| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
Metabolic Research Group, VA Medical Center and University of Kentucky (HMR) Weight Management Program (J.W.A., S.V.),
Department of Statistics (W.Q., R.J.K.), University of Kentucky, Lexington, Kentucky
Address reprint requests to: James W. Anderson, MD, Medical Service, 111C, 2250 Leestown Road, Lexington, KY 40511
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Subjects: Individuals who had completed the 12-week core education program and lost 
10 kg were recruited.
Results: Of 154 eligible subjects, follow-up weights were obtained at 2 years in 112 subjects (72.7%, 72 women, 40 men). Subjects had an average initial body mass index of 37.3 kg/m2 and an average weight loss of 29.7 kg in five months. Six hundred and forty-five follow-up weights (median, five per subject) were obtained over two to seven years of follow-up from clinic visits (70%) and self-report by telephone or mail (30%). Subjects regained an average of 2.5% per month of their lost weight during the first two to three years of follow-up; however, their weight stabilized over the next four years. Subjects regained an average of 73.4% of their weight loss during the first three years. The average weight loss maintained for 112 subjects was 22.8% of initial weight loss after an average of 5.3 years of follow-up. When successful weight maintenance was defined as maintaining a weight loss of 5% or 10% of initial (pre-treatment) body weight, 40% were maintaining a 5% weight loss at five years and 25% were maintaining a weight loss of 10% at 7 years. Multiple regression analyses suggested that age had a significant (p=0.004) and positive effect on weight maintenance.
Conclusions: This study suggests that weight maintenance after an intensive VLCD program is improving but still needs intensive efforts to enable most individuals to maintain a substantial percentage of their weight loss long-term.
Key words: weight loss, body mass index, very low calorie diet
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Current treatment programs for obese individuals are not effective over the long-term. While intensive weight loss programs enable many individuals to lose 20 kg to 25 kg or 20% to 25% of their initial weight [6], long-term weight maintenance is poor [7]. Common dogma indicates that persons who successfully lose weight will regain all of their weight loss within five years [8]. Even rather heroic surgical interventions do not enable most morbidly obese individuals to achieve and maintain non-obese body weights [9].
The combination of very-low-calorie diets (VLCD) with behavior modification represents an important advance in enabling obese individuals to lose substantial amounts of weight [6]. Weight losses after VLCD and low-calorie diets (LCD) combined with behavioral approaches average 20 kg to 25 kg [6]. However, long-term weight maintenance appears to be suboptimal after these interventions, although limited follow-up information is available. Recently Wadden and colleagues [10] reported long-term maintenance data for individuals who participated in an intensive VLCD program. This study is a prospective study assessing five- to seven-year weight maintenance of 112 individuals after an intensive VLCD and behavioral modification program. At seven years, 25% of these individuals were maintaining a weight loss of more than 10% of their initial body weight.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Program
During the weight-loss phase (the VLCD program), all subjects were instructed to abstain from food and consume at least 520 kcal/day in the form of five chocolate or vanilla liquid supplements plus two vitamin-mineral tablets daily. This intake provided 100% to 150% of the US Recommended Dietary Allowance [11] of most vitamins and minerals. The nonfat, dry-milk-based supplement, HMR 500 (Health Management Resources, Inc., Boston, MA) provided 50 g protein, 79 g carbohydrate, 1 g fat, and 1.5 g fiber daily [6].
In addition to weekly medical monitoring, trained health educators conducted weekly 90-minute group classes including a 12- to 14-week core education program that all subjects were encouraged to complete [12]. These classes emphasized the skills needed for long-term weight maintenance, including physical activity, environmental control and estimation of energy and portion sizes of foods. Participants were instructed to keep records of supplement and energy intake and energy expended in physical activity and to complete weekly totals with calculation of energy balance. These records were reviewed weekly by program staff. When subjects reached their self-determined goal weight or wished to discontinue the weight-loss phase, low-fat foods were gradually reintroduced over a four- to six-week period while medical monitoring and group education classes were continued.
After completing the weight-loss phase, subjects were encouraged to participate in the maintenance phase of the program for a minimum of 18 months [12]. Weekly small-group sessions focused on individual procedures for preventing weight regain by using records to confront perceived "failures," such as food binges, stress eating and snacking. Subjects also attended monthly large-group seminars and participated in periodic restaurant meals and grocery shopping tours designed for practical application of weight management skills. Subjects were encouraged to practice healthy behaviors, which included a low-fat, high-fiber diet, regular exercise and record keeping of food intake and physical activity.
Subjects failing to maintain their weight loss or desiring further weight loss were encouraged to reenter special "restart" classes tailored to the needs of former VLCD participants. These participants are termed "restarts."
Follow-up
Between June, 1993, and June, 1996, we actively sought information on the current weight of each subject. All follow-up weights in the Weight Management medical record were recorded. We also reviewed other Medical Center charts to obtain weights of participants receiving medical care at this facility. We requested weights by telephone and mail from participants who were no longer active in the program. All subjects reachable by phone were interviewed, current weight information was requested, and they were invited to visit the clinic for measurement of current weight. We wrote personal letters to subjects and requested that they mail us information on their current weight. As a final follow-up, we mailed a $10 check to all subjects for whom we had a current address and asked them to record their weight on the comment section of the check before cashing it.
Subjects were classified as "successful" if they maintained a weight loss of 5% or 10% of their enrollment (pre-treatment) weight at the time interval being analyzed. If multiple weights were available, the last available weight during each time period was assessed. Since most of the weights used for analysis were measured in the clinic, as opposed to "self-reported," we did not distinguish between measured weights and "self-reported" weights and did not make adjustments for "self-reported" weights.
Statistical Analysis
Baseline characteristics among groups based on BMI, age, enrollment weight, weight loss, duration in VLCD program and duration in maintenance program were compared by one-way analysis of variance.
Preliminary analysis indicated that weight maintenance was best described by a two-segment line indicating that the rate of weight regain was more rapid during the early period, but was more stable during the later period. The equation for this two-segment line was:
 |
where y=(weight loss in follow-up)/(weight loss in weight loss phase) x 100; x=months of follow-up; d=month at which rate of weight change changed; a=rate of weight change before d; b=intercept of the first line segment; c=rate of weight change after d; f=intercept of the second line segment. While multiple weights were available for most subjects, the two-segment analysis was based on these three weights: the weight at completion of the weight loss phase, the weight closest to three years of follow-up, since this was the average "d" or time when the rate of weight change changed, and the weight at the last follow-up point. All weights were expressed as percent of initial weight.
Two-tailed t tests assuming unequal variances were used to compare differences in percentage weight maintenance among groups at three and five years. Mean age, BMI and other weight loss endpoints were compared among these four groups: non-restart women, restart women, non-restart men and restart men. These comparisons were done by using a one-way analysis of variance (ANOVA) procedure with post hoc comparison of means for pre-selected constructs based on Fisher’s protected least significant differences procedure. Covariates potentially affecting long-term weight maintenance (gender, age, initial BMI, weight loss, length of time in weight loss, length of time in maintenance and whether or not the patient participated in restart) were examined by multiple regression analysis.
Statistical analyses were performed by using Statistica for Windows (StatSoft Inc., Tulsa, OK) and SAS software (SAS Institute, Inc., NC). A p value of <0.05 was considered to be statistically significant.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Table 1 provides the baseline characteristics of the subjects that included 72 women and 40 men. Eighty-nine subjects (79.5%, 59 women and 30 men) were classified as "non-restart" because they did not participate in an intensive weight loss program during the period of follow-up. Twenty-three subjects (20.5%, 13 women and 10 men) were classified as "restarts" because they enrolled at least one time in our intensive weight loss "restart" program during the period of observation. BMIs averaged 37.3 kg/m2 for the entire group; men had higher average BMIs than women, and the BMIs of men who restarted were significantly higher than those of the other groups. The average age was 45.8 years and was similar in both subgroups of women and in the male non-restart group; the male restart subgroup was significantly younger than other subgroups. Enrollment weights were significantly higher in men than in women.
|
Follow-up Weights
After completion of weight loss, 645 follow-up weights were obtained for the 112 subjects (median of five weights/subject, range of two to 23 weights/subject). Follow-up weights were obtained in clinic for 70.2% of measurements and were self-reported for 29.8% of measurements. Fig. 1 illustrates the follow-up weights. Weight maintenance was plotted in kilograms, indicating an initial weight loss of 29.7 kg (100%) at zero years. Over the first three years subjects regained weight rapidly. However, between three and five years the weight loss in kilograms was fairly stable at approximately eight kilograms. The best linear fit for this pattern was a polynomial equation. Because of the variable number of measurements available at various time points, there is some fluctuation in the percentage weight maintenance that was especially noteworthy at five years.
|
60 months of follow-up (the average was 73 months). At this time point, 24 women were maintaining an average of 19.1% (±9.5%, SEM) of their weight loss and 28 men were maintaining an average of 16.3% (±11.0%) of their weight loss (p=0.85). For these 52 subjects there were no significant differences in weight maintenance for restarts versus non-restarts. The last available weight for the 112 subjects at an average of 5.3 years of follow-up showed that subjects were maintaining an average of 22.8% of their weight loss.
Successful Weight Maintenance
Defining successful weight maintenance as maintaining a weight loss of more than 10% of initial body weight which represents an average of approximately 11 kg (see Table 1), Fig. 2 illustrates success rates between one and seven years. Women and men had approximately the same success rates, with variations related primarily to variable numbers of subjects available at each observation point. By three years, approximately 31% remained successful, and this success rate was maintained fairly well through seven years with values being 25% successful for both women and men. Success rates were also calculated at the 5% weight maintenance level. Weight maintenance at 5% loss of initial body weight were as follows: 50.4% at three years (n=76); 61.8% at four years years (n=38); 40% at five years (n=15); 49.2% at six years (n=42); and 34.7% at seven years (n=32). Meaningful differences in success rates between women and men were not noted.
|
Multiple Regression Analysis
A multiple regression analysis was performed using weight maintenance at three and five years as the dependent variables. The two-segment model calculated weights at years three and five; thus, for these analyses 94 weights (61 women and 33 men) were available at years three and five. The independent variables assessed were gender, age, initial BMI, weight loss in VLCD program, months in VLCD program, months in maintenance program and group (non-restart vs. restart). None of these variables were significantly related to weight maintenance at three years. Although the regression lines were not statistically significant at three years (p=0.062 to p=0.118), the association between weight loss and weight maintenance at three years were significant (p=0.031 to p=0.040). Multiple analyses using different combinations of these variables consistently indicated that age had a significant and positive correlation with weight maintenance at five years (p=0.012 to p=0.005). Because of the large variance in percentage weight maintenance, however, the regression lines were not statistically significant with the lowest p value being 0.052. These analyses also suggested that group (non-restart vs. restart) approached statistical significance (p=0.053 to p=0.064). Although neither regression model appeared to be statistically significant, the only strong association observed was that older individuals appears to maintain weight loss better than younger individuals.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
10 kg. Initial BMIs averaged 37.3 kg/m2 and were similar for women and men. Weight loss averaged 29.7 kg over an average period of five months. Weight regain was rapid and averaged 73% of the weight lost over three years after completing the weight-loss program. However, between three to seven years, the body weights were fairly stable with weight maintenance averaging 26.6% to 17.9% of weight loss. Successful weight maintenance at levels of 10% and 5% of initial body weight were examined. At the 10% level, 31% at three years and 25% at seven years were maintaining a weight loss 10% of their initial body weight; values were similar in women and men. At the 5% level, 50% at three years and 40% at five years were maintaining a weight loss 5% of their initial body weight. Women and men did not differ significantly for these values. When we analyzed the last available weight for 112 subjects, they were maintaining a weight loss of 23% of their initial weight loss at an average of 5.3 years. Our results were appropriate for analysis with a two-segment linear model. This indicated that subjects regained an average of 2.5%/month of their initial weight loss over a 24- to 36-month period after weight loss. Subsequently, they did not gain further weight, on average, since the second slope indicated a weight loss of 0.01%/month. Multivariate analyses for weight maintenance data at three and five years did not yield statistically significant regression lines. However, older individuals appeared to maintain their weight loss significantly better (p=0.004) than younger individuals. Furthermore, duration of time in the maintenance program and whether the individual participated in a restart program appeared to be related to percentage of weight loss maintained (p=0.06). Gender, initial body weight and amount of weight loss did not appear to be related to weight maintenance in this analysis.
The two-segment linear model appears to be useful in describing the weight change pattern after an intensive weight-loss program. Simple linear regression analyses for weight changes in the first three years and for years three to seven yielded similar data, with a similar intercept at approximately three years. These observations suggest that most patients rapidly regain weight over the first three years and, on average, maintain their weight fairly well over the next several years. Further analyses incorporating more extensive long-term data are required to confirm our observations that weight remains fairly stable after three years. Previous data from Bjorvell and Rossner [13] and Walsh and Flynn [14] are consistent with our observations.
Wadden and Frey [10] recently reported five-year follow-up data on patients completing a similar VLCD program. Their analysis differs from ours in these respects: their inclusion criteria required completion of 26 weeks of treatment compared to our twelve weeks; all of their weights were self-reported whereas >70% of our weights were measured and <30% were self-reported; and they obtained follow-up weights on 48% of eligible patients while we obtained values on 72% of eligible patients. As they [10] point out, long-term studies are likely to be favorably biased because unsuccessful patients are less willing to participate. Because their patient participation level was lower than ours, their results possibly are more favorably biased. Nevertheless, at five years their men were maintaining 34.6% of their weight loss and their women were maintaining 18.2% of their weight loss (data from Table 1 [10]), while our patients were maintaining an average of 19.9% of their weight loss after five years. The five-year success rates for maintaining 10% and 5% reductions in body weight can also be compared as follows: 57.9% of their men, 47.8% of their women and 40.2% of our patients were maintaining a 5% reduction at five years; 28.1% of their men, 31.3% of their women were maintaining a reduction of 10% at five years while 25% of our patients were maintaining a 10% reduction at seven years.
Our current study is consistent with many previous studies [7,15–22] indicating that, even after successful weight loss of 10 kg, most individuals are unable to maintain substantial amounts of weight loss over five years. However, this report and the report of Wadden and Frey [10] suggest that long-term results currently being achieved are better than those reported by Andersen and colleagues [8] and others [7] in the previous decade. The recent report from the National Weight Control Registry [23] also indicates that selected individuals are successfully maintaining substantial amounts of weight loss over the long-term. This unique Registry sought information on individuals who had lost 13.6 kg and maintained that loss for at least one year. These individuals lost an average of 30 kg and maintained a minimum weight loss of 13.6 kg (>45% of initial weight loss) for an average of 5.5 years.
Predictors of long-term weight maintenance are not clearly identified because, in large part, there is a paucity of long-term data to test hypotheses. Our observations provide data for further hypothesis development. The strong positive correlation between age and long-term weight maintenance at five years is of note. Further, suggestions that the duration of time in maintenance and the participation in restart approach statistical significance require further evaluation. Our clinical observations suggesting that the amount of weight lost in the weight loss phase might correlate with percentage of weight maintained were not supported by this analysis.
As others have previously documented [24], the amount of weight being lost in intensive weight loss programs is increasing steadily. While the percentage of weight loss being maintained may not be increasing dramatically, the absolute amount of weight loss being sustained is also rising steadily. When individuals lose 9% of their initial body weight, the average for behavioral programs [24], their chances of being successful at maintaining a loss of 10% or even 5% of initial body weight are substantially less than after losing the average of 26% of initial body weight, as observed in this study. Because we excluded persons who lost <10 kg from these analyses, we may have diminished the chances of seeing a stronger correlation between initial weight loss and long-term weight maintenance.
Although the weight maintenance in this study, reflecting an average of 6.8 kg kept off at 5.3 years, is clearly suboptimal, several additional factors should be considered. First, the available data indicate that obese individuals who are not treated gain weight over the following five to ten years, although the magnitude of weight gain is not well delineated. One study [25] of middle-aged obese women in the US reported an average weight gain of 2.1 kg over a ten-year period; in another study [23], more than half the "superobese" women gained over 20% of their ideal body weight in at least one decade of life. In a research study [27], untreated control women gained 1.5 kg to 4 kg over three years. Thus, the average woman who participated in our intensive VLCD program might be expected to gain between one kilogram [25] to three kilograms [27] to six kilograms [26] during the five years of follow-up in our study. One could speculate that this hypothetical woman might weigh between eight and fifteen kilograms less at five years than she might have weighed if she had not participated in this program.
A second factor is that modest weight losses of 5% to 10% of body weight are accompanied by significant reductions in risk for developing several diseases. Recent reports persuasively indicate that modest weight gains are associated with significantly increased risks for developing coronary heart disease [28], hypertension [24] and diabetes [29]. Consistent with these observations regarding the health risks of weight gain, emerging data strongly suggest that maintenance of modest amounts of weight loss significantly decreases risk for developing diabetes [29]. A third factor relates to the significant improvements in several co-morbid diseases including diabetes [30,31], hypertension [32–34] and dyslipidemia [33–35], which are associated with modest weight loss.
A fourth benefit of maintaining modest amounts of weight loss may be important changes in lifestyle associated with this weight maintenance. As recently reviewed [24], predictors of good maintenance of weight loss include regular physical activity and low-fat intakes [23]. While we encourage individuals in our maintenance program to maintain physical activities levels of 2000 kcal/week, to maintain a low fat intake, to eat five servings of vegetables and fruits daily and to keep records of physical activity and food intake until these behaviors are incorporated into their lifestyle [6], we do not have data on achievement of these goals by patients in this current study. These behaviors, however, are similar to those used by successful maintainers of significant weight loss in the National Weight Control Registry [23].
Because successful maintenance of weight loss over the long-term is difficult, additional strategies should be explored. We recently reported [36] that subjects given orlistat, a lipase inhibitor, enhanced weight maintenance over a one-year period after successful weight loss. Long-term adjunctive drug therapy may be required to enable some individuals to successfully maintain their weight loss.
In conclusion, our study indicated that individuals who successfully lost 30 kg in an intensive VLCD weight loss program rapidly regained almost 75% of their lost weight over three years after completing the program. However, weight stabilized after three years and was maintained over the next two to four years. Older patients maintained their weight significantly better than younger patients. Our 112 patients were maintaining an average of 22.3% of their weight loss after an average of 5.3 years of follow-up. When successful weight loss was defined as maintaining a weight reduction of 5% or 10% of their initial (pre-treatment) weight, 40% were maintaining a 5% weight reduction at five years and 25% were maintaining a 10% weight reduction at seven years.
|
FOOTNOTES |
|---|
Received December 1, 1998. Accepted June 1, 1999.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  K. S. Dale, K. A. McAuley, R. W. Taylor, S. M. Williams, V. L. Farmer, P. Hansen, S. M. Vorgers, A. W. Chisholm, and J. I. Mann Determining optimal approaches for weight maintenance: a randomized controlled trial Can. Med. Assoc. J., May 12, 2009; 180(10): E39 - E46. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. W Anderson, S. B Conley, and A. S Nicholas One hundred pound weight losses with an intensive behavioral program: changes in risk factors in 118 patients with long-term follow-up Am. J. Clinical Nutrition, August 1, 2007; 86(2): 301 - 307. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. R. Wilson and D. D. McAlpine The Effectiveness of Screening for Obesity in Primary Care: Weighing the Evidence Med Care Res Rev, October 1, 2006; 63(5): 570 - 598. [Abstract] [PDF]
|
|
|
|
|
|
N. Vogels, E. C. Mariman, F. G Bouwman, A. D. Kester, K. Diepvens, and M. S Westerterp-Plantenga Relation of weight maintenance and dietary restraint to peroxisome proliferator-activated receptor {gamma}2, glucocorticoid receptor, and ciliary neurotrophic factor polymorphisms Am. J. Clinical Nutrition, October 1, 2005; 82(4): 740 - 746. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. G. Tsai and T. A. Wadden Systematic Review: An Evaluation of Major Commercial Weight Loss Programs in the United States Ann Intern Med, January 4, 2005; 142(1): 56 - 66. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. A van Baak, E. van Mil, A. V Astrup, N. Finer, L. F Van Gaal, J. Hilsted, P. G Kopelman, S. Rossner, W P. James, and W. H. Saris Leisure-time activity is an important determinant of long-term weight maintenance after weight loss in the Sibutramine Trial on Obesity Reduction and Maintenance (STORM trial) Am. J. Clinical Nutrition, August 1, 2003; 78(2): 209 - 214. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
