Abstract

Background: To examine whether neighbourhood non-employment is associated with daily smoking after adjustment for individual characteristics, such as employment status. Methods: Cross-sectional study of a simple, random sample of 31,164 women and men aged 25–64, representative of the entire population in Sweden. Data were collected from the years 1993–2000. The individual variables included age, sex, employment status, occupation and housing tenure. Logistic regression was used in the analysis with neighbourhood non-employment rates measured at small area market statistics level. Results: There was a significant association between neighbourhood non-employment rates and daily smoking for both women and men. After adjustment for employment status and housing tenure the odds ratios of daily smoking were 1.39 (95% CI = 1.22–1.58) for women and 1.41 (95% CI = 1.23–1.61) for men living in neighbourhoods with the highest non-employment rates. The individual variables of unemployment, low occupational level and renting were associated with daily smoking. Conclusion: Neighbourhood non-employment is associated with daily smoking. Smoking prevention in primary health care should address both individuals and neighbourhoods.

Even though there has been a substantial decline in cigarette smoking during the last few decades in, for example, the USA1 and Sweden,2 smoking still remains widespread in many Western countries. In Sweden, about 19% of the population smoked in 2000–2001. Smoking causes mortality through a variety of diseases, e.g. coronary heart disease (CHD),35 one of the major causes of death in Western countries. The well-known association between CHD and low socio-economic status (SES)610 could partly be explained by a higher prevalence of CHD risk factors among people with low SES.1113 For example, the risk of smoking is higher among individuals with low SES, measured as educational status.6

The cigarette epidemic in many industrialised countries has been described in four stages.14 In the first two stages the prevalence of smoking increases among both men and women and is higher among men. During these initial stages of the cigarette epidemic the upper social classes have a slightly higher smoking prevalence. In the last two stages the smoking prevalence declines among both men and women and particularly among the more highly educated as they respond more favourably to health promotion campaigns. The current approach in health promotion is focused on individuals, and this approach has been somewhat disappointing both in the USA1517 and in Sweden.18 However, during the last decade a growing number of studies have examined the impact of neighbourhood characteristics on different health outcomes. Some of these studies have shown an association between neighbourhood characteristics and CHD risk factors,1921 such as smoking. A study from Sweden showed that living in a deprived neighbourhood is associated with an increased risk of smoking.22

In this study we examined the association between neighbourhood characteristics and daily smoking in a simple, random sample, representative of the whole Swedish adult population. Women and men from all socio-economic groups were included in the study, which improves the possibilities of generalising our findings.

The first aim was to examine whether there is an association between neighbourhood non-employment and the risk of daily smoking among women and men. The second aim was to examine whether this association remains after adjustment for individual characteristics, such as employment status.

Methods

Study population and data sources

We based our analysis on a simple, random sample of 31,164 women and men, aged 25–64, representative of the entire Swedish population. The sample was drawn from the Swedish Annual Level of Living Survey (SALLS), a survey conducted by Statistics Sweden since 1974. The survey includes several questions concerning socio-economic and demographic conditions and lifestyle factors, e.g. smoking habits. Participants were interviewed between 1993 and 2000 face-to-face by well-trained interviewers. The response rate was approximately 80% on average during the studied years. We used small area market statistics (SAMS) in order to define neighbourhoods. SAMS are small geographic units with boundaries defined by homogeneous types of buildings. Each SAMS consists of approximately 2000 people in Stockholm and 1000 people in the rest of Sweden. The whole of Sweden consists of 9230 SAMS. We were able to identify the SAMS neighbourhood in which the participants lived because the home addresses of the participants had been previously geocoded. Participants with missing geocodes (1.1%) were excluded. In the final sample 30,826 participants in 6817 SAMS remained for analysis.

Dependent variable

Smoking habits was based on the following question in SALLS: ‘Do you smoke daily?’ Those who answered ‘yes’ were considered to be smokers and those who answered ‘no’ were considered to be non-smokers.

Independent variables

Neighbourhood non-employment

Employment status for the entire Swedish population was measured in one week in November 1997. Those who were salaried for at least one hour during the measurement week were counted as employed. All others were counted as non-employed, which means that those with early retirement pension, housewives and some students were also included in this group. Data used to calculate neighbourhood non-employment were obtained from a national database for the entire Swedish adult population, with annual data for each individual on, for example, socio-economic characteristics. To calculate neighbourhood non-employment we used data for women and men aged 25–64. The proportions of non-employed people were calculated for each neighbourhood. Neighbourhood non-employment rates were calculated and then linked to the SAMS neighbourhoods where the participants lived by the use of geocodes. The distribution was then divided into quintiles. Quintile 1 represented neighbourhoods with the lowest proportion of non-employed people and quintile 5 represented neighbourhoods with the highest proportion of non-employed people.

Age was used as a categorical variable and divided into the following groups: (i) 25–34, (ii) 35–44, (iii) 45–54 and (iv) 55–64.

Gender was categorised as female or male.

Employment status comprised two groups: (i) employed and (ii) unemployed.

Socio-economic status (SES) was operationalised as occupation and housing tenure.

Occupation was categorised into four groups: (i) middle level employees and professionals, (ii) self-employed and farmers, (iii) lower level employees and (iv) manual workers. Highly structured questions about position in work and occupationally related activities were asked in order to capture occupational level.

Housing tenure was divided into two groups: (i) ownership and (ii) renting.

Data analysis

The SAS software package was used in the statistical analyses.23 Population data were obtained from SALLS for the years of the data collection. Prevalence rates (per 100 persons) of daily smoking were obtained by calculating the mean prevalence during the study years of 1993–2000. A logistic regression model24 was used to estimate the odds ratios (ORs) of daily smoking in the different variables. Separate models were calculated for women and men. The results are shown as OR with 95% confidence intervals (CI). Significant interactions (P < 0.05) between occupation and age were found. Therefore ORs were calculated for all occupational groups by the different age categories.

Ethics

The ethics committee at Huddinge University Hospital, Karolinska Institutet, Stockholm, has approved this study (registration number 12/00 with an additional registration April 4, 2002).

Results

Table 1 shows the sample by neighbourhood non-employment rates (in quintiles) and the different explanatory variables. In quintile 1 and quintile 5 the mean non-employment rates were 14.5% and 40.1%, respectively. There were small differences in the age distribution in the five quintiles. In the neighbourhoods with the lowest non-employment rates (quintile 1), the highest proportions of middle level employees and professionals (43.1%) and the lowest proportions of manual workers (29.3%) were found. In quintile 1 and quintile 5, 10.1% and 59.0% rented their home, respectively.

Table 1

Distribution of the sample, in percentages, by neighbourhood non-employment rates (quintiles 1–5) and the individual variables (women and men, aged 25–64, 1993–2000, Sweden; n = 30,826)


 
Quintile 1 (lowest rates)
 
Quintile 2
 
Quintile 3
 
Quintile 4
 
Quintile 5 (highest rates)
 
Neighbourhood non-employment rates      
Mean (range) % 14.5 (0–17.5) 19.6 (>17.5–21.4) 23.2 (>21.4–25.2) 27.6 (>25.2–30.7) 40.1 (>30.7–100) 
Sample sizes n = 6053 n = 5859 n = 6537 n = 6449 n = 5928 
 
Age      
    25–34 19.9 24.0 27.6 30.1 32.8 
    35–44 29.0 26.6 26.0 24.4 24.7 
    45–54 32.0 29.2 26.3 25.5 24.3 
    55–64 19.1 20.1 20.2 20.0 18.1 
Sex      
    Women 51.3 50.4 50.2 50.9 49.3 
    Men 48.7 49.6 49.8 49.0 50.7 
Employment status      
    Employed 89.0 85.5 83.3 80.2 70.0 
    Unemployed 11.0 14.5 16.7 19.8 30.0 
Occupation      
    Middle level employees and professionals 43.1 34.5 32.3 30.1 23.7 
    Self-employed and farmers 12.5 13.5 15.4 14.3 16.3 
    Lower level employees 15.1 14.5 13.2 12.8 12.0 
    Manual workers 29.3 37.5 39.1 42.8 48.0 
Housing tenure      
    Ownership 89.9 80.2 70.0 61.6 41.0 
    Renting 10.1 19.8 30.0 38.4 59.0 

 
Quintile 1 (lowest rates)
 
Quintile 2
 
Quintile 3
 
Quintile 4
 
Quintile 5 (highest rates)
 
Neighbourhood non-employment rates      
Mean (range) % 14.5 (0–17.5) 19.6 (>17.5–21.4) 23.2 (>21.4–25.2) 27.6 (>25.2–30.7) 40.1 (>30.7–100) 
Sample sizes n = 6053 n = 5859 n = 6537 n = 6449 n = 5928 
 
Age      
    25–34 19.9 24.0 27.6 30.1 32.8 
    35–44 29.0 26.6 26.0 24.4 24.7 
    45–54 32.0 29.2 26.3 25.5 24.3 
    55–64 19.1 20.1 20.2 20.0 18.1 
Sex      
    Women 51.3 50.4 50.2 50.9 49.3 
    Men 48.7 49.6 49.8 49.0 50.7 
Employment status      
    Employed 89.0 85.5 83.3 80.2 70.0 
    Unemployed 11.0 14.5 16.7 19.8 30.0 
Occupation      
    Middle level employees and professionals 43.1 34.5 32.3 30.1 23.7 
    Self-employed and farmers 12.5 13.5 15.4 14.3 16.3 
    Lower level employees 15.1 14.5 13.2 12.8 12.0 
    Manual workers 29.3 37.5 39.1 42.8 48.0 
Housing tenure      
    Ownership 89.9 80.2 70.0 61.6 41.0 
    Renting 10.1 19.8 30.0 38.4 59.0 

Table 2 shows the prevalence of smoking, defined as daily smoking, in the different quintiles and by the individual variables. The youngest age group (25–34) had the lowest prevalence of smoking. Women had a higher prevalence of smoking than men. Unemployed people had a higher prevalence of smoking than people who were employed. Manual workers had the highest and middle level employees and professionals the lowest prevalence of smoking. Those who rented their home had a higher prevalence of smoking than those who owned their home. There was an apparent gradient by neighbourhood non-employment rates: when non-employment rates increased, prevalence rates of smoking increased. The highest prevalence of smoking was found among manual workers, unemployed people and people aged 45–54 in quintile 5.

Table 2

Prevalence of daily smoking, in percentages, by neighbourhood non-employment rates (quintiles 1–5) and the individual variables (women and men, aged 25–64, 1993–2000, Sweden; n = 30,826)


 
Quintile 1 (lowest rates)
 
Quintile 2
 
Quintile 3
 
Quintile 4
 
Quintile 5 (highest rates)
 
Total 18.5 21.0 23.3 25.4 30.6 
Age      
    25–34 15.3 17.8 17.5 19.7 24.4 
    35–44 18.3 21.7 24.6 27.4 33.3 
    45–54 20.9 24.4 27.7 29.8 36.9 
    55–64 18.1 18.8 23.9 26.0 29.6 
Sex      
    Women 20.2 22.9 25.6 28.1 33.1 
    Men 16.7 19.1 21.0 22.6 28.2 
Employment status      
    Employed 18.0 19.9 22.2 23.9 28.9 
    Unemployed 22.8 27.2 28.9 31.5 34.6 
Occupation      
    Middle level employees and professionals 14.1 13.9 17.1 17.2 18.9 
    Self-employed and farmers 19.2 24.3 22.4 25.0 28.3 
    Lower level employees 22.2 23.3 24.6 28.2 29.4 
    Manual workers 22.8 25.5 28.3 30.5 37.4 
Housing tenure      
    Ownership 17.9 19.9 20.6 22.5 24.8 
    Renting 24.3 25.3 29.7 30.1 34.6 

 
Quintile 1 (lowest rates)
 
Quintile 2
 
Quintile 3
 
Quintile 4
 
Quintile 5 (highest rates)
 
Total 18.5 21.0 23.3 25.4 30.6 
Age      
    25–34 15.3 17.8 17.5 19.7 24.4 
    35–44 18.3 21.7 24.6 27.4 33.3 
    45–54 20.9 24.4 27.7 29.8 36.9 
    55–64 18.1 18.8 23.9 26.0 29.6 
Sex      
    Women 20.2 22.9 25.6 28.1 33.1 
    Men 16.7 19.1 21.0 22.6 28.2 
Employment status      
    Employed 18.0 19.9 22.2 23.9 28.9 
    Unemployed 22.8 27.2 28.9 31.5 34.6 
Occupation      
    Middle level employees and professionals 14.1 13.9 17.1 17.2 18.9 
    Self-employed and farmers 19.2 24.3 22.4 25.0 28.3 
    Lower level employees 22.2 23.3 24.6 28.2 29.4 
    Manual workers 22.8 25.5 28.3 30.5 37.4 
Housing tenure      
    Ownership 17.9 19.9 20.6 22.5 24.8 
    Renting 24.3 25.3 29.7 30.1 34.6 

Table 3 shows men and women and their relative risks for daily smoking, expressed as ORs with 95% CI. Model 1 is crude and Model 2 is adjusted for neighbourhood non-employment and the individual variables of employment status and housing tenure. For both men and women there was an apparent gradient: when neighbourhood non-employment rates increased, the risk of daily smoking increased. The increased risk of daily smoking remained significant for both sexes in quintiles 2–5 after adjustment for employment status and housing tenure. For example, the ORs of daily smoking were 1.41 (CI 1.23–1.61) for men and 1.39 (CI 1.22–1.58) for women in quintile 5 (highest non-employment rates). For both sexes, unemployment, low occupational level and renting were associated with daily smoking.

Table 3

Odds ratios (OR) for daily smoking with 95% confidence intervals (CI) by sex and the independent variables (men and women, aged 25–64, 1993–2000, Sweden; n = 30,826; logistic regression)

Variable Level Men: OR (95% CI)
 
 Women: OR (95% CI)
 
 

 

 
Model 1a
 
Model 2a
 
Model 1a
 
Model 2a
 
Neighbourhood non-employment Quintile 1 (Low) Reference Reference Reference Reference 
 Quintile 2 1.18 (1.03–1.35) 1.09 (0.95–1.25) 1.17 (1.04–1.32) 1.07 (0.94–1.21) 
 Quintile 3 1.33(1.17–1.51) 1.17 (1.02–1.33) 1.35 (1.21–1.53) 1.19 (1.06–1.35) 
 Quintile 4 1.46 (1.29–1.66) 1.23 (1.07–1.40) 1.54 (1.37–1.73) 1.28 (1.13–1.44) 
 Quintile 5 (High) 1.96 (1.73–2.22) 1.41 (1.23–1.61) 1.95 (1.73–2.19) 1.39 (1.22–1.58) 
Age 25–34 0.59 (0.53–0.67) – 1.02 (0.96–1.14) – 
 35–44 0.92 (0.82–1.03) – 1.27 (1.14–1.41) – 
 45–54 1.11 (0.99–1.24) – 1.40 (1.26–1.54) – 
 55–64 Reference – Reference – 
Employment status Employed Reference Reference Reference Reference 
 Unemployed 1.92 (1.74–2.11) 1.53 (1.37–1.70) 1.23 (1.13–1.34) 1.08 (0.98–1.19) 
Occupation Middle level employees and professionals Reference – Reference – 
 Self-employed and farmers 1.67 (1.49–1.88) – 1.69 (1.49–1.93) – 
 Lower level employees 1.50 (1.30–1.74) – 1.88 (1.68–2.10) – 
 Manual workers 2.02 (1.84–2.22) – 2.41 (2.20–2.64) – 
Housing tenure Ownership Reference Reference Reference Reference 
 Renting 1.68 (1.55–1.82) 1.68 (1.53–1.84) 1.75 (1.62–1.88) 1.66 (1.53–1.81) 
Age × occupation   See table 4  See table 4 
Variable Level Men: OR (95% CI)
 
 Women: OR (95% CI)
 
 

 

 
Model 1a
 
Model 2a
 
Model 1a
 
Model 2a
 
Neighbourhood non-employment Quintile 1 (Low) Reference Reference Reference Reference 
 Quintile 2 1.18 (1.03–1.35) 1.09 (0.95–1.25) 1.17 (1.04–1.32) 1.07 (0.94–1.21) 
 Quintile 3 1.33(1.17–1.51) 1.17 (1.02–1.33) 1.35 (1.21–1.53) 1.19 (1.06–1.35) 
 Quintile 4 1.46 (1.29–1.66) 1.23 (1.07–1.40) 1.54 (1.37–1.73) 1.28 (1.13–1.44) 
 Quintile 5 (High) 1.96 (1.73–2.22) 1.41 (1.23–1.61) 1.95 (1.73–2.19) 1.39 (1.22–1.58) 
Age 25–34 0.59 (0.53–0.67) – 1.02 (0.96–1.14) – 
 35–44 0.92 (0.82–1.03) – 1.27 (1.14–1.41) – 
 45–54 1.11 (0.99–1.24) – 1.40 (1.26–1.54) – 
 55–64 Reference – Reference – 
Employment status Employed Reference Reference Reference Reference 
 Unemployed 1.92 (1.74–2.11) 1.53 (1.37–1.70) 1.23 (1.13–1.34) 1.08 (0.98–1.19) 
Occupation Middle level employees and professionals Reference – Reference – 
 Self-employed and farmers 1.67 (1.49–1.88) – 1.69 (1.49–1.93) – 
 Lower level employees 1.50 (1.30–1.74) – 1.88 (1.68–2.10) – 
 Manual workers 2.02 (1.84–2.22) – 2.41 (2.20–2.64) – 
Housing tenure Ownership Reference Reference Reference Reference 
 Renting 1.68 (1.55–1.82) 1.68 (1.53–1.84) 1.75 (1.62–1.88) 1.66 (1.53–1.81) 
Age × occupation   See table 4  See table 4 

a: Model 1 is crude, Model 2 is adjusted for neighbourhood non-employment, individual employment status, and housing tenure and an interaction term between age and occupation

Table 4 shows the interactions between age and occupation. There was an apparent gradient for all occupational groups: with increasing age the risk of daily smoking decreased. The highest risks of daily smoking were found among women aged 25–34 in the lowest occupational groups, i.e. manual workers and lower level employees. ORs were 3.56 (CI 2.89–4.41) and 2.59 (CI 1.99–3.36), respectively.

Table 4

Interactions between age and occupation for daily smoking, after adjustment for the individual and neighbourhood variables (men and women, aged 25–64, 1993–2000, Sweden; n = 30,826; logistic regression)

Age
 
Manual workers
 
Lower level employees
 
Self-employed and farmers
 
Middle level employees and professionals
 
Men     
    25–34 2.06 (1.70–2.49) 1.75 (1.31–2.33) 1.76 (1.39–2.23) 1 (ref) 
    35–44 2.31 (1.84–2.92) 1.34 (0.92–1.95) 1.86 (1.41–2.46) 1 (ref) 
    45–54 1.98 (1.67–2.35) 1.41 (1.08–1.85) 1.38 (1.11–1.71) 1 (ref) 
    55–64 1.30 (1.06–1.60) 1.24 (0.91–1.69) 1.28 (0.99–1.66) 1 (ref) 
Women     
    25–34 3.56 (2.89–4.41) 2.59 (1.99–3.36) 1.98 (1.52–2.58) 1 (ref) 
    35–44 2.54 (2.14–3.04) 1.82 (1.47–2.27) 1.96 (1.54–2.50) 1 (ref) 
    45–54 2.04 (1.73–2.40) 1.78 (1.47–2.17) 1.21 (0.93–1.57) 1 (ref) 
    55–64 1.64 (1.31–2.05) 1.45 (1.11–1.88) 1.29 (0.90–1.83) 1 (ref) 
Age
 
Manual workers
 
Lower level employees
 
Self-employed and farmers
 
Middle level employees and professionals
 
Men     
    25–34 2.06 (1.70–2.49) 1.75 (1.31–2.33) 1.76 (1.39–2.23) 1 (ref) 
    35–44 2.31 (1.84–2.92) 1.34 (0.92–1.95) 1.86 (1.41–2.46) 1 (ref) 
    45–54 1.98 (1.67–2.35) 1.41 (1.08–1.85) 1.38 (1.11–1.71) 1 (ref) 
    55–64 1.30 (1.06–1.60) 1.24 (0.91–1.69) 1.28 (0.99–1.66) 1 (ref) 
Women     
    25–34 3.56 (2.89–4.41) 2.59 (1.99–3.36) 1.98 (1.52–2.58) 1 (ref) 
    35–44 2.54 (2.14–3.04) 1.82 (1.47–2.27) 1.96 (1.54–2.50) 1 (ref) 
    45–54 2.04 (1.73–2.40) 1.78 (1.47–2.17) 1.21 (0.93–1.57) 1 (ref) 
    55–64 1.64 (1.31–2.05) 1.45 (1.11–1.88) 1.29 (0.90–1.83) 1 (ref) 

DISCUSSION

Our main findings showed that there is an association between neighbourhood non-employment and daily smoking among women and men. This association remained after adjustment for individual employment status and housing tenure.

Our study is in agreement with a few previous studies, demonstrating a significant association between neighbourhood characteristics and the individual's risk of smoking.2527 For example, a study from California showed that living in a deprived neighbourhood increased the individual's probability of smoking, independent of individual SES.28 In contrast, a study from southern Sweden found no association between neighbourhood characteristics and daily smoking,29 a finding in agreement with a Scottish study. However, the Scottish study revealed a significant association between neighbourhood characteristics and other CHD risk factors such as elevated cholesterol levels.30 It is possible that these inconsistencies in previous studies are due to the use of different types of neighbourhood measures; for example, the study from southern Sweden did not include neighbourhood non-employment rates, which was the case in the present study.

Even though this is a cross-sectional study, some pathways between neighbourhoods and daily smoking are possible. For example, neighbourhoods may have different social norms that may in turn differentially influence health behaviours, e.g. smoking habits. There might even be a normative pressure to smoke in deprived neighbourhoods where smoking provides a means of coping with the poor conditions in such neighbourhoods.25 Under these circumstances it is easy for tobacco advertising to undermine health promotion campaigns against smoking. This is salient, especially among people with low socio-economic status, who tend to respond less favourably to smoking cessation programmes.14 Our study showed that the highest risks of daily smoking were found among young women in the lowest occupational groups, which is alarming evidence of the strength of the social forces in society.

The level of social participation might also vary between neighbourhoods. A Swedish study showed that low social participation, measured as attendance at study circles, organisations, theatre, cinema etc., was associated with an increased risk of smoking.31 In addition, the ORs of low levels of social participation were significantly higher among the unemployed than among those who were employed on the labour market. Thus, it is possible that low levels of social participation might act as a mediator between living in neighbourhoods with high non-employment rates and increased risk of daily smoking. Another possible mechanism is that it is difficult to recruit doctors and nurses to primary health care centres located in neighbourhoods with high non-employment rates. For example, in Sweden all general practitioners ranked, in a questionnaire, the impact of seven demographic and socio-economic factors on their workload. Unemployed people were considered to have a great impact on the general practitioners' workload.32 Thus, the inhabitants in neighbourhoods with high non-employment rates would receive limited primary health care and less smoking prevention. This can further deteriorate their exposed position because smoking cessation is a very important part in both primary and secondary prevention of CHD.3335 In addition, there are a number of levels of organisation above the neighbourhood level, e.g. social, economic and political contexts of regions and countries that can affect both the neighbourhood and the individual. Moreover, the neighbourhood is only one of many contexts to which an individual belongs. For example, for some people the work context can be more important for health than the neighbourhood.36 However, the neighbourhood context may have a larger impact on individuals spending more time in that context, e.g. unemployed, housewives and those with early retirement pension, than the working population of the neighbourhood.37 Consequently, neighbourhoods, workplaces and the socio-economic structure of society can all influence the individual's health status.

Limitations and strengths

There were some limitations to our study. First, we did not use multi-level modelling because, while the number of SAMS was large for each group, there were too few participants per SAMS to allow a correct calculation of the variance measure. Traditional contextual models may in these circumstances be an adequate alternative to the more complex multilevel models when examining the association between group-level variables and individual-level variables with appropriate adjustment for residual correlation.36 Therefore, we consider that this analytical approach is appropriate, and previous studies have used a similar analytical approach.28,3840 Secondly, if the non-respondents' (approximately 20%) smoking habits differed from those of the respondents a response bias may have occurred. We examined this possible bias in one of our previous studies.41 The non-respondents consisted of approximately 70% who refused to participate, 20% who could not be located and 10% who were too ill to participate. We included both non-respondents and respondents in a proportional hazard model that adjusted for age, sex, marital status and region, with all-cause mortality as the outcome. This showed that the 70% who refused to participate had the same mortality risk as the respondents, but the other two groups had significantly higher mortality risk. Third, residual confounding probably exists because individual socio-economic status cannot be measured precisely and completely.42,43

This study also has several strengths. The key strengths of our study are the large sample size, the small area neighbourhood units and the high reliability of the survey questions. The random sample of the entire population was simple (not stratified) and included 31,164 women and men from all socio-economic groups. SAMS neighbourhoods are relatively small and include about 1000–2000 people per neighbourhood. Moreover, our neighbourhood measures were based on the entire population and were almost 100% complete. The survey questions were collected in face-to-face interviews by well-trained interviewers. The reliability was high; reinterviewing a sample of the participants (test–retest method) yielded kappa coefficients of 0.96–0.99 for smoking.44 Misclassification rates for self-reported smoking habits were also low in a previous study, which examined the cotinine levels of body fluid.45

Conclusions and policy implications

Our study showed that neighbourhood characteristics are associated with smoking habits. Interventions concerning smoking cessation are highly cost-effective46 and a reduction in risk factors, especially smoking, is shown to be three times as effective as medical and surgical treatment together in CHD mortality reduction.47 The change of lifestyle factors, e.g. smoking, has proved to be a difficult task48 and the doctors and nurses in primary care often play a central role in the detection and management of risk factors.49 An ideal smoking cessation programme should include both an individual and a neighbourhood approach.1 Implementing such a strategy in primary health care can constitute a valuable instrument in smoking prevention.

Key points

  • This study examined whether neighbourhood non-employment rates is associated with daily smoking after adjustment for individual characteristics.

  • A significant association between neighbourhood non-employment rates and daily smoking was found for both women and men, after adjustment for individual characteristics.

  • The individual variables unemployment, low occupational level, and renting were associated with daily smoking.

  • Smoking prevention should address both individuals and neighbourhoods.

The authors wish to thank Sanna Sundquist at the University of California, San Diego, for technical assistance. This work was supported by grants from the National Institutes of Health (1 R01 HL71084-01), the Swedish Council for Working Life and Social Research (2001–2373), the Swedish Research Council (K2001-27X-11651-06C), the Knut and Alice Wallenberg Foundation and the Stockholm County Council.

References

1
Marlow SP, Stoller JK. Smoking cessation.
Respir Care Dec
 
2003
;
48
:
1238
–54.
2
The National Public Health Report. Stockholm: The National Board on Health and Welfare,
2001
.
3
Khot UN, Khot MB, Bajzer CT, et al. Prevalence of conventional risk factors in patients with coronary heart disease.
J Am Med Assoc
 
2003
;
290
:
898
–904.
4
De Bacquer D, De Backer G, Ostor E, Simon J, Pyorala K. Predictive value of classical risk factors and their control in coronary patients: a follow-up of the EUROASPIRE I cohort.
Eur J Cardiovasc Prev Rehabil
 
2003
;
10
:
289
–95.
5
Morris RW, Whincup PH, Lampe FC, Walker M, Wannamethee SG, Shaper AG. Geographic variation in incidence of coronary heart disease in Britain: the contribution of established risk factors.
Heart
 
2001
;
86
:
277
–83.
6
Bucher HC, Ragland DR. Socioeconomic indicators and mortality from coronary heart disease and cancer: a 22-year follow-up of middle-aged men.
Am J Public Health
 
1995
;
85
:
1231
–6.
7
Kunst AE, Groenhof F, Andersen O, et al. Occupational class and ischemic heart disease mortality in the United States and 11 European countries.
Am J Public Health
 
1999
;
89
:
47
–53.
8
Hemingway H, Shipley M, Macfarlane P, Marmot M. Impact of socioeconomic status on coronary mortality in people with symptoms, electrocardiographic abnormalities, both or neither: the original Whitehall study 25 year follow up.
J Epidemiol Community Health
 
2000
;
54
:
510
–6.
9
Salomaa V, Niemela M, Miettinen H, et al. Relationship of socioeconomic status to the incidence and prehospital, 28-day, and 1-year mortality rates of acute coronary events in the FINMONICA myocardial infarction register study.
Circulation
 
2000
;
101
:
1913
–8.
10
Wamala SP. [Large social inequalities behind women's risk of coronary disease. Unskilled work and family strains are crucial factors].
Lakartidningen
 
2001
;
98
:
177
–81.
11
Winkleby M, Jatulis DE, Frank E, Fortmann SP. Socioeconomic status and health: how education, income, and occupation contribute to risk factors for cardiovascular disease.
Am J Public Health
 
1992
;
82
:
816
–20.
12
Dennis BH, Pajak A, Pardo B, Davis CE, Williams OD, Piotrowski W. Weight gain and its correlates in Poland between 1983 and 1993.
Int J Obes Relat Metab Disord
 
2000
;
24
:
1507
–13.
13
Luepker RV, Rosamond WD, Murphy R, et al. Socioeconomic status and coronary heart disease risk factor trends. The Minnesota Heart Survey.
Circulation
 
1993
;
88
:
2172
–9.
14
Lopez AD, Collishaw NE, Piha T. A descriptive model of the cigarette epidemic in developed countries.
Tob Control
 
1994
;
3
:
242
–7.
15
Cooper R, Cutler J, Desvigne-Nickens P, et al. Trends and disparities in coronary heart disease, stroke, and other cardiovascular diseases in the United States: findings of the national conference on cardiovascular disease prevention.
Circulation
 
2000
;
102
:
3137
–47.
16
Winkleby MA. Community-Based Interventions to Prevent Cardiovascular Disease. In: Anderson N, editor. Encyclopaedia of Health and Behavior. Thousand Oaks, CA: Sage Publications, 2004: 142–6.
17
Winkleby MA, Feldman HA, Murray DM. Joint analysis of three U.S. community intervention trials for reduction of cardiovascular disease risk.
J Clin Epidemiol
 
1997
;
50
:
645
–58.
18
Health in Sweden—Sweden's Public Health Report 2001. Stockholm, Sweden: The National Board of Health and Welfare,
2001
.
19
Yen IH, Kaplan GA. Poverty area residence and changes in physical activity level: evidence from the Alameda County Study.
Am J Public Health
 
1998
;
88
:
1709
–12.
20
Ellaway A, Anderson A, Macintyre S. Does area of residence affect body size and shape?
Int J Obes
 
1997
;
21
:
304
–8.
21
Diez-Roux AV, Nieto FJ, Caulfield L, Tyroler HA, Watson RL, Szklo M. Neighbourhood differences in diet: the Atherosclerosis Risk in Communities (ARIC) Study.
J Epidemiol Community Health
 
1999
;
53
:
55
–63.
22
Sundquist J, Malmstrom M, Johansson SE. Cardiovascular risk factors and the neighbourhood environment: a multilevel analysis.
Int J Epidemiol
 
1999
;
28
:
841
–5.
23
SAS Institute Inc. SAS/STAT User's Guide, Version 6, Fourth Edition, Volume 1. Cary, NC: SAS Institute Inc.,
1989
.
24
Kleinbaum D. Survival analysis. New York: Springer Verlag,
1996
.
25
Stead M, MacAskill S, MacKintosh AM, Reece J, Eadie D. “It's as if you're locked in”: qualitative explanations for area effects on smoking in disadvantaged communities.
Health Place
 
2001
;
7
:
333
–43.
26
Kleinschmidt I, Hills M, Elliott P. Smoking behaviour can be predicted by neighbourhood deprivation measures.
J Epidemiol Community Health
 
1995
;
49
(Suppl 2):
S72
–7.
27
Duncan C, Jones K, Moon G. Smoking and deprivation: are there neighbourhood effects?
Soc Sci Med
 
1999
;
48
:
497
–505.
28
Cubbin C, Hadden WC, Winkleby MA. Neighborhood context and cardiovascular disease risk factors: the contribution of material deprivation.
Ethn Dis
 
2001
;
11
:
687
–700.
29
Lindstrom M, Moghaddassi M, Bolin K, Lindgren B, Merlo J. Social participation, social capital and daily tobacco smoking: a population-based multilevel analysis in Malmo, Sweden.
Scand J Public Health
 
2003
;
31
:
444
–50.
30
Hart C, Ecob R, Smith GD. People, places and coronary heart disease risk factors: a multilevel analysis of the Scottish Heart Health Study archive.
Soc Sci Med
 
1997
;
45
:
893
–902.
31
Lindstrom M. Social capital and the miniaturization of community among daily and intermittent smokers: a population-based study.
Prev Med
 
2003
;
36
:
177
–84.
32
Sundquist K, Malmstrom M, Johansson SE. Neighbourhood deprivation and incidence of coronary heart disease: a multilevel study of 2.6 million women and men in Sweden.
J Epidemiol Community Health
 
2004
;
58
:
71
–7.
33
Whitlock EP, Williams SB. The primary prevention of heart disease in women through health behavior change promotion in primary care.
Womens Health Issues
 
2003
;
13
:
122
–41.
34
Hensrud DD. Clinical preventive medicine in primary care: background and practice: 1. Rationale and current preventive practices.
Mayo Clin Proc
 
2000
;
75
:
165
–72.
35
Critchley J, Capewell S. Smoking cessation for the secondary prevention of coronary heart disease.
Cochrane Database Syst Rev
 
2003
;
4
:
CD003041
.
36
Kawachi I, Berkman LF. Neighborhoods and Health. New York: Oxford University Press,
2003
.
37
van Lenthe FJ, Mackenbach JP. Neighbourhood deprivation and overweight: the GLOBE study.
Int J Obes Relat Metab Disord
 
2002
;
26
:
234
–40.
38
Winkleby MA, Cubbin C. Influence of individual and neighbourhood socioeconomic status on mortality among black, Mexican–American, and white women and men in the United States.
J Epidemiol Community Health
 
2003
;
57
:
444
–52.
39
Diez Roux AV, Merkin SS, Arnett D, et al. Neighborhood of residence and incidence of coronary heart disease.
N Engl J Med
 
2001
;
345
:
99
–106.
40
Soobader MJ, LeClere FB. Aggregation and the measurement of income inequality: effects on morbidity.
Soc Sci Med
 
1999
;
48
:
733
–44.
41
Sundquist K, Winkleby M, Ahlén H, Johansson S-E. Neighborhood socioeconomic environment and incidence of coronary heart disease: a follow-up study of 25,319 women and men in Sweden. Am J Epidemiol, 2004;159(7):655–62.
42
Kaufman JS, Cooper RS, McGee DL. Socioeconomic status and health in blacks and whites: The problem of residual confounding and the resiliency of race.
Epidemiology
 
1997
;
8
:
621
–8.
43
Braveman P, Cubbin C, Marchi K, Egerter S, Chavez G. Measuring socioeconomic status/position in studies of racial/ethnic disparities: maternal and infant health.
Public Health Rep
 
2001
;
116
:
449
–63.
44
Statistics Sweden. Re-interviews in the Survey of Living Conditions (ULF). Appendix No 12. Stockholm: Statistics Sweden,
1989
.
45
Wells AJ, English PB, Posner SF, Wagenknecht LE, Perez-Stable EJ. Misclassification rates for current smokers misclassified as nonsmokers.
Am J Public Health
 
1998
;
88
:
1503
–9.
46
Eagles CJ, Martin U. Non-pharmacological modification of cardiac risk factors: part 3. Smoking cessation and alcohol consumption.
J Clin Pharm Ther
 
1998
;
23
:
1
–9.
47
Critchley JA, Capewell S, Unal B. Life-years gained from coronary heart disease mortality reduction in Scotland: prevention or treatment?
J Clin Epidemiol
 
2003
;
56
:
583
–90.
48
Clinical reality of coronary prevention guidelines: a comparison of EUROASPIRE I and II in nine countries. EUROASPIRE I and II Group. European Action on Secondary Prevention by Intervention to Reduce Events.
Lancet
 
2001
;
357
:
995
–1001.
49
Block DE, Hutton KH, Johnson KM. Differences in tobacco assessment and intervention practices: a regional snapshot.
Prev Med
 
2000
;
30
:
282
–7.

Comments

0 Comments