You are here

  • Home
  • Archive
  • Volume 65, Issue 10
  • Is the association between job strain and carotid intima-media thickness attributable to pre-employment environmental and dispositional factors? The Cardiovascular Risk in Young Finns Study

Article Text

Article menu

Download PDFPDF

Original article
Is the association between job strain and carotid intima-media thickness attributable to pre-employment environmental and dispositional factors? The Cardiovascular Risk in Young Finns Study
  1. T Hintsa1,
  2. M Kivimäki1,2,
  3. M Elovainio3,
  4. J Vahtera4,
  5. M Hintsanen1,
  6. J S A Viikari4,
  7. O T Raitakari5,
  8. L Keltikangas-Järvinen1
  1. 1
    Department of Psychology, University of Helsinki, Finland
  2. 2
    Department of Epidemiology and Public Health, University College London Medical School, London, UK
  3. 3
    National Research and Development Centre for Welfare and Health (STAKES), Helsinki, Finland
  4. 4
    Finnish Institute of Occupational Health, Turku, Finland
  5. 5
    Department of Medicine, University of Turku, Finland
  6. 6
    Department of Clinical Physiology, University of Turku, Finland
  1. Taina Hintsa, Department of Psychology, PO Box 9, FIN-00014 University of Helsinki, Finland; taina.hintsa{at}helsinki.fi

Abstract

Objectives: Most previous studies of job strain and cardiovascular risk have been limited to adult data. It remains unclear whether this association might be explained by factors already present before entering the labour market. This study examined whether pre-employment family factors and participants’ own dispositional factors contribute to the relationship between job strain and carotid intima-media thickness (CIMT) among male employees.

Methods: The sample consisted of 494 men from the Cardiovascular Risk in Young Finns Study. Parental socioeconomic position and parental life dissatisfaction were assessed at 9–21 years of age and components of type A behaviour (Hunter-Wolf) were assessed at 12–24 years of age before the participants had entered the labour market. Job strain, education and CIMT were assessed at 27–39 years of age when all participants were employed.

Results: There was an association between higher job strain and increased CIMT in adulthood (mean 0.59 mm; 95% CI 0.42 to 0.76) which was only slightly affected on adjustment for parental socioeconomic position and parental life dissatisfaction as well as participants’ education. However, the job strain/CIMT relationship attenuated by 17% to non-significant after taking into account the effect of the participants’ type A behaviour components.

Conclusions: In this contemporary cohort of men, lack of leadership (a type A behaviour component) contributed to the association between job strain and CIMT 15 years later, whereas pre-employment family factors had only a modest effect on this association.

https://doi.org/10.1136/oem.2007.037622

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    Job strain is suggested to be a risk factor for cardiac events or coronary heart disease.1 2 Recently, an association between job strain and an early atherosclerotic process, as indicated by carotid intima-media thickness (CIMT), has been documented in men aged under 40 years of age.3 Considering that atherosclerosis develops over a long period of time and that by 40 years of age people have had a relatively short working life, the question of confounding by pre-employment environmental and dispositional factors is raised. However, as most previous studies have been limited to adult data, it remains unclear whether the association between job strain and cardiovascular risk is explained by environmental and dispositional factors already present before entering the labour market that affect both job strain and coronary health.

    Although work stress is suggested to originate mainly from work,4 it is also possible that work stress is partly affected by pre-existing factors. For example, previous literature has found a link between earlier neuroticism and a tendency to report stress5 and between pre-employment family factors and adult work stress.4 6 7

    Previous studies of cardiovascular disease have examined personality characteristics present before individuals enter the labour market. In the late 1950s, Friedman and Rosenman found that a type A behaviour pattern, characterised by feelings of urgency, competitiveness and anger/hostility, was associated with increased cardiovascular disease risk.8 Prospective investigations in the 1970s, such as the Western Collaborative Group study9 and the Framingham Heart study,10 provided further support for this finding, but subsequent failure to replicate these results in the 1980s11–13 focused interest on the components of type A behaviour instead of the global rating.14 Among adolescents, a type A behaviour pattern has been found to consist of several components, such as aggression, leadership, assumption of responsibility and eagerness-energy.15 16 A recent study from the population-based Cardiovascular Risk in Young Finns (CRYF) cohort showed that the eagerness-energy component of the type A behaviour pattern measured at 12–24 years of age was positively associated with CIMT among men aged 27–39.17

    Other pre-employment factors may also be important. Low socioeconomic position has been associated with job strain.18 A negative family atmosphere is related to less schooling19 and to the hostility component of type A behaviour among adult men.20 Furthermore, maternal life dissatisfaction has been related to the total type A behaviour score in adolescence.21 Because pre-employment family and dispositional factors, including family history of coronary heart disease, may directly or indirectly affect stress-related exposures and perceptions in adulthood,6 19 21 22 these factors may contribute to the relationship between job strain and early atherosclerosis.

    We examined the role of environmental and dispositional factors already present before entering the labour market in the relationship between job strain and CIMT in men because the job strain/CIMT relationship has previously been observed only in men in CRYF data.3 Among the men participating in the CRYF study, job strain was associated with increased CIMT even when a family history of coronary heart disease and childhood and adulthood biological coronary risk factors were taken into account.3 23 However, the effects of developmental context and emotional support indicated by socioeconomic conditions, family emotional atmosphere and personality factors on this relationship have not been examined so far. In this study, we hypothesised that pre-employment environmental and dispositional factors, such as parental socioeconomic position, parental life satisfaction, family history of coronary heart disease and participants’ type A behaviour components, would contribute to the job strain/CIMT relationship.

    METHODS

    Participants

    The Cardiovascular Risk in Young Finns Study is an ongoing follow-up study of coronary heart disease risk factors in Finnish children, adolescents and young adults.24 The first cross-sectional study was conducted in 1980 when cohorts of 3-, 6-, 9-, 12-, 15- and 18-year-olds were randomly sampled on the basis of social security numbers, resulting in a total of 3596 participants. Approximately 34% of participants have been lost after 21 years of follow-up due to lack of time, unwillingness to participate or absence from place of residence at the time of assessment.25

    In the present study, we included only men, as the relationship between job strain and CIMT is found in men but not in women in the Young Finns study.3 As the youngest cohort (n = 295) was excluded because they did not self-report type A behaviour components, the eligible population consisted of 1469 men (ie, 83% of all male participants at baseline). Furthermore, because we were examining pre-employment effects on the job strain/CIMT relationship, the inclusion criteria specified that the participants should not be working full time in 1986 when the pre-employment factors were assessed but should be employed in 2001 at the time job strain and CIMT were measured. Therefore, we excluded participants already working full time in 1986 (n = 278), those who were not employed in 2001 (n = 145) and those with missing information on current employment in 2001 (n = 552). Thus, the final cohort comprised 494 men.

    pre-employment factors

    Participants’ education was measured as completed school years. Parental socioeconomic position was measured in 1983 as household income and school years completed by the mother and father. Parental life satisfaction was measured in 1983 using a scale taken from the Operation Family Study questionnaire,26 the participants being 6–21 years of age. The questionnaire covered both maternal and paternal satisfaction as a mother/father, as a spouse and with her/his work role. The average score was calculated to form an index of parental life satisfaction, with larger values indicating higher levels of life satisfaction. The scale reliability (Cronbach’s α) was 0.72 for maternal life satisfaction and 0.76 for paternal life satisfaction. Family history of coronary heart disease was assessed using baseline and follow-up data. Family history was considered positive if either the participant’s father or mother had been diagnosed with coronary heart disease, had a myocardial infarction, or had percutaneous coronary intervention or coronary by-pass surgery at or before the age of 55 years.23

    Components of type A behaviour

    Components of type A behaviour were self-rated with the Hunter-Wolf A–B Rating Scale (HWolf)16 27 in 1986. The type A behaviour scale consists of 23 items and four subscales: (1) impatience-aggression, seven items, eg, “I lose my temper easily”, (2) leadership, six items, eg, “I am always a leader in activities”, (3) hard-driving, three items, eg, “I am hard-driving and competitive”, and (4) eagerness-energy, seven items, eg, “I always feel hurried”. The response was marked using a seven-point scale in a form of a ladder (put an “X” on the step you are on). We calculated the means of response scores across the scale items, with higher values indicating a higher level of type A behaviour in each subscale. Cronbach’s α reliabilities of the subscales vary between 0.55 and 0.69. The HWolf has been documented as being especially suitable for measuring type A behaviour in adolescents.28

    Job strain

    Job control was measured using the Job Content Questionnaire,29 which contains nine items, eg, “I can make independent decisions at my work” (1 = agree, 5 = do not agree) (Cronbach’s α 0.85). Job demands (α 0.59) were measured using a three-item scale from the Occupational Stress Questionnaire,30 which has been validated in Finland in 25 000 employees. The items included “Does your work have phases that are too difficult?”, “I must hurry to get my work done” and “Is your work mentally straining?”. The responses were given on a scale ranging from 1 = strongly disagree to 5 = strongly agree. The continuous linear job strain indicator was obtained by subtracting demands from control.31 Other possible job strain formulations suggested in the literature include a quotient term (demand score divided by control score), a tertile term (distributions of demands and control divided into tertiles, then the highest two tertiles of demands combined with the lowest two tertiles of control form the high strain category, and the lowest two tertiles of demands combined with the highest two tertiles of control form the low strain category, while all other combinations form the intermediate strain category), a quadrant term (dichotomising demands and control at the medians, low demands and high control forming the strain category and all other combinations a no-strain category), and a multiplicative interaction term (control×demands).31 Of these job strain formulations, the tertile, linear and quotient terms of job strain have been reported to be associated with CIMT in men, but not the quadrant term or the multiplicative interaction term.3

    CIMT

    The measurement of CIMT has previously been described in detail.25 In brief, CIMT was measured using Sequoia 512 ultrasound mainframes (Acuson, Mountain View, CA) with 13.0 MHz linear array transducers. The left common carotid artery was scanned by ultrasound technicians according to a standardised protocol. A minimum of four measurements of the common carotid wall were taken to derive mean CIMT. The between-observer coefficient of variation was 5.2% and the between-visit coefficient of variation was 6.4%.25 The measurements were carried out between September 2001 and January 2002.

    Statistical analyses

    The differences between participants and excluded individuals were examined using t tests, and the associations between predictor variables were studied by calculating Pearson correlations.

    In multivariate analyses, the number of participants with full information on all study variables can be small relative to the actual sample size. In addition, considering that the participants were required not to be working in 1986 but to be working in 2001, the number of participants included was very low. Therefore, a missing value analysis was carried out, and the missing data were completed by the expectation-maximisation (EM algorithm) method.32 The EM algorithm has two steps. First, in the estimation step the missing data are imputed by predicted scores in a series of regressions where every missing variable is regressed on the remaining variables for a particular case. Second, in the maximisation step, the imputed data set is submitted for maximum likelihood estimation.33 The two steps are repeated until a stable solution is obtained. This method has been considered to produce unbiased estimates of the mean and standard error (SE).34 35 We imputed 9–30% of the data depending on the variable (table 1). An age adjusted job strain/CIMT relationship exists in unimputed data among 478 men with measurements of both job strain and CIMT.3

    View this table:
    Table 1 Sample characteristics

    The relationships between pre-employment factors and job strain were studied using linear regression analyses controlling for age and education. To examine the contribution of parental socioeconomic position, life satisfaction and participants’ type A behaviour components to the relationship between job strain and CIMT, a series of multivariate linear regression analyses were carried out using pre-employment factors as covariates. Statistical analyses were carried out with SPSS statistical software v 15.0.

    RESULTS

    The participants were not different from excluded individuals with respect to biological coronary risk factors in 1980, such as body mass index (BMI), HDL- and LDL-cholesterol, triglycerides, and systolic and diastolic blood pressure (all p values >0.05). Mean BMI was 17.16 (SD 2.68), HDL-cholesterol 1.53 (SD 0.30) mmol/l, LDL-cholesterol 3.37 (SD 0.71) mmol/l, triglycerides −0.75 (SD 0.50) mmol/l, systolic blood pressure 112 (SD 11) mm Hg and diastolic blood pressure 69 (SD 10) mm Hg. Participants did not differ from excluded individuals as regards parental life satisfaction in 1983, type A behaviour subscales in 1986 or participants’ job strain or CIMT in 2001 (all p values >0.05). The educational level of participants’ mothers and fathers in 1983 was somewhat higher (10.5 vs 10.0 years and 10.1 years vs 9.7 years, respectively; p values <0.05) than that of excluded individuals, and participants’ parents had a somewhat higher income in 1983 than the parents of excluded individuals (60K vs 56K FIM per year, p<0.01). Participants were more educated (15.1 vs 13.4 years, p<0.001) than the excluded individuals.

    The sample characteristics are shown in table 1. Descriptive statistics were almost identical between original and imputed datasets. Further analyses were carried out on the imputed dataset.

    Pearson correlations showed that the leadership component of type A behaviour correlated positively with family income (r = 0.10, p<0.05), fathers’ education (r = 0.14, p<0.01) and participants’ educational attainment (r = 0.12, p<0.01) (table 2). The job strain components, job demands and job control, correlated positively with each other (r = 0.21, p<0.001).

    View this table:
    Table 2 Bivariate correlations (Pearson) between independent variables

    In the age adjusted model, father’s life dissatisfaction and participant’s low leadership predicted higher job strain in men. In the age and education adjusted model, both maternal and paternal life dissatisfaction as well as participants’ low leadership skills predicted higher job strain (table 3).

    View this table:
    Table 3 Unstandardised regression coefficients of preemployment factors predicting job strain in men at 27–39 years of age

    The results of multivariate linear regression analyses showed that adjustment for pre-employment parental socioeconomic position weakened the relationship between job strain and CIMT by 4%, and that parental life satisfaction adjustment strengthened the job strain/CIMT relationship by 8%, whereas adjustment for pre-employment type A behaviour components attenuated the job strain/CIMT relationship by 17% to non-significant. Despite the fact that adjustments for pre-employment factors had an effect on the job strain/CIMT relationship, it remained significant after full adjustment (table 4).

    View this table:
    Table 4 Unstandardised regression coefficients of job strain and preemployment factors predicting CIMT in men at 27–39 years of age

    Furthermore, we repeated the analyses using different formulations of job strain, as well as the components of job strain. The quotient term of job strain was associated with CIMT in age adjusted analyses but not in fully adjusted analyses (table 5). The regression β coefficient reduction in the age adjusted model compared to the fully adjusted model was 14% (the tertile term of job strain) and 26% (the quotient term). In addition, multivariate regression analyses shown in tables 4 and 5 were repeated with these alternative operationalisations of job strain controlling for biological coronary risk factors, but this did not change the results.

    View this table:
    Table 5 Unstandardised regression coefficients of different job strain formulations and its components predicting CIMT in men at 27–39 years of age

    DISCUSSION

    In this study of 494 healthy young men aged 27–39 years, higher job strain was associated with higher CIMT. pre-employment, family factors did not markedly attenuate the job strain/CIMT relationship, although parental socioeconomic position slightly weakened it, while parental life satisfaction slightly strengthened it. pre-employment the leadership component of type A behaviour attenuated the association between job strain and CIMT by 17% to non-significant. These results suggest that pre-employment factors may contribute to the job strain/CIMT relationship in healthy men, and furthermore that pre-employment disposition may have a greater influence on this relationship than environmental family factors.

    In the present study, age adjusted analyses of family factors and job strain showed that paternal life dissatisfaction predicted job strain in men. When an additional adjustment for participants’ education was made, both maternal and paternal life dissatisfaction were predictive of job strain. This finding is in line with previous studies that have indicated that parental life dissatisfaction is linked to various negative outcomes in adolescence and adulthood.21 36 Parents’ marital life dissatisfaction has been related to behaviour problems and psychopathology in adolescent offspring.36 In the CRYF sample, childhood maternal life dissatisfaction as such predicted depressive tendencies in adolescent girls, and a combination of maternal life dissatisfaction and hostility predicted depressive tendencies in boys.20

    Although it has been documented that the relationship between job strain and CIMT is explained by conventional risk factors for atherosclerosis, such as age, socioeconomic status, childhood risk factors, health behaviour or biological factors in our cohort,3 a recent meta-analysis showed that the association between job strain and cardiovascular disease can be substantially attenuated after adjustment for adult socioeconomic status, biological factors and health behaviour related factors.2 Our finding that pre-employment family factors had only a modest effect on the job strain/CIMT relationship is not completely in accordance with previous findings reporting an impact of socioeconomic status on longitudinal accumulation of cardiovascular risks.37

    We found that of the four components of type A behaviour, low leadership was associated with increased risk of job strain and early atherosclerosis. This finding is in line with previous reports from the Cardiovascular Risk in Young Finns study38 linking low leadership as measured by the AFMS (type A behaviour questionnaire for the Finnish Multicenter Study) to somatic risk factors of coronary heart disease, such as blood pressure and BMI. The AFMS leadership measure38 and the HWolf leadership measure are highly correlated.39 It is not known exactly what the lack of leadership implies and why low leadership predicts a risk. High leadership is related to a low level of risk factors as well as somatic well-being, for example, it is related to high social and general self-esteem, popularity among peers, high achievement standards and perseverance.39 This suggests that high leadership may represent a non-risk component of type A behaviour.

    Previous literature shows that the role of leadership with respect to stress and disease is far from clear. In the 1950s, high leadership was linked to high physiological stress in both humans and other primates, but in the 1960s low leadership was associated with stress-related diseases.40 In a recent review by van Vugt, leadership in humans is linked with initiative taking, task competency and (social) intelligence.41 Lower leadership predicted lower educational attainment in our sample, as well as in the whole CRYF population. This link is a possible explanation for our finding that low leadership predisposes to the selection of high strain employment later in life, as lower education is known to increase the possibility of having a job with low control and high demands in adulthood.4 42

    In addition, high eagerness-energy had a direct symptomatic relationship with CIMT. This finding is in accordance with previous studies as eagerness-energy has been linked to various biological risk factors for coronary heart disease in the current as well as in other samples: the eagerness-energy component has been associated with high levels of serum cholesterol and triglycerides in children and adolescents,27 longer physiological recovery time after exposure to stress,43 diastolic blood pressure in 18-year-old girls,38 and thicker adulthood CIMT in men over different developmental periods during a 15-year follow-up.17 The HWolf eagerness-energy component has been suggested as being conceptually close to Jenkin’s Activity Survey speed and impatience scale (JAS-S).17 Although eagerness-energy has been documented to be related to risk factors for cardiovascular heart disease, it was not an independent predictor of job strain in the present study, which may explain why only a symptomatic association with early atherosclerosis was found with these data.

    Main messages

    In this contemporary cohort of men, lack of leadership (a type A behaviour component) contributed to the association between job strain and carotid intima-media thickness 15 years later, whereas preemployment family factors had only a modest effect on this association

    Policy implications

    The results suggest that type A behaviour components, such as preemployment lack of leadership, contribute to the association between job strain and CIMT in later life.

    The current findings by adopting a life-course perspective add to the strain/health literature and suggest that in addition to traditional risk factors, personality factors from the period before first entering the labour force may contribute to the job strain–health associations. Hence, job strain may not originate entirely from occupational settings. Job strain is documented to increase the risk for cardiovascular events/disease,2 but in addition, personality may contribute to perceived job strain and the job strain/CIMT association in men. Our findings do not fully explain the impact of personality versus (work) environment on health. However, we suggest that the success of organisational attempts to modify job strain has thus far been relatively modest possibly because the effects of personality are largely ignored when work stress is considered to originate mainly from work.4 44

    There are at least two important advantages to this study. First, it combines population-based data with a follow-up period of up to 18 years, a length of time very rarely available for study. Second, we applied a longitudinal study design in which type A behaviour components were measured before the participants entered the labour force for the first time. Hence, neither work nor job strain could have exerted any effect on the type A behaviour component (eg, increasing or decreasing type A behaviour).

    There are, however, several limitations that should be taken into account when interpreting these findings. First, missing data can be a problem in epidemiological studies. In multivariate analyses, the number of participants with full information on all study variables can be small relative to the actual sample size. Therefore, data imputation is used to take into account at least some of the lost information. Furthermore, information missing at random is taken into account by the EM method, whereas complete case analysis ignores this. The EM algorithm used in this study is documented to provide unbiased estimates.34 35 The studied sample was representative of the baseline population with respect to cardiovascular risk factors. However, the participants’ parents and the participants themselves were slightly more educated than the excluded individuals. Thus, it is not possible to completely exclude the possibility of selection bias in this study. Second, despite the 18-year follow-up period, the study population was under 40 years of age when job strain and CIMT were assessed. It is possible that the relationships between pre-employment factors, job strain and CIMT may differ from those observed in this study later in life when clinical conditions become more common. In addition, job strain was obtained by a self-report measure, which response style may partly explain our findings. Furthermore, as our analysis of the job strain/CIMT association was based on a relatively small number of participants, including a wide range of covariates in the model decreased the stability of the job strain effect and widened confidence intervals. Third, components of type A behaviour may have different implications for educational performance45 and for work stress and health.39 More research is needed to clarify this issue. Furthermore, given that we assessed personality with four scales but work stress with only two scales, a more comprehensive assessment of personality than work stress might have overestimated the effect of personality on the association between job strain and CIMT. Finally, our sample corresponded to the vast majority of the Finnish population in 1980, when the study was launched. As the study sample was thus racially and culturally homogeneous, the present results cannot therefore be generalised to other ethnic and cultural groups.

    CONCLUSIONS

    This study provides prospective evidence that excess risk for CIMT in employees perceiving job strain may be partly related to their pre-employment disposition but is less likely be related to pre-employment family factors. Thus, the results suggest that type A behaviour components, such as pre-employment lack of leadership, contribute to the association between job strain and CIMT in later life.

    REFERENCES

    1. ↵
      1. Karasek RA
      . Job demands, job decision latitude, and mental strain: implications for job redesign.Admin Sci Quart1979;24:285–308.
      OpenUrlCrossRef
    2. ↵
      1. Kivimäki M,
      2. Virtanen M,
      3. Elovainio M,
      4. et al
      . Work stress in the etiology of coronary heart disease: a meta-analysis.Scand J Work Environ Health2006;32:431–42.
      OpenUrlCrossRefPubMedWeb of Science
    3. ↵
      1. Hintsanen M,
      2. Kivimäki M,
      3. Elovainio M,
      4. et al
      . Job strain and early atherosclerosis: the Cardiovascular Risk in Young Finns Study.Psychosom Med2005;67:740–7.
      OpenUrlAbstract/FREE Full Text
    4. ↵
      1. Lange AH,
      2. Taris TW,
      3. Kompier MAJ,
      4. et al
      . “The very best millennium”: longitudinal research and the demand-control-(support) model.J Occup Health Psychol2003;8:282–305.
      OpenUrlCrossRefPubMed
    5. ↵
      1. Fergusson DM,
      2. Horwood L
      . Vulnerability to life events exposure.Psychol Med1987;17:739–49.
      OpenUrlCrossRefPubMedWeb of Science
    6. ↵
      1. Elovainio M,
      2. Kivimäki M,
      3. Ek E,
      4. et al
      . The effect of pre-employment factors on job control, job strain and psychological distress: a 31-year longitudinal study.Soc Sci Med2007;65:187–99.
      OpenUrlCrossRefPubMedWeb of Science
    7. ↵
      1. Hintsa T,
      2. Kivimäki M,
      3. Elovainio M,
      4. et al
      . Parental socioeconomic position and parental life satisfaction as predictors of job strain in adulthood: 18-year follow-up of the Cardiovascular Risk in Young Finns Study.J Psychosom Res2006;61:243–9.
      OpenUrlCrossRefPubMedWeb of Science
    8. ↵
      1. Friedman M,
      2. Rosenman RH
      . Association of specific overt behavior pattern with blood and cardiovascular findings.JAMA1959;169:1286–96.
      OpenUrlCrossRefPubMedWeb of Science
    9. ↵
      1. Rosenman RH,
      2. Brand RJ,
      3. Sholtz RI,
      4. et al
      . Multivariate prediction of coronary heart disease during 8.5 year follow-up in the Western Collaborative Group Study.Am J Cardiol1976;37:903–10.
      OpenUrlCrossRefPubMedWeb of Science
    10. ↵
      1. Haynes S,
      2. Feinleib M,
      3. Levine S,
      4. et al
      . The relationship of psychosocial factors to coronary heart disease in the Framingham study II. Prevalence of coronary heart disease.Am J Epidemiol1978;107:384–402.
      OpenUrlAbstract/FREE Full Text
    11. ↵
      1. Matthews KA,
      2. Haynes SG
      . Type A behavior pattern and coronary heart disease risk. Update and critical evaluation.Am J Epidemiol1986;123:923–60.
      OpenUrlFREE Full Text
      1. Ragland DR,
      2. Brand RJ
      . Type A behavior and mortality from coronary heart disease.N Engl J Med1988;318:65–9.
      OpenUrlCrossRefPubMedWeb of Science
    13. ↵
      1. Shekelle RB,
      2. Gale M,
      3. Norusis M
      . Type A score (Jenkins Activity Survey) and risk of recurrent coronary heart disease in the aspirin myocardial infarction study.Am J Cardiol1985;56:221–5.
      OpenUrlCrossRefPubMedWeb of Science
    14. ↵
      1. Dembroski TM,
      2. MacDougall JM,
      3. Williams RB,
      4. et al
      . Components of type A, hostility, and anger-in: relationship to angiographic findings.Psychosom Med1985;47:219–33.
      OpenUrlAbstract/FREE Full Text
    15. ↵
      1. Räikkönen K
      . Predictive associations between type A behavior of parents and their children: a 6-year follow-up.J Genet Psychol1993;154:315–28.
      OpenUrlPubMedWeb of Science
    16. ↵
      1. Wolf TM,
      2. Sklov MC,
      3. Wenzl PA,
      4. et al
      . Validation of a measure of type A behavior pattern in children: Bogalusa heart study.Child Dev1982;53:126–35.
      OpenUrlCrossRefPubMedWeb of Science
    17. ↵
      1. Keltikangas-Järvinen L,
      2. Hintsa T,
      3. Kivimäki M,
      4. et al
      . Type A eagerness-energy across developmental periods predicts adulthood carotid intima-media thickness: the Cardiovascular Risk in Young Finns Study.Arterioscler Thromb Vasc Biol2007;27:1638–44.
      OpenUrlAbstract/FREE Full Text
    18. ↵
      1. Karasek RA,
      2. Baker D,
      3. Marxer F,
      4. et al
      . Job decision latitude, job demands, and cardiovascular disease: a prospective study of Swedish men.Am J Public Health1981;71:694–705.
      OpenUrlPubMedWeb of Science
    19. ↵
      1. Bereczkei T,
      2. Csanaky A
      . Stressful family environment, mortality, and child socialisation: life history strategies among adolescents and adults from unfavourable circumstances.Int J Behav Dev2001;25:501–8.
      OpenUrlCrossRef
    20. ↵
      1. Katainen S,
      2. Räikkönen K,
      3. Keskivaara P,
      4. et al
      . Maternal child-rearing attitudes and role satisfaction and children’s temperament as antecedents of adolescent depressive tendencies.J Youth Adolesc1999;28:139–63.
      OpenUrlPubMed
    21. ↵
      1. Bäckman L,
      2. von Hofsten C
      1. Keltikangas-Järvinen L
      . Psychology and the risk for cardiovascular disorder: a developmental perspective.In:Bäckman L, von Hofsten C, eds. Psychology at the turn of the millennium. Cognitive, biological, and health perspectives. Vol 1.Hove, East Sussex: Psychology Press, 2002:335–53.
    22. ↵
      1. Power C,
      2. Stansfeld SA,
      3. Matthews S,
      4. et al
      . Childhood and adulthood risk factors for socioeconomic differentials in psychological distress: evidence from the 1958 British birth cohort.Soc Sci Med2002;55:1989–2004.
      OpenUrlCrossRefPubMedWeb of Science
    23. ↵
      1. Kivimäki M,
      2. Hintsanen M,
      3. Keltikangas-Järvinen L,
      4. et al
      . Early risk factors, job strain, and atherosclerosis among men in their 30s: the Cardiovascular Risk in Young Finns Study.Am J Public Health2007;97:450–2.
      OpenUrlCrossRefPubMed
    24. ↵
      1. Åkerblom HK,
      2. Uhari M,
      3. Pesonen E,
      4. et al
      . Cardiovascular risk in young Finns.Ann Med1991;23:35–40.
      OpenUrlCrossRefPubMedWeb of Science
    25. ↵
      1. Raitakari OT,
      2. Juonala M,
      3. Kähönen M,
      4. et al
      . Cardiovascular risk factors in childhood and carotid artery intima-media thickness in adulthood.JAMA2003;290:2277–83.
      OpenUrlCrossRefPubMedWeb of Science
    26. ↵
      1. Makkonen T,
      2. Ruoppila I,
      3. Rönkä T,
      4. et al
      . Operation family study. Helsinki: Mannerheim League of Child Welfare, 1981.
    27. ↵
      1. Hunter SM,
      2. Wolf TM,
      3. Sklov MC,
      4. et al
      . Type A coronary-prone behavior pattern and cardiovascular risk factor variables in children and adolescents: the Bogalusa Heart Study.J Chronic Dis1982;35:613–21.
      OpenUrlCrossRefPubMedWeb of Science
    28. ↵
      1. Wolf TM,
      2. Hunter SM,
      3. Webber LS,
      4. et al
      . Self-concept, locus of control, goal blockage, and coronary-prone behavior pattern in children and adolescents: Bogalusa Heart Study.J Gen Psychol1981;105:13–26.
      OpenUrlPubMedWeb of Science
    29. ↵
      1. Karasek RA
      . Job content questionnaire and user’s guide. Revision 1.1.Los Angeles: University of Southern California, 1985.
    30. ↵
      1. Elo A-L,
      2. Leppänen A,
      3. Lindström K,
      4. et al
      . Occupational Stress Questionnaire: user’s instructions (Reviews 19). Helsinki: Finnish Institute of Occupational Health, 1992.
    31. ↵
      1. Landsbergis PA,
      2. Schnall PL,
      3. Warren K,
      4. et al
      . Associations between ambulatory blood pressure and alternative formulations of job strain.Scand J Work Environ Health1994;20:349–63.
      OpenUrlCrossRefPubMedWeb of Science
    32. ↵
      1. Schafer JL,
      2. Graham JW
      . Missing data: our view of the state of art.Psychol Methods2002;7:147–77.
      OpenUrlCrossRefPubMedWeb of Science
    33. ↵
      1. Kline RB
      . Principles and practice of structural equation modeling. New York: Guilford Press, 2005.
    34. ↵
      1. Oostenbrink JB,
      2. Maiwenn JA
      . The analysis of incomplete cost data due to dropout.Health Econ2005;14:763–76.
      OpenUrlCrossRefPubMedWeb of Science
    35. ↵
      1. Enders CK
      . Using the expectation maximization algorithm to estimate coefficient alpha for scales with item-level missing data.Psychol Methods2003;8:322–37.
      OpenUrlCrossRefPubMedWeb of Science
    36. ↵
      1. King CA,
      2. Radpour L,
      3. Naylor MW,
      4. et al
      . Parents’ marital functioning and adolescent psychopathology.J Consult Clin Psychol1995;63:749–53.
      OpenUrlCrossRefPubMedWeb of Science
    37. ↵
      1. Karlamangla AS,
      2. Singer BH,
      3. Williams DR,
      4. et al
      . Impact of socioeconomic status on longitudinal accumulation of cardiovascular risk in young adults: the CARDIA Study (USA).Soc Sci Med2005;60:999–1015.
      OpenUrlCrossRefPubMedWeb of Science
    38. ↵
      1. Keltikangas-Järvinen L,
      2. Jokinen J
      . Type A behavior, coping mechanisms and emotions related to somatic risk factors of coronary heart disease in adolescents.J Psychosom Res1989;33:17–27.
      OpenUrlCrossRefPubMedWeb of Science
    39. ↵
      1. Keltikangas-Järvinen L,
      2. Räikkönen K
      . Healthy and maladjusted type A behavior in adolescents.J Youth Adolesc1990;19:1–18.
      OpenUrlCrossRef
    40. ↵
      1. Sapolsky RM
      . The influence of social hierarchy on primate health.Science2005;308:648–52.
      OpenUrlAbstract/FREE Full Text
    41. ↵
      1. van Vugt M
      . Evolutionary origins of leadership and followership.Pers Soc Psychol Rev2006;10:354–71.
      OpenUrlAbstract/FREE Full Text
    42. ↵
      1. Belkic KL,
      2. Landsbergis PA,
      3. Schnall PL,
      4. et al
      . Is job strain a major source of cardiovascular disease risk?Scand J Work Environ Health2004;30:85–128.
      OpenUrlPubMedWeb of Science
    43. ↵
      1. Palmero F,
      2. Diez JL,
      3. Asensio AB
      . Type A behavior pattern today: relevance of the JAS-S factor to predict heart rate reactivity.Behav Med2001;27:28–36.
      OpenUrlPubMedWeb of Science
    44. ↵
      1. Strike PC,
      2. Steptoe A
      . Psychosocial factors in the development of coronary artery disease.Prog Cardiovasc Dis2004;46:337–47.
      OpenUrlCrossRefPubMedWeb of Science
    45. ↵
      1. Spence JT,
      2. Helmreich RL,
      3. Pred RS
      . Impatience versus achievement strivings in the type A pattern: differential effects on students’ health and academic achievement.J Appl Psychol1987;72:522–8.
      OpenUrlCrossRefPubMedWeb of Science

    Footnotes

    • Funding: TH was supported by a grant from the Finnish Cultural Foundation, MK and JV by the Academy of Finland (grants 117604, 124322 and 124271), LK-J by the Academy of Finland (grants 111056, 209514, 209518) and the Yrjö Jansson Foundation, OTR by the Academy of Finland (grants 77841 and 210283) and the Finnish Foundation of Cardiovascular Research, and OTR and JSAV by Special Federal Grants for Turku University Central Hospital.

    • Competing interests: None.