The European Journal of Public Health Advance Access originally published online on May 31, 2008
The European Journal of Public Health 2008 18(5):504-508; doi:10.1093/eurpub/ckn037
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Health-related behaviours |
Employment trajectory as determinant of change in health-related lifestyle: the prospective HeSSup study
Pekka Virtanen1, Jussi Vahtera2, Ulla Broms3,4, Lauri Sillanmäki3, Mika Kivimäki5 and Markku Koskenvuo3
1 Tampere School of Public Health, University of Tampere, Finland
2 Finnish Institute of Occupational Health, Turku, Finland
3 Department of Public Health, University of Helsinki, Finland
4 Department of Mental Health and Alcohol Research, National Public Health Institute, Helsinki, Finland
5 Department of Epidemiology and Public Health, University College London, the United Kingdom
Correspondence: Pekka Virtanen, Tampere School of Public Health, 33014 University of Tampere, Finland, tel: +358-3-35516007, fax: +358-3-35516057, e-mail: pekka.j.virtanen{at}uta.fi
Received December 17, 2007, accepted April 18, 2008
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionAcknowledgementReferences |
|---|
Background: Changes in employment status may be associated with changes in health-related lifestyle, but population level research of such associations is very limited. This study aimed to determine associations between lifestyle and five employment trajectories, i.e. ‘stable’, ‘unstable’, ‘upward’ ‘downward’ and ‘chronic unemployment’. Methods: A cohort of 10 100 employees was followed up for 5 years. Associations of the employment trajectories with changes in smoking, alcohol drinking, body weight, physical activity and sleep duration were assessed with analysis of variance for repeated measures and pairwise post hoc comparisons. Results: Smoking was the only lifestyle component that was not associated with employment trajectory. In both genders, sleep duration decreased during chronic unemployment and among those on a downward employment trajectory. In men, alcohol consumption also increased in these two groups and body weight increased in the latter group. In women, physical activity decreased among those on a downward trajectory. In contrast, an upward labour market trajectory was associated with healthy or no changes in lifestyle both in men and women. Conclusion: Changes in lifestyle may contribute to development of the health gradients between the employed and unemployed, whereas unstable employment versus permanent employment does not incur risk of unhealthy lifestyle changes. In order to prevent widening of employment-related health inequalities, passages into employment should be facilitated and opportunities for health promotion should be improved among those trapped in or moving towards the labour market periphery.
Keywords: alcohol drinking, career mobility, sleep, smoking, unemployment
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAcknowledgementReferences |
|---|
The conventional dichotomous classification into employed and unemployed seems not to be sensitive enough to explain health gradients in the modern labour force. Rather, the major gradients on the labour market core-periphery axis are located within the employed and within the unemployed work force.1 As non-standard and irregular employment is related to poorer health, there is a reason to ask whether such employment also means non-standard and irregular, or more risky, lifestyle.
Risky lifestyle contributes to poor health and excess mortality among the unemployed.2–4 Unemployment is associated with smoking and heavy alcohol consumption,5–12 obesity7,8,13–15 and sleeping problems.16 Some studies suggest that loss of employment is associated with decreased rather than increased alcohol consumption and smoking.8,13,17
The lifestyle of non-permanent employees has been paid little attention. A study of hospital employees found no associations between career direction and changes in smoking, drinking, body mass index and physical activity.18 However, fixed-term employment has been associated with deaths from alcohol-related causes in both sexes and smoking-related cancer in men.4
To date, we are aware of no population-based longitudinal studies on lifestyle changes in relation to labour market status. The existing studies on lifestyle with respect to unemployment are mainly limited to men8,13 and with respect to non-permanent job mainly on female fixed-term employees.18 In addition, previous research has rarely included potentially important details on the characteristics of the unemployment, such as the duration and recurrence of unemployment and the level and conditions of unemployment benefits. The Health and Social Support (HeSSup) Study provided us an opportunity to address the career in the labour market as a potential determinant of health-related lifestyle in a contemporary population-based cohort. We are expected to find associations between labour market trajectories and changes in smoking, alcohol consumption, physical activity, body weight and sleep duration during a 5-year follow-up.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAcknowledgementReferences |
|---|
The HeSSup Study is a longitudinal study of a sample representative of the Finnish population of four age groups (20–24, 30–34, 40–44 and 50–54 years). The Time 1 postal survey in 1998 yielded, with a response rate of 40.0%, 25 901 participants. By the follow-up survey 5 years later (Time 2) 216 participants had died, 234 had moved abroad and 969 could not be reached due to unknown addresses. Thus, the follow-up questionnaire was sent to 24 482 people, of which 19 269 (80.2%) responded. The respondents at Time 1 represented Finnish population in terms of age and gender.19 The spectrum of employment statuses among respondents corresponded to the proportions of the respective groups among the Finnish labour force.20 Taking into account the high participation at Time 2, the cohort described below is likely to represent the active labour force and its trajectories on the national labour market.
We excluded economically inactive population, i.e. those who were not working or seeking a job at Time 2 (n = 3806) or Time 1 (n = 5363). The remaining cohort of 10 100 participants was first classified into three groups both at Time 1 and Time 2: Group 1, permanent and fixed-term employees; Group 2, atypical employees, unemployed people receiving income-related daily unemployment benefit and those participating in re-employment programs; and Group 3, respondents receiving the basic unemployment benefit. The classification was based on previous findings of health gradients in the labour market core-periphery axis: poor health is least common in the core work force, or Group 1, and most common in the farthest periphery, or Group 3, while Group 2 is situated in between the two with respect to health as well (for a detailed definition of the labour market status, see Ref. 1). The employment trajectory was defined as ‘stable’ (respondents belonging at both times to Group 1, n = 8235), ‘unstable’ (those belonging at both times to Group 2, n = 59), ‘chronic unemployment’ (those belonging at both times to Group 3, n = 160), ‘upward’ (respondents who moved towards the core, n = 579) or ‘downward’ (those who moved towards the periphery, n = 867).
Health-related lifestyle
The respondents reported their habitual frequency and amount of beer, wine and spirits consumed per week. This information was transformed into grams of absolute alcohol.21,22
Intensity of physical activity during leisure and the amount of time (hours per week) spent on activity of each intensity were used to calculate a weighted sum of weekly physical activity energy expenditure [Metabolic Equivalent Task (MET) hours per week].23
The number of cigarettes smoked daily was elicited with the question ‘How many cigarettes do you smoke a day on average?’ (eight response categories, 1 = 0, 2 = 0–5, ..., 8 = 36 or more).
In statistical analyses the categories were given respective weights of 0, 2.5, 7.5, 12.5, 17.5, 22.5, 30 and 40. Sixty-three percent of the respondents were non-smokers at both time points. They were not included in the analysis of change in smoking intensity.
Sleep duration was measured with the question ‘How many hours do you normally sleep during the day and night?’ (1 = 6 h or less, 2 = 6.5 h, ..., 9 = 10 h or more).24
In addition, we used self reported body weight as a proxy for eating habits.
To obtain measures of change in lifestyle, the value of the lifestyle variables at Time 1 was subtracted from the value at Time 2.
Statistical analysis
Lifestyle at baseline and changes in lifestyle during the follow-up by employment trajectory were studied with general linear models (GLM) with and without age group as a covariate. Base-line level of outcome was not adjusted for in the analyses of change-score, as this strategy is more likely to provide unbiased causal effect estimates than baseline-adjusted estimates.25 If statistically significant, the interaction between age and employment trajectory was included in the model. In pairwise comparisons the Tukey–Kramer adjustment was used to control for Type I error. Men and women were studied separately. All statistical analyses were conducted with SAS 9.1.3 for Windows.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAcknowledgementReferences |
|---|
Descriptive statistics of the cohort are presented in table 1. There were slightly more women (56%) than men and the majority (62%) of the participants were older than 40 years. No major differences were seen in the employment trajectories between men and women. From Time 1 to Time 2, smoking intensity decreased, body weight increased, physical activity decreased and sleep duration decreased slightly both in men and in women, while alcohol intake increased in men and remained unchanged in women.
|
There were differences in the lifestyles at baseline between employment trajectories (table 2). In men, those facing chronic unemployment scored highest on smoking and alcohol consumption, those facing an upward career had highest Body Mass Index (BMI), and stable trajectory was associated to the shortest sleep duration. The only variable not associated with employment trajectory was physical activity. Among women, stable employment was associated with the shortest sleep duration. Chronic unemployment predicted smoking and high BMI, but there were no differences in alcohol intake between the employment trajectories.
|
Lifestyle changes are presented in table 3. Among men, differences by employment trajectory were seen in alcohol consumption, body weight and sleep duration, and among women in physical activity and sleep duration. According to the pairwise comparisons in men (table 4) the upward trajectory, compared to stable trajectory, was associated with decreased alcohol intake whereas the downward trajectory and chronic unemployment were associated with increased alcohol intake. The increase of body weight was slower for the upward than the downward trajectory. There was an increase in sleep duration among those on the upward trajectory whereas stable employment, chronic unemployment and the downward trajectory were associated with decrease in sleep duration.
|
|
In women (table 4) the pairwise comparisons revealed that those on the downward employment trajectory had a greater decline in physical activity than those on the stable trajectory. The pattern of changes in sleep was similar to that in men, i.e. the upward career stood out as the one with increased sleep duration while decreasing sleep duration was associated with the downward employment trajectory and chronic unemployment.
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAcknowledgementReferences |
|---|
This is apparently the first population-based prospective study to examine lifestyle changes among employees with different trajectories on the labour market. We found that being in or moving towards the labour market periphery was associated with increased risk in all lifestyle components studied except smoking. In contrast, an upward labour market trajectory was associated with healthy lifestyle changes, with these associations being more pronounced among men than women.
The figures for alcohol consumption concur with an earlier finding of contrasting patterns between genders.26 In women the level of and change in alcohol consumption was relatively independent of labour market trajectory. In men, there was an overall increasing trend confirming the findings of previous population-based studies.27 The only exception to this was those on an upward employment trajectory, as they had a decreasing figure in alcohol use. Marginalization in the labour market was strongly associated with increasing drinking. This is in contrast to two earlier studies reporting decline in alcohol consumption among marginalizing British men13 and Finnish construction workers.8 Differences in social security and time frame between these studies and the present investigation could have contributed to inconsistent findings. Furthermore, both previous studies targeted non-population-based samples.
Excess bodyweight is a major public health concern, contributing to the overall burden of disease worldwide.28 Height does not substantially change in the age groups assessed in this study; therefore we used simple body weight rather than body mass index as a proxy of change in eating habits and other factors affecting weight. The reported weight gain of 
3 kg during the 5-year follow-up is an expected finding in light of other population studies.29 Our results show that the increase in body weight does not slow down during unemployment; on the contrary, the downward labour market trajectory entails a risk of weight gain among men.
The prevalence of physical inactivity is high in contemporary populations, but there is no agreement about its trends.30 This study showed an increase in sedentary life style during the 5-year follow-up across the labour market trajectories. Although increased leisure during unemployment would give opportunities for physical activity, this opportunity was not the case. In contrast, women those on a downward labour market trajectory actually showed larger reductions in physical activity than those with stable employment.
Sleep duration in the population seems to be declining. In the United States, for instance, the self-reported sleep duration has been estimated to be about 8 h in 1960s, while more recent studies have yielded estimates about 7 h.31 The change is important from the public-health perspective, as short self-reported sleep duration predicts increased mortality and ill-health.32,33 Our results may be interpreted as follows: the slight decrease in sleep duration in the cohort reflects the population level trend, and the contrasting patterns for upward and downward career trajectories reflect the sensitivity of sleep to changes in working life. Among women, in particular, instability and marginalization from work were associated with risks for reduced sleep.
In agreement with other Finnish population-based studies,34 our results showed a decrease in smoking. The changes of smoking intensity by employment trajectory did not differ significantly. In the same vein, there were no differences in quitting of smoking between trajectories (figures of the regression analyses not shown). However, an indication of increased smoking in women, chronically unemployed trajectory, gives a reason to recommend further studies on this issue, in particular because those women were already the heaviest smokers at baseline.
Economically inactive respondents were excluded from this study. The reasons for inactivity were manifold, including e.g. from sabbatical, studying, maternity and parent leave, sick leave and to permanent retirement due to disability. Part of the inactivity may be ‘forced’, i.e. a hidden form of unemployment. In all, however, these people could not be treated as a uniform group of ‘flexible’ labour force with a uniform lifestyle. However, changes in lifestyle while entering into and out of the work force would be worth to study in the future.
As indicated in the descriptive statistics (table 1), the participation of women in work force is quite high in Finland. Thus, from the public-health perspective, the findings of women may be considered equally important as those for men.
The analyses were adjusted for age but not for socioeconomic position. This does not mean that the latter was irrelevant. On the contrary, employment trajectories are likely to belong to the mechanisms linking socioeconomic position with health outcomes. This is well illustrated in our data, where only 5% of women and 6%of men with university degree belonged to the cohorts with adverse trajectories (unstable, downward or ‘chronic unemployment’) whereas the corresponding figures for those with no vocational education were 20 and 16%, respectively. Thus, we feel that adjustment for socioeconomic position would have presented over-controlling.
Although our findings of lifestyle changes were neither universal, as regards the lifestyle components, nor uniform, as regards the employment trajectories and sexes, they show more favourable changes in lifestyle among those on upward labour market trajectory and a deteriorating lifestyle among those on downward trajectories. Regarding to the socioeconomic correlates of the trajectories, the results justify two major conclusions. First, unstable trajectories include, by definition, unemployment with relatively high income and gainful, even if atypical, employment. This kind of instability seems not to incur risks of unhealthy behaviour. Second, chronic unemployment indicated a minimum level fixed basic allowance paid out by the national unemployment insurance scheme. These participants were in the ‘hard core’ of unemployment, characterized by several years of unemployment experiences, poor prospects for re-employment and poverty. Also part of those on downward labour market trajectory received this basic allowance at follow-up. It is obvious that the socioeconomic environments and the structures of day-to-day life offer them very rarely resources for changing their lifestyle healthier.
The findings may furthermore be converted into policy implications. By denominating the measured variables as ‘lifestyle’, we aimed to emphasize their duality: although an individual is relatively free to choose how to act, there are strong structural determinants in the society that limit the choices.35 Our recommendations to policymakers are also two-fold: in addition to structural reforms facilitating passages towards the core of the labour market, there is a need to develop services in order to increase the opportunities for individual level health promotion among the long-term unemployed and those experiencing marginalization towards the labour market periphery.
|
Acknowledgement |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAcknowledgementReferences |
|---|
J.V. and M.K. were supported by the Academy of Finland (grants 117604, 124271 and 124322).
Conflicts of interest: None declared.
Key points
|
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAcknowledgementReferences |
|---|
1 Virtanen P, Liukkonen V, Vahtera J, et al. Health inequalities in the workforce: the labour market core-periphery structure. Int J Epidemiol (2003) 32:1015–21.
2 Martikainen P, Valkonen T. Excess mortality of unemployed men and women during period of rapidly increasing unemployment. Lancet (1996) 348:909–12.[CrossRef][Web of Science][Medline]
3 Morrell S, Taylor R, Quine S, et al. A case-control study of employment status and mortality in a cohort of Australian youth. Soc Sci Med (1999) 49:383–92.[CrossRef][Web of Science][Medline]
4 Kivimäki M, Vahtera J, Virtanen M, et al. Temporary employment and risk of overall and cause-specific mortality. Am J Epidemiol (2003) 158:663–8.
5 Hammarström A, Janlert U. Early unemployment can contribute to adult health problems: results from a longitudinal study of school leavers. J Epidemiol Community Health (2002) 56:624–30.
6 Reine I, Novo M, Hammarstrom A. Does the association between ill health and unemployment differ between young people and adults? Results from a 14-year follow-up study with a focus on psychological health and smoking. Public Health (2004) 118:337–45.[CrossRef][Web of Science][Medline]
7 Kivimäki M, Kinnunen M, Pitkanen T, et al. Contribution of early and adult factors to socioeconomic variation in blood pressure: thirty-four-year follow-up study of school children. Psychosom Med 2004;66:184-9. Erratum in:. Psychosom Med (2004) 66:3b.
8 Leino-Arjas P, Liira J, Mutanen P, et al. Predictors and consequences of unemployment among construction workers: prospective cohort study. Br Med J (1999) 319:600–5.
9 Morrell S, Taylor R, Kerr CB. Jobless. Unemployment and young people's health. Med J Aust (1998) 168:236.–40. Erratum in: Med J Aust 1998;168:412.[Web of Science][Medline]
10 Kaprio J, Koskenvuo M. A prospective study of psychological and socioeconomic characteristics, health behavior and morbidity in cigarette smokers prior to quitting compared to persistent smokers and non-smokers. J Clin Epidemiol (1988) 41:139–50.[CrossRef][Web of Science][Medline]
11 Montgomery S, Cook D, Bartley M, et al. Unemployment, cigarette smoking, alcohol consumption and body weight in young British men. Eur J Public Health (1998) 8:21–7.
12 Khlat M, Sermet C, Le Pape A. Increased prevalence of depression, smoking, heavy drinking and use of psycho-active drugs among unemployed men in France. Eur J Epidemiol (2004) 19:445–51.[CrossRef][Web of Science][Medline]
13 Morris JK, Cook DG, Shaper AG. Non-employment and changes in smoking, drinking, and body weight. Br Med J (1992) 304:536–41.
14 Sarlio-Lähteenkorva S, Silventoinen K, Lahti-Koski M, et al. Socio-economic status and abdominal obesity among Finnish adults from 1992 to 2002. Int J Obes (2006) 30:1653–60.[CrossRef][Web of Science][Medline]
15 Laitinen J, Power C, Ek E, et al. Unemployment and obesity among young adults in a northern Finland 1966 birth cohort. Int J Obes Relat Metab Disord (2002) 26:1329–38.[CrossRef][Web of Science][Medline]
16 Paine SJ, Gander PH, Harris R, et al. Who reports insomnia? Relationships with age, sex, ethnicity, and socioeconomic deprivation. Sleep (2004) 27:1163–9.[Web of Science][Medline]
17 Fagan P, Shavers V, Lawrence D, et al. Cigarette smoking and quitting behaviors among unemployed adults in the United States. Nicotine Tob Res (2007) 9:241–8.
18 Virtanen M, Kivimäki M, Elovainio M, et al. From insecure to secure employment: changes in work, health, health related behaviours, and sickness absence. Occup Environ Med (2003) 60:948–53.
19 Korkeila K, Suominen S, Ahvenainen J, et al. Non-response and related factors in a nation-wide health survey. Eur J Epidemiol (2001) 17:991–9.[CrossRef][Web of Science][Medline]
20 Statistics Finland. Työvoimatilasto 1998 (Labor force statistics 1998). (1999) Helsinki: Statistics Finland.
21 Kaprio J, Koskenvuo M, Langinvainio H, et al. Genetic influences on use and abuse of alcohol: a study of 5638 adult Finnish twin brothers. Alcohol Clin Exp Res (1987) 11:349–56.[CrossRef][Web of Science][Medline]
22 Vahtera J, Poikolainen K, Kivimäki M, et al. Alcohol intake and sickness absence: a curvilinear relation. Am J Epidemiol (2002) 156:969–76.
23 Kujala U, Kaprio J, Koskenvuo M. Modifiable risk factors as predictors of all-cause mortality: the roles of genetics and childhood environment. Am J Epidemiol (2002) 156:985–93.
24 Vahtera J, Pentti J, Helenius H, Kivimäki M. Sleep disturbances as a predictor of long-term increase in sickness absence among employees after family death or illness. Sleep (2006) 29:673–82.[Web of Science][Medline]
25 Glymour M, Weuve J, Berkman L, et al. When is baseline adjustment useful in analyses of change? An example with education and cognitive change. Am J Epidemiol (2005) 162:267–78.
26 Lahelma E, Kangas R, Manderbacka K. Drinking and unemployment: contrasting patterns among men and women. Drug Alcohol Depend (1995) 37:71–82.[CrossRef][Web of Science][Medline]
27 Koski A, Siren R, Vuori E, et al. Alcohol tax cuts and increase in alcohol-positive sudden deaths: a time series intervention analysis. Addiction (2007) 102:362–8.[CrossRef][Web of Science][Medline]
28 Haslam D, James W. Obesity. Lancet (2005) 366:1197–209.[CrossRef][Web of Science][Medline]
29 Silventoinen K, Sans S, Tolonen H, et al. Trends in obesity and energy supply in the WHO MONICA project. Int J Obes Relat Metab Disord (2004) 28:710–8.[CrossRef][Web of Science][Medline]
30 Livingstone M, Robson P, Wallace J, et al. How active are we? Levels of routine physical activity in children and adults. Proc Nutr Soc (2003) 62:681–701.[CrossRef][Web of Science][Medline]
31 Girardin J-L, Kripke D, Ancoli-Israel S. Sleep and quality of well-being. Sleep (2000) 23:1115–21.[Web of Science][Medline]
32 Cappuccio F, Stranges S, Kandala N, et al. Gender specific associations of short sleep duration with prevalent and incident hypertension: the Whitehall II Study. Hypertension (2007) 50:693–700.
33 Hublin C, Partinen M, Koskenvuo M, et al. Sleep and mortality: a population based 22-year follow-up study. Sleep (2007) 30:1245–53.[Web of Science][Medline]
34 Helakorpi S, Martelin T, Torppa J, et al. Did Finland's Tobacco Control Act of 1976 have an impact on ever smoking? An examination based on male and female cohort trends. J Epidemiol Community Health (2004) 58:649–54.
35 Giddens A. Modernity and self-identity. In: Self and society in the late modern age. (1991) Cambridge: Polity Press.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  L. A. MacDonald, A. Cohen, S. Baron, and C. M. Burchfiel MacDonald et al. Respond to "Search for Preventable Causes of Cardiovascular Disease" Am. J. Epidemiol., June 15, 2009; 169(12): 1426 - 1427. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||