1 Department of Psychology, Division of Applied Psychology, University of Helsinki, PO Box 9, FIN-00014 University of Helsinki, Finland.
2 Centre of Applied and Preventive Cardiology, Turku University, PO Box 52, 20251 Turku, Finland.
3 Department of Medicine, Turku University, PO Box 52, 20251 Turku, Finland.
Correspondence: Liisa Keltikangas-Järvinen. E-mail: liisa.keltikangas-jarvinen{at}helsinki.fi
![]() |
Abstract |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Methods Prospective cohort study with a population-based random sample of 477 men and 648 women, aged 1221 years at baseline. Baseline data included information on pathogenic and protective components of type A behaviours (impatience, aggression, hard-driving, and engagement-involvement) and parental education. The 9-year follow-up data included information on the participants educational level and health behaviours (smoking, alcohol consumption, physical inactivity, butter use).
Results After adjustment for parental education, high levels of impatience and low levels of hard-driving in adolescence and early adulthood predicted low educational level in adulthood (Ps < 0.01 for men, Ps < 0.001 for women). Adulthood education was inversely associated with smoking in women and men (odds ratios [OR] = 8.5 and 7.9, 95% CI: 3.418.4 and 3.123.9, respectively), and with physical inactivity in women (OR = 5.4, 95% CI: 2.611.4). In men, components of type A behaviour explained 28.5% of the inverse association between education and smoking, even after controlling for parental education. In women, the corresponding proportions were 20.5% and 17.7% for smoking and physical inactivity, respectively.
Conclusions The inverse associations of adulthood education with smoking in men and women and physical inactivity in women may be partly rooted in personality-related factors present earlier in life. Our evidence suggests that personality should be studied as a potential contributor to socioeconomic differences in health behaviours.
Accepted 24 January 2003
The continually increasing cardiovascular risk with decreasing level of socioeconomic position, as indicated by education, occupational status, and income, is well-established.13 Understanding factors contributing to this socioeconomic gradient is important for planning disease prevention. It has been suggested that early-life social circumstances influence employment and social position in adulthood which, in turn, influence cardiovascular risk (the pathway model).4 Cumulative lifetime exposures to adverse social and material circumstances, beginning in early life might also have a role in the development of socioeconomic inequalities in cardiovascular health in adulthood (the accumulative model).5,6
Not enough is known about the extent to which stable psychological attributes in childhood and adolescence determine ones socioeconomic trajectory. Indirect evidence suggests that components of type A behaviour pattern (TABP) may be relevant in this context. Type A behaviour pattern is a multidimensional, temperament-related construct,7,8 which is assumed to be partly inherited.9 Type A behaviour pattern includes psychologically distinct components which have both pathogenic and protective aspects in terms of health and adjustment.10,11 High levels of impatience and aggressiveness have been associated with poor school achievement, downward occupational mobility, long-term unemployment, and a high level of somatic and behavioural cardiovascular risks.1116 These and other pathogenic components of TABP have their antecedents in childhood hyperactivity, which has been shown to predict adjustment problems such as antisocial personality, poor school achievement, and onset of substance use.1719 The protective components of TABP, including engagement-involvement and hard-driving (responsibility), are associated with high self-esteem, high level of aspirations, and low level of somatic cardiovascular risk factors.11,20
The on-going population-based Cardiovascular Risks in Young Finns (CRYF) study provided an opportunity to prospectively examine the extent to which components of TABP in adolescence and early adulthood are associated with educational level later in adulthood and the extent to which these personality traits contribute to the inverse association between educational level and cardiovascular health risk behaviours in adulthood. We hypothesized that components of impatience and aggression predict low educational level and high levels of health risk behaviours. Components of hard-driving and engagement-involvement were hypothesized to predict high educational level and low levels of behavioural risks.
![]() |
Methods |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Information about TABP and parents socioeconomic position was collected in 1983, when the participants were 12, 15, 18 and 21 years old. The follow-up measurement 9 years later in 1992 assessed participants socioeconomic status and cardiovascular health risk behaviours. Complete data on TABP and participants educational level were available from 648 men and 477 women, 57% of the participants of the sample in 1983. Sample attrition in the CRYF follow-up studies has not been found to be systematic in regard to health behaviours or personality factors.2224
Measures
Participants TABP was both self-rated and evaluated by their mothers using the TABP questionnaire for the Finnish Multicenter Study (AFMS).25 The AFMS consists of 17 items on a five-point scale of 1 (totally disagree) to 5 (totally agree), of which 14 items are derived from the Matthews Youth Test for Health (MYTH).26 To make the AFMS more suitable for adolescents and young adults, four items of the original MYTH were replaced by four items derived from the Swedish version of the Jenkins Activity Survey for Students (JAS).27 The details of the development of this method have been given by Keltikangas-Järvinen and Räikkönen.25 As in previous studies on the associations of TABP with school achievement and cardiovascular risk,14,25 we used non-correlated linear composite scales based on factor analysis to measure the following dimensions of TABP: impatience (sample item: When I have to wait for others I become impatient (from the self-report scale), When my child has to wait for others he/she becomes impatient (from the mother-rating scale); aggression-competitiveness (sample item: I tend to get easily into fights (from the self-report scale), My child tends to get easily into fights (from the mother-rating scale); engagement-involvement (sample item: I am frequently asked to be a leader, (from the self-report scale), My child is frequently asked to be a leader (from the mother-rating scale); and hard-driving (sample items: I consider myself more responsible than the average person, (from the self-report scale), I consider my child to be more responsible than the average person (from the mother-rating scale). The general coefficients of reliability of the different components have been shown to range from 0.75 to 0.85.11 Correlations between self-ratings and mother-ratings were 0.34, 0.22, 0.30, and 0.22 (Ps < 0.01) for impatience, aggression-competitiveness, engagement-involvement, and hard-driving, respectively. The AFMS correlates significantly with another well-known type A test, i.e. the Wolf-Hunter A-B Rating Scale.25 The AFMS shows significant continuity over different developmental periods,28 and the 3-year test-retest stability is 0.32 (P < 0.05) for young adult men and 0.42 (P < 0.01) for young adult women.28
Parental education was requested from both parents of the participants. Three categories were used: I academic (parents studying at or graduated from university); II secondary education but not academic (parents reporting high school or vocational school as their highest education); and III comprehensive school (parents reporting comprehensive school as their highest education). Information on the parent with the higher education was used in the analyses. Participants education in the follow-up was measured with a similar classification.
Health risk behaviours in the follow-up comprised smoking (daily smokers versus others), regular alcohol consumption (alcohol use at least once a week versus less than once a week), and butter use in diet (preference for butter or butter-based mixtures versus preference for margarine or vegetable oil). Physical inactivity was assessed by a physical activity index consisting of frequency, intensity, and duration of exercise.29 Participants in the lowest quartile were defined as physically inactive. These self-reported measures (and cutoff points) have been associated with serum lipids, insulin concentrations, and blood pressure in previous studies of the CRYF sample.30
Statistical analyses
Positive findings in all the following steps were considered to support the hypothesis that TABP contributes to the generation of educational differences in health risk behaviours: (1) impatience and aggression-competitiveness should be inversely associated with subsequent educational level and positively associated with subsequent health risk behaviours, while hard-driving and engagement-involvement should be positively associated with subsequent educational level and inversely associated with health risk behaviours; (2) educational level should be associated with health risk behaviours; and (3) this association should attenuate after controlling for components of TABP.
In the first step, differences in the levels of TABP dimensions between the three educational level groups of participants were tested by analysis of variance. As parental education is a known predictor of adulthood education,4,6 age and parental education were used as covariates. Only those dimensions of TABP that were significantly associated with educational level were considered when testing the age-adjusted associations of TABP dimensions with health risk behaviours.
In the second step, logistic regression analysis was used to study the associations between participants educational level and health risk behaviours. Odds ratios (OR) and their 95% CI were adjusted for age and parental education. In the third step, TABP, as a covariate, was entered into these models. Following Bosma et al.31 and van de Mheen et al.32 the contribution of TABP was estimated by the percentage reduction in the OR of the participants educational level using the following formula:
![]() |
![]() |
Results |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
|
The relations of the relevant TABP dimensions to smoking and physical inactivity are shown in Table 3 (a table showing the relations between all TABP dimensions and all health risk behaviours is available from the first author on request). Male and female smokers scored higher on impatience (mother-rated) and lower on hard-driving (mother- and self-rated for men, mother-rated for women) than non-smokers. In addition, female smokers scored lower on engagement-involvement (self-rated) than non-smokers. As regards physical inactivity in women, inactive participants scored higher on impatience (mother-rated) and lower on engagement-involvement (self-rated) than active participants.
|
|
![]() |
Discussion |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Our study replicated findings which have shown that impatience and inability to tolerate frustration predict poor health, adverse health behaviours, and poor school achievement.1214,18,19 Our evidence may be interpreted according to both the pathway model4 and the accumulation model.5,6 In line with the first model, high levels of impatience and low levels of hard-driving predicted drift to lower educational level, which in turn was associated with smoking. Thus, personality earlier in life may affect adulthood health behaviours through its impact on adult social circumstances. In accordance with the latter model, high levels of pathogenic and low levels of protective components of TABP may be an antecedent of a cycle of maladaptation which leads to the accumulation of adverse experiences, e.g. school maladjustment, health risk behaviours, and subsequent erratic worklife,12,13 the combination of which may contribute to socioeconomic inequalities in adulthood health.
Bosma et al.31 showed that personality is partially rooted in childhood social class, and that personality may act as a mediator between childhood social class and adulthood health. According to our findings, educational level in adulthood may be partially rooted in biologically based, temperament-related personality characteristics, and such personality may explain the association between low educational level and adverse health behaviours, irrespective of parental socioeconomic position. It should be noted, however, that the inverse associations of education with smoking and physical inactivity remained highly significant after adjustment for TABP. This implies that other factors besides personality contribute to the relationship between educational level and health risk behaviours. Prior research shows that low social class in childhood is associated with multiple chronic stressors, such as a high level of family conflicts, unresponsive parenting, threatening neighbourhood environments, and poor school achievements.33 It has also been suggested that the mothers low social class may predispose the fetus to an unfavourable growing environment, thus programming the childs health even before birth.34
Personality, as rated by mothers, seemed to be a stronger predictor of educational level and health risk behaviours than self-rated personality. It is possible that mother-ratings of personality reflect the mothers perception of the childs behaviour rather than the childs actual behaviour. Mothers characterized by a hostile parenting attitude35 tend to reject their children emotionally and to evaluate them as difficult, burdensome and in need of disciplinary action. Such hostile maternal attitude towards the child may be a marker of dysfunctional family relationships and a high level of family stress.36 Thus, mother-ratings of the childs personality as difficult may in fact reflect abnormalities in family relationships which, in turn, are known to contribute to the childs adjustment problems and poor physical health.36
Smoking and physical inactivity are well-known risk factors for cardiovascular disease.29,30,38 Smoking is strongly associated with current socioeconomic position,2,3,22 while other health risk behaviours, such as dietary practices, have been shown to be largely determined by parental socioeconomic position. The propensity to smoke may also be partly inherited,37 and evidence shows that both individual differences in smoking behaviour and certain personality characteristics have their genetic basis in the same dopaminergic function-related alleles.39 Therefore, the association between TABP and smoking may in fact be based on a common biological predisposition.
Type A behaviour pattern is assumed to be formed by the interplay between genetic and environmental factors. In this study, TABP was measured in adolescence and young adulthood, the participants being 12 to 21 years of age. By that age, parental behaviours such as the use of strict discipline and control are known to foster pathogenic aspects of TABP in children.7,40 An earlier assessment of TABP would have been less affected by environmental influences.
It should also be pointed out that the youngest participants were only 21 years of age at the measurement of educational level. At that age, ones educational status could still be in flux. However, compulsory education in Finland ends when the pupil turns 17. Upper secondary education comprises upper secondary school and vocational study programmes. Upper secondary school culminates in the matriculation examination (comparable to the British A-Levels), which is a prerequisite for studies at the tertiary level, i.e. at universities and polytechnics. Thus, the age of 17 is the critical point at which an individual chooses his/hers educational trajectory. In this study, the participants educational level was determined by completed education or ongoing education, i.e. it was not required for the 21 year olds to have completed university education to be classified as academic.
![]() |
Conclusions |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
KEY MESSAGES
|
![]() |
Acknowledgments |
---|
![]() |
References |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
2 Marmot MG, Davey Smith G, Stansfeld S et al. Health inequalities among British civil servants: the Whitehall II study. Lancet 1991;337: 138792.[CrossRef][ISI][Medline]
3 Blane D, Hart CL, Davey Smith G, Gillis CR, Hole DJ, Hawthorne VM. Association of cardiovascular disease risk factors with socioeconomic position during childhood and during adulthood. BMJ 1996;313: 143438.
4 Marmot M, Shipley M, Brunner E, Hemingway H. Relative contribution of early life and adult socioeconomic factors to adult morbidity in the Whitehall II study. J Epidemiol Community Health 2001;55:30107.
5 Davey Smith G, Hart C, Blane D, Gillis C, Hawthorne V. Lifetime socioeconomic position and mortality: prospective observational study. BMJ 1997;314:54752.
6 Power C, Matthews S, Manor O. Inequalities in self rated health in the 1958 birth cohort: lifetime social circumstances or social mobility? BMJ 1996;313:44953.
7 Räikkönen K, Keltikangas-Järvinen L. Childhood hyperactivity and the mother-child relationship as predictors of risk type A behavior in adolescence: a six year follow-up. Pers Indiv Differ 1992;13:32127.[CrossRef][ISI]
8 MacEvoy B, Lambert WW, Karlberg P et al. Early affective antecedents of adult type A behavior. J Pers Soc Psychol 1988;54:10816.[CrossRef][ISI][Medline]
9 Carmelli D, Rosenman RH, Chesney MA, Fabsitz R, Lee M, Borhani N. Genetic heritability and shared environmental influences of type A measures in the NHLBI Twin Study. Am J Epidemiol 1988;127:104152.[Abstract]
10 Steinberg L. Stability (and instability) of type A behavior from childhood to young adulthood. Dev Psychol 1986;22:393402.[CrossRef][ISI]
11 Keltikangas-Järvinen L, Räikkönen K. Pathogenic and protective factors of type A behavior in adolescents. J Psychosom Res 1989;33: 591602.[CrossRef][ISI][Medline]
12 Caspi A, Elder GH, Bem DJ. Moving against the world: Life-course patterns of explosive children. Dev Psychol 1987;23:30813.[CrossRef][ISI]
13 Kokko K, Pulkkinen L. Aggression in childhood and long-term employment in adulthood. A cycle of maladaptation and some protective factors. Dev Psychol 2000;36:46372.[CrossRef][ISI][Medline]
14 Keltikangas-Järvinen L. Type A behaviour and school achievement. Eur J Pers 1992;6:7181.[ISI]
15 Johnson CC, Hunter SM, Amos CI, Elder ST, Berenson GS. Cigarette smoking, alcohol, and oral contraceptive use by type A adolescentsthe Bogalusa Heart Study. J Behav Med 1989;12:1324.[ISI][Medline]
16 Booth-Kewley S, Friedman H. Psychological predictors of heart disease: a quantitative review. Psychol Bull 1987;101:34362.[CrossRef][ISI][Medline]
17 Tremblay RE, Phil RO, Vitaro F, Dobkin PL. Predicting early onset of male antisocial behavior from preschool behavior. Arch Gen Psychiatry 1994;51:73239.[Abstract]
18 Tremblay RE, Masse DP, Leblanc M. Early disruptive behavior, poor school achievement, delinquent behavior, and delinquent personality: longitudinal analyses. J Cons Clin Psychol 1992;60:6472.[CrossRef][ISI][Medline]
19 Masse LC, Tremblay RE. Behavior of boys in kindergarten and the onset of substance use during adolescence. Arch Gen Psychiatry 1997; 51:73239.
20 Keltikangas-Järvinen L, Räikkönen K. Healthy and maladjusted Type A behavior in adolescents. J Youth Adol 1990;19:118.[ISI]
21 Åkerblom HK, Uhari M, Pesonen E et al. Cardiovascular risk in young Finns. Ann Med 1991;23:3540.[ISI][Medline]
22 Leino M, Raitakari O, Porkka KVK, Taimela S, Viikari J. Associations of education with cardiovascular risk factors in young adults: the Cardiovascular Risk in Young Finns Study. Int J Epidemiol 1999;28:66775.[Abstract]
23 Raitakari O, Leino M, Räikkönen K et al. Clustering of risk habits in young adults. Am J Epidemiol 1995;142:3644.[Abstract]
24 Räikkönen K, Katainen S, Keskivaara P, Keltikangas-Järvinen L. Temperament, mothering, and hostile attitudes: a 12-year longitudinal study. Pers Soc Psychol Bull 2000;26:312.
25 Keltikangas-Järvinen L, Räikkönen K. Type A factors as predictors of somatic risk factors of coronary heart disease in young Finnsa six-year follow-up study. J Psychosom Res 1990;34:8997.[CrossRef][ISI][Medline]
26 Matthews KA, Angulo J. Measurement of the type A behavior pattern in children: assessment of childrens competitiveness, impatience-anger, and aggression. Child Dev 1980;51:46675.[ISI][Medline]
27 Lundberg U. Type A behavior and its relation to personality variables in Swedish male and female university students. Scand J Psychol 1980; 21:13338.[ISI]
28 Keltikangas-Järvinen L. Stability of Type A behavior during adolescence, young adulthood, and adulthood. J Behav Med 1989;12: 38796.[ISI][Medline]
29 Raitakari OT, Porkka KVK, Taimela S, Telama R, Räsänen L, Viikari JSA. Effects of persistent physical activity and inactivity on coronary risk factors in children and young adults. Am J Epidemiol 1994;140: 195205.[Abstract]
30 Raitakari OT, Porkka KVK, Räsänen L, Viikari J. Relations of life-style with lipids, blood pressure and insulin in adolescents and young adults. The Cardiovascular Risk in Young Finns Study. Atherosclerosis 1994;111:23746.[ISI][Medline]
31 Bosma H, van de Mheen HD, Mackenbach JP. Social class in childhood and general health in adulthood: questionnaire study of contribution of psychological attributes. BMJ 1999;318:1822.
32 van de Mheen H, Stronks K, Looman CWN, Mackenbach JP. Does childhood socioeconomic status influence adult health through behavioural factors? Int J Epidemiol 1998;27:43137.[Abstract]
33 Chen E, Matthews KA, Boyce WT. Socioeconomic differences in childrens health: how and why do these relationships change with age? Psychol Bull 2002;128:295329.[CrossRef][ISI][Medline]
34 Barker DJP. Fetal origins of coronary heart disease. BMJ 1995;311:17174.
35 Schaefer ES. A circumplex model for maternal behavior. J Abnorm Soc Psychol 1959;59:226335.[ISI][Medline]
36 Repetti RL, Taylor SE, Seeman TE. Risky families: family social environments and the mental and physical health of offspring. Psychol Bull 2002;128:33036.[ISI][Medline]
37 Carmelli D, Swan GE, Robinette D, Fabsitz R. Genetic influence on smokinga study on male twins. N Engl J Med 1992;327:82933.[Abstract]
38 Howard G, Wagenknecht LE, Burke GL et al. Cigarette smoking and progression of atherosclerosis: the Atherosclerosis Risk in Communities (ARIC) Study. JAMA 1998;279:11924.
39 Sabol SZ, Nelson ML, Fisher C et al. A genetic association for cigarette smoking behavior. Health Psychol 1999;18:713.[CrossRef][ISI][Medline]
40 Woodall K, Matthews KA. Familial environment associated with type A behaviors and psychophysiological responses to stress in children. Health Psychol 1989;8:40326.[CrossRef][ISI][Medline]