I appreciate the opportunity to respond to Professor Gross, since it seems, from his reply, that several things may not have been clear in my original rejoinder. However, before beginning, it is worth mentioning that another prospective study of CHD and ETS exposure recently appeared (Kawachi et al. 1997). The study involved 32,046 women aged 36 to 61 in 1982, who had never smoked. After 10 years of follow-up, the relative risk for total CHD, adjusted for a broad range of cardiovascular risk factors, was 1 . 71 (1 . 03, 2 . 84) for those reporting exposure to ETS at home or work. The corresponding age-adjusted relative risk for fatal CHD was 1 . 87 (0 . 56, 6 . 20). Those reporting occasional exposure at home or work had lower relative risks than those reporting regular exposure.
I will discuss the points raised by Gross grouped according to the issues he raises.
ONE-TAILED TESTS
The suggestion that a one-tailed test might be appropriate is based on the following:
- There is a strong and consistent link between active smoking and heart disease (this is hardly in dispute, at this point in time). 2. Environmental tobacco smoke contains substances known to have cardiovascular eects. 3. Experimental evidence indicates that ETS can have eects on time to angina and other acute eects. 4. Studies show eects of chronic exposure to ETS. 5. Eects of ETS on heart disease morbidity are similar to those of ETS on CHD deaths. 6. ETS has been implicated in other diseases such as lung cancer and respiratory disease, indicating that exposure to ETS can have both short and long term eects. 7. The evidence from studies from outside the US, which Gross (1998) summarizes in his Table , indicates that in ยฎfteen study groups where relative risks are presented only two are less than 1 (both are 0 . 97).
Is there consensus among those who do research in heart disease on all of the above? Major research journals representing those with interests in heart disease have taken deยฎnite stands on CCC 1180ยฑ4009/98/020229ยฑ05$17 . 50