Food Components & Health summary
2024 - 2025
,Lecture 1:
Study design:
Experimental: Can reveal causation
• Intervention
Observational: Can only reveal correlation, NOT causation
1. Ecological (Epidemiology) Population based
2. Cross-sectional studies
3. Case-Control studies
4. Cohort studies
Ecological studies (Observational)
• Population based / per capita
• Looks at existing data
Advantages:
• Large number of individuals
Disadvantages:
• Expensive
• Does not give information on an individual level
• Can lead to incorrect conclusions. Correlation is NOT causation
Ecological studies or population-based studies can be useful to generate hypotheses, but cannot be used to
actually test hypotheses. The major limitation is that a group association determined in a population-based
study cannot automatically be extrapolated to the individual level
Cross-sectional studies (Observational)
• Looks at large groups of individuals
• Looks at existing data for two conditions. Between an 'exposure' and a 'outcome'
*example. Body weight and blood pressure
• Looks at data at ONE POINT IN TIME 'snapshot'.
Advantages:
• Easy to perform.
• Cheap, as you just look at one point in time
Disadvantages:
• Reverse Causality
• Confounding
• Correlation is NOT CAUSATION
,Case-Control Studies (Observational)
• Retrospective (history focussed, looks back in time)
• Mainly to looks at the relation between diet and a disease
• Case: group of people with the disease
• Control: group of people that does not have this disease
• Both groups should have the same demographic background and preferably, the same lifestyle, but a
different eating habbit.
Advantages:
• Allows for the linkage between diet and a disease
• Still correlation NOT CAUSATION
Disadvantages:
• Selection bias: controls are not selected from the same population as the cases
• Recall bias: systematic difference in the accuracy /completeness of the dietary information recalled by
participants
• Confounding
• Still correlation NOT CAUSATION
Cohort study (Observational)
• Look in the 'future'. participants are followed over a certain moment of time. Can be months or years.
Advantages:
• Most powerful observational study:
• Followed over time (prospective)
• Large number of individuals
Disadvantages:
• Expensive
• Self-reported dietary assessments are not always accurate (for example alcohol consumption)
• Confounding
• Still correlation NOT CAUSATION
Intervention study (Experimental)
An intervention experimental study design involves researchers actively introducing a treatment or intervention to
observe its effects on an outcome. Participants are typically divided into two or more groups, such as a treatment
group (receiving the intervention) and a control group (receiving no intervention or a placebo). The outcomes in these
groups are then measured and compared to determine if the intervention caused any significant changes.
Advantages:
• CAN CONCLUDE CAUSATION
• Randomization
Disadvantages:
• Observer bias: requires blinding of researcher
• Difficult to study effect of intervention on disease risk
• Ethical considerations
, Intervention study parralel:
Two groups, one follow one diet the other another diet (Two groups following a different diet)
ex. 1 group high fat and 1 high sugar > iso-caloric
Measure the outcome in the subjects and see if there is a difference (Individual person bias)
Intervention study crossover:
two groups follow both diet with a washout time in between
ex. group 1 first high fat, group 2 first high carb > iso-caloric
flush time in between
then group 1 high fat, group 2 high carb > isocaloric
Measure the outcome in the subjects and see if there is a difference
Cohort study:
Indeed, the barrage of data from prospective cohort studies is not consistently or conclusively in or against a
stimulatory effect of SFA on CHD risk when compared with the intake of carbohydrates. In simple terms, the studies
don’t clearly show that SFA intake is linked to CHD, leading to two possible conclusions: 1) a high SFA intake is not
associated with an elevated risk of CHD, or 2) the epidemiological research methodologies available are insufficient
and unable to detect a link between SFA and CHD, even if there was one.
Let's analyse the second option in a bit more detail. In prospective cohort studies, individuals write down their dietary
habits at the beginning of the study – allowing estimation of their SFA intake - and are subsequently followed for years
or decades for the occurrence of illness or death. As pointed out in module 1, prospective cohort studies have
intrinsic limitations that affect the conclusions that can be reached from them. For example, people with a
high intake of SFA may have other dietary or non-dietary habits that differ from people with a low intake of
SFA, thereby confounding the analysis. Moreover, the recording of dietary intake may be problematic. People
have difficulty remembering what they ate, or they consciously or subconsciously write down something
else. The problem of accurately recording the intake of nutrients is a major caveat in nutritional epidemiology and is
not just limited to the link between dietary fat and heart disease.
Another serious limitation of contemporary prospective cohort studies is that the overall level of intake of SFA
in the general population is considerably lower than it was 30-40 years ago and that the difference in SFA
intake between individuals is rather small. To put it simply, if everybody consumes similar amounts of SFA,
how can you study the association between high or low intake of SFA and heart disease?
An additional key issue that has created a lot of noise in the entire discussion on SFA and CHD disease is what you
compare SFA with. In practice, people will replace SFA with unsaturated fatty acids or carbohydrates. The analysis of
the connection between SFA and CHD risk only has meaning if we take into account these replacements. Essentially,
the entire analysis of SFA and CHD becomes a comparative analysis between SFA on the one hand, and unsaturated
fatty acids or (refined) carbohydrates on the other hand.
The SFA that most potently raise LDL levels are myristic acid, lauric acid, and palmitic acid.
2024 - 2025
,Lecture 1:
Study design:
Experimental: Can reveal causation
• Intervention
Observational: Can only reveal correlation, NOT causation
1. Ecological (Epidemiology) Population based
2. Cross-sectional studies
3. Case-Control studies
4. Cohort studies
Ecological studies (Observational)
• Population based / per capita
• Looks at existing data
Advantages:
• Large number of individuals
Disadvantages:
• Expensive
• Does not give information on an individual level
• Can lead to incorrect conclusions. Correlation is NOT causation
Ecological studies or population-based studies can be useful to generate hypotheses, but cannot be used to
actually test hypotheses. The major limitation is that a group association determined in a population-based
study cannot automatically be extrapolated to the individual level
Cross-sectional studies (Observational)
• Looks at large groups of individuals
• Looks at existing data for two conditions. Between an 'exposure' and a 'outcome'
*example. Body weight and blood pressure
• Looks at data at ONE POINT IN TIME 'snapshot'.
Advantages:
• Easy to perform.
• Cheap, as you just look at one point in time
Disadvantages:
• Reverse Causality
• Confounding
• Correlation is NOT CAUSATION
,Case-Control Studies (Observational)
• Retrospective (history focussed, looks back in time)
• Mainly to looks at the relation between diet and a disease
• Case: group of people with the disease
• Control: group of people that does not have this disease
• Both groups should have the same demographic background and preferably, the same lifestyle, but a
different eating habbit.
Advantages:
• Allows for the linkage between diet and a disease
• Still correlation NOT CAUSATION
Disadvantages:
• Selection bias: controls are not selected from the same population as the cases
• Recall bias: systematic difference in the accuracy /completeness of the dietary information recalled by
participants
• Confounding
• Still correlation NOT CAUSATION
Cohort study (Observational)
• Look in the 'future'. participants are followed over a certain moment of time. Can be months or years.
Advantages:
• Most powerful observational study:
• Followed over time (prospective)
• Large number of individuals
Disadvantages:
• Expensive
• Self-reported dietary assessments are not always accurate (for example alcohol consumption)
• Confounding
• Still correlation NOT CAUSATION
Intervention study (Experimental)
An intervention experimental study design involves researchers actively introducing a treatment or intervention to
observe its effects on an outcome. Participants are typically divided into two or more groups, such as a treatment
group (receiving the intervention) and a control group (receiving no intervention or a placebo). The outcomes in these
groups are then measured and compared to determine if the intervention caused any significant changes.
Advantages:
• CAN CONCLUDE CAUSATION
• Randomization
Disadvantages:
• Observer bias: requires blinding of researcher
• Difficult to study effect of intervention on disease risk
• Ethical considerations
, Intervention study parralel:
Two groups, one follow one diet the other another diet (Two groups following a different diet)
ex. 1 group high fat and 1 high sugar > iso-caloric
Measure the outcome in the subjects and see if there is a difference (Individual person bias)
Intervention study crossover:
two groups follow both diet with a washout time in between
ex. group 1 first high fat, group 2 first high carb > iso-caloric
flush time in between
then group 1 high fat, group 2 high carb > isocaloric
Measure the outcome in the subjects and see if there is a difference
Cohort study:
Indeed, the barrage of data from prospective cohort studies is not consistently or conclusively in or against a
stimulatory effect of SFA on CHD risk when compared with the intake of carbohydrates. In simple terms, the studies
don’t clearly show that SFA intake is linked to CHD, leading to two possible conclusions: 1) a high SFA intake is not
associated with an elevated risk of CHD, or 2) the epidemiological research methodologies available are insufficient
and unable to detect a link between SFA and CHD, even if there was one.
Let's analyse the second option in a bit more detail. In prospective cohort studies, individuals write down their dietary
habits at the beginning of the study – allowing estimation of their SFA intake - and are subsequently followed for years
or decades for the occurrence of illness or death. As pointed out in module 1, prospective cohort studies have
intrinsic limitations that affect the conclusions that can be reached from them. For example, people with a
high intake of SFA may have other dietary or non-dietary habits that differ from people with a low intake of
SFA, thereby confounding the analysis. Moreover, the recording of dietary intake may be problematic. People
have difficulty remembering what they ate, or they consciously or subconsciously write down something
else. The problem of accurately recording the intake of nutrients is a major caveat in nutritional epidemiology and is
not just limited to the link between dietary fat and heart disease.
Another serious limitation of contemporary prospective cohort studies is that the overall level of intake of SFA
in the general population is considerably lower than it was 30-40 years ago and that the difference in SFA
intake between individuals is rather small. To put it simply, if everybody consumes similar amounts of SFA,
how can you study the association between high or low intake of SFA and heart disease?
An additional key issue that has created a lot of noise in the entire discussion on SFA and CHD disease is what you
compare SFA with. In practice, people will replace SFA with unsaturated fatty acids or carbohydrates. The analysis of
the connection between SFA and CHD risk only has meaning if we take into account these replacements. Essentially,
the entire analysis of SFA and CHD becomes a comparative analysis between SFA on the one hand, and unsaturated
fatty acids or (refined) carbohydrates on the other hand.
The SFA that most potently raise LDL levels are myristic acid, lauric acid, and palmitic acid.
