A prospective cohort study was used to test the association between body mass index (BMI) and cancer death (Calle EE, et. al. NEJM, 2003). 107,030 non-smoking males were followed for up to 16 years, and 5,314 cancer deaths were observed. Below is a partially completed table of the results from this study. (Please only use information provided in the following table to answer below questions)

Table 2: Cancer Death Rates among 107,030 Male Non-smokers by Body Mass Index

	Body Mass Index (BMI)
	18.5-24.9 (non-exposed)	25.0-39.9 (exposed)
Cancer Deaths (N)	2119	3195
Total observed person-years	699,155	887,736
Relative Risk ( RR = Incidence Ratio)	1.0 (Reference)
Attributable Rate	0.0 (Reference)
% Attributable Rate	0.0% (Reference)
Population Prevalence of BMI	24.1%	75.9%
Population % AR	0.0 (Reference)

1. For this population of male non-smokers, calculate the relative risk of cancer deaths for BMI 25.0-39.9 compared to the reference group BMI 18.5 - 24.9 and 95% confidence interval. Please interpret the relative risk and its confidence interval.
2. For this population of male non-smokers, calculate the attributable rate and the percent attributable rate of cancer deaths for BMI 25.0-39.9 compared to the reference group BMI 18.5 - 24.9. That is, complete the appropriate rows in the table above. And provide the interpretation of the % Attributable Rate of cancer deaths for the BMI 25.0-39.9 group. [Note, these “crude” estimates are based on the assumption that there are no other confounding factors…ideally, we should adjust for age].
3. Based on the population prevalence provided and the numbers that you calculated above, calculate the population % attributable rate (%PAR) for BMI 25.0-39.9 compared to the reference group BMI 18.5 - 24.9. And provide the interpretation of the Population Attributable Rate of cancer deaths for the BMI 25.0-39.9 group.

Answer rating: 100% (QA)

2 0 185241 106249 for 9 B 250 250399 2119 699155 3195 887736 Co... View the full answer