Postmenopausal women's experiences of weight maintenance following a very low energy diet

Abstract Introduction Very low energy diets (VLEDs) effectively induce substantial weight loss in people with obesity, yet they are rarely used as a first line treatment. There is a belief that such diets do not teach the lifestyle behavior changes needed for long‐term weight maintenance. However, little is known about the lived experiences of people who have lost weight on a VLED in the long term. Methods This study aimed to explore the behaviors and experiences of postmenopausal women who had followed a 4‐month VLED (using total meal replacement products [MRPs]), followed by a food‐based, moderately energy‐restricted diet for an additional 8 months, as part of the TEMPO Diet Trial. Qualitative in‐depth semi‐structured interviews were conducted with 15 participants at 12 or 24 months (i.e., at 8 or 20 months post diet completion). Transcribed interviews were analyzed thematically using an inductive approach. Results Undertaking a VLED was reported by participants to confer advantages in weight maintenance that previous weight loss attempts had not been able to do for them. Firstly, the rapid and significant weight loss, in conjunction with ease of use, was motivational and helped instill confidence in the participants. Secondly, the cessation of a normal diet during the VLED was reported by participants to break weight gain‐inducing habits, allowing them to abandon unhelpful habits and to introduce in their place more appropriate attitudes toward weight maintenance. Lastly, the new identity, helpful habits and increased self‐efficacy around weight loss supported participants during weight maintenance. Additionally, participants reported that ongoing occasional use of MRPs provided a useful and easy new strategy for countering weight regain and supporting their weight maintenance regimen. Conclusion Among the participants in this qualitative study, most of whom had maintained a loss of over 10% of their baseline body weight at the time of interview, using a VLED in the context of a clinical weight loss trial conferred confidence, motivation and skills for weight maintenance. These findings indicate that VLEDs with clinical support could be successfully leveraged to set up behaviors that will support weight maintenance in the long term.


| INTRODUCTION
To treat obesity, meta-analyses indicate that the most effective longterm non-surgical, non-pharmacological individual-level weight loss treatments are very low energy diets (VLEDs). [1][2][3] Internationally, VLEDs are defined by the Food Standards of Australia and New Zealand, the World Health Organization, the international CODEX standards of the WHO and Food and Agriculture Organization, as well as, the European Union as formula-based foods used in weight control diets that provide between 2090 to 3350 kJ (500-800 kcal) daily. [4][5][6][7] This involves replacing all meals and snacks with prepackaged, nutritionally complete shakes or soups (to which water or milk is added), bars or other pre-packaged meal replacement products (MRPs). Often, an allowance of low starch vegetables, black tea or coffee and no sugar products (e.g., diet drinks, diet jelly) is also used in this regimen. 8,9 The terms very low-calorie diet (VLCD) and very low energy total diet replacement are sometimes used interchangeably with the term VLED. For clarity, we have used the term VLED throughout this paper, to comply with the most common international nomenclature.
VLEDs are safe in the short term and have been used for up to 5 months with no significant negative effects. 10 They effectively induce fast weight loss in people with obesity. 11 Furthermore, weight regain after these diets is no faster than after any other diet, [12][13][14] and the greater initial weight loss after VLEDs compared to other diets has been shown to be predictive of long-term maintenance of a lower body weight. 3,12,15 Additionally, those who remain engaged with support after a VLED can retain clinically significant weight loss, of greater than or equal to 5% of initial body weight, in the long term (3 years). 16 A meta-analysis of long-term weight maintenance, after either a VLED or a conventional food-based diet, showed weight loss at 4-5 years after commencing the intervention of 6.6% of initial body weight versus 2.1%, respectively, demonstrating that greater long-term weight loss can be achieved with a VLED. 3 In a recent randomized controlled trial, participants who underwent a 16-week VLED, compared to participants who underwent a conventional food-based diet, were 2.6 times more likely (42% vs. 16%) to have lost 10% or more of their initial body weight at 3 years. 17 Given that just 5% weight loss is known to confer metabolic advantages, VLEDs can also ameliorate obesity-related chronic disease.
Despite the greater weight reductions and longer-term maintenance of a reduced body weight, VLEDs are not routinely recommended as a first line weight loss method. Practice guidelines worldwide recommend gradual weight loss for overweight and obesity. [18][19][20] The reasons for these recommendations are based on the assertion that gradual weight reduction in small increments is a more realistic goal, and making small changes slowly, confers betterlasting weight maintenance results. [18][19][20] Furthermore, healthcare professionals are reported to be reluctant to use VLEDs as a treatment for obesity. [21][22][23][24] While quantitative evidence shows better weight loss outcomes in the short-and long-term, concerns from healthcare professionals about using VLEDs are often qualitative in nature. These concerns include beliefs that VLEDs may be difficult to adhere to, unpleasant to undertake, have low long-term efficacy, and could promote weight cycling with concomitant potential negative psychological impact. [21][22][23][24] To address these qualitative concerns and inform healthcare professionals on how VLEDs are experienced by users, we undertook a systematic review of qualitative research on experiences during and after undertaking a VLED using total meal replacement. [25][26][27][28] Across all three studies available for inclusion in that review, VLED was generally reported by participants to be a positive experience, easy to follow, with decreased requirements to make decisions about food, and rapid weight loss increasing motivation for, and adherence to, the intervention. 28 Another study on using a LED with total diet replacement for weight loss, also showed that total diet replacement is well tolerated by participants. 29 Whilst these studies highlighted important aspects of using VLEDs for weight loss and reported that VLEDs and total diet replacement low energy diets are well tolerated in the specific populations studied, the significance of a VLED to the weight maintenance period was poorly described. [25][26][27][28] Therefore, this study was conducted to explore the experiences and behaviors of women who have undertaken a VLED intervention, with a specific focus on the weight maintenance phase. Acknowledgment by health researchers that the perspectives and experiences of participants and consumers have been neglected and are an important aspect of intervention acceptability, has progressed qualitative research as an important adjunct to quantitative findings of clinical research. [30][31][32] Accordingly, this research was a sub-study of the TEMPO Diet Trial (Type of Energy Manipulation for Promoting optimum metabolic health and body composition in Obesity), a randomized controlled trial comparing the long-term effects of severe energy restriction versus moderate energy restriction in postmenopausal women. 9 The primary aim of the TEMPO Diet Trial was 306to make a head-to-head comparison of the long-term (3-year) effects of fast versus slow weight loss on body composition, including bone mineral density (an important predictor of osteoporotic fractures).
Postmenopausal women were studied because the estimated lifetime risk of osteoporotic fractures is 2-fold higher in women than in men (40% vs. 13%). 33 The rationale and protocol of the TEMPO Diet Trial can be found in a previous publication. 9 2 | METHODS

| Participants
This qualitative/exploratory study included 15 participants who were randomized to the severely energy-restricted diet arm of the TEMPO Diet Trial. Participants had a mean (SD) age of 58.1 (3.8) years and mean (SD) body mass index of 34.0 (2.7) kg/m 2 at baseline (i.e., at week 0, prior to commencement of the severely energy-restricted diet), were at least 5 years postmenopausal at the time of recruitment, and sedentary (defined here as undertaking less than 3 h of structured physical activity per week). The full inclusion and exclusion criteria and the rationale for these have been detailed in our published protocol. 9

| Severely energy-restricted diet and weight maintenance diet
Details of the study protocol and 12-month intervention for the severely energy-restricted arm of the TEMPO Diet Trial have been reported in detail previously. 9 Briefly, participants were randomized to four months (16 weeks) of severe energy restriction of 65%-75% relative to estimated energy expenditure, using a total meal replacement diet, followed by moderate energy restriction of 35%-25% for an additional 8 months (36 weeks, to a total of 12 months [52 weeks]), achieved with a food-based diet based on the Australian Guide to Healthy Eating. 9,34 The severe energy restriction prescription for participants in this study consisted of three meal replacement shakes, two cups of nonstarchy vegetables, and one teaspoon of oil daily for 16 weeks. All MRPs (Kicstart TM shakes and soups) were sourced from Prima Health Solutions Pty Ltd., Brookvale, NSW, Australia and were supplied to participants at no cost to them. The participants had regular one-onone contact with a dietitian (approximately once every 2 weeksboth face to face and over the phone for the first 20 weeks), which decreased in intensity thereafter to approximately once monthly until 12 months, and then to once yearly after the 12-month time point. Participants had the option of attending monthly face-to-face group support meetings after they had completed 12 months on the trial. Participants were then followed-up at 24 and 36 months from the date they started the diet.

| Recruitment
Women were eligible to participate in this sub-study if they had

| Study design
This qualitative exploratory study was based on descriptive phenomenology, 35,36 as we explored both the common and uncommon experiences of postmenopausal women who had lost weight via a VLED and were attempting to maintain a reduced body weight. We also used research methods from grounded theory for interview questioning and to provide structure to the analytical process. 36,37

| Data collection
Semi-structured interviews were held at the Charles Perkins Centre Royal Prince Alfred Clinic on the University of Sydney campus in Camperdown, New South Wales, Australia, the same location where quantitative data were collected for the TEMPO Diet Trial, as this was convenient for both participants and researchers. Interviews were 1-1.5 h in duration, and were audio recorded. An interview discussion guide (Appendix A) was developed through discussions between researchers that included the TEMPO Diet Trial dietitians, chief investigators, and co-authors with qualitative research HARPER ET AL.
-307 experience, but who had no affiliation with the TEMPO Diet Trial.
The interviews were conducted by a Masters student and a PhD student (and accredited practicing dietitian) who administered the weight loss intervention for 3 of the 15 participants (20%) in this substudy. All participants were encouraged to talk about their experiences, thoughts, beliefs and attitudes regarding their weight loss, undertaking a VLED intervention, and their weight maintenance efforts and experiences in the months after losing weight on a VLED.
The interviews evolved over time. Concepts brought up by participants during earlier interviews were tested and expanded in later interviews. Data collection ceased when data saturation was reached (i.e., when no new major concepts around the topic were being discussed in the interviews). 38

| Data analysis
The interviews were transcribed verbatim and analyzed in-depth

| Sample overview
Most of the 15 women participating in this sub-study were in fulltime or part-time employment, with only two of them being retired, albeit maintaining active volunteering roles. They all described a history of numerous previous weight loss attempts with gradual weight regain over their lifetimes. All participants could pinpoint at least one prior successful weight loss effort, either with a F I G U R E 1 Recruitment of participants for this sub-study of the TEMPO Diet Trial, and flow of participants through the sub-study. commercial weight loss program or via individual efforts. The most common commercial weight loss program previously used was Weight Watchers. The primary reasons stated for joining the current trial were health and weight loss. All 15 of the participants stated that joining a clinical trial in a university environment gave them a sense of safety and legitimacy about the dietary intervention, which some viewed as "drastic". Participants also reported that being part of a clinical trial also increased their compliance and sense of responsibility to adhere to the intervention. None of the participants had attempted a total diet replacement before, however, all of them had either heard of meal replacements or had used them in an ad hoc manner prior to volunteering to be a participant in the TEMPO Diet Trial.  (Table 1). Detailed weight loss results, as well as overall dropout rates at 12 months (4 out of 50 participants) for the TEMPO Diet

| Weight loss results
Trial are available elsewhere.
The mean (SD) weight loss at 16 weeks of the participants in this sub-study compared with those who did not respond or who declined to participate (n = 8) was 18.2 (2.5) kg versus 15.5 (4.9) kg, respectively ( Table 1). The weight gain at the time of interviews at 12 months in our participants versus those who did not respond or who declined to participate was 2.6 (6.5) kg versus 4.1 (5.4) kg, respectively. For ease of reading, individual weight losses are shown after the participant's identity code as a percentage of initial body weight at 4 months (end of the phase of the VLED intervention where total meal replacement diets were used), and at the time of the interview. For example, P01 (20.8% at 4 months, 21.6% at 12 months) shows that participant P01 had lost 20.8% of her initial (baseline or pre-diet) body weight at the end of the 4-month VLED, and 21.6% at 12 months, which was the time of her interview.

| Overview of results-Weight maintenance
This study revealed three major themes of how the VLED intervention-as experienced by the participants-influenced their subsequent approach to and thoughts about weight maintenance.
These three themes are summarized in Figure 3. Two of these three major themes relate specifically to aspects of undertaking a VLED using meal replacements. These two themes are: Rapid, significant weight loss and ease of use of the VLED; and Cessation of a normal diet. These two aspects of the VLED were considered by participants to lead to the third of the three major themes, which we have termed "Weight maintenance". This theme of Weight maintenance comprised three sub-themes of experiences during the weight maintenance phase of the intervention. These three sub-themes were: New identity; New habits and increased self-efficacy; and Learned strategies to combat weight regain. In brief, the rapid, significant weight loss and ease of use of the VLED, in conjunction with the cessation of normal diet that was otherwise usually consumed by the participants, led to the perception of gaining a new identity, new habits and increased self-efficacy, as well as having learned a strategy to combat weight regain during the weight maintenance phase of the intervention. These concepts are described below, in conjunction with participant quotes to give insight into the participants' viewpoints and feelings. Examples of further quotes from participants can be found in Appendix C.

F I G U R E 2
Weight change at 16 weeks and at time of interview (i.e., 12 or 24 months). Overall, the participants reported that the VLED intervention had unique aspects which they had not experienced during other attempts at weight loss. These factors were highly rewarding and conferred a high level of motivation and self-efficacy which persisted during weight maintenance. Recognizing bad habits during the VLED and changing unhelpful habits after the cessation of the VLED, combined with continued use of MRPs, helped the participants to feel empowered to maintain their weight loss for the long term.

| DISCUSSION
Our research has extended previous findings and demonstrated that the rapid weight loss and the stark contrast of a VLED based on meal replacement compared with usual eating patterns provided motivation and new approaches to food which served to support participants in the weight maintenance phase. Further, the amount of weight lost by participants during the VLED allowed them to reengage with a positive perception of themselves, as signaled through new clothes and self-image.
According to our interviewees, the rapid and significant weight loss they experienced conferred considerable motivation to not only adhere to the total meal replacement VLED, but also to continue to maintain that weight loss in the months after the VLED phase of the intervention had ended. This stands in contrast to worldwide Practice Guidelines for the management of obesity, which consistently recommend slow weight loss and small, achievable goals. [18][19][20] For example, in Australia, the National Health and Medical Research Council (NHMRC) recommends that "a focus on sustainable (rather than restrictive) changes in dietary behavior may support motivation" and that a weight loss of 5% is a realistic and achievable goal. 19 Setting "realistic" weight loss goals (less than 10% of initial body weight) prior to weight loss has been a cornerstone of weight loss interventions, however there is no evidence to support setting minimal weight loss goals and, setting the goal higher has been shown to be successful. 12,14 The commonly-held assumption and oft cited recommendations 18-20 that making small, slow changes over time is a better way to create more permanent lifestyle changes has not been supported by the narrative of our participants' experiences.
Buying a new wardrobe or being able to fit back into favorite items of clothing was described as a real and significant turning point by our participants, as well as an ongoing source of motivation for maintenance of reduced body weight. Shopping for clothing as a larger person can be confronting, 39 and keeping favorite clothing has also been shown to be a motivator to commence weight loss efforts in women. In our study this was also a motivator for maintaining weight lost. The participants expressed positive self-esteem and Through choosing our daily wardrobe, we control the presentation of our identity, selfhood and well-being. While identity shift has previously been identified as an enabler of weight loss maintenance, 42 our research suggests participants' physical reality became aligned with their pre-existing self-identity through losing and maintaining weight lost. This served to reinforce their identity through being able to select clothing they liked and enjoyed wearing.
Practical outcomes such as fitting into preferred clothing could be a goal-setting point to either help people develop a desire to lose weight, or to continue to maintain a reduced body weight. We note that this perspective is very well represented, and seemingly well understood by commercial weight loss providers, as a legitimate goal or as a major theme of testimonials from their clients.
Despite our participants having tried to lose weight at least once-and in most cases, numerous times-previously, it seems the persistence of counterproductive habits remained, until meals and opportunities to eat were removed by the VLED and MRPs. This absence of their everyday food intake highlighted to participants a myriad of contexts in which they would reach for food, either consciously or unconsciously. These habits were newly understood as being unnecessary and detrimental to the participants' goal of permanent weight loss. For example, the assertion by our participants that they were for the first time recognizing what real hunger feels like, and that they are far more comfortable going without food than they had previously thought, made their previous failures to connect with their natural hunger cues more understandable. Taken together, participants' experiences and results during and after the VLED worked together to allow participants to critically evaluate their previous lifestyle habits and either cease old habits or introduce new habits, in order not to circumvent the results they had gained.
Identifying cues and triggers that lead to unhelpful eating habits is necessary in order to address unhelpful eating habits and develop alternative, intention-driven habits. 43 For example, knowing that stress or boredom leads a person to habitually reach for chocolate is considered to be a first step in eliminating that habit. Once identified, a disruption to that habit can be set in place to slowly cause the old habit to cease and a new habit to take its place. This is usually done by preplanning an alternative behavior in the same contextual situation that -with time and persistence-should replace existing undesirable habits. 44 According to one estimate, it takes an average of 66 days to change an old habit or form a new habit. 45 The VLED phase of the TEMPO Diet Trial being 4 months (120 days), presumably gave the participants time to recognize and eliminate unhelpful habits.
The VLED acted not only as an identifier and but also as a disruptor to unhelpful habits. Because the participants had to respond to usual triggers to snack or eat by using different (non-eating) strategies for 4 months, this appeared to embed new responses to old triggers and forced the participants to actively re-program behaviors that had previously contributed to their inability to lose weight. Additionally, recent analysis of food diaries of participants of the TEMPO Diet Trial has shown that diet quality improved for all participants from baseline, and this was mostly driven by a reduction in the intake of discretionary foods. 46 The VLED intervention group also showed higher levels of physical activity and lower levels of sedentary behavior at 6 and 12 months, in comparison to the non-VLED group. 47 Taken together, these factors likely contribute to the longer-term successful weight loss that is seen in response to VLEDs.
Participants continued to have feelings of self-efficacy in their own ability to either maintain their reduced body weight (in those who were at a stable weight), or to re-lose weight that had been regained (in those who had started to regain weight). Contrary to beliefs by healthcare professionals that VLEDs will induce yo-yo dieting with negative psychological effects, we found that participants who had regained some of their weight continued to feel confident that they could successfully return to their goal weight, despite feeling disappointed about their weight regain. The success and discipline they had shown to themselves during the VLED conferred conviction that they had undertaken actions that could subsequently be repeated. While the majority of studies have shown an inverse relationship between higher number of past weight loss attempts and subsequent weight loss attempts, other studies have failed to show a relationship between these two variables. [48][49][50] Although the number of previous weight loss attempts may have a deleterious effect on future efforts, a history of larger amounts of weight loss has shown to be associated with greater weight loss in subsequent attempts. 48 This suggests that people who have lost large amounts of weight in the past retain higher levels of self-efficacy and confidence for the success of future weight loss attempts. 48 However, while it has been shown that higher self-efficacy can be predictive of future successful weight loss attempts, and successful weight loss attempts can increase self-efficacy, making weight maintenance more likely, to our knowledge, a continued feeling of confidence and self-efficacy has not been researched or demonstrated in individuals who have subsequently re-gained weight after a weight loss attempt. [51][52][53][54][55] It is plausible however, as found by Myers et al., 48 that continuing high self-efficacy among the participants in this research is due to the large amount of weight lost with a VLED, despite their weight re-gain.
Finally, weight maintenance in our participants was supported by the ongoing use of MRPs, either as a daily supplement instead of a meal, or intermittently to lose weight or offset times of overindulging. The use of MRPs has shown to be helpful during weight maintenance, 56 and is as effective as medication for maintaining a reduced body weight. 57 Our participants chose to use MRPs of their own accord and purchased them from local retailers and found ways to use them effectively in the context of their own lifestyles. This appears to be a strategy that is easy to adopt, and that allows for easier maintenance of a reduced body weight. 58,59 To our knowledge, this is the first qualitative study to explore post-VLED weight maintenance experiences. Given only three studies qualitatively explore the use of VLEDs to date, as outlined in a systematic review, 28 this study added information not only about the use of VLED in a previously unstudied group (post-menopausal women) but also, regarding the post-diet weight maintenance phase.
With consistent and known timing of interviews, the length of the weight maintenance phase is explicitly stated, and we were able to 312gather experiences of those who had maintained their weight and those who had regained weight after the VLED intervention.
Because our study was conducted in post-menopausal Caucasian women, the results may not be transferable to other populations, including men. It is also possible, despite our sub-study results closely resembling the whole group's weight loss results, that our respondents were participants who felt more positively about their experiences of weight maintenance. Indeed, the participants who were interviewed 3 for this qualitative sub-study had lost a mean of 2.7 and 5.6 kg more at 4 and 12 months respectively, compared to participants 8 who either did not respond or declined to participate in this sub-study. Further, our participants were part of a clinical trial which involved intensive support and strict inclusion criteria, and therefore their experiences may differ significantly from people using VLEDs for self-directed weight loss and/or who would not have met the criteria for inclusion in the larger TEMPO Diet Trial.

| CONCLUSION
Overall, the VLED intervention was well received and apparently easily executed by participants in this sub-study. The significant weight loss increased motivation to maintain the reduced body