PH 601 San Diego State University Epidemiology Worksheet
THIS QUESTION CONTAINS A CHART I WAS UNABLE TO PASTE HERE. SEE ATTACHEMENT
The following table from a 2001 JAMA paper by Pradhan and colleagues (JAMA 2001; 286:327-334) summarizes findings from a nested case-control study. The purpose of this study was to determine if elevated levels of inflammatory markers interleukin 6 (IL6) and C-reactive protein (CRP) are associated with the development of Type-2 Diabetes Mellitus (T2DM) in healthy middle-aged women.
- Using Table 2, select the correct answer (only one) (3 points)
- In the crude analyses, there is no association between T2DM and any of the IL6 quartiles because none the confidence intervals include 1.
- There is a 4.1-fold increase in the fully-adjusted risk of T2DM among persons with an CRP in the third quartile, but this association is not statistically significant.
- Compared to the crude model, the adjusted model including all risk factors results in a 3.7-fold decrease in the relative risk among persons with a CRP in the fourth quartile compared to the those with a CRP in the first quartile.
- Findings suggest no overall dose response of IL6 and risk of T2DM in the crude analysis.
- The unadjusted relative risk (RR) for persons with an CRP in the third quartile is 8.7, but the model adjusted for all risk factor yields a RR of 4.1. How do you explain the difference? (3 points)
- This is most likely an error—as adjustment for confounding should always increase the magnitude of the relative risk.
- The change is irrelevant since both relative risks are statistically significant anyway.
- The unadjusted relative risk was an over-estimate—which could happen if some of the confounders were directly related to exposure or disease.
- Since both relative risks are statistically significant, this indicates that there is little confounding going on.
- Discuss the differences between the BMI-adjusted model and the model that was adjusted for all risk factors in the analysis of IL6 and T2DM only. In your discussion, explain the difference between the two approaches (in brief) and discuss the findings and how they differ. Limited your discussion to 2-3 sentences. (3 points)