Age is a variable that is often responsible for confounding in epidemiology, in part because the occurrence of many diseases changes with age. The change in disease risk with age is oft en referred to as the effect of age. Does it make sense to think of age as having an effect on disease risk, or is it more sensible to think that the effect of age is itself confounded by other factors?
More people in Los Angeles die from cardiovascular disease each year than do people in San Francisco. What is the most important explanation for this difference? What additional factors would you consider to explain the difference in the number of deaths?
In Table 1–2, which age group would you say shows the greatest effect of smoking on the risk of death during the 20-year interval? How have you defined “greatest effect”? What other way could you have defined it? Does your answer depend on which definition you use?
On a piece of graph paper, use the data in Table 1–2 to plot the 20-year risk of death against age. Put age on the horizontal axis and the 20-year risk of death on the vertical axis. Describe the shape of the curve. What biological forces account for the shape?
A physician who was interested in jazz studied the age at death of jazz musicians, whom he identified from an encyclopedia of jazz. He found that the average age at death of the jazz musicians was about the same as that in the general population. He concluded that this finding refuted the prevailing wisdom that jazz musicians tended to live dissolute lives and therefore experienced greater mortality than other people. Explain his error.
A researcher determined that being left -handed was dangerous, because he found that the average age at death of left -handers was lower than that of right-handers. Was he correct? Why or why not?
What is the underlying problem in comparing the average age at death, or the average age at which a person gets a specific disease, between two populations? How should you avert this problem?