A retrospective case-control study is conducted to identify dietary factors that may prevent the development of lung cancer among current or former smokers. Cases were ascertained from the New York State Cancer Registry for 2000 through 2004, and controls were identified from random digit dialing of numbers in the same telephone exchanges as the cases. Personal interviews were conducted to verify smoking status and to assess Vitamin A intake from dietary histories. The results are summarized below.

Average Daily Vitamin A Intake Cases Controls

High (> 10000 IU) 10 100

Medium (3500 - 10000 IU) 35 200

Low (

Total 145

1) Calculate the odds ratio for high Vitamin A intake compared to low intake. Interpret the results in a sentence. (Note that if vitamin A is indeed protective, then the OR should be less than 1.0, that is, the risk for high vitamin A should be lower than the risk for low vitamin A).

2) Calculate the odds ratio for medium Vitamin A intake compared to low intake. Interpret the results in a sentence.

Imagine that this is a "nested" case-control study, that is, the cases and controls were obtained from a defined cohort of 10,000 former or current smokers. The 145 cases represented all persons in the cohort who developed lung cancer from 2000 to 2004, while the 600 controls were randomly selected from all of those without lung cancer in 2004. The data for the entire cohort of 10,000 persons are summarized below. Assume that there is no loss to follow-up.

Average Daily Vitamin A Intake Lung Cancer No Lung Cancer Total

High (> 10000 IU) 10 1640 1650

Medium (3500 - 10000 IU) 35 3285 3320

Low (

Total 145 9855 10000

3) Calculate the risk ratios for high and medium Vitamin intake respectively compared to low intake. Do the risk ratios approximately equal the odds ratios? Why or why not?

4) Twenty-nine cases (20%) died from lung cancer by the time that the study was conducted in 2005. Consequently, their next-of- kin were interviewed and asked about the dietary habits of the deceased case. All of the controls were alive, but the investigators decided to randomly select 120 (20%) of the controls and interview their next-of-kin about the control's dietary information instead of asking the control directly. What are the advantages and disadvantages of this approach?

3) Calculate the risk ratios for high and medium Vitamin intake respectively compared to low intake. Do the risk ratios approximately equal the odds ratios? Why or why not? 4) Twentynine cases (20%) died from lung cancer by the time that the study was conducted in 2005. Consequently, their nextof kin were interviewed and asked about the dietary habits of the deceased case' All of the controls were alive, but the investigators decided to randomly select 120 (20%) of the controls and interview their nextofkin about the control' s dietary information instead of asking the control directly. What are the advantages and disadvantages of this approach? A retrospective casecontrol stud).r is conducted to identify dietary factors that may prevent the development of lung cancer among current or former smokers. Cases were ascertained from the New.r York State Cancer Registry for 2000 through 2004, and controls were identied from random digit dialing of numbers in the same telephone exchanges as the cases. Personal interviews were conducted to verify smoking status and to assess Vitamin A intake from dietary histories. The results are summarized below. - verage I a: y Vitamin A Intake cases "-eium II 1 00 00 IU) ow