Background and aims Standard epidemiological measures of the risk of premature death from smoking might be unsuitable for risk communication in actual counselling situations. The rate advancement period (RAP) is an epidemiologic metric that could be useful for conveying information on the benefits of quitting. More effective risk communication could motivate older smokers to make an attempt at quitting. We provide empirical evidence on the impact of smoking, and the benefits of quitting on all-cause mortality and RAPs for people aged 60 years and older in a large cohort of older adults.
Methods Smoking information was obtained from 6545 participants aged 60–74 years of ESTHER, a population-based German cohort. Cox proportional hazards regression was applied to estimate associations of smoking status, amount of smoking and time since smoking cessation with all-cause mortality. Premature mortality was quantified by RAPs.
Results Current smokers had a 2.5-fold increased risk for all-cause mortality (adjusted HR: 2.53, 95% CI 2.10 to 3.03) and an RAP of 10.7 years when compared with never smokers. Strong dose-response relationships were seen with both current and life-time amount of smoking. Compared with current smokers, significant mortality reductions by 30%, 39% and 59%, and rate advancement reductions of 4.0, 5.6 and 10.0 years within <10 years, 10–19 years and ≥20 years after cessation were found for former smokers, respectively.
Conclusions Smoking remains a strong risk factor for premature mortality, and smoking cessation remains highly beneficial also at older ages.
- Global health
- Smoking Caused Disease
Statistics from Altmetric.com
The harms of smoking and the benefits of quitting are well established among middle-aged adults,1–3 while evidence for older ages is more limited.4 Support for smoking cessation at older ages is particularly important given the ageing populations of many countries. The differing goals and motivations of older smokers compared with younger adults5 might necessitate age-tailored interventions to improve this population's willingness to quit smoking.
An epidemiologic measure that could be of particular use in communicating the benefits of cessation for older smokers is the rate advancement period (RAP), which conveys information on the impact of risk factors on the age dimension of chronic disease occurrence or related mortality.6 More specifically, RAP quantifies the age difference by which a risk factor advances disease incidence or mortality among exposed compared with unexposed individuals. It is applicable to outcomes which monotonically increase with age, as happens with most chronic diseases, and total mortality.
In this study, we estimate the association of smoking with all-cause mortality in people aged 60 years and older. We provide covariate-adjusted HRs and RAPs by intensity and duration of smoking for current and former smokers compared with never smokers, and for quitters compared with continuing smokers.
This investigation is based on 6545 residents of Saarland, Germany, aged 60 years and older at baseline (2000–2002) from ESTHER, an ongoing cohort study (mean age: 66.1 years, SD: 4.1 years). Details of the ESTHER study are described elsewhere.7 ,8 The ESTHER study has been approved by the ethics committees of the medical faculty of the University of Heidelberg and the Medical Association of Saarland, and is being conducted in accordance with the declaration of Helsinki. Written informed consent was obtained from all participants.
Smoking status, amount of smoking and age at smoking cessation were assessed via a self-administered questionnaire at baseline. Pack-years for current and former smokers were calculated by multiplying the average number of cigarette packs smoked per day (number of cigarettes divided by 20) by the number of years the person smoked.
Study participants were contacted again 2, 5 and 8 years after baseline. Information on vital status as of 31 December 2010 was obtained for 99.9% of the cohort from the central population registries and the Saarland state statistical office.
Age, sex, alcohol consumption, education and history of cancer, stroke and myocardial infarction were measured in the baseline questionnaire. Systolic blood pressure, diabetes, height and weight were assessed by general practitioners. Blood samples were taken and stored at –80°C. Total cholesterol levels were measured in the laboratory of the university clinics of Heidelberg by a standard HPLC method.
Cox proportional hazards regression was used to estimate adjusted HR and 95% CI. Point estimates of RAPs were derived from multivariable regression models by the ratio of the regression coefficients for the indicators of smoking exposure and for age, and 95% CIs were derived as previously described.6 When derived from the Cox model, both HRs and RAPs among older adults reflect a larger absolute difference in mortality than the same HRs and RAPs among younger adults. All analyses were conducted with SAS V.9.2, and statistical tests were two-sided with p<0.05.
Multiple imputation was employed. Imputation models comprised the predictor variables and covariates (proportion of missing values ranged from 0.1% for body mass index to 10.6% for alcohol consumption). The SAS V.9.2 procedure proc MI was used to impute 20 datasets and proc MIANALYZE was used for estimation of HRs and RAP. Therefore, complete data from 6545 participants could be used in the analyses.
During a mean follow-up of 9 years (r=1.8 years), 944 participants died. This is lower than the number expected based on life tables for the general population of Saarland (n=1223; standardised mortality ratio=0.77).9 ,10
HR and RAP of associations of smoking with all-cause mortality are presented in table 1. By comparison with never smokers, current smokers had a 2.5-fold mortality and an RAP of 10.7 years. For former smokers, mortality was significantly increased by approximately 30% compared with never smokers, and the RAP was 3.3 years. Strong dose-response relationships were found for number of cigarettes smoked per day as well as for pack-years of smoking. Current smokers of more than 20 cigarettes a day had almost 4-fold mortality (HR, 3.91; 95% CI 2.87 to 5.34), with a RAP of 14.4 years. For every 10 cigarettes smoked per day, and for every 10 pack-years of smoking, the RAPs were 4.5 years and 1.9 years, respectively. Results of former smokers according to time since smoking cessation are presented both with reference to never smokers and to current smokers. For former smokers who quit less than 10 years and 10–19 years ago, significantly increased HRs were found when compared with never smokers. No excess mortality was seen for former smokers who had quit 20 or more years ago. The longer the time since smoking cessation, the stronger is the reduction in premature mortality compared with a current smoker. Quitting smoking less than 10, 10–19 and 20 or more years ago was associated with significant mortality reductions by 30%, 39% and 59%, and rate advancement reductions of 4.0, 5.6 and 10.0 years, respectively, compared with persistent smokers.
To our knowledge, this is the first study to provide estimates of RAPs of mortality for continued and former smoking at old age. Even though the detrimental impact of smoking is well known, the worldwide smoking epidemic continues.11 More effective risk communication could assist the fight against the smoking epidemic. Standard epidemiological measures, such as risk ratios or ORs, may be difficult for the public to understand. To enhance quantification and communication of risk factor impact on chronic age-related diseases, risk and RAPs were proposed some years ago6 and applied to a variety of risk factors.12–15 The message to a 60-year-old smoker that he or she has the same mortality risk as a 70-year-old never smoker if he or she continues to smoke (as suggested by our RAP estimate of 10.7 years) might increase motivation to quit. Furthermore, the results of current smokers compared with former smokers might be of high interest in the clinical context. A clinician working with an older current smoker could inform him that a large proportion of the smoking-related excess risk could still be reversed if he quit. This message only takes a few seconds and is simple, direct and strong. Older smokers often mistakenly believe that they are too old to receive a benefit from quitting because the damage is already done.16 Our results suggest that a substantial proportion of excess risk can still be avoided by quitting at older ages, even though the benefits of quitting are greater the earlier a smoker quits. Smoking cessation should be encouraged and supported at any age. Age-tailored smoking cessation programmes for older people should consider their special goals and motivations, such as maintenance of current functioning and slowing decline which could lead to greater willingness to engage in cessation efforts.
Our results for all-cause mortality are in line with those of a systematic review and meta-analysis reporting on smoking and mortality among older adults.4 Our estimates of relative mortality of current and former smokers were also similar to previously reported values for middle-aged populations.1 ,2 ,17 However, in a population of people at older ages, some specifics should be noted. The general mortality rate is substantially higher than in middle-aged populations, hence, the same relative mortality at older age corresponds to a much higher absolute difference in mortality. Furthermore, the ‘depletion of susceptibles’ effect18 ,19 might play a role. Those most susceptible to smoking-related diseases might have already died and left the population at risk.
Our study has specific strengths and limitations. Strengths include the large sample size, detailed assessment of smoking histories, and relevant confounders and completeness of mortality follow-up. However, smoking data were self-reported and, therefore, some misclassification bias is possible, particularly the underreporting of smoking. Also, some recall bias with respect to dates of smoking initiation and cessation cannot be excluded. Although the overall sample size in this study was large and none of the cells had fewer than 50 events, some cells comprised a small number of individuals. Furthermore, results found in the current cohort of 60–74-year-olds might differ from a younger cohort when they reach this age.
In this large population-based cohort study, a strong association of smoking status, amount of smoking, and time since smoking cessation with all-cause mortality was found in people aged 60–74 years at baseline. Current smokers showed a substantially higher risk for premature death compared with former smokers. A clear dose-response relationship of the number of cigarettes smoked daily, and pack-years of smoking, with mortality was shown. These patterns suggest smoking cessation to be beneficial up to the highest ages. Motivation and support for smoking cessation should have priority at any age.
What is already known on this topic?
The harms of smoking and the benefits of quitting are well established among middle-aged adults, while evidence for older ages is more limited. Standard epidemiological measures of the risk of premature death from smoking might be of limited use for risk communication in actual counselling situations.
What this study adds?
This study shows that smoking remains a strong risk factor for premature death at older age. However, a substantial proportion of excess risk can still be avoided by quitting at advanced ages. Rate advancement periods for current and former smoking are provided, which may serve as a simple, direct and effective way to communicate the risks of smoking and benefits of quitting at old age.
The ESTHER study was funded by the Baden-Württemberg state Ministry of Science, Research and Arts (Stuttgart, Germany), the Federal Ministry of Education and Research (Berlin, Germany) and the Federal Ministry of Family Affairs, Senior Citizens, Women and Youth (Berlin, Germany). This analysis was conducted in the context of the CHANCES project funded in the FP7 framework programme of DG-RESEARCH in the European Commission. The project is coordinated by the Hellenic Health Foundation, Greece.
Contributors Carolin Gellert wrote the statistical analysis plan, cleaned and analysed the data, drafted and revised the paper. She is the guarantor. Ben Schöttker helped in writing of the statistical analysis plan, in analysis and interpretation of data, revised the paper for important intellectual content, and approved the final version. Bernd Holleczek and Christa Stegmaier collected data, provided administrative, technical or material support, revised the paper for important intellectual content, and approved the final version. Heiko Müller monitored data collection, provided administrative, technical or material support, revised the paper for important intellectual content, and approved the final version. Hermann Brenner designed the ESTHER study, helped in writing the statistical analysis plan, in analysis and interpretation of data, designed data collection tools, monitored data collection for the whole cohort, obtained funding, supervised the study, revised the draft paper for important intellectual content, and approved the final version. He is also the guarantor.
Competing interests None.
Ethics approval Ethics Committees of the University of Heidelberg, Germany and the Medical Association of Saarland, Germany.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.