There has been a spate of confusion about the proper number for the deaths due to secondhand smoke. This message puts all the numbers that are floating around in context.
All the numbers that are being quoted come from Table 1.1 of the CalEPA report on secondhand smoke which was republished at NCI Smoking and Health Monograph 10. You can access the full report at
The different numbers are due to two things:
1. The CalEPA report presents a range of risk for some of the end points, most importantly for heart disease.
2. Some numbers that are being used do not include all causes of death due to secondhand smoke, in particular SIDS.
The CalEPA estimates that passive smoking causes 35,000 to 62,000 deaths annually due to heart disease. Put in plain English, this means that secondhand smoke causes at least 35,000 deaths and as many as 62,000 deaths from heart disease. The CDC has used both of these numbers in different places. Both are acceptable, so long as one understands that these are the two ends of the range of risk. The best single number is the middle of this range (what statisticians call the maximum likelihood estimate), 48,500 deaths.
CalEPA has a single estimate of 3000 deaths for lung cancer.
Cal EPA has a range of risk for SIDS of 1900 to 2700 deaths; the best single estimate is the midpoint, 2300 deaths.
(There are many other diseases that secondhand smoke causes, but death estimates have not been produced).
So, here where how the numbers come from:
38,000 deaths (from CDC MMWR) is the lower bound estimate of heart disease deaths (35,000) plus 3,000 lung cancer deaths. This is an estimate of the minimum number of deaths from just these two causes.
65,000 deaths is the upper bound of heart disease (62,000) plus 3000 lung cancer deaths. This is an estimate of the maximum number of deaths from just these two causes.
53,000 is based on the midpoints for heart disease (48,500) plus lung cancer (3,000) plus SIDS (2300 deaths), which adds up to 53,800. As the midpoint and including all the diseases with death estimates, it is the most meaningful estimate.
October 23, 2002
From an earlier article on this subject -
The 38,000 number, included in the 2002 CDC MMWR includes deaths from lung cancer and heart disease. The lung cancer estimates are consistent with the earlier US EPA estimate.
The 53,000 number is from the 1991 Glantz and Parmley analysis, which updates an earlier one by Wells. It includes lung cancer, heart disease, and other cancers. The lung cancer and heart disease numbers are similar to the ones that the CDC came up with in 2002.
The 62,000 number is the upper bound of the estimate of the number of heart disease deaths that the 1997 CalEPA report estimated for heart disease. The lower bound, 35,000, is similar to that used by the CDC and Glantz and Parmley. The CalEPA report is the most complete analysis of heath effects of secondhand smoke available at this time and includes estimates of the deaths from many other disease caused by secondhand smoke (SIDS, etc) .
The 53,000 number includes more than just lung cancer and heart disease, but is a conservative (low) estimate of the number of deaths. Based on the CalEPA analysis, 53,000 is almost certainly an underestimate of the true burden of disease due to involuntary smoking.
None of these estimates include breast cancer. There is ongoing debate on about whether passive smoking causes breast cancer. If the 40% elevation in risk published in a recent meta-analysis holds up, that would mean than secondhand smoke at work is responsible for about 20% of all breast cancer in waitresses and female bartenders. That would make passive smoking the leading controllable risk factor for breast cancer in this group.