A study to explore the role of coherence in promoting the intentions to quit smoking (Hall et al. 2004).
This paper presents the results from two studies to examine the relationship between women’s beliefs about cervical cancer and their intentions to quit smoking. In particular the study explored whether, by making beliefs about threat and behaviour more coherent with each other, people are more likely to be motivated to change their behaviour.
Background
Research shows that women who smoke have twice the chance of developing cervical cancer than those who do not. Most women, however, are unaware of this association and, when told that smoking can increase the risk of cervical cancer, report finding this information confusing and nonsensical (‘how [can] smoking a cigarette in your mouth cause you problems down-stairs?’). Leventhal’s self-regulatory model illustrates that people represent their illness in the form of illness representations. He also argues that if people are to act on threats to their health they need to have a coherent model whereby their beliefs about the nature of the threat are coherent with their beliefs about any action that could be taken. The relationship between smoking and cervical cancer does not immediately make sense, suggesting that most people do not have a coherent model about the link between these factors. The present study used an experimental design to present women with a coherent model of how smoking is linked to cer-vical cancer and to explore whether a more coherent model was associated with a greater inten-tion to quit smoking.
Study 1
Methodology DesignThe study used an experimental design with women receiving either a detailed leaflet about cer-vical cancer and smoking, a less detailed leaflet or no leaflet.
Sample
The sample consisted of female smokers aged between 20 and 64 years who were recruited from two general practices in the UK.
Procedure
Women who received a leaflet in the post were then asked to complete a questionnaire a week later. Those who did not receive a leaflet were just sent the questionnaire.
Measures
Women completed measures of response efficacy (the extent to which stopping smoking would reduce vulnerability to cervical cancer), self-efficacy for smoking cessation, severity, coherence (the extent to which they believed they had a coherent explanation for the link between smoking and cervical cancer) and intentions to quit smoking in the next month. In addition, measures of smoking behaviour were taken.
The individual processes involved in the self-regulatory model will now be examined in greater detail.
Stage 1: interpretation Symptom perception
Individual differences in symptom perception
Symptoms such as a temperature, pain, a runny nose or the detection of a lump may indicate to the individual the possibility of illness. However, symptom perception is not a straightforward process (see Chapter 12 for details of pain perception and Chapter 16 for details of menopausal symptoms). For example, what might be a sore throat to one person could be another’s tonsilli-tis, and whereas a retired person might consider a cough a serious problem, a working person Data analysis
The results were analysed to assess the impact of the leaflets on women’s level of coherence, beliefs and intentions. The results were then analysed to assess the relationship between level of coherence, beliefs and intentions.
Results
The results showed that the detailed and less detailed leaflet were equally as effective at produc-ing a coherent model of the relationship between smokproduc-ing and cervical cancer and were both more effective than receiving no leaflet. The results also showed that those who received a leaflet (regardless of level of detail) reported higher vulnerability to cervical cancer, greater response efficacy and higher intentions to quit smoking. Finally, in terms of the relationship between coherence, beliefs and intentions the results showed that greater intentions, to quit smoking were predicted by greater coherence, greater perceptions of severity, higher response efficacy and higher self-efficacy. Furthermore, greater perceptions of vulnerability to cervical cancer only predicted intentions to quit smoking in those women who showed a coherent model of the link between smoking and cervical cancer.
Conclusion
The results therefore show the importance of a coherent model in creating a link between beliefs and intentions. The results also show that coherence can be changed by a simple leaflet intervention.
Study 2
Because of the weakness of some of the measures in Study 1, the authors replicated their study with a further 178 women smokers and included an improved measure of vulnerability.
The results directly replicated the findings of Study 1.
Conclusion
This study illustrates a role for coherence between beliefs about threat and subsequent action and supports Leventhal’s model. Therefore people would seem more likely to intend to change their behaviour if they have a coherent model as to how this behaviour may impact upon their health status.
might be too busy to think about it. Pennebaker (1983) has argued that there are individual differences in the amount of attention people pay to their internal states. Whereas some individuals may sometimes be internally focused and more sensitive to symptoms, others may be more externally focused and less sensitive to any internal changes. However, this difference is not always consistent with differences in accuracy. Some research suggests that internal focus is related to overestimation. For example, Pennebaker (1983) reported that individuals who were more focused on their internal states tended to overestimate changes in their heart rate com-pared with subjects who were externally focused. In contrast Kohlmann et al. (2001) examined the relationship between cardiac vigilance and heart-beat detection in the laboratory and reported a negative correlation; those who stated they were more aware of their heart underesti-mated their heart rate. Being internally focused has also been shown to relate to a perception of slower recovery from illness (Miller et al. 1987) and to more health-protective behaviour (Kohlmann et al. 2001). Being internally focused may result in a different perception of symptom change, not a more accurate one.
Mood, cognitions, environment and symptom perception
Skelton and Pennebaker (1982) suggested that symptom perception is influenced by factors such as mood, cognitions and the social environment.
Mood
The role of mood in symptom perception is particularly apparent in pain perception with anxiety increasing self-reports of the pain experience (see Chapter 12 for a discussion of anxiety and pain). In addition, anxiety has been proposed as an explanation for placebo pain reduction as taking any form of medication (even a sugar pill) may reduce the individual’s anxiety, increase their sense of control and result in pain reduction (see Chapter 13 for a discussion of anxiety and placebos and Chapter 16 for a discussion of anxiety and birth and menopausal symptoms). Cropley and Steptoe (2005) directly explored the relationship between recent life stress and general symptom reporting and found that higher stress was associated with an increased frequency of a range of symptoms. Stegen et al. (2000) explored the impact of negat-ive affectivity on both the experience of symptoms and attributions for these symptoms. In an experimental study, participants were exposed to low intensity somatic sensations induced by breathing air high in carbon dioxide. They were then told that the sensation would be either positive, negative or somewhere between and were asked to rate both the pleasantness and intensity of their symptoms. The results showed that what the participants were told about the sensation influenced their ratings of its pleasantness. The results also showed that although people who rated high on negative affectivity, showed similar ratings of pleasantness to those low on negative affectivity they did report more negative meanings and worries about their symptoms. This indicates that expectations about the nature of a symptom can alter the experience of that symptom and that negative mood can influence the attributions made about a symptom. Similarly, Mora et al. (in press) explored the role of negative affect on symptom perception and the processes underlying this relationship. Their study involved both a cross-sectional and longitudinal design and assessed trait and state negative affect in adults with mod-erate and severe asthma. The results showed that higher trait negative affect was related to higher reports of all symptoms whether or not they were related to asthma. In addition, the results showed that only those who were worried about their asthma attributed their asthma symptoms to asthma. This suggests that negative affect increases symptom perception; and further, that worrying about asthma enables the individual to associate their symptoms with
STAGE 1: INTERPRETATION 57
their illness. In line with this relationship between mood and symptoms, a recent study explored the impact of manipulating psychological stress on symptom perception (Wright et al.
2005). Using an experimental design, 42 patients with heartburn and reflux were exposed either to a psychological stressor or a no-stress control condition. They then rated their state anxiety and symptom perception. In addition, objective ratings of reflux symptoms were taken. The results showed that the stressor resulted in increased subjective ratings of symptoms. The stres-sor, however, did not result in any increase in actual reflux. Therefore the stressor resulted in a greater dissociation between subjective and objective symptoms. This study is important as it not only illustrates the impact of stress on symptom perception but also illustrates that gap between objective and subjective accounts of symptoms.
Cognition
An individual’s cognitive state may also influence their symptom perception. This is illustrated by the placebo effect with the individual’s expectations of recovery resulting in reduced symptom perception (see Chapter 13). It is also illustrated by Stegen et al.’s (2000) study of breathing symptoms with expectations changing symptom perception. Ruble (1977) carried out a study in which she manipulated women’s expectations about when they were due to start menstruating. She gave subjects an ‘accurate physiological test’ and told women either that their period was due very shortly or that it was at least a week away. The women were then asked to report any premenstrual symptoms. The results showed that believing that they were about to start menstruating (even though they were not) increased the number of reported premenstrual symptoms. This indicates an association between cognitive state and symptom perception.
Pennebaker also reported that symptom perception is related to an individual’s attentional state and that boredom and the absence of environmental stimuli may result in over-reporting, whereas distraction and attention diversion may lead to under-reporting (Pennebaker 1983).
One study provides support for Pennebaker’s theory. Sixty-one women who had been hospitalized during pre-term labour were randomized to receive either information, distraction or nothing (van Zuuren 1998). The results showed that distraction had the most beneficial effect on measures of both physical and psychological symptoms, suggesting that symptom perception is sensitive to attention. Symptom perception can also be influenced by the ways in which symptoms are elicited. For example, Eiser (2000) carried out an experimental study whereby students were asked to indicate their symptoms, from a list of 30 symptoms, over the past month and the past year and also to rate their health status. However, whereas half were asked to endorse their symptoms (i.e. mark those they had had), half were asked to exclude their symptoms (i.e. mark those they had not had). The results showed that those in the
‘exclude’ condition reported 70 per cent more symptoms than those in the ‘endorse’ condition.
In addition, those who had endorsed the symptoms rated their health more negatively than those who had excluded symptoms. This suggests that it is not only focus and attention that can influence symptom perception but also the ways in which this focus is directed.
Environment
Symptom perception is therefore influenced by mood and cognition. It is also influenced by an individual’s social context. Cross-cultural research consistently shows variation in the presenta-tion of psychiatric symptoms such as anxiety, psychosis and depression. For example, Minsky et al. (2003) explored diagnostic patterns in Latino, African American and European American psychiatric patients and reported that not only did the diagnoses of major depression and schiz-ophrenic disorders vary by ethnic group, but so did symptom presentation, with Latinos report-ing a higher frequency of psychotic symptoms than the other groups. Similarly, a consensus
statement by the International Consensus Group of Depression and Anxiety (Ballenger et al.
2001) concluded that there was wide cultural variation not only in the diagnosis and respon-siveness to treatment for depression and anxiety but also significant variation in symptom pres-entation. A similar pattern of variation can also be found for somatic symptoms such as headaches, fatigue, constipation and back pain although research in this area is less extensive.
For example, epidemiological studies indicate that while headache is a common symptom in the USA and Western Europe, its prevalence remains much lower in China and in African and Asian populations (e.g. Ziegler 1990; Stewart et al. 1996; Wang et al. 1997). Similarly, large surveys of primary care attenders report that those from less developed countries and from Latin America tend to report more somatic symptoms in general (Gureje et al. 1997; Piccinelli and Simon 1997). One study explored cataract patients’ reports of visual function and the extent to which they were bothered by their cataract and explored differences by culture (Alonso et al. 1998). The results showed that after controlling for clinical and sociodemographic characteristics, patients from Canada and Barcelona reported less trouble with their vision than patients from Denmark or the USA, suggesting cultural variation in the perception of visual symptoms. Symptom perception and diagnosis are therefore highly influenced by the indi-vidual’s context and cultural background.
Mood, cognition and environment therefore influence symptom perception. These different factors are illustrated by a condition known as ‘medical students’ disease’, which has been described by Mechanic (1962). A large component of the medical curriculum involves learning about the symptoms associated with a multitude of different illnesses. More than two-thirds of medical students incorrectly report that at some time they have had the symptoms they are being taught about. Perhaps this phenomenon can be understood in terms of the following:
■ Mood: Medical students become quite anxious due to their workload. This anxiety may heighten their awareness of any physiological changes, making them more internally focused.
■ Cognition: Medical students are thinking about symptoms as part of their course, which may result in a focus on their own internal states.
■ Social: Once one student starts to perceive symptoms, others may model themselves on this behaviour.
Therefore, symptom perception influences how an individual interprets the problem of illness.