Change and stability of sense of coherence in adulthood: Longitudinal evidence from the Healthy Child study

https://doi.org/10.1016/j.jrp.2006.07.001Get rights and content

Abstract

The main aim of this three-wave 35-year follow-up study among Finnish employees (n = 532) was to investigate whether Sense of Coherence (SOC) is more stable among those with high SOC compared to those with low SOC, as hypothesized by Antonovsky [Antonovsky, A. (1987). Unraveling the mystery of health: how people manage stress and stay well. San Francisco, CA, US: Jossey-Bass], using two measurement points over 13 years. The participants were first studied in adolescence in 1961–1963, after which they responded to two postal questionnaires measuring SOC, first in 1985, and again in 1998. The results of the Factor Mixture Modeling identified two groups of individuals with different levels and mean changes of SOC: high-SOC (n = 344) and low-SOC (n = 188) individuals. The results supported Antonovsky’s hypothesis: SOC was more stable among high-SOC individuals (stability coefficient 0.57) than among low-SOC individuals (0.31). The results further supported the factor structure of the 11-item SOC scale. However, out of the five generalized resistance resources investigated in adolescence, only the level of home care was associated with SOC in adulthood.

Introduction

Recent years have witnessed considerable interest in Antonovsky’s (1987) concept of sense of coherence (SOC) in the conceptualization of successfully coping. SOC is at the core of a complex salutogenic model that seeks to explain health rather than disease, focuses on coping rather than stressors, salutary factors rather than risk factors, survivors rather than the defeated, and the invulnerable rather than the damaged. Like many other personality dispositions, such as optimism (Scheier and Carver, 1985, Scheier and Carver, 1992) and locus of control (Rotter, 1966) SOC is assumed to remain relatively stable in adulthood, especially after the age of 30. However, Antonovsky, 1987, Antonovsky, 1991 hypothesis; that SOC is more stable among high-SOC individuals than among those with a low SOC, has until now remained totally without empirical evidence.

The present study focused on this issue using unique 13-year SOC follow-up data of participants consisting of Finnish employees over 30 years of age from different occupations. The main goal was to investigate whether the stability of SOC differs according to its strength. A related aim was to investigate whether the structure of the SOC instrument, i.e., the Orientation to Life Questionnaire (OLQ; Antonovsky, 1987) is invariant across time. Demonstration of the structural invariance of the OLQ is necessary as it allows us to conduct stability analysis of SOC without having to be concerned about whether observed differences in the stability of SOC are due to structural change in the OLQ over time. Previous studies investigating the stability and structural invariance of SOC have not covered a time-span longer than 5 years. In addition, by utilizing the whole three-wave 35-year study design, we investigated the childhood antecedents of SOC in adulthood.

According to Antonovsky’s definition, SOC is “a global orientation that expresses the extent to which one has a pervasive, enduring though dynamic feeling of confidence that (1) the stimuli deriving from one’s internal and external environments in the course of living are structured, predictable, and explicable; (2) the resources are available to one to meet the demands posed by these stimuli; and (3) these demands are challenges, worthy of investment and engagement” (Antonovsky, 1987, p. 19). Accordingly, a high SOC is a dispositional orientation to life that protects people from stress by the way they perceive life events as challenges (sense of meaningfulness), occurring for a reason (sense of comprehensibility) and that, even if not under their own personal control, they can be handled by some other resource at their disposal (sense of manageability). According to the SOC theory, the components of meaningfulness, comprehensibility and manageability are highly interrelated.

SOC is assumed to promote an individual’s health through its central role in generating a more resilient choice of strategies for coping with life stressors. Thus, people with a high SOC are assumed to occupy a more favourable position on the health ease/disease continuum than those scoring low in SOC (Antonovsky, 1987, Antonovsky, 1991). SOC-health studies performed so far support this assumption; people with a high SOC tend to enjoy greater health than those with a low SOC (Feldt et al., 2005a, Kivimäki et al., 2000, Pallant and Lae, 2002, Reid et al., 2005). Furthermore, low SOC has also been associated with mortality (Poppius et al., 2003, Surtees et al., 2003). For example, a prospective (⩽6 years) general population study with over 20,000 participants by Surtees and his colleagues (2003) showed that after controlling for the effect of e.g. demographics, prevalent chronic disease, cigarette smoking history, systolic blood pressure, hostility and neuroticism, high SOC (based on median split) was still associated with a reduced risk for mortality from all causes.

As a construct of resiliency, SOC bears a resemblance to other personal key resources, e.g., high self-esteem, self-efficacy, hardiness, and dispositional optimism (Hobfoll, 2002). Indeed, SOC has been positively interrelated with self-esteem (Pallant & Lae, 2002), self-efficacy (Amirkhan & Greaves, 2003), and dispositional optimism (Feldt, Mäkikangas, & Aunola, 2006). In addition, although SOC is not construed as a personality trait it has been found to correlate with the Big Five traits (Ebert et al., 2002, Ruiselova, 2000), particularly low neuroticism.(Amelang, 1997, Ebert et al., 2002). Conceptually this association is easy to understand, since high SOC refers to a person’s capacity to respond to stressful situations, whereas individuals with high neuroticism are more likely to use ineffective coping mechanisms (McCrae & Costa, 1990).

The three components of SOC, i.e., sense of meaningfulness, comprehensibility and manageability, are also included in the OLQ, the instrument originally developed for measuring SOC (see Antonovsky, 1987). The OLQ exists in two forms. The longer version of the scale includes 29 items and the shortened version 13 items. We chose the latter form of the scale for the present study. The previous confirmatory factor analysis studies have supported the theory-based correlated three-factor structure of the 13-item OLQ as well as its equivalent second-order factor solution (Feldt et al., 2004, Feldt et al., 2005b, Feldt and Rasku, 1998, Gana and Garnier, 2001). The supported correlated three-factor structure of the scale indicates that the scale includes highly interrelated latent factors of meaningfulness (four items), comprehensibility (five items) and manageability (four items). Its equivalent second-order factor solution in turn explains the high inter-correlations of the first-order factors by a second-order factor. These structures have been found to be invariant across different age groups (Feldt, Leskinen, Kinnunen, & Ruoppila, 2003), employment groups (Feldt et al., 2005b) and occupational groups (Feldt & Rasku, 1998). Factorial invariance over time has also been supported (Feldt et al., 2003, Feldt et al., 2004, Feldt et al., 2005b, Veenstra et al., 2005).

According to Antonovsky (1987), the origins of SOC are to be found in the so-called generalized resistant resources (GRRs). Starting from early childhood, GRRs (e.g., money, social support, ego strength, intelligence) shape everyday encounters in life to become positive experiences, which gradually build up high SOC, whereas lack of various GRRs will be conducive to a low SOC. By the age of 30, the level of SOC should have stabilized, and, as stated previously, from then on it is hypothesized to remain as a relatively stable personality disposition throughout adulthood (Antonovsky, 1987, Antonovsky, 1993). Previous empirical studies with follow-up periods of up to five years have indeed shown high test–retest stability for SOC with stability coefficients ranging from 0.67 to 0.82 (Feldt et al., 2005b, Feldt et al., 2000, Kivimäki et al., 2000, Schnyder et al., 2000, Veenstra et al., 2005).

Some issues still remain unclear regarding the development and stability of SOC. Firstly, the hypothesized stabilization with age of SOC has remained without empirical evidence. Antonovsky (1987) stated that SOC remains relatively stable as a developmental construct after the formative years, with eventual stabilization occurring at about the age of 30. To the best of our knowledge, there are only two studies that have investigated this “age-hypothesis”. By using male-dominated Finnish technical designer samples the results obtained by Feldt and her colleagues (Feldt et al., 2003) did not support Antonovsky’s (1987) assumption: the stability of SOC was not higher among older individuals (over 30 years) than those under the age of 30. Similarly, in a study representing Canadian labour force (Smith, Breslin, & Beaton, 2003) the test–retest correlations over a 4-year period were practically on the same level for those aged less than 30 (r = .42) and for those aged over 30 (r = .45).

Secondly, according to Antonovsky, 1987, Antonovsky, 1991, the level of SOC plays a significant role in its stability: among high-SOC individuals stability is assumed to be higher than among low-SOC individuals, who will tend to move to an even lower level of SOC over time. This increasing disparity in adulthood is due to GRRs, which high-SOC individuals are more capable of bringing into play when facing stressors. Thus, when threatened, the equilibrium for high-SOC individuals will gradually return (Antonovsky, 1987, Antonovsky, 1991). On the other hand, the opposite will hold true for low-SOC persons, and therefore for them, life experiences may, in time, become more chaotic, less manageable, and more meaningless. The question of this “level hypothesis” has never been empirically tested, thus we conducted the present 13-year follow-up study in order to investigate it.

Finally, evidence of the role of childhood GRRs in the development of SOC is sparse. Feldt and her colleagues (Feldt et al., 2005a) found that child-centered parenting in adolescence was directly, although rather weakly, associated with a high SOC in adulthood. However, socio-economic background and school success were not directly related to adult SOC. In Lundberg’s study (1997), none of the childhood GRRs directly predicted SOC in adulthood. Only economic hardship in childhood had a small indirect effect, through adult class position, on SOC in adulthood.

Methodologically, SOC stability study results have often been based on simple test–retest correlations between two unidimensional SOC summary scales in two time points, and sometimes changes in the mean level of SOC have also been reported. Longitudinal stability and change can be analyzed by examining (1) how the particular psychological construct’s absolute level, averaged across all individuals, is affected by the change in time (absolute stability), and (2) what happens to the individual’s rank-order, within a group, relative to other individuals over time (differential stability; Caspi, 1998). In this study, we focused both on mean differences in SOC (absolute stability) and rank-order stability (differential stability) of SOC. The specific aims and hypothesis of the present study are as follows:

The first aim of our study was to test, by confirmatory factor analyses (CFA), the factor structure and factorial invariance of the 13-item OLQ scale over 13 years among Finnish employees aged over 30 at the beginning of the study.

Hypothesis 1

A correlated three-factor structure consisting of the three highly inter-correlated latent factors of comprehensibility, manageability, and meaningfulness, and its equivalent second-order factor structure, will fit the data at both times, and these latent structures are invariant over time.

Our second aim was to investigate the rank-order stability of SOC over 13 years in the whole data.

Hypothesis 2

SOC is relatively stable in adulthood after the age of 30.

The third aim was to investigate by Factor Mixture Modeling, whether it is possible to identify latent groups with different mean levels of SOC. Furthermore, if latent groups could be identified in the data, a related aim was to investigate the stability coefficients (indicating rank-order stability) and changes in the means (indicating absolute stability) of SOC in different groups over a 13-year follow-up period.

Hypothesis 3

Those with high SOC will have more stable SOC over 13 years than those with low SOC. Thus, the stability coefficient for high-SOC individuals will be higher than for low-SOC individuals.

Finally, our fourth aim was to investigate whether GRRs in adolescence predict high SOC in adulthood.

Hypothesis 4

The GRRs in adolescence are associated with high SOC, whereas lack of GRRs is associated with low SOC in adulthood.

Section snippets

Participants

The data of the present 13-year prospective study consisted of 532 employees from different occupations and professional sectors. In the last phase of the study, the mean age of the participants was 49 (range 44–57). The data represented both genders evenly; there were 270 women and 262 men among the respondents. The present study is part of the research project focusing on work-related well-being from a life-course perspective (Hakanen, 2004, Vuori, 1993).

The data originates from the third

Measurement models for SOC: Confirmatory factor analysis

By using confirmatory factor analysis, we first tested whether the second-order factor model (with the three first-order meaningfulness, comprehensibility and manageability factors) for the 13-item SOC-scale would be applicable to the data. The fit indices for this model were relatively poor at both measurement times (Time 1: χ2(62) = 203.10, p < .001, CFI = 0.89, TLI = 0.87, RMSEA = 0.065; SRMR = 0.054; Time 2: χ2(62) = 312.62, p < .001, CFI = 0.85, TLI = 0.81, RMSEA = 0.087; SRMR = 0.063). The results showed that

Discussion

This three-wave 35-year follow-up study was the first study to use the relatively new statistical technique of Factor Mixture Modeling to investigate the level and the mean changes and stability of SOC in a sample of 532 Finnish employees. Our major finding was that the stability of SOC after the age of 30 depended strongly on its level. Statistically it was possible to identify two latent groups in the data: high-SOC and low-SOC individuals. As hypothesized on the basis of Antonovsky’s (1987)

Conclusions

This study, with its unique follow-up period of SOC over 13 years, brings new information to Antonovsky’s often debated (e.g., Geyer, 1997, Schnyder et al., 2000, Smith et al., 2003) assumption of the stability of SOC in adulthood. Supporting Antonovsky’s salutogenic theory (1987), we found SOC to be relatively stable among high-SOC individuals, whereas for the low-SOC individuals the stability was weak. The theoretically based factorial structure and time invariance over 13 years of the

References (51)

  • M. Amelang

    Using personality variables to predict cancer and heart disease

    European Journal of Personality

    (1997)
  • J. Amirkhan et al.

    Sense of coherence and stress: the mechanics of a healthy disposition

    Psychology and Health

    (2003)
  • A. Antonovsky

    Health, stress, and coping: new perspectives on mental and physical health

    (1979)
  • A. Antonovsky

    Unraveling the mystery of health: how people manage stress and stay well

    (1987)
  • A. Antonovsky

    The structural sources of salutogenic strengths

  • M.W. Browne et al.

    Alternative ways of assessing model fit

  • Caspi, A. (1998). Personality development across the life course. In: W. Damon (Series Ed.) & N. Eisenberg (Vol. Ed.),...
  • S.A. Ebert et al.

    Psychological resistance factors as predictors of general health status and physical symptom reporting

    Psychology, Health and Medicine

    (2002)
  • M. Eriksson et al.

    Validity of Antonovsky’s sense of coherence scale: a systematic review

    Journal of Epidemiology and Community Health

    (2005)
  • T. Feldt et al.

    Sense of coherence and work characteristics: a cross-lagged structural equation model among managers

    Journal of Occupational and Organizational Psychology

    (2004)
  • T. Feldt et al.

    The role of family background, school success, and career orientation in the development of sense of coherence

    European Psychologist

    (2005)
  • T. Feldt et al.

    Structural invariance and stability of sense of coherence: a longitudinal analysis of two groups with different employment experiences

    Work and Stress

    (2005)
  • T. Feldt et al.

    Sense of coherence and optimism: A more positive approach to health

  • A.W. Frenz et al.

    Psychometric evaluation of Antonovsky’s sense of coherence scale

    Psychological Assessment

    (1993)
  • J. Hakanen

    Työuupumuksesta työn imuun - työhyvinvointitutkimuksen ytimessä ja reuna-alueilla [From burnout to work engagement: In the core and on the fringes of work-related well-being]

    (2004)
  • Cited by (73)

    • The relationship between life events and sense of coherence in adolescence. A longitudinal twin study

      2022, Journal of Research in Personality
      Citation Excerpt :

      This is also in line with the empirical literature on the stability of personality traits, which finds that the rank-order stability increases throughout adolescence and peaks in adulthood (Costa, McCrae, & Lockenhoff, 2019). Supporting its trait-like nature, empirical studies have found moderate rank-order stability of SOC measured with a one-year interval and up to 13 years between measurements, both in adolescence and adulthood (Eriksson & Lindström, 2005; Hakanen, Feldt, & Leskinen, 2007; Honkinen et al., 2008), similar to the rank-order stability found for personality traits (Costa et al., 2019; Ferguson, 2010; Roberts & Delvecchio, 2000). The introduction of the theory of SOC represented a paradigm shift from a pathogenic focus on risk factors and causes of disease, to factors promoting and maintaining good health (A. Antonovsky, 1993; Eriksson & Lindström, 2005).

    • Salutogenic childbirth education raises sense of coherence: A longitudinal survey

      2021, Midwifery
      Citation Excerpt :

      Although, Habroe et al. (2007) found that achieving childbirth after ART was associated with increased sense of coherence at one year follow-up, this result was only significant among participants who had low sense of coherence at baseline. Other studies confirm the finding that sense of coherence is more stable among people with a high baseline sense of coherence compared to those people with a low baseline sense of coherence (Hakanen et al., 2007; Nilsson et al., 2003). Our study found that the conventional childbirth education increased sense of coherence for participants with a weak baseline sense of coherence only while the Salutogenic Childbirth Education program increased sense of coherence for participants with weak, moderate and strong sense of coherence.

    View all citing articles on Scopus
    View full text