Current issues and new directions in Psychology and Health: The potential contribution of health psychology to developing effective interventions to reduce tobacco smoking

This issue of Psychology and Health illustrates well the breadth of research being undertaken in the study of the psychology of smoking. The studies are diverse but from the perspective of a researcher whose main area of study is tobacco use, but who is also a Health Psychologist, a common thread is evident: there is considerable promise in the ideas emerging from Health Psychology for the study of smoking, but so far it has not delivered interventions that are demonstrably better than those that derive from a simple view of smoking as a conflict between concern over the harms of smoking and the often strong want or need to smoke arising from social and pharmacological reward and punishment. To establish this common thread this editorial will briefly summarise what seem to be the key findings of each study and set this in the context of past and possible future research.

Balmford, Borland, and Burney (2010) report secondary analyses from a study of the effectiveness of telephone counselling which appear to show that predictors of quit attempts at 3-month follow-up differ in those who had recently made a quit attempt from those who had not. There is a question of how robust these findings are given that there was no direct statistical test of the interaction between past quit attempts and the predictors, only an interpretation of differences in the size of associations. Also, the nature of the sample may have played a role in that all the participants had volunteered to take part in an intervention study. Considering this in the context of past research, there appear to be some strongly held views in the field about predictors of quit attempts in population samples (e.g. self-efficacy, gender), but the results are not fully consistent and no systematic review has yet been published to determine whether these views are warranted. Such a review needs to take longitudinal studies of population samples and predict separately the incidence of quit attempts and the success of quit attempts. Such a review will need to acknowledge the possibility of the kind of moderator effect reported in this study.

de Leeuw, Scholte, Vermulst, and Engels (2010) report results from a longitudinal study of adolescents which confirm previous findings that parental discussions about smoking with children that were respectful and constructive were associated with non-smoking status in the children at later time points. This association was over and above a generally supportive parenting style. Even with longitudinal studies, establishing causal associations is problematic in this area. There are now a large number of such studies of adolescent smoking and impressionistically a picture is emerging of vulnerability factors and triggers that promote or protect against smoking uptake. The picture is complicated by different definitions of uptake, but as with the previous study the time is now ripe for a systematic review so that we can establish with confidence what predictors of uptake are robust and begin to construct a comprehensive model.

Farrimond, Joffe, and Stenner (2010) report findings from a study in which a sample of smokers sorted statements about their relationship to smoking in order to establish a typology of identities. The authors labelled the types of identity they found as: ‘addicted’, being ‘in control’, smoking being ‘no big deal’ and being ‘proud’. As the authors point out, there is very little research into adult smoker identities and it seems reasonable to postulate that identities play some causal role in maintaining or stopping smoking. However, this remains to be seen and indeed there is a substantial research agenda examining how far identity as opposed to specific responses to, concerns and desires relating to smoking predicts cessation.

Murnaghan et al. (2010) report significant correlations between Theory of Planned Behaviour measures (attitude, subjective norm, perceived behavioural control and intention) and smoking behaviour 1 month later in a sample of adolescents. Previous research has found similar sizes of association cross-sectionally but predicting take up of smoking or its cessation has proved much more difficult. Given that smoking is quite a stable behaviour pattern even at the age studied, the findings could reflect reverse causation which is what has been presumed to underlie the discrepancy between longitudinal and cross-sectional findings in the literature.

Oettingen, Mayer, and Thorpe (2010) report the results of an experiment which appeared to find that smokers who had high expectations of success in reducing their smoking had more immediate intentions to stop in response to being encouraged to imagine the positive consequences of stopping and negative consequences of continued smoking than those with lower expectations of success. The idea of encouraging visualisation of outcomes to generate motivation for change is something that has a long history as is the view that people are more likely to try to do something if they believe that they will succeed. Self-reported smoking reduction bears little or no relationship to reduction in exposure to smoke constituents, so will be interesting to see whether the use of imagery will show similar results when it comes to actual attempts at cessation and success of those attempts.

Reimer, Gerrard, and Gibbons (2010) examined ethnic differences in the US concerning beliefs about the harms of smoking. They found that ‘African Americans’ were more likely than whites to believe that lung cancer could be cured and that physical activity and vitamins could help protect against the disease. ‘African Americans’ also, understandably, were not aware of the evidence that they were statistically at greater risk of smoking-related diseases than ‘European Americans’. Differences between different cultural and ethnic groups in terms of psychological factors in tobacco addiction have not been widely studied and considering the number of different ethnic groups in countries across the world it is difficult to imagine that anything like a general set of principles could be adduced. However, this study illustrates how descriptive research can potentially lead to ways of targeting interventions to benefit particular groups.

Rivis, Sheeran, and Armitage (2010) take perhaps the broadest theoretical position of the series of papers. Their paper recognises that behaviours at a given moment can arise from prior intentions, desires that are not blocked by prior intentions (labelled ‘willingness’), identification with a positively evaluated stereotype, and habit. They go further and argue that the more reflective sources of motivation will dominate in individuals who tend to spend more time and effort reflecting on their actions. This captures some elements of the PRIME Theory of motivation (West, 2006) which has been developed to gain a more comprehensive understanding of motivation in general and has been applied to tobacco dependence (West, 2009). This proposes that conscious intentions can only influence our actions at a given moment if they generate sufficiently strong ‘wants or needs’ at that moment to overcome wants and needs arising from the immediate environment and impulses driven by stimulus–impulse associations (habit). PRIME Theory further proposes that a strong-coherent identity can be a very important source of stable wants and needs and, depending on the individual, can bind us strongly to our intentions. Rivis et al.'s (2010) findings appear to show the importance of habit, willingness and what I would call identity, but they call stereotype activation, in predicting smoking behaviour in adolescents. The significance of the finding, as with Murnaghan et al.'s (2010) study, is arguably weakened by the very short follow-up period, but this appears to be a line of research with considerable promise.

This diverse collection of studies illustrates a common thread in the psychology of smoking: the research is full of promise but has yet to ‘deliver’. It lies largely in the realms of developing ideas and finding partial support for interesting hypotheses. As stated at the start of this editorial, to date we do not have interventions arising from it that are more effective than ones that derive from a simple model of smoking based on the conflict between concerns about negative aspects and the desire (often very strong) to smoke because of social and pharmacological rewards and punishments. The Cochrane Library provides a comprehensive set of systematic reviews of intervention evaluations. We need a comparable set of reviews of observational studies and tests of theories. With this and the richness of ideas evident in the articles in this issue of Psychology and Health, there is every reason to be optimistic that this situation will change within the next 5 years.