Is There Something Beyond Stages of Change in the Transtheoretical Model? The State of Art for Physical Activity Ahmed Jérôme Romain Centre de Recherche du Centre Hospitalier de lUniversité de Montréal, Montréal, Québec, Canada Johan Caudroit University of Lyon 1 Marie Hokayem Lebanese University Paquito Bernard Université du Québec a` Montréal and University Institute of Mental Health at Montréal, Montreal, Quebec, Canada Over the past 30 years, there has been a growing interest in the application of the transtheoretical model (TTM) in the domain of physical activity (PA). Even though this model has been widely used to implement PA interventions, most of these interventions have not used all of the TTMs theoretical constructs. Indeed, several studies focused exclusively on the stages of change although this construct is only descriptive. Thus, in the present review, we wanted to encourage researchers to go beyond stages of change when they use the TTM. To do so, we aimed to provide an overview of the TTM, its constructs and to present on one hand, longitudinal studies examining the association between PA and TTM constructs and, on the other hand, summarising the efficacy of TTM-based interventions as to present future TTM challenges. Public Significance Statement Theory-based interventions, including those based on the transtheoretical model, showed their efficacy in physical activity promotion. However, although the transtheoretical model proposes key regulatory components (namely, processes of change, self-efficacy, decisional balance, and temptation) to implement interventions, most of research remains focused on the stages of change. It should be reiterated that stages of change are a construct, not a theory, and therefore should not be used to tailor physical activity interventions. The key regulatory components of the transtheoretical model should be used to individualize counseling to physical activity. Moreover, the specific role of the processes of change in health education towards a more physically active lifestyle represents a future area of research. Keywords: physical activity, transtheoretical model, stages of change, processes of change, mediators of change Supplemental materials: http://dx.doi.org/10.1037/cbs0000093.supp As recently evidenced, the world actually faces an increasing prevalence of physical inactivity (Andersen, Mota, & Di Pietro, 2016), which partly explains the ever-rising worldwide prevalence of major noncommunicable diseases (Sallis et al., 2016). This physical inactivity epidemic constitutes an economic burden to the international health care systems of up to US$53.8 billion in 2013 (Ding et al., 2016). Thus, there is a pressing urgency to promote physical activity (PA) by implementing interventions that take into account the reasons favouring its adoption. In this context, theory-based interventions constitute an interesting option, not only because they are assumed to be better than nontheoretical interventions but also because of the fact that they provide a framework that makes interventions easier to replicate and disseminate in real-life settings. A recent metaanalysis of randomized controlled trials (RCTs) concluded that theory-based interventions in PA promotion effectively increase PA and that none of the psychological theories included were found to be superior in PA promotion (Gourlan et al., 2016). This meta-analysis of 31 interventions (over the 82 included) using the transtheoretical model (TTM) also highlighted that this model is among the most used theories to promote PA Ahmed Jérôme Romain, Département de neurosciences, Centre de Recherche du Centre Hospitalier de lUniversité de Montréal, Montréal, Québec, Canada; Johan Caudroit, Laboratoire sur la Vulnérabilité et Linnovation dans le Sport, Département des sciences de lactivité physique, University of Lyon 1; Marie Hokayem, Department of Nutrition, Faculty of Public Health, Lebanese University; Paquito Bernard, Département des sciences de lactivité physique, Université du Québec a` Montréal, and University Institute of Mental Health at Montréal, Montreal, Quebec, Canada. Correspondence concerning this article should be addressed to Ahmed Jérôme Romain, Département de neurosciences, Centre de Recherche du Centre Hospitalier de lUniversité de Montréal, 900 rue St-Denis,
(Prochaska & DiClemente, 1983). The TTM conceptualizes the process of intentional behaviour change by assuming that (a) a single theory cannot account for the complexity of human behaviour change, (b) behaviour change is a process that unfolds over time through several stages, (c) these stages are stable and open to change, and (d) specific processes should be used at specific stages to facilitate the efficacy of behaviour change (Prochaska, Redding, & Evers, 2008). It is interesting that the TTM speculates a nonlinear transition between the stages of change (SOC) with patterns of discontinuity (Lippke & Plotnikoff, 2006). Even though the TTM is widely used, it remains poorly implemented in interventions seeking to improve PA behaviour (Romain, Bortolon, et al., 2016), and this can be explained by the fact that many researchers are probably unaware of all the TTMs constructs. In view of these observations, it was therefore necessary to explain how to use the TTM and its underlying mediators in research and intervention contexts. The objectives of the present narrative review were to (a) briefly provide an overview of the TTM and the mediators of change on which it is based, (b) discuss studies examining longitudinal associations between PA change and TTM mediators, (c) outline TTM-validated questionnaires, (d) examine the efficacy of TTM-based interventions for PA promotion, and (e) debate future TTM challenges. The Transtheoretical Model and Its Mediators The TTM identifies change as a progressive process through a series of five different SOC over time (Prochaska & DiClemente, 1983). Although the SOC are the most popular part of the TTM, notably because of their ease of use and scoring, they also constitute its most descriptive construct. The five SOC are precontemplation (not ready; not intending to change in the next six months), contemplation (getting ready; intention to change within the next six months), preparation (ready; intention to change within 30 days), action (new behaviour is initiated within the last six months), and maintenance (behaviour is sustained for more than six months). Thus, while going through the SOC, an individual starts by intending to adopt the behaviour criteria in the early preaction SOC (precontemplation, contemplation, and preparation) to later adopt and maintain this newly acquired behaviour throughout the action and maintenance stages (Prochaska & Velicer, 1997). If SOC represent the most descriptive part of the TTM, it is mainly because they explain where people are in terms of motivation but not how to motivate them or why they move across stages. Indeed, according to the TTM, the transition between the different SOC is influenced by its mediators of change (its theoretical constructs) that include decisional balance, temptation, self-efficacy, and processes of change (POC; Prochaska, DiClemente, & Norcross, 1992). Decisional balance is defined as the perception of advantages (pros) and/or disadvantages (cons) related to the decision of undertaking or not a behaviour (Prochaska et al., 1994). Temptation is the urge to engage in a specific behaviour in the midst of difficult barriers (Hausenblas et al., 2001). Self-efficacy, a component of social cognitive theory (Bandura, 1977), is defined as a persons judgment of his or her capabilities to organise and execute courses of action required to attain designated types of performance (Bandura, 1997). Finally, there are the POC that help clarify how behaviour changes take place, and SOC help pinpoint when those modifications occur. POC are comprised of a total of five experiential processes and five behavioural processes that need to be executed to ensure a certain progress through the SOC and achieve the desired behaviour change. Experiential processes are defined as processes in which individuals obtain information based on their own experiences, and behavioural processes regroup strategies used to modify the environment to help change the behaviour (Burkholder & Nigg, 2001; Romain, Chevance, Caudroit, & Bernard, 2016; see Table 1 for a definition of POC). In the TTM, the relationship between its mediators and the SOC has been tested extensively (Burkholder & Nigg, 2001; Marshall & Biddle, 2001), and was found to be consistent throughout different types of behaviour (e.g., smoking, diet). However, contrary to the assumptions formulated in tobacco cessation, the POC by SOC sequence was found to be different. Indeed, in smoking cessation, experiential and behavioural POC act sequentially, with experiential POC used in the early stages and behavioural POC in the later stages (action and maintenance). Inversely, in PA, this sequential order was not found with experiential and behavioural POC acting in tandem, with the use of both increasing across stages (Marshall & Biddle, 2001; Table 1 Processes of Change and Their Definitions Processes of change Definition Experiential processes of change Consciousness raising Efforts to better understand the problematic behavior Dramatic relief Affective aspects of behavior change Self-reevaluation Cognitive or emotional appraisal of the impact of the behavior on the individual Environmental reevaluation Impact of negative or positive behavior on individuals social and physical environment Social liberation Recognition that actual social norms encourage individuals to reach/sustain their healthier lifestyle Behavioral processes of change Self liberation Committing to change and believing in this commitment Helping relationships Using the support of caring others to modify behaviour Counterconditioning Substituting unhealthy for healthy behaviour Reinforcement management Use of reinforcement and reward to support/sustain healthy behaviour Stimulus control Modifying the environment to encourage healthy behaviour A
osen, 2000; see Figure 1 for an illustration). This crucial point will be discussed further subsequently. In the TTM, mediators explain why people modify their behaviour. In order to better understand how changes occur, it is essential to focus on longitudinal, interventional, or observational studies (Rhodes & Quinlan, 2015) rather than cross-sectional research designs. What Do Longitudinal Observational Studies Using TTM Mediators Tell Us About the Transition Between SOC of Physical Activity? Observational studies provide a primary insight to understand the complex associations between SOC and the mediators of the TTM. For this purpose, in this section, only observational studies having investigated the role of these mediators in the transition between SOC were included. Plotnikoff, Hotz, Birkett, and Courneya (2001) assessed whether self-efficacy, decisional balance, and POC predicted the transition between exercise SOC within a 12-month period among 1,602 adults. Results showed that self-efficacy, decisional balance, and both experiential and behavioural POC were predictors of the transition between SOC. To be more precise, the transition out of the precontemplation and contemplation stages was predicted by higher levels of self-efficacy, perception of advantages (pros), and behavioural POC. Also, the transition out of the preparation stage was predicted by higher levels of self-efficacy and pros. Moreover, retention in postaction stages was predicted by higher levels of pros versus lower cons, and by the activation of both experiential and behavioural POC. Thus, Plotnikoff, Hotz, et al. (2001) study partially supports the validation of TTM in exercise. A similar study testing the TTMs capacity for predicting PA transitions was performed among 1,674 adults with Type 1 or Type 2 diabetes over six months (Plotnikoff, Lippke, Johnson, & Courneya, 2010). Findings provided moderate support for the TTM constructs in predicting PA stage transitions, with very few differences between Type 1 and 2 diabetic groups. Indeed, the transition from precontemplation to contemplation was predicted by the pros and the experiential POC. The transition out of preparation was only predicted by higher self-efficacy. Transition out of the action stage was predicted by the pros and the behavioural POC, whereas remaining in the maintenance stage was predicted by higher levels of self-efficacy, pros, and experiential and behavioural POC. Analogous results were found in a prospective investigation in which TTM showed significant potential for motivating women with multiple sclerosis to increase their PA over a period of 12 months (Levy, Li, Cardinal, & Maddalozzo, 2009). In addition, Dishman, Vandenberg, Motl, and Nigg (2010) assessed TTM constructs relating to the 2010 guidelines for regular moderate or vigorous PA, at 6-month intervals three or more times over 24 months, among a cohort of 497 multiethnic participants. The results provided great support for core TTM constructs by showing that people meeting, or partially meeting, PA guidelines had a decrease in temptation, an increase in self-efficacy, and also a higher use of both experiential and behavioural POC. Only decisional balance was not associated with PA guidelines. Nevertheless, the absence of results regarding decisional balance is not supported by one of the first longitudinal studies on TTM showing that pros, cons, and self-efficacy, but not POC, were associated with leisure exercise three years after initial assessment in adolescents (Nigg, 2001). The aforementioned studies provide important information to consider, including the fact that all TTM constructs were predictors of the transition between the different SOC, but to different extents. Thus, regarding PA, to progress through SOC, people need to find more reasons to exercise (the pros) than not to (the cons), and to feel more confident (self-efficacy) by increasing the use of both experiential and behavioural strategies (POC). These arguments are corroborated by findings from a previous meta-analysis of cross-sectional studies on TTM applications to PA (Marshall & Biddle, 2001). The precited research supports the use of TTM interventions in the context of PA by demonstrating that all TTM constructs are necessary in order to adopt or sustain a physically active lifestyle. However, it should be noted that these observations were drawn from observational studies, so to confirm them it is necessary to analyse results from interventional studies. 1 2 3 4 5 PC C P A M Behavioral Experiential 1 2 3 4 5 PC C P A M Behavioral Experiential Figure 1. Schematic representation of the relationship between processes and stages of change in tobacco use (left figure; sequential association) and physical activity (right figure; tandem association). This figure has been adapted from Romain, Chevance et al. (2016) with their authorization. PC precontemplation; C contemplation; P preparation; A action; M maintenance. See the online article for the color version of this figure. 44 R
that after 24 months, behavioural POC were the only mediators of the relationship between the TTM intervention and PA/cardiorespiratory fitness relationship in sedentary adults. Thus, these two studies support the use of TTM, and more particularly behavioural POC in PA behaviour modulation. Consequently, although the TTM provides information about its mediators, few studies have really addressed this relevant issue (Rhodes & Pfaeffli, 2010). Although some research failed to show any mediation effects, most studies showed that TTM mediators significantly modified PA level even though the sample size was too small to provide any robust conclusion (Fahrenwald, Atwood, Walker, Johnson, & Berg, 2004; Rabin, Pinto, & Frierson, 2006). Among the most prominent TTM mediators, self-efficacy and behavioural POC were found to be of great importance in PA interventions, even though further explanations are necessary to understand the extent of their impact. The consistent association between these variables/mediators may be explained by the fact that they are often well correlated. Nevertheless, other assumption can be drawn from their significant relation. Loprinzi and Cardinal (2013) performed a study on the supposition that behavioural POC and self-efficacy are important in PA behaviour change and that the literature does not really provide any clear explanation. Thus, among breast cancer patients, they examined the mediation link between PA, behavioural POC, and self-efficacy, and highlighted that behavioural POC were related to PA and that this relationship was mediated by self-efficacy. This result was also confirmed by the Training Interventions and Genetics of Exercise Response study, in which self-efficacy and experiential and behavioural POC were correlated with PA at the baseline period of their trial, with only behavioural POC mediating the relationship between selfefficacy and adherence to exercise (defined as the number of exercise sessions attended compared with the possible number of exercise sessions offered; Dishman, Jackson, & Bray, 2014). Consequently, even though these findings further solidify our understanding of the TTM when relating to PA change, they do not negate the role of experiential POC that can trigger the intention to exercise in different populations (Nigg, 2005). Initial TTM Instruments Development for Physical Activity Behaviour Change In the TTM, one of the undeniable limitations is that most assessment tools are presented in English, which restricts their use to English-speaking countries; in addition, not all studies used validated questionnaires in their surveys. Thus, to overcome this issue, in the following paragraphs, we present a systematic overview of the different worldwide validations that exist. TTM research in the context of PA was initiated by Marcus, Rakowski, and Rossi (1992), who published three validation studies for assessing the four key TTM constructs with cross-sectional designs across work-site samples in Rhode Island. Except for the Temptation scale, these questionnaires have been extensively used, examined, and adapted. As recommended by Reed, Velicer, Prochaska, Rossi, and Marcus (1997), the SOC measure was developed as an algorithm to categorise individuals in one of five SOC. This scale consists of one item with five statements representing each a stage, going from the precontemplation to the maintenance stage. Thus, a reliable SOC algorithm should include a clear definition of PA and its frequency and duration. PA defined as a 30 min session at least four times per week is generally recommended (Nigg et al., 2005; Romain et al., 2012), and the validity of this SOC algorithm has been shown with self-reported PA and anthropometrical measures in adults (Hellsten et al., 2008; Nigg et al., 2005). Regarding POC, Marcus et al. (1992) adapted the initial scale developed for smoking cessation by Prochaska, Velicer, DiClemente, and Fava (1988) for the context of PA. The scale contained 39 items measuring both experiential and behavioural POC ( .62.88). Later, Nigg, Norman, Rossi, and Benisovich (1999) created a new and shorter measure of POC, which contains 30 items measuring the 10 POC for PA ( .62.85). Regarding self-efficacy, Marcus et al. (1992) validated a fiveitem-measure scale assessing self-efficacy for PA ( .82). In addition, Benisovich, Rossi, Norman, and Nigg (1998) developed the multidimensional self-efficacy questionnaire, which comprises 18 items measuring the individuals confidence in his ability to overcome PA-related barriers (e.g., excuse making, bad weather; .77.85). Finally, Marcus et al. (1992) validated a 16-item Decisional Balance scale for PA, with 10 items for the perceived benefits of PA (pros; .95) and six items for the perceived costs (cons; .79). Plotnikoff, Blanchard, Hotz, and Rhodes (2001) updated this scale by using 10 items (five pros, .79; five cons, .71) for PA. The temptation measure was validated by Hausenblas et al. (2001). In their initial development and validation, two factors were reported: affect (five items; .81) and competing demands (five items; .86). Another seven-item version showed a similar structure (Geller, Nigg, Motl, Horwath, & Dishman, 2012). Regarding the validation of TTM scales in PA, several studies have investigated the validity, adaptation, translation, and application of TTM constructs in different populations and languages (see Table 2 for summary and Supplementary File 1 of the online supplemental materials for the complete table). TTM Questionnaires Available in 11 Different Languages Among studies presented in the Table 2, several researchers have used the original TTM questionnaires validated in English (Blaney et al., 2012; Carnegie et al., 2002; Dishman, Jackson, et al., 2010; Geller et al., 2012; Kearney, de Graaf, Damkjaer, & Engstrom, 1999; Maddison & Prapavessis, 2006; Norman, Velicer, Fava, & Prochaska, 1998; Pickering & Plotnikoff, 2009; Rhodes, Berry, Naylor, & Wharf Higgins, 2004; Sallis, Pinski, Grossman, Patterson, & Nader, 1988; Skaal, 2013; Skaal & Pengpid, 2012; Vita & Owen, 1995). TTM scales were then translated into 11 different languages (see Table 2). Psychometric studies have validated TTM constructs from English to French (Bernard et al., 2014; Eeckhout, Francaux, Heeren, & Philippot, 2013; Eeckhout, Francaux, & Philippot, 2012a, 2012b; Romain, Bernard, Hokayem, Gernigon, & Avignon, 2016), Finnish (Cardinal, Tuominen, & Rintala, 2003, p. 200), Dutch (Ronda, Van Assema, & Brug, 2001), German (Bucksch, Finne, & Kolip, 2008; Fuchs & Schwarzer, 1994; Kanning, 2010; Tergerson & King, 2002), Greek (Bebetsos & Papaioannou, 2009; Korologou, Barkoukis, Lazuras, & Tsorbatzoudis, 2015), Persian (Farmanbar, Niknami, Lubans, & Hidarnia, 2013; Sanaeinasab, Saffari, Nazeri, Karimi Zarchi, & Cardinal, 2013), Korean (Y. Kim, Cardinal, & Lee, 2006; Y.-H. Kim, 2007), Chinese (Si et al., 2011; 46 R
Tung, Gillett, & Pattillo, 2005; Yang & Chen, 2005), Malaysian (Phing, 2014), Japanese (Horiuchi, Tsuda, Kobayashi, Fallon, & Sakano, 2017; Oka, 2000, 2003), Taiwanese (Sechrist, Walker, & Pender, 1987), and Spanish (Gonzalez & Jirovec, 2001). No psychometric investigation has, to our knowledge, interpreted or adapted the temptation scale in other languages. Investigating Invariance of TTM Questionnaires The different types of invariance (configural, metric, and scalar) of TTM questionnaires (see Table 2) have been investigated across various time sets and subgroup characteristics, with results showing that TTM constructs were invariant according to sex, student status, ethnicity, age, body mass index, employment, PA level, protocol adherence, level of education, and diabetes type (Bernard et al., 2014; Dishman, Jackson, & Bray, 2010; Geller et al., 2012; Paxton et al., 2008; Pickering & Plotnikoff, 2009). These analyses were performed with English and French versions of TTM questionnaires (Bernard et al., 2014; Geller et al., 2012). Moreover, the longitudinal invariance of TTM constructs has also been provided across 3- and 6-month periods, with studies showing that any temporal differences or modifications identified can be interpreted as changes related to time or intervention mistakes, but not measurement errors (Dishman, Jackson, et al., 2010; Geller et al., 2012). Are TTM-Based Interventions Effective in Promoting PA? Over the last decade, interventional researchers in health psychology and behavioural medicine have gradually integrated the specific methodological requirements of evidence-based medicine (Keefe & Blumenthal, 2004). In this methodological paradigm, the RCT design is recognised as the highest level of investigative methodology to establish the efficacy or effectiveness of health behaviour change interventions (Davidson et al., 2003). In this context, several critics have questioned the worth of TTM interventions in promoting PA, arguing that SOC may not be applied to PA change because of the complexity of this behaviour, the lack of validated staging algorithms, and the possibility that the most reliable determinants of PA change are not included in the TTM (Adams & White, 2005; Armitage, 2009; Brug et al., 2005). However, two systematic reviews including only RCTs examined the efficacy of TTM interventions on PA promotion, with findings indicating that TTM-based interventions induce a small to medium effect size for PA behaviour change. The most recent review (Romain, Bortolon, et al., 2016) included 33 RCTs, with 4,950 and 5,400 participants in the interventional and control groups, respectively. Fourteen studies included exclusively adults with chronic illness (e.g., multiple sclerosis). The length of intervention ranged from 2 to 100 weeks, and PA level was an inclusion criterion but stage progression was not. In addition, all constructs related to PA were self-reported. This review obtained an overall effect size of d 0.33 (95 % confidence interval [CI] [0.22, 0.43]) for PA behaviour change, which was consistent with Gourlan et al. (2016) (d 0.31, 95 % CI [0.20, 0.42]). These effect sizes need to be interpreted in the context of public health (Prentice & Miller, 1992), seeing that even a slight PA increase may lead to a major health impact (Khan et al., 2012). Evidence-Based Rather Than Evidence-Inspired TTM Interventions to Change Physical Activity Behaviour In line with previous recommendations (Michie & Johnston, 2012), a thorough analysis of theoretical moderators of TTMbased interventions has been performed in the present narrative review. Romain, Bortolon et al. (2016) observed that TTM-based interventions implementing at least three constructs (e.g., selfefficacy, decisional balance, POC) obtained a 3-times larger effect size (d 0.49, 95 % CI [0.29, 0.69]) versus applying two constructs or less (d 0.16, 95 % CI [0.06, 0.25]) regarding PA promotion. Moreover, bivariate metaregressions showed that selfefficacy and POC were the most active and effective components T
Theories play a critical role in health-promotion program planning, implementation, and evaluation. They also help answer important questions such as why people dont engage in healthy behaviors. However, before theories officially became theories, they were tested through research for their ability to predict and explain health behavior. They were also examined through research for their efficacy in health promotion program development. It is through research that we examine current theories, create new theories, and collect evidence concerning the effectiveness of theory-based public health and health promotion programs. Therefore, understanding how to analyze and interpret research is important to our professional practice. This Assignment will also help you analyze research in other courses in your program.
This week, you examine research that applied a theory or model of health behavior at the intrapersonal level.
Review Chapters 3, 4, and 6 in your textbook.
Consider the constructs of the different theories and models and how they relate to the intrapersonal level of influence on health behavior.
Review your assigned article.
Review the Research Article Review worksheet located in the Learning Resources for the specific article you were assigned.
Familiarize yourself with the questions on the worksheet prior to reading the article.
Complete the assigned Research Article Review Worksheet.
Use critical thinking to analyze, assess, and evaluate the article and to respond to the questions in the worksheet. Be specific and use examples in your answers.
HLTH 2500: Theories of Health Behavior
Research Article Review (RAR) Worksheet
Week 2: Transtheoretical Model
Name: [Type your name here.]
Romain, A. J., Caudroit, J., Hokayem, M., & Bernard, P. (2018). Is there something beyond
stages of change in the transtheoretical model? The state of art for physical activity.
Canadian Journal of Behavioural Science, 50(1), 42 53. http://dx.doi.org/10.1037/
1. Read the research article, which you will access this article from the Walden Library databases.
2. Type your answers under each question. The space between questions will expand as you write your response.
3. Be sure to type your full name above.
4. Save the worksheet as: WK2_RAR_TTM [your last name_first name initial]
(for example: WK2_RAR_TTM_Gallien_T)
1. What is the purpose of the article?
2. Respond to the following three questions:
a. Which theoretical constructs of TTM are discussed in the article?
b. Which constructs are considered the mediators of change?
c. Explain what is meant by mediators of change.
3. Choose only one (1) of the following questions to answer.
a. What do studies using TTM mediators tell us about the transition between stages of change (SOC) of physical activity?
b. What did the authors conclude after reviewing the research about the effectiveness of TTM-based interventions on physical activity level?
4. Even though the Transtheoretical Model (TTM) is widely used in research and in the development of interventions, the authors of the article noted several weaknesses of TTMs implementation. What are some of the weaknesses identified by the authors? Be specific.
5. What suggestions do the authors give for creating a physical-activity-specific Transtheoretical Model?
In addition, answer the following:
1. What was the most challenging aspect of this assignment?
2. Describe a strategy you might use to overcome this challenge.
- Assignment status: Already Solved By Our Experts
- (USA, AUS, UK & CA PhD. Writers)
- CLICK HERE TO GET A PROFESSIONAL WRITER TO WORK ON THIS PAPER AND OTHER SIMILAR PAPERS, GET A NON PLAGIARIZED PAPER FROM OUR EXPERTS