No. Correlation is not the same as causation. In order to find a causal link, controlled clinical trials must be performed during which people are put on diets with differing proportions of trans fat.
2.2 Applications and skills
2.2.3 Nature of Science: Correlation and causation
- Epidemiology is the study of patterns, trends and causes of disease and health conditions in a population.
- The main methods used to gather data for epidemiological studies are statistical. Large volumes of data over long time periods are analysed and trends are identified.
- Results gathered from epidemiological studies can show correlations between the investigated variable and a possible cause.
- The limitations of epidemiological research are illustrated in the following example.
Application: Trans fats and saturated fatty acids
- Diets high in saturated fats are positively correlated with high blood cholesterol and higher risk of coronary heart disease.
- Trans fats are artificially hydrogenated vegetable oils. When hydrogen is added to unsaturated oils, they become solid and are used in processed food products as a substitute for animal based saturated fats.
- Margarine is made from trans fats. Analyse the graphs in Figure 2.2.3a.
Figure 2.2.3a – Positive correlation between trans fat consumption (left) and CHD mortality (right) in the USA
- Is it reasonable to conclude that margarine causes CHD based on this and similar data?
- What other lifestyle factors may be responsible for the pattern observed?
- Give examples of health claims that you have heard that are based on epidemiological research.
- Determine which type of correlation (shown in Figure 2.2.3b) matches the claim.
Here is an example to get you started:
Health claim: Regular exercise decreases risk of heart attack and stroke. (Negative correlation)
Skill: Determining body mass index (BMI)
- The body mass index is a diagnostic tool used to quickly determine whether a person is at a healthy weight, or at risk for diseases associated with being either overweight or underweight.
Less than 18
More than 30
- BMI can be calculated using the following formula:
or determined using the nomogram shown in Figure 2.2.3c.
- To use the nomogram, connect the points between the height (right line) and mass (left line)of the subject with a straight line. The BMI is indicated on the point of intersection on the middle line.
Figure 2.2.3d – Shortening-lard
Cooking with fats: vegetable shortening (left) is made of trans fat, while lard (right) is rendered pig fat.
Nature of Science
Correlation and causation: It is important to distinguish between correlation and causation when evaluating epidemiological claims. In different parts of the world, high fat diets are not correlated with CHD. There may be genetic or other lifestyle factors to consider.
- Specialists rarely claim that CHD is caused by bad fat. Instead, they talk of ‘risk factors’ and ‘associations’. Is this type of language any less convincing? What are the advantages and disadvantages of ‘overstating the case’?
- There are conflicting views on the harms and benefits of fat in diets. How do we decide between conflicting views?
Could a medium to large-scale epidemiological study be an appropriate choice for your IA? Brainstorm ideas with a classmate.
Figure 2.2.3e – Anesthetics
Nomograms are important tools for medical workers. For example, doctors may consult specific nomograms to determine the appropriate dosage of drugs.
Print a copy of Figure 2.2.3c and determine your BMI at home.