soda in an
effort to lose body weight. And their consumption of diet soda is not a cause, but a
consequence of obesity.
Limitation: The inability to determine which come first, the exposure or the outcome.
However, if the outcome is a parameter that people cannot sense or don’t know about, (like
blood pressure), the direction of the relation is more or less fixed.
Prone for confounding.
Case-control study:
retrospective studies: meaning that they look back
Disadvantage: prone to different types of bias.
- Selection bias: selecting groups of two different social classes for example.
- Recall bias: systematic differences between the cases and controls in the way they
remember or report their dietary intake.
to minimise recall bias use a control group that has a different disease.
Cohort study:
Many people are followed over a long time.
Participants report their dietary habits at the beginning of the study and are asked to report
when they develop a certain disease.
Disadvantages:
- They rely on self reported dietary assessment to acquire information about people’s
diets may not be very accurate.
- Confounding
Cohort studies are the most powerful observational studies. By being generally prospective,
they do not suffer from reverse causality. However, they are still observational studies which
means they reveal associations and nothing more.
The experimental study: interventions/ trials they involve a treatment
Ideally, interventions should be double-blind (neither the participants, nor the investigator
knows what treatment each subject is getting).
Crossover design: All participants get both types of treatment, half of participants gets
treatment A followed by treatment B. The other half gets it in reverse order.
When it comes to interventions, almost anything is possible except causing harm to people.
Any type of dietary intervention needs to be evaluated by the ethical committee for an
approval. Even putting people on a weight gain diet can be problematic. However, this can be
allowed if you also guide them back to their start weight.
Main advantage of intervention studies over observational studies: They can establish whether
a particular dietary treatment causes a change in certain parameters.
Observational studies can reveal associations or correlations, but they do not provide
definitive clues as to whether the associations are causal.
Study designs used in nutrition research:
- Experimental
o Intervention/ trials
This is the best research design because only this method, can reveal a causal relation between
a particular exposure and outcome.
- Observational
With observational studies, the researchers don’t change anything, but just observe the dietary
habits of a large number of people and examine whether there is relationship with health or
disease. People are thus in their natural context.
Observational studies cannot lead to conclusions. They can rather conclude that certain diet
components are associated with health/diseases.
o Cohort study (prospective)
o Case-control study
o Cross sectional study
o Ecological (population-based) study
Dietary Assessment: If in an observational study we want to make a link between dietary
intake and risk of disease, we need to have solid info on the composition of the diet of the
people in the study.
Self-reported Dietary Assessment: when participants report diet information themselves.
done with:
- Food Frequency Questionnaire (FFQ) (different for different countries)
- Dietary recall/ diary: recall all the foods participants ate for e.g. 24h or a longer
period.
o Advantage: can be quite accurate if done well
o Disadvantage: People tend to change their food consumption
Expensive and time consuming
People lose motivation and become sloppy
From diet to nutrient intake food composition tables
Disadvantage: many food products are not listed, which forces researchers to select an
alternative food product to calculate nutrient intake.
Main advantage of these methods: fairly efficient and can be used to collect the nutrient intake
data of a lot of people.
Main disadvantage: people tend to be less than 100% truthful
, Cross-sectional study:
dietary intake and the outcome are assessed at same time.
Reverse causality: e.g. people with a high BMI are more likely to switch to diet soda in an
effort to lose body weight. And their consumption of diet soda is not a cause, but a
consequence of obesity.
Limitation: The inability to determine which come first, the exposure or the outcome.
However, if the outcome is a parameter that people cannot sense or don’t know about, (like
blood pressure), the direction of the relation is more or less fixed.
Prone for confounding.
Case-control study:
retrospective studies: meaning that they look back
Disadvantage: prone to different types of bias.
- Selection bias: selecting groups of two different social classes for example.
- Recall bias: systematic differences between the cases and controls in the way they
remember or report their dietary intake.
to minimise recall bias use a control group that has a different disease.
Cohort study:
Many people are followed over a long time.
Participants report their dietary habits at the beginning of the study and are asked to report
when they develop a certain disease.
Disadvantages:
- They rely on self reported dietary assessment to acquire information about people’s
diets may not be very accurate.
- Confounding
Cohort studies are the most powerful observational studies. By being generally prospective,
they do not suffer from reverse causality. However, they are still observational studies which
means they reveal associations and nothing more.
The experimental study: interventions/ trials they involve a treatment
Ideally, interventions should be double-blind (neither the participants, nor the investigator
knows what treatment each subject is getting).
Crossover design: All participants get both types of treatment, half of participants gets
treatment A followed by treatment B. The other half gets it in reverse order.
When it comes to interventions, almost anything is possible except causing harm to people.
Any type of dietary intervention needs to be evaluated by the ethical committee for an
approval. Even putting people on a weight gain diet can be problematic. However, this can be
allowed if you also guide them back to their start weight.
Main advantage of intervention studies over observational studies: They can establish whether
a particular dietary treatment causes a change in certain parameters.
Observational studies can reveal associations or correlations, but they do not provide
definitive clues as to whether the associations are causal.
