Active but not consistent: dietary behaviour and the stages of change model

Abstract

In Stages of Change studies on dietary fat, standard algorithms allocate many respondents to the maintenance stage and their stage allocations are not always supported by objective criteria. In defining the preparation stage standard instruments emphasise inactivity allied to confident intention to change. In contrast Lamb and Joshi's (1996) staging instrument emphasises inconsistent effort, and thereby allocates inconsistently active respondents to preparation when a standard instrument might place them in action or maintenance. In the current cross-sectional study 375 respondents completed a standard algorithm, the Lamb and Joshi instrument and a 24-hour recall food diary. They also gave judgements of their current weight and diet. Unlike the standard instrument, the Lamb and Joshi instrument: (a) allocated fewer respondents to action/maintenance, (b) discriminated key behavioural stages by percent fat in diet, and (c) captured non-linear discontinuities in the respondents’ weight and diet self-perceptions. Such discontinuities in the psychosocial processes are predicted by the transtheoretical model, and support the claim that its stages are categorically distinct rather than artificially isolated points on a continuum. Distinguishing between respondents who describe themselves as consistently active and those who describe themselves as inconsistently active may help to overcome some apparent problems in the application of the transtheoretical model to behavioural change.

Keywords:

• Transtheoretical model
• Stages of Change
• Stage instruments
• Dietary fat
• Discontinuity patterns

Acknowledgements

We would like to thank the Health Education Authority for funding this project; the employees of ICI Slough for participating in the study; Kaushik Bose for data entry, and Jane Thomas (Department of Nutrition and Dietetics, King's College, London) for assistance in designing the dietary measure and overseeing the analysis of the nutritional information.

Notes

 Precontemplation is defined by two statements in order to capture the two aspects of either lacking information or being unwilling to give it serious consideration.

 In a different area of health, (viz. exercise) Prochaska and his colleagues have themselves recognised the importance of clarity and brevity in the design of staging instruments (Reed et al., 1997). After extensive comparisons of eight staging instruments they recommended an instrument very much like the Oxford instrument in that respondents allocate themselves to one of five categories each defined by a single short statement. They are asked “Do you exercise three times a week for at least 20 min each time?”, and are allowed one of five responses, for example, for preparation “No, but I am planning to start in the next 30 days”. Note that this phrasing continues to restrict preparation to its inactive component. Inspired by the work on staging instruments for exercise, Armitage and Conner (2001) devised a similar simple instrument for dietary fat. Although it defines preparation as “some attempt to eat a low fat diet”, they still found only a statistically significant binary distinction between the dietary fat of those in precontemplation and those in action and maintenance.

 We are confident in our measure for a variety of other reasons. Firstly, the fact that the 24-hour recall questionnaire was administered by post rather than in a face-to-face interview and was not administered in the context of an intervention is likely to have minimised social desirability response bias (Kristal et al., 1998). Secondly, the respondents were offered a nutritional analysis of their diet which prima facie should enhance the truthfulness of replies. Fifty per cent of respondents did ask for such feedback. Thirdly, the fact that 48% of respondents offered a “typical” day's 24-hour recall in addition to “yesterday's” diet suggests that the task was taken seriously.

 Even precontemplators had a more negative view of the health of their diet than those in action and maintenance despite the fact that they did not think the amount of fat in their diet needed changing (see Fig. 4). Some authors (e.g., Kristal et al., 1999) have discussed the possibility that some apparent precontemplators may have no plans to change the fat in their diet because they have really been in maintenance for a considerable length in time. The relatively negative self-perceptions of diet among precontemplators in the current study reassure us that these respondents are unlikely to be in maintenance. In addition the wording of the Oxford instrument makes in very unlikely that someone who is really in maintenance and therefore being “vigilant” about their diet (Kristal et al., 1999) would tick either of the statements that would allot them to precontemplation (see Fig. 2).