Search tips
Search criteria 


Logo of bmjThis ArticleThe BMJ
BMJ. 2007 April 21; 334(7598): 841.
PMCID: PMC1853166

Living with a smoker increases mortality among non-smokers


Research question Is passive smoking at home associated with an increased risk of death?

Answer Yes.

Why did the authors do the study? Living with a smoker is bad for your health. These researchers wanted to know if it is also life threatening. There is relatively little published information linking passive smoking with increased mortality, and most of it comes from the US. These authors wanted to strengthen the evidence by looking at non-smokers from New Zealand.

What did they do? They linked entries from two New Zealand censuses (1981 and 1996) with national mortality records three years after each census to create two cohorts of non-smokers followed up for three years. They then compared the death rates of cohort members living with smokers and cohort members living in a smoke-free household at the time of the censuses. Together, the cohorts included 668 262 adults aged between 45 and 74.

The authors looked at all cause mortality and mortality from smoking related diseases, including cardiovascular disease and lung cancer. They adjusted their findings for age, ethnicity, and various socioeconomic variables, including education and income, that might confound any link between passive smoking and mortality. They presented their findings separately for the earlier and later cohort and for men and women.

What did they find? Overall, adults living with smokers had a higher risk of death from any cause than adults living in smoke-free homes. The association was significant for men from both cohorts (adjusted relative risk 1.17 (95% CI 1.05 to 1.30) in 1981, and 1.16 (1.04 to 1.3) in 1996). But in women the association was significant only for the later cohort (1.28 (1.16 to 1.42)).

Men and women in the later cohort who were living with smokers were significantly more likely to die from cardiovascular disease (1.25 (1.06 to 1.47) for men, 1.35 (1.11 to 1.64) for women. The researchers also found a significant association between passive smoking and death from respiratory disease and cerebrovascular disease in men from the later cohort (1.81 (1.00 to 3.28), 1.82 (1.20 to 2.77)), but not in women. They found no association between passive smoking and death from lung cancer in either cohort.

What does it mean? These findings confirm that passive smoking is a threat to life as well as health. The authors found a link between reduced survival and living with a smoker that was unexplained by social factors such as poverty, education, or neighbourhood of residence. They are fairly certain the association they found is causal, not least because passive smoking was associated with smoking related diseases, not just all cause mortality. They didn't find a link between passive smoking and lung cancer deaths, but the follow-up was too short and number of cancer deaths too small to be certain of this finding.


  • Hill et al. Mortality among lifelong nonsmokers exposed to secondhand smoke at home: cohort data and sensitivity analyses. American Journal of Epidemiology 2007;165:530-40 [PubMed]

Articles from The BMJ are provided here courtesy of BMJ Publishing Group