In the Bulkow et al study, ETS exposure was highly prevalent in the community (in both cases and controls) and indoor visiting was frequent during the winter RSV season [ 12]; therefore, the ETS exposure variable "smoker in household" was not an adequate proxy for exposure in this population.
But after controlling for other variables, smokers had double the risk of death before 90 compared with nonsmokers, those with diabetes increased their risk of death by 86percentt, obese men by 44percentt, and those with high blood pressure by 28percentt.
Smoking at baseline was treated as a categorical variable: never smokers, former smokers and current smokers.
Among 3 categories of smoking variable (non-smoker, past smoker, current smoker), the age-adjusted mortality rate of those who smoked in the past was highest while of those who never smoked was lowest.
There were no statistically significant differences for the variables smoker, fever, cough, new auscultatory findings or chronic liver disease.
We also controlled for the variables smoker, former smoker, and risk drinker, but they were not significant and the results are not included in the tables.
We classified each variable into smokers and non-smokers.
Maternal smoking was used as a three-level factor variable: non-smoker, moderate smoker (1 9 cigarettes/day) and heavy smoker (>10 cigarettes/day).
Smoking heaviness was categorized into a three-level variable [non-smoker, light smoker (<10 smoked per day) and heavy smoker (≥10 smoked per day)].
The binary response variable (current/recent smokers vs. former and non-smokers) and 0.5 probability cut-point were prospectively defined for the analysis of the validation set.
