The effect of pre-employment factors on job control, job strain and psychological distress: A 31-year longitudinal study

ARTICLE IN PRESS

Social Science & Medicine 65 (2007) 187–199 www.elsevier.com/locate/socscimed

The effect of pre-employment factors on job control, job strain and psychological distress: A 31-year longitudinal study Marko Elovainioa,, Mika Kivima¨kib, Ellen Ekb, Jussi Vahterab, Teija Honkonenb, Anja Taanilac, Juha Veijolac, Marjo-Riitta Ja¨rvelinc a

Department Psychology, University of Helsinki, P.O. Box 9, 00014 University of Helsinki, Helsinki, Finland b Finnish Institute of Occupational Health, Finland c University of Oulu, Finland Available online 8 May 2007

Abstract This study examined the role of pre-employment factors, such as maternal antenatal depression, low birth weight, childhood socioeconomic position, early adolescence health risk behaviours and academic performance, in the relationship between work characteristics (low job control and high job demands, or job strain) and psychological distress at age 31. The data of 2062 women and 2231 men was derived from the prospective unselected population-based Northern Finland 1966 Birth Cohort study. Results of linear regression models showed that being female, father’s low socioeconomic position, and poor academic achievement in adolescence were linked to low control and high job strain jobs at age 31, and that low control and high job strain were associated with psychological distress at age 31. Although having lower school grades, high absence rate from school, and moderate alcohol consumption at age 14 were significant predictors of psychological distress at age 31, the associations between job control, job strain and psychological distress remained after controlling for these and other pre-employment effects. As such, pre-employment factors do seem to link people to risky work environments, which in turn seem to relate strongly to psychological distress. However, the relationship between pre-employment factors and later psychological distress in adulthood is not completely explained by job environment. r 2007 Elsevier Ltd. All rights reserved. Keywords: Mental health; Life course; Job strain; Job control; Birth weight; Finland; Psychological distress

Introduction Mental health problems have been recognized as the most rapidly growing reasons for early retireCorresponding author. Tel.: +358 503020621.

E-mail addresses: [email protected].fi (M. Elovainio), mika.kivimaki@ttl.fi (M. Kivima¨ki), ellen.ek@ttl.fi (E. Ek), jussi.vahtera@ttl.fi (J. Vahtera), teija.honkonen@ttl.fi (T. Honkonen), anja.taanila@oulu.fi (A. Taanila), juha.veijola@oulu.fi (J. Veijola), marjo-riitta.jarvelin@oulu.fi (M.-R. Ja¨rvelin). 0277-9536/$ - see front matter r 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.socscimed.2007.02.052

ment in many Western countries. Although a number of factors, such as genetic factors, disturbed family environment, pre-disposing personality traits, exposure to traumatic events, low social support and recent stressful life events, are known to increase the risk of affective disorders (Bruce, 2002; Kendler, Neale, Kessler, Heath, & Eaves, 1993) it has also been suggested that factors related to work life may predispose individuals toward such disorders (Kivimaki, Elovainio, Kokko et al., 2003; Kivimaki, Elovainio, Vahtera, Virtanen, &

ARTICLE IN PRESS 188

M. Elovainio et al. / Social Science & Medicine 65 (2007) 187–199

Stansfeld, 2003; Stansfeld, Fuhrer, Head, Ferrie, & Shipley, 1997; Stansfeld, Head, & Ferrie, 1999). Also evidence on the relationship between job characteristics and affective disorders has recently begun to emerge (Niedhammer et al., 1998; Stansfeld et al., 1997; Wall et al., 1997; Weinberg & Creed, 2000; Ylipaavalniemi et al., 2005). The Job Strain Model (Karasek, 1990; Karasek, Baker, Marxer, Ahlbom, & Theorell, 1981), also known as the Demand-Control Model, has been one of the most widely tested models in research on the relationship between work and health. The model states that employees working under high strain (a combination of high work demands and low job control) have a higher risk of health problems than those with no such strain. Indeed, previous evidence, which is mainly cross-sectional, suggests that high job strain, high demand or low control are associated with depression, poor mental health, and psychological distress (de Lange, Taris, Kompier, Houtman, & Bongers, 2003; Karasek, 1990; Landsbergis et al., 1998; Mausner-Dorsch & Eaton, 2000; Stansfeld et al., 1997; Williams et al., 1997). Previous studies on job strain and mental health have largely been focused on adulthood factors. We would like to suggest that pre-employment factors might influence job strain and mental health and therefore their role merits further consideration. For example, exposure to adversities, such as work stress, is not necessarily random. A genetically influenced set of traits may both increase an individuals’ probability of selecting themselves into a high risk environment likely to produce job strain and increase their vulnerability to affective disorders (Kendler et al., 1995). Socioeconomic status during childhood may select people to lower academic careers (Osborn & Milbank, 1989) and related highly stressful work environment. It is also possible that poor health, especially poor mental health, which may largely be affected by early psychosocial factors (Wang, 2004), make one vulnerable to stressful work environment. Problems at school and negative health risk behaviour during adolescence may also restrict ones possibilities of choosing ones occupational status (Bradley & Corwyn, 2002). Life course epidemiology has emphasized various developmental stages and processes defining the relationships between early risk factors, adulthood circumstances and health. The hypothesis defining the relationship between socioeconomic status and

health, emphasize the role of various developmental stages differently. They seem to underlie early risks as etiological factors, later risks as independent factors, later risk as pathways from early risks (selective hypothesis) or early and later risks as independent or additive factors. All of these hypotheses, which may be considered to be rather complementary than contradictory, have gained some empirical support (Wilkinson & Pickett, 2006). The hypothesis behind the idea of critical period or early origin implies a stage in the individual’s development in which an increased sensitivity to the influence of external agents may have crucial effects on later health. The foetal origin’s hypothesis is well-known example which assumes that adverse social circumstances in early life and especially poor maternal nutrition during pregnancy lead to impaired foetal growth, biological programming of the foetus and increased risk of coronary hear disease, hypertension and diabetes later in life (Barker, Forsen, Uutela, Osmond, & Eriksson, 2001). Thereafter, it is hypothesized that the individual can function only within the parameters set during this unique developmental opportunity. Barker (1991) describes this process as biological programming. The research on the effects of early factors, such as birth weight, on psychological problems or mental health, have mainly been restricted to research on antecedents of schizophrenia (Isohanni et al., 2000). The results of Cheung, Ma, Machin, & Karlberg (2004) showed that low birth weight for gestational age was related to psychological distress at age 42. It has also been shown that low birth weight may predict depressive disorders (Gale & Martyn, 2004; Patton, Coffey, Carlin, Olsson, & Morley, 2004). Although the relationship between low socioeconomic position and poor health is widely reported (Adler et al., 1994; Marmot, 1997), less is known about the developmental pathways through which low SES in childhood leads to morbidity in adulthood. Childhood and adolescence are times of dramatic developmental changes, and these changes may result in differences in SES relationships. It has been, for example, suggested that development of biological organs that is less than optimal may not become a problem unless the system is stressed. The idea behind this pathway model is that although SES effects initially are large they gradually decrease over time, unless those early SES circumstances select individuals to poor circumstances or

ARTICLE IN PRESS M. Elovainio et al. / Social Science & Medicine 65 (2007) 187–199

health risks later in life. Typical factor selecting people is education. Children in poorer home circumstances, and parents with low educational attainment, have reduced chances of pre-school experience, which provides a disadvantaged beginning to the school years (Wadsworth, 1997). There are clear SES differences in age at school leaving and in going on to further education (Halsey, Collin, & Anderson, 1996) and parental concern for educational attainment have been shown to be a powerful predictor of occupational status, and this in turn is strongly associated with health (Sweeting & West, 1995). Health-related habits, such as smoking, nutrition and alcohol consumption tend to be associated with education, which generally has the effect of reducing adverse health behaviours (Kivimaki et al., 2006). Pathway hypothesis is clearly related to research on the factors that predict work-related health and health behaviours (Kouvonen et al., 2005; Pulkki et al., 2003; Pulkki-Raback, Elovainio, Kivimaki, Raitakari, & Keltikangas-Jarvinen, 2005) Evidence for this model comes mainly from social mobility studies suggesting that downward intergenerational social mobility seems to be a risk for unhealthy behaviour in women (Pulkki, Kivimaki, Elovainio, Viikari, & Keltikangas-Jarvinen, 2003). Less is known, however, about the processes behind this mobility and the evidence on intra-generational social mobility is scarce. The few follow-up studies have obtained results suggesting that SES may act both as a cause and consequence in the development of psychosomatic symptoms and intergenerational social mobility, especially among women (Huurre, Aro, & Rahkonen, 2003). It is also suggests that observed socioeconomic differences in risk of morbidity is mainly explained by cumulative differential lifetime exposure to risk of chronic disease and mortality. The accumulation hypothesis has been tested in several studies (Davey et al., 1998; Hallqvist, Lynch, Bartley, Lang, & Blane, 2004; Holland et al., 2000; Lynch, Kaplan, & Shema, 1997; Power, Matthews, & Manor, 1996; Wamala, Lynch, & Kaplan, 2001). Typically, the model is tested using two or more measures at different stages of life and to analyse whether there is a difference between those having several low SES points and those having only one low SES point. For example, Power et al. (1996)) measured social class in birth, at ages 16, 23 and 33. The outcome measure was self-rated health at ages 23 and 33. They concluded that cumulative lifetime exposures,

189

as represented by social class at birth and at ages 16 and 23 and prior ill health, seemed to have a major role in creating health inequalities at age 33. According to Power and others (2002) gradient in psychological distress reflect the cumulative effect of multiple adversities experienced from childhood to adulthood. They found that early life factors contributing to differences in adulthood psychological distress include childhood socioeconomic status, material, educational and family factors. The current study Although previous studies have found factors related to mental health at each life stage onwards from birth (Power & Matthews, 1997), only few studies on the relationship between job strain and psychological distress have been able to take into account for the effects of pre-employment risk factors for mental health (Bourbonnais, Comeau, & Vezina, 1999; Matthews & Power, 2002; Matthews, Power, & Stansfeld, 2001; Matthews, Stansfeld, & Power, 1999; Power, Stansfeld, Matthews, Manor, & Hope, 2002; Vermeulen & Mustard, 2000). Previously found risk factors for distress or depression, such as maternal antenatal depression (Hellgren, Gillberg, Bagenholm, & Gillberg, 1994), low birth weight (Cheung, Low, Osmond, Barker, & Karlberg, 2000; Cianfarani et al., 2000; Ogdendahl et al., 2006), childhood socioeconomic status (Brooks-Gunn, Klebanov, Liaw, & Spiker, 1993; Wilkinson, 1997), poor educational achievement, teenage heavy alcohol use and smoking (Anda et al., 1990; Brown, Croft, Anda, Barrett, & Escobedo, 1996; Haarasilta, Marttunen, Kaprio, & Aro, 2004; Kandel et al., 1999; Merikangas, Dierker, & Szatmari, 1998; Riala, Hakko, Isohanni, Jarvelin, & Rasanen, 2004). We tested the role of preemployment factors in the relationship between work characteristics (job control, job demands and job strain) and psychological distress using the 31year longitudinal three-wave study. The testing procedure included following steps: (1) To test the extent to which pre-employment childhood risk factors (mothers’ depressive mood during pregnancy, low parental SES and low birth weight) and adolescence risk factors (low academic performance, absence from school, smoking, alcohol consumption) predict (A) psychological distress in adulthood and (B) poor adulthood circumstances, such as high strain job; (2) To test the potential relationship between work characteristics and

ARTICLE IN PRESS 190

M. Elovainio et al. / Social Science & Medicine 65 (2007) 187–199

psychological distress; and (3) whether the potential association between pre-employment factors and psychological distress is attributable to job characteristics (through selective processes) or whether early and later work-related risks have independent effects on psychological distress. Methods Procedure and participants The participants were derived from the prospective epidemiological study of Northern Finland birth cohort of 1966. The cohort is based upon 12,055 pregnant women and their 12,058 live-born children (6169 boys and 5889 girls) in the provinces of Lapland and Oulu with an expected delivery date during 1966, representing 96% of all births in the region (Isohanni et al., 2001; Rantakallio, 1988). Data on the biological, socioeconomic, and health conditions, living habits, and family characteristics of cohort members were collected prospectively from pregnancy up to the age 31. At the 14-year follow-up in 1980 the subjects, who were then at age 14, were requested to complete a questionnaire concerning parents’ socioeconomic status, other family-related factors, school-related factors and health behaviour. In 1997 (at age 31), the subjects completed a questionnaire on work related factors. The present study is based on persons who participated in the postal and clinical examination at age 31 and gave a written consent for the use of data for research purposes (N ¼ 5956, 71% of those invited). The final study sample consisted of those 4293 (2231 men and 2096 women) who had answered the work-related questions at age 31. Previous studies on sample attrition in the studies between 1966 and 1998 has found that dropouts were more often men (43% vs. 33%), but otherwise the attrition was not systematic (Rantakallio, 1988). No serious attrition related to mental health issues (evaluated using data from hospitalization registers) has been previously reported (Isohanni et al., 1998). Pre-employment measures Childhood factors As part of the antenatal data collection, the mothers were asked by the interviewing nurse at the antenatal clinic during midgestation (mainly between the 24th and 28th gestational week) whether they felt that their mood had been 1 ¼ normal,

2 ¼ depressed or 3 ¼ very depressed during pregnancy (Ma¨ki et al., 2003). In the analyses, the two latter categories were considered as ‘‘depressed’’. Birth weight was measured at birth and recorded from the hospital files (Rantakallio & Koiranen, 1987). Previously, mothers’ antenatal depression measure using this question has been shown to associate with criminality of the offspring (Ma¨ki et al., 2003). Parental socioeconomic position was defined using fathers’ occupation at the time of birth in 1966. Occupations were classified into two categories (1) professionals and skilled workers and (2) unskilled workers, workers with no occupation and farmers (Rantakallio, 1979). This classification has been related a wide variety of negative childhood outcomes including smoking and mild mental retardation (Rantakallio, 1979, 1983). Adolescence factors At age 14, information on school attending was obtained from a questionnaire filled in by the child and his or her parents (response rate 98%): repeating class (a proxy measure of serious learning difficulties in at least two subjects, classified as no/ yes), and absence from school (number of absence days derived from school registers, classified as no, 1 day or 2 days or more). Information on the adolescents’ school marks (teacher ratings) was obtained from the national application register for upper secondary education including vocational schools and upper secondary schools until age 16. These data were extracted until the year 1985 (Isohanni et al., 2000). The health risk behaviours measured at age 14 were smoking status, and alcohol consumption. Smoking status was assessed by a direct question ‘‘Do you smoke?’’ and classified in 8-point range from more than 10 cigarettes every day to nonsmokers. The participants were further classified as current smokers (once a week or more often) and non-smokers (others). In the assessment of alcohol consumption, the participants were asked to report their habitual frequency of consuming alcohol (never, merely tasted, occasionally, monthly and more often) (Rantakallio, 1983). The participants were then classified as non-drinkers, sometimes and regular users (monthly or more often). Work characteristics The scale of job control (Karasek et al., 1981) comprised nine items (e.g., ‘‘My job allows me to

ARTICLE IN PRESS M. Elovainio et al. / Social Science & Medicine 65 (2007) 187–199

make a lot of decisions on my own’’, ‘‘I have a lot of say about what happens on my job’’, ‘‘My job requires a high level of skill’’, ‘‘My job requires that I learn new things’’, ‘‘I get to do a variety of different things on my job’’, and ‘‘I have an opportunity to develop my own special abilities’’). Responses were given along a five-point Likert-type scale ranging from 1 ¼ strongly disagree to 5 ¼ strongly agree. The job control index was the sum of response scores (Cronbach’s alpha reliability (a) ¼ 0.87). Job demands scale (Karasek et al., 1981) comprised four items. (e.g., ‘‘My job requires working very fast.’’) Responses were given along a five-point Likert-(type) scale ranging from 1 ¼ strongly disagree to 5 ¼ strongly agree. The job demands index was the sum of response scores (Cronbach’s alpha reliability (a) ¼ 0.82). Psychological distress The dependent variable psychological distress was assessed by the SCL-25. It is a 25-item instrument assessing global psychological distress derived from the SCL-90 (Derogatis, Lipman, & Covi, 1973). The SCL-25 is composed of depression items (e.g. ‘‘Have you been distressed by feeling lonely?’’), somatization items (e.g. ‘‘Have you been distressed by feeling weak in your body?’’) and phobic anxiety items (e.g. ‘‘Have you been distressed by feeling tense or keyed-up?’’) from SCL90. Participants indicated how well each item described the psychological distress they experienced within the week prior to participation using 1 (not at all) to 4 (extremely) scale. A single global score was used as an index of psychological distress as recommended by Nguyen, Attkisson, and Stegner (1983). The predictive validity of the SCL25 has found to be good in terms of clinical response, clinical remission, and criterion symptom remission (Karlsson, Joukamaa, & Lehtinen, 2000; Williams, 2004). Statistical analysis We studied the relationships between pre-employment factors, work characteristics (job control, job demands and job strain) and psychological distress by linear regression analyses. Psychological distress at age 31 was the dependent variable. Psychological distress assessed by the SCL-25 was used as a

191

continuous variable. All statistical models were adjusted for gender. At the first step, all of the pre-employment factors, such as fathers’ socioeconomic status, birth weight, mothers’ depressive mood during the pregnancy, school grade, repeated class, absence from school, alcohol consumption and smoking were modeled as predictors of psychological distress at age 31separately (univariate analysis) and then simultaneously (multivariate analysis). In the second step, assessed the associations of pre-employment factors with work characteristics. In the third step, assessed the associations of job control, job demands, and job strain with psychological distress at age 31. In the fourth step, statistically significant associations between work characteristics and psychological distress were adjusted for pre-employment factors one by one and finally for all of the pre-employment factors at the same time. Finally we showed the model, where the all pre-employment factors and work characteristics acted all simultaneously (multivariate model) as predictors of psychological distress. Results were expressed with standardized regression coefficients (b). For all analyses, we used the SAS 9.1 statistical program package. Results Descriptive statistics of the study variables are presented in Table 1. There were slightly more men than women in the sample. Twenty-three percent of the mothers had experienced antenatal depressive symptoms and most fathers were classified as professionals or skilled workers. Repeated absences from school and especially repeating school classes were relatively rare. Larger share of the subjects experienced high than low job control. Similar result was obtained in job demands. In Table 2, the relationships between pre-employment factors and psychological distress at age 31 are reported. Female gender, mothers’ depressiveness during pregnancy, high absence rate from school (2 days or more), moderated and regular alcohol consumption and smoking at age 14 were significantly related to psychological distress at age 31. In the multivariate model only low mean academic score, high absence rate from school and moderate alcohol consumption were significantly related to psychological distress. Table 3 shows the multivariate relationships between pre-employment factors and work

ARTICLE IN PRESS 192

M. Elovainio et al. / Social Science & Medicine 65 (2007) 187–199

Table 1 Characteristics of respondents (N ¼ 4293–3994) N

%

Men Women

2231 2062

52 48

Childhood factors Mother’s antenatal depressed mood Birth weight (g) Father’s socioeconomic status Professional and skilled workers Unskilled workers, workers with no occupation and farmers

542 4293 3994 2253 1741

23

Adolescence factors (measured at age 14) Mean academic score Repeated class No One to three times Absence from school No absences One day Two or more High alcohol consumption Never Some times Smoking status Current smokers at age 14 Non smokers

4009 4124 4045 79 4101 2136 1455 510 4116 1676 2339 101 4124 1333 2791

Work characteristics (measured at age 31) Job control Job demands Job strain Psychological distress

4292 4292 4291 4249

characteristics at age 31. Male gender, father’s high socioeconomic status, and having high mean academic scores at school up to age 16 were all significantly related to high job control. Female gender, having high-grade mean and low probability of repeating class at school at age 14 were related high job demands. Only female gender and father’s low socioeconomic position were related to high job strain. Table 4 shows the results of regression analysis with work characteristics as independent and psychological distress at age 31 as the dependent variable. Because job demands were not associated with psychological distress (B ¼ 0.00, b ¼ 0.03, t ¼ 1.87, p ¼ 0.06), only models for job control and job strain are presented. After adjustment for gender the relationship between job control (B ¼ 0.07, b ¼ 0.15, t ¼ 9.66, po0.001) and distress, and between job strain and distress

Mean (SD)

Range

1.15 (0.40) 3506.8 (521.1) 1.90 (0.72)

1–3 1200–6080 1–3

56 44

7.74 (0.84)

4.5–9.9

2.56 3.03 1.24 1.31

0.50–4.00 0.36–4.00 0.22–5.09 1–4

98 2 52 36 12 41 57 2 32 78 (0.59) (0.61) (0.35) (0.29)

(B ¼ 0.13, b ¼ 0.16, t ¼ 9.79, po0.001) were statistically highly significant. Additional adjustments for the pre-employment childhood risk factors did not attenuate these relationships. Table 5 shows the results of regression analysis with all of the pre-employment factors and work characteristics as independent and psychological distress at age 31 as the dependent variable. Again only job control and job strain are presented as work characteristics. In the multivariate model, only absence from school and alcohol consumption at age 14 and the work characteristics, were statistically significantly associated with psychological distress at age 31. Also gender (B ¼ 0.08, b ¼ 0.15, t ¼ 8.40, po0.001) was significantly related to psychological distress even after adjusted for work characteristics. All the effects were clearly attenuated (see Table 2.), however, when adjusted for work characteristics.

ARTICLE IN PRESS M. Elovainio et al. / Social Science & Medicine 65 (2007) 187–199

193

Table 2 Regression (B) and standardized regression coefficients (b) of pre-employment risk factors on psychological distress in men and women Pre-employment factors

Univariate models

Multivariate models

B

b

t-value

p-value

B

b

t-value

p-value

0.09 0.04 0.00 0.01

0.16 0.05 0.00 0.03

10.3 3.56 1.03 1.69

o0.001 o0.001 0.301 0.092

0.10 0.02 0.00 0.01

0.18 0.03 0.01 0.02

14.47 1.82 0.55 0.93

o0.001 0.060 0.580 0.352

Adolescence factors Mean academic score Repeated class (no/yes)

0.00 0.04

0.02 0.02

0.95 1.36

0.340 0.175

0.00 0.07

0.04 0.03

2.35 1.71

0.019 0.083

Absence from schoola One Two or more

0.02 0.05

0.02 0.05

1.56 3.26

0.119 0.001

0.01 0.02

0.03 0.05

1.94 3.08

0.052 o0.001

Alcohol consumptiona Some times Regularly Current smokers at age 14 (no/yes)

0.02 0.06 0.02

0.04 0.03 0.03

2.95 2.16 2.04

0.003 0.031 0.042

0.03 0.05 0.01

0.06 0.03 0.01

3.16 1.51 0.42

o0.001 0.134 0.674

Childhood factors Gender (1 ¼ men, 2 ¼ women) Mother’s antenatal depressed mood Birth weight Father’s socioeconomic status

a

Never as a reference group.

Table 3 Regression (B) and standardized regression coefficients (b) of pre-employment risk factors on job factors, such as job control, job demands and job strain in men and women (multivariate models, all variables entered simultaneously) Pre-employment factors

Job control B

Gender*

b

Job demands t-value p-value B

0.20 0.18 9.96

Childhood factors Mother’s antenatal depressed mood 0.02 0.01 0.86 Birth weight 0.00 0.03 1.73 Father’s socioeconomic status 0.03 0.05 2.71

o0.001

Job strain t-value p-value B

b

t-value p-value

b

11.22

o0.001

0.392 0.04 0.03 1.64 0.084 0.00 0.00 0.16 0.004 0.01 0.01 0.50

0.105 0.00 0.00 0.01 0.874 0.00 0.00 0.20 0.617 0.02 0.05 2.78

0.996 0.843 0.006

o0.001 0.00 0.01 0.28 0.001 0.07 0.02 1.40

0.784 0.167

0.07

0.05

3.14

0.002

0.14

0.20

Adolescence factors Mean academic score Repeated class (no/yes)

0.01 0.13 7.27 0.12 0.03 1.58

o0.001 0.01 0.18 10.01 0.111 0.27 0.06 3.99

Absence from school One Two or more

0.02 0.02 1.15 0.01 0.01 0.48

0.253 0.02 0.02 1.00 0.632 0.03 0.02 1.07

0.311 0.282

0.20 0.54

0.832 0.588

Alcohol consumption Some times 0.02 0.02 0.98 Regularly 0.00 0.00 0.02 Current smokers ate age 14 (no/yes) 0.02 0.02 0.87

0.322 0.00 0.00 0.15 0.980 0.03 0.01 0.47 0.388 0.02 0.01 0.63

0.880 0.01 0.02 1.02 0.646 0.01 0.01 0.33 0.522 0.01 0.01 0.71

0.303 0.744 0.478

0.00 0.01

0.01 0.01

*1 ¼ men, 2 ¼ women.

Discussion This study extended prior evidence by applying a life course perspective to the associations between stressful working conditions and psychological

distress. Previously Harper et al. (2002) showed that childhood socioeconomic position was associated with adulthood psychosocial functioning, such as hostility and hopelessness. Our results suggest that work-related psychosocial factors are

ARTICLE IN PRESS M. Elovainio et al. / Social Science & Medicine 65 (2007) 187–199

194

Table 4 Regression (B) and standardized regression coefficients (b) of job factors on psychological distress in men and women Work characteristics

Adjusted factors in addition to sex

B

b

t-value

p-value

Job control

None Mother’s antenatal depressed mood Birth weight Father’s socioeconomic status Mean academic score Repeated class Absence from school Alcohol consumption Smoking (current smokers at age 14) All

0.07 0.07 0.07 0.07 0.07 0.07 0.08 0.07 0.07 0.07

0.15 0.15 0.15 0.16 0.15 0.15 0.15 0.15 0.15 0.14

9.66 9.61 9.67 9.84 9.34 9.59 9.61 9.63 9.64 8.56

o0.001 o0.001 o0.001 o0.001 o0.001 o0.001 o0.001 o0.001 o0.001 o0.001

Job strain

None Mother’s antenatal depressed mood Birth weight Father’s socioeconomic status Mean academic score Repeated class Absence from school Alcohol consumption Smoking (current smokers at age 14) All

0.13 0.13 0.13 0.13 0.13 0.13 0.13 0.13 0.13 0.12

0.16 0.16 0.16 0.16 0.16 0.16 0.16 0.16 0.16 0.15

9.79 9.79 9.79 9.82 9.84 9.86 9.73 9.76 9.81 9.11

o0.001 o0.001 o0.001 o0.001 o0.001 o0.001 o0.001 o0.001 o0.001 o0.001

Adjusted for pre-employment factors (N ¼ 3873).

Table 5 Regression (B) and standardized regression coefficients (b) of preemployment risk factors and job factors on psychological distress in men and women Risk factors B

Childhood factors Mother’s antenatal depressed mood Birth weight Father’s socioeconomic status Adolescence factors Mean academic score Repeated class (no/yes) Absence from school One Two or more Alcohol consumption Some times Regularly Smoking (current smokers at age 14) Work characteristics Job Control Job strain

t-value

b

0.02

0.03

p-value

1.67

0.091

0.00 0.01 0.55 0.00 0.01 0.51

0.582 0.601

0.00 0.03 1.82 0.07 0.03 1.76

0.065 0.070

0.02 0.05

0.03 0.05

1.98 3.17

0.044 o0.001

0.03 0.05 0.01

0.06 0.03 0.01

3.03 1.57 0.61

o0.001 0.111 0.547

0.04 0.08 3.72 0.08 0.10 4.86

0.003 o0.001

(N ¼ 3873) (adjusted for gender).

one of pathways between childhood conditions and emotional disorders. In addition, we were able to test the effects of a wide variety of childhood and adolescent risk factors, in addition to childhood SES, on adulthood emotional disorders. Our results supported and extended prior evidence on the relationship between job strain and psychological distress (Power et al., 2002) by showing that this relationship is not confounded by a variety of pre-employment risk factors for adult psychological distress. Those who scored low on job control and high on job strain were at higher risk for the development of psychological distress than others and that this result is not explained by mother’s depressive mood during pregnancy, low birth weight, parental socioeconomic status, or early markers of social exclusion, such as problems at school and health risk behaviour. However, the results of the present study suggest that some of the early childhood risk factors select people to risky adulthood environments, such as low control and high strain jobs (Karasek et al., 1981). Especially female gender, father’s low socioeconomic position and poor academic achievements in adolescence seemed to predict low job control and high job strain later in life. Taking into account the later work characteristics attenuated the associations between early risk factors and

ARTICLE IN PRESS M. Elovainio et al. / Social Science & Medicine 65 (2007) 187–199

psychological distress, but some of the effects remained significant. This suggests that part of the effects of pre-employment risk factors may be mediated by job characteristics. The life course perspective on the aetiology of various health problems including affective disorders emphasises either the importance of early life or adult behaviours (Kuh, Ben-Shlomo, Lynch, Hallqvist, & Power, 2003). The early origins hypothesis (critical period model) suggests that early psychosocial circumstances have long-lasting effects on the structure or function of the psychological and physiological development (Lynch et al., 1997; Power, Matthews, & Manor, 1998). According to this idea, childhood represents a critical developmental period and experiences during that period have negative or positive consequences that persist to adulthood and later life. This model implies that effects of childhood social circumstances are relatively independent of adult circumstances. Our results did not offer strong support for this idea. It is also possible to suggest that childhood circumstances represent risks for ones mental health only if they select one to poor adulthood circumstances, such as low job control or high job strain (the pathway model) (Marmot & Wilkinson, 2001). According to this model, adult psychosocial environment contributes to mental health risk independently of childhood circumstances. According to our results, early childhood factors predicted adulthood work characteristics and weaker effects between childhood factors and psychological distress was detected when adulthood work characteristics were entered in the models. Thus, the results of the present study provide some support for the pathway hypothesis. It is also suggested that psychosocial risks during life simply have an additive effect on ones health (Davey Smith et al., 1998). This line of reasoning got also some support from the present study. Both pre-employment risk factors in childhood and the negative work characteristics in adulthood, such as low job control and high strain were independently related to later psychological distress. Although this is not an sufficient indicator of an additive effect, it suggest that risk factors measured in various stages associate with psychological distress and that these associations are not attributable to not any single group of risk factors. In interpreting the present results, it is important to note some limitations. Although the design of the study is longitudinal

195

with a large cohort, the sample was restricted to men and women aged 31. Thus our findings may not be generalized to the Finnish population. We measured both job characteristics and psychological distress using self-reports. While psychiatric interviews would provide a more valid identification of psychiatric disorders at a given point of time (Eaton, Neufeld, Chen, & Cai, 2000; Murphy & Schachar, 2000), such methods are impractical for large cohorts and long follow-up periods. In observational studies, confounding, reversed causality and selection bias are potential threats to validity (Kraemer, Stice, Kazdin, Offord, & Kupfer, 2001). In the present study, we adjusted all findings for demographic and socioeconomic background to minimize the possibility of confounding. It is also possible that we did not choose the right preemployment variables although the ones used in the current study have all been related to health-related outcomes previously. Variables, such as personality, cognitive capacity, social support, living conditions and peer influence may be among those that should be studied in the area of life course epidemiology. In addition, future studies should use birth weight corresponding to gestational age as a measure of birth weight. This may be one of reasons we could not find association between birth weight and psychological distress, although similar results to ours have been gained before. Although we did find a stronger relationship between work characteristics and psychological distress than that of between preemployment factors and psychological distress, this difference may be due the fact that the relationship between work characteristics and psychological distress was cross-sectional. However, the associations were not considerably stronger between work characteristics and distress than those between preemployment factors and psychological distress in univariate models. Although the original sample attrition was not significant (except in gender) it is possible that those with highly educated may have more likely to have been enrolled. This may have reduced the variation in later socioeconomic status/ work characteristics and probably caused less significant associations between later risk factors and psychological distress. One reason for small number of previous studies on the relationship between work conditions as mental health risk is due to the fact that the protective aspects of work may mask effects of work-related risks. Major psychiatric disorders are a cause of work disability and a potential selective

ARTICLE IN PRESS 196

M. Elovainio et al. / Social Science & Medicine 65 (2007) 187–199

factor for unemployment (Goldberg & Steury, 2001; Lecrubier, 2000). Unemployment, in turn, is predictive of psychological distress, increased suicide rates and mortality due to external causes (Kivimaki, Elovainio, Kokko et al., 2003; Voss, Floderus, & Diderichsen, 2001) and does not prevent isolation among depressive individuals as employment might do. However, measurements within working populations and sensitive to minor psychiatric disorders reveal that work might include risk factors for mental health (Kivimaki et al., 2003). In conclusion, our study of Finns at age 31 lends support for the effects of factors related to adulthood work characteristics on adulthood psychological distress. The large data set provided us with the possibility simultaneously to test the relative effect of other known risk factors and possible mediators as well. Our results do not offer strong support for the idea of critical period behind the development of psychological distress. In contrast, our results strengthen the view that work characteristics such as low job control and high job demands have an important role in development psychological distress among young adults. The identification of low job control and high job strain as common pathways between adverse childhood/adolescent conditions and adulthood emotional disorders has substantial public health relevance for turning life course epidemiology into practice. Interventions that target important pathways, such as job characteristics, have the potential to reduce morbidity related to multiple conditions in multiple stages of life. Acknowledgement The study was supported by grants from the Academy of Finland (Projects 105195, 105168 and 117604). References Adler, N. E., Boyce, T., Chesney, M. A., Cohen, S., Folkman, S., Kahn, R. L., et al. (1994). Socioeconomic status and health. The challenge of the gradient. American Psychologist, 49(1), 15–24. Anda, R. F., Williamson, D. F., Escobedo, L. G., Mast, E. E., Giovino, G. A., & Remington, P. L. (1990). Depression and the dynamics of smoking. A national perspective. Journal of American Medical Association, 264(12), 1541–1545. Barker, D. J. (1991). The foetal and infant origins of inequalities in health in Britain. Journal of Public Health Medicine, 13(2), 64–68.

Barker, D. J., Forsen, T., Uutela, A., Osmond, C., & Eriksson, J. G. (2001). Size at birth and resilience to effects of poor living conditions in adult life: Longitudinal study. British Medical Journal, 323(7324), 1273–1276. Bourbonnais, R., Comeau, M., & Vezina, M. (1999). Job strain and evolution of mental health among nurses. Journal of Occupational Health Psychology, 4(2), 95–107. Bradley, R. H., & Corwyn, R. F. (2002). Socioeconomic status and child development. Annual Review of Psychology, 53, 371–399. Brooks-Gunn, J., Klebanov, P. K., Liaw, F., & Spiker, D. (1993). Enhancing the development of low-birthweight, premature infants: changes in cognition and behavior over the first three years. Child Development, 64(3), 736–753. Brown, D. R., Croft, J. B., Anda, R. F., Barrett, D. H., & Escobedo, L. G. (1996). Evaluation of smoking on the physical activity and depressive symptoms relationship. Medicine and Science in Sports and Exercise, 28(2), 233–240. Bruce, M. L. (2002). Psychosocial risk factors for depressive disorders in late life. Biological Psychiatry, 52(3), 175–184. Cheung, Y. B., Low, L., Osmond, C., Barker, D., & Karlberg, J. (2000). Fetal growth and early postnatal growth are related to blood pressure in adults. Hypertension, 36(5), 795–800. Cheung, Y. B., Ma, S., Machin, D., & Karlberg, J. (2004). Birth weight and psychological distress in adult twins: A longitudinal study. Acta Paediatrica, 93(7), 965–968. Cianfarani, S., Bonfanti, R., Bitti, M. L., Germani, D., Boemi, S., Chiumello, G., et al. (2000). Growth and insulin-like growth factors (IGFs) in children with insulin-dependent diabetes mellitus at the onset of disease: evidence for normal growth, age dependency of the IGF system alterations, and presence of a small (approximately 18-kilodalton) IGF-binding protein-3 fragment in serum. Journal of Clinical Endocrinology and Metabolism, 85(11), 4162–4167. Davey Smith, G., Hart, C., Hole, D., MacKinnon, P., Gillis, C., Watt, G., et al. (1998). Education and occupational social class: Which is the more important indicator of mortality risk? Journal of Epidemiology and Community Health, 52(3), 153–160. de Lange, A. H., Taris, T. W., Kompier, M. A., Houtman, I. L., & Bongers, P. M. (2003). The very best of the millennium’’: Longitudinal research and the demand-control-(support) model. Journal of Occupational Health Psychology, 8(4), 282–305. Derogatis, L. R., Lipman, R. S., & Covi, C. (1973). SCL-90: An outpatient psychiatric rating scale—preliminary report. Psychopharmacology Bulletin, 9, 13–27. Eaton, W. W., Neufeld, K., Chen, L. S., & Cai, G. (2000). A comparison of self-report and clinical diagnostic interviews for depression: Diagnostic interview schedule and schedules for clinical assessment in neuropsychiatry in the Baltimore epidemiologic catchment area follow-up [see comment]. Archives of General Psychiatry, 57(3), 217–222. Gale, C. R., & Martyn, C. N. (2004). Birth weight and later risk of depression in a national birth cohort. British Journal of Psychiatry, 184, 28–33. Goldberg, R. J., & Steury, S. (2001). Depression in the workplace: Costs and barriers to treatment. Psychiatric Services, 52(12), 1639–1643. Haarasilta, L. M., Marttunen, M. J., Kaprio, J. A., & Aro, H. M. (2004). Correlates of depression in a representative nationwide sample of adolescents (15–19 years) and young adults

ARTICLE IN PRESS M. Elovainio et al. / Social Science & Medicine 65 (2007) 187–199 (20–24 years). European Journal of Public Health, 14(3), 280–285. Hallqvist, J., Lynch, J., Bartley, M., Lang, T., & Blane, D. (2004). Can we disentangle life course processes of accumulation, critical period and social mobility? An analysis of disadvantaged socio-economic positions and myocardial infarction in the Stockholm Heart Epidemiology Program. Social Science & Medicine, 58(8), 1555–1562. Halsey, C. L., Collin, M. F., & Anderson, C. L. (1996). Extremely low-birth-weight children and their peers. A comparison of school-age outcomes. Archives of Pediatrics and Adolescent Medicine, 150(8), 790–794. Harper, S., Lynch, J., Hsu, W. L., Everson, S. A., Hillemeier, M. M., Raghunathan, T. E., et al. (2002). Life course socioeconomic conditions and adult psychosocial functioning [see comment]. International Journal of Epidemiology, 31(2), 395–403. Hellgren, L., Gillberg, I. C., Bagenholm, A., & Gillberg, C. (1994). Children with deficits in attention, motor control and perception (DAMP) almost grown up: Psychiatric and personality disorders at age 16 years. Journal of Child Psychology and Psychiatry and Allied Disciplines, 35(7), 1255–1271. Holland, P., Berney, L., Blane, D., Smith, G. D., Gunnell, D. J., & Montgomery, S. M. (2000). Life course accumulation of disadvantage: Childhood health and hazard exposure during adulthood. Social Science & Medicine, 50(9), 1285–1295. Huurre, T., Aro, H., & Rahkonen, O. (2003). Well-being and health behaviour by parental socioeconomic status: A followup study of adolescents aged 16 until age 32 years. Social Psychiatry and Psychiatric Epidemiology, 38(5), 249–255. Isohanni, I., Ja¨rvelin, M. R., Nieminen, P., Jones, P., Rantakallio, P., Jokelainen, J., et al. (1998). School performance as a predictor of psychiatric hospitalization in adult life. A 28year follow-up in the Northern Finland 1966 Birth Cohort. Psychological Medicine, 28(4), 967–974. Isohanni, M., Jones, P., Kemppainen, L., Croudace, T., Isohanni, I., Veijola, J., et al. (2000). Childhood and adolescent predictors of schizophrenia in the Northern Finland 1966 birth cohort—A descriptive life-span model. European Archives of Psychiatry and Clinical Neuroscience, 250(6), 311–319. Isohanni, M., Jones, P. B., Moilanen, K., Rantakallio, P., Veijola, J., Oja, H., et al. (2001). Early developmental milestones in adult schizophrenia and other psychoses. A 31-year follow-up of the Northern Finland 1966 Birth Cohort. Schizophrenia Research, 52(1–2), 1–19. Kandel, D. B., Johnson, J. G., Bird, H. R., Weissman, M. M., Goodman, S. H., Lahey, B. B., et al. (1999). Psychiatric comorbidity among adolescents with substance use disorders: Findings from the MECA Study. Journal of the American Academy of Child and Adolescent Psychiatry, 38(6), 693–699. Karasek, R. (1990). Lower health risk with increased job control among white collar workers. Journal of Organizational Behavior, 11(3), 171–185. Karasek, R., Baker, D., Marxer, F., Ahlbom, A., & Theorell, T. (1981). Job decision latitude, job demands, and cardiovascular disease: A prospective study of Swedish men. American Journal of Public Health, 71(7), 694–705. Karlsson, H., Joukamaa, M., & Lehtinen, V. (2000). Differences between patients with identified and not identified psychiatric

197

disorders in primary care. Acta Psychiatrica Scandinavica, 102(5), 354–358. Kendler, K. S., Kessler, R. C., Walters, E. E., MacLean, C., Neale, M. C., Heath, A. C., et al. (1995). Stressful life events, genetic liability, and onset of an episode of major depression in women. American Journal of Psychiatry, 152(6), 833–842. Kendler, K. S., Neale, M. C., Kessler, R. C., Heath, A. C., & Eaves, L. J. (1993). Major depression and phobias: The genetic and environmental sources of comorbidity. Psychological Medicine, 23(2), 361–371. Kivimaki, M., Elovainio, M., Kokko, K., Pulkkinen, L., Kortteinen, M., & Tuomikoski, H. (2003). Hostility, unemployment and health status: Testing three theoretical models. Social Science & Medicine, 56(10), 2139–2152. Kivimaki, M., Elovainio, M., Vahtera, J., Virtanen, M., & Stansfeld, S. A. (2003). Association between organizational inequity and incidence of psychiatric disorders in female employees. Psychological Medicine, 33(2), 319–326. Kivimaki, M., Smith, G. D., Juonala, M., Ferrie, J. E., Keltikangas-Jarvinen, L., Elovainio, M., et al. (2006). Socioeconomic position in childhood and adult cardiovascular risk factors, vascular structure, and function: Cardiovascular risk in young Finns study. Heart, 92(4), 474–480. Kouvonen, A., Kivimaki, M., Elovainio, M., Virtanen, M., Linna, A., & Vahtera, J. (2005). Job strain and leisure-time physical activity in female and male public sector employees. Preventive Medicine, 41(2), 532–539. Kraemer, H. C., Stice, E., Kazdin, A., Offord, D., & Kupfer, D. (2001). How do risk factors work together? Mediators, moderators, and independent, overlapping, and proxy risk factors. American Journal of Psychiatry, 158(6), 848–856. Kuh, D., Ben-Shlomo, Y., Lynch, J., Hallqvist, J., & Power, C. (2003). Life course epidemiology [erratum appears in J Epidemiol Community Health. 2003 Nov;57(11):914]. Journal of Epidemiology and Community Health, 57(10), 778–783. Landsbergis, P. A., Schnall, P. L., Deitz, D. K., Warren, K., Pickering, T. G., & Schwartz, J. E. (1998). Job strain and health behaviors: Results of a prospective study. American Journal of Health Promotion, 12(4), 237–245. Lecrubier, Y. (2000). Depressive illness and disability. European Neuropsychopharmacology, 10(Suppl. 4), S439–S443. Lynch, J. W., Kaplan, G. A., & Shema, S. J. (1997). Cumulative impact of sustained economic hardship on physical, cognitive, psychological, and social functioning. New England Journal of Medicine, 337(26), 1889–1895. Ma¨ki, P., Veijola, J., Rasanen, P., Joukamaa, M., Valonen, P., Jokelainen, J., et al. (2003). Criminality in the offspring of antenatally depressed mothers: A 33-year follow-up of the Northern Finland 1966 Birth Cohort. Journal of Affective Disorders, 74(3), 273–278. Marmot, M. G. (1997). Early life and adult disorder: Research themes. British Medical Bulletin, 53(1), 3–9. Marmot, M., & Wilkinson, R. G. (2001). Psychosocial and material pathways in the relation between income and health: A response to Lynch et al. British Medical Journal, 322(7296), 1233–1236. Matthews, S., & Power, C. (2002). Socio-economic gradients in psychological distress: A focus on women, social roles and work-home characteristics. Social Science & Medicine, 54(5), 799–810.

ARTICLE IN PRESS 198

M. Elovainio et al. / Social Science & Medicine 65 (2007) 187–199

Matthews, S., Power, C., & Stansfeld, S. A. (2001). Psychological distress and work and home roles: A focus on socio-economic differences in distress. Psychological Medicine, 31(4), 725–736. Matthews, S., Stansfeld, S., & Power, C. (1999). Social support at age 33: The influence of gender, employment status and social class. Social Science & Medicine, 49(1), 133–142. Mausner-Dorsch, H., & Eaton, W. W. (2000). Psychosocial work environment and depression: Epidemiologic assessment of the demand-control model [erratum appears in Am J Public Health 2001 May;91(5):828]. American Journal of Public Health, 90(11), 1765–1770. Merikangas, K. R., Dierker, L. C., & Szatmari, P. (1998). Psychopathology among offspring of parents with substance abuse and/or anxiety disorders: A high-risk study. Journal of Child Psychology and Psychiatry and Allied Disciplines, 39(5), 711–720. Murphy, P., & Schachar, R. (2000). Use of self-ratings in the assessment of symptoms of attention deficit hyperactivity disorder in adults. American Journal of Psychiatry, 157(7), 1156–1159. Nguyen, T. D., Attkisson, C. C., & Stegner, B. L. (1983). Assessment of patient satisfaction: Development and refinement of a service evaluation questionnaire. Evaluation and Program Planning, 6(3–4), 299–313. Niedhammer, I., Goldberg, M., Leclerc, A., David, S., Bugel, I., & Landre, M. F. (1998). Psychosocial work environment and cardiovascular risk factors in an occupational cohort in France. Journal of Epidemiology and Community Health, 52(2), 93–100. Ogdendahl, B. K., Agerbo, E., Byrne, M., Licht, R. W., Eaton, W. W., & Mortensen, P. B. (2006). Indicators of fetal growth and bipolar disorder: A Danish national register-based study. Psychological Medicine, 36(9), 1219–1224. Osborn, A. F., & Milbank, J. E. (1989). The effects of early education. Manchester: Blackwell. Patton, G. C., Coffey, C., Carlin, J. B., Olsson, C. A., & Morley, R. (2004). Prematurity at birth and adolescent depressive disorder. British Journal of Psychiatry, 184, 446–447. Power, C., & Matthews, S. (1997). Origins of health inequalities in a national population sample. Lancet, 350(9091), 1584–1589. Power, C., Matthews, S., & Manor, O. (1996). Inequalities in self rated health in the 1958 birth cohort: Lifetime social circumstances or social mobility? British Medical Journal, 313(7055), 449–453. Power, C., Matthews, S., & Manor, O. (1998). Inequalities in selfrated health: Explanations from different stages of life. Lancet, 351(9108), 1009–1014. Power, C., Stansfeld, S. A., Matthews, S., Manor, O., & Hope, S. (2002). Childhood and adulthood risk factors for socioeconomic differentials in psychological distress: Evidence from the 1958 British birth cohort. Social Science & Medicine, 55(11), 1989–2004. Pulkki, L., Kivimaki, M., Elovainio, M., Viikari, J., & Keltikangas-Jarvinen, L. (2003). Contribution of socioeconomic status to the association between hostility and cardiovascular risk behaviors: A prospective cohort study [see comment]. American Journal of Epidemiology, 158(8), 736–742. Pulkki, L., Kivimaki, M., Keltikangas-Jarvinen, L., Elovainio, M., Leino, M., & Viikari, J. (2003). Contribution of adolescent and early adult personality to the inverse associa-

tion between education and cardiovascular risk behaviours: Prospective population-based cohort study [see comment]. International Journal of Epidemiology, 32(6), 968–975. Pulkki-Raback, L., Elovainio, M., Kivimaki, M., Raitakari, O. T., & Keltikangas-Jarvinen, L. (2005). Temperament in childhood predicts body mass in adulthood: The cardiovascular risk in young Finns study. Health Psychology, 24(3), 307–315. Rantakallio, P. (1979). Social background of mothers who smoke during pregnancy and influence of these factors on the offspring. Social Science & Medicine, 13A(4), 423–429. Rantakallio, P. (1983). Family background to and personal characteristics underlying teenage smoking. Background to teenage smoking. Scandinavian Journal of Social Medicine, 11(1), 17–22. Rantakallio, P. (1988). The longitudinal study of the northern Finland birth cohort of 1966. Paediatric and Perinatal Epidemiology, 2(1), 59–88. Rantakallio, P., & Koiranen, M. (1987). Neurological handicaps among children whose mothers smoked during pregnancy. Preventive Medicine, 16(5), 597–606. Riala, K., Hakko, H., Isohanni, M., Jarvelin, M. R., & Rasanen, P. (2004). Teenage smoking and substance use as predictors of severe alcohol problems in late adolescence and in young adulthood. Journal of Adolescent Health, 35(3), 245–254. Stansfeld, S. A., Fuhrer, R., Head, J., Ferrie, J., & Shipley, M. (1997). Work and psychiatric disorder in the Whitehall II Study. Journal of Psychosomatic Research, 43(1), 73–81. Stansfeld, S., Head, J., & Ferrie, J. (1999). Short-term disability, sickness absence, and social gradients in the Whitehall II Study. International Journal of Law and Psychiatry, 22(5–6), 425–439. Sweeting, H., & West, P. (1995). Family life and health in adolescence: A role for culture in the health inequalities debate? Social Science & Medicine, 40(2), 163–175. Vermeulen, M., & Mustard, C. (2000). Gender differences in job strain, social support at work, and psychological distress. Journal of Occupational Health Psychology, 5(4), 428–440. Voss, M., Floderus, B., & Diderichsen, F. (2001). Physical, psychosocial, and organisational factors relative to sickness absence: A study based on Sweden Post. Occupational and Environmental Medicine, 58(3), 178–184. Wadsworth, M. E. (1997). Health inequalities in the life course perspective. Social Science & Medicine, 44(6), 859–869. Wall, T. D., Bolden, R. I., Borrill, C. S., Carter, A. J., Golya, D. A., Hardy, G. E., et al. (1997). Minor psychiatric disorder in NHS trust staff: Occupational and gender differences [see comment]. British Journal of Psychiatry, 171, 519–523. Wamala, S. P., Lynch, J., & Kaplan, G. A. (2001). Women’s exposure to early and later life socioeconomic disadvantage and coronary heart disease risk: The Stockholm female coronary risk study. International Journal of Epidemiology, 30(2), 275–284. Wang, J. (2004). A longitudinal population-based study of treated and untreated major depression. Medical Care, 42(6), 543–550. Weinberg, A., & Creed, F. (2000). Stress and psychiatric disorder in healthcare professionals and hospital staff. Lancet, 355(9203), 533–537. Wilkinson, R. G. (1997). Socioeconomic determinants of health. Health inequalities: relative or absolute material standards? BMJ, 314(7080), 591–595.

ARTICLE IN PRESS M. Elovainio et al. / Social Science & Medicine 65 (2007) 187–199 Wilkinson, R. G., & Pickett, K. E. (2006). Income inequality and population health: A review and explanation of the evidence. Social Science & Medicine, 62(7), 1768–1784. Williams, G. (2004). Management of patients who have relapses in multiple sclerosis. British Journal of Nursing, 13(17), 1012–1016. Williams, R. B., Barefoot, J. C., Blumenthal, J. A., Helms, M. J., Luecken, L., Pieper, C. F., et al. (1997). Psychosocial

199

correlates of job strain in a sample of working women. Archives of General Psychiatry, 54(6), 543–548. Ylipaavalniemi, J., Kivimaki, M., Elovainio, M., Virtanen, M., Keltikangas-Jarvinen, L., & Vahtera, J. (2005). Psychosocial work characteristics and incidence of newly diagnosed depression: A prospective cohort study of three different models. Social Science & Medicine, 61(1), 111–122.