That article reads a lot like a press release. It also includes things like: > The only category of exposure that showed a trend toward increased risk was livi

That article reads a lot like a press release. It also includes things like: > The only category of exposure that showed a trend toward increased risk was living in the same house with a smoker for 30 years or more. In that group, the hazard ratio for developing lung cancer was 1.61, but the confidence interval included 1.00, making the finding of only borderline statistical significance. "Well, it could cause a 61% increase in cancer cases, but our error bars are so wide that there's really no way of knowing, so it's probably fine."