روان رنجوری،تعصبات پیش توجه و توجهی نسبت به تهدید و اضطراب قبل و بعد از یک عامل استرس زای شدید: یک مطالعه آینده نگر

This study was designed to prospectively investigate the role of neuroticism, trait anxiety and attentional biases towards threat in the development of anxiety after a failed IVF or ICSI treatment. A subliminal and supraliminal Stroop task was administered to 49 women entering IVF or ICSI treatment as well as self-report measures of neuroticism, trait anxiety and state anxiety. The assessment of state anxiety was repeated after failed treatment. Results of concurrent analyses have indicated that there was no relationship between Stroop factors and state anxiety. As hypothesised, prospective analyses showed a significant effect for neuroticism and Stroop interference on anxiety response to failed IVF or ICSI treatment. This effect was only apparent for interference on stressor related words and not for words related to general threat. The effect of subliminal Stroop interference was more pronounced in a subsample of women showing an increase in anxiety as a result of failed treatment. It can be concluded that, in addition to neuroticism, subliminal Stroop interference is a predictor for emotional response to an aversive stressor, such as failed IVF or ICSI treatment.

Stress vulnerability models differ in their attention to factors supposed to be the main predictors of the development of anxiety disorders. Models of personality dimensions stress the importance of neuroticism and trait anxiety, whereas cognitive models stress the importance of information processing. Until now, studies have mainly focussed on one of these two models, as a result of which relatively less attention has been paid to the additional contribution of factors from both models in the development of anxiety disorders. Information about this would increase our understanding of how to identify which individuals are more vulnerable than others to developing anxiety. The object of the present study is to illuminate this by investigating the concurrent and prospective relationship between factors of both models and state anxiety, before and after the occurrence of an aversive stressor. Personality theories have stressed the importance of negative affectivity and underlying constructs, such as neuroticism and trait anxiety. It is assumed that individuals vulnerable to developing anxiety disorders are characterised by enhanced sensitivity to aversive stimuli (Clark et al., 1994, Eysenck, 1970, Eysenck, 1998 and Watson & Clark, 1984). This enhanced sensitivity is the main characteristic of negative affectivity. Neuroticism can be defined as a general sensitivity to negative information, trait anxiety as an enhanced sensitivity to threatening information. Neuroticism and trait anxiety are generally assessed by self-report measures, like the Eysenck Personality Questionnaire (Eysenck & Eysenck, 1975) and the Spielberger Trait and State Anxiety Inventory (Spielberger, 1983). Cognitive theories have proposed that enhanced vulnerability is explained by the selective processing of threatening information (Williams et al., 1996 and Mogg & Bradley, 1998). In anxiety, selective processing is shown as an attentional bias towards threatening information. Pathological fear occurs when a fear structure is activated without further processing of threatening information (Foa & Kozak, 1986). Further processing involves incorporation of new information that is incompatible with the fear structure and results in its modification. Activation of the fear structure causes an orientation towards threat in the initial phase of information processing. Relaxed subjects are better able to entirely process threatening information, whereas this entire processing takes too much effort when subjects are under stress (Foa & Kozak, 1986). Foa and Kozak did not differentiate between preconscious and conscious activation of fear structures. Later theories, like those of Beck and Clark (1997) and Mogg and Bradley (1998), proposed that biases in the processing of threatening information take place on preattentive (preconscious) and attentional (conscious) levels. Preattentive and attentional biases towards threatening information are most frequently identified with an emotional Stroop task. In the emotional Stroop task, subjects are asked to name the colour of a word as quickly as possible, while ignoring the meaning of the word. Colour-naming interference is positively related to the emotional valence of the word. Attentional biases towards general threat words are most frequently identified in clinical samples (see for reviews: Mathews & Macleod, 1994, Mogg & Bradley, 1998 and Williams et al., 1996). A few studies have also reported a preattentive bias towards threatening information in clinical samples ( Lunch et al., 1999 and Mogg, Mentish et al., 1993). In non-clinical samples, there is less evidence of a relationship between anxiety and attentional biases towards threatening information. MacLeod and Ruthenford (1992) proposed that this could be due to the fact that highly anxious subjects in non-clinical samples are able to strategically override a possible processing bias for threatening information. This capacity is assumed to be lacking in patients with anxiety disorders. This might explain why an attentional bias towards threat-related words is indicated most consistently in clinical groups. However, it is possible that, while non-clinical subjects strategically override a processing bias on the supraliminal level, they are not able to do this on the subliminal level. Thus non-clinical subjects with high levels of trait anxiety could show a preattentive bias towards threat. Indeed, there is evidence for a relationship between preattentive biases towards threat and trait anxiety in non-clinical groups (Macleod & ruthenford, 1992 and Mogg, Mentish et al., 1993), but studies on this issue were based on small samples with high trait anxiety and revealed equivocal results. Accordingly, in clinical samples, there seems to be support for a concurrent relationship between anxiety and preattentive and attentional biases towards threat. In non-clinical samples, however, the relationship between both of these factors has been insufficiently demonstrated. So, the question concerning the concurrent relationship between preattentive biases towards threat and anxiety in a normal sample with respect to their anxiety levels remains unanswered. Up to now, prospective studies on the relationship between preattentive and attentional biases towards threat and anxiety have been rare. A prospective design is warranted in order to exclude the possibility of a preattentive or attentional bias being a consequence of anxiety, and not being a mediating factor in the course of anxiety. MacLeod and Hagan (1992) used a quasi prospective design in a sample of 32 gynaecological patients due for a colpescopy after a positive cervical test. Based on the results of their study, they concluded that the subliminal attentional bias before the stressor was a better predictor for emotional response to that stressor than self-report measures on state and trait anxiety. Supraliminal Stroop interference was not a predictor for emotional response. This supports the assumption of possible strategic overriding in non-clinical subjects with high trait anxiety. Caution in drawing firm conclusions should be exercised, however, because of the small sample size and the consequent limited power of their study. Until now, the prospective results of MacLeod and Hagan's study have only been partly replicated (Pury, 2002 and Van den Hout et al., 1995). Accordingly, the relative contribution of subliminal threat interference along with dispositional factors as assessed by self-report measures is still unclear. The present study was carried out for two purposes: firstly, to explore the concurrent relationship between neuroticism, trait and state anxiety and attentional and preattentive biases to threat in a non-clinical sample. Secondly, it was conducted to investigate the relative contributions of neuroticism, trait anxiety and Stroop interference to anxiety response to an aversive stressor. The study sample consisted of women about to start in vitro fertilisation (IVF) or intracytoplasmic sperm injection (ICSI) treatment. Before the start of the treatment, IVF or ICSI patients showed trait and state anxiety levels comparable to sex and age specific norm groups (Verhaak, Smeenk, Eugster, Van Minnen, Kremer, & Kraaimaat, 2001). The nature of the IVF/ICSI treatment provides a good opportunity to prospectively study the predictive value of neuroticism and processing biases for the development and consolidation of anxiety in a natural setting with non-clinical subjects before and after experiencing a severe stressor. The success rate of an IVF treatment is low (first cycle only about 25%), while its emotional and physical strain is considerable (for a review see: Eugster & Vingerhoets, 1999). A treatment cycle takes about 6 weeks, in which women first inject themselves with hormones to stimulate follicle growth. When follicles are sufficiently present, retrieval of oocytes takes place. Two days after fertilisation occurs, two embryos are replaced in the womb. A pregnancy test is performed two to three weeks after the embryo transfer. In the Netherlands, insurance companies pay for three treatment cycles. According to vulnerability models for the aetiology of clinical anxiety, a failed treatment cycle can be considered as a severe stressor able to trigger relatively high levels of anxiety in more vulnerable women. This study was undertaken to investigate both the concurrent and prospective relationship between self report measures of personality characteristics, state anxiety and Stroop interference. The results of studies into the concurrent relationship between these self report measures and Stroop interference revealed equivocal results, partly due to differences in characteristics of the samples of the studies. In the present study, we explored this relationship in a non-clinical sample of women entering fertility treatment. Theories on the processing of threatening information assume that preattentive and attentional biases towards threat is a determinant of enhanced vulnerability to develop an anxiety response as a result of a stressor (Mogg & Bradley, 1998 and Williams et al., 1996). Empirical evidence has indicated that attentional biases towards threat in non-clinical samples, seem to be cognitively overridden (MacLeod & Ruthenford, 1992), but there appears to be evidence for preattentive biases towards stress as vulnerability factors in non-clinical groups (MacLeod & Hagen, 1992, Pury, 2002 and Van den Hout et al., 1995). Accordingly, in the non-clinical sample of women with fertility problems, we expected a positive relationship between preattentive biases towards threat and change in anxiety after a failed treatment.