Weight cycling in a very low-calorie diet programme has no effect on weight loss velocity, blood pressure and serum lipid profile

Transcription

1 ORIGINAL ARTICLE doi: /j x Weight cycling in a very low-calorie diet programme has no effect on weight loss velocity, blood pressure and serum lipid profile Z. Li, K. Hong, E. Wong, M. Maxwell* and D. Heber David Geffen School of Medicine at University of California, Los Angeles, UCLA Center for Human Nutrition, Los Angeles, CA, USA Background: Many dieters lose and regain weight many times. It is unclear whether weight cycling is associated with adverse metabolic alterations or becomes more difficult with each attempt. Methods: From 1988 to 2000, 2474 obese patients enrolled in our outpatient weight loss programme using a very low-calorie diet. Caloric intake consisted of meal replacement supplying cal/day. Results: Our search distinguished 480 patients who had restarted the weight reduction programme at least once and up to four times (restarts). All subjects remained on the programme for 2 weeks or more each time. Mean initial weight loss was 21.3 kg for women and 28.8 kg for men. Rate of weight loss on first restart was not different from initial weight loss for women [1.6 vs. 1.4 kg/week; not significant (NS)] or for men (2.2 vs. 2.1 kg/week; NS). Of the 480 patients, 85 women and 51 men entered the programme three times. Rate of weight loss was similar for all three entries (1.4, 1.6 and 2.1 kg/week for first, second and third entry in women, NS; 2.1, 2.1 and 2 kg/week for men, NS). Only 20 women and 18 men entered the programme four times. The rate of weight loss was again similar for both men (NS) and women (NS) during each re-entry. Cardiovascular risk factors including lipid profile and blood pressure were not adversely affected by weight cycling. In fact, lipid levels were lower at each restart. Conclusions: The present study refutes the hypothesis that repeated dieting makes further dieting efforts more difficult. Keywords: VLCD, weight cycling Received 9 March 2005; returned for revision 28 March 2006; revised version accepted 28 March 2006 Introduction Obesity has become epidemic in the United States, with two of three Americans being either obese or overweight [1]. It has been controversial whether weight loss, per se, is an appropriate recommendation for all overweight individuals. Although weight loss improves health risk factors, the majority of people who lose weight regain it [2,3] with associated worsening of health risk measures [4,5]. A clinical issue of utmost importance is whether it is better to have lost and regained than never to have lost at all. Several recent studies have suggested that weight cycling can have negative health consequences, with increased total mortality and, specifically, with increased cardiovascular mortality and morbidity due to coronary heart disease [6 9]. However, conclusions from these epidemiological studies are limited by the inability to distinguish voluntary from involuntary weight change and by the lack of data regarding plausible biological mechanisms underlying these negative effects [10 12]. The present study compares repetitive treatment outcomes in a large group of 480 patients who participated in the UCLA Risk Factor Obesity (RFO) weight reduction Correspondence: Dr Zhaoping Li, David Geffen School of Medicine at University of California, Los Angeles, UCLA Center for Human Nutrition, 900 Veteran Avenue, , Los Angeles, CA 90095, USA. zli@mednet.ucla.edu *Deceased Diabetes, Obesity and Metabolism, 9, 2007, j 379

2 OA j Effect of weight cycling Z. Li et al. programme for up to four times, having regained near or all of their initial weight loss between treatments. Subjects and Methods Subjects This is a retrospective chart review of subjects who participated in weight loss programme at the University Obesity Center between 1988 and The study was approved by the Institutional Review Board at the University of California, Los Angeles. Over a period of 12 years, 2474 obese patients participated in a multidisciplinary weight loss programme using a very low-calorie diet (VLCD) at the UCLA RFO programme. All study patients were prescribed a VLCD, an exercise regimen, and participated in group classes on behavioural modification. Caloric intake consisted of a commercially prepared meal replacement powder supplying cal/day. Each formula packet provided 100 cal and 15 g of high biological value protein, as well as allowance for required minerals and vitamins. Patients who were unsuccessful in their initial treatment programme or who regained weight were permitted to re-enter the programme (restarts). The restart patients were combined with groups of new patients and participated in the same programme of caloric restriction and educational and behavioural instruction. Dieticians and psychologists were available for consultation at each visit. For exercise, all patients were instructed to have 30 min of aerobic exercise combined with min of light-weight training three times per week. Patients attended the clinic weekly until they reached their target weight or until they voluntarily discontinued the programme. All patient data were programmed into the UCLA mainframe computer, stored and upgraded periodically. Original patient charts were available for a detailed review, if indicated. A computer search distinguished 480 patients who had restarted the weight reduction programme at least once and remained on the programme 2 weeks or more each time. All these restart patients were included in the present study. Statistical Analysis The principle investigator had full access to all the data in the study and was responsible for the integrity of the data and the accuracy of the data analysis. The data of this study were examined using analysis of variance, specifically utilizing a repeated measurement model. This model enabled us to separate the large variability due to differences between individual patients from the treatment effect, which is the outcome examined. Each subject served as his/her own control, increasing the sensitivity of the analysis. Pair-wise t-tests, utilizing Fisher s least significant difference, were used for further comparison of means for entries in the programme. The standard errors reported are based on the variance among patients. For this study a p < 0.01 was considered significant. The significance of percentages in the various crosstabulations was evaluated using a chi-square statistic for tests of independence or a z-statistic as an approximation to the binomial distribution. Independent t-tests were used for comparison of sample means where appropriate. All calculations were made using the STATISTICAL ANALYSIS SYSTEM (SAS, Cary, NC, USA). Results Four hundred and eighty (19.4%) of the 2474 patients who participated in the RFO programme restarted the programme at least once. For all patients, the mean weight loss during the initial course of treatment was kg, with a range of 3 88 kg. The median duration of the initial weight loss period was 17 weeks. Table 1 shows the comparison of the patients with multiple restarts (multiple starts, MS) with the patients who entered the programme only once (single start, SS). It can be seen in table 1(A) that baseline demographics were similar in both groups, except that a larger proportion of men volunteered to restart the programme as compared to women. Table 1(B) shows that MS patients were more successful in their initial weight loss than SS patients. This corresponded to a longer duration of initial programme participation. A greater success rate, defined as weight loss greater than 15 kg, was also observed for MS patients than for SS patients. The rate of weight loss was similar in both groups. In the MS group, the median interval between initial treatment and the first restart was 68 weeks. Patients had lower baseline weights at the first restart than when they initially entered the programme; the mean weight difference was 8.2 kg. There was considerable scatter, however, and 25% of the patients in the restart group were at least 0.5 kg above their initial baseline weight at the first restart. A comparison of initial vs. first restart treatment outcomes, using the criteria of greater or less than 15 -kg weight loss to define success or failure respectively, is presented in tables 2 and 3. It should be noted that of the 69 patients who had failed initially (<15 -kg weight loss), 15 patients were able to succeed on re-entry 380 j Diabetes, Obesity and Metabolism, 9, 2007,

3 Z. Li et al. Effect of weight cycling j OA Table 1 Comparison of single-entry and restart patients Table 3 Weight regain on restart weight loss Group Restart results Single start Multiple starts Weight gain prior to restart Success Failure (A) Demographic data Number Race (%) Black White Others 3 4 Gender (%) Women 73 65* Men 27 35* Age (years) Women Men Initial BMI (kg/m 2 ) Women Men Total (B) Results during initial therapy Weight loss (kg) Women * Men * Total * Duration (weeks) Women 16 22* Men 14 16* Total * Loss rate (kg/week) Women Men Total Success rate (%) Women 61 83* Men 74 90* Total 65 85* *Comparison between group: p < BMI, body mass index. treatment (21.7%, table 2). Of the 411 patients who were initially successful but regained weight, 169 patients (41.1%) were successful again during restart. Analysis of these 169 patients (table 3) revealed that 94 had partial weight regain before restart (defined as regain >10% of initial weight loss), 29 had complete weight regain before Table 2 Initial weight loss on restart success Restart results Initial weight loss Success Failures Patient with success (n 5 411) 169 (41.1%)* 242 (58.9%) Patient with failure (n 5 69) 15 (21.7%) 54 (78.3%) Total patients (n 5 480) *Percentage of initial patient group with either success or failure. Success, >15- kg weight loss; failure, <15- kg weight loss. Patient with partial weight regain (n 5 278) Patient with complete weight regain (n 5 49) Patient with weight on restart > initial weight (n 5 84) 94 (33.8%) 184 (66.2%) 29 (59.2%) 20 (40.8%) 46 (54.7%) 38 (45.3%) Success, >15 kg weight loss; failure, <15 kg weight loss. restart (defined as regain >90% of initial weight loss) and 46 had weight at restart greater than their initial weight. Among patients who were successful in their initial weight loss, the correlation coefficient between first and second weight loss was The overall success rate for all 480 patients (regardless of their success/failure status on initial therapy) during restart regimen was 38.3% (184 patients). Initial success or failure or interval weight changes between the two treatments do not bear any correlation with initial degree of obesity as measured by body mass index. However, the correlation coefficient between initial weight loss and subsequent weight gain in the interval between treatment periods was 0.56 (p < 0.05), that is, patients who lost more weight originally were less likely to regain all of their lost weight before restarting the programme. As shown in table 4, cardiovascular risk factors, such as blood pressure and serum cholesterol and triglycerides, were not adversely affected by weight cycling, but in fact were lower prior to restart than initial entry. High-density lipoprotein was not significantly affected by weight loss in all the attempts for weight loss in women and men. Success at achieving reduction in systolic and diastolic blood pressures as well as in total cholesterol and triglyceride levels was similar at the end of initial and restart therapy. There were also no differences in success rate between men and women. In assessing the changes in weight following each treatment period, the final weight after restart therapy was more than the final weight following initial therapy. Compared to initial weight at the start of the programme, women achieved a 27.6% weight loss and men achieved a 25.4% weight loss. When this is compared to the final weight following restart treatment, a reduction of 20.9% in weight was seen in women and 13.5% in men (as compared to initial weight). However, the rate of weight loss during the two treatment periods was identical (table 4). Additionally, the total weight loss following Diabetes, Obesity and Metabolism, 9, 2007, j 381

4 OA j Effect of weight cycling Z. Li et al. Table 4 Patients who entered the programme twice Women (N 5 314) Men (N 5 166) Mean values Initial Restart Initial Restart Baseline weight (kg) * * Final weight (kg) 76 83* * % Change from start Loss weight (kg) 29 14* 32 17* Duration (weeks) 22 11* 16 10* Rate of loss (kg/week) Baseline BMI (kg/m 2 ) 39 36* 40 37* Final BMI (kg/m 2 ) 28 31* 29 31* Baseline cholesterol * (mmol/l) Final cholesterol (mmol/l) % Change from start Baseline triglycerides * * (mmol/l) Final triglycerides (mmol/l) % Change from start Baseline glucose (mmol/l) Final glucose (mmol/l) % Change from start Baseline systolic BP * * (mmhg) Final systolic BP (mmhg) % Change from start Baseline diastolic BP 82 78* 88 85* (mmhg) Final diastolic BP (mmhg) % Change from start Comparisons between first and second entry into the programme *p < BMI, body mass index; BP, blood pressure. restart was less when compared to the initial treatment period, which was attributable to a shorter duration of participation on the restart programme as compared to the initial course of treatment. A smaller number of patients entered the programme three or four times (tables 5 and 6). The median interval between the second and third entry into the programme was 50 weeks, and between the third and fourth entry was 39 weeks. It can be seen in tables 5 and 6 that the rate of weight loss did not decrease with multiple treatments. However, similar to patients who only reentered the programme once, the duration of treatment with each subsequent re-entry into the programme became progressively shorter. This contributes to a higher final weight at the end of each treatment period when compared to the previously achieved lowest weight. This trend was seen for both male and female patients. Discussion Obesity is a major independent health hazard leading to increased overall mortality [13]. It is also closely associated with known cardiovascular risk factors such as hypertension, dyslipidaemia and diabetes mellitus [14]. Despite the fact that 25% of men and almost 50% of women are dieting at any given time [15], the average weight of the American population continues to increase [1]. In a consensus document, the National Institutes of Health stressed that the treatment of obesity is a national health priority [16]. Most people who diet lose and regain weight many times [17]. Repeated dieting, or weight cycling, has been impugned in the medical literature [18] and the popular press as a barrier to successful permanent weight loss as well as an independent hazard to general health [6 9]. It is hypothesized that weight loss followed by weight regain results in a metabolic adaptation that inhibits subsequent weight loss and promotes regain [19]. Presumably, there is a stimulus to conserve ingested energy, so that despite the same caloric intake, weight regain is greater because food efficiency is increased and energy expenditure decreased. If this is true, it may help explain the poor long-term results with respect to maintenance of weight loss [20], and repeated dieting becomes an exercise in futility and should be discouraged. Indeed, a 1987 British government report [21] stated: The end result of repeated cycles of rapid weight loss and regain may be to create the self-defeating situation of progressive loss of metabolically active tissue and its replacement by energy rich adipose tissue. Research is needed into this important sequel to weight loss. Table 5 Weight loss and rate of weight change in men and women with multiple programme restarts (patients who entered the programme three times) Women (N 5 85) Men (N 5 51) Initial First restart Second restart Initial First restart Second restart Baseline weight (kg) Final weight (kg) Weight loss (kg) Duration (weeks) Rate of loss (kg/week) j Diabetes, Obesity and Metabolism, 9, 2007,

5 Z. Li et al. Effect of weight cycling j OA Table 6 Weight loss and rate of weight change in men and women with multiple programme restarts (patients who entered the programme four times) Women (N 5 20) Men (N 5 18) Initial First restart Second restart Third Restart Initial First Restart Second restart Third restart Baseline weight (kg) Final weight (kg) Weight loss (kg) Duration (weeks) Rate of loss (kg/week) Animal experiments are contradictory. Brownell et al. studied adult male obese rats through two cycles of food restriction and refeeding [19]. These animals showed significant increases in food efficiency (weight gain/kcal food intake) during the second restriction and feeding period compared to the first, that is, weight loss occurred at half the rate and regain at three times the rate during the second cycle. Gray et al., however, using a similar study design in adult female ovariectomized rats, reported that repeated cycles of weight loss and regain did not change the rate of weight loss or percentage body fat [22]. These results confirmed a prior study by Hill et al. [23]. The relevance of these findings to weight loss and regain in humans has not been established. There is a paucity of human studies on weight cycling. Steen et al. [24] found that college wrestlers who repeatedly lose weight in order to achieve required weight for their matches have lower resting metabolic rates than their noncycling team-mates. These wrestlers are not obese, and it is unclear whether weight cycling leads to lower metabolic rates or whether lower metabolic rates promote weight cycling in this discreet, lean male population. In a retrospective study similar to this report, Blackburn et al. [25] found 43 outpatients and 14 inpatients who had participated in the same weight loss programme ( kcal/day) twice during a 9-year period. Mean interdiet interval was 810 days, and patients had regained an average of 120% of their weight lost during the first cycle. Weight loss velocity was significantly less during restart than during the first cycle. Deeson et al. [26], on the other hand, compared four subjects who restarted an 8-week VLCD (405 kcal/day) 18 months after initial participation and found no diminution in rate of weight loss. The present study population of 480 weight cycling obese individuals is by far the largest group reported. This has the advantage of permitting more accurate statistical estimates of the rates of weight loss reflected in a smaller standard error. The even larger number of the nonrestart control group (1954 subjects) permits comparison of population subgroups and treatment outcomes. In addition, this is the only study of subjects who have participated in more than two cycles of weight loss and regain. It is clear that the rate of weight loss in response to caloric restriction does not diminish with repeated episodes of weight loss and regain (tables 4 6). This implies that there is no long-term metabolic adaptation to weight cycling, which would make subsequent weight loss more difficult than that achieved initially. It also suggests that repeated weight loss and regain does not permanently change resting metabolic rate or relative fat-free mass, as has been posited by some [19,27,28] and denied by others [22,29 31]. We could not distinguish any demographic characteristics to predict which patients would re-enter the programme (table 1), except that proportionally more men than women chose to restart, and patients who restarted had been more successful in their initial weight loss than those who did not restart. We did not have sufficient information to stratify socioeconomic or psychological status of our patient population as a predictor for re-entry into the programme. It is of interest that most patients maintained some weight loss between the first and second dietary periods. This persistent difference of only 8.5 kg in body weight was associated with an improvement in cardiovascular risk factors such as elevated cholesterol and triglycerides, and a reduction in blood pressure (table 4). The weight loss rate is 1.5 for women and 2.2 for men. The weight loss for VLCD programme is average 0.8 kg/ week independent of the body weight [32,33]. In such short duration of diet, absolute weight loss would be a more fair assessment for success. Considering the average duration of patient treatment in our weight loss programme at 10 weeks, we used the criterion of greater or less than 15- kg weight loss to define success or failure, respectively. There was only a weak correlation between initial and subsequent weight loss (R ¼ 0.29). One of the five patients who failed initially was successful on reentry, whereas four of the 10 patients who were initially Diabetes, Obesity and Metabolism, 9, 2007, j 383

6 OA j Effect of weight cycling Z. Li et al. successful were successful again on restart (tables 2 and 3). The major challenge for weight loss is adherence. For patients who were motivated to restart the programme, the success rates of repeated dieting would have been greater if the subsequent treatment period was as long as the initial attempt, given that the rate of weight loss is similar during each treatment period. A tendency towards shorter participation duration is a behavioural pattern that we are trying to correct. Thus, all patients who regain a substantial amount of weight after dieting should be encouraged to try again. The study population of this analysis was patients in self-support VLCD programme. These conclusions are generalizable to obese individuals who lose weight, regardless of the degree of caloric restriction. Two recent studies compared short-term [34] and long-term [35] metabolic consequences of rapid weight loss with VLCD vs. slower weight loss with conventional higher calorie reducing diets. Both studies found no difference in resting metabolic rate normalized to fat-free mass between the two dietary regimens. Neither rapid nor slower weight loss was associated with reductions in resting metabolic rate that exceeded decreases anticipated with the achievement of a lower body weight. The patients who re-enter the programmes were probably more motivated than the general population. This does not detract from the physiology contained in the findings, and the conclusions clearly do have some implications for application to the wider population. The present study refutes the idea that repeated dieting, which is what most obese people do, results in metabolic alterations that are either harmful per se or render further dietary efforts more difficult. References 1 Mokdad AH, Bowman BA, Ford ES, Vinicor F, Marks IS, Koplan JP. The continuing epidemics of obesity and diabetes in the United States. JAMA 2001; 286: Technology assessment panel, NIH. Methods of voluntary weight loss and control. Ann Intern Med 1993; 119: Consumer s Union. Losing weight: what works, what doesn t. Consum Rep 1993; June: Ashley F, Kannal W. Relation of weight change to change in atherogenic traits: the Framingham study. J Chronic Dis 1974; 27: Sedgwick A, Thomas D, Davies M, Baghurst K. Relationships between weight change and blood lipids in men and women: The Adelaide Int J Obes 1990; 14: Hamm P, Shekell RB, Stamler J. Large fluctuations in body weight during young adulthood and twenty-fiveyear risk of coronary death in men. Am J Epidemiol 1989; 129: Lissner L, Odell PM, D Agostino EB et al. Variability of body weight and health outcomes in the Framingham population. N Engl J Med 1991; 324: Blaisr SN, Shaten J, Brownell K, Collins G, Lissner L. Body weight change, all-cause mortality, and causespecific mortality in the Multiple Risk Factor Intervention Trial. Ann Intern Med 1993; 119: National Task Force on the prevention and treatment of obesity. Weight cycling. JAMA 1994; 272: Dyer AR, Stamler J, Greenland P. Associations of weight change and weight variability with cardiovascular and all-cause mortality in the Chicago Western Electric Company Study. Am J Epidemiol 2000; 152: Lissner L, Andres R, Muller DC, Shimokata H. Body weight variability in men: metabolic rate, health and longevity. Int J Obes 1990; 14: Iribarren C, Sharp DS, Burchfiel CM, Petrovitch H. Association of weight loss and weight fluctuation with mortality among Japanese American men. N Engl J Med 1995; 333: Manson JE, Stampfer MJ, Hennekens CH, Willett WC. Body weight and longevity a reassessment. JAMA 1987; 257: Bray GA. Complications of obesity. Ann Intern Med 1985; 103: National Center for Health Statistics. Provisional data from the Health Promotion and Disease Prevention supplement to the National Health Interview Survey. Adv Data 1985; 109: National Institutes of Health Consensus Development Conference Statement. Health implications of obesity. Ann Intern Med 1985; 103: Garrow J. Energy Balance and Obesity in Man. Amsterdam: Elsevier, 1978: Jeffery RW, Bjornson-Benson WM, Rosenthal BS, Lindquist RA, Kurth CL, Johnson SI. Correlates of weight loss and its maintenance over two years of follow-up among middle-aged men. Prev Med 1984; 13: Brownell KD, Greenwood MRC, Stellar E, Shrager BE. The effects of repeated cycles of weight loss and regain in rats. Physiol Behav 1986; 38: Wing RR, Jeffery RW. Outpatient treatment in obesity: a comparison of methodology and clinical results. Int J Obes 1979; 3: Committee on Medical Aspects of Food Policy. The Use of Very Low Calorie Diets in Obesity. DHSS Report 31. London: HMSO, Gray DS, Fisler JS, Bray GA. Effects of repeated weight loss and regain on body composition in obese rats. Am J Clin Nutr 1988; 47: Hill JO, Thacker S, Newby D, Nickel M, Digirolamo M. A comparison of constant feeding with bouts of 384 j Diabetes, Obesity and Metabolism, 9, 2007,

7 Z. Li et al. Effect of weight cycling j OA fasting-refeeding at three levels of nutrition in the rat. Int J Obes 1987; 11: Steen SN, Oppliger RA, Brownell KD. Metabolic effects of repeated weight loss and regain in adolescent wrestlers. JAMA 1988; 260: Blackburn GL, Wilson GT, Kanders BS et al. Weight cycling: the experience of human dieters. Am J Clin Nutr 1989; 49: Deeson V, Ray C, Coxon A, Kreitzman S. The myth of the yo-yo: consistent rate of weight loss with successive dieting by VLCD. Int J Obes 1989; 13 (Suppl. 1): Bjorntorp P, Yang MU. Refeeding after fasting in the rat: effects on body composition and food efficiency. Am J Clin Nutr 1982; 736: Leibel RL, Hirsch J. Diminished energy requirements in reduced obese persons. Metabolism 1984; 33: Weigle DS, Brunzell JD. Assessment of energy expenditures in ambulatory reduced-obese subjects by the techniques of weight stabilization and exogenous weight replacement. Int J Obes 1990; 14 (Suppl. l): de Groot LCPCM, van Es AJH, van Raaij JMA, Vogt JE, Hautvast JCAJ. Energy metabolism of overweight women 1 mo and 1 y after 8-wk slimming period. Am J Clin Nutr 1990; 51: Kreitzman S, Coxon A, Brodie O, Eston R, Howard A. Restoration to normal body composition by VLCD: a comparison of post obese and lean subjects questions DM1 25 as diet limit. Int J Obes 1989; 13 (Suppl. 2): Saris WH. Very-low-calorie diets and sustained weight loss. Obes Res 2001; 9 (Suppl. 4): 295S 301S. 33 Very low-calorie diets. National Task Force on the Prevention and Treatment of Obesity, National Institutes of Health. JAMA 1993; 270: Coxon A, Kreitzman S, Brodie D, Howard A. Rapid weight loss and lean tissue: evidence for comparable body composition and metabolic rate in differing rates of weight loss. Int J Obes 1998; 13 (Suppl. 2): Wadden TA, Foster CD, Letizia KA, Mullen JL. Longterm effects of dieting on resting metabolic rate in obese outpatients. JAMA 1990; 264: Diabetes, Obesity and Metabolism, 9, 2007, j 385