According to the job demand-control (JD-C) model, job demands and job control have an effect on employee well-being. The model suggests that the work environment can be characterised by a combination of high job demands and low job control (Karasek, 1979; Karasek and Theorell, 1990).
A similar approach has been developed by Warr (1987a) who in his vitamin model suggests that job characteristics affect mental health. Warr suggests nine (later 10) broad categories of job characteristics, grouping together the principal variables. Warrs vitamin model was based on the analogy that as vitamins are required for physical health, a similar pattern can be observed with the environmental features on the mental health and well-being of individuals. Although the JD-C model suggests an interaction effect, Warr (1987a, 1990), however, suggests a non-linear relationship between the job characteristics and mental health.
According to the vitamin model, the availability of vitamins is important for physical health up to, but not beyond, a certain level. At low levels of intake vitamin deficiency gives rise to physiological impairment and ill health, but after attainment of specified levels, there are no benefits from additional quantities. Principal features of the environment (see table 1) are important to mental health in a similar manner, their absence leads to impairment in mental health, and their presence beyond a required level does not yield further benefits. There are several types of vitamins; some in large doses can become harmful, similarly desirable environmental features come to have a negative impact on mental health at extremely high levels (Warr, 1994). For example vitamins A and D in very large quantities can be toxic and therefore act as an additional decrement (AD). While vitamins C and E have no additional impact but rather have a constant effect and therefore no ill-effects are observed in large doses.
According to Warr (1987a) low levels or extremely high levels of certain environmental features can have a harmful effect and give rise to an additional decrement in mental health, depending on the environmental feature; some can have a constant effect (CE). Within the vitamin model there are six AD features and three with CE (see table 1) (Warr, 1994). For example job autonomy (opportunity for control) is assumed to follow an inverted U-shape or AD pattern; very high levels of job autonomy are harmful to the employees level of mental health, since it involves high job responsibility, uncertainty and difficulty in decision making (Warr, 1987a).
The nine environmental features seem to represent sources of variation in mental health; they can both promote and impair mental health (Warr, 1987a, 1987b). Warr (1987a, 1994) suggests that the nine environmental features should be treated separately, although there may be some associations between the sub-categories of variables; for example role conflict (in category 3) and role ambiguity (in category 5) are similar. However, the model suggests that the nine categories of job characteristics do act in isolation; this has been supported to a certain extent (Warr, 1994; De Jonge and Schaufeli, 1998), however, these studies have failed to consider all nine job categories. They have considered only a number of the environmental features (i.e. job control, job autonomy and social support). Therefore, there is a need for strong multivariate research, examining several job features simultaneously in relation to mental health and well-being (Warr, 1987b).
The vitamin model assumes a non-linear relationship between the nine environmental features and mental health. Warr (1987, 1994) suggests there are five components to mental health, these include affective well-being, competence, autonomy, aspiration and integrated functioning. Psychological research has mainly focused on the first component of affective well-being as an indicator of job-related mental health (Warr, 1994; De Jonge and Schaufeli, 1998). There are three main axes in measuring job-related affective well-being: discontented-contented, anxious-comfortable and depressed-actively pleased (Warr, 1987a). Warr (1987b, 1990) has developed psychometrically sound scales to measure the full range of the three axes; measures such as job satisfaction, job-related tension, occupational burnout and fatigue are a few such examples.
Depression and anxiety although empirically interdependent are often considered to arise primarily from separate sources. Depression is more likely to be associated with loss or deprivation, while anxiety is more a function of threat or danger. Environmental features such as opportunity for control and opportunity for skill use are usually considered important at levels representing deprivation, whereas job demands often have their impact at high levels through presenting threats to the employee. Therefore, the vitamin model predicts that job demands are more closely linked to anxiety-comfort, whereas decision latitude is more strongly associated with depression-enthusiasm. This prediction was supported by Warr (1990) who found a non-linear relationship between job demands and decision latitude and employee well-being. The strength of this study was the large sample size, as non-linear relationships can only be examined with a large heterogeneous sample (Warr, 1994).
This prediction was further supported by De Jonge and Schaufeli (1998) who also found support for a non-linear relationship between job demands and job-related anxiety, job autonomy with emotional exhaustion (depression), and workplace social support with job satisfaction (discontented-contended) and emotional exhaustion. The results of this study have shown the validity of the three-axial framework predicted by the vitamin model in relation with the job characteristics (nine environmental features), indicating that different aspects of well-being are differentially associated with various job characteristics. This study was of a cross-sectional nature, which limits causal interpretations of the relationship between job characteristics and employee well-being.
Xie and Johns (1995) also investigated the curvilinear relationship and found support for the vitamin model. Xie and Johns showed a non-linear relationship between job scope (five job characteristics were measured here skill variety, task identity, task significance, autonomy and feedback) and emotional exhaustion and anxiety. Again, this study was based on cross-sectional methodology and therefore limits causal inferences.
Rydstedt et. al. (2006) used the Whitehall II study, which was a longitudinal investigation looking at the degree and causes of the social gradient in morbidity and mortality (Marmot et. al., 1991) to test the curvilinear relationship predicted by the vitamin model. However, Rydstedt et. al. found little support for the non-linear relationship between the job characteristics (social support, decision latitude, job demands and job satisfaction) and mental health. This study meets the requirements of a large sample size (Warr, 1994); however, the sample consisted mainly of office workers which may partly explain the contrasting results (Warr, 1990; De Jonge and Schaufeli, 1998). Participants were followed for 5.8 years, (with dropout rates of 13% between phases 3 and 5); it could be that the timespan may be too long to explain non-linear relationships. Due to a lack of longitudinal research on the vitamin model, it remains to be seen whether a shorter timespan may support the non-linear component.
Jeurissen and Nyklicek (2001) argue that little attention has been given to individual differences between people in studies testing the vitamin model. Therefore, in their study they controlled the effects of personality traits negative affectivity (NA) and positive affectivity (PA) on the relationship between job characteristics and well-being. According to Warr (1996) NA and PA refer to individual differences in emotional style and feelings about oneself, with both traits having a general influence on an individuals affective responses to features and events in the environment. Jeurissen and Nyklicek found weak support for the non-linear relationship between job characteristics and mental health, and found partial support for the assumption that job demands and autonomy are more strongly related to well-being than personality characteristics NA and PA. Job demands remained as the major predictor of well-being when NA and PA were controlled; however, autonomy was found to be non-significant when controlling NA and PA. The inconsistent results were due to the small, predominantly female and homogenous sample.
In summary, the vitamin model is similar to the JD-C model as both suggest that job characteristics affect the mental well-being of employees. Although the vitamin model embraces broader job characteristics, there is limited evidence on all job categories in relation to mental health (Warr, 1987a, 1990, 1996). Demands in the JD-C model correspond to externally generated goals in the vitamin model, while control is similar to opportunity for control and opportunity for skill use, and social support in the JD-C model corresponds to opportunity for interpersonal contact in the vitamin model. Most research has also mainly focused on these job characteristics which are also present in the JD-C model (Warr, 1990; Xie and Johns, 1995; De Jonge and Schaufeli, 1998 and Rydstedt et. al. 2006).
The vitamin model assumes a curvilinear relationship between job characteristics and mental health (affective well-being) suggesting that, with extreme increases in job features, there is likely to be a decrease in affective well-being of individuals. While increase in some job features (availability of money, physical security) may produce a constant effect, for example an increase in salary may produce beneficial effects at first, however, once a plateau (a certain required level according to need) has been reached, the job characteristic has no further beneficial effect and the level of health remains constant. The non-linear stance of the vitamin model has been supported to a certain extent, however, due to the majority of research being of a cross-sectional nature, it limits the statistical power of the results. Although Rydstedt et. al. (2006) did try to address this issue in their longitudinal research, their study found no support for the non-linear relationship. Due to there being paucity of longitudinal research on the vitamin model, the assumption of the vitamin model of a non-linear relationship remains incomplete.
Therefore, there is a need for a more multivariate and longitudinal study, taking into consideration the broad job characteristics introduced by the vitamin model in examining the relation to mental health in order to determine the curvilinear association of the model. The longitudinal methodology should consider a shorter time-lag in comparison to Rydstedt et. al. (2006) who found no support for the non-linear association. However, although this study measured the appropriate job characteristics in relation to mental well-being, this study was based on the degree and causes of the social gradient in morbidity and mortality (Marmot et. al., 1991). Therefore, there is a need for a longitudinal study specifically designed to test the vitamin model in mind.
De jonge, J. & Schaufeli, W.B. (1998) Job characteristics and employee well-being: a test of Warrs Vitamin Model in health care workers using structural equation modelling. Journal of Organizational Behaviour, 19, pp.387-407.
Jeurissen, T. & Nyklicek, I (2001). Testing the Vitamin Model of job stress in Dutch health care workers. Work & Stress, 15 (3), pp.254-264.
Karasek Jr., R. A. Job demands, job decision latitude, and mental strain: Implications for job design. Administrative Science Quarterly, 1979, 24, pp285-308.
Karasek, R. A. & Theorell, T. Healthy work: Stress, productivity and the reconstruction of working life. New York: Basic Books, 1990.
Marmot, M.G., Smith, G.D., Stansfeld, S., Patel, C., North, F., Head, J., et. al. (1991). Health inequalities among British civil servants: The Whitehall II study. Lancet, 337, pp.1387-1393.
Rydstedt, L.W., Ferrie, J., & Head, J. (2006). Is there support for curvilinear relationships between psychosocial work characteristics and mental well-being? Cross-sectional and long-term data from the Whitehall II study. Work & Stress, 20 (1), pp.6-20.
Warr, P. (1987a). Work, unemployment, and mental health. Oxford: Clarendon Press.
Warr, P.B. (1987b). Job characteristics and Mental Health. In P. B. Warr (Ed.), Psychology at Work (3rd ed., pp.247-269). Harmondsworth: Penguin Books.
Warr, P. B. (1990) Decision latitude, job demands, and employee well-being. Work & Stress, 4 (4), pp.285-294.
Warr, P. (1994). A conceptual framework for the study of work and mental health. Work and stress, 8 (2), pp.84-97.
Warr, P. B. (1996). Employee well-being. In P. B. Warr (Ed.), Psychology at work (3rd ed., pp.224-235). Harmondsworth: Penguin Books Ltd.
Xie, J. L. & Johns, G. (1995). Job scope and stress: can job scope be too high? Academy of Management Journal, 38, pp.1288-1309.