Introduction
The traditional way of testing coping is through questionnaires with a large number of questions based on definitions of coping as strategies (e.g. “Ways of Coping”) [
1]. Whilst these questionnaires yield differentiated assessments of the coping styles used by the respondents, such long questionnaires are often impractical in epidemiological research and clinical settings. Furthermore, it can be argued that the most important aspect of coping for health outcomes is not
how a person copes but rather
if the person tries to cope at all.
In contrast to other theories of coping, the Cognitive Activation Theory of Stress (CATS) [
2] stipulates a formal set of definitions for the mechanisms that dampen, eliminate, or reinforce the stress response to a challenging situation. All individuals have acquired such expectancies in relation to stimuli and to the outcome of the responses that are available. The response outcome expectancies are categorised as either positive (coping, expected to lead to a positive outcome), negative (hopelessness), or uncertain (helplessness). Based on CATS, we have developed Theoretically Originated Measure of the Cognitive Activation Theory of Stress (TOMCATS), a brief questionnaire aiming to measure response outcome expectancies.
There have been a number of studies comparing single-item questions with established measures. Good predictive and converging validity has been demonstrated on a number of concepts such as general health [
3], burnout [
4], patient satisfaction [
5], self-esteem [
6] and anxiety [
7]. With depression, the experience with single-item questions is inconsistent. Asking “are you depressed” to cancer patients worked well compared with a full diagnostic interview in North America [
8], but not in Japanese [
9] or UK [
10] cancer patients. Reme and Eriksen [
11] found that a single depression question identified most of the depressive symptoms measured by the Hopkins Symptom Checklist-25 in chronic pain patients. In study 1, we tested the concurrent validity of TOMCATS by comparing it with a traditional test of coping strategies. In study 2, we examined the relations between the questionnaire and socioeconomic differences in health in order to test the validity of TOMCATS.
The presence of substantial socioeconomic differences in health is well established [
12] and is often manifested as gradients rather than differences between distinct classes [
13]. Explanations offered for socioeconomic differences in health may be classified into two, possibly interacting, categories: structural vs. individual factors. Structural factors are external to the individual, such as the social and societal context, for instance differences in wealth, access to education and physical environment. Individual factors are internal to the individual, such as health behaviours, expectancies, intelligence, or social skills. TOMCATS measures the individually acquired expectancies of being able, or unable, to handle the stressors and challenges of everyday life. In this design, we test whether this brief questionnaire reveals any new perspective on the relations between socioeconomic factors and health.
Our main hypothesis is that differences in socioeconomic status (SES) over the life course lead to differences in reinforcement contingencies, which in turn lead to differences in response outcome expectancies. Furthermore, we hypothesise that individual differences in response outcome expectancies contribute to the socioeconomic differences in health, for instance through differences in health behaviours [
14]. In most work on SES and health, objective measures of SES have been used. In this article, we added a scale measuring the individual’s subjective evaluation of his or her place in society. This scale should be more sensitive to the learning history that we believe to be an essential factor for differences in health, particularly for subjective evaluation of health.
Discussion
The results from study 1 showed the expected significant correlations between the coping item in TOMCATS and the Utrecht Coping List (UCL): positive with overall instrumental mastery-oriented coping as well as with the active coping subscale and negative with passive coping and depressive scores. There were also negative correlations between instrumental mastery-oriented coping and the TOMCATS helplessness and hopelessness scores, and moderate but significant correlations between the TOMCATS helplessness and hopelessness scores and the UCL passive and depressive scores. However, the helplessness and hopelessness factors had similar correlations with passive and depressive reaction patterns, and contrary to expectation, the hopelessness factor was not more strongly correlated with the passive than the depressive reaction pattern.
Previous validation studies with coping instruments have shown a wide range of results. Correlations in the order of 0.78 were found when measuring the same coping concept (general self-efficacy, GSE) with two different general self-esteem scales [
27]. When measuring against similar but not identical concepts, correlations between GSE and the positive emotions factor of the “Positive and Negative Affect Scale” were about 0.40. The correlation with the less general concept of “health locus of control” was 0.23, a low but significant correlation [
28].
The UCL was chosen as a validation instrument because the instrument specifies a structure similar to the TOMCATS inventory by separating a passive avoidance strategy from a depressive reaction pattern. However, there are important distinctions as the UCL measures strategies and TOMCATS measures expectations, so we did not expect to see very high correlations between the factors. The fact that TOMCATS showed meaningful and moderate correlations indicates a relation to the coping strategies without measuring exactly the same phenomena. However, the correlations were somewhat low (between 0.27 and 0.47), which indicates that the TOMCATS factors are less closely related to the UCL than expected, but the results were generally in line with previous studies of closely related but different concepts.
Study 2 showed a strong association between the subjective expectation of coping, SES and self-rated general health. This supports the assumption that individual-learned expectancies matter for socioeconomic health differences [
14,
29]. As expected, low social status was also associated with individual experiences of failure to cope with the challenges of life (hopelessness) and the expectancy that there is no predictable relationship between what the individual does and what happens to him or her (helplessness). The gradient for helplessness appears even more pronounced than for hopelessness. This may be because the small number of individuals reporting a high degree of hopelessness creates a floor effect. Furthermore, the moderate amount of explained variance in our models indicates that there is reason for some caution when interpreting the importance of the results.
The two studies indicate the usefulness of a very brief questionnaire testing general response outcome expectancies. Important relations are revealed without the use of long and tedious forms. In a previous report, Odéen et al. [
30] analysed two questionnaires: one based on the UCL [
15] which is a development of the Lazarus Ways of Coping scales [
1] and one based on the Bandura self-efficacy concept [
31], the General Perceived Self-Efficacy Scale [
32]. None of the questionnaires were able to predict return to work in patients in a rehabilitation clinic. There were also difficulties with the theoretical bases as the questionnaires identify general trends rather than specific strategies. Given this lack of precision of the two instruments and the theoretical problems with them, the authors felt that a moderate degree of caution is warranted when inferring from results from these questionnaires to CATS or self-efficacy theory. The general overarching brief questions used in TOMCATS may be a better way to catch general trends and attitudes.
Our data support the individual explanations of socioeconomic differences in health as coping outcome expectancy is more strongly associated with self-rated health than both objective and subjective measures of socioeconomic status. However, structural factors in the social environment influence the learning history of an individual through differences in reinforcement contingencies. Those who grow up in high social strata have more resources available, and the chances of experiencing positive outcomes of coping attempts are probably higher. There is evidence that a low socioeconomic status has negative effects on health from early in life, and there may be “vicious circles” where adverse circumstances contribute to the development of expectancies of no or negative response outcomes of attempted coping [
14]. This in turn inhibits motivation to engage in behaviours that could lead to better health [
14] and could also mean that the individual is more likely to remain in an unfavourable social position.
If outcome expectancies can explain differences in health, a systematic effort to change the response outcome expectancies early in life could potentially be of great long-term benefit for individuals and reduce the social inequalities in health. It may be this very learning process that determines later behaviour, later optimism and later motivation to take care of one’s own health. A positive response outcome expectancy improves the chances that the individual will choose positive health behaviours. Further confirmation of the theoretical position would be to show that interventions aiming at improving coping skills and expectancies attached to coping strategies improve the health status of individuals. There is a possibility that the relationship between coping and subjective and objective socioeconomic status is reciprocal, in other words that coping is important in key behaviours that may advance or impede social mobility, such as children’s perceived vocational outcomes [
33], educational perseverance and performance [
34], and job satisfaction and performance [
35]. The associations between health, SES and coping suggest a common underlying factor, such as a tendency to view the world in an optimistic or pessimistic way. This may be a crucial element in the many cognitive interventions available to improve both subjective health and loss of function and working capacity.
The main strength of the two studies in this paper is that they are based on large, representative population samples. Mechanisms underlying social gradients could vary between countries and replications in other population samples would strengthen the evidence. It should be noted that in the first study, the sample was fairly homogenous; 80% were women and all were public sector employees. The SLOSH sample, however, was larger and representative of the national working population. Also, low test–retest reliability has been reported for a single global question on health [
36], and this tendency was stronger in subjects with low SES.
At this point in time, TOMCATS has been used solely as an explorative tool in epidemiological research. For the instrument to be used for other purposes, such as screening for low coping, clinical use, or as an indicator of effect of interventions, more validation research is needed. Especially, the test–retest reliability of the scale, as well as the sensitivity to change (the smallest detectable change as well as the minimal important change), needs to be established and reported. The CATS theory [
2] makes clear predictions of coping as a result of learned expectancies. In order to be a valid instrument based on this theory, TOMCATS must show stability in periods where no or minimal learning of new expectancies takes place, as well as sensitivity to acquisition of new expectancies. This would be a natural next step in the development of the TOMCATS inventory.
The major limitation of both studies in this article is that they are cross-sectional. In order to fully study the associations between environmental factors, coping expectancy and health, life course data would be needed. In addition, the use of objectively assessed, prospective health outcomes would further strengthen the evidence.