• E-mail
• Print
• Comment
• Font Size
• Digg
• del.icio.us
• Discuss article

Weight Loss and Biomedical Health Improvement on a Very Low Calorie Diet: The Moderating Role of History of Weight Cycling

Posted on: Friday, 10 June 2005, 03:00 CDT

In this study, the authors examined biomedical consequences of participation in a professionally delivered, multifaceted very low calorie diet (VLCD) program and whether the degree of benefit associated with treatment was moderated by history of weight cycling. The authors monitored body weight and biomedical health indicators in 66 severely obese outpatients on a VLCD liquid fast. Participants remained on the VLCD for a median of 55 (range 9 to 247) days. Treatment was associated with significant pre-to-post improvements on body weight, systolic and diastolic blood pressure, triglycerides, and cholesterol. History of weight cycling (independent of age) was inversely related to the magnitude of absolute pre-to-post treatment changes in systolic and diastolic blood pressure, as well as to the rate of weight change. More intensive, longer term, and explicit maintenance components, especially aimed at individuals with multiple weight loss-regain episodes, may be necessary to facilitate weight loss and attain optimal health benefits from VLCDs.

Index terms: health, medical, obesity, relapse, treatment, very low calorie diets, weight cycling, weight loss

Although weight-loss programs appear to result in significant short-term improvements in health,1-3 only a small percentage (less than 25%) of individuals actually maintain their lowered weight over substantial time periods during the post-loss follow-up period.4-6 As a result of widespread relapsing, a large percentage of dieters attempt numerous dieting programs and experience repetitive bouts of short-term weight loss followed by subsequent weight regain. This is commonly referred to as weight cycling or yo-yo dieting. Because of the high level of recidivism or negative rebounding, controversy has ensued about the desirability of losing weight only to regain it.

Results from a seminal study by Brownell et al.7 suggested that history of weight cycling may adversely impact the outcome of future diets. In this early animal study, obese male rats with a history of weight cycling had more difficulty losing weight (2 times slower) when subsequently placed on a calorie-restricted diet. These yo-yo- dieting rats also regained their weight more quickly following the test diet than did their non-cycling counterparts. In a more recent animal study, Gerardo-Gettens et al.8 studied obese female rats using a repeated-measures design and found a decrease in the amount of weight lost for the second weight-loss attempt as compared with the first weight-loss attempt.

Research studies of obese human participants have also found a reduced amount and rate of weight loss across consecutive weight- loss attempts,9,10 suggesting findings based on animal models may be able to be generalized for humans as well. In this connection, the data are mixed. Other studies, for example, have failed to find an association between the history of weight cycling and greater difficulties in weight loss over the course of a particular treatment program.11-14 Other studies involving humans have shown that a greater history of weight cycling may increase the probability and amount of eventual weight regain.15

Longitudinal studies consistently report a higher risk of mortality resulting from coronary heart disease (CHD) for individuals with a history of greater variability in their weight. This mortality effect of (apparent) weight cycling remains significant even after adjusting for age and other known CHD risk factors.16-19 The physiological mechanisms underlying this increased risk and the extent to which these exaggerated levels of weight fluctuation were actually due to voluntary dieting efforts still remains unclear.

One proposed mechanism may be that repeated weight loss attempts may lead to reductions in metabolic rates that may render the body more efficient at storing fats, which in turn makes subsequent weight loss more difficult and regain easier.20,21 In this connection, research shows that higher levels of body fat, especially in centrally located adipose tissue, can greatly increase the risk of morbidity and mortality associated with CHD.22 Other supporting evidence comes from findings that indicate lower resting and exercise energy expenditure in samples of nonobese female cyclic dieters relative to nondieting controls.23,24 However, successive decreases in resting metabolic rate were not evident in a prospective study of obese women followed over the course of 3 consecutive weight-loss and regain cycles.10

A second mechanism involves the suggestion that yo-yo dieting may promote the accumulation of higher levels of body fat. Such an association has been observed in weight cycling studies of rats8,25,26 and normal-weight humans.23,27 Also, in comparison to obese controls of stable weight, increased preferences for food with higher contents of sugar and fat have been reported in rats with a history of weight cycling8 and in obese human weight cyclers.28,29 There are inconsistencies in the literature, however, and it should be noted that no association has been found linking cycling to body fat in the vast majority of animal studies30 or in samples of obese men and women.14,31

Yet another possible mechanism to account for the apparent deleterious health effects of repeated bouts of weight loss and regain may be that weight cycling contributes to alterations in serum glucose or lipid metabolism.32 In this regard, Sea et al.33 reported that weight cycling in rats resulted in significant alterations in measures of both lipid (serum cholesterol and triglyceride levels) and glucose metabolism (serum glucose, insulin, and glucagon levels). Similarly, Oison et al.34 found that reduced levels of high-density lipoprotein cholesterol were associated with a positive history of weight cycling in a sample of 485 women at risk for CHD. There are some inconsistencies, however, given that earlier research with obese adults did not show that history of weight fluctuations (a proxy measure of weight cycling) was associated with disturbances in glucose tolerance or levels of total serum cholesterol.31,35-37

Finally, individuals with a positive history of weight cycling may be at elevated risk of CHD by virtue of the mediating influence of higher levels of blood pressure. Ernsberger and Nelson38 reported that rats with a history of weight cycling had significantly higher systolic blood pressure than control rats without a history of cycling fed either a high sucrose or regular diet. However, other animal studies have been unable to replicate these results.39,40

Studies of the impact of weight cycling on blood pressure among human samples have also yielded mixed results. Kajioka et al.24 employed a repeated measures design and reported significant elevations in systolic and diastolic blood pressure in nonobese young women subjected to a cycle of weight loss, regain, and subsequent loss. Recent prospective studies suggest that weight cycling may serve as a risk factor (or marker) for the development of hypertension among obese women.41,42 In contrast, other research has failed to find associations between the history of weight cycling and hypertension in obese adults35,36 and in mixed normal- weight and overweight samples of women43 and men.44

Although longitudinal studies consistently report a higher risk of mortality resulting from CHD for individuals with greater variability in their weight, mixed results have emerged from existing research of authors who have attempted to pinpoint the possible physiological mechanisms for this increased risk. Inconsistent findings may be attributable to differences in sample characteristics (eg, age, sex, and starting weight), variable adherence to the treatment programs investigated, or heterogeneity in operational definitions of weight cycling or outcomes.21,45

OBJECTIVES AND PREDICTIONS OF THE PRESENT INVESTIGATION

In the present investigation, we sought to extend and refine previous weight cycling research that has examined clinical samples of obese weight loss patients. In particular, we examined the association of the history of weight cycling to a variety of treatment-related outcome measures in the context of an aggressive multifaceted weight-loss program based on a liquid supplemented fast. Refining previous human research involving clinical samples, we statistically controlled for a host of possible confounding factors-influences that are likely correlated with weight cycling. In particular, we sought to examine whether the medical health benefits of professionally assisted intentional weight loss are moderated by the history of weight cycling independent of (a) age, (b) gender, (c) preexisting medical conditions, (d) depression, (e) starting weight, and (f) treatment program adherence. Because previous studies have employed different means of assessing outcomes, we examined the effects of weight cycling history on treatment outcome using 2 commonly applied operations: (1) absolute and percent difference between baseline and posttreatment and (2) rate of change.

In the present study, we monitored severely obese weight-loss clients with varying degrees of historical involvement in weight cycling before and after participation in an aggressivemultifaceted very low calorie diet (VLCD) that employed a supplemented fast consisting only of Optifast liquid milkshakes (Sandoz Nutrition Co., Minneapolis, MN). No solid food was permitted during the loss phase of the program, which serves as the focus for the present study. On the basis of previous research evaluating health effects of the Optifast VLCD diet,46 our first general expectation was that clients would lose a clinically significant amount of body weight as a result of treatment participation. To extend research on liquid fast- based VLCDs,46,47 we also examined whether the health benefits of participation in a VLCD would extend beyond mere weight loss. In particular, we examined a host of biomedical health indicators, including blood pressure, triglycerides, cholesterol, and glucose levels.

Our second major objective was to extend previous research suggesting weight cycling may moderate the health benefits associated with intentional efforts to lose weight. On the basis of the literature reviewed, we predicted that individuals who started treatment with a positive history of weight cycling would lose less weight and lose weight more slowly than their non-cycling counterparts.10,24 Furthermore, we predicted that cyclers would show lesser improvements in biomedical health indicators (ie, lesser change in blood pressure, triglycerides, cholesterol, and glucose).

METHOD

Participants

The sample consisted of 66 obese outpatients (women, n = 45 and men, n = 21) attending a multifaceted VLCD (M-VLCD) treatment program at a university-based weight-loss clinic in Long Island, New York. At baseline, the mean startweight (kg) of the participants was 98.55 (SD = 25.62), 114.60 (SD = 23.08) and 91.06 (SD = 23.40) for the entire sample, men and women, respectively. The mean age was 43.97 years (SD = 12.82) for the entire sample, and ages ranged between 18 and 73 years. After completing an informed consent agreement, clients responded to a brief battery of self-report questionnaires. They also participated in a medical screening conducted by a physician as part of their ongoing treatment program. For the purposes of the present study, we extracted data from the pretreatment and posttreatment medical screenings.

Measures and Intervention Program

Pretreatment Measures

Assessment of Weight Cycling. We assessed the subjects' history of weight cycling at treatment intake using a self-report measure that excluded informal natural change efforts to lose weight. Specifically, we based the cycling index on the number of previously attended weight loss programs that were organized and structured. We provided clients with a vertical list of numbers 1 through 10. Above the list, they were given the following instruction: "We would like to know which dieting clinics or weight loss agencies/programs you have attended in the five years prior to coming to this program. Do not include idiosyncratic self-change efforts to lose weight that you, yourself, devised (e.g., a Grapefruit diet you found in a magazine or self-help book). List only programs that were organized and that involved 'formal' treatment run by trained health care professionals. You may also list organized dieting or lifestyle change programs lead by semi-professional for-profit organizations (e.g., Jenny Craig) or organized programs based on support groups such as TOPS, Weight Watchers, Overeaters Anonymous, etc." The weight cycling index simply consisted of a count indicating the number of organized programs attended in the previous 5 years.

Recent studies have employed a similar measure of weight cycling by having participants quantify the frequency of diets undertaken over their lifetime.48-50 The median number of substantial weight loss and regain cycles reported by participants was 1 and ranged from 0 to 4 cycles.

Assessment of General Physical and Mental Health. Participants completed a questionnaire indicating whether they had any of the following medical illness symptoms or conditions: diabetes, kidney disease, high blood pressure, coronary heart disease, shortness of breath, swelling of the legs, back pain, odd sleep times, or any other medical condition. We calculated a total score for illness symptoms by summing the number of "yes" answers. We measured depressed mood at entrance into the program with the 21-item Beck Depression Inventory (BDI).51

Assessment of Treatment-Related Weight Loss. We determined weight loss by calculating the difference between weight (kg) prior to entering the treatment program and weight (kg) at the end of Phase 1 of the treatment program (ie, the end of the liquid fast diet). A physician obtained each participant's weight during the pretreatment and posttreatment medical screenings. We computed treatment-related changes in body weight in 2 different ways: (1) absolute change score, where we subtracted posttreatment values from pretreatment values and (2) rate of percent change score, calculated using the formula: {[(posttreatment weight - pretreatment weight) pretreatment weight] 100%} time involved in the supplemental fast. The use of rate of percent change permits the adjustment for individual variability in both starting weight (baseline weight) and time in treatment.

Assessment of Biomedical Health Indicators. Prior to entering the treatment program, participants received a physician-administered medical screening that yielded the following measures: (a) systolic and diastolic blood pressure (mm Hg), (b) blood levels of triglycerides (mg/dl), (c) total cholesterol (mg/dl), and (d) blood glucose (mg/dl) levels. SmithKline Bio-Science Laboratories of Lake Success, New York performed blood assays. We completed posttreatment assessment of these variables at the end of Phase 1 of the treatment program (ie, at the end of the liquid fast diet).

The M-VLCD Treatment Program

We placed participants on a M-VLCD weight control program. The liquid diet fast aspect of the program consisted of a highly restrictive diet of Optifast 70, a very-low calorie commercially prepared liquid-formula diet (Sandoz Nutrition Co., Minneapolis, MN) that provides a total of only 420 calories daily. This aggressive VLCD is designed to produce very rapid weight loss while preserving levels of body protein and was supplemented by psychological support and an exercise program. Consumption of any sort of solid food (even gum chewing) is strictly forbidden on the Optifast 70 diet. Because of the rapidity of weight loss, we monitored participants' vital signs medically on a weekly basis. Moreover, we obtained their blood profiles on a biweekly basis. In regards to psychological support, the VLCD treatment program involved weekly group therapy support sessions that emphasized cognitive-behavioral skills training. Master's level graduate students facilitated most groups as part of their supervised training required for a doctorate in clinical psychology. Finally, clients received a progressive walking program supervised by exercise physiologists. We strongly encouraged treatment adherence to all components of the weight management program and monitored participants closely. For purposes of the present study, we assessed adherence using client attendance records and based it on the percentage of participants' attendance at weekly group therapy support sessions. Clients who dropped out prior to 7 days were excluded from the present dataset.

When clients reached 10% of their goal weight, they began to receive instructional sessions on nutritional counseling and food purchase and preparation. This marked the end of Phase 1 of the program. At that point, we progressively readied participants for graduation from the VLCD supplemented fast phase, which involved nutrition only from Optifast 70 milkshakes, to Phase 2, the maintenance phase. Phase 2 involved progressive consumption of greater and greater amounts of a low-fat solid food diet based on prepared frozen meals. Phase 3 involved weaning from the prepared frozen meals to food purchased or prepared by the client. In the present study, we limited our attention to Phase 1 only (55 days, on average, from start to finish).

RESULTS

Pretreatment Descriptive Data

We performed all statistical analyses using SPSS Statistical Package, version 10.0 for Windows. On the general physical and mental health questionnaires, participants reported a median number of 1 (range = 0 to 8) medical illness symptoms and a mean depression score of 11.11 (SD = 4.90). Participants remained on the supplemented fast diet (ie, Optifast 70) for a median number of 55 days (range = 9 to 247). Mean treatment adherence, as expressed in terms of percentage of possible attendance, was 86.45% (SD = 21.42), indicating good adherence. We noted higher mean weight cycling scores for women as compared with men, F (1, 65) = 5.124, p = .027. Sex differences were not evident on any of these other descriptive variables.

Main Effects of Treatment on Weight and Biomedical Health Indicators

Table 1 shows the comparisons of pre- and posttreatment means on the 6 outcome measures. Two-tailed group t tests revealed significant pretreatment to posttreatment changes in body weight and all of the biomedical health indicators, except serum glucose levels. As can be seen in Table 1, participation in the M-VLCD was associated with significant and potentially clinically meaningful improvements in systolic blood pressure, diastolic blood pressure, triglycerides, and cholesterol. When we examined treatment effects separately by sex, we found that reductions in triglycerides were no longer significant for women, n = 45; t (44) = .304, p = .763.

Pretreatment Correlates of History of Weight Cycling

Both Pearson's correlations and partial correlations that were adjusted for age showed that the history of weight cycling was not related to pretreatment weight or any of the pretreatment values on the other biomedical health indicators\. Two-tailed Pearson's correlations showed that weight cycling was significantly correlated with age (r = .303, p = .014), but unrelated to pretreatment (total) medical illnesses, depression, number of days on the fast supplement, or treatment adherence. Hence, we used partial correlations controlling for age to examine the relationship between weight cycling and changes in body weight and biomedical indicators over the course of the treatment program.

Moderating Effects of Weight Cycling on Treatment Impact

Descriptive analyses showed that the distribution of 5 of the dependent variables used in the following analyses deviated notably from normal. We made an attempt to obtain normalized distributions by applying the appropriate transformations to scores on the medical illness questionnaire (square root), glucose absolute change scores (reflect and square root), systolic blood pressure rate of percent change scores (reflect and inverse), and triglyceride rate of percent change scores (inverse). When we refer to these variables from this point forward, the values used in the correlations are the transformed scores. Cholesterol rate of percent change, however, could not be normalized by any of the commonly applied transformations, and subsequent analyses used the original data.

TABLE 1. Pretreatment and Posttreatment Descriptive Data and t Tests for Measures of Weight and Biomedical Health Indicators Showing Main Effects of Treatment (N = 66)

TABLE 2. Partial Correlation Coefficients (Controlling for Age) Relating Weight Cycling to Treatment-related Changes in Body Weight and Biomedical Health Indicators (N = 66)

Partial correlations relating weight cycling to baseline values of the 6 dependent variables were not significant. Table 2 shows the results of the partial correlations relating weight cycling to treatment-related changes for the 6 outcome measures. We computed treatment-related changes in body weight and biomedical health indicators in 2 different ways: (1) absolute change score, where we subtracted posttreatment values from pretreatment values (left-hand column) and (2) rate of percent change score (right-hand column).

The history of weight cycling was significantly inversely associated with absolute changes in both systolic (r = -.428, p < .001) and diastolic blood pressure (r = -.330, p < .01), suggesting that individuals with lesser improvements in blood pressure had higher weight cycling scores (see Table 2). In addition, greater levels of weight cycling were significantly and inversely related with the rate of percent weight loss during the supplemental fast (r = -.286, p < .05). When we repeated these correlations separately by sex, we found similar results for the women (n = 45), although weight cycling was not significantly associated with changes in weight or blood pressure in men (n = 21), possibly because of reduced power related to the smaller sample size. To test for the possibility that significant effects for women might be due to greater power from the larger number of women in the sample, we chose 21 women randomly and reexamined the associations. Results based on this subsample of women revealed weaker correlations but a similar pattern. The association between cycling was marginally significant for absolute systolic blood pressure (r = -.455, p < .05), and we noted a trend for rate of weight loss (r = -.413, p < .07). However, absolute diastolic blood pressure was no longer significant in the subsample of women.

If the history of weight cycling has a deleterious effect on treatment outcome, the number and magnitude of treatment-related changes should be less in clients with a positive history of cycling relative to clients with a negative history. Accordingly, we examined treatment-related changes in weight loss and biomedical health indicators separately in a subsample of participants without a history of weight cycling and in a subsample of participants with greater than the median number of weight cycling episodes. We identified 27 clients who had no history of weight cycling in the past 5 years, whereas we identified 14 clients who had a positive history of cycling. Negative and positive history of weight cycling groups did not differ significantly on any of the baseline values for the 7 dependent variables, nor in terms of their duration or adherence on treatment. To adjust for the potentially biasing effects of unequal sample sizes, we extracted a subsample of 14 clients randomly from the larger sample of 27 negative history clients. Table 3 shows repeated measure t tests of the pretreatment and posttreatment means on the 7 outcome measures. As can be seen in Table 3, treatment-related changes in weight loss, systolic blood pressure, and diastolic blood pressure remained significant (p < .001) for the group of individuals with no history of weight cycling, but were not significant for individuals with a positive history of weight cycling.

TABLE 3. Repeated-measures t Tests for Pretreatment to Posttreatment Weight and Biomedical Health Change Indicators Separately for Individuals With a Negative (n = 14) and Positive History of Weight Cycling (n = 14)

Table 4 shows between-group t tests comparing individuals without a history of weight cycling (n= 14) in terms of the magnitude of their improvement in weight and biomedical health indicators resulting from their participation in the treatment program. Individuals with a negative and positive history of weight cycling differed significantly in treatment-related mean absolute change in systolic blood pressures, t (26) = 2.848, p < .01, and mean rate of percent change in weight loss, t (26) = 2.444, p < .05.

DISCUSSION

We had 2 goals in this study: (1) to examine the impact of participation in a multifaceted VLCD treatment program on weight loss and other indicators of biomedical health status and (2) to examine the potential moderating influence of history of weight cycling on treatment-related biomedical health benefits.

TABLE 4. Between-groups t Tests Comparing Clients With a Positive (n = 14) and Negative (n = 14) History of Weight Cycling in Terms of Weight and Health Improvement Resulting From Participation in Treatment

In terms of our first objective, we observed significant pre- to posttreatment reductions in body weight. This finding replicates previous studies and provides further evidence to support the clinical effectiveness of VLCDs. 1-3,46 Indeed, the typical client in the present study lost 14 kg during the course of their involvement in the supplemented fast, with a mean rate of loss approximating 1.7 kg per week. Extending previous research, we were also able to demonstrate that participation in the VLCD yielded a broad array of health benefits. Over the course of treatment, the average client enjoyed statistically significant and potentially clinically meaningful improvements on 4 of the 5 additional indicators of biomedical health status. Specifically, participants who stayed in treatment showed, on average, a pre-to-post reduction in systolic blood pressure of 11.32 points and a mean reduction in diastolic blood pressure of 5.20 points. On average, serum triglyceride levels were reduced by 14.97 mg/dl and total serum cholesterol levels were reduced by 37.38 mg/dl.

Although a variety of short-term health benefits seem clearly documented in this study, the impact of these improvements on the risk of future illness or medical problems is unknown. Among those participants who are able to subsequently sustain their lowered weight over an extended period of time, the blood pressure and blood chemistry improvements may have some positive medical impact in terms of coronary disease (or other) risk. However, the prognostic significance of blood pressure and blood chemistry changes may be quite different for those participants who lost weight on the program but who went on to relapse. The possible existence of differential health benefits of weight loss over the long term awaits empirical validation and thus represents a direction for future research. Given that repeated bouts of losing and then regaining weight (ie, a history of yo-yo dieting) may confer increased risk of subsequent development of health problems, such as hypertension, CHD, and diabetes,15,41,42 it seems prudent to suggest that individuals who attend weight loss programs for health reasons should be careful to select programs that give special attention to relapse prevention.

In terms of our second objective, we found that weight loss participants with a positive history of weight cycling showed a decreased rate of weight loss and lesser improvements in blood pressure over the course of the VLCD. They also showed trends toward losing lesser amounts of body weight. Because we controlled for a host of possible confounding variables, the observed moderating effects of history of yo-yo dieting are independent of gender, age, preexisting medical problems, depression, time in treatment, and treatment program adherence. Thus, our findings pertaining to the moderating effects of history of yo-yo dieting on health improvement cannot easily be explained away as statistical artifacts of correlated factors. Because we adjusted for these other influences, our results provide the strongest evidence to date to suggest that past involvement in weight cycling affects the likelihood of enjoying medical health benefits typically derived from completion of professionally assisted weight-loss programs. In conclusion, when attempting to diet we found that clients with a positive history, relative to clients with a negative history, of yo-yo dieting appear to fair more poorly in terms of how much they lose, how easy it is for them to lose this weight, and how much medical benefit they derive from losing the weight.

The clinical implications seem clear enough. Take for example the case of the obese hypertensive patient who is prescribed \a weight- loss regimen for health reasons. On the basis of our findings, we would argue that it is prudent to take care to screen these types of patients so as to refer those with a positive history of cycling to programs that provide extra support during both the loss phase and the maintenance phase of treatment. In this way, knowledge about status as a yo-yo dieter may have some clinical utility. Specifically, the quality of health care provided may be improved by optimizing patient-treatment matching. This is similar to the philosophy of stepped-care in the area of substance abuse in which treatment intensity differs for alcohol abusers and alcoholics.

In general, our conclusions pertaining to weight cycling are consistent with previously published longitudinal studies of obese adults whose weight was tracked in a naturalistic context over consecutive diets.9,10,41,42 Previously published studies also suggest that multiple cycles of weight loss and regain may have longer term negative health consequences. When considered in the aggregate, yo-yo dieters appear to fair poorly both in the short term and long term. In the short term, on any particular weight- loss program, yo-yo dieters struggle more to lose weight and enjoy fewer health benefits from their hard-earned losses. In the long term, because they subsequently relapse, they increase lifetime chances of experiencing negative health outcomes. It is ironic that the desire to lose weight coupled with ineffectual maintenance of gains may bring negative health consequences, but this is the picture emerging from research on weight cycling.

It should be noted that in the present study, weight cycling was not found to moderate the degree of treatmentrelated benefit associated with 3 indicators of biomedical health status. Specifically, the history of yo-yo dieting was unrelated to pre- to posttreatment improvements in serum levels of glucose, triglycerides, and total cholesterol. These null findings are consistent with previous investigations that have also unsuccessfully attempted to relate weight fluctuations in obese adults to measures of serum glucose and total cholesterol levels.31,35-37 It is possible that methodological limitations may account for these null effects. In this connection, results from a recent study by Olson et al.34 suggest it may be more appropriate to conduct analyses that involve more specific blood chemistries, such as levels of high-density versus low-density lipoprotein cholesterol. Advancements in the sensitivity of biomedical assays provide the opportunity for future studies to explore the relationship of weight cycling to more specific products of glucose and fatty acid metabolism.

Study Limitations and Directions for Future Research

In the present study, we defined history of weight cycling as the frequency of significant weight loss associated with participation in formal weight-loss programs. Historical frequency of participation in community-based weight-loss initiatives appears a more precise way of assessing weight cycling than merely asking about "weight fluctuations." Attendance at a program signals intentional weight-loss efforts, whereas reporting weight fluctuations may signal either intentional or unintentional weight loss (eg, loss attributed to medical or psychiatric illness). Nevertheless, our operationalization captured only the frequency dimension of weight cycling associated with publicly enacted efforts to lose weight. Future studies may consider refining further the operational definition of weight cycling to include both the frequency of formal and informal intentional weight loss attempts (caloric restrictions and expenditures pursued in private), as well as the magnitude of weight lost and regained during each weight- cycling episode.50,52 Information on the time course of these cycles of loss and regain may also prove useful.

In future research, increased attention needs to be given to understanding the maintenance of weight loss in clinical and nonclinical samples with and without a history of weight cycling. Research involving clinical samples would examine patients who have sought professional assistance. To compliment this line of investigation, studies could also be done with obese people who are pursuing informal and individualized programs of "Natural Change." There is very little work with humans to identify the influence of weight cycling history on patterns of posttreatment weight regain (Kroke et al.15 provide an exception). Establishing such a link with improved measures of weight cycling would clearly represent a fruitful direction for future research. In this connection, care should be given to examining the potential health impairment associated with weight regain. If, as we have documented, weight loss is associated with widespread biomedical health benefits, weight regain may be associated with the opposite. By way of speculation, it may be found that mortality risk attributed to weight cycling may operate, in part, by way of one or more of the particular types of biomedical health processes identified in the present research to be associated uniquely with a history of cycling.

NOTE

For comment or further information, please address correspondence to Kenneth E. Hart, PhD, Director, Center for Psychological Intervention and Research, Department of Psychology, University of Windsor, 401 Sunset Avenue, Windsor, ON, Canada N9B 3P4 (e-mail: kenhart@uwindsor.ca). Erin Warriner is now Dr Warriner and is now affiliated with Hamilton Health Sciences.

REFERENCES

1. Blackburn GL. Benefits of weight loss in the treatment of obesity. Am J Clin Nutr. 1999;69:341-349.

2. Brownell KD, Wadden TA. Etiology and treatment of obesity: understanding a serious, prevalent, and refractory disorder. J Consult Clin Psychol. 1992;60:505-517.

3. Dattilo AM, Kris-Etherton PM. Effects of weight reduction on blood lipids and lipoproteins: a meta-analysis. Am J Clin Nutr. 1992;56:320-328.

4. Goodrick GK, Foreyt JP. Why treatments for obesity don't last. J Am Diet Assoc. 1991;91:1243-1247.

5. McGuire MT, Wing RR, Hill JO. The prevalence of weight loss maintenance among American adults. Int J Obes. 1999;23:1314-1319.

O.Wilson GT. Behavioral treatment of obesity: thirty years and counting. Adv Behav Rese Ther. 1994;16:31-75.

7. Brownell KD, Greenwood MRC, Stellar E, Shrager EE. The effects of repeated cycles of weight loss and regain in rats. Physiol Behav. 1986;38:459-464.

8. Gerardo-Gettens T, Miller GD, Horwitz BA, et al. Exercise decreases fat selection in female rats during weight cycling. Am J Physiol. 1991;260:R518-R524.

9. Blackburn GL, Wilson GT, Kanders BS, et al. Weight cycling: the experience of human dieters. Am J Clin Nutr. 1989;49:1105-1109.

10. Jebb SA, Goldberg GR, Coward WA, Murgatroyd PR, Prentice AM. Effects of weight cycling caused by intermittent dieting on metabolic rate and body composition in obese women. Int J Obes. 1991;15:367-374.

11. Gray DS, Fisler JS, Bray GA. Effects of repeated weight loss and regain on body composition in obese rats. Am J Clin Nutr. 1988;47:393-399.

12. Reed GW, Cox G, Yakubu F, Ding L, Hill JO. Effects of weight cycling in rats allowed a choice of diet. Am J Physiol. 1993;264:R35- R40.

13. van Dale D, Saris WHM. Repetitive weight loss and weight regain: effects on weight reduction, resting metabolic rate, and lipolytic activity before and after exercise and/or diet treatment. Am J Clin Nutr. 1989;49:409-416.

14. Wadden TA, Bartlett S, Letizia KA, Foster GD, Stunkard AJ, Conill A. Relationship of dieting history to resting metabolic rate, body composition, eating behavior, and subsequent weight loss. Am J Clin Nutr. 1992;56:S203-S208.

15. Kroke A, Liese AD, Schulz M, et al. Recent weight changes and weight cycling as predictors of subsequent two year weight change in a middle-age cohort. Int J Obes. 2002;26:403-409.

16. Blair SN, Shaten J, Brownell KD. Collins G, Lissner L. Body weight fluctuation, all-cause mortality, and cause-specific mortality in the Multiple Risk Factor Intervention Trial. Ann Intern Med. 1993;119:749-757.

17. Hamm P, Shekelle RB, Stamler J. Large fluctuations in body weight during young adulthood and twenty-five year risk of coronary death in men. Am J Epidemiol. 1989;129:312-318.

18. Lissner L, Odell PM, D'Agostino RB. et al. Variability of body weight and health outcomes in the Framingham population. N Engl J Med. 1991;324:1839-1844.

19. Reynolds MW, Fredman L, Langenberg P, Magaziner J. Weight, weight change, and mortality in a random sample of older community- dwelling women. J Am Geriutr Soc. 1999;47:1409-1414.

20. Berg FM. Is yo-yo dieting dangerous? In: Nolen-Hoeksema S, ed. Clashing Views on Abnormal Psychology: A Taking Sides Custom Reader. Guilford, CT: Dushkin/McGraw-Hill; 1998. 156-173.

21. Wing RR. Weight cycling in humans: a review of the literature. Ann Behav Med. 1992;14:113-119.

22. Pi-Sunyer FX. Health implications of obesity. Am J Clin Nutr. 1991;53:1595S-1603S.

23. Manore MM, Berry TE, Skinner JS, Carroll SS. Energy expenditure at rest and during exercise in non-obese female cyclical dieters and in non-dieting control subjects. Am J Clin Nutr. 1991;54:41-46.

24. Kajioka T, Tsuzuku S, Shimokata H, Sato Y. Effects of intentional weight cycling on non-obese young women. Metabolism. 2002;51:149-154.

25. Reed DR, Contreras RJ, Maggio C, Greenwood MRC. Rodin J. Weight cycling in female rats increases dietary fat selection and adiposity. Physiol Behav. 1988;42:389-395.

26. Szepesi B. Effect of frequency of caloric deprivation on the success of growth compensation. Nutr Rep Inter. 1980;21:479-486.

27. Rodin J, Radke-Sharpe N, Rebuffe-Scrive M, Greenwood MRC. Weight cycling and fat distribution. Int J Obes. 1990;14:303-310.

28. Drewnowski A, Holden-Wiltse J. Taste responses and food preferences in obese women: effects of weight cycling. Int J Obes. 1992;16:639-648.

29. Drewnowski A, Kurth CL, Rahaim JE. \Taste preferences in human obesity: environmental and familial factors. Am J Clin Nutr. 1991;54:635-641.

30. Reed GW, Hill JO. Weight cycling: a review of the animal literature. Obes Res. 1993;1:392-402.

31. Jeffery RW, Wing RR, French SA. Weight cycling and cardiovascular risk factors in obese men and women. Am J Clin Nutr. 1992;55:641-644.

32. Pasanisi F, Contaldo F, de Simone G, Mancini M. Benefits of sustained moderate weight loss in obesity. Nutr Metab Cardiovasc Dis. 2001;11:401-406.

33. Sea MM, Fong WP, Huang Y, Chen ZY. Weight cycling-induced alteration in fatty acid metabolism. Am J Physiol. 2000;279:R1145- R1155.

34. Olson MB, Kelsey SF, Bittner V, et al. Weight cycling and high-density lipoprotein cholesterol in women: evidence of an adverse effect-A report from the NHLBI-sponsored WISE study. J Am Coll Cardiol. 2000;36:1565-1571.

35. Barlow CE, Kohl HW, Blair SN. Weight fluctuation and cardiovascular risk factors in obese men and women. Med Sci Sports Exerc. 1991;23:S53.

36. Lissner L, Andres R, Muller DC, Shimokata H. Body weight variability in men: metabolic rate, health and longevity. Int J Obes. 1989;14:373-383.

37. Podar T, Solntsev A, Vali M, Vinogradova T, Podar I. No deterioration of glucose tolerance in weight cycling obese. Int J Obes. 1996;20:921-924.

38. Ernsberger P, Nelson DO. Refeeding hypertension in dietary obesity. Am J Physiol. 1988;254:R47-R55.

39. Contreras RJ, Williams A. High fat/sucrose feeding attenuates the hypertension of spontaneously hypertensive rats. Physiol Behav. 1989;46:285-291.

40. Contreras RJ, King S, Rives L, Williams A, Wattleton T. Dietary obesity and weight cycling in rats: a model of stress- induced hypertension? Am J Physiol. 1991;262:R848-R857.

41. Guagnano MT, Ballone E, Pace-Palitti V, et al. Risk factors for hypertension in obese women: the role of weight cycling. Eur J Clin Nutr. 2000;54:356-360.

42. Guagnano MT, Pace-Palitti V, Carrabs C, Merlitti D, Sensi S. Weight fluctuations could increase blood pressure in android obese women. Clin Sci. 1999;96:677-680.

43. Field AE, Byers T, Hunter DJ, et al. Weight cycling, weight gain, and risk of hypertension in women. Am J Epidemiol. 1999;150:573-579.

44. Petersmarck KA, Teitelbaum HS, Bond JT, Bianchi L, Hoerr SM, Sowers MF. The effect of weight cycling on blood lipids and blood pressure in the Multiple Risk Factor Intervention Trial Special Intervention Group. Int J Obes. 1999;23:1246-1255.

45. Brownell KD, Rodin J. Medical, metabolic, and psychological effects of weight cycling. Arch Intern Med. 1994;154:1325-1330.

46. Raymond NC, de Zwaan M, Mitchell JE, Ackard D, Thuras P. Effects of a very low calorie diet on the diagnostic category of individuals with binge eating disorder. Int J Eat Disord. 2002;31:49- 56.

47. Wadden TA, Van Itallie TB, Blackburn GL. Responsible and irresponsible use of very-low calorie diets in the treatment of obesity. JAMA. 1990;263:83-85.

48. Bartlett SJ, Wadden TA, Vogt RA. Psychosocial consequences of weight cycling. J Consult Clin Psychol. 1996;64:587-592.

49. Foster GD, Wadden TA, Kendall PC, Stunkard AJ, Vogt RA. Psychological effects of weight loss and regain: a prospective evaluation. J Consult Clin Psychol. 1996;64:752-757.

50. Simkin-Silverman LR, Wing RR, Plantinga P, Mathews KA, Kuller LH. Lifetime weight cycling and psychological health in normal- weight and overweight women. Int J Eat Disord. 1998;24:175-183.

51. Beck AT, Ward CH, Mendelson M, Mock J, Erbaugh J. An inventory for measuring depression. Arch Gen Psychiatry. 1961;4:561- 571.

52. Friedman MA, Schwartz MB, Brownell KD. Differential relation of psychological functioning with the history and experience of weight cycling. J Consult Clin Psychol. 1998;66:646-650.

Kenneth E. Hart, PhD, and Erin M. Warriner, MA

Dr Hart and Ms Warriner are with the University of Windsor, Ontario, Canada.

Copyright HELDREF PUBLICATIONS Winter 2005

Source: Behavioral Medicine

More News in this Category