خصومت بدبینانه و مدل آسیب پذیری روانی اجتماعی خطر ابتلا به بیماری: مخدوش کننده اثرات روان رنجوری سوگیری(عواطف منفی)

Results obtained from a sample of young adults showed that, independent of the effects of neuroticism/negative affectivity (NA), Cook–Medley hostility (Ho) scores were significantly associated with life stress, trait anger, loneliness and irrational beliefs but not social support. The strength of relationships linking Ho scores to these risk and protective factors was noticeably stronger when the confounding effects of neuroticism-stability were not statistically controlled. When neuroticism was covaried, there was a 67.7% relative reduction in mean effect size. We tentatively conclude that self-report bias associated with dispositional neuroticism/NA may represent a serious threat to the internal validity of self-report studies that test the psychosocial vulnerability model of disease risk associated with cynical hostility. Results from our partial correlations provide only partial support for the theory that Ho scores confer increased health risk through a negative psychosocial `profile' characterised by the relative imbalance between factors that are stressors and factors that are coping resources.

Numerous studies over the past 30 years have provided evidence to support the conclusion that cynical hostility, as measure by the Cook–Medley Hostility (Ho) scale, is a psychosocial risk factor for death, medically-related disability, and ill-health (for reviews, see Miller, Smith, Turner, Guijarro, & Hallet, 1996; Smith & Ruiz, 2002). Considerable research has also been done to uncover the mediating biopsychosocial mechanisms or processes through which hostility impacts upon health outcomes. As articulated by Smith and his colleagues (Smith, 1992; Smith & Christensen, 1992; Smith & Gallo, 1994; Smith & Pope, 1990), the psychosocial vulnerability model of health risk posits cynical hostility is associated with a unique `clustering' of risk and protective factors. In particular, the model argues Ho scores are related to an unhealthy constellation of psychosocial factors that work synergistically to confer increased susceptibility to disease, ill health and medical dysfunction. Essentially, the model represents a stress-vulnerability framework because it argues that high hostile people are at elevated risk for medical problems because they experience an excessive number of external and internal stressors (illness vulnerability factors) concurrent with deficiencies in social support and other types of coping resources (illness protective factors). As noted by Smith ( Smith & Christensen, 1992; Smith & Ruiz, 2002), this combination of heightened vulnerability and deficient coping resources may be involved in the onset and maintenance of ill-health in a number of ways: (1) by increasing the frequency, intensity and duration of episodes of physiological reactivity; (2) by increasing health impairing behaviour patterns involving consumption of alcohol/drugs or food or cigarette smoking, (3) by increasing the probability of maladaptive illness behaviours, or (4) by impeding adjustment to and recovery from illness or medically-related disability. Evidence from a number of studies supports the contention that Ho scores are associated with the presence of stress-related vulnerability factors (Davidson, Prakachin, Lefcourt, & Mills, 1996; Greenglass & Julkunen, 1991; Hardy & Smith, 1988; Houston & Vavak, 1991; McCann, Russo, & Benjamin, 1997; Pope, Smith, & Rhodewalt, 1990; Rosenberg, Ekman, & Blumenthal, 1998; Smith, Pope, Sanders, Allred, & O'Keefe, 1988; Smith, Sanders, & Alexander, 1990) and the absence of coping-related protective factors such as social support (Davidson et al., 1996; Elovainio, Kivimaecki, Kortteinen, & Toumikoski, 2001; Hardy & Smith, 1988; Hart, 1996 and Hart, 1999; Houston & Vavak, 1991; Kivimaeki et al., 2002; McCann et al., 1997; Raynor, Pogue-Geile, Kamarck, McCaffery, & Manuck, 2002; Smith et al., 1988; Swan, Carmelli, & Rosenman, 1990). Although the Ho scale has been shown to be related to the presence of risk factors and absence of resource factors in a manner consistent with expectations drawn from the psychosocial vulnerability model, the results of this research are difficult to interpret because of ambiguity regarding the construct validity of the 50-item Ho scale. Empirical findings from a number of studies question the psychological meaning of Ho scores (Contrada & Jussim, 1992; Han, Weed, Calhoun, & Butcher, 1995; Steinberg & Jorgensen, 1996), leading some researchers to reconsider how well the scale actually measures `cynical hostility'. Of special concern to the present study are findings suggesting the Ho scale contains surplus construct irrelevancies associated with neuroticism/negative affectivity. In particular, previous research has found the Ho scale is so highly correlated with other MMPI-2 subscales as to support the interpretation that Cook–Medley items are confounded by neuroticism or `negative affectivity'1 (Han et al., 1995). This finding is consistent with results reported by Denollet (1991), who compared the Ho scale with a measure of negative affectivity (NA), showing the Ho scale shared 16% common variance with the personality trait of neuroticism (assessed via trait anxiety). Other research has also provided evidence that Ho scores are significantly and positively correlated with measures of NA/neuroticism (Carmody, Crossen, & Weins, 1989; Gallo & Smith, 1999; Smith & Pope, 1990; Swan et al., 1990). These and other empirical findings have led McCann and her colleagues to question “whether the Cook–Medley Ho scale is simply a reflection of the more general trait of negative affectivity” (McCann et al., 1997, p. 181). This ambiguity in the meaning of Ho scores raises the possibility that significant findings reported by previous research may have less to do with the specific effects of `cynical hostility' per se and more to do with the broader personality trait of neuroticism/NA. To the extent to which both the Ho scale and the psychosocial factors to which it has been correlated share variance in common with neuroticism, effect size estimates based on previously reported relationships may be spuriously inflated due to shared error variance. This threat to internal validity suggests increased methodological rigor is needed in research examining the risk and resource factors associated with the Cook–Medley Ho scale. The purpose of the present study was to replicate, extend and refine previous research pertaining to the psychosocial vulnerability model of hostility-related health risk. Using improved methodology, we tested the suggestion that Ho scores might possibly confer increased health risk through a negative psychosocial `profile' characterised by the relative imbalance between stress-related factors and coping-related factors. In particular, we compared the pattern of psychosocial associations to the Ho scale before and after statistically covarying the influence of neuroticism bias. If neuroticism acts as a confounding `nuisance variable' as we suspect it will, relationships that were formerly significant at zero-order should diminish in magnitude and significance after covarying neuroticism at first-order. Thus, we anticipate the mean effect size linking Ho scores to the psychosocial correlates should shrink after removing common error variance due to neuroticism bias. Consistent with previous research, we predicted that Ho scores would be significantly and positively related to a host of vulnerability variables: specifically, stress, negative emotional arousal (anger and loneliness) and irrational beliefs. We also expected Ho scores would be significantly and inversely related to protective variables such as total social support and components of support availability.