The relationship between sense of coherence and mental health problems from childhood to young adulthood: A meta-analysis

Stress constitutes a threat to well-being and health development (Downey and Crummy, 2022; Grant et al., 2003; Hamby et al., 2021). Its effects can be particularly pervasive in early-to-mid-life spanning from childhood to young adulthood (Björkenstam et al., 2015). This assumption is supported by research showing that macro stressors like the COVID-19 pandemic have multifarious adverse effects on mental health at younger ages (Ma et al., 2021; Newnham et al., 2022; Ravens-Sieberer et al., 2021). However, research into mental health consequences of stress also highlights the importance of focusing not only on mental health problems, but equally on resilience (Chen and Bonanno, 2020; Schäfer et al., 2022a). Resilience factors may indirectly affect resilient outcomes (i.e., the maintenance or quick regain of mental health during or after stressor exposure) via higher-level resilience mechanisms (e.g., positive appraisal style or regulatory flexibility; Bonanno, 2021; Kalisch et al., 2015). Thereby, resilience factors are potential targets of interventions to foster resilience.

One of the best evidenced resilience factors is sense of coherence (SOC1; Kunzler et al., 2020a, Kunzler et al., 2020b; Schäfer et al., 2022a). SOC is the key component of the salutogenesis framework developed by Aaron Antonovsky, 1979, Antonovsky, 1987, which aimed at changing the perspective from risk and disease to health and well-being (Rivera et al., 2013). Within this model, health is assumed to be modulated by SOC, which is defined as a global orientation in life that “expresses the extent to which one has a pervasive, enduring though dynamic feeling of confidence that one's internal and external environments are predictable and that there is a high probability that things will work out as well as can reasonably be expected” (Antonovsky, 1979, p. 10). Individuals with higher levels of SOC perceive their environment as comprehensible and manageable and believe that their lives are meaningful. While many systematic reviews pointed to a substantial negative association of SOC and mental health problems in middle- to old-aged adult populations (e.g., del-Pino-Casado et al., 2019; Eriksson and Lindström, 2006, Eriksson and Lindström, 2007; Schäfer et al., 2019), research into younger age groups is rarer, even though, these life phases were supposed to be of major importance for the development and shaping of SOC (Braun-Lewensohn et al., 2013, Braun-Lewensohn et al., 2022). However, within the last years, emerging evidence provided support for a negative relationship between SOC and mental health problems also at younger ages (Länsimies et al., 2017). Yet, individual studies yielded heterogeneous effect size estimates ranging from small to large effect sizes (e.g., for general distress: r = −0.10 in Al-Yagon, 2011, vs. r = −0.78 in Anmyr et al., 2015).

Research into the link between mental health problems and SOC at younger ages is closely related to the question of whether and when SOC stabilizes in life (Honkinen et al., 2008). The salutogenic model assumes that SOC's stability develops over the lifespan (see Fig. 1) and that SOC assessed during youth or earlier only allows for short-term prediction of responses to stressor exposure (Antonovsky, 1987). According to these theoretical considerations, the stability of SOC, that is, temporal stability, mean level stability, and - in a broader sense - internal consistency, increases with age (Antonovsky and Sagy, 1986), whereby instability at younger ages results from frequent changes in the use of coping resources. While empirical research on this (in)stability is still rare, a closer look at the SOC scales may provide indirect evidence.

The SOC scales (Antonovsky, 1993) not only capture total SOC but also its components comprehensibility, manageability, and meaningfulness. Of these, comprehensibility is supposed to be shaped first by life experiences followed by manageability and meaningfulness (Idan et al., 2022). Thus, it is plausible to assume that also the internal consistency of SOC measures increases with age, as its components show a stronger association from adolescence to young adulthood. Though this assumption has not yet been tested empirically for SOC, indirect support comes from empirical studies investigating the internal consistency of personality traits finding internal consistencies to increase with participants' age (Finn, 1986; McFarland and Sparks, 1985).

SOC's (in)stability is also relevant for investigating SOC's link to health outcomes as it may limit potential cross-sectional and longitudinal associations with health outcomes at younger ages. Low internal and temporal stability (Holden and Bernstein, 2013; McCrae, 2015) can bias regression slopes towards zero (regression dilution; Hutcheon et al., 2010). In case of SOC, lower internal and temporal stability might not only reflect measurement error, but rather lower ‘true’ stability at younger ages. However, previous studies on the stability of SOC remained limited to absolute changes in SOC (Braun-Lewensohn and Sagy, 2010; Honkinen et al., 2008) and rank stability (Feldt et al., 2006), while no study yet examined the course of internal consistency or age-related changes in the associations with health outcomes.

Besides SOC's (in)stability, research into SOC in adult samples pointed to conceptual problems (Bachem and Maercker, 2016). On the one hand, SOC's strong negative relationship with mental health problems provides support for its role as resilience factor (Kunzler et al., 2020a, Kunzler et al., 2020b; Schäfer et al., 2022a), on the other hand, this association raised the question whether SOC is a unique construct related to responses to external stressors or simply a measure of mental health (Bachem and Maercker, 2016; Geyer, 1997). Similar concerns were raised due to strong associations found with personality traits, especially neuroticism, in adult samples (Barańczuk, 2021; Feldt et al., 2007; Strümpfer et al., 1998). In line with this notion, a recent study on 20-to-27-year-old twins provided evidence for a substantial proportion of the genetic variation in SOC being shared with mental health and personality traits (Silventoinen et al., 2022), thereby, suggesting that the strong associations partly derive from a common neurobiological underpinning. Similarly, longitudinal research in adult samples found a substantial overlap between SOC and mental health measures as well as personality traits (e.g., Grevenstein et al., 2016; Schäfer et al., 2022b; Schäfer et al., 2020a, Schäfer et al., 2020b). At the same time, these studies provided evidence for SOC having incremental predictive power beyond baseline measures of mental health and personality traits for later changes in mental health. So far, these findings were limited to mostly middle-aged adults, while less is known on these conceptual concerns during childhood and adolescence.

To date, no systematic review aimed at quantitatively summarizing empirical evidence on the relationship between SOC and mental health problems from childhood to young adulthood. The present review aimed at addressing this gap by providing an estimation of the population effect and its robustness across different symptom types and samples. Moreover, our aim was to investigate the factors that underlie heterogeneity at primary study level with a special focus on participants' age. Building on these findings, we aimed at providing new insights into the (in)stability of SOC at younger ages by examining the association of internal consistencies and participants' mean age. Moreover, with respect to the conceptual criticism on SOC (Bachem and Maercker, 2016; Geyer, 1997), we aimed at deriving preliminary evidence on whether SOC may simply constitute a measure of mental health. As correlations of r ≥ 0.70 with other measures assessing the same construct are often viewed as evidence for construct validity of psychological measures (Nunnally, 1967; Taras et al., 2014), relationships between SOC and mental health problems larger than 0.70 would have challenged SOC's quality as a unique construct.

To shed light on heterogeneity, we examined sociodemographic characteristics and symptom types as moderators and examined whether differences between clinical and non-clinical populations as well as the use of different SOC measures may act as confounders.

Based on the salutogenic theory (Antonovsky, 1979, Antonovsky, 1987), we expected that SOC's association with mental health problems increases from childhood to young adulthood. Moreover, some but not all studies pointed to gender-related differences in SOC (Bronikowski et al., 2017; Silventoinen et al., 2022), thus, we examined whether the relationship between SOC and mental health problems depended on differences in gender distribution.

Previous systematic reviews found SOC to be related to different symptom types, for example, depressive and anxiety symptoms, or posttraumatic stress (del-Pino-Casado et al., 2019; Schäfer et al., 2019). As the present meta-analysis was the first to examine a broad range of mental health problems including internalizing and externalizing symptoms, we aimed at examining whether there was evidence for between-outcome differences.

While some studies investigated (mostly) healthy populations, others focused on clinical samples (e.g., participants with somatic diseases, learning disabilities, or mental disorders; Idan and Margalit, 2014; Moons et al., 2020; Sehlin et al., 2020). As stressor exposure is likely to be higher in clinical samples (Butler et al., 2018), we examined whether these differences may also impact on the association of SOC and mental health problems.

Research on SOC mostly uses different versions of the Antonovsky (1993) SOC scales. On one hand, this is a strength as the consistent use of one measure reduces unwanted variance. On the other hand, it may negatively impact the validity of results via common method bias (Reio, 2010). This bias may be particular important since the SOC scales (Antonovsky, 1993) have been criticized for methodological (Flannery et al., 1994; Frenz et al., 1993) and conceptual problems (Bachem and Maercker, 2016; Gruszczynska, 2006). Apart from the short version and the long version of SOC scales (Antonovsky, 1993), there is another version adapted to children's lives (CSOC; Margalit, 1995; Margalit and Efrati, 1996). Thus, we examined whether the use of different measures affected effect size estimates.

留言 (0)

沒有登入
gif