SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. SUPPLEMENTARY MATERIAL
  8. Acknowledgments
  9. Disclosure
  10. REFERENCES
  11. Supporting Information

The present study aimed to establish, by a consensus of experts, the stages and processes of change for weight management in overweight and obese people. The first step involved developing two questionnaires aimed at assessing stages and processes of change for weight loss in overweight and obese people. The processes-of-change questionnaire consisted of 12 subscales, and contained 107 items. A three-round Delphi study was carried out through a website, where participants were asked to give their opinion about the representativeness and clarity of the scale items. The stages-of-change questionnaire consisted of five items and was presented in the final round of the study. A team of 66 experts in the obesity field from 29 countries participated in the study. They were selected either because they belonged to the organizing committee of international associations related to obesity, or because of their research career. The required changes in the questionnaire were made according to the opinions of the participants. Some of these were the result of the group statistical response, whereas others were due to the suggestions made by the participants. A final version of the questionnaire consisting of 63 items was eventually obtained. The present study produced two questionnaires to assess stages and processes of change for weight management. The strength of the study lies in the consensus reached by the panel of experts in order to establish the required content of the questionnaires. The two measures provide useful tools for practitioners who wish to tailor weight-management interventions according to transtheoretical model constructs.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. SUPPLEMENTARY MATERIAL
  8. Acknowledgments
  9. Disclosure
  10. REFERENCES
  11. Supporting Information

The transtheoretical model of behavior change (TTM) (1) is currently one of the most promising models related to the acquisition of healthy habits. Systematic reviews in this field show how the TTM has been widely applied to multiple health-change behaviors, such as substance abuse (2), diabetes mellitus (3), or exercise (4). However, more recently, there has been growing interest in applying the TTM to weight management in overweight and obese patients (5).

The model suggests that individuals engaging in a new behavior move through a series of stages of change, the first dimension of the model. Five stages have been defined by the literature (6). The first of them, precontemplation, is that at which there is no intention to change behavior in the foreseeable future (at least not within the next 6 months). Many individuals in this stage are unaware or under-aware of their problems. The next stage, contemplation, is that in which people are aware that a problem exists and are seriously thinking about overcoming it, but have not yet made a commitment to take action in the next 6 months. Serious consideration of problem resolution is the central element of this stage. Preparation is the stage that combines intention and behavioral criteria. Individuals at this stage are intending to take action in the next month and have unsuccessfully taken action in the past year. Action is a stage in which individuals have modified their behavior in order to overcome their problem for a period of no >6 months. Finally, maintenance is the stage in which people work to prevent relapse and consolidate the gains attained during action. The stages represent a temporal dimension that allows us to understand when particular shifts in attitudes, intentions, and behaviors occur.

The second major dimension of this model refers to processes of change. These processes represent basic change principles in which therapeutic approaches have been considered to be responsible for behavioral change (7). In other words, change processes are covert and overt activities and experiences that individuals engage in when they attempt to modify problem behaviors (8). Researchers have defined up to 12 processes of change that are powerful predictors of behavior change, some of which are experiential processes (consciousness raising, dramatic relief, environmental reevaluation, self-reevaluation, and social liberation), however, others are behavioral processes (counterconditioning, helping relationships, reinforcement management, self-liberation, stimulus control, substance abuse, and interpersonal control systems).

Despite the extensive research carried out on the TTM as applied to the promotion of behavior change, some authors argue against its validity. Their reasons are set out in West (9) and can be summarized as follows: the arbitrary boundaries between stages of change, the assumption of the model that behavior change is coherent, stable, conscious and planned, and the fact that stage definitions represent a mixture of different types of construct that do not fit together coherently. In the specific area of weight management the TTM has been applied in a wide variety of settings and samples. However, there is no agreement about its effectiveness when applying it to weight loss interventions. Indeed, those studies in which a weight-management intervention based on the TTM has been applied to adults have produced heterogeneous data. Results about the effectiveness of TTM interventions in terms of reducing BMI between baseline and follow-up are not very promising. Some studies have reported a medium effect size as regards a reduction in BMI (10,11), whereas others have shown no significant differences in weight reduction (12,13,14). Consequently, further research is needed to clarify the influence of the TTM in weight-management interventions.

Processes-of-change have been shown to be predictors of behavior change in interventions aimed at promoting healthy behaviors such as smoking cessation, eating behavior, exercise, and weight management (15,16,17,18,19,20). Processes of change are factors that produce transitions between stages. The change that occurs at different stages is qualitatively different, requiring different cognitive, emotional, and behavioral activities (21). Therefore, an effective treatment involves the differential employment of selected change processes at strategically critical times in the course of change (19).

Processes-of-change have been shown to be a valid model when applied to weight management (22). However, their use is mediated by other variables such as self-efficacy or decisional balance. The construct of self-efficacy (23) refers to the specific confidence that people have in their ability to make and maintain a behavior. The term decisional balance (24) concerns the way in which people choose to engage in a particular behavior change based on the perceived pros and cons that it represents. Both self-efficacy and decisional balance have been shown to be associated with weight loss outcome (25,26), as well as with the stages-of-change classification for weight loss (26,27). The use of processes of change can also be mediated by positive emotional states such as happiness and euphoria, or by negative emotional states such as anxiety, depression, or high levels of perceived stress. The cognitive effects of different emotional states can contribute positively or negatively to people's decision making or perceived self-efficacy. Consequently, it is also important to assess all these variables in weight-management interventions.

Despite the wide range of assessment tools designed to assess different aspects of the TTM, there remains a need for valid questionnaires to measure TTM constructs, particularly processes of change (21). Furthermore, there is the need for expert judges who can establish the item content validity of these questionnaires (28). However, at present no questionnaire for measuring the main constructs of the TTM has been developed by consensus of experts. Therefore, it would be useful to develop a questionnaire to assess stages and processes of change for weight management that was based on consensus among experts in this field. In this regard, the Delphi method could be useful in gathering the opinions of expert judges.

The Delphi technique may be considered as a method for structuring a group communication process so that the process is effective in allowing a group of individuals, as a whole, to deal with a complex problem (29). The main characteristics of this method were defined by Dalkey and Helmer (30): it is an iterative process in which the people taking part have to give their opinion about a subject (on which they are experts) more than once. In this way, experts have the option of reconsidering their opinion. It is also an anonymous process, so experts do not know who the other participants are, nor are they aware of their individual responses; this minimizes the potential inhibition of some participants. Another characteristic is that participants receive controlled feedback between rounds by means of a group statistical response. Then, from one round to the next, participants receive the general group response about the problem they had to solve.

Given the advantages of the Delphi method, when it comes to gathering the opinions of experts it has been widely used in medical sciences to obtain consensus about diagnosis (31,32,33,34). However, more recently it has been applied to analyze the content validity of questionnaires in multiple settings (35,36,37).

The present study aims to validate, by means of the Delphi method, the content of a questionnaire specifically designed to assess processes and stages of change for weight loss in overweight and obese people.

Methods and Procedures

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. SUPPLEMENTARY MATERIAL
  8. Acknowledgments
  9. Disclosure
  10. REFERENCES
  11. Supporting Information

Participants

Participants in this study, who formed the panel of experts, were mainly specialists in the field of obesity. Also, experts in the TTM were invited to join the panel of experts. Specialists were invited to take part in the study if they belonged to the organizing committee of international associations related to obesity (International Association for the Study of Obesity, International Obesity Task Force, European Society of Endocrinology). As these are international organizations, people from the organizing committee of all the headquarters around the world were invited to participate.

In order to be as exhaustive as possible, authors of specialized literature in the obesity field were also invited to take part. Finally, in order to have a delegation of specialists in the TTM, a bibliographic search was conducted in the ISI Web of Science and authors of articles related to the TTM applied to obesity were also invited. The outcome of this process was that 341 people from 65 countries were invited to participate in the present study.

Instruments

A questionnaire to assess processes of change and a brief scale to assess stages of change were specifically developed to be applied with overweight and obese people regarding weight loss. The questionnaire Processes of Change in Overweight and Obese People was developed according to the definition of the 12 processes of change put forward by Prochaska and DiClemente in 1985 (38). The procedure for test development was as follows. First, we collated several definitions of processes of change found in the literature (8,38,39,40,41,42,43,44). As these publications apply the TTM to many different health problems, new definitions regarding weight loss were developed (see Supplementary Data online). From these new definitions, items for each subscale were developed. Finally, we checked that items from other validated processes-of-change questionnaires (41,45) were included in the new scale, whenever possible. A final pool of 107 items thus formed the basis of the questionnaire.

The questionnaire Stages of Change in Overweight and Obese People was developed according to published definitions of stages (8,46,47). Five items corresponding to the five stages of change were developed to assess the readiness of overweight and obese patients to change, in order to lose weight. Both questionnaires were developed simultaneously in English and Spanish according the back translation process.

Procedure

The study was carried out through a website. As the questionnaires were developed in English and Spanish; the website was also presented in the two languages and participants could choose the language in which they would assess the questionnaires. Although there were two questionnaires to assess, the Processes of Change in Overweight and Obese People was presented first, whereas the Stages of Change in Overweight and Obese People was presented in the final phase of the study.

The process began with an email being sent to all selected experts, encouraging them to visit the website. The task they were given was to assess the content of the questionnaire Processes of Change in Overweight and Obese People, across the 12 subscales (Figure 1).

image

Figure 1. Assessment of the questionnaire through a website.

Download figure to PowerPoint

By selecting one of these subscales, its definition and items were shown (Figure 2). The task consisted in assessing, for each item, the following aspects:

image

Figure 2. First round: assessing representativeness and clarity of the items.

Download figure to PowerPoint

  • Representativeness. The extent to which the item represented the meaning of the subscale. Each item had to be ranked between 0 and 5. Thus, if experts believed that the item was very representative of the subscale definition (as a whole or in part), they had to score it as 5 (item very representative of the subscale). In contrast, if they considered that the item was not remotely representative of the subscale, they had to give it a score of 0.

  • Clarity. Experts also had to decide whether the item, as presented, was clearly expressed, i.e., that it was neither ambiguous nor confusing. Each item had to be ranked between 0 (the item was confusing, unclear) and 5 (the item could be understood by the scale's target population). In the event that the item was assessed as unclear, participants were encouraged to state their reasons in the “Observations” section.

  • Observations. Finally, they had the option of adding particular comments regarding the items or the whole subscale.

After the deadline for participating in this first round, all the opinions were gathered and a group statistical analysis was carried out. In the second round, only those experts who had participated in the first round were asked to reassess the questionnaire. Here, they had access to their own responses given in the previous round, as well as the overall responses of the panel of experts. The group response was pooled in two measures:

  • Median (Md). Obtained by placing all the expert responses in ascending order and selecting the middle one. Thus, half the values will be lower than the Md and the other half, higher.

  • Interquartile range (IR). Obtained from the difference between the third and the first quartile. Between these two quartiles we will find the central 50% of responses, and thus the IR tells us how dispersed the data are.

With this information, participants were asked to reassess the items regarding representativeness and clarity (Figure 3). As before, a data analysis was then conducted. In the third and final round, only those experts who had participated in the previous round were invited to take part in the final phase of the study. This time, they had access both to their own responses given in the second round and the group response (Md and IR), although the required task was the same as in the previous rounds. However, in addition to the items from the processes-of-change scale, participants were now asked to give their opinion about the questionnaire Stages of Change in Overweight and Obese People. Once again, they were asked to assess the representativeness and clarity of these items.

image

Figure 3. Second round: assessing representativeness and clarity of the items.

Download figure to PowerPoint

Data analysis

The data analysis was carried out between rounds and considered for both quantitative and qualitative aspects. The quantitative analysis was based on Md and IR in order to delete items that were poorly rated. We also applied the consensus criterion proposed by Hagen et al. (35) and defined a priori as >50% of respondents being in agreement with a statement. Here, consensus was calculated as the percentage of experts who agreed that the item was representative or clear (those who scored the item with 4 or 5 points), and consensus was then calculated for both variables (representativeness and clarity) and across all three rounds. At the same time, a qualitative analysis was performed on the basis of the comments and suggestions made by participants. Both the quantitative and qualitative analyses were complementary and were conducted during the three rounds as well as at the end of the study. This iterative analysis allowed us to obtain a refined final version of the questionnaire.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. SUPPLEMENTARY MATERIAL
  8. Acknowledgments
  9. Disclosure
  10. REFERENCES
  11. Supporting Information

First round

Participants in this first round were 66 experts in the obesity field. Therefore, the participation rate was 19.35% of the 341 experts originally invited. The panel of experts comprised 71% men and 29% women and had a worldwide representation (Table 1).

Table 1.  Representation of the panel of experts in the first and second round
inline image

The formal education of participants was mainly medicine (60.6%), dietetics and nutrition (33.3%), and psychology (15.2%). However, other experts were trained in nursing (3.0%), sociology (1.5%), and biology (1.5%). Almost 30% of participants were formed in more than one discipline. As shown in Figure 4, the experience of the panel of experts in the obesity research field and clinical practice was, on the whole, >15 years in both areas.

image

Figure 4. Professional experience of the panel of experts in the first round (n = 66) and in the second round (n = 22).

Download figure to PowerPoint

Participants were also asked to rate their knowledge of the TTM on a scale from 0 to 10. The mean score for knowledge about the model was 5.4 (s.d. = 3.1). Furthermore, 27.4% of them stated that they had published at least one paper about the TTM.

Once the opinions of all participants had been gathered the suggested changes to the questionnaire were made. First of all, some of them were deleted after calculating the Md and IR regarding representativeness and clarity of the 107 items. The criteria for deleting items were as follows:

  • Global score of representativeness ≤3. This meant that experts believed that the content of the item did not fit in the subscale for which it was developed.

  • IR ≥2. This meant that there was no agreement about whether the item was suitable or not in the subscale.

  • By suggestion of some of the participants.

In addition, a consensus of experts was calculated (see Supplementary Data online). The required vocabulary changes were also made in the event that an item was regarded as not clearly comprehensible by the target population (global score of clarity ≤3, or by suggestion of some of the participants). Moreover, some content needed to be added and other changes were made as requested by the experts. A new version of the questionnaire comprising 87 items was obtained.

Finally, the average consensus of experts was calculated by taking into consideration those items that had been assessed in this first round and which had also been selected to form the questionnaire that would be shown in the second round. The average consensus was 72.2 and 78.4% for representativeness and clarity, respectively.

Second round

Only participants from the first round were invited to continue participating in the study. The participation rate in this round was 33.3%, and the panel consisted of 22 experts (68% men and 32% women). The representation of the panel of experts for this second round is shown in Table 1. The formal education of the panel was similar to that in the previous stage: mainly trained in medicine, dietetics and nutrition, and psychology (54.4, 36.8, and 36.4%, respectively). A further 4.5% had trained in nursing and 1.6% in sociology. Again, some of participants (22.7%) were trained in more than one discipline. Similar to the first round, most of the participants had extensive experience in both the research field and clinical practice (Figure 4).

The level of knowledge about the TTM was slightly higher in this round. The panel of experts had a mean knowledge score of 6.7 (s.d. = 3.0) and 41% had published articles about this model.

Once again, some items were deleted according to the same criteria as in the first round. At the end of the second round a new version of the scale was obtained that consisted of 67 items. Consensus was again analyzed regarding those items that were assessed in this second round and which were selected either to be shown in the third round or to form part of the final version of the questionnaire. This time, the average consensus was slightly higher compared to the previous round: 83.5% for representativeness and 85.1% for clarity.

Third round

In this third and final round, only a few items from the scale Processes of Change in Overweight and Obese People were presented on the website. This time, participants had to reassess only those items that had not been assessed twice, in other words, items that had been added in the first or second rounds. Also, the five items of the questionnaire Stages of Change in Overweight and Obese People were now included.

Only the 22 participants from the second round were invited to take part in this final phase, and all of them gave their opinion about the items (participation rate = 100%). Once again, any items that met the required exclusion criteria were deleted and a number of other changes related to inappropriate vocabulary or content were also made. None of the experts proposed adding content in this final phase. After the changes made in the third round a final version of the questionnaire Processes of Change in Overweight and Obese People was obtained, consisting of 63 items. The average consensus of participants in this round was 87.3% for representativeness and 88.2% for clarity (taking into consideration only those items about which participants expressed an opinion and which were also selected to be in the final version of the questionnaire).

As regards the questionnaire Stages of Change in Overweight and Obese People, some vocabulary changes were made following suggestions by the panel of experts. Consensus about these items was favorable for both representativeness (ranging from 72.7 to 90.9%) and clarity (ranging from 78.2 to 91.3%). Finally, the average consensus was 80% for representativeness and 80.9% for clarity.

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. SUPPLEMENTARY MATERIAL
  8. Acknowledgments
  9. Disclosure
  10. REFERENCES
  11. Supporting Information

The present study describes the first questionnaires to be developed by a consensus of experts in the obesity field. As processes of change are powerful predictors of behavior change, their assessment is important in weight-management interventions, as they provide information about how the patient is tackling his or her weight problem. The stages-of-change classification also complements the assessment of processes of change by identifying when behavior change occurs. Consequently, this study has sought to establish the processes and stages of change for weight loss in overweight and obese people by means of a consensus of experts in the obesity and TTM field.

The present research also shows how a Delphi study can be used as a tool to assess the content validity of a questionnaire. Furthermore, although there is a lack of valid questionnaires to assess TTM constructs when the model is applied to weight loss, this study provides two questionnaires based on a consensus of experts and which have been specifically developed to assess processes and stages of change in overweight and obese people.

The Delphi method used here has a number of advantages over traditional content validity analysis. First, it enabled opinions to be obtained from a broad panel of experts from many countries and disciplines, many of whom were distinguished figures in the fields of obesity and the TTM. Furthermore, participants had the opportunity to assess the questionnaire and change their opinion across consecutive rounds, and this dynamic iterative process allowed us to refine the questionnaires in an iterative way. Another strong point of this study is the simultaneous validation of the questionnaires in both English and Spanish, with linguistically equivalent versions being obtained. Although it is not possible to establish the optimum and exact number of participants for a panel of experts (48) it has been shown that the larger the panel size the more accurate the group estimation (49). Consequently, we can state that the panel size of this study (n = 66) was highly adequate, both alone and in comparison with the panel size of other similar studies (35,37,50).

Another strength of the present study is the wide representation of experts from many countries and disciplines. Most of them were experts in obesity field with a extensive experience both in research field and clinical practice. The level of knowledge about the TTM was higher in the second and third rounds compared to the first. In consequence, participants who had more knowledge about the model seemed to be more likely to participate until the end of the study.

The consensus obtained in assessing the processes-of-change questionnaire was very high, as it was near 80% in each round for both representativeness and clarity. Furthermore, the consensus increased slightly across the study (from 72.2% for representativeness and 78.4% for clarity in the first round to 87.3 and 88.2%, respectively in the third one). Participants also reached a consensus regarding the stages of change questionnaire (agreement of 80.0 and 80.9% for representativeness and clarity, respectively). These results are similar to that found by Hagen et al. (35). These authors obtained a level of agreement with the adequacy of items definitions ranging from 87 to 88% and the clarity agreement scores ranged from 81 to 93%.

In sum, we have developed two questionnaires to assess the main constructs of the TTM based on a consensus of experts. Now that the content of these questionnaires has been shown to be appropriate to measure the processes and stages of change in overweight and obese people, future research will be directed toward their application. The questionnaires should be applied to both clinical samples (overweight and obese people) and normal weight people in order to analyze their practical functioning and psychometric properties.

The questionnaires developed by consensus in this study will allow us to analyze whether overweight and obese people use different processes of change across the stages. Their use in clinical practice will allow practitioners to tailor weight-management interventions according to the patient's stage of change and promote those processes of change that will facilitate movement toward the next stage.

Further research will be directed to the application of these questionnaires to overweight and obese people from clinical and community sample, as well as to normal weight sample. Its application will allow us to obtain its psychometric properties regarding discrimination indexes of items, reliability, and validity. Then, it will be possible to depurate the questionnaire again, based on its results on applied field. Then, a final version of the questionnaire will be obtained.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. SUPPLEMENTARY MATERIAL
  8. Acknowledgments
  9. Disclosure
  10. REFERENCES
  11. Supporting Information

We thank Juan José Fábregas and Joan M. León from the Centre de la Imatge i la Tecnologia Multimèdia, Fundació Universitat Politècnica de Catalunya for their work in preparing and maintaining the website. Also, we specially thank Claudio R. Nigg, both for its participation and suggestions made about the writing of this manuscript. Finally, we thank all the participants who helped during the study (alphabetically listed those who agreed to publish their name): R. Allyn, Philippines; I. Amigó, Spain; J. Aranda, Spain; G.L. Blackburn, US; V. Burke, Australia; R. Carraro, Spain; K. Cox, Australia; R. Figueredo, Paraguay; N. Finer, UK; K. Fontaine, US; E. García, Spain; A.R. Ghione, Argentina; J. Jorga, Serbia; N. Katsilambros, Greece; D.D. Micic, Serbia; A. Miján, Spain; J. Montero, Argentina; J.M. Peri, Spain; T. Rathmanner, Austria; D. Sánchez-Carracedo, Spain; S.G. Sapp, US; J.B. Schorling, US; S. Tonstad, Norway; H. Toplak, Austria; and P. Zelissen, The Netherlands. This study was partially supported by grants 2008FIC 00119 and 2005SGR00365 from the “Departament d'Universitats, Recerca i Societat de la Informació de la Generalitat de Catalunya,” and SEJ2005-09144-C02-02 from Spain's “Ministerio de Ciencia y Tecnología” under the European Regional Development Fund.

REFERENCES

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. SUPPLEMENTARY MATERIAL
  8. Acknowledgments
  9. Disclosure
  10. REFERENCES
  11. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. SUPPLEMENTARY MATERIAL
  8. Acknowledgments
  9. Disclosure
  10. REFERENCES
  11. Supporting Information

supporting Information

FilenameFormatSizeDescription
oby_1532_sm_1.doc341KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.