In one of its main areas of inquiry, nutrition science is stuck in a rut. Weight loss research has become Einstein’s definition of insanity; repeat the same thing over and over and expect different results each time.
Diet A vs. Diet B. Low-carb vs. low-fat. Control protein. 5-7% target bodyweight loss to be “significant”. And one thing we know: it doesn’t matter. Diet A, B, C, D, or Y and Z: they all work. Except they don’t. Results are published as “significant” despite the fact that the true significance in what we have learnt from weight loss research is the paucity of evidence for long-term maintenance. We’re all familiar with the statistics by now: up to 35% of weight lost in an intervention will be regained after 1-year, and by 5-years practically all – or a majority – of weight lost will have been regained .
Which brings us to the subject of this article: is there any evidence for long-term maintenance of weight loss? What is the difference in the types of intervention that are successful in achieving long-term maintenance? What are the characteristics of successful participants? And is there anything we can glean from this limited data set to inform where investigations might look further?
This article will attempt to examine these questions.
Evidence for Long-term Weight Loss Success
Evidence for long-term maintenance is not absent the literature, but it is scarce. Two factors which should be considered is the definition of ‘successful weight loss’, and definition of ‘long-term’. The framing of these two factors may often lead to a misleading interpretation of the data. ‘Success’ is typically defined as a 5-10% loss of initial bodyweight; these targets are not necessarily arbitrary, and are established by reference to clinically significant reductions in risk factors for type-2 diabetes and heart disease that are also considered achievable for the individual [this latter premise will be discussed further below]. 
‘Long-term success’ is often defined as maintenance of lost weight for 1-year . This is potentially misleading for two reasons. The first is the misrepresentation of the time-course of weight regain, given that weight regain over 1-year is highly predictive of continuing weight regain beyond that timepoint . Using 1-year as a time-point thus misrepresents the definition of ‘successful’ weight maintainers. The second, and more important factor for our purposes, is that the crucial period for maintaining weight loss over the long-term also appears to be the period going beyond 2 years and up to 4-5 years . In analysis comparing participants in a weight loss trial whom maintained weight loss beyond 1-year against participants who regained weight, the single greatest predictor of weight regain was duration of weight maintenance: successfully maintaining weight lost beyond 2-years exponentially increased the likelihood of success, with a 50% reduction in risk for weight regain between 2-years and 4-years . Maintenance over 5-years reduced risk of regain by 71% . The picture that emerges is that duration of time spent in maintenance is a critical factor. Where weight has already started to be regained within a 1 to 2-year period, the odds of weight regain continuing in an almost linear fashion are significant .
This trend toward almost immediate weight regain must also be considered in the context of absolute weight lost. Indeed, one of the observations from diabetes literature is that there may be a legacy benefit to weight loss, even if a significant proportion of initial weight loss is regained . In the Look AHEAD trial, a diabetes prevention lifestyle intervention program, mean weight loss in the intensive intervention group was 8.5% bodyweight after 1 year, and 4.7% at year 4 . However, by year 8 of the program there had been a stabilising effect on weight in the intensive intervention group, with mean weight loss remaining on average 4-4.7% between years 4 and 8 . Scrutinising the means closer, it emerged that subjects whom lost >10% of bodyweight during the initial 1-year intervention were significantly more likely to maintain between 5-10% weight loss at 4-years: 42% had maintained 10% of weight loss [a staggering number relative to the wider literature], while 70% had maintained >5% loss . At year 8 of the intervention, those numbers remained 39% and 65%, respectively, rendering this trial the single most successful long-term intervention of its kind .
Two observations emerge from scrutinizing the evidence for long-term weight loss maintenance. The first is that the period between 2 and 5-years post weight loss appears to be critical, with duration of maintenance a key determining factor increasing the odds of maintaining significantly beyond two years and linearly improving up to and beyond 5-years . In this context, 5-years is a more appropriate time point to consider weight loss maintenance ‘successful’. The second is that we need to consider both the absolute level of initial weight loss and weight regain, within the context of these timeframes. One possibility is that greater initial weight loss may increase the likelihood of successful maintenance over the long-term; the other is that greater initial weight loss may mean that a certain degree of weight regain still results in a net benefit over the long-term .
Does Rate of Weight Loss Influence Long-term Success?
This question has been one of the few areas of weight loss research opened to further scrutiny. The effects of rapid weight loss vs. gradual weight loss on long-term maintenance is not a new observation, but the historical ideology emphasized that rapid weight loss led to rapid weight regain . It is important to distinguish whether rate of initial weight loss influences weight regain from whether greater initial weight loss improves likelihood of long-term maintenance. In relation to whether rate of initial weight loss influences weight regain, this has not been supported by the evidence. Rather, the overall body of research shows that the rate of weight loss has no impact on weight regain, i.e., weight is ultimately regained over the long-term whether initial weight loss is gradual or rapid . In this respect, weight regain remains the status quo overall.
However, recently it has been suggested that greater initial weight loss may increase the chances of long-term maintenance, or result in less absolute weight regained over time. For example, in a 2010 study by Nackers et al., analysis of rate of weight loss indicated that subjects who lost 13.5kg in 6-months were 5 times more likely to retain >10% weight loss at 18-months [maintaining 10.9kg loss], compared to those who lost only 5.1kg at 6-months [who maintained 3.7kg loss] . This latter point illustrates where statistics may not necessarily give an accurate portrayal of the total picture. In relation to weight loss research, we have the well-known statistics that, for example “35% of weight will be regained after 1-year”. But taking the slow weight loss group referenced above, a 5.1kg weight loss with a 1.3kg weight regain is around 30%: but it is more a reflection of the relatively small amount of absolute weight lost in the first place. If we take the average weight loss in trials over 6-months of 5kg to 8.5kg, and average maintenance of 3kg to 6kg over 48 months , that can easily become a statistic of anywhere between “66% to 100% of weight lost will be regained”. This is not discounting the stats: we know that a majority of weight loss fail over the long-term. What this is illustrating is that what the statistics are hiding is how little absolute weight is often lost in the first place. And this is another crucial point in the wider conversation, reflecting the dogma of the paradigm of weight loss research: a majority of the interventions used are plain awful.
Having regard to this fact, that the average initial weight loss is often so minor such that any weight regain as a percentage appears quite substantial, it is not surprising that there has been a focus on whether greater initial weight loss may increase successful maintenance. These suggestions have emerged from a number of studies, including the Nackers et al. study above, however, it is important to note that these studies were all post-hoc, secondary analyses of weight loss trials . The significance of this is that a post-hoc analysis is an analysis that occurs after the fact, i.e., it is based on data from a trial in which the question being examined by the post-hoc analysis was not an outcome of the original trial. This means a post-hoc analysis is essentially an observational study borne out of an intervention, and is not directly experimental. This means that the studies often cited in support of the contention that greater initial weight loss improves long-term maintenance cannot demonstrate whether greater initial weight loss causes greater maintenance . Instead, we have to look to interventions which have directly tested this hypothesis.
An early study to test the difference between rate of weight loss on long-term maintenance by Toubro and Astrup randomised 43 adults with obesity to achieve the same total weight loss: one group used a very low-calorie diet [‘VLCD’] for 8-weeks compared to a group using a conventional diet for 17-weeks . Both groups lost 13.6kg in their respective timeframes, thus the VLCD lost 1.6kg/week vs. 0.8kg in the conventional group. After the weight loss phase, subjects were re-randomised to either an ad libitum low-fat, high-carbohydrate diet or a fixed energy restricted diet, and both groups received intensive follow-up support. At 2-years follow-up, the fixed energy restriction group had regained 11.3kg, while the ad lib group had regained 5.4kg. Thus, while rate of initial weight loss had no influence of maintenance at 2-years, what this study highlighted was that dietary characteristics coupled with behaviour therapy did have a significant positive effect, with the ad lib group retaining 60% of initial weight loss.
Another recent RCT by Purcell et al. also directly tested whether greater initial weight loss improved long-term maintenance outcomes. 200 subjects were randomly assigned to achieve 12.5% bodyweight loss either in 12-weeks or 36-weeks ; the 12-week rapid group consumed an 800kcal/d VLCD, while the 36-week gradual group consumed a conventional 500kcal/d fixed energy restricted diet. After the weight loss phase in which subjects had lost 14.6kg and 14.3kg in the rapid and gradual groups, respectively, subjects then entered a 144-week maintenance phase: the entire study duration was thus 3.5 years. However, almost all subjects began to regain weight during the maintenance phase; by week 48 all participants had regained a similar amount of weight, and experienced linear weight regain until the study’s end, irrespective of weight loss group. Average weight regain at 144-weeks was 10.3kg in the rapid group and 10.4kg in the gradual group.
What comparisons can we make between the latter trial and successful trials? A number of factors appear pertinent. First, in the Purcell et al. trial the attempts at maintenance centered on deliberate attempts to control/restrict energy intake and prevent weight regain [ironic perhaps, this will be discussed further below]. Secondly, there was no emphasis on behaviours in this trial, and there was weak intensity of follow-up. In contrast, the Look AHEAD trial was an intensive lifestyle program encompassing weekly behaviour treatment sessions for the first year, followed by monthly thereafter . The trial by Toubro et al. had similar intensity of behaviour therapy over the course of the follow-up period . And, in analysis of trials going back to the 1960’s, passive follow-up was consistently associated with lower success rates than active follow-up; active follow-up combined with behaviour therapy improved outcomes even further . In person follow-up appears to be a key determinant, as the Weight Loss Maintenance Randomised Controlled Trial, which compared maintenance between personal follow-up, interactive Web-based follow-up, and self-directed follow-up, found that while the Web-based follow-up had early benefit, this waned over the course of 30-months; the personal contacts on a monthly basis yielded the best outcomes .
We can make a number of conclusions from the foregoing. First, despite popular citing in the fitness industry of studies indicating a benefit to rapid weight loss, these studies are primarily post-hoc analyses, and do not demonstrate cause-effect . The studies that have directly tested this hypothesis experimentally have not found any effect – positive or negative – on long-term weight loss maintenance . Secondly, when we do scrutinize the factors that do seem to influence long-term maintenance it becomes evident that behaviors are what dictate outcomes (7,9,10,14,18). This wouldn’t be surprising to anyone familiar with the psychology literature on implementation intentions and goal-setting, but it appears to be lost on much of the wider nutrition research community. Thus, let’s scrutinize in more detail the characteristics of successful long-term maintenance of weight loss.
Characteristics of Successful Weight Loss Maintenance
There are a number of consistent characteristics that emerge when we look closer at the factors identified in the literature. The first and most obvious relates to the intensity of the intervention itself, i.e., the degree of practitioner support, the proximity and type of contact that participants are given not only during the intervention, but during the crucial period following weight loss. A contrast between two studies – one highly successful in achieving long-term maintenance, the other highly abysmal in its failure to do so – illustrates this point. The first is the aforementioned Look AHEAD Study, which assessed the difference between standard general advice for diabetes and an intensive lifestyle intervention, over 8-years. The general advice group had three 1-hour meetings a year from Year 1 to 4 with general advice on diet, lifestyle, and physical activity education, which reduced to 1 meeting per year for Years 5 to 8; the results in this group were underwhelming . However, in the intensive support group in the Look AHEAD Study, participants received:
Small groups of 10-20 which met 3 times per month for the first 6-months; 2 times per month thereafter to 1-year; then 1 time per month up to Year 8.
In addition to these monthly group sessions, subjects had individual sessions with an ‘interventionist’ (which included registered dietitians, psychologists, and exercise specialists) every month in person, with a second contact by phone or email 2-weeks thereafter.
The interventionists were counseled to factor in cultural differences, and employ different modalities including motivational interviewing, self-regulation theory, relapse prevention, and problem solving.
The results, highlighted above, remain the most impressive long-term weight loss maintenance results published to date: of participants who lost >10% of bodyweight during the initial 1-year intervention 39% and 65% maintained 10% or 5%, respectively, of weight lost at Year 8.
Now let’s look in contrast at another previously mentioned trial which compared the effects of rapid weight loss vs. gradual weight loss on long-term maintenance over 3-years. Subjects met with a dietitian every 2-weeks during the weight loss phases: 6 contacts for the rapid group [12-weeks] and 18 for the gradual group [36-weeks] . During the maintenance phase that lasted 144-weeks in total, all participants met with a dietitian at 4-weeks, 12-weeks, and then every 3-months thereafter until the end of the study. Practically all participants, irrespective of rate of initial weight loss, had regained a majority of weight lost by week 48 of maintenance. The contrast between this trial and the Look AHEAD trial is instructive: Look AHEAD was multimodal and had a high level and frequency of practitioner contacts, while this trial had no behavioural targets and sparse practitioner contact for diet advice alone. Number of contacts emerges as a pivotal factor, with increasing contacts consistently associated with improved outcomes, effects which also appear to be more pronounced with in person, face-to-face contacts . In the Look AHEAD Study, those who had maintained >10% weight loss over 8-years were those who had significantly more recorded practitioner contacts than those who regained their baseline weight .
It is important to look beyond the contacts per se to the purpose which they serve: behaviour modification. Ultimately when we dig into the research to distinguish someone who is successful in maintenance over 2-years from those who begin to regain weight almost immediately following the end of an intervention, one word emerges from the literature: behaviours. In the Lean Habits Study, Westenhoefer et al. quantified 8 behavioural characteristics at baseline: flexible vs. rigid dietary control, meal regularity, eating situations [i.e., sitting down, undistracted, taking time], food choice, restriction of food quantity, physical activity, and stress coping . At 3-years follow-up, those participants whom had made between 5 to 8 behavioural improvements had the highest percentage of successful weight loss maintenance. The only two behaviours not associated with any increased probability of successful maintenance of rigid control of food intake and restriction of food quantity. This is consistent with the wider literature which implicates both of these behaviours in disinhibited eating, and studies which quantify eating behaviours at baseline indicate that higher levels of dietary disinhibition and eating restraint predict weight regain.  This may relate to dichotomous thinking about food in ‘good or bad’, ‘black or white’ terms, which is strongly related to rigid control of food intake and predictive of weight regain . In this regard, dichotomous thinking is itself a ‘cognitive dysfunctionality’, and it has been suggested that dichotomous thinking and food and dieting, rather than dietary restraint per se, is the underlying factor predicting weight regain .
In terms of adverse behavioural effects that may be observed in diet studies, researchers that have quantified baseline behavioural characteristics have found that subjects with higher disinhibited eating, restrained eating, depression and binge eating, and number of previous attempts at intentional weight loss, are all strongly predictive of weight regain . The crucial distinction here is that these behaviours appear to be present prior to a weight loss intervention. Given a majority of the research in this area does not screen for such behavioural characteristics prior to an intervention, it is difficult to tease out whether adverse behavioural factors are a direct result of the intervention. It is arguably a case of chicken and egg, but the studies that do quantify potentially problematic behaviours that could predispose participants to rebound weight gain indicate that such characteristics are often present before the intervention.  A crucial point to make here is that we have known such characteristics predict weight regain for over 20-years. The failure to implement this knowledge in trial designs and exclude at-risk participants amounts to negligence in standards of care.
What behaviours, other than frequency of contacts, are positively associated with weight loss maintenance? Taking the literature as a whole, factors including increasing physical activity [including less screen time], fewer intentional weight loss episodes, greater time spent in maintenance, less dietary restraint, and frequency of self-weighing.  Physical activity levels are perhaps the strongest and most consistent predictor of successful maintenance (10,18,19,20). Time spent in maintenance is a consistent factor in dictating successful long-term maintenance.  This is important to distinguish from the trends that characterise weight regain, which is almost immediate relapse with no weight stabilisation: the magnitude of weight regained at the end of 1-year strongly predicts outcomes at 2-years . This is the period in which the current dogma of weight loss research is failing people, and one has to seriously question the purpose of diet trials that consider “success” as weight loss at 6-months or 12-months. In relation to fewer intentional weight loss episodes being associated with long-term maintenance, this is simply another way of saying that increasing number of deliberate weight loss attempts are associated with weight regain . Repeated, deliberate dieting attempts may reflect lower self-esteem, which has been associated with unhealthy weight control behaviours . This may be a domino effect from the negative effects of internalised weight stigma, a factor overlooked in much of the literature but which is strongly predictive of adverse behavioural effects associated with dieting .
Given the recent focus on perceived negative behaviours, perhaps the most controversial behaviour that is consistently positively associated with weight loss maintenance is frequency of self-weighing.  How do we reconcile this against the suggestions that such behaviours are associated with disordered body image and disordered eating? The first potential explanation is to recall that studies that do assess potentially problematic behaviours at baseline have found that such behaviours are predisposing factors, i.e., they are present prior to the intervention.  It is arguable that in these predisposed person’s, behaviours such as self-weighing may have a negative effect on psychological well-being. This would be supported by research finding poor self-esteem associated with weight-controlling behaviours.  These factors are important to tease out against positive psychological predictors of success which have been identified: higher self-motivation, which is consistent with the inclusion of motivational interviewing strategies employed by the Look AHEAD trial, self-efficacy, and locus of control.  It could be argued that individuals for whom frequent self-weighing is an effective behaviour tool for long-term maintenance may exhibit higher levels of these characteristics, rendering them less likely to experience it as a negative, ‘triggering’ behaviour. But we have no data on this, so it is speculative. One retrospective analysis found that weight regain was associated with a tendency to evaluate self-worth in terms of weight and body shape.  However, it is clear that regular self-monitoring of weight does not predispose all individuals to rebound weight gain, evident in the fact that many are successful over periods of up to nearly a decade. 
Taken as a whole, it is not surprising that there is an overall lack of quality evidence for long-term maintenance given the level of behaviour change that is required, and the number of contacts, particularly in-person contacts, required to increase the odds of success. However, there is at least some evidence for long-term maintenance, and those nuances warrant further investigation in a more pragmatic RCT design model, particularly for management and prevention of Type-2 Diabetes, and Fatty Liver Disease, the two most problematic chronic lifestyle conditions associated with adiposity.
Ultimately, the literature on psychological predictors of weight loss is inconsistent . It does appear that weight regain may be predicted by high levels of problematic behaviours at baseline, indicating that many participants are already experiencing difficulties with diet and/or body image prior to the intervention.  That these factors are not assessed as part of a trial’s inclusion/exclusion criteria is a gross failing of the research community. What we do see from the characteristics of successful long-term weight loss maintenance is that the interventions are labour intensive, involving active follow-up, with more positive outcomes from in-person contacts, and overall more total contacts and frequency of contacts improving outcomes.  An emphasis on behaviours, as opposed to deliberating attempting to solely focus on and prevent weight regain, also emerges as an important determinant.  However, it is not surprising given the level of follow-up intensity, and the fact that behaviour change is difficult due to the ingrained nature of habits, that long-term weight loss maintenance is so difficult for many people.
- Weiss, E. C., Galuska, D. A., Kettel Khan, L., Gillespie, C., & Serdula, M. K. Weight Regain in U.S. Adults Who Experienced Substantial Weight Loss, 1999-2002. American Journal of Preventive Medicine. 2007;33(1):34–40.
- Hamman, R., Wing, R., Edelstein, S., Lachin, J., Bray, G., Delahanty, L., Hoskin, M., Kriska, A., Mayer-Davis, E., Pi-Sunyer, X., Regensteiner, J., Venditti, B. and Wylie-Rosett, J. Effect of Weight Loss With Lifestyle Intervention on Risk of Diabetes. Diabetes Care. 2006;29(9):2102-2107.
- Lindstrom, J. (2003). Prevention of Diabetes Mellitus in Subjects with Impaired Glucose Tolerance in the Finnish Diabetes Prevention Study: Results From a Randomized Clinical Trial. Journal of the American Society of Nephrology. 2003;14(90002):108S-113.
- The Look AHEAD Research Group. (2006). The Look AHEAD Study: A Description of the Lifestyle Intervention and the Evidence Supporting It. Obesity. 2006;14(5):737–752.
- Lowndes, J., Kawiecki, D., Pardo, S., Nguyen, V., Melanson, K., Yu, Z. and Rippe, J. The effects of four hypocaloric diets containing different levels of sucrose or high fructose corn syrup on weight loss and related parameters. Nutrition Journal. 2012;11(1).
- Wing RR, Hill JO. Successful weight loss maintenance. Annu Rev Nutr 2001;21:323–41.
- Wing RR, Phelan S. Long-term weight loss maintenance. Am J Clin Nutr. 2005;82(1):222S–225S.
- Mcguire MT, Wing RR, Klem ML, Lang W, Hill JO. What Predicts Weight Regain in a Group of Successful Weight Losers ? 1999;67(2):177–85.
- Wadden TA, Neiberg RH, Wing RR, Clark JM, Delahanty LM, Hill JO, et al. Four-year weight losses in the look AHEAD study: Factors associated with long-term success. Obesity [Internet]. 2011;19(10):1987–98.
- The Look AHEAD Research Group. Eight-Year Weight Losses with an Intensive Lifestyle Intervention: The Look AHEAD Study. Obesity. 2014;22(1):5–13.
- Dave R, Davis R, Davies JS. The impact of multiple lifestyle interventions on remission of type 2 diabetes mellitus within a clinical setting. Obes Med. 2019;13(October 2017):59–64.
- Casazza K, Fontaine KR, Astrup A, Birch LL, Brown AW, Bohan Brown MM, et al. Myths, Presumptions, and Facts about Obesity. N Engl J Med. 2013;368(5):446–54
- Astrup A, Rossner S. Lessons from obesity management programmes: greater initial weight loss improves long-term maintenance. Int J Dermatol. 2000;53(8):952–5.
- Toubro S, Astrup A. Randomised comparison of diets for maintaining obese subjects’ weight after major weight loss: ad lib, low fat, high carbohydrate diet v fixed energy intake. Br Med J. 1997;314(7073):13–8.
- Nackers LM, Perri MG. The Association Between Rate of Initial Weight Loss and Long-Term Success in Obesity Treatment: Does Slow and Steady Win the Race? Int J Behav Med. 2010;17(3):161–7.
- Purcell K, Sumithran P, Prendergast LA, Bouniu CJ, Delbridge E, Proietto J. The effect of rate of weight loss on long-term weight management: A randomised controlled trial. Lancet Diabetes Endocrinol. 2014;2(12):954–62.
- Caplan W, Bowman JD, Pronk NP. Weight-Loss Outcomes: A Systematic Review and Meta-Analysis of Weight-Loss Clinical Trials with a Minimum 1-Year Follow-Up. 2007;67(2):177-185.
- Ayyad C, Andersen T. Long-term efficacy of dietary treatment of obesity: a systematic review of studies published between 1931 and 1999. Obes Rev. 2000;1:113–9.
- Svetkey L, Stevens VJ, Brantley PJ, Appel LJ, Hollis JF, Loria CM, et al. Comparison of Strategies for Sustaining Weight Loss. JAMA. 2008;299(10):1139–48.
- Wing RR, Papandonatos G, Fava JL, Gorin AA, Phelan S, Mccaffery J, et al. Maintaining Large Weight Losses : The Role of Behavioral and Psychological Factors. 2008;76(6):1015–21.
- Wadden TA, Foster GD, Letizia KA. One-Year Behavioral Treatment of Obesity : Comparison of Moderate and Severe Caloric Restriction and the Effects of Weight Maintenance Therapy. 1994;62(1):165–71.
- Westenhoefer J, Falck B Von, Stellfeldt A, Fintelmann S. Behavioural correlates of successful weight reduction over 3 y . Results from the Lean Habits Study. 2004;334–5.
- Palascha A, Van Kleef E, Van Trijp HCM. How does thinking in Black and White terms relate to eating behavior and weight regain? J Health Psychol. 2015;20(5):638–48.
- Daniali, S. Azadbakht, L. Mostafavi, F. Relationship between body satisfaction with self-esteem and unhealthy body weight management. J Educ Health Promot. 2013;2:29.
- Pearl RL, Wadden TA, Hopkins CM, Shaw JA, Hayes MR, Bakizada ZM, et al. Association between weight bias internalization and metabolic syndrome among treatment-seeking individuals with obesity. Obesity. 2017;25(2):317–22.
- Lazzeretti L. Assessment of psychological predictors of weight loss: How and what for? World J Psychiatry [Internet]. 2015;5(1):56.
- Byrne S, Cooper Z, Fairburn C. Weight maintenance and relapse in obesity: a qualitative study. Int J Obes Relat Metab Disord 2003;27:955-962.