Maintenance Following a Very-Low-Calorie Diet W. Stewart Agras Department of Psychiatry Stanford University School of Medicine Robert I. Berkowitz Department of Psychiatry Stanford University School of Medicine Bruce A. Arnow Department of Psychiatry Stanford University School of Medicine Christy F. Telch Department of Psychiatry Stanford University School of Medicine Margaret Marnell Department of Psychiatry Stanford University School of Medicine Judith Henderson Department of Psychiatry Stanford University School of Medicine Yvonne Morris Department of Psychiatry Stanford University School of Medicine Denise E. Wilfley Department of Psychiatry Stanford University School of Medicine ABSTRACT The authors posed 2 questions in this randomized study of maintenance procedures in which participants were followed for 15 months after completion of a very-low-calorie diet: Would stimulus narrowing during the reintroduction of solid food, achieved by the use of prepackaged foods, improve weight losses and the maintenance of those losses as compared with the use of regular food? Would reintroduction of foods dependent on progress in losing or maintaining weight be superior to reintroduction on a time-dependent basis? Neither the stimulus narrowing condition nor the reintroduction procedure enhanced either maximum weight loss or maintenance of those losses. The stimulus narrowing condition appeared to be poorly tolerated; compliance and attendance were poorer in this condition than in the regular food condition. This research was supported in part by National Institutes of Health Grant DK 39673. Optifast-800 was generously provided by the Sandoz Corporation, Minneapolis, Minnesota. Correspondence may be addressed to W. Stewart Agras, Department of Psychiatry, Stanford University School of Medicine, 401 Quarry Road, Room 1322, Stanford, California, 94305-5544. Received: April 20, 1995 Revised: August 11, 1995 Accepted: August 17, 1995 As Stamler (1993) pointed out in a recent editorial, obesity has reached epidemic proportions in the United States. Moreover, obesity is accompanied by a variety of serious health problems as well as increased morbidity and mortality ( Bray, 1981 ; Manson et al., 1990 ; Must, Jacques, Dallai, Bajema, & Dietz, 1992 ). The introduction of the very-low-calorie diet (VLCD), which results in relatively large Journal of Consulting and Clinical Psychology © 1996 by the American Psychological Association June 1997 Vol. 64, No. 3, 610-613 For personal use only--not for distribution. Page 1 of 711/4/2000http://spider.apa.org/ftdocs/ccp/1996/june/ccp643610.htmlinitial weight losses averaging 15-20 kg in 12 weeks ( Vertes, Genuth, & Hazelton, 1977 ; Wadden, Van Itallie, & Blackburn, 1990 ), raised hopes that an effective treatment was available for the moderately and severely obese. Unfortunately, as Wilson (1994) pointed outand a recent study with the VLCD underlined ( Wadden, Foster, & Letizia, 1994 ), weight regain tends to be inexorable. In the latter study patients treated with a VLCD lost 21.45 kg but regained almost half of that amount in the year following the end of the fast. It may be that the element leading to the success of the VLCD, that is, caloric restriction in the context of much narrowed food cues, also leads to its failure during the reintroduction of normal foods when patients are faced with an increasing variety of food cues. This possibility led to the principal hypothesis underlying the present study of weight maintenance following a VLCD, namely, that continued narrowing of food stimuli during the reintroduction of normal foods might be associated with continued weight loss and enhanced maintenance of losses. This reasoning was based on the principle of sensory-specific satiety ( Rolls, 1986 ). Both animal and human studies have demonstrated that increasing the variety of foods leads to increased food consumption, which if maintained may lead to obesity. Conversely, decreasing variety leads to decreased consumption ( Kanarak & Hirsch, 1977 ; Rolls, Rowe, Kingston, Megson, & Gunary, 1981 ). Hence, one aim of the present study was to compare reintroduction of solid food following a VLCD under two conditions: with a narrowed range of food stimuli (prepackaged foods) and with regular food. On the basis of the existing literature, it was hypothesized that the narrowed food stimuli would result in greater weight losses and improved maintenance of weight losses as compared with regular food&period It was also hypothesized that in both conditions individualized pacing of the reintroduction on the basis of weight maintenance or further loss (i.e., if weight gain occurred the reintroduction process would be slowed) would prove more effective than a time-based procedure in which food reintroduction would occur regardless of weight changes. These hypotheses led to the following experimental design. All participants in this study first received a 12-week VLCD. If they lost 5% or more of their initial weight, they were allocated at random to one of the following four conditions: (a) refeeding with standard food–time dependent; (b) refeeding with standard food–weight dependent; (c) refeeding with prepackaged food (stimulus narrowing)–time dependent; and (d) refeeding with prepackaged food (stimulus narrowing)–weight dependent. All participants received behavior therapy for 9 months following completion of the VLCD and were then followed for 6 months. Method Participants Two hundred one overweight women were entered into this study. Their mean age was 43.7 years SD = 10 , the average age of reported onset of overweight was 19.4 years SD = 11.7 , and their weight on entry to the VLCD phase of the study was 100.3 kg SD = 14 with a body mass index (BMI) of 36.6 kg/ m 2 SD = 4.4 . Forty-nine percent of the sample had a college degree and more than half of these had attended graduate school; 40% had some college education; and the remainder had completed high school. Before beginning treatment all participants were medically screened, including a full history, physical examination, mental status examination, electrocardiogram, and appropriate blood tests. Exclusion criteria included recent myocardial infarction, major cardiac arrhythmia, or stroke; type-II diabetes not controlled with oral hypoglycemic agents; bleeding peptic ulcer; or other serious disorders that may complicate dieting (such as liver or kidney disease; evidence of