A brief behavioural intervention to promote regular self-weighing to prevent weight regain after weight loss: a RCT

Article history

The research reported in this issue of the journal was funded by the PHR programme as project number 12/179/09. The contractual start date was in June 2014. The final report began editorial review in June 2017 and was accepted for publication in October 2017. The authors have been wholly responsible for all data collection, analysis and interpretation, and for writing up their work. The PHR editors and production house have tried to ensure the accuracy of the authors’ report and would like to thank the reviewers for their constructive comments on the final report document. However, they do not accept liability for damages or losses arising from material published in this report.

Declared competing interests of authors

Kate Jolly was an investigator on a trial in which the weight loss intervention was donated to the NHS by Slimming World^® (Alfreton, UK; www.slimmingworld.co.uk) and Rosemary Conley Health and Fitness Clubs (Tharston, UK; www.rosemaryconley.com). Paul Aveyard reports grants from Weight Watchers^® (Maidenhead, UK; www.weightwatchers.com/uk) and Cambridge Weight Plan^® (Cambridge Weight Plan Ltd, Corby, UK) and non-financial support from Slimming World, Weight Watchers and Rosemary Conley Health and Fitness Clubs, outside the submitted work.

Copyright statement

© Queen’s Printer and Controller of HMSO 2019. This work was produced by Daley et al. under the terms of a commissioning contract issued by the Secretary of State for Health and Social Care. This issue may be freely reproduced for the purposes of private research and study and extracts (or indeed, the full report) may be included in professional journals provided that suitable acknowledgement is made and the reproduction is not associated with any form of advertising. Applications for commercial reproduction should be addressed to: NIHR Journals Library, National Institute for Health Research, Evaluation, Trials and Studies Coordinating Centre, Alpha House, University of Southampton Science Park, Southampton SO16 7NS, UK.

2019 Queen’s Printer and Controller of HMSO

Obesity

Obesity is the most common cause of premature mortality in the UK1 and a significant cause of morbidity in terms of increased risk of type 2 diabetes mellitus, cardiovascular disease and many cancers. 2,3 In the UK, the rates of obesity have more than doubled in the last 25 years, and being overweight has become the norm for adults. 2 Over one-quarter of adults in the UK are obese, and > 63% are either overweight or obese. 4 Obesity is associated with a reduced life expectancy of up to 14 years. 5 Direct costs of obesity are about £5.1B per year, thus placing a significant economic burden on the NHS and society. Given its prevalence and costs, obesity is a significant public health priority in the UK, as well as in other developed countries. Although many behavioural weight loss treatments are effective,6–8 long-term maintenance of weight loss remains a critical challenge. The period after initial weight loss is when people are at the highest risk of weight regain. Few people (1 in 10) recover from even minor lapses of 1–2 kg of regain in weight. 9 On average, people will regain one-third to half of their lost weight within the first year following treatment and will return to their baseline weight within 3–5 years after treatment. 10,11 These data clearly indicate that weight regain is common and efforts are needed to prevent it. Given the large numbers of people who need to lose weight and who need to maintain this weight loss, there is a critical need for low-cost, practical and scalable interventions that can effectively help adults to maintain their weight loss throughout their lives.

Maintenance of weight loss

Compared with weight loss trials, relatively few randomised controlled trials (RCTs) have focused on weight maintenance, and those trials that do exist have tended to evaluate intensive interventions. Over time, it becomes increasingly difficult for individuals to continue to follow the weight management strategies learnt during their attendance at a weight loss programme. This is because most people’s commitment wanes and they drift away from their weight loss eating pattern. This is a problem because a return to former eating and activity habits is associated with weight regain. Ultimately, weight loss will be effective and cost-effective only if weight loss is maintained, but weight loss maintenance is more difficult than weight loss. Therefore, interventions that can successfully help adults to manage their weight throughout their lives are required. The strategies required for weight loss may be different from those required for weight loss maintenance or the prevention of weight gain, and research is required to establish what these particular strategies or approaches might be. Furthermore, there is no consensus about the most effective weight maintenance strategies, the intensity of these strategies and/or the best mode or timing of delivery. Novel interventions are needed to capitalise on the success of individuals’ initial weight loss efforts.

An early meta-analysis12 of 42 RCTs that focused on the effects of extended care on weight loss maintenance reported on strategies that appeared to be useful for weight loss maintenance: taking medication, following a low-fat diet, doing physical activity, having continued contact with others, undergoing problem-solving therapy, increasing protein intake, increasing caffeine intake for those consuming < 100 mg per day, and receiving acupressure. The review authors noted that interventions lacked a theory base and the attrition rate was high in many of the trials (> 35%). Ramage et al. 13 systematically reviewed the evidence (all study designs) for healthy strategies for successful weight loss and weight maintenance. The review concluded that a combination of energy and fat restriction, regular physical activity and behavioural strategies, such as self-monitoring and adhering to a calorie goal, were required for successful weight loss maintenance.

A more recent review14 of RCTs of weight loss maintenance interventions for adults who are obese after clinically significant weight loss found that behavioural interventions that focused on both diet and physical activity led to an average difference of –1.56 kg [95% confidence interval (CI) −2.27 to −0.86 kg; 22 studies, 25 comparisons, 2949 participants] in weight regain in intervention participants compared with control participants 12 months after randomisation. At 18 months (7 studies, 13 comparisons), the mean difference in weight was –1.96 kg (95% CI –2.73 to –1.20 kg). At 24 months, only two studies were found to have reported outcomes and the mean difference in weight change was –1.48 kg (95% CI –2.27 to –0.09 kg). At 30 months, the effect was not significant (–0.85 kg, 95% CI –1.81 to 0.11 kg). No adverse events were reported for any of the behavioural or lifestyle weight loss maintenance trials included in the review. There was no evidence that more intensive interventions were more effective than less intensive interventions, that there were any differences between internet-delivered interventions and comparator groups, or that delivery of the same intervention face to face was more effective than remote delivery (i.e. telephone or via the internet). None of the included review studies had been conducted in the UK. The findings from this review demonstrate that behavioural lifestyle interventions have the potential to be effective in reducing, or slowing down, weight regain after initial weight loss for up to 2 years, but the strength of the evidence is limited because of methodological shortcomings in the included studies. Of particular concern is the fact that many trials reported data only for participants who completed their intervention, introducing the possibility of bias. The review identified that the greatest need for further research was in the area of lifestyle interventions focused on supporting people to manage their weight over the longer term by regulating food intake and increasing physical activity.

A systematic review and synthesis of qualitative research15 identified a wide range of processes and modifying variables involved in weight loss maintenance. Qualitative studies were included if they focused on the experiences of adults who had previously been or were currently obese or overweight. The authors of this review thematically synthesised studies and, from this, developed a model of weight maintenance. The model proposed that making the behaviour changes required for weight loss maintenance generates psychological ‘tension’ because of the need to over-ride existing habits and the incompatibility of the new behaviours with the fulfilment of psychological needs. If successful weight loss maintenance is to be achieved, then management or resolution of this ‘tension’ needs to occur. Management of the tension can be achieved through self-regulation, renewing motivation and managing external influences, but this is likely to require constant effort and resolution through changing habits, finding non-obesogenic methods for addressing needs and, possibly, changes in self-concept.

The reviews presented here concluded that further research is needed on the question of how to successfully help people maintain their weight loss. In particular, authors commented that most trials in the reviews were of low or moderate quality and that many trials of longer-term weight maintenance had been inconclusive. In addition, the review by Dombrowski et al. 14 noted that trials had been heterogeneous in relation to setting, type and duration of interventions and length of time for follow-up. Many trials had methodological issues that could have introduced bias (i.e. inadequate reporting of the randomisation process, blinding or a lack of use of the intention-to-treat principle). Substantial dropout from follow-up was also evident in many trials; this is typically associated with participants’ inability to maintain their weight loss and, therefore, a reluctance to provide data about their weight at follow-up. Loss to follow-up is highly problematic in weight loss maintenance trials because it can lead to an overestimation of the true effects of an intervention.

Since the present study was commissioned, four other trials of weight loss maintenance interventions have been published; three were conducted in England16–18 and one was conducted in the USA. 19 Simpson et al. 16 conducted a three-group RCT of a motivational interviewing-based intervention for weight loss maintenance in adults (n = 170) who had lost at least 5% of their body weight in the previous 12 months. The intervention was based on three main components: motivational interviewing (incorporating implementation intentions), social support and self-regulation/self-monitoring. The intervention used a number of techniques and strategies central to motivational interviewing, for example, promoting intrinsic motivation and encouraging goal-setting and action-planning. Participants were randomised to an intensive intervention group, a less intensive group or a control comparator group. This trial was originally designed as an effectiveness trial with the primary outcome of body mass index (BMI) assessed 3 years after randomisation. Owing to difficulties with participant recruitment, however, the trial was closed early and was reported as a feasibility RCT with the 12-month follow-up as the end point. A total of 170 participants were recruited; retention (84%) and intervention adherence were high (intensive, 83%; less intensive, 91%). The intensive intervention had mean BMI scores 0.96 kg/m² lower than controls (95% CI –2.2 to 0.2 kg/m²) at the 12-month follow-up. A difference of –0.21 kg/m² (95% CI –1.4 to 1.0 kg/m²) was reported between the less intensive intervention group and the control group. The average difference in weight was 2.8 kg (95% CI –6.1 to 0.5 kg) between the intensive intervention group and the control group, with a smaller difference noted for the less intensive intervention group (–0.70 kg, 95% CI –4.10 to 2.70 kg). The intensive and less intensive interventions costed £623 and £198, respectively, to deliver. The results of this study are promising, but it was not adequately powered to detect the outcome of interest and the cost of the intensive intervention is likely to mean that it is unlikely to be funded as a routine public health intervention.

The authors of this report completed a RCT (Lighten Up Plus)17 to evaluate a 12-week Short Message Service (SMS) text message weight maintenance intervention that encouraged weekly weighing. The comparator group comprised 380 men and women who were overweight or obese and had completed a 12-week commercial weight loss programme. Participants were eligible to take part in Lighten Up Plus if they had a 12-week weight recorded at their 12-week weight loss programme or had attended a minimum of 9 out of 12 of their weight loss programme sessions and had had their weight measured at their weight loss service in the previous 2 weeks. Participants also had to have access to scales to weigh themselves and needed to own a phone that could receive SMS text messages. The comparator group received a brief weight maintenance support telephone call from non-specialist call centre staff (lasting 10–15 minutes), which gave lifestyle information on a balanced diet, portion control and regular exercise. A brief ‘hints and tips’ leaflet for weight loss maintenance was also mailed to this group. The comparator group received a final call 12 weeks after the initial call, which recapped the hints and tips given during the first call. The intervention group received the same as the comparator group and, additionally, received a SMS-text messaging intervention of 12 weeks’ duration to encourage weekly weighing to prevent weight regain. Text messages that asked participants to reply to the text by sending their weight and whether they had gained or lost weight in the previous week were sent weekly. A reply text was then sent to participants depending on whether or not they had gained weight: a congratulatory message if weight had been maintained or lost; advice about diet and increasing physical activity if weight of < 2 kg had been gained and if this had occurred for < 3 consecutive weeks; or an offer of telephone support if weight had been gained 3 weeks in succession. If participants continued to gain weight, they were offered referral to receive specialist support. The primary outcome was change in weight at the 9-month follow-up, with a follow-up also at 3 months. There was no evidence that the intervention was effective, although the direction of change favoured the intervention group. Data showed that the intervention group adhered well to weekly weighing, with a median of nine weights texted back to the research team out of a possible 13.

Voils et al. 19 assessed the effectiveness of a weight loss maintenance intervention in people (n = 222) who had lost ≥ 4 kg during a 16-week group weight loss programme (nutrition training). The maintenance intervention was delivered primarily by telephone and focused on satisfaction with outcomes, preventing relapse, self-monitoring and obtaining social support. The maintenance intervention was delivered by dietitians and involved both group (three sessions) and individual telephone support (eight sessions) delivered over 42 weeks, followed by 14 weeks of no contact. This was compared with usual care of no further intervention after the initial nutrition training. Follow-up took place at 56 weeks post study randomisation. The mean weight gain was statistically lower in the intervention group (0.75 kg) than in the usual-care group (2.36 kg) (mean difference 1.60 kg, 95% CI 0.07 to 3.13 kg; p = 0.04). Of interest here is that the intervention costs were US$276 (£184) per participant.

Very recently, Sniehotta18 completed the NU Level trial, which investigated an automated remote weight monitoring and feedback intervention using participants’ mobile phones as the mode of delivery over 12 months (n = 288). Participants received wirelessly connected weighing scales that, every time they were used, transmitted participants’ weight in real time to the research team, and these data formed the basis of the automated feedback to participants on their weight maintenance progress. Participants also received an initial individual face-to-face weight loss maintenance intervention session with a psychologist. This intervention was compared with a control intervention comprising quarterly text messages with links to evidence-based weight management guidance. The control group also received the same wireless data transfer weighing scales, but no additional feedback on weight progress was provided. The intervention was ineffective in preventing weight regain, although data indicated that those in the intervention group were able to weigh themselves regularly throughout the intervention (an average of 4.4 times per week).

The trials by Volis and Simpson16,19 highlight that intensive weight loss maintenance interventions might be effective but such interventions would be expensive to deliver at scale. The trial by Sniehotta18 also evaluated an intensive intervention that was ineffective, although this trial does demonstrate that participants are able to adhere to daily weighing. A relatively brief intervention was evaluated by Sidhu et al. 17 and was found to be ineffective, although the results were in the direction that favoured the intervention group and the process data showed that participants were able to engage well with a text messaging system that encouraged them to weigh themselves weekly and to record this weight.

Self-regulation and self-monitoring of weight

Self-monitoring is a key component of self-regulation and has been shown to be important for successful health behaviour change. One promising specific behavioural strategy for weight management is regular self-monitoring to check progress against a target. The potential efficacy of self-monitoring has been based on the principles of self-regulation theory20 and the relapse prevention model. 21 Self-regulation has been described as a process that has three distinct stages: self-monitoring, self-evaluation and self-reinforcement. 20 Self-monitoring is a method of systematic self-observation, periodic measurement and recording target behaviours, with the goal of increasing self-awareness. The awareness fostered during self-monitoring is considered an essential initial step in promoting and sustaining behaviour change. Self-monitoring in the context of self-weighing can show individuals how their behaviour affects their weight and allows them to adjust their behaviour to achieve their goals. In other words, self-regulation theory postulates that motivation for behavioural change results from self-monitoring and observing the comparison of the current recorded behaviour with the desired outcome, along with the interplay between awareness, self-observation, recording and self-evaluation.

Several quantitative and qualitative studies22–25 have reported that individuals who are successful with weight loss and weight loss maintenance frequently monitor their weight. More specifically, there are observational data indicating that a lower frequency of self-weighing is associated with higher fat intake, increased disinhibition and decreased cognitive restraint,26 all of which are behavioural strategies associated with weight gain and regain. Self-monitoring of weight may act as a reward for individuals who control their food intake and physical activity behaviour and who are provided with positive feedback from the scales, thereby enhancing their motivation and reducing the potential for relapse. Frequent self-weighing and reflection of weight progress may also improve self-efficacy for weight maintenance, which, in turn, could improve body image. Encouragement of self-monitoring and recording of weight is a simple concept for a health professional or public health communication to advocate. It is simple for people to understand and implement, and it is the kind of behaviour that could become habitual and, thus, performed throughout life, in the way that cleaning one’s teeth becomes automatic and effortless. Trials have shown that participants can adhere to daily self-monitoring of weight. 16,18,27,28

Data from the US National Weight Control Registry29 indicate that the only factor that people losing weight have in common is that they use both diet and exercise to achieve weight loss. When examining weight loss maintenance, however, there appear to be several common factors, including self-monitoring, having high levels of physical activity, consuming a low-fat diet and eating breakfast. As suggested earlier, strategies needed for successful weight loss and weight loss maintenance may be different because weight loss requires individuals to produce a negative energy balance but weight loss maintenance requires continued and ongoing energy balance. Most importantly, this energy balance needs to be sustained by engaging in behaviours and strategies that can be demonstrated over the long term. Regular self-weighing is a self-monitoring behaviour that can be easily and readily used by people throughout their lives, regardless of whether they are aiming to lose weight or maintain their weight loss.

Data from RCTs30 also show that self-weighing may be an effective strategy within weight loss programmes. Our own systematic review of the effectiveness of self-weighing for weight loss showed that adding advice to self-weigh to a behavioural programme improved its effectiveness when compared with no intervention, but only four trials had assessed this, and the estimate of effect was imprecise and clouded by the use of other self-regulatory components. Another systematic review31 of self-weighing in both weight loss and weight gain prevention interventions reported that 75% of self-weighing-only interventions and 67% of interventions that combined self-weighing with other strategies demonstrated improved weight outcomes. No negative psychological effects were reported.

Potential risks and harms of self-weighing

Although self-weighing may help people control their weight, there may be concerns that it will have negative psychological consequences or lead to the adoption of unhealthy weight control practices. Some researchers have suggested that feedback about body size may result in psychological distress and that self-weighing may worsen body image and/or mood by continuously reinforcing to people that their current body size is not appropriate or ideal32,33 or lead to unhealthy dietary behaviours, such as binge eating and skipping meals. 34 However, there is little evidence to support these concerns. A very recent systematic review and meta-analysis (RCTs and observational studies)35 of the psychological impact of self-weighing found minimal evidence that self-weighing was associated with adverse psychological outcomes in relation to affect, body-related attitudes or disordered eating. There was, however, a small negative association between self-weighing and psychological functioning. Of interest here is that duration significantly moderated the relationship between self-weighing and body-related attitudes: self-weighing over longer durations compared with over shorter durations was associated with more positive and/or less negative body-related attitudes. Study design also moderated the relationship: observational studies tended to report more negative associations between self-weighing and psychological outcomes, and RCTs, which are less prone to bias, tended to report a positive relationship. In a very recent study that post-dates this review,27 49 female university students who were randomised to a daily weight monitoring intervention or a control condition with a 20-week follow-up did not report any harmful effects from regular weighing, and showed high acceptability of and adherence to the intervention. Nevertheless, it is important that trials continue to monitor whether or not self-weighing leads to adverse psychological events.

Rationale for the study

After the pilot work was completed, the effectiveness of the intervention needed to be tested robustly before it could be implemented. Moreover, at the time of the study, commissioning groups throughout England had begun to offer patients weight loss programmes, which were often provided by commercial weight management providers but had no or minimal provision for weight maintenance support thereafter. If effective, the study would provide information to public health agencies, commissioners or services that may be considering adopting a similar model. The need for evidence about effective weight maintenance interventions delivered within public health arenas will become even more important in years to come as cuts to public spending increase. The need for evidence about effective weight maintenance programmes is also critical to ensure that weight loss is maintained following weight loss interventions given that regain is the norm. This study would provide the public and commissioners with a simple, ‘ready to go’, low-cost weight maintenance intervention that could be implemented immediately, if effective.

The primary aim of this study was to evaluate the effectiveness and cost-effectiveness of a theoretically informed brief behavioural intervention delivered by non-specialist staff to promote regular self-weighing to prevent weight regain after intentional weight loss. The intervention was compared with usual care. The research tested a weight maintenance intervention in participants who had successfully completed standard, widely available commercial and NHS weight loss programmes. Although complex, multicomponent, high-intensity interventions may be more effective than the brief intervention we planned to test here, we deliberately chose to test a simple self-weighing intervention, as this has not been done before in a high-quality trial. There are several advantages of testing simple and low-cost interventions. If we test unicomponent interventions and show that they are effective, we can build up to a more complex, multicomponent intervention. It is often difficult to go the other way. Many people in the UK and in other developed countries would benefit from losing weight, and it would not be affordable to deliver intensive interventions to this number of people. Even if our strategy leads to a smaller effect than a complex intervention, the higher reach of the simple, cheap intervention would still result in substantial gains to public health. If we have a range of simple, evidence-based self-help strategies that may prevent weight regain, we can encourage the public to use these outside formal programmes.

Aims and objectives

Study design

The study design was a RCT in which adults who had received a weight loss programme and had lost ≥ 5% of their initial body weight were allocated to receive a standard weight loss maintenance intervention (a leaflet, the EatWell Plate and a list of useful websites) or an intervention group who received the same written materials plus a 12-week behavioural weight maintenance intervention that promoted daily self-weighing and the recording of weight on a record card. The full trial protocol has been published. 39

Ethics approval and research governance

Ethics approval for the study was obtained from the Science, Technology, Engineering and Mathematics Ethical Review Committee at the University of Birmingham (ERN-13-1380). The study was conducted in accordance with the recommendations for physicians involved in research on human subjects adopted by the 18th World Medical Assembly, Helsinki, 1964, and later revisions. The trial was hosted by the Primary Care Clinical Research Trials Unit (PCCRTU) at the University of Birmingham and was conducted in accordance with the Research Governance Framework and the standard operating procedures for the PCCRTU. The University of Birmingham acted as the sponsor for the study. Research and development approval from Birmingham City Council was also obtained. An independent Trial Steering Committee was formed to oversee the conduct of the study.

Identifying potential participants and recruitment

In 2008, South Birmingham Primary Care Trust commissioned the Lighten Up service, which provided NHS patients who were obese with a free course of weight loss treatment for 3 months. The Lighten Up service commissioned three commercial weight loss programmes to offer treatment to these patients: Weight Watchers^® (Maidenhead, UK; www.weightwatchers.com/uk), Rosemary Conley (Tharston, UK; www.rosemaryconley.com) and Slimming World^® (Alfreton, UK; www.slimmingworld.co.uk). Patients could choose their weight loss treatment, and the primary care trust commissioned a telephone call centre to administer the service (Gateway Family Services). Adults who took part in the Lighten Up weight loss programmes were sent an invitation letter and information leaflet when they reached week 9 of their weight loss programme. This letter informed participants about our weight maintenance study and notified them that they would be asked whether they would be willing to take part in a study to prevent them regaining the weight they might have recently lost. During an end-of-programme courtesy telephone call from a member of the study team at Gateway Family Services, participants were asked to report their current weight and their total amount of weight loss since starting their weight loss programme. Those who had lost ≥ 5% of their starting weight were asked to participate in this study about maintaining weight loss. In other sites, namely the Public Health-funded weight management services where recruitment took place, participants were sent a letter, an information sheet, a reply form and a prepaid envelope. The research team at the university contacted these participants on receipt of the reply form to assess their eligibility and, if the participant was eligible, booked a home visit for a baseline assessment.

The eligibility criteria for access to each weight loss programmes can be found in Appendix 1.

Intervention

Outcome measures

Patient and public involvement

We approached previous users of the Lighten Up weight loss programme, who then formed the public and patient involvement (PPI) group (KB and PD); they had successfully lost weight using Lighten Up and, therefore, were similar to the individuals we wanted to recruit. The study was presented to the PPI group and their comments were incorporated into the submitted funding request proposal. Their main feedback was that it was a good idea to involve Gateway Family Services in delivering the intervention because participants would already have had contact with this organisation when attending their weight loss programme. There was some concern that people who did not own scales would be expected to buy some. The PPI group felt that it was important to give participants a choice about when the intervention calls would be made. One member commented that those at risk of relapse often have other issues going on their lives and it would therefore be important to ask about this during the calls. We did not incorporate this suggestion because to do so would be beyond the competence of the telephonists and the scope of the study. However, we included a question on major life events in the study questionnaire. Throughout the trial, Karen Biddle and Polly Dixon attended Trial Steering Committee meetings and we consulted our PPI representatives about any issues on which we needed their input.

Sample size calculation

The standard approach to sample size calculation is to aim to detect a worthwhile intervention effect. This was difficult for this study because there is a linear relationship between being overweight and mortality (30% increase in mortality per 5 kg/m² increase in BMI). 59 This kind of intervention may be applied broadly in contexts outside the specific behavioural intervention we propose here; in any case, even here, likely at its maximal intensity and cost, it is likely to cost around ≤ £10 per person in public health practice. According to the National Institute for Health and Care Excellence, even very small decrements in weight are likely to be cost-effective if they are maintained throughout a person’s life. 60 Consequently, instead of an approach by which we specified a sample size based on a clinically important difference, we proposed a sample size based on the likely size of effect we expected to achieve, namely a 2-kg difference in change in weight at the 12-month follow-up. A total of 280 participants randomised to each group (n = 560) was sufficient to detect a 2-kg difference in change in weight at the 12-month follow-up between the intervention and usual-care groups, with 90% power and 5% significance level. This estimate was based on our pilot trial, in which we found that the standard deviation (SD) of the difference from baseline (i.e. the end of weight loss/start of maintenance programmes) to the 12-month follow up in those who lost at least 5% of their starting weight was 6.3 kg. Having 560 participants also allowed for 20% loss to follow-up at 12 months. In our pilot, the intention-to-treat analysis included all participants regardless of whether or not they had lost weight in their weight loss programme or whether or not they had accepted or received the maintenance intervention; the adjusted mean difference in weight between intervention and control was 0.68 kg. However, in the per-protocol analysis of those who had accepted the maintenance intervention (i.e. those motivated to change), the mean group difference was larger, at 2.96 kg. Similarly, in this study, we planned to recruit participants who had agreed to accept the weight maintenance programme and who showed good adherence to their weight loss programme by losing at least 5% of their starting weight; therefore, we conservatively estimated the likely difference in weight change to be about 2 kg, and more on a par with the findings from the per-protocol analysis in our pilot study. In addition, the intervention group in this study received more contacts than the intervention group in the pilot did, potentially increasing the difference we expected to find.

Changes to the protocol

By March 2015, recruitment had been marginally lower than expected, and we made the decision to recruit from other local programmes. As this study was focused on testing a weight loss maintenance intervention, as long as people purposefully lost weight the method of that weight loss was not important. Participants in programmes funded by Dudley and Solihull City Councils were eligible to join this study if they met the study eligibility criteria. These programmes were either commercial weight loss programmes, as per Lighten Up in Birmingham, or other community programmes (Fit Blokes and SHAPES). Participants recruited from Dudley and Solihull were sent a screening questionnaire and asked to return this to the study team using a Freepost envelope if they were interested in taking part. The research team determined eligibility, and thereafter flow through the trial occurred in the same manner as for those recruited via Lighten Up. Gateway delivered the intervention to all participants.

We also planned a cost-effectiveness analysis with modelling of the long-term health consequences that may follow successful weight maintenance, but this was not completed as the intervention was ineffective.

Partner collaboration

This study was based on a partnership between the study team (Universities of Birmingham and Oxford), Gateway Family Services and Birmingham Public Health.

Statistical analysis

Recruitment and retention of randomised participants

A total of 813 potential participants were screened, of whom 230 (28.3%) were excluded because either they were ineligible (n = 49) or they declined to participate (n = 181). A total of 583 participants were randomised (usual care: n = 292; intervention: n = 291). Once randomised, the vast majority of participants provided follow-up data at 3 months (usual care, 90.4%; intervention, 97.6%) and at 12 months (usual care, 88.4%; intervention, 90.0%) (primary end point), well below the 20% loss to follow-up rate that we had planned for in the study. The Consolidated Standards of Reporting Trials (CONSORT) participant flow diagram is presented in Figure 1. At 3 months, seven weights were self-reported (usual care, n = 6; intervention, n = 1). At 12 months, 14 weights were self-reported (usual care, n = 10; intervention, n = 4). Additional data at the 6-month follow-up (postal questionnaire with no incentive) were collected for process evaluation purposes only, and 399 (68.4%) participants provided data (usual care, n = 204; intervention, n = 195). The first participant was randomised on 14 July 2014 and the last on 21 January 2016.

FIGURE 1.

The CONSORT recruitment flow chart.

Effectiveness outcomes

Effect of the intervention: primary analysis of weight at 12 month’s follow-up

The mean unadjusted change in weight was 0.39 kg (SD 7.12 kg) in the intervention group and –0.17 kg (SD 7.57 kg) in the usual-care group. A non-significant adjusted treatment difference of 0.53 kg (95% CI –0.64 kg to 1.71 kg) was found in favour of the usual-care group (Table 5). The distribution of the residuals from the ANCOVA analysis was plotted using a histogram for each group (Figure 2). The residuals appeared to be sufficiently normally distributed for the analysis to proceed. The mean change in weight at assessment points, with standard error bars, is shown in Figure 3.

TABLE 5 - Adjusted and unadjusted results for weight change (kg) at 12 month's follow-up

ANCOVA with covariates for randomised group, baseline weight and intention at baseline to continue to attend weight loss programme. A negative value indicates that weight reduced from baseline. Difference is intervention – usual care.

Weight outcome	Randomised group
	Usual care (n = 292)	Intervention (n = 291)
Unadjusted mean change in weight from baseline (SD) (units = kg)	–0.17 kg (7.57 kg)	0.39 kg (7.12 kg)
Adjusted difference in treatment effect (CI)a	0.53 kg (–0.64 to 1.71 kg)
p-value	0.375

FIGURE 2.

Histogram of the residuals from the primary ANCOVA analysis by randomised group. (a) Usual care; and (b) intervention.

FIGURE 3.

Mean change in weight at each assessment point, with standard error bars.

The imputation for objective weight at 12 months used self-reported weight as the preferred value, followed by imputation with the baseline weight plus 0.3 kg per month. A sensitivity analysis was specified in which self-reported weight would be set as a missing weight. No imputation for missing weights was used (i.e. completers-only analysis) (Table 6). Compared with the primary analysis, the weight reduction in both groups was larger, but the treatment difference was smaller. The p-value of the treatment difference was larger, indicating no significant difference.

TABLE 6 - Adjusted and unadjusted results for the primary outcome of weight change from baseline to 12 months (missing objective weights at 12 months not imputed)

ANCOVA with covariates for randomised group, baseline weight and intention at baseline to continue to attend weight loss programme. A negative value indicates that weight reduced from baseline. Difference is intervention – usual care.

Weight outcome	Randomised group
	Usual care (n = 248)	Intervention (n = 258)
Unadjusted mean reduction in weight from baseline to the 12-month follow-up (SD)	–0.84 kg (7.99 kg)	–0.35 kg (7.81 kg)
Adjusted difference in treatment effect (95% CI)a	0.45 kg (–0.90 to 1.81 kg)
p-value	0.51

Maintenance of weight loss

A total of 234 (80.1%) and 235 (80.8%) participants regained ≤ 1 kg at the 3-month follow-up in the usual-care and intervention groups, respectively. The logistic regression model produced a non-significant odds ratio of 1.05 (95% CI 0.69 to 1.58) comparing the odds of regaining ≤ 1 kg in the intervention group with the odds in the usual-care group at 3 months. At 12 months, 134 (45.9%) and 130 (44.7%) participants regained ≤ 1 kg when compared with baseline weight in the usual-care and intervention groups, respectively. The logistic regression model produced a non-significant odds ratio of 0.96 (95% CI 0.69 to 1.33) comparing the odds of regaining ≤ 1 kg in the intervention group with the odds in the usual-care group (Table 9).

Subgroup analyses for primary outcome of change in weight

The results of the prespecified subgroup analyses performed are presented in Figure 4. The subgroup variable included whether or not participants at baseline stated that they would continue to attend their weight loss programme, whether participants at baseline stated they wanted to lose or maintain their weight, the previous weight loss programme that the participant attended and the participants’ gender. An ANCOVA model was fitted, similar to the model performed for the primary analysis, but including the subgroup categorical variable and an interaction term between the randomised group and the subgroup categorical variable. There was no evidence that these subgroups modified the effectiveness of treatment.

FIGURE 4.

Forest plot of the results of change in weight from baseline to 12 months, obtained from the subgroup analyses.

Continued attendance at weight loss programme and weight loss maintenance

At 3 months, about 6 out of 10 participants in both the intervention and usual-care groups had chosen to continue attending the weight loss group at their own expense. By the 12-month follow-up, around 4 out of 10 participants in both the intervention and the usual-care groups were doing so (Table 10). Paying to attend a weight loss group during the first 3 months after baseline was associated with a higher likelihood of staying ≤ 1 kg above baseline weight at 12 months (Table 11).

In a series of analyses, we examined whether or not intent to use, and use of, weight loss programmes after baseline modified the effect of the intervention on weight change at 12 months, using appropriate multiplicative interaction terms. This is because we hypothesised that the maintenance intervention would be most effective for people who were planning to stay at their current weight and that the effect may be lost in people continuing to attend a weight loss programme, which itself increases restraint. On the whole, there was no evidence that intent to use or continued use of a behavioural weight management programme modified the effect of the intervention (Figure 5). We examined whether or not the number of sessions attended during the first 3 months after baseline and between 6 and 12 months after baseline modified the effect of the analysis. There was a significant interaction at 3 months, when the coefficient was 0.26 (p = 0.017), but not at 12 months, when the coefficient was –0.02 (p = 0.789).

FIGURE 5.

Forest plot of the results of change in weight from baseline to 12 months, obtained from the subgroup analyses, including the self-weighing category.

Objective recording of weight and weight change at 3 months in the intervention group

The relationships between frequency of self-weighing in the intervention group and changes in weight at the 3- and 12-month follow-up points are displayed in Table 12. At 3 months, participants who weighed themselves on the scales on ≥ 90% of occasions reduced their weight by –3.2 kg and those who weighed themselves on < 50% of occasions gained 0.2 kg. At 12 months, those weighing themselves on ≥ 90% of occasions lost 3.1 kg and those weighing on < 50% of occasions gained 3.3 kg.

Psychological harm and body image

There was no significant difference between the groups at either follow-up point with regard to uncontrolled eating, emotional eating or body image scores (Table 13).

Serious adverse events requiring hospitalisation

No serious adverse events requiring hospitalisation were recorded in either group. There was no evidence that the intervention increased the frequency of symptoms of bulimia, assessed using the SCOFF questionnaire (Table 14).

Intervention delivery by telephone calls

The intervention was delivered mainly through the initial telephone calls, and most people received all three intervention calls. In total, out of 291 participants, 251 (86.2%) received all three intervention calls, 34 (11.6%) received two calls, 3 (1.0%) received one call and 3 (1.0%) received zero calls. A total of 824 calls were delivered as part of the intervention. Thirty-six intervention calls were audio-recorded and checked for fidelity in accordance with the intervention manual. Using the intervention checklist [see project web page: www.journalslibrary.nihr.ac.uk/programmes/phr/1217909/#/ (accessed 12 February 2015)], calls were delivered per protocol as follows: 71.9% for week 1, 60.3% for week 2 and 97.0% for week 4.

Intervention delivery by text message

A total of 21 out of 291 (7.1%) participants requested not to receive the intervention support text messages during the study.

Self-weighing frequency

There was clear evidence that the intervention achieved its goal of increasing the proportion of people who weighed themselves. In the usual-care group, around 16% of participants weighed themselves on almost every day of the week at baseline. This was similar at 3 months and fell to around 10% at 6 and 12 months. In the intervention group, around 18% of participants weighed themselves almost every day, rising to 61% by 3 months and remaining at > 40% at 6 and 12 months (Table 15).

Weight control strategies and cognitive restraint of eating

There were no significant group differences in the Weight Control Strategies Scale (WCSS) subscales at either follow-up point, except in the case of the self-monitoring strategies subscale of the WCSS questionnaire, for which the intervention group had a significantly higher score than the usual-care group (Table 16). There was no significant difference between the groups for cognitive restraint of eating scores (see Table 16).

Perceptions of daily self-weighing (intervention group)

The intervention group was assessed in terms of their perception of self-weighing and the formation of the habit of self-weighing at 3 and 12 months (Table 17). Overall, participants felt that weighing themselves regularly was easy to do, easy to remember and helpful. On average, participants did not report that regular weighing was frustrating or that it made them feel anxious or self-conscious. On average, participants reported that it was likely that they would weigh themselves regularly beyond the end of the study. For all items, scores were marginally higher at the 3-month follow-up than at the 12-month point.

Automaticity of regular weighing/habit strength score (intervention group)

Habit scores were moderate, at around 2.6 out of 4, indicating that many people were not in a habit of weighing themselves, which reflects the fact that a large minority at 3 months and a majority at 12 months were not weighing themselves daily (see Table 17).

Self-reported frequency of self-weighing at 3 months and weight change at 12 months

We wanted to assess whether or not self-reported frequency of self-weighing, as reported at 3 months by all participants (categorical variable), was related to weight change at 12 months; it was not significantly related (see Figure 5). Summary data on the frequency of self-weighing in both groups at each follow-up point are presented in Table 15.

Frequency of recording of self-reported weight using the record card

One hundred and sixty-two people (55.6%) returned their written weighing records at 3 months. Of these participants, most had recorded their weight on nearly every day of the week throughout the intervention period (Table 18).

Objective recording of weight (intervention group)

As the scales recorded whether or not a person had weighed themselves, and this information was transmitted remotely back to the research team, we were able to assess whether or not people in the intervention group had weighed themselves. Participants in the intervention group weighed themselves on 57% of 365 days (a mean of 207.4 times/365 days). At 3 months, 55% of participants were weighing themselves on 5 or 6 days per week, and by 12 months 31% were doing so (Table 19), figures that are similar to the self-reported data (see Table 18).

Relationship between frequency of objective recording of weight and self-reported frequency of weight

The association between self-reported and objective recording of weight was assessed at weeks 1 and 12 of the intervention using Bland–Altman plots. This is based on data only for those who submitted self-report cards (n = 162) (Table 20 and Figures 6 and 7).

FIGURE 6.

Bland–Altman plot of the agreement between frequency of objective weight recordings and self-reported weights on the weight record card during intervention week 1.

FIGURE 7.

Bland–Altman plot of the agreement between objective weights obtained from scales and self-reported recorded weights at the week of the 3-month follow-up.

Automaticity (frequency of self-weighing) and change in weight

Regression analyses were performed to determine if the frequency of self-weighing predicted the habit score, and whether or not the habit score predicted a change in weight, to lend support to the hypothesised causal pathway presented in Figure 8. Higher frequency of weighing was associated with higher habit strength scores and lower weights at 3 and 12 months. Higher habit strength reported at 6 and 12 months was associated with a larger reduction in weight at 12 months, but habit strength at 3 months was not significantly associated with weight change at 3 or 12 months (Table 21).

FIGURE 8.

Pathway diagram of how the intervention may lead to weight change.

Did participants reapply the lessons learnt in their weight loss programme?

We did not complete an exploration of whether the intervention worked by encouraging participants to reapply lessons and techniques they learnt during their weight loss programme or whether it may have worked by encouraging participants to re-enrol in a weight loss programme, as the intervention was ineffective.

Health economics evaluation

A long-term evaluation of the intervention was not undertaken as the intervention was shown to be ineffective. The costs of delivering the intervention and the comparator intervention were £9.52 and £0.16, respectively.

Summary of main findings

This study evaluated the effectiveness of a brief behavioural intervention delivered by non-specialist staff employed by a third-sector community organisation to promote target-setting and regular self-weighing to prevent weight regain after intentional weight loss. We examined the effect of the intervention on weight and other outcomes of interest at the 3- and 12-month follow-up. Process outcome analyses were also performed and the fidelity of intervention delivery and content was assessed. A long-term economic analysis was not performed as the intervention was ineffective.

A total of 583 participants were randomised (usual care, n = 292; intervention, n = 291). Most participants provided follow-up data at 3 months (usual care, 90.4%; intervention, n = 97.6%) and at 12 months (usual care, 88.4%; intervention, 90.0%) (primary end point). Approximately 80% of participants were living in the lowest two Index of Multiple Deprivation quartiles, and 19% were from non-white ethnic groups. In relation to demographic characteristics, baseline weight and BMI were generally balanced between the groups. At baseline, most participants (≈82%) intended to continue to attend a weight loss programme. At the 12-month follow-up for the primary outcome, the mean unadjusted change in weight was 0.39 kg (SD 7.12 kg) for the intervention group and –0.17 kg (SD 7.57 kg) for the usual-care group, with a non-significant adjusted treatment difference of 0.53 kg (95% CI –0.64 to 1.71 kg) in favour of the usual-care group. At 3 months, the mean unadjusted change in weight was –1.58 kg (SD 3.90 kg) for the usual-care group and –1.71 kg (SD 3.51 kg) for the intervention group. A non-significant adjusted treatment difference of –0.17 kg (95% CI –0.76 to 0.42 kg) in favour of the intervention group was found, indicating that an average of 0.17 kg more weight was lost by 3 months in the intervention group than in the usual-care group. A total of 234 (80.1%) and 235 (80.8%) participants regained ≤ 1 kg at 3 months in the usual-care group and intervention group, respectively. At 12 months, 134 (45.9%) and 130 (44.7%) participants in the usual-care and intervention groups, respectively, regained ≤ 1 kg of their baseline weight (odds ratio 0.96, 95% CI 0.69 to 1.33). There was no evidence that any variables related to continued efforts to lose weight or that frequency of self-weighing moderated the effect of the intervention on weight change at the 12-month follow-up. There was no evidence that the intervention caused psychological harm. There were no significant differences between the groups in terms of uncontrolled eating, emotional eating, body image state scores or symptoms of bulimia at follow-up. Intervention delivery was very high; 86.2% of participants received all three intervention calls. Objective weight recording for the intervention group showed that, on average, the intervention group weighed themselves on 57% of the 365 days on which they were involved in the study. There was no evidence that the intervention met its process targets, namely motivating greater efforts as self-regulation of energy and expenditure. Nor was there a significant difference in conscious energy restraint scores between the groups. There were no significant group differences in the WCSS subscales at follow-up except for the self-monitoring strategies subscale of the WCSS questionnaire, for which, as expected, the intervention group had a significantly higher score than the usual-care group.

Strengths and limitations

The findings from this study should be interpreted in light of its strengths and limitations. We exceeded our recruitment target of 560 participants and recruited 583 over 18 months. Most participants lived in the lowest two deprivation quartiles and about one-fifth were from non-white ethnic groups, demonstrating that we were able to reach those who are likely to be most in need of intervention. We achieved this by working in partnership with public health departments across the West Midlands and by collaborating with a third-sector health and community organisation that facilitated recruitment. The follow-up rates at 3 and 12 months were excellent, demonstrating that the strategies we employed to reduce attrition and, thus, bias worked very well. To our knowledge, this is the first trial in the world that has evaluated an intervention for weight loss maintenance where the intervention has been delivered by lay call centre staff, and is one of very few trials that has investigated an intervention specifically for weight loss maintenance for adults who have successfully lost weight before commencing a study such as ours. Rather than relying on self-reported data, which are often inaccurate and incomplete, we used a real-time objective measure of self-weighing frequency throughout the 12-month study in the intervention group to inform our interpretation of the results. The vast majority of the intervention group received their intervention calls as planned.

There were some limitations. To make the intervention relatively cheap to deliver, and to ease the problem of implementation, we used telephone operators from a third-sector community organisation to deliver the intervention, supplemented by short text messages. Fidelity to delivery of the intervention was moderate. That said, most people were weighing themselves daily during the intervention period and so moderate fidelity perhaps played a modest role in limiting the effectiveness of the intervention. As required by the funder, we recruited people who were continuing to attend their weight loss programme. Setting a target for weight loss maintenance of 1 kg more than a person’s current weight may feel like it has little relevance for a person who is actively pursuing weight loss. Moreover, continuing to attend a weight loss programme with daily targets for energy intake and expenditure is likely to invoke conscious cognitive restraint to a greater extent than in our minimal intervention. Given that most people in both groups were attending a weight loss programme, perhaps it is unsurprising that the intervention was ineffective in this context.

In addition, a high level of continued engagement with weight loss programmes during the study may have meant that participants received mixed messages about self-weighing, as commercial weight loss programmes discourage self-weighing between weekly classes. We undertook a post hoc analysis to assess whether participants who did not attend their weight loss programme benefited from our intervention more than those who continued. This showed that there was no difference in weight change between intervention and usual-care participants, suggesting that this may not be the explanation for the ineffectiveness of the intervention. Only eight participants indicated at baseline that their goal was to maintain their weight loss. This raises an important practical point regarding the best time to intervene to facilitate weight loss maintenance. The intervention text messages were generic, and future studies should consider sending more personalised messages, although this will have additional time and cost implications, particularly if interventions are to be delivered at scale. We recruited mainly women, but this is not uncommon in trials of weight management.

Comparisons with other studies

Other than our pilot study,38 this is the first RCT to have used lay staff from a community organisation to deliver a weight maintenance intervention, and so a direct comparison with other studies is not possible. Reviews of interventions have shown that it is possible to reduce weight regain through extended or intensive interventions, but overall success has been mixed and modest. 61,62 Unlike the findings from previous systematic reviews that have reported that self-weighing was effective for weight management, we did not find that a brief intervention focused on encouraging regular self-weighing and recoding of weight was an effective intervention for weight loss maintenance. The recent feasibility trial by Simpson et al. [Weight Loss Maintenance in Adults (WILMA)]16 demonstrated promising results for an intensive, individually tailored, motivational-interviewing-based intervention on weight loss maintenance. The WILMA trial involved six 1-hour face-to-face sessions and nine 20-minute telephone sessions. Although the interventions in the WILMA trial (intensive and less intensive) contained some of the behavioural techniques tested in our study here (e.g. self-monitoring, implementation intentions and goal-setting), both the intensive and less intensive versions of the interventions used in WILMA were very comprehensive and substantially more intensive than the intervention investigated in this study.

The intervention tested here is most aligned to the intervention tested by Sidhu et al. 17 in that both were brief and offered telephone support and SMS text messages. The trials differed in that the intervention tested by Sidhu et al. 17 encouraged weekly weighing, whereas this study encouraged daily weighing. Both trials showed good adherence to daily or weekly weighing, but both interventions were ineffective in maintaining weight loss, highlighting that more intensive intervention may be required.

Consonant with the findings of systematic reviews of trials35 assessing the harms of self-weighing, we found no evidence to indicate that encouraging people to weigh themselves each day resulted in psychological harms or concerns about body image. Systematic reviews of weight management interventions have also commented that past studies have tended not to be based on behavioural theory; here, we tested an intervention that was embedded within self-regulation theory, the relapse prevention model and habit formation theory. Reviews have reported that previous trials of weight maintenance have contained methodological flaws, such as inadequate reporting of the randomisation process, lack of blinding of the primary outcome, inadequate statistical power or high rates of loss to follow-up, which could introduce bias. 14 This trial was designed not to have these limitations, to ensure that the findings were robust. We are confident that the null findings in this study do not result from limitations in the trial design or conduct. The results of this trial might also raise questions about the usefulness of self-regulation theory as a guiding theory in the development of weight maintenance interventions.

In this study, most people achieved weight maintenance at 3 months by not gaining more than 1 kg of their baseline weight (≈80%), and at 12 months approximately 45% of participants (intervention, 44.7%; usual care, 45.9%) had maintained their weight. In the WILMA trial,16 a binary assessment of whether or not participants had gained any weight (weight maintenance) between baseline and 12 months showed that about 54% had not done so (control, 54%; less intensive, 63%; intensive, 44%). The pragmatic nature of the WILMA trial meant that participants were free to use other weight management strategies during the intervention. Data on the numbers still attending their weight loss programme at follow-up are not available for the WILMA study, but in this trial, at the 3-month follow-up, 62.3% of participants were still attending their weight loss programme, and at 12 months 41.4% were still attending. Although the majority of participants were from the lowest deprivation quartiles, many continued to attend a weight loss programme at their own expense. Taken together, these data may suggest that people can achieve further weight loss or weight maintenance over time if they continue to attend their weight loss programme. Furthermore, this might also suggest that what is needed is not a greater emphasis on weight maintenance interventions per se, but that members of the public be encouraged to keep attending weight loss programmes, or to reattend a weight loss programme periodically if they regain some of the weight that they had lost. A recent systematic review13 of health strategies for successful weight loss and weight maintenance concluded that, to achieve weight loss maintenance, the strategies that had led to initial weight loss had to be adhered to over the longer term. The results from the recent WRAP trial (Weight loss Referrals for Adults in Primary care) are relevant here. 63 WRAP showed that, at the 2-year follow-up, attendance at a commercial weight loss programme for 52 weeks produced greater weight loss and other clinical benefits than attendance for 12 weeks. However, attending for a whole year led to no better weight loss maintenance than attending for 12 weeks. The rate of regain was the same on ceasing the programme whether it had lasted 12 weeks or 52 weeks.

Recommendations for future research

In this study, regular self-weighing was ineffective in promoting cognitive restraint of eating. Studies that aim to test other strategies to engage people in long-term conscious cognitive control over energy balance would be very worthwhile. Participants were recruited immediately after completing a weight loss programme and many continued to attend their weight loss programme. Data show that many participants were focused on continuing to lose weight, rather than maintaining the weight they had lost thus far. This raises questions for future research: when is the best time to intervene to assist participants with weight loss maintenance, and, more specifically, should weight loss programmes include a strong(er) emphasis on weight maintenance from the outset, or at least greater emphasis on explaining that continued weight loss maintenance will take considerable effort? Research that investigates whether or not participants are ready for, or accepting of, a weight maintenance approach following weight loss would also be worthwhile. Given that most participants indicated that their goal was to achieve additional weight loss, it might have been the case that our weight maintenance intervention did not meet the psychological needs of participants, and future exploratory research around this question would be useful. Additionally, the focus of participants to continue to lose weight could highlight the importance of managing individuals’ expectations of weight management. This may be an interesting area for future research to explore.

Preliminary evidence from the WILMA trial suggests that more intensive interventions that deliver and encourage the use of multiple weight loss maintenance behavioural skills may be more effective than the brief intervention tested here. 16 However, WILMA failed to recruit participants, and it is possible that intensive interventions of this kind are impractical to deliver at scale and unpopular with potential users. Perhaps brief interventions for weight loss maintenance may be ineffective because maintaining weight loss is a complex behaviour that, to be successful, requires the enactment of multiple behavioural skills. Research that focuses on evaluating other approaches to weight loss maintenance are also needed. Currently, there is no clear evidence about the most effective intensity, mode of delivery or behaviour change techniques for facilitating weight loss maintenance, and this should be a priority area for future research.

Research that evaluates interventions focused on changing policy and the environment (particularly food environments), and their impact on weight loss maintenance, may also be worthwhile. It is also essential that future studies are based on sound theoretical frameworks. Given that weight loss maintenance is a complex behaviour, this field of research would benefit from more qualitative work to explore how behaviour change techniques are perceived and then operationalised by participants to bring about weight loss maintenance. Once effective interventions are identified, cost-effectiveness analyses are essential.

In this trial, a successful maintainer was defined as a participant who regained ≤ 1 kg, but across studies definitions of weight loss maintenance vary; for example, some researchers have used the cut-off point of no weight gain at 12 months to define succes,16 whereas others19 have used weight gain of no more than 1.36 kg (3 pounds) or maintenance of 5% weight loss at 12 months. 13 It is also important that researchers move towards a common and agreed definition of what constitutes weight loss maintenance to enable meaningful comparisons between interventions. In recent years, the delivery of behavioural weight management interventions has moved from mostly face-to-face contact interventions to technology- or remote-based interventions. Future research should explore what contribution technological interventions might make to the question of weight loss maintenance, either as stand-alone interventions or as part of face-to-face interventions.

Conclusions

A brief telephone intervention delivered by non-specialist workers from a community interest organisation that promoted target-setting, regular self-weighing and recording of weight to encourage reflection on weight gain progress was ineffective in preventing weight regain after intentional weight loss. Encouraging participants to weight themselves every day did not cause psychological harm.

Contributions of authors

Dr Amanda Daley (Reader, Behavioural Medicine) conceived the idea for the study and led the design and implementation of the study. She led the writing of the report.

Professor Kate Jolly (Professor, Public Health and Primary care) contributed to study and intervention design. She contributed to writing the report.

Dr Claire Madigan (Trial Manager, Primary Care) contributed to study design, intervention development and implementation of the study. She contributed to writing the report. She was employed to manage the trial as a research fellow in year 1.

Ryan Griffin (Research Associate, Primary Care) managed the trial in years 2 and 3 and contributed to drafting the report.

Andrea Roalfe (Senior Lecturer, Medical Statistics) contributed to study design, calculated the sample size, created the randomisation lists, contributed to the development of the statistical analysis plan and commented on a draft of the report.

Dr Amanda Lewis (Trial Manager, Social and Community Medicine) contributed to the study design. She commented on a draft of the report.

Alecia Nickless (Trial Statistician, Primary Care) conducted the statistical analyses and contributed to drafting the report.

Professor Paul Aveyard (Professor, Behavioural Medicine) contributed to the design of the study, intervention development and implementation of the study. He contributed to writing the report.

Data-sharing statement

We shall make data available to the scientific community with as few restrictions as feasible, while retaining exclusive use until the publication of major outputs. Data can be obtained from the University of Oxford Primary Care Clinical Trials Unit.

Disclaimers

This report presents independent research funded by the National Institute for Health Research (NIHR). The views and opinions expressed by authors in this publication are those of the authors and do not necessarily reflect those of the NHS, the NIHR, NETSCC, the PHR programme or the Department of Health and Social Care. If there are verbatim quotations included in this publication the views and opinions expressed by the interviewees are those of the interviewees and do not necessarily reflect those of the authors, those of the NHS, the NIHR, NETSCC, the PHR programme or the Department of Health and Social Care.

Birmingham Gateway Lighten Up adult weight management programme entry criteria

• Have a raised BMI, defined as a BMI of ≥ 25 kg/m² for South Asian people with no comorbidities or a BMI of ≥ 23 kg/m² for people with comorbidities.

• All other ethnic groups are eligible if they have a BMI of ≥ 30 kg/m² with no comorbidities or a BMI of ≥ 28 kg/m² with comorbidities.

Solihull weight management programme entry criteria

• Have a BMI of ≥ 30 kg/m².

• Be aged ≥ 16 years.

• Live in Solihull or have a Solihull GP.

Dudley weight management service entry criteria