In: Statistics and Probability
A number of studies have shown a relationship between obesity or body mass index (BMI) and endometrial cancer. A study was conducted among post-menopausal women aged 50-74 years. BMI was used to measure the obesity. BMI >= 30 were interpreted as obese and BMI<30 were as non-obese. Of the total 1055 eligible cases identified through the five regional cancer registries in Sweden, 789 participated in the study but only 592 had complete data to be used in the analysis of obesity and endometrial cancer. Of the 4,268 controls randomly selected from a population register that included all residents of Sweden, 3368 participated in the study, but only 2521 had complete data for the analysis of obesity and endometrial cancer. Following table provides further information about BMI distribution:
BMI |
Cases |
Control |
22.5-24.99 |
154 |
929 |
25-27.99 |
144 |
894 |
28-29.99 |
89 |
348 |
30-33.99 |
120 |
268 |
>34 |
85 |
112 |
<22 (unexposed) |
117 |
772 |
Answer the following questions:
Answer:
Given Data
A) What makes this study a case-control study
This ia a case control study because this study proceeds backward from the effect to cause.. It is a retrospective study based on the disease exposure of the cases and contols. The Outcome (Endometrial cancer) and cause (obisity) has already occured before the study.
B) When are case-control studies an appropriate study design.
Common used study to test an casual hypothesis
Backward study involving exposure histories of both case and control
When needed to investigate about the etiology
To investigate about the rare disease
When follow up is difficult. no follow up is required in future.
It is realtively easy to conduct.
Moderate or inexpensive compared to other studies. can be carried out with minimal resources.
C) What is our outcome and exposure in this study.
Case control study provides a direct estimation of the exposure rates to the suspected factor (obisity) in diseased and non diseased group. Also provide associated disease risk (endometrial cancer) with the exposure (obese)
D) Calculate the odds ratio by BMI category, comparing each BMI category to those less than 22.
Odd's ratio: strength of association of obisity and cancer
lets calculate foe the first group:
group 1 (22.5 -24.99) with < 22
Cases | Control | |
BMI 22.5 - 24.99 |
154 a |
929 b |
BMI < 22 |
117 c |
772 d |
odd's ratio :
= 1.09 ( risk of cancer is 1.9 times more for BMI 22.5 to 24.99 than <22)
similary calculating each category
c and d remains same for all group. c=117 d=772
144*772/ 894*117 =1.06
89*772 /348*117 = 1.6
120*772/268*117= 2.95
85*772 /117*112 = 5
E) Create a 2x2 tables based on BMI >=30 (obese) and BMI < 30.
for making table first calculate cases and control in each section ie < 30 and > 30. for that add up all the values in cases and control given in the question
cases in BMI > 30
= 120+85
= 205
control in BMI > 30
= 268+112
= 380
cases in < 30
= 154 + 144 + 89 + 117
= 504
control in < 30
= 929 + 894 + 348 + 772
= 2,943
lets create the table
Cases | Control | |
BMI > 30 |
205 a |
380 b |
BMI < 30 |
504 c |
2,943 d |
on computing
Odd's ratio =
=
= 3.15
ie Risk of endometrail cancer is 3 times more in women having BMI > 30 than having BMI < 30
F) Calculate the odds ratio for the table created in question e.
Obisity is an risk factor for endometrial cancer
obese women are more prone to cancer than non obese.
BMI is also an confounding factor for cancer.
***Please like it...