Baldness and myocardial infarction, interviewer assessments.

Exercises 18.17 and 18.18 considered a case–control study of baldness and heart attacks in men using self-assessments of baldness.

This same study also had the interviewers assess the extent of baldness of study subjects using a Hamilton baldness scale (Figure 18.13).

(a) Which method of ascertaining baldness do you think would be preferable in this study, the self-assessment or interviewer-based assessment?

(b) Table 18.32 cross tabulates the data according to the ascertainments made by the interviewers. Calculate the odds ratios of myocardial infarction associated with each level of baldness according to this classification.

TABLE 18.32 Data for Exercise 18.29. Baldness and myocardial infarction, case–control data. Baldness was assessed by interviewers according to the Hamilton baldness scale (Figure 18.13).

Baldness Cases Controls Nonea 238 480 Frontalb 44 82 Mild vertexc 108 137 Mod vertexd 40 46 Severe vertexe 35 23 Total 465 768

(c) Compare results from this analysis to the results achieved with the self-assessed baldness in Exercise 18.17.