Skip Navigation


Age and Ageing Advance Access originally published online on September 28, 2007
Age and Ageing 2007 36(6):638-643; doi:10.1093/ageing/afm119
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
36/6/638    most recent
afm119v1
Right arrow E-letters: Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when E-letters are posted
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Lang, I. A.
Right arrow Articles by Melzer, D.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Lang, I. A.
Right arrow Articles by Melzer, D.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

Copyright © The Author 2007. Published by Oxford University Press on behalf of the British Geriatrics Society.

Smoking cessation and transition into retirement: analyses from the English Longitudinal Study of Ageing

Iain A. Lang1,, Neil E. Rice2, Robert B. Wallace3, Jack M. Guralnik4 and David Melzer2

1 Epidemiology & Public Health Group, Peninsula Medical School, Exeter, UK
2 Epidemiology & Public Health Group, Peninsula Medical School, RD&E Wonford Site, Barrack Road, Exeter EX2 5DW, UK
3 Department of Epidemiology, College of Public Health, The University of Iowa, Iowa City, IA, USA
4 Laboratory of Epidemiology, Demography and Biometry, National Institute on Aging, National Institutes of Health, Bethesda, MD, USA

Address correspondence to: Iain A. Lang. Tel: +44 (0) 1392 406749. Email: iain.lang{at}pms.ac.uk


   Abstract
Top
 Abstract
Introduction
 Methods
 Results
 Discussion
 Conclusion
 References
 
Background: transitions such as retirement may represent points at which changes in health behaviour occur.

Objective: to assess whether transition into retirement is associated with increased rates of smoking cessation.

Design: population-based prospective cohort study in England.

Setting and Participants: one thousand seven hundred and twelve smokers aged 50 years and over, followed up for 5 to 6 years.

Measurements: work status (working/retired) and smoking status (non-smoker/smoker) at baseline and follow-up.

Results: at baseline, 381 (22.2%) of our respondents had retired, 444 (25.9%) were working and remained in work at follow-up, and 167 (9.8%) transitioned from work to retirement. Seven hundred and twenty (42.1%) had some other status (e.g. unpaid work/unemployment). A total of 42.5% (95% CI 34.9–50.1) of those who retired quit smoking; for those remaining in employment this figure was 29.3% (95% CI 25.0–33.6), and for those already retired it was 30.2% (95% CI 25.5–34.9). In adjusted regression analyses, those aged 55–70 who retired were more than twice as likely (fully adjusted odds ratio 2.50 (95% CI 1.35–4.62)) to quit smoking as those who continued to work. Results were robust when those who retired for reasons of ill-health were excluded.

Conclusions: our results suggest individuals who undergo the transition into retirement are more likely to quit smoking than those who do not. Interventions should be developed to specifically target those who are retiring, or soon to retire, and those who are due to retire should be helped to incorporate smoking cessation into their retirement planning.

Keywords: retirement, smoking, smoking cessation, health behaviour, elderly


   Introduction
Top
 Abstract
 Introduction
Methods
 Results
 Discussion
 Conclusion
 References
 
The promotion of positive changes in behaviour is central to improving the health of individuals and communities. Evidence indicates timing is important in intervening effectively to bring about such changes, i.e. that interventions must be appropriate to an individual's stage of thinking about change [1]. Sociologists have highlighted the importance of life-course transitions and have observed that, as people live longer, participation in late-life transitions, such as retirement, has become an expectation rather than a possibility [2]. Such transitions are often accompanied by changes in identity and in behaviour.

Our thesis is that there is a relationship between behaviour change and life-course transitions in later life, specifically, that individuals who retire are likely to change their health behaviours at the same time. Although the period around retirement has been identified as one at which positive health-behaviour changes could be made [3], there has been relatively little study of peri-retirement health-behaviour changes.

In this study, we focus on smoking cessation. Quitting smoking is known to be of benefit to adults of peri-retirement age: it is beneficial in relation to mortality [4] and a range of conditions including myocardial infarction [5], coronary heart disease [6], stroke [7], diabetes [8] and poor pulmonary function [9, 10]. In contrast, smoking in middle and old age is strongly and consistently associated with a reduction in years of healthy life [11]. In this study we used data from a population-representative sample of older adults in England, followed for 5–6 years, to assess whether self-reported smoking cessation is more common in those who transition into retirement compared to those who stay in work and those who have already retired.


   Methods
Top
 Abstract
 Introduction
 Methods
Results
 Discussion
 Conclusion
 References
 
The English Longitudinal Study of Ageing (ELSA) sample was drawn from households that responded to the Health Survey for England (HSE) in years 1998, 1999 and 2001, and our baseline data come from these interviews. Households were included in ELSA if one or more individuals living there was going to be 50 or over in 2002. There were 19,924 eligible individuals in households that responded to HSE, though not all these individuals participated in HSE. Two thousand five hundred and ninety six of these older individuals died or were ineligible for follow-up; of the remainder, 11,392 (65.7%) became ELSA respondents. (The final ELSA sample contained 708 additional respondents who had joined eligible households since HSE, but these do not form part of our analysis). Analyses of socio-demographic characteristics against census results indicated the ELSA sample remained population representative [12].

Measures of smoking and smoking cessation
We were interested in looking at those who said they were smokers at baseline and in examining whether they still smoked or had quit at follow-up waves 1 or 2. Based on their responses to a range of smoking-related questions at baseline, principally ‘Do you smoke cigarettes at all nowadays?’, we classed respondents as ‘current smokers’ and ‘current non-smokers’. At follow-up in 2002 and 2004, respondents were asked the same questions and we again classed them as smokers or non-smokers. Those who reported that they smoked at baseline and who reported that they did not smoke at one or both of the follow-up waves we classified as quitters.

There were 11,203 ELSA respondents to the baseline smoking questions of whom 19.38% (2,171) were smokers. Of these baseline smokers, 459 (21.14%) were lost to follow-up or provided incomplete data at follow-up and were omitted from our analysis. This left 1,712 individuals who were smokers at baseline, 1,148 of whom were still smokers at follow-up and 564 (32.94%) of whom reported that they no longer smoked. We used baseline cotinine measures to assess accuracy of self-reported smoking status. Mean cotinine levels amongst those who reported that they had never smoked was 4.6 ng/ml (95% CI 2.8–6.4 ng/ml), and among smokers was 308.6 ng/ml (95% CI 296.0–321.3 ng/ml).

Measures of work/retirement status and of reasons for retiring
We were interested in whether the transition from being in employment at baseline to being retired at follow-up had an effect on whether an individual quit smoking. Respondents selected their employment status from a card with various options (including being in paid employment and having retired). At baseline, respondents were asked, ‘Which of these applies to what you were doing last week?’ At follow up, respondents were asked, ‘Which one of these best describes your current situation?’ Respondents were assigned to one of four retirement-transition categories:

  • Employed at baseline, retired at follow-up
  • Employed at baseline, employed at follow-up
  • Retired at baseline, retired at follow-up
  • Other (e.g. long-term sick; full-time caregiver).

Because we were interested in those who were either in work, and therefore, potentially going to make the transition from working to retirement (along with those people who were already retired for comparison), our analysis excluded those who said that they were permanently sick or disabled, or who stated their occupation at baseline as looking after the home or family.

Of the 1,712 baseline smokers, 167 (9.75%) were employed at baseline and made the transition into retirement by follow-up; 444 (25.93%) remained in employment; and 381 (22.25%) were already retired at baseline. The other 720 baseline smokers were in other employment activity groups either at baseline or follow-up. Median length of follow-up was 69 months.

Health events, such as those that may lead to retirement on the grounds of ill-health, are known to have substantial impacts on older smokers' smoking cessation rates [13, 14]. To take this into account, we wanted to differentiate respondents who retired for reasons of ill health from those who retired for other reasons. At follow-up, respondents who reported being retired were asked to select which of the following was their main reason for retiring: reached retirement age; own ill health; ill health of a relative/friend; made redundant/dismissed/had no choice; could not find another job; to spend more time with partner/family; to enjoy life while still young and fit enough; fed up with job and wanted a change; to retire at the same time as husband/wife/partner; to retire at a different time to husband/wife/partner; to give the young generation a chance; other. Of the smokers who entered retirement in the course of our study, 11 (6.6%) reported ‘own ill health’ as their main reason for retirement and we excluded them from the second part of our analyses.

Statistical analysis
We used adjusted logistic regression to estimate the effects on smoking cessation behaviour of being in different retirement-transition categories. Analyses were conducted using STATA SE Version 9.2 and were weighted for non-response.

We controlled for the following factors known to influence smoking behaviour [15]: age; gender; education level, categorised by highest educational qualification attained (degree/higher; intermediate; no qualifications); total household wealth (including financial, physical, and housing wealth, but not pension wealth); body mass index (BMI), measured by a nurse who took height and weight measurements, (categorised as low or normal (BMI <25), overweight (BMI 25 to <30) and obese (BMI >=30)); mean level of alcohol consumption (non-drinkers, <1 drink (14 g of alcohol)/day, > = 1 and <2 drinks/day, and > = 2 drinks/day); presence of a longstanding limiting illness at baseline. We present results separately for those of all ages and for those in the peri-retirement period i.e. aged 55–70.


   Results
Top
 Abstract
 Introduction
 Methods
 Results
Discussion
 Conclusion
 References
 
Table 1 shows the baseline characteristics of smokers in relation to their work and retirement status over the follow-up period. Significant differences between the socio-demographic and health characteristics of those with differing work-retirement patterns over the period are indicated.


View this table:
[in this window]
[in a new window]

  		Table 1. Baseline characteristics of study subjects who were baseline smokers and respondents at follow-up, by work status change over time

 
Figure 1 shows rates of smoking cessation between baseline and follow-up by work status and retirement status over this period, with 42.5% (95% confidence interval 34.9–50.1%) of respondents who transitioned from employment to retirement quit smoking, compared with 29.3% (95% CI 25.0–33.6%) of those who remained in work and 30.2% (95% CI 25.5–34.9%) of those already retired at baseline.


Figure 1
View larger version (17K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Figure 1. (a)Smoking cessation rates by baseline and follow-up work status, all ages; (b)Smoking cessation rates by baseline and follow-up work status, ages 55–70.

 
Table2 shows the outcomes of logistic regressions of work and retirement status on smoking cessation, adjusted for the confounders listed above. Because of the collinearity between age and likelihood of retirement, we present outcomes with and without age as a control variable.


View this table:
[in this window]
[in a new window]

  		Table 2. Outcomes of adjusted logistic regression: effect of change in work status on smoking cessation

 
When those of all ages are included, the odds ratio (OR) associated with quitting among those who moved into retirement, compared to those who stayed in work, is 1.87 (95% CI 1.27–2.81). Amongst those aged 55–70 at baseline, the OR of quitting smoking among retirers compared to those remaining in work is 2.67 (95% CI 1.48–4.84), adjusting for all confounders except age; adjusting for all confounders including age, the OR is 2.50 (95% CI 1.35–4.62). With those who retired because of ill-health excluded, the corresponding ORs are 2.47 (95% CI 1.34–4.60), adjusting for all confounders except age, and 2.33 (95% CI 1.24–4.38), adjusting for all confounders includingage.

Compared to those who remain in work, there is a significantly higher likelihood of quitting among those in other employment groups, and a higher likelihood of quitting among those already retired (not significant). When age is controlled for, effect sizes are reduced but show the same overall pattern.

As a sensitivity analysis we repeated our models controlling for partner's smoking status (where the respondent had a partner who lived with them). The outcomes of this analysis were very similar to those reported. (Results available from authors on request.)


   Discussion
Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
Conclusion
 References
 
Our results show a statistically significant positive association between transition into retirement and smoking cessation. Over the follow-up period of our study, those who transitioned from work to retirement were significantly more likely to quit smoking than those who remained in work. These outcomes are robust to changes in the age range included and to the exclusion of respondents who retired because of their own ill-health. Our findings are in keeping with our hypothesis that quitting smoking is more likely among those who experience the transition from work to retirement than among those who do not.

A number of points should be kept in mind when considering our findings. First, our measures of work and retirement rely on respondents' self-reports, and we recognise that transition into retirement may not be clear-cut or straightforward. For example, retirees may retire gradually, with a period of semi-retirement, rather than switching directly from economic activity to inactivity. Our measures are, however, suitable for establishing general patterns of change in status.

Second, a similar issue affects our assessment of smoking, which also relies on self-reported data. Our categorisation of behaviour into smokers and non-smokers is relatively crude and ignores the psychological complexities of giving up smoking. In fact, when assessed at baseline this categorisation was associated with significant differences in cotinine levels when smokers were compared with those who had never smoked or were former smokers. Evidence suggests incidence of quitting is sensitive to shifts in tobacco use in population subgroups and can be used for evaluation and monitoring [16]. Self-report might also fail to identify those who failed to report that poor health was their reason for retirement, but when we do exclude those who report they retired for health reasons the reduction in OR is very modest, so this is unlikely to have biased our results.

Third, because they rely on cross-sectional data these findings cannot be relied upon to establish causality. It would be difficult, however, to carry out a trial on this topic, and clearly problematic to allocate participants randomly to continue working or to retire. We recognise individuals have a range of attitudes towards retirement and that these attitudes play an important part in shaping their post-retirement behaviours [17, 18]. Recent findings suggest identifying appropriate interventions is key to promoting smoking cessation in older adults [19–21], and our findings about the significance of retirement in relation to quitting highlight a potential opportunity to improve quit rates among peri-retirement adults.

Our study is based on population-representative data from a prospective cohort designed to assess economic and health behaviours in later life. Nonetheless, generalising these findings to other countries should be approached with caution because differences in retirement policy and provision are likely to alter individuals' thinking about retirement. In England there is currently a statutory retirement age, and in countries where this is not the case there may be differences in individual expectations and behaviour around retirement and retirement decisions. The only other longitudinal study that looked at smoking cessation in the retirement transition was based in the United States and found no difference according to whether individuals retired or stayed in work [22]. Why our results differ from these US findings is unclear, but we speculate that the difference may relate to the absence of a statutory retirement age in the United States. It is conceivable that reaching retirement age may, in England, have a heightened significance as a point of transition that brings with it the motivation to make other lifestyle changes, including those relating to health behaviours.

Only a relatively small number of those who were smokers retired during the follow-up period of our study. Further research is needed to assess whether the findings we report here are observed in studies with longer periods of follow-up, and whether they hold in societies where the legal and social status of retirement differs.


   Conclusion
Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 Conclusion
References
 
Individuals participating in a cohort study in England who transitioned into retirement were more likely to quit smoking than those who remained in work or were already retired. Further research is necessary to establish whether similar dynamics influence other health-behaviour changes, and to assess whether the same applies to those who transition from work to retirement in other countries. Given the increased incidence of smoking cessation around retirement that we have identified, consideration should be given to targeting interventions to those planning their retirement and to ensuring that pre-retirement planning incorporates advice about positive health-behaviour change.

Key points

  • Major life-course transitions such as retirement may be points at which people make other changes in their lives, including changes in health-related behaviours, but there has been little empirical assessment of this
  • we assessed smoking cessation in a group of peri-retirement-age smokers and compared quit rates over a 5–6-year period among those who retire, those who stay in work, and those who are already retired
  • adults who transitioned into retirement were approximately two-and-a-half times as likely to give up smoking as those who remained in work
  • our results suggest the peri-retirement period represents a key point at which smoking cessation interventions should be targeted.


   Conflicts of interest
 
None declared.


   References
Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 Conclusion
 References
 

  1. Prochaska JO, Velicer WF. The transtheoretical model of health behavior change. Am J Health Promot (1997) 12:38–48.[Web of Science][Medline]
  2. George LK. Sociological-perspectives on life transitions. Annu Rev Sociol (1993) 19:353–73.[CrossRef][Web of Science]
  3. Nooyens AC, Visscher TL, Schuit AJ, et al. Effects of retirement on lifestyle in relation to changes in weight and waist circumference in Dutch men: a prospective study. Public Health Nutr (2005) 8:1266–74.[CrossRef][Web of Science][Medline]
  4. Rose G, Colwell L. Randomised controlled trial of anti-smoking advice: final (20 year) results. JEpidemiol Community Health (1992) 46:75–7.
  5. Hjermann I, Velve BK, Holme I, et al. Effect of diet and smoking intervention on the incidence of coronary heart disease. Report from the Oslo Study Group of a randomised trial in healthy men. Lancet (1981) 2:1303–10.[CrossRef][Web of Science][Medline]
  6. Kawachi I, Colditz GA, Stampfer MJ, et al. Smoking cessation and time course of decreased risks of coronary heart disease in middle-aged women. Arch Intern Med (1994) 154:169–75.[Abstract/Free Full Text]
  7. Kawachi I, Colditz GA, Stampfer MJ, et al. Smoking cessation and decreased risk of stroke in women. JAMA (1993) 269:232–6.[Abstract/Free Full Text]
  8. Wannamethee SG, Shaper AG, Perry IJ. Smoking as a modifiable risk factor for type 2 diabetes in middle-aged men. Diabetes Care (2001) 24:1590–95.[Abstract/Free Full Text]
  9. Anthonisen NR, Connett JE, Kiley JP, et al. Effects of smoking intervention and the use of an inhaled anticholinergic bronchodilator on the rate of decline of FEV1. The Lung Health Study. JAMA (1994) 272:1497–505.[Abstract/Free Full Text]
  10. Burchfiel CM, Marcus EB, Curb JD, et al. Effects of smoking and smoking cessation on longitudinal decline in pulmonary function. Am J Respir Crit Care Med (1995) 151:1778–85.[Abstract]
  11. Ostbye T, Taylor DH. The effect of smoking on years of healthy life (YHL) lost among middle-aged and older Americans. Health Serv Res (2004) 39:531–52.[CrossRef][Web of Science][Medline]
  12. Taylor R, Conway L, Calderwood L, Lessof C. Methodology. In: Health, wealth and lifestyles of the older population in England: The 2002 English Longitudinal Study Of Ageing—Marmot M, Banks J, Blundell R, Lessof C, Nazroo J, eds. (2003) London: Institute for Fiscal Studies. 357–74.
  13. Falba T. Health events and the smoking cessation of middle aged Americans. J Behav Med (2005) 28:21–33.[CrossRef][Web of Science][Medline]
  14. Salive ME, Cornoni-Huntley J, LaCroix AZ, et al. Predictors of smoking cessation and relapse in older adults. Am J Public Health (1992) 82:1268–71.[Abstract/Free Full Text]
  15. Giovino GA, Henningfield JE, Tomar SL, et al. Epidemiology of tobacco use and dependence. Epidemiol Rev (1995) 17:48–65.[Free Full Text]
  16. Gilpin EA, Pierce JP. Demographic differences in patterns in the incidence of smoking cessation: United States 1950–1990. Ann Epidemiol (2002) 12:141–50.[CrossRef][Web of Science][Medline]
  17. Hornstein GA, Wapner S. Modes of experiencing and adapting to retirement. Int J Aging Hum Dev (1985) 21:291–315.[Web of Science][Medline]
  18. Hopkins CD, Roster CA, Wood CM. Making the transition to retirement: Appraisals, post-transition lifestyle, and changes in consumption patterns. J Consum Mark (2006) 23:89–101.
  19. Schroeder K, Lawlor DA, Montaner D, et al. Self-reported smoking cessation interventions were not associated with quitting in older women. J Clin Epidemiol (2006) 59:622–8.[Web of Science][Medline]
  20. Donze J, Ruffieux C, Cornuz J. Determinants of smoking and cessation in older women. Age Ageing (2007) 36:53–7.[Abstract/Free Full Text]
  21. Tait RJ, Hulse GK, Waterreus A, et al. Effectiveness of a smoking cessation intervention in older adults. Addiction (2007) 102:148–55.[CrossRef][Web of Science][Medline]
  22. Midanik LT, Soghikian K, Ransom LJ, et al. The effect of retirement on mental health and health behaviors: the Kaiser Permanente Retirement Study. J Gerontol B Psychol Sci Soc Sci (1995) 50:S59–61.[Abstract]
Received 5 December 2006; accepted in revised form 21 May 2007.


Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?


This article has been cited by other articles:


Home page
 Eur J Public HealthHome page
K. Henkens, H. van Solinge, and W. T. Gallo
Effects of retirement voluntariness on changes in smoking, drinking and physical activity among Dutch older workers
Eur J Public Health, December 1, 2008; 18(6): 644 - 649.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
36/6/638    most recent
afm119v1
Right arrow E-letters: Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when E-letters are posted
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Lang, I. A.
Right arrow Articles by Melzer, D.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Lang, I. A.
Right arrow Articles by Melzer, D.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?