- Split View
-
Views
-
Cite
Cite
Chris Power, Hilary Graham, Pernille Due, Johan Hallqvist, Inez Joung, Diana Kuh, John Lynch, The contribution of childhood and adult socioeconomic position to adult obesity and smoking behaviour: an international comparison, International Journal of Epidemiology, Volume 34, Issue 2, April 2005, Pages 335–344, https://doi.org/10.1093/ije/dyh394
- Share Icon Share
Abstract
Background Our objective was to investigate the contribution of childhood and adult socioeconomic position (SEP) to adult obesity and smoking behaviour, in particular to establish the role of childhood circumstances across different studies in Europe and the US.
Methods Seven population-based surveys in six Western countries (Britain, Denmark, Finland, Netherlands, Sweden, US) were examined, with participants aged 30–50 yr and born between 1910 and 1960. Adult smoking was analysed using three outcomes (ever, current, or ex-) and adult obesity was defined as body mass index (kg/m2) ≥30.
Results A strong effect of adult social position was observed for smoking outcomes and obesity. For example, manual SEP in adulthood increased the risk of ever smoking (adjusted odds ratio (OR) 1.47–2.00 for men; 0.94–1.81 for women), and obesity (adjusted OR 1.06–2.24 for men, 1.21–3.26 for women). In most studies, childhood position was not associated with ever-smoking. For current smoking, manual childhood position was associated among women (adjusted OR 1.09–1.54), but no consistent pattern was seen for men. For ex-smoking, manual childhood origins lowered the chance of quitting among women (adjusted OR 0.64–0.81) except in the US (OR = 1.17); among men this association was seen in fewer studies (adjusted OR 0.74–1.09). For obesity, manual origins increased the risk for women (adjusted OR 0.96–2.50); effects were weaker among men but mostly in the same direction (adjusted OR 0.79–1.42).
Conclusions As expected, adult SEP was an important influence on smoking behaviour and obesity. In addition, factors related to disadvantaged social origins appeared to increase the risk of obesity and reduce the probability of quitting smoking in adulthood, particularly in women.
It is well-established that socioeconomic position (SEP) in adulthood is associated with mortality, healthy life expectancy and several important disease risk factors.1,2 Evidence is now growing that (SEP) in childhood also contributes to some types of mortality and morbidity in adult life.3 Many so-called ‘adult’ risk factors—such as cholesterol, blood pressure, smoking, and obesity—may develop in childhood and track into adulthood. For instance, the Bogalusa Heart Study showed that 77% of obese children became obese adults with worse heart disease risk factor profiles.4 An extensive review showed that blood cholesterol, blood pressure, and body mass index (BMI) measurements taken in adolescence or early adulthood were predictive of heart disease up to 50 yr later.5 For example, cholesterol measured in childhood predicted carotid artery intima-media thickness (a marker of atherosclerosis) at least as strongly as cholesterol measured in adulthood, and the childhood measures remained predictors after adulthood measures of cholesterol were taken into account.6
The influence of childhood SEP on the development and trajectories of adult risk factors has been identified as a likely route through which early social circumstances exert an effect on adult health.3,7–9 For example, childhood SEP can affect cognitive development and educational achievement, which in turn, are major influences on health related behaviour, as well as adult SEP. However, there are few studies to date that investigate the link between childhood SEP and specific health behaviours such as adult smoking and obesity, and existing studies are predominantly from the UK.10–19 The limited range of studies in which this potentially important link has been investigated leaves open the question of whether childhood effects are specific to particular populations or whether they are found in populations born at different times and places.
We identified several studies in Europe and north America spanning cohorts born throughout the last half of the twentieth century to examine the role of both childhood and adult SEP on smoking and obesity in adult life. Smoking and adult obesity were selected because of their public health importance.20,21 Our aim was to establish whether childhood SEP influenced adult smoking and obesity and, if so, whether it acts mainly through an association with adult SEP. In addition, we examined whether the well-known association between smoking and obesity and adult SEP was largely due to circumstances in childhood. Associations between childhood SEP and adult smoking and obesity were investigated separately for men and women because previous work had suggested gender differences in these associations.12,16,19,22
Subjects and methods
Study samples
Seven population-based studies in six countries, across a range of time periods and cohorts, were included. These studies are representative of their original birth samples,23,24 national25,26 or local populations27–29 and have information on smoking and obesity in mid-life, and SEP in childhood and adulthood (Table 1). All except the British birth cohort studies used retrospective recall for childhood SEP. To ensure a focus on a comparable life stage, studies with a broad age range were restricted to 30–50 yr (to 54 yr in Finland); for the British cohort studies, one age within this range was selected.
Country/study . | Samplea . | Year of birth . | Age (yr) range in analyses . | N in studyc (n inanalyses) . | Survey instrument . | Year of baseline data collection . | Year of outcome measurement . | Additional study details . |
---|---|---|---|---|---|---|---|---|
Finland | Area-based | 1930–47 | 42–54 | 2284 men (1918) | Personal interview/clinical exam | 1984–89 | 1989 | 3433 men aged 42, 48, 54 or 60 yr eligible for the study. Exclusions because of death, serious disease or migration = 198. |
Kuopio ischaemic heart disease risk factor study27 | ||||||||
Sweden | National (annual survey) | 1930–59 | 30–50 | 9330 (9218) | Personal interview | 1980/81 and 1988/89 | 1980/81 and 1988/89 | Response rate across all ages in Survey of Living Conditions: 86% (1980/81), 78–80% (1988/89). Sub-category in ages 30–50 yr = 9330, with 112 excluded from analyses due to missing data. |
Survey of living conditions (Undersökning av levnadsförhållanden)25 | ||||||||
Denmark | National | 1950–60 | 40 and 50 | 7299b (6143) | Postal questionnaire | 2000 | 2000 | Sample from AKF Longitudinal Register at Statistics Denmark, includes 10% of adults. Total response rate: 66% (out of 11 082 adults) |
Danish Longitudinal studyon work, unemploymentand Health26 | ||||||||
Britain | National | 1946 | 36 | 5362∼ 3322 (3266) | Personal interview with mother & cohort member | 1946 | 1982 | Socially stratified sample of single, legitimatebirths in 1 week in March 1946, followed through-out childhood to adulthood. Attrition to age 36 yr:death (323), refusal (520), emigrated (644), untraced (553). Missing data = 52 |
1946 birth cohort23 | ||||||||
Britain | National | 1958 | 33 | 17 000∼ 11 405 (9350) | Personal interview with mother and cohort member | 1958 | 1991 | Baseline sample is 98% of all births in 1 week in March 1958 (∼17 000), followed through-out childhood to adulthood (11 405 at 33 yr), Attrition to 33 yr: death (315), refusal (1898), emigrated (269), untraced (1927), traced but not interviewed (529). Missing data = 2055 |
1958 birth cohort24 | ||||||||
Netherlands | Area-based (south-east) | 1942–61 | 30–49 | 5857 (5057) | Postal questionnaire | 1991 | 1991 | Sample aged 15–74 yr with over-sampling of adults >45 yr and in the lowest and highest SEP. Total response rate 70% (n = 18 973/27 070). Eligible in age group 30–49 yr (n = 5857). Exclusions: missing values (n = 800). Adjustment for overrepresentation was made by proportionally weighting data to resemble the composition of base population. |
Globe Study28 | ||||||||
US | Area-based | 1910–35 | 30–50 | 3087∼ (2727) | Postal questionnaire | 1965 | 1965 | Total response rate: 86% (6928/8083), followed to 2000. Exclusions: women ‘working in the home’ not assigned to an occupation. |
Alameda County29 |
Country/study . | Samplea . | Year of birth . | Age (yr) range in analyses . | N in studyc (n inanalyses) . | Survey instrument . | Year of baseline data collection . | Year of outcome measurement . | Additional study details . |
---|---|---|---|---|---|---|---|---|
Finland | Area-based | 1930–47 | 42–54 | 2284 men (1918) | Personal interview/clinical exam | 1984–89 | 1989 | 3433 men aged 42, 48, 54 or 60 yr eligible for the study. Exclusions because of death, serious disease or migration = 198. |
Kuopio ischaemic heart disease risk factor study27 | ||||||||
Sweden | National (annual survey) | 1930–59 | 30–50 | 9330 (9218) | Personal interview | 1980/81 and 1988/89 | 1980/81 and 1988/89 | Response rate across all ages in Survey of Living Conditions: 86% (1980/81), 78–80% (1988/89). Sub-category in ages 30–50 yr = 9330, with 112 excluded from analyses due to missing data. |
Survey of living conditions (Undersökning av levnadsförhållanden)25 | ||||||||
Denmark | National | 1950–60 | 40 and 50 | 7299b (6143) | Postal questionnaire | 2000 | 2000 | Sample from AKF Longitudinal Register at Statistics Denmark, includes 10% of adults. Total response rate: 66% (out of 11 082 adults) |
Danish Longitudinal studyon work, unemploymentand Health26 | ||||||||
Britain | National | 1946 | 36 | 5362∼ 3322 (3266) | Personal interview with mother & cohort member | 1946 | 1982 | Socially stratified sample of single, legitimatebirths in 1 week in March 1946, followed through-out childhood to adulthood. Attrition to age 36 yr:death (323), refusal (520), emigrated (644), untraced (553). Missing data = 52 |
1946 birth cohort23 | ||||||||
Britain | National | 1958 | 33 | 17 000∼ 11 405 (9350) | Personal interview with mother and cohort member | 1958 | 1991 | Baseline sample is 98% of all births in 1 week in March 1958 (∼17 000), followed through-out childhood to adulthood (11 405 at 33 yr), Attrition to 33 yr: death (315), refusal (1898), emigrated (269), untraced (1927), traced but not interviewed (529). Missing data = 2055 |
1958 birth cohort24 | ||||||||
Netherlands | Area-based (south-east) | 1942–61 | 30–49 | 5857 (5057) | Postal questionnaire | 1991 | 1991 | Sample aged 15–74 yr with over-sampling of adults >45 yr and in the lowest and highest SEP. Total response rate 70% (n = 18 973/27 070). Eligible in age group 30–49 yr (n = 5857). Exclusions: missing values (n = 800). Adjustment for overrepresentation was made by proportionally weighting data to resemble the composition of base population. |
Globe Study28 | ||||||||
US | Area-based | 1910–35 | 30–50 | 3087∼ (2727) | Postal questionnaire | 1965 | 1965 | Total response rate: 86% (6928/8083), followed to 2000. Exclusions: women ‘working in the home’ not assigned to an occupation. |
Alameda County29 |
All samples are representative, including both men and women, except Finland with men only; in all studies cross-sectional data are used, except for the two prospective British birth cohorts.
Sample at baseline.
Differences between N and n vary partly because of differences in study design: for example, cross-sectional studies do not generally include deaths or emigrations in their sampling frame, whereas these appear as attrition in birth cohort studies.
Country/study . | Samplea . | Year of birth . | Age (yr) range in analyses . | N in studyc (n inanalyses) . | Survey instrument . | Year of baseline data collection . | Year of outcome measurement . | Additional study details . |
---|---|---|---|---|---|---|---|---|
Finland | Area-based | 1930–47 | 42–54 | 2284 men (1918) | Personal interview/clinical exam | 1984–89 | 1989 | 3433 men aged 42, 48, 54 or 60 yr eligible for the study. Exclusions because of death, serious disease or migration = 198. |
Kuopio ischaemic heart disease risk factor study27 | ||||||||
Sweden | National (annual survey) | 1930–59 | 30–50 | 9330 (9218) | Personal interview | 1980/81 and 1988/89 | 1980/81 and 1988/89 | Response rate across all ages in Survey of Living Conditions: 86% (1980/81), 78–80% (1988/89). Sub-category in ages 30–50 yr = 9330, with 112 excluded from analyses due to missing data. |
Survey of living conditions (Undersökning av levnadsförhållanden)25 | ||||||||
Denmark | National | 1950–60 | 40 and 50 | 7299b (6143) | Postal questionnaire | 2000 | 2000 | Sample from AKF Longitudinal Register at Statistics Denmark, includes 10% of adults. Total response rate: 66% (out of 11 082 adults) |
Danish Longitudinal studyon work, unemploymentand Health26 | ||||||||
Britain | National | 1946 | 36 | 5362∼ 3322 (3266) | Personal interview with mother & cohort member | 1946 | 1982 | Socially stratified sample of single, legitimatebirths in 1 week in March 1946, followed through-out childhood to adulthood. Attrition to age 36 yr:death (323), refusal (520), emigrated (644), untraced (553). Missing data = 52 |
1946 birth cohort23 | ||||||||
Britain | National | 1958 | 33 | 17 000∼ 11 405 (9350) | Personal interview with mother and cohort member | 1958 | 1991 | Baseline sample is 98% of all births in 1 week in March 1958 (∼17 000), followed through-out childhood to adulthood (11 405 at 33 yr), Attrition to 33 yr: death (315), refusal (1898), emigrated (269), untraced (1927), traced but not interviewed (529). Missing data = 2055 |
1958 birth cohort24 | ||||||||
Netherlands | Area-based (south-east) | 1942–61 | 30–49 | 5857 (5057) | Postal questionnaire | 1991 | 1991 | Sample aged 15–74 yr with over-sampling of adults >45 yr and in the lowest and highest SEP. Total response rate 70% (n = 18 973/27 070). Eligible in age group 30–49 yr (n = 5857). Exclusions: missing values (n = 800). Adjustment for overrepresentation was made by proportionally weighting data to resemble the composition of base population. |
Globe Study28 | ||||||||
US | Area-based | 1910–35 | 30–50 | 3087∼ (2727) | Postal questionnaire | 1965 | 1965 | Total response rate: 86% (6928/8083), followed to 2000. Exclusions: women ‘working in the home’ not assigned to an occupation. |
Alameda County29 |
Country/study . | Samplea . | Year of birth . | Age (yr) range in analyses . | N in studyc (n inanalyses) . | Survey instrument . | Year of baseline data collection . | Year of outcome measurement . | Additional study details . |
---|---|---|---|---|---|---|---|---|
Finland | Area-based | 1930–47 | 42–54 | 2284 men (1918) | Personal interview/clinical exam | 1984–89 | 1989 | 3433 men aged 42, 48, 54 or 60 yr eligible for the study. Exclusions because of death, serious disease or migration = 198. |
Kuopio ischaemic heart disease risk factor study27 | ||||||||
Sweden | National (annual survey) | 1930–59 | 30–50 | 9330 (9218) | Personal interview | 1980/81 and 1988/89 | 1980/81 and 1988/89 | Response rate across all ages in Survey of Living Conditions: 86% (1980/81), 78–80% (1988/89). Sub-category in ages 30–50 yr = 9330, with 112 excluded from analyses due to missing data. |
Survey of living conditions (Undersökning av levnadsförhållanden)25 | ||||||||
Denmark | National | 1950–60 | 40 and 50 | 7299b (6143) | Postal questionnaire | 2000 | 2000 | Sample from AKF Longitudinal Register at Statistics Denmark, includes 10% of adults. Total response rate: 66% (out of 11 082 adults) |
Danish Longitudinal studyon work, unemploymentand Health26 | ||||||||
Britain | National | 1946 | 36 | 5362∼ 3322 (3266) | Personal interview with mother & cohort member | 1946 | 1982 | Socially stratified sample of single, legitimatebirths in 1 week in March 1946, followed through-out childhood to adulthood. Attrition to age 36 yr:death (323), refusal (520), emigrated (644), untraced (553). Missing data = 52 |
1946 birth cohort23 | ||||||||
Britain | National | 1958 | 33 | 17 000∼ 11 405 (9350) | Personal interview with mother and cohort member | 1958 | 1991 | Baseline sample is 98% of all births in 1 week in March 1958 (∼17 000), followed through-out childhood to adulthood (11 405 at 33 yr), Attrition to 33 yr: death (315), refusal (1898), emigrated (269), untraced (1927), traced but not interviewed (529). Missing data = 2055 |
1958 birth cohort24 | ||||||||
Netherlands | Area-based (south-east) | 1942–61 | 30–49 | 5857 (5057) | Postal questionnaire | 1991 | 1991 | Sample aged 15–74 yr with over-sampling of adults >45 yr and in the lowest and highest SEP. Total response rate 70% (n = 18 973/27 070). Eligible in age group 30–49 yr (n = 5857). Exclusions: missing values (n = 800). Adjustment for overrepresentation was made by proportionally weighting data to resemble the composition of base population. |
Globe Study28 | ||||||||
US | Area-based | 1910–35 | 30–50 | 3087∼ (2727) | Postal questionnaire | 1965 | 1965 | Total response rate: 86% (6928/8083), followed to 2000. Exclusions: women ‘working in the home’ not assigned to an occupation. |
Alameda County29 |
All samples are representative, including both men and women, except Finland with men only; in all studies cross-sectional data are used, except for the two prospective British birth cohorts.
Sample at baseline.
Differences between N and n vary partly because of differences in study design: for example, cross-sectional studies do not generally include deaths or emigrations in their sampling frame, whereas these appear as attrition in birth cohort studies.
Measures
Socioeconomic position (SEP)
For childhood, father's occupation was used to identify manual/non-manual backgrounds. Childhood SEP was reported retrospectively, except in the British studies. For adult SEP, occupation was used to identify manual/non-manual groups, with women in paid work classified by their own occupation and housewives by their husbands' occupation. Table 2 shows that most study participants could be allocated to a SEP in childhood and adulthood, but for a proportion there was insufficient information to allocate individuals to a manual or non-manual SEP. The latter included housewives with no data on husband's occupation (US and Denmark) and farmers (Finland, Netherlands, Sweden). In these countries, the SEP of farmers—a category which traditionally included farm workers and self-employed farmers—changed as the agricultural sector declined. Where appropriate (childhood SEP in the Finnish study), farmers were allocated to the manual group; otherwise, where they represented a significant group (Finland, Sweden, Netherlands), they were included in a separate category. As the comparative focus of our study is on manual/non-manual groups, data are not presented for farmers in these studies. However, farmers were included in the analyses in order to avoid substantial reductions in sample sizes (see footnote for Table 2 for further details).
. | Males . | . | . | . | Females . | . | . | . | ||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
. | Childhood SEP . | . | Adult SEP . | . | Childhood SEP . | . | Adult SEP . | . | ||||||
Study . | Manual . | Non-manual . | Manual . | Non-manual . | Manual . | Non-manual . | Manual . | Non-manual . | ||||||
Finlandc | 72 | 28 | 44 | 39 | – | – | – | – | ||||||
Swedend | 48 | 23 | 48 | 50 | 49 | 23 | 44 | 52 | ||||||
Denmarke | 66 | 18 | 36 | 51 | 63 | 19 | 35 | 45 | ||||||
Britain, 1946f | 73 | 25 | 52 | 46 | 72 | 26 | 37 | 61 | ||||||
Britain, 1958 | 62 | 30 | 47 | 47 | 63 | 29 | 28 | 63 | ||||||
Netherlandsg | 53 | 32 | 41 | 54 | 54 | 29 | 36 | 59 | ||||||
US | 54 | 46 | 51 | 49 | 50 | 38 | 30 | 58 |
. | Males . | . | . | . | Females . | . | . | . | ||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
. | Childhood SEP . | . | Adult SEP . | . | Childhood SEP . | . | Adult SEP . | . | ||||||
Study . | Manual . | Non-manual . | Manual . | Non-manual . | Manual . | Non-manual . | Manual . | Non-manual . | ||||||
Finlandc | 72 | 28 | 44 | 39 | – | – | – | – | ||||||
Swedend | 48 | 23 | 48 | 50 | 49 | 23 | 44 | 52 | ||||||
Denmarke | 66 | 18 | 36 | 51 | 63 | 19 | 35 | 45 | ||||||
Britain, 1946f | 73 | 25 | 52 | 46 | 72 | 26 | 37 | 61 | ||||||
Britain, 1958 | 62 | 30 | 47 | 47 | 63 | 29 | 28 | 63 | ||||||
Netherlandsg | 53 | 32 | 41 | 54 | 54 | 29 | 36 | 59 | ||||||
US | 54 | 46 | 51 | 49 | 50 | 38 | 30 | 58 |
Childhood SEP (based on measures of father's occupation or vocational training: manual vs non-manual).
Adult SEP (own occupation; manual vs non-manual).
For childhood SEP, all those whose fathers were farmers were classified as manual and they are included in the analyses as such; in adulthood, farmers were not classified as manual and are omitted from the table, although they were included in analyses as a separate category.
For childhood and adult SEP, farmers were not classified as manual and are omitted from the table, although they were included in analyses as a separate category.
Housewives, persons with unclassifiable occupations, and farmers are not included in manual/non-manual groups.
Weighted to allow for original sampling procedure.
Persons with unclassifiable occupations, farmers, and self-employed people are not included in manual/non-manual groups, although they were included in Analyses as a separate category.
. | Males . | . | . | . | Females . | . | . | . | ||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
. | Childhood SEP . | . | Adult SEP . | . | Childhood SEP . | . | Adult SEP . | . | ||||||
Study . | Manual . | Non-manual . | Manual . | Non-manual . | Manual . | Non-manual . | Manual . | Non-manual . | ||||||
Finlandc | 72 | 28 | 44 | 39 | – | – | – | – | ||||||
Swedend | 48 | 23 | 48 | 50 | 49 | 23 | 44 | 52 | ||||||
Denmarke | 66 | 18 | 36 | 51 | 63 | 19 | 35 | 45 | ||||||
Britain, 1946f | 73 | 25 | 52 | 46 | 72 | 26 | 37 | 61 | ||||||
Britain, 1958 | 62 | 30 | 47 | 47 | 63 | 29 | 28 | 63 | ||||||
Netherlandsg | 53 | 32 | 41 | 54 | 54 | 29 | 36 | 59 | ||||||
US | 54 | 46 | 51 | 49 | 50 | 38 | 30 | 58 |
. | Males . | . | . | . | Females . | . | . | . | ||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
. | Childhood SEP . | . | Adult SEP . | . | Childhood SEP . | . | Adult SEP . | . | ||||||
Study . | Manual . | Non-manual . | Manual . | Non-manual . | Manual . | Non-manual . | Manual . | Non-manual . | ||||||
Finlandc | 72 | 28 | 44 | 39 | – | – | – | – | ||||||
Swedend | 48 | 23 | 48 | 50 | 49 | 23 | 44 | 52 | ||||||
Denmarke | 66 | 18 | 36 | 51 | 63 | 19 | 35 | 45 | ||||||
Britain, 1946f | 73 | 25 | 52 | 46 | 72 | 26 | 37 | 61 | ||||||
Britain, 1958 | 62 | 30 | 47 | 47 | 63 | 29 | 28 | 63 | ||||||
Netherlandsg | 53 | 32 | 41 | 54 | 54 | 29 | 36 | 59 | ||||||
US | 54 | 46 | 51 | 49 | 50 | 38 | 30 | 58 |
Childhood SEP (based on measures of father's occupation or vocational training: manual vs non-manual).
Adult SEP (own occupation; manual vs non-manual).
For childhood SEP, all those whose fathers were farmers were classified as manual and they are included in the analyses as such; in adulthood, farmers were not classified as manual and are omitted from the table, although they were included in analyses as a separate category.
For childhood and adult SEP, farmers were not classified as manual and are omitted from the table, although they were included in analyses as a separate category.
Housewives, persons with unclassifiable occupations, and farmers are not included in manual/non-manual groups.
Weighted to allow for original sampling procedure.
Persons with unclassifiable occupations, farmers, and self-employed people are not included in manual/non-manual groups, although they were included in Analyses as a separate category.
Smoking and obesity
Data on smoking were self-reported. Regular smoking was defined as smoking ≥1 cigarette/day; ex-smokers are ex-regular smokers (i.e. those who previously smoked ≥1 cigarette a day). Using this definition, we examined three smoking outcomes—ever smoking, current smoking, and ex-smoking (quitting). For obesity, the British and Finnish studies used measurements of height and weight; all other studies obtained self-reported data. Obesity was defined as BMI (kg/m2) ≥30 according to WHO recommendations.30 Analyses using overweight (BMI ≥ 25) showed similar results.
Data analysis
We first examined the separate associations of childhood and adult SEP on the three smoking measures and on obesity, separately for men and women, using logistic regression. Odds ratios (OR) and 95% confidence intervals (CIs) were calculated, age standardized where necessary. For smoking, we calculated the OR of ever smoking vs never smoking, current smokers vs others (never smokers and ex-smokers) and ex-smokers vs current smokers. Next, in multivariate analyses including both SEP measures, we obtained mutually adjusted estimates for childhood and adult SEP. Where farmers were omitted from the manual/non-manual classification (Finland, Netherlands, Sweden) they were included in multivariate analyses as a third socioeconomic category. Significant effects are indicated by 95% CI, though our primary focus is on consistency of results across studies. For studies in which farmers were allocated to a separate socioeconomic category, we undertook analyses with and without this group and found similar results (data not presented).
Results
Table 1 presents basic information for the seven studies. Most individuals came from a manual socioeconomic background (range 48–73%) (Table 2). As expected, in each study a larger proportion was in a non-manual group in adulthood (range 39–63%), showing secular changes in inter-generational social mobility and the overall decline in all these countries of manual jobs.
Smoking
The prevalence of ever smoking ranged from 44 to 82% (men) and from 45 to 75% (women). Rates of ever smoking were consistently higher in manual groups, except among US females (Table 3) but the US data were collected in 1965, prior to the peak of the US smoking epidemic among women. Age adjusted ORs in Table 4 show no consistent effect of childhood position on ever smoking, whereas there is a strong association between adult position and ever smoking, except in US women (range in adjusted ORs 1.47–2.00 for men; 0.94–1.81 for women).
. | Ever smokers . | . | . | . | Current smokers . | . | . | . | Ex-smokers (% of ever smokers) . | . | . | . | |||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
. | M . | . | F . | . | M . | . | F . | . | M . | . | F . | . | |||||||||
Study . | Manual . | Non-manual . | Manual . | Non-manual . | Manual . | Non-manual . | Manual . | Non-manual . | Manual . | Non-manual . | Manual . | Non-manual . | |||||||||
Finland | 80 | 68 | – | – | 40 | 32 | – | – | 40 | 36 | – | – | |||||||||
Sweden | 71 | 63 | 63 | 56 | 42 | 31 | 42 | 30 | 41 | 50 | 34 | 46 | |||||||||
Denmark | 72 | 62 | 66 | 56 | 48 | 33 | 43 | 30 | 33 | 46 | 34 | 46 | |||||||||
Britain, 1946b | 80 | 72 | 73 | 64 | 44 | 27 | 46 | 30 | 45 | 62 | 37 | 52 | |||||||||
Britain, 1958 | 61 | 44 | 60 | 45 | 42 | 22 | 43 | 26 | 31 | 50 | 29 | 43 | |||||||||
Netherlands | 82 | 74 | 75 | 68 | 53 | 40 | 47 | 34 | 36 | 46 | 37 | 49 | |||||||||
US | 81 | 72 | 58 | 61 | 61 | 48 | 51 | 49 | 25 | 33 | 13 | 19 |
. | Ever smokers . | . | . | . | Current smokers . | . | . | . | Ex-smokers (% of ever smokers) . | . | . | . | |||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
. | M . | . | F . | . | M . | . | F . | . | M . | . | F . | . | |||||||||
Study . | Manual . | Non-manual . | Manual . | Non-manual . | Manual . | Non-manual . | Manual . | Non-manual . | Manual . | Non-manual . | Manual . | Non-manual . | |||||||||
Finland | 80 | 68 | – | – | 40 | 32 | – | – | 40 | 36 | – | – | |||||||||
Sweden | 71 | 63 | 63 | 56 | 42 | 31 | 42 | 30 | 41 | 50 | 34 | 46 | |||||||||
Denmark | 72 | 62 | 66 | 56 | 48 | 33 | 43 | 30 | 33 | 46 | 34 | 46 | |||||||||
Britain, 1946b | 80 | 72 | 73 | 64 | 44 | 27 | 46 | 30 | 45 | 62 | 37 | 52 | |||||||||
Britain, 1958 | 61 | 44 | 60 | 45 | 42 | 22 | 43 | 26 | 31 | 50 | 29 | 43 | |||||||||
Netherlands | 82 | 74 | 75 | 68 | 53 | 40 | 47 | 34 | 36 | 46 | 37 | 49 | |||||||||
US | 81 | 72 | 58 | 61 | 61 | 48 | 51 | 49 | 25 | 33 | 13 | 19 |
Adult SEP (own occupation; manual vs non-manual)
Weighted to allow for original sampling procedure.
. | Ever smokers . | . | . | . | Current smokers . | . | . | . | Ex-smokers (% of ever smokers) . | . | . | . | |||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
. | M . | . | F . | . | M . | . | F . | . | M . | . | F . | . | |||||||||
Study . | Manual . | Non-manual . | Manual . | Non-manual . | Manual . | Non-manual . | Manual . | Non-manual . | Manual . | Non-manual . | Manual . | Non-manual . | |||||||||
Finland | 80 | 68 | – | – | 40 | 32 | – | – | 40 | 36 | – | – | |||||||||
Sweden | 71 | 63 | 63 | 56 | 42 | 31 | 42 | 30 | 41 | 50 | 34 | 46 | |||||||||
Denmark | 72 | 62 | 66 | 56 | 48 | 33 | 43 | 30 | 33 | 46 | 34 | 46 | |||||||||
Britain, 1946b | 80 | 72 | 73 | 64 | 44 | 27 | 46 | 30 | 45 | 62 | 37 | 52 | |||||||||
Britain, 1958 | 61 | 44 | 60 | 45 | 42 | 22 | 43 | 26 | 31 | 50 | 29 | 43 | |||||||||
Netherlands | 82 | 74 | 75 | 68 | 53 | 40 | 47 | 34 | 36 | 46 | 37 | 49 | |||||||||
US | 81 | 72 | 58 | 61 | 61 | 48 | 51 | 49 | 25 | 33 | 13 | 19 |
. | Ever smokers . | . | . | . | Current smokers . | . | . | . | Ex-smokers (% of ever smokers) . | . | . | . | |||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
. | M . | . | F . | . | M . | . | F . | . | M . | . | F . | . | |||||||||
Study . | Manual . | Non-manual . | Manual . | Non-manual . | Manual . | Non-manual . | Manual . | Non-manual . | Manual . | Non-manual . | Manual . | Non-manual . | |||||||||
Finland | 80 | 68 | – | – | 40 | 32 | – | – | 40 | 36 | – | – | |||||||||
Sweden | 71 | 63 | 63 | 56 | 42 | 31 | 42 | 30 | 41 | 50 | 34 | 46 | |||||||||
Denmark | 72 | 62 | 66 | 56 | 48 | 33 | 43 | 30 | 33 | 46 | 34 | 46 | |||||||||
Britain, 1946b | 80 | 72 | 73 | 64 | 44 | 27 | 46 | 30 | 45 | 62 | 37 | 52 | |||||||||
Britain, 1958 | 61 | 44 | 60 | 45 | 42 | 22 | 43 | 26 | 31 | 50 | 29 | 43 | |||||||||
Netherlands | 82 | 74 | 75 | 68 | 53 | 40 | 47 | 34 | 36 | 46 | 37 | 49 | |||||||||
US | 81 | 72 | 58 | 61 | 61 | 48 | 51 | 49 | 25 | 33 | 13 | 19 |
Adult SEP (own occupation; manual vs non-manual)
Weighted to allow for original sampling procedure.
. | . | Childhood SEP . | . | . | . | Adult SEP . | . | . | . | ||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
. | . | Ever smokers vs never smokers . | . | Current smokers vs others (never and ex smokers) . | . | Ever smokers vs never smokers . | . | Current smokers vs others (never and ex smokers) . | . | ||||||
Study . | . | Unadjusteda . | Adjustedb . | Unadjusteda . | Adjustedb . | Unadjusteda . | Adjustedc . | Unadjusteda . | Adjustedc . | ||||||
Finland | M | 0.95 (0.77–1.17) | 0.84 (0.67–1.04) | 0.89 (0.73–1.08) | 0.82 (0.67–1.00) | 1.80 (1.46–2.22) | 1.88 (1.51–2.32) | 1.43 (1.18–1.73) | 1.50 (1.23–1.82) | ||||||
Sweden | M | 1.42 (1.26–1.60) | 1.35 (1.20–1.53) | 1.20 (1.06–1.35) | 1.12 (0.99–1.26) | 1.45 (1.28–1.64) | 1.47 (1.29–1.66) | 1.59 (1.41–1.79) | 1.60 (1.41–1.81) | ||||||
F | 1.30 (1.15–1.46) | 1.26 (1.12–1.42) | 1.47 (1.30–1.67) | 1.39 (1.23–1.58) | 1.33 (1.18–1.50) | 1.33 (1.18–1.50) | 1.65 (1.46–1.87) | 1.59 (1.41–1.80) | |||||||
Denmark | M | 0.81 (0.66–0.98) | 0.73 (0.60–0.89) | 1.09 (0.92–1.32) | 0.99 (0.82–1.19) | 1.49 (1.27–1.74) | 1.54 (1.31–1.80) | 1.63 (1.41–1.90) | 1.62 (1.39–1.88) | ||||||
F | 0.93 (0.79–1.11) | 0.84 (0.71–1.01) | 1.23 (1.03–1.47) | 1.09 (0.91–1.31) | 1.43 (1.24–1.65) | 1.45 (1.25–1.69) | 1.55 (1.34–1.80) | 1.52 (1.30–1.77) | |||||||
Britain, 1946 | M | 1.21 (0.96–1.51) | 0.98 (0.77–1.24) | 1.53 (1.24–1.89) | 1.14 (0.91–1.43) | 1.80 (1.43–2.27) | 1.82 (1.42–2.32) | 2.40 (1.95–2.95) | 2.29 (1.84–2.86) | ||||||
F | 1.12 (0.91–1.38) | 1.04 (0.84–1.28) | 1.62 (1.31–2.00) | 1.42 (1.15–1.77) | 1.43 (1.14–1.78) | 1.41 (1.12–1.77) | 1.94 (1.57–2.40) | 1.78 (1.43–2.22) | |||||||
Britain, 1958 | M | 1.24 (1.10–1.40) | 0.98 (0.86–1.11) | 1.59 (1.39–1.82) | 1.20 (1.04–1.38) | 1.99 (1.77–2.23) | 2.00 (1.77–2.26) | 2.46 (2.17–2.79) | 2.33 (2.04–2.66) | ||||||
F | 1.39 (1.23–1.57) | 1.25 (1.11–1.42) | 1.75 (1.52–2.00) | 1.54 (1.34–1.77) | 1.88 (1.66–2.13) | 1.81 (1.60–2.05) | 2.20 (1.93–2.50) | 2.04 (1.80–2.33) | |||||||
Netherlands | M | 1.19 (0.99–1.43) | 1.03 (0.85–1.26) | 1.13 (0.96–1.32) | 0.95 (0.80–1.12) | 1.56 (1.30–1.86) | 1.54 (1.28–1.86) | 1.63 (1.41–1.89) | 1.66 (1.42–1.94) | ||||||
F | 1.26 (1.06–1.50) | 1.15 (0.96–1.38) | 1.45 (1.23–1.72) | 1.26 (1.06–1.50) | 1.45 (1.23–1.72) | 1.40 (1.17–1.67) | 1.73 (1.48–2.02) | 1.62 (1.38–1.91) | |||||||
US | M | 1.08 (0.87–1.36) | 0.97 (0.76–1.22) | 1.32 (1.09–1.60) | 1.17 (0.96–1.44) | 1.66 (1.32–2.08) | 1.64 (1.30–2.08) | 1.73 (1.42–2.10) | 1.62 (1.33–1.91) | ||||||
F | 1.21 (0.93–1.57) | 1.23 (0.94–1.60) | 1.13 (0.87–1.46) | 1.12 (0.86–1.46) | 0.93 (0.96–1.00) | 0.94 (0.71–1.24) | 1.10 (0.85–1.44) | 1.13 (0.86–1.49) |
. | . | Childhood SEP . | . | . | . | Adult SEP . | . | . | . | ||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
. | . | Ever smokers vs never smokers . | . | Current smokers vs others (never and ex smokers) . | . | Ever smokers vs never smokers . | . | Current smokers vs others (never and ex smokers) . | . | ||||||
Study . | . | Unadjusteda . | Adjustedb . | Unadjusteda . | Adjustedb . | Unadjusteda . | Adjustedc . | Unadjusteda . | Adjustedc . | ||||||
Finland | M | 0.95 (0.77–1.17) | 0.84 (0.67–1.04) | 0.89 (0.73–1.08) | 0.82 (0.67–1.00) | 1.80 (1.46–2.22) | 1.88 (1.51–2.32) | 1.43 (1.18–1.73) | 1.50 (1.23–1.82) | ||||||
Sweden | M | 1.42 (1.26–1.60) | 1.35 (1.20–1.53) | 1.20 (1.06–1.35) | 1.12 (0.99–1.26) | 1.45 (1.28–1.64) | 1.47 (1.29–1.66) | 1.59 (1.41–1.79) | 1.60 (1.41–1.81) | ||||||
F | 1.30 (1.15–1.46) | 1.26 (1.12–1.42) | 1.47 (1.30–1.67) | 1.39 (1.23–1.58) | 1.33 (1.18–1.50) | 1.33 (1.18–1.50) | 1.65 (1.46–1.87) | 1.59 (1.41–1.80) | |||||||
Denmark | M | 0.81 (0.66–0.98) | 0.73 (0.60–0.89) | 1.09 (0.92–1.32) | 0.99 (0.82–1.19) | 1.49 (1.27–1.74) | 1.54 (1.31–1.80) | 1.63 (1.41–1.90) | 1.62 (1.39–1.88) | ||||||
F | 0.93 (0.79–1.11) | 0.84 (0.71–1.01) | 1.23 (1.03–1.47) | 1.09 (0.91–1.31) | 1.43 (1.24–1.65) | 1.45 (1.25–1.69) | 1.55 (1.34–1.80) | 1.52 (1.30–1.77) | |||||||
Britain, 1946 | M | 1.21 (0.96–1.51) | 0.98 (0.77–1.24) | 1.53 (1.24–1.89) | 1.14 (0.91–1.43) | 1.80 (1.43–2.27) | 1.82 (1.42–2.32) | 2.40 (1.95–2.95) | 2.29 (1.84–2.86) | ||||||
F | 1.12 (0.91–1.38) | 1.04 (0.84–1.28) | 1.62 (1.31–2.00) | 1.42 (1.15–1.77) | 1.43 (1.14–1.78) | 1.41 (1.12–1.77) | 1.94 (1.57–2.40) | 1.78 (1.43–2.22) | |||||||
Britain, 1958 | M | 1.24 (1.10–1.40) | 0.98 (0.86–1.11) | 1.59 (1.39–1.82) | 1.20 (1.04–1.38) | 1.99 (1.77–2.23) | 2.00 (1.77–2.26) | 2.46 (2.17–2.79) | 2.33 (2.04–2.66) | ||||||
F | 1.39 (1.23–1.57) | 1.25 (1.11–1.42) | 1.75 (1.52–2.00) | 1.54 (1.34–1.77) | 1.88 (1.66–2.13) | 1.81 (1.60–2.05) | 2.20 (1.93–2.50) | 2.04 (1.80–2.33) | |||||||
Netherlands | M | 1.19 (0.99–1.43) | 1.03 (0.85–1.26) | 1.13 (0.96–1.32) | 0.95 (0.80–1.12) | 1.56 (1.30–1.86) | 1.54 (1.28–1.86) | 1.63 (1.41–1.89) | 1.66 (1.42–1.94) | ||||||
F | 1.26 (1.06–1.50) | 1.15 (0.96–1.38) | 1.45 (1.23–1.72) | 1.26 (1.06–1.50) | 1.45 (1.23–1.72) | 1.40 (1.17–1.67) | 1.73 (1.48–2.02) | 1.62 (1.38–1.91) | |||||||
US | M | 1.08 (0.87–1.36) | 0.97 (0.76–1.22) | 1.32 (1.09–1.60) | 1.17 (0.96–1.44) | 1.66 (1.32–2.08) | 1.64 (1.30–2.08) | 1.73 (1.42–2.10) | 1.62 (1.33–1.91) | ||||||
F | 1.21 (0.93–1.57) | 1.23 (0.94–1.60) | 1.13 (0.87–1.46) | 1.12 (0.86–1.46) | 0.93 (0.96–1.00) | 0.94 (0.71–1.24) | 1.10 (0.85–1.44) | 1.13 (0.86–1.49) |
Unadjusted = adjusted for age only.
Adjusted for adult SEP (own occupation; manual vs non-manual (reference category)) and age.
Adjusted for childhood SEP (father's occupation: manual vs non-manual (reference category)) and age.
. | . | Childhood SEP . | . | . | . | Adult SEP . | . | . | . | ||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
. | . | Ever smokers vs never smokers . | . | Current smokers vs others (never and ex smokers) . | . | Ever smokers vs never smokers . | . | Current smokers vs others (never and ex smokers) . | . | ||||||
Study . | . | Unadjusteda . | Adjustedb . | Unadjusteda . | Adjustedb . | Unadjusteda . | Adjustedc . | Unadjusteda . | Adjustedc . | ||||||
Finland | M | 0.95 (0.77–1.17) | 0.84 (0.67–1.04) | 0.89 (0.73–1.08) | 0.82 (0.67–1.00) | 1.80 (1.46–2.22) | 1.88 (1.51–2.32) | 1.43 (1.18–1.73) | 1.50 (1.23–1.82) | ||||||
Sweden | M | 1.42 (1.26–1.60) | 1.35 (1.20–1.53) | 1.20 (1.06–1.35) | 1.12 (0.99–1.26) | 1.45 (1.28–1.64) | 1.47 (1.29–1.66) | 1.59 (1.41–1.79) | 1.60 (1.41–1.81) | ||||||
F | 1.30 (1.15–1.46) | 1.26 (1.12–1.42) | 1.47 (1.30–1.67) | 1.39 (1.23–1.58) | 1.33 (1.18–1.50) | 1.33 (1.18–1.50) | 1.65 (1.46–1.87) | 1.59 (1.41–1.80) | |||||||
Denmark | M | 0.81 (0.66–0.98) | 0.73 (0.60–0.89) | 1.09 (0.92–1.32) | 0.99 (0.82–1.19) | 1.49 (1.27–1.74) | 1.54 (1.31–1.80) | 1.63 (1.41–1.90) | 1.62 (1.39–1.88) | ||||||
F | 0.93 (0.79–1.11) | 0.84 (0.71–1.01) | 1.23 (1.03–1.47) | 1.09 (0.91–1.31) | 1.43 (1.24–1.65) | 1.45 (1.25–1.69) | 1.55 (1.34–1.80) | 1.52 (1.30–1.77) | |||||||
Britain, 1946 | M | 1.21 (0.96–1.51) | 0.98 (0.77–1.24) | 1.53 (1.24–1.89) | 1.14 (0.91–1.43) | 1.80 (1.43–2.27) | 1.82 (1.42–2.32) | 2.40 (1.95–2.95) | 2.29 (1.84–2.86) | ||||||
F | 1.12 (0.91–1.38) | 1.04 (0.84–1.28) | 1.62 (1.31–2.00) | 1.42 (1.15–1.77) | 1.43 (1.14–1.78) | 1.41 (1.12–1.77) | 1.94 (1.57–2.40) | 1.78 (1.43–2.22) | |||||||
Britain, 1958 | M | 1.24 (1.10–1.40) | 0.98 (0.86–1.11) | 1.59 (1.39–1.82) | 1.20 (1.04–1.38) | 1.99 (1.77–2.23) | 2.00 (1.77–2.26) | 2.46 (2.17–2.79) | 2.33 (2.04–2.66) | ||||||
F | 1.39 (1.23–1.57) | 1.25 (1.11–1.42) | 1.75 (1.52–2.00) | 1.54 (1.34–1.77) | 1.88 (1.66–2.13) | 1.81 (1.60–2.05) | 2.20 (1.93–2.50) | 2.04 (1.80–2.33) | |||||||
Netherlands | M | 1.19 (0.99–1.43) | 1.03 (0.85–1.26) | 1.13 (0.96–1.32) | 0.95 (0.80–1.12) | 1.56 (1.30–1.86) | 1.54 (1.28–1.86) | 1.63 (1.41–1.89) | 1.66 (1.42–1.94) | ||||||
F | 1.26 (1.06–1.50) | 1.15 (0.96–1.38) | 1.45 (1.23–1.72) | 1.26 (1.06–1.50) | 1.45 (1.23–1.72) | 1.40 (1.17–1.67) | 1.73 (1.48–2.02) | 1.62 (1.38–1.91) | |||||||
US | M | 1.08 (0.87–1.36) | 0.97 (0.76–1.22) | 1.32 (1.09–1.60) | 1.17 (0.96–1.44) | 1.66 (1.32–2.08) | 1.64 (1.30–2.08) | 1.73 (1.42–2.10) | 1.62 (1.33–1.91) | ||||||
F | 1.21 (0.93–1.57) | 1.23 (0.94–1.60) | 1.13 (0.87–1.46) | 1.12 (0.86–1.46) | 0.93 (0.96–1.00) | 0.94 (0.71–1.24) | 1.10 (0.85–1.44) | 1.13 (0.86–1.49) |
. | . | Childhood SEP . | . | . | . | Adult SEP . | . | . | . | ||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
. | . | Ever smokers vs never smokers . | . | Current smokers vs others (never and ex smokers) . | . | Ever smokers vs never smokers . | . | Current smokers vs others (never and ex smokers) . | . | ||||||
Study . | . | Unadjusteda . | Adjustedb . | Unadjusteda . | Adjustedb . | Unadjusteda . | Adjustedc . | Unadjusteda . | Adjustedc . | ||||||
Finland | M | 0.95 (0.77–1.17) | 0.84 (0.67–1.04) | 0.89 (0.73–1.08) | 0.82 (0.67–1.00) | 1.80 (1.46–2.22) | 1.88 (1.51–2.32) | 1.43 (1.18–1.73) | 1.50 (1.23–1.82) | ||||||
Sweden | M | 1.42 (1.26–1.60) | 1.35 (1.20–1.53) | 1.20 (1.06–1.35) | 1.12 (0.99–1.26) | 1.45 (1.28–1.64) | 1.47 (1.29–1.66) | 1.59 (1.41–1.79) | 1.60 (1.41–1.81) | ||||||
F | 1.30 (1.15–1.46) | 1.26 (1.12–1.42) | 1.47 (1.30–1.67) | 1.39 (1.23–1.58) | 1.33 (1.18–1.50) | 1.33 (1.18–1.50) | 1.65 (1.46–1.87) | 1.59 (1.41–1.80) | |||||||
Denmark | M | 0.81 (0.66–0.98) | 0.73 (0.60–0.89) | 1.09 (0.92–1.32) | 0.99 (0.82–1.19) | 1.49 (1.27–1.74) | 1.54 (1.31–1.80) | 1.63 (1.41–1.90) | 1.62 (1.39–1.88) | ||||||
F | 0.93 (0.79–1.11) | 0.84 (0.71–1.01) | 1.23 (1.03–1.47) | 1.09 (0.91–1.31) | 1.43 (1.24–1.65) | 1.45 (1.25–1.69) | 1.55 (1.34–1.80) | 1.52 (1.30–1.77) | |||||||
Britain, 1946 | M | 1.21 (0.96–1.51) | 0.98 (0.77–1.24) | 1.53 (1.24–1.89) | 1.14 (0.91–1.43) | 1.80 (1.43–2.27) | 1.82 (1.42–2.32) | 2.40 (1.95–2.95) | 2.29 (1.84–2.86) | ||||||
F | 1.12 (0.91–1.38) | 1.04 (0.84–1.28) | 1.62 (1.31–2.00) | 1.42 (1.15–1.77) | 1.43 (1.14–1.78) | 1.41 (1.12–1.77) | 1.94 (1.57–2.40) | 1.78 (1.43–2.22) | |||||||
Britain, 1958 | M | 1.24 (1.10–1.40) | 0.98 (0.86–1.11) | 1.59 (1.39–1.82) | 1.20 (1.04–1.38) | 1.99 (1.77–2.23) | 2.00 (1.77–2.26) | 2.46 (2.17–2.79) | 2.33 (2.04–2.66) | ||||||
F | 1.39 (1.23–1.57) | 1.25 (1.11–1.42) | 1.75 (1.52–2.00) | 1.54 (1.34–1.77) | 1.88 (1.66–2.13) | 1.81 (1.60–2.05) | 2.20 (1.93–2.50) | 2.04 (1.80–2.33) | |||||||
Netherlands | M | 1.19 (0.99–1.43) | 1.03 (0.85–1.26) | 1.13 (0.96–1.32) | 0.95 (0.80–1.12) | 1.56 (1.30–1.86) | 1.54 (1.28–1.86) | 1.63 (1.41–1.89) | 1.66 (1.42–1.94) | ||||||
F | 1.26 (1.06–1.50) | 1.15 (0.96–1.38) | 1.45 (1.23–1.72) | 1.26 (1.06–1.50) | 1.45 (1.23–1.72) | 1.40 (1.17–1.67) | 1.73 (1.48–2.02) | 1.62 (1.38–1.91) | |||||||
US | M | 1.08 (0.87–1.36) | 0.97 (0.76–1.22) | 1.32 (1.09–1.60) | 1.17 (0.96–1.44) | 1.66 (1.32–2.08) | 1.64 (1.30–2.08) | 1.73 (1.42–2.10) | 1.62 (1.33–1.91) | ||||||
F | 1.21 (0.93–1.57) | 1.23 (0.94–1.60) | 1.13 (0.87–1.46) | 1.12 (0.86–1.46) | 0.93 (0.96–1.00) | 0.94 (0.71–1.24) | 1.10 (0.85–1.44) | 1.13 (0.86–1.49) |
Unadjusted = adjusted for age only.
Adjusted for adult SEP (own occupation; manual vs non-manual (reference category)) and age.
Adjusted for childhood SEP (father's occupation: manual vs non-manual (reference category)) and age.
Current smoking, ranging from 22 to 61% in men and from 26 to 51% in women, was also more prevalent among manual groups (Table 3). The absolute difference between manual and non-manual groups was generally large, though ranging from 2% (US women) to 20% (British men in the 1958 cohort). There was a consistent effect of childhood position in women, such that manual background increased the risk of current adult smoking (Table 4). The childhood effect on current smoking remained in four studies (adjusted ORs 1.26–1.54) although it attenuated with adjustment for adult SEP. In men, only one study showed a childhood effect after adjustment for adult position. In general, adult SEP showed strong effects on the odds of current smoking and very little of the effect was explained by childhood social background (adjusted ORs for men 1.50–2.33; for women 1.13–2.04) (Table 4).
The prevalence of ex-smoking (among ever smokers) ranged from 25 to 62% among men and from 13 to 52% among women (Table 3). Ex-smoking was more prevalent among the non-manual group, except in the Finnish study. Figure 1 shows that women with manual childhood origins had a reduced chance of ex-smoking, except in the US, and this was mostly unexplained by adjustment for adult position (adjusted ORs 0.64–0.81 and for the US 1.17). For men, the pattern was less consistent, but four of the seven studies (the US, Danish, and two British studies) show that manual childhood position reduced the chance of being an ex-smoker, even after allowing for adult SEP. Figure 1 shows that for males and females in all countries except Finland, manual adult SEP reduced the probability of being an ex-smoker. This effect remained after control for childhood SEP (adjusted ORs 0.48–0.70). So for ex-smoking there was a contribution of both childhood and adulthood SEP to the chance of quitting.
Obesity
The prevalence of obesity ranged from 3 to 17% among men and from 2 to 16% among women, with the highest prevalence found among those with manual occupation (Table 5). The absolute socioeconomic differences in obesity were modest (range 1–10%), mainly due to low prevalence in these populations, while the relative differences were fairly large, especially among women from the US and Sweden, where three times as many women in manual occupations, when compared with non-manual, were obese.
. | BMI 30+ . | . | . | . | |||
---|---|---|---|---|---|---|---|
. | M . | . | F . | . | |||
Study . | Manual . | Non-manual . | Manual . | Non-manual . | |||
Finland | 17 | 16 | – | – | |||
Sweden | 6 | 4 | 6 | 2 | |||
Denmark | 15 | 10 | 12 | 8 | |||
Britain, 1946b | 6 | 5 | 10 | 5 | |||
Britain, 1958 | 13 | 10 | 16 | 10 | |||
Netherlands | 6 | 3 | 9 | 5 | |||
US | 7 | 4 | 15 | 5 |
. | BMI 30+ . | . | . | . | |||
---|---|---|---|---|---|---|---|
. | M . | . | F . | . | |||
Study . | Manual . | Non-manual . | Manual . | Non-manual . | |||
Finland | 17 | 16 | – | – | |||
Sweden | 6 | 4 | 6 | 2 | |||
Denmark | 15 | 10 | 12 | 8 | |||
Britain, 1946b | 6 | 5 | 10 | 5 | |||
Britain, 1958 | 13 | 10 | 16 | 10 | |||
Netherlands | 6 | 3 | 9 | 5 | |||
US | 7 | 4 | 15 | 5 |
Adult SEP (own occupation; manual vs non-manual).
Weighted to allow for original sampling procedure.
. | BMI 30+ . | . | . | . | |||
---|---|---|---|---|---|---|---|
. | M . | . | F . | . | |||
Study . | Manual . | Non-manual . | Manual . | Non-manual . | |||
Finland | 17 | 16 | – | – | |||
Sweden | 6 | 4 | 6 | 2 | |||
Denmark | 15 | 10 | 12 | 8 | |||
Britain, 1946b | 6 | 5 | 10 | 5 | |||
Britain, 1958 | 13 | 10 | 16 | 10 | |||
Netherlands | 6 | 3 | 9 | 5 | |||
US | 7 | 4 | 15 | 5 |
. | BMI 30+ . | . | . | . | |||
---|---|---|---|---|---|---|---|
. | M . | . | F . | . | |||
Study . | Manual . | Non-manual . | Manual . | Non-manual . | |||
Finland | 17 | 16 | – | – | |||
Sweden | 6 | 4 | 6 | 2 | |||
Denmark | 15 | 10 | 12 | 8 | |||
Britain, 1946b | 6 | 5 | 10 | 5 | |||
Britain, 1958 | 13 | 10 | 16 | 10 | |||
Netherlands | 6 | 3 | 9 | 5 | |||
US | 7 | 4 | 15 | 5 |
Adult SEP (own occupation; manual vs non-manual).
Weighted to allow for original sampling procedure.
All studies show that women from manual childhood backgrounds had an elevated risk of obesity in adulthood (age adjusted ORs 1.15–2.94) (Figure 2). This finding is irrespective of differences in prevalence; it is strongest in studies with measured body-size (the British studies) and remains for most countries, though attenuated, after adjustment for adult SEP (adjusted ORs 0.96–2.50). Among men, the association with childhood position was mostly in the same direction, although effects were weaker (adjusted ORs 0.79–1.42).
Most studies show a strong association between obesity and adult SEP: men and women in manual social classes had an increased risk of obesity (Figure 2). This risk was largely unaffected by adjustment for childhood SEP (adjusted ORs 1.06–2.24 for men; 1.21–3.26 for women). In the two British studies where childhood position was measured prospectively, the adult effect showed some attenuation with adjustment.
Discussion
In line with other studies, our analyses across six countries show that, in general, adult SEP influences adult smoking behaviour and obesity in mid-life. Associations with adult position tended, in general, to be more pronounced than, and little explained by, childhood SEP. Building on this well-established finding, we examined whether childhood SEP influences these two major risk factors for chronic disease in adult life and if so, whether this acts mainly through an association with adult position. A supplementary question was whether the associations for smoking and obesity with adult social position could be explained by childhood position. Socioeconomic gradients in smoking and obesity are known to vary over time, across cohorts and between societies.22,31–34 Heterogeneity is therefore likely in the effects of childhood and adult social position on smoking and obesity measured in different countries among different birth cohorts. However, this study provides some evidence, at least in women, that manual social origins increased the risk of obesity and reduced the chance of quitting smoking in adulthood, irrespective of adult social position. The effect of childhood SEP was inconsistent for ever-smoking, and in men also for current smoking. It is noteworthy that there is some consistency across studies in the effect of child SEP for women, given the diversity of year of birth and geographical location. This may be, at least in part, because the adult socioeconomic measure, based on occupation, is less adequate for women than for men. It also points to the importance of studying the associations between childhood SEP, adult SEP and adult health risks separately for men and women. There are few such studies. However, these studies suggest that both the social pathways from childhood SEP to adult SEP and the childhood influences on adult health behaviours may be gender dependent. For men and women, education has been identified as a major pathway in the intra-generational continuities in SEP and one which mediates the association between childhood SEP and adult health behaviours like smoking.13,16 Also, reproductive pathways and early and lone parenthood in particular, have been found to have an additional effect on women's SEP and smoking status in adulthood.16
Methodological considerations
The selection of studies was arbitrary, but all were population-based, representative of the areas from which they were drawn, included adults in mid-life born at different times in different places, and unusually, had information on childhood and adult-life SEP. To optimize comparability of outcome, the age range in some studies, such as The Globe and Alameda County studies, was restricted, resulting in a reduction of sample size. While for most studies smoking outcomes and BMI were obtained within the last 20 yr, in the US study information was gathered in the 1960s when obesity was less prevalent and smoking among men had peaked but among women was still rising to its zenith in the mid 1970s.31,35 The prospective cohorts, where participants were all of the same age, height, and weight were measured, showed the most consistent patterns.
Standard measures of SEP appropriate for quantitative cross-national comparisons are difficult to obtain. Other studies have used measures based either on occupation1 or on education.32 For our purposes where we are seeking to disentangle childhood influences from those in adult life, education is ambiguous because of its position on the pathway between childhood and adult socioeconomic circumstances. Hence, in line with other comparative studies, we relied on a simple dichotomous manual/non-manual classification of occupational class.1 This is likely to under-estimate associations for both childhood and adult SEP that would be seen with better differentiated occupational measures. For example, we report a weak effect of childhood SEP on obesity in the 1946 birth cohort, whilst analyses based on a six-category measure shows stronger associations.19 An additional consideration is that adult SEP was obtained contemporaneously, whilst data on childhood position were usually obtained retrospectively. This may result in an under-estimation of the effects of childhood SEP compared with adult SEP since the latter may be better measured.7,36 A further possibility is that smoking and obesity are linked to social mobility and that this, in turn, influences the associations observed in our study. Some evidence exists that adolescent health-related behaviours, including smoking, physical activity, and diet, are related to an individual's social trajectory.16,37,38 Similar findings have been reported for obesity, at least among women.39 It is plausible therefore, that social mobility would conflate associations for adult SEP. The same argument does not apply to childhood SEP, which pre-dates any social mobility.
Comparison with other studies: smoking
In line with other studies of northern Europe and the US, we found that manual adult social position generally increased the odds of ever smoking and current smoking and reduced the odds of ex-smoking.31,32,41,42 Despite the fact that smoking is initiated in adolescence, to our knowledge, no previous studies have examined the influence of childhood position on these three major components of adult smoking status. Previous work focussing only on current smoking shows no effect of childhood position for men, though some effects are reported for women.10,12–14,42 We have broadened this focus to examine a potential influence of social origins on other aspects of smoking behaviour. Our findings suggest that the effects of childhood may be most evident in relation to quitting. Individuals with non-manual social class of origin are more likely to quit, with childhood influences stronger among women than men. Quitting holds the key to reducing the public health burden of smoking.20 There is evidence that poor childhood origins are linked to higher levels of cigarette consumption (and therefore nicotine addiction) in adolescence and adulthood, which mitigate against quitting later in life.43 Factors in the home environment associated with childhood SEP, including parental smoking, may have an influence on the persistence of smoking into adulthood;44,45 in addition, childhood position may act through an effect on educational and social pathways.13,16 However, becoming a smoker (ever smoking) was not associated with SEP in childhood, despite its proximity to adolescence and early adulthood, which are the key life stages for the uptake of regular smoking.
It is well-established that socioeconomic gradients in smoking change as populations move through the smoking epidemic, first among men and then among women. We would therefore not expect the influence of childhood and adult SEP to be uniform across time, place, and gender. Variability of childhood effects was seen across our studies, and also in the magnitude of adulthood effects. There is nonetheless a pervasive influence of adult position on all three aspects of smoking behaviour, which is not explained by childhood background.
Comparison with other studies: obesity
We find, as do others,22 that manual social position in adulthood increases the risk of obesity, most notably in women and less consistently in men. There are now several studies also showing that childhood social origins have long-lasting effects on adult obesity,34 and that, as in our study, this childhood influence persists after allowing for adult position.46 Explanations for cumulative effects of both childhood and adult SEP are likely to include both biological and social mechanisms and pathways. Education level is a strong determinant of health-related behaviours, such as diet and physical activity. Elsewhere we have shown that education is related to adult BMI in the two British studies and that this largely accounts for the relationship with adult social position, whereas it does not explain the association between social origins and adult obesity.17,18 The underlying mechanisms are still to be identified. One possibility is that there may be prenatal or early post-natal influences that have long-term effects on body composition.47,48 Further research is needed on these complex patterns and how they relate to the dramatic secular trend in obesity in recent decades.49
To conclude, this study confirms that, across time and place, adult SEP is important for smoking and obesity in middle age. In addition, our findings suggest that influences from childhood may also contribute, in particular, to quitting smoking and to obesity. Future research should investigate these influences using other datasets and examine childhood position in relation to other adult risk factors. In terms of public health policy, our study points to the importance of interventions targeted not only at changing health behaviours in adulthood, for example through voluntary and statutory agreements on food and tobacco advertising, fiscal regulations, and investments in health promotion and support services. In line with other studies, it points to the potential public health gain of a life-course approach to tackling smoking and obesity. This would include policies which address inequalities in socioeconomic circumstances in childhood as well as adult life.8,16,50
Socioeconomic inequalities in adult risk factors such as smoking and obesity are well-established. Less is known about the influence of childhood socioeconomic position across time and place, and whether it explains some of these adult inequalities.
Seven studies in six countries showed adult social position to be an important influence on obesity and quitting smoking.
Manual social background in childhood increased the risk of obesity and reduced the chance of quitting smoking in adulthood in most, but not all, populations.
The authors are all part of the ESF programme on Social Variations in Health Expectancy (Working Group on life course influences on health inequalities), and the Programme facilitated the scientific exchange on which the paper is based. We thank Barbara Jefferis for her help with analyses of the 1958 cohort and in the preparation of the manuscript, tables, and figures.
References
Kunst AE, Groenhof F, Mackenbach JP, EU Working Group on Socioeconomic Inequalities in Health. Occupational class and cause specific mortality in middle aged men in 11 European countries: comparison of population based studies.
Drever F, Whitehead M (eds). Health Inequalities. Office for National Statistics. DS No. 15. London: The Stationery Office,
Galobardes B, Lynch JW, Davey Smith G. Childhood socioeconomic circumstances and cause-specific mortality in adulthood.
Freedman DS, Dietz WH, Srinivasan SR, Berenson GS. The relation of overweight to cardiovascular risk factors among children and adolescents: the Bogalusa Heart Study.
McCarron P, Davey Smith G. Physiological measurements in children and young people, and risk of coronary heart disease in adults. In: Giles A (ed). A Lifecourse Approach to Coronary Heart Disease Prevention. London: National Heart Forum, The Stationery Office,
Raitakari OT, Juonala M et al. Cardiovascular risk factors in childhood and carotid artery intima-media thickness in adulthood: the Cardiovascular Risk in Young Finns Study.
Kuh D, Ben-Shlomo Y (eds). A Lifecourse Approach to Chronic Disease Epidemiology. Oxford: Oxford University Press,
Lynch JW, Kaplan GA, Salonen JT. Why do poor people behave poorly? Variation in adult health behaviours and psychosocial characteristics by stages of the socioeconomic lifecourse.
Graham H, Power C. Childhood Disadvantage and Adult Health: A Lifecourse Framework. London: Health Development Agency,
Blane D, Hart CL, Davey Smith G, Gillis CR, Hole DJ, Hawthorne VM. Association of cardiovascular disease risk factors with socioeconomic position during childhood and during adulthood.
Wannamethee SG, Whincup PH, Shaper G, Walker M. Influence of fathers’ social class on cardiovascular disease in middle-aged men.
Brunner E, Shipley MJ, Blane D, Davey Smith G, Marmot MG. When does cardiovascular risk start? Past and present socioeconomic circumstances and risk factors in adulthood.
Graham H, Der G. Smoking and women's health. Influences on women's smoking status. The contribution of socioeconomic status in adolescence and adulthood.
Graham H, Hunt K. Socio-economic Influences on Women's Smoking Status in Adulthood: Insights from the West of Scotland Twenty-07 Study.
Heslop P, Davey Smith G, Macleod J, Hart C. The socioeconomic position of employed women, risk factors and mortality.
Jefferis JMH, Power C, Graham G, Manor O. Effects of childhood socio-economic circumstances on persistent smoking.
Hardy R, Wadsworth M, Kuh D. The influence of childhood weight and socioeconomic status on change in adult body mass index in a British national birth cohort.
Power C, Manor O, Matthews S. Child to adult socioeconomic conditions and obesity in a national cohort.
Langenberg C, Hardy R, Kuh D, Brunner E, Wadsworth M. Central and total obesity in middle aged men and women in relation to lifetime socioeconomic status: evidence from a national birth cohort.
Doll R, Peto R, Jillian Boreham J, Sutherland I. Mortality in relation to smoking: 50 years' observations on male British doctors
WHO. Obesity: Preventing and Managing the Global Epidemic. Report of a WHO Consultation on Obesity. Geneva: World Health Organization,
Sobal J, Stunkard AJ. Socioeconomic status and obesity: a review of the literature.
Wadsworth ME, Butterworth SL, Hardy RJ et al. The life course prospective design: an example of benefits and problems associated with study longevity.
Ferri E (ed). Life at 33 The Fifth Follow-up of the National Child Development Study. London: National Children's Bureau,
Statistics Sweden. The Swedish Survey of Living Conditions. Design and Methods. Living Conditions. Stockholm: SCB (Appendix 16.
Christensen U, Lund R, Damsgaard MT et al. Cynical hostility, socioeconomic position, health behaviours and symptom load: a cross-sectional analysis in a Danish population-based study.
Lynch JW, Kaplan GA, Cohen RD et al. Childhood and adult socioeconomic status as predictors of mortality in Finland.
Mackenbach JP, Mheen H van de, Stronks K. A prospective cohort study investigating the explanation of socio- economic inequalities in health in The Netherlands.
Kaplan GA, Camacho T. Perceived health and mortality: a nine-year follow-up of the human population laboratory cohort.
WHO Expert Committee. Physical Status: The Use and Interpretation of Anthropometry. Geneva: WHO, WHO Technical Report,
Pierce JP. International comparisons of trends in cigarette smoking prevalence.
Cavelaars AE, Kunst AE, Geurts JJ et al. Educational differences in smoking: international comparison.
Evandrou M, Falkingham J. Smoking behaviour and socio-economic status: a cohort analysis, 1974 to 1998.
Parsons TJ, Power C, Logan S, Summerbell CD. Childhood predictors of adult obesity: a systematic review. Int J Obes Relat Metab Disord
Lynch J, Davey-Smith G. A life course approach to chronic disease epidemiology.
Kuh D, Hardy R, Langenberg C, Richards M, Wadsworth ME. Mortality in adults aged 26–54 years related to socioeconomic conditions in childhood and adulthood: post war birth cohort study.
Glendinning A, Shucksmith J, Hendry L. Social class and adolescent smoking behaviour.
Karvonen S, Rimpela AH, Rimpela MK. Social mobility and health related behaviours in young people.
Gortmaker SL, Must A, Perrin JM, Sobol AM, Dietz WH. Social and economic consequences of overweight in adolescence and young adulthood.
Graham H. Smoking prevalence among women in the European Community 1950–1990.
Jarvis MJ. Patterns and predictors of smoking cessation in the general population. In: Bollinger CT, Fagerstrom KO (eds). The Tobacco Epidemic. Progr Respir Res. Basel: Karger,
Davey-Smith G, Hart C, Hole D et al. Education and occupational social class: which is the more important indicator of mortality risk?
Jefferis BJ, Graham H, Manor O, Power C. Level of cigarette smoking and socio-economic circumstances in adolescence: how do they affect adult smoking?
Rossow I, Rise J. Concordance of parental and adolescent health behaviours
Green G, Macintyre S, West P, Ecob R Like parent like child? Associations between drinking and smoking behaviour of parents and their children.
Poulton R, Caspi A, Milne BJ et al. Association between children's experience of socioeconomic disadvantage and adult health: a life-course study.
Power C, Jefferis BJ. Fetal environment and subsequent obesity: a study of maternal smoking.
Gillman MW, Rifas-Shiman SL, Camargo CA, Jr et al. Risk of overweight among adolescents who were breastfed as infants.
Mokdad AH, Serdula MK, Dietz WH, Bowman BA, Marks JS, Koplan JP. The spread of the obesity epidemic in the United States, 1991–1998.