Question

Epidemiologists want to explore a potential association between exposure to a certain soap and development of dermatitis (a skin rash). In their initial cross-sectional surveys, they find a strong association between the exposure and outcome (OR = 3.0). They also see evidence that age may be strongly associated with both exposure and outcome. After stratifying their participants into those under age 40 and those ages 40+, they calculate stratum-specific odds ratios, and discover that both ORs = 3.0. Based on these results, is age a confounder?

• A. Yes, because the association remains significant in the total population and in the stratum-specific populations as well.
• B. No, because there is less than a 10% difference between the crude ORs and the stratum-specific ORs.
• C. Unknown, because there may be unmeasured confounding from different variables.

It is well established that alcohol can pass through the placenta to the developing baby, and excess alcohol consumption is known to have a number of adverse effects on the body. Based on this evidence, bottles of wine and other alcoholic beverages include the following statement: “According to the Surgeon General, women should not drink alcoholic beverages during pregnancy because of the risk of birth defects.” The evidence provided above to support the warning label on alcoholic beverages represents which of the following Bradford-Hill causal guidelines:

• A. Strength of association
• B. Biological Plausibility
• C. Consistency of association
• D. Temporality

Answer 1

C. Unknown, there may be unmeasured confounding from different variables.

Here studies are conducted based on, relation between soap and development of skin rashes. Here the rotio was 3.0. And also conducting studies on age above 40 and below 40 , it given rotio 3.0. So here, it is unknown that any other factors are influenced in this. For knowing that the study should be based on different variables with age and then finally can able to come to a conclusion, whether age is a confound or not.

Why answeres are not, because

A. Yes, because the association remains significant in the total population and in the stratum-specific populations as well.

It can't be correct due to other variables are influenced or not is not clear here.

B. No, because there is less than a 10% difference between the crude ORs and the stratum-specific ORs.

Here also not clear, how other variables are influenced or not.  

********************************"""""***********

A. Strength of association

It shows, how an independent variable affects on dependent variable. Or it can also explain as how increasing cause affects on increased risk in outcome.

B. Biological plausibility.

It shows cause and effect relationship. How a a factor leads to effect in biological system. Eg. Continuous cigarette smoking will lead to deposition of carcinogen in lungs, then it will cause lung cancer  

So here pregnant women should not drink because, it will cause birth defects. This is come under biological plausibility.

C. Consistency of association.

It deals also cause and effective relationship. Bt it strongly studying under how the effect varies in different times, place etc. It checks association between certain factors.

D. Temporality.

It is also a type of study in which it checks the character of effect, or how it affected by time, place etc.