Advanced Search

Last Name First Name Company Name City Advanced Search

Actuary of the Future Education & Research Entrepreneurial Actuaries Financial Reporting Forecasting and Futurism Health International Investment Joint Risk Management Long-Term Care Insurance Management & Personal Marketing & Distribution Pension Product Development Reinsurance Smaller Insurance Company Social Insurance & Public Finance Taxation Technology

Research

Research Studies in Pension

Factor Affecting Retirement Mortality (FARM)

This FARM site consists of

• an Introduction
• an Abstract
• a Bibliography of research papers
• a collection of Summaries of the research papers.

Use the Table of Summaries to link to the summaries either by author or by risk factor.

Abstract | Bibliography | Introduction | Table of Summaries

Smoking

Kennedy et al (1996) found smoking to be an independent predictor of total mortality and deaths from cancer. Thun et al (1997) commented that smoking approximately doubles the risk of death. Rogers et al (1999) also found the odds of dying for both heavy and light smokers were about twice that of those who never smoked. This also applied to those who quit smoking but at a late stage. These results were robust to controls for demographic, socioeconomic and other behavioral factors. These conclusions support Vaillant and Mukamal (2001) who state that the absence of cigarette abuse was one of the most important protective factors for successful aging.

Rogers et al (1999) comment that although all smoking categories had significantly higher odds of dying compared with non–smokers, the highest were for current heavy smokers. The mortality of former smokers depended on the volume of their previous smoking and whether they quit before or after contracting a serious illness.

There were high odds ratios of deaths from a large number of diseases, but circulatory diseases was by far the disproportionate underlying cause of death associated with smoking. This may reflect the fact that circulatory disease is a very common cause of death. Smokers are also more likely to have multiple causes of deaths. These findings were largely independent of age and sex.

Women were shown to be less likely to be smokers or have smoked at all, and those that did, smoked fewer cigarettes than men. Rogers et al (1999) suggested that smoking is a major contributor to the current sex gap in mortality.

Rogers et al (1999) comment that earlier research stated that non–Hispanic Whites are more likely to have smoked or to smoke a larger number of cigarettes than Blacks or Hispanics. However, they found that race had a negligible effect on the effects of smoking.

Williams and Collins (1995) states that evidence suggests smoking is increasingly concentrated among the lowest socioeconomic groups and minority populations. Although earlier research showed that those with less education and lower incomes were more likely to be smokers and less likely to quit, in Roger et al (1999) the socioeconomic factors of education and income did not alter the odds significantly. This study also found that marital status had a negligible impact on the effects of smoking.

There is some suggestion that smoking is related to other health behaviors. Thun et al (1997) commented that alcohol is positively correlated to cigarette smoking, but Rogers et al (1999) found that high–risk smokers were not very likely to be high–risk drinkers, nor did alcohol have any significant impact on smoking patterns. Allison et al (1999) found that mortality hazard rates generally increased with BMI1. However, in the higher BMI categories hazard rates tended to be higher in non– smokers or never–smokers. Rogers et al (1999) also found that high–risk smokers were not abnormally overweight. However, being overweight or underweight increased the effects of smoking on mortality.

Lantz et al (1998) found that the distribution of four behavioral risk factors (cigarette smoking, alcohol drinking, sedentary lifestyle and relative body weight) significantly varied by educational attainment and annual household income. Those with the least education and lowest income were significantly more likely to be current smokers, overweight and in the lowest quintile for physical activity. However, they found that the effects of smoking were no longer significant once they were adjusted for demographic, socioeconomic and other health behavior variables. This contradicts Roger et al's (1999) final conclusion as mentioned in the first paragraph.

Wald and Watt (1997) considered smoking from a slightly different perspective. They found that all smokers have a greater risk of lung cancer than lifelong non–smokers. They also found that those who switch from cigarettes do not achieve the lower risk of pipe and cigar smokers who have never smoked cigarettes. Pipe and cigar smokers smoked less tobacco than cigarette smokers, but former cigarette smokers who had switched inhaled tobacco to a greater extent and therefore were at a higher risk. It was shown that men who had stopped smoking cigarettes for at least 20 years had death rates similar to lifelong non–smokers.

Similarly, Rogers et al (1999) concluded that as well as smoking being a crucial mortality risk, the volume of cigarettes smoked also has a significant impact. They suggested that other aspects of smoking status, e.g. age at initiation, age at quitting, type of cigarettes smoked and extent of inhaling may have an effect, but they were not captured.

1BMI (Body Mass Index) is calculated as weight (in kilograms) divided by height (in metres) squared. The unit of measurement is kg/m2.