Academic journal article Demographic Research

A Modified New Method for Estimating Smoking-Attributable Mortality in High-Income Countries

Academic journal article Demographic Research

A Modified New Method for Estimating Smoking-Attributable Mortality in High-Income Countries

Article excerpt


Preston, Glei, and Wilmoth (2010) recently proposed an innovative regression-based method for estimating smoking-attributable mortality in developed countries based on observed lung cancer death rates. Their estimates for females, however, differ appreciably from some published estimates. This article presents a modified version of the Preston, Glei, and Wilmoth method that includes an age-period interaction term in its model. This modified version produces improved estimates of smoking-attributable mortality that are consistent with results from a modified version of the Peto-Lopez indirect method.

(ProQuest: ... denotes formulae omitted.)

1. Introduction and review of existing methods for estimating smoking-attributable mortality

Attention has recently been directed to methods that estimate smoking-attributable mortality because of their usefulness in examining the role of smoking in mortality differences among developed countries. Several studies have shown a divergence in mortality trends at older ages among developed countries in recent decades, particularly for females. Meslé and Vallin (2006) observed that life expectancy at age 65 for females increased only slightly from 1984 to 2000 in the U.S. and the Netherlands, even as female e65 increased steadily in France and Japan during this period. Janssen, Kunst, and Mackenbach (2007) found similar differences in mortality declines at ages 80 and over in seven European countries, with declines steadily continuing in France, but stagnating in countries such as Denmark and the Netherlands. They concluded that these differences among countries were largely due to differences in smoking-related mortality. The effect of previous cigarette smoking on these mortality trends deserves particular attention. Decades of medical and epidemiological research have demonstrated that cigarette smoking is the leading cause of preventable mortality in the U.S. and most other developed countries (DHHS 2000; Ezzati et al. 2002). More specifically, Staetsky (2009) recently showed that smoking is a principal cause of observed mortality differences among selected developed countries. Staetsky also used the indirect Peto-Lopez method to estimate mortality in these countries in the absence of smoking-attributable deaths.

Various methods have been proposed and used to measure smoking-attributable mortality in populations. Pérez-Ríos and Montes (2008) provided a useful systematic review of these methods. They found that two general types of methods were most commonly used in the research literature. Both types of methods generally calculate the fraction of deaths in a population that are attributable to smoking (smoking-attributable fraction or SAF). The first group of methods, which Pérez-Ríos and Montes called "prevalence-based analysis," calculate SAF from data for current and former smoker prevalence, usually obtained from surveys, and relative risks for smoking, usually obtained from prospective cohort studies, such as the American Cancer Society's Cancer Prevention Study II (CPS II). Pérez-Ríos and Montes cited the method used by the U.S. Centers for Disease Control and Prevention (CDC) to estimate smoking-attributable mortality as an example of a prevalence-based analysis. The CDC method (2008) estimates adult smoking-attributable mortality using current and former smoker prevalence data from the National Health Interview Survey (NHIS), with relative risks for 19 causes obtained from the CPS-II. It estimates infant smoking-attributable mortality using maternal smoking prevalence from birth certificate data and relative risks for four causes for infants of mothers who smoke obtained from a meta-analysis of epidemiologic studies. Separate methods are used by the CDC to estimate smoking-attributable mortality due to secondhand smoke and burns.

Pérez-Ríos and Montes identified the indirect method presented by Peto et al. (1992, 1994) as the other commonly used method for estimating smoking-attributable mortality. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.