Question

Consider a case-control study of the possible association between alcohol consumption and the risk of esophageal cancer as reported by Tyns et al. (1977). The study entailed 200 males diagnosed with esophageal cancer in one of the regional hospitals between the years1972 and 1974, and controls were a sample of 775 adult males selected from electoral lists in each community. The data on alcohol consumption dichotomized at a value of 80 g/d for cases and controls are summarized below:

Average alcohol consumption (g)                  Cases   Controls          Total

>80                                                      96        109                  205

<80                                                      104      666                  770

Total                                                    200      775                  975

Carry out a test of possible association between the average daily alcohol consumption and the risk of esophageal cancer. Calculate 95% CIs of the measure of association and interpret your findings.

Answer 1

Alcohol consumpton	Cases	Control	Total
> 80	96	109	205
< 80	104	666	770
Total	200	775	975

Below is the contngency table :

Row_1	Col._1	fo (Observed freq.)	Expected Freq. (fe)	(fo - fe)	(fo - fe)^2	(fo - fe)^2/fe
1	1	96	42.05128205	53.9487	2910.4642	69.21225766
1	2	109	162.9487179	-53.949	2910.4642	17.86122778
2	1	104	157.9487179	-53.949	2910.4642	18.42664003
2	2	666	612.0512821	53.9487	2910.4642	4.755261942
					Sum	110.2553874