| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Department of Internal Medicine, University of Kentucky, Lexington, Kentucky (J.W.A.)
NutriPharma, Oslo, Norway (L.H.H.)
Address reprint requests to: James W. Anderson, MD, 1030 South Broadway, Lexington, KY 40504-2681. E-mail: jwandersmd{at}aol.com
 |
ABSTRACT |
|---|
 TOPFOOTNOTES ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSREFERENCES |
|---|
Methods: Overweight or obese women and men (body mass indices 27–40 kg/m2) were randomly assigned to LED providing 1200 kcal/day, with consumption of five soy-based or two milk-based liquid MR for a 12-week weight loss trial. Serum lipoprotein measurements were obtained at baseline, 6 and 12 weeks.
Results: For soy and milk MR groups, subject numbers were, respectively, 51 and 39 randomized and 30 and 22 completers. MR use averaged: soy, 3.7/day; and milk, 1.9/day. Weight losses for completers at 12 weeks were: soy MR, 9.0% of initial body weight (95% confidence intervals, 7.3–10.6%); and milk MR, 7.9% (5.8–8.8%) with no significant differences. Reductions from baseline in serum cholesterol and LDL-cholesterol values, respectively, at six weeks were significantly greater (P < 0.015) with soy MR (15.2% and 17.4%) than with milk MR (7.9% and 7.7%). Soy MR use was associated with significant reductions in serum triglycerides at 6 and 12 weeks while milk MR use was not.
Conclusions: Soy MR use, as part of a low-energy diet, was associated with slightly but not significantly greater weight loss over a 12-week period than milk MR use. These observations confirm previous studies documenting the effectiveness of MR use for weight loss.
Key words: clinical trial, obesity, meal replacements, serum cholesterol, soy protein, weight loss
|
INTRODUCTION |
|---|
|
TOPFOOTNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSREFERENCES |
|---|
Soy protein has been used as a principal ingredient of liquid MR for treatment of obese individuals for two decades [9,10]. Soy protein has been the protein component of a widely used VLED in Scandinavia over the past 15 years [11,12]. Laboratory studies—in vitro and with animal models—suggest that soy protein has selective effects on upregulation of genes involved in glucose and lipid metabolism, enhances insulin sensitivity, and promotes a select loss of visceral adipose tissue [10]. Furthermore, studies in humans indicate the soy protein has specific effects on serum lipoproteins [13], increases insulin sensitivity [14] and may protect from development of diabetes [15].
The purpose of this study was to compare weight loss and serum lipid changes with two popular MR that are widely available and are purchased by consumers for self-help weight loss efforts. One MR shake was milk based and the other soy based. These products were provided to volunteers with minimal lifestyle counseling or behavioral intervention to mimic the self-help activities that these individuals might perform after purchase of these products.
|
MATERIALS AND METHODS |
|---|
|
TOPFOOTNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSREFERENCES |
|---|
4 kg) in the past 6 months; current use of medication or herbal product for weight loss; prior surgical intervention for the treatment of obesity; history of severe renal, hepatic, neurological, chronic pulmonary disease, or any other unstable medical disorder; depression, panic disorder, psychosis, or bipolar disorder; history of alcohol or substance abuse; or risk of non-compliance with study procedures. Each subject signed and dated an informed consent form that was witnessed before undertaking any screening procedures. This protocol was approved by the Institutional Review Board at the University of Kentucky.
Protocol
This was a 12-week, randomized, controlled, parallel group study. Because of differences in the number of shakes used daily and the preparation of the shakes, blinding of the study was not possible. Subjects volunteered for the study after seeing newspaper advertisements, hearing about the study or being contracted because of a prior interest in this type of research. After screening by telephone, subjects were scheduled for a screening visit where the informed consent was reviewed and signed. A detailed history and medication review was performed; blood and urine specimens were obtained for screening. Eligible subjects returned for a second visit for review of laboratory results and physical examination. Eligible subjects were scheduled for the randomization visit. At the second visit subjects were given a lifestyle diary and instructed to record food intake and physical activity. At randomization, baseline physical and laboratory measurements were completed. Subjects were assigned screening numbers consecutively at the initial clinic visit. They were randomly assigned to treatment groups based their screening number and a random number series. Study staff were blinded to assignment until the subject was randomized to the study product group.
Subjects were randomized to use either five soy-based meal replacements (Soy MR) or two milk-based meal replacements (Milk MR) per day. A registered dietitian instructed each subject in use of a 1200 kcal diet to include the MR. Subjects randomized to the Soy MR were encouraged to include fruits and vegetables while subjects using Milk MR were instructed to include fruits, vegetables, and two servings of lean meat such as breast of chicken or turkey. Soy MR (Scan-DietTM) is a widely available powder that was mixed with water to form a shake; subjects were instructed to take five Soy MR shakes daily. The Milk MR (Slim-Fast®) is a widely available powder that is mixed with 8 ounces of skim milk to make a shake; subjects were instructed to take two shakes daily. The products were used as directed by the product information. Nutrient information for the two powdered MR is presented in Table 1. To more closely mimic the practice that consumers would follow, only limited nutrition information was provided to subjects. They were given a lifestyle diary and instructed to record MR use, food intake, and physical activity in 10 minute blocks. The main purpose of using the lifestyle diary was to collect information about MR use since detailed food intake information was not collected.
|
Quest Diagnostic Laboratories, Louisville, KY, performed all laboratory procedures.
Statistical Analyses
The primary endpoints were the absolute and percentage change from baseline body weight at 6 and 12-weeks and change in serum cholesterol and LDL-cholesterol values between baseline and 6 and 12 weeks. The conservative power analysis approach led to an unpaired two-way t-test analysis. Sample size calculations were based on data from previous clinical trials at the University of Kentucky [16]. Since we did not anticipate large differences in weight loss between groups we did power calculations based on differences between serum cholesterol changes (6 or 12-week minus baseline) comparing changes for Soy MR with those for Milk MR. We calculated the minimum sample size to detect a difference in serum cholesterol of 3% and assumed a standard deviation of 5%. We calculated that 30 subjects would be required per treatment group to detect a 3% difference with a significance of 
= 0.05 and a power of 80%.
All continuous variables were analyzed as change from baseline. We used the mixed model approach and analyzed data for completers and for the last-observation-carried-forward (LOCF) or intention-to-treat (ITT) group. Triglyceride values were analyzed without and with logarithmic transformation.
|
RESULTS |
|---|
|
TOPFOOTNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSREFERENCES |
|---|
Randomization
Because of high early dropout rates in the Soy MR group, subjects were randomized at a 2:1 ratio to Soy MR vs. Milk MR during the middle third of the study. Fifty-one subjects were assigned to Soy MR and 39 subjects were assigned to Milk MR.
Demographic Characteristics of Subjects
The Soy MR subjects included 47 women and 4 men while the Milk MR included 32 women and 7 men. The average ages were: Soy MR, 47.3 ± 9.4 (SD) years and Milk MR, 47.6 ± 9.4 years. The average BMIs were: Soy MR, 34.8 ± 3.2 kg/m2 and Milk MR, 34.2 ± 2.7 kg/m2. Mean ages and BMIs did not differ significantly between diets.
Early Withdrawals
Thirty of 51 subjects (58.8%) completed the Soy MR intervention while 22 of 39 subjects (56.4%) completed the Milk MR intervention. Subjects withdrew from the Soy MR intervention earlier than with Milk MR. Before the return visit at 2 weeks 9 Soy MR subjects had withdrawn while only 3 Milk MR subjects had withdrawn. The early withdrawals from the Soy MR intervention were related to intolerance to the product. However, more Milk MR withdrew later in the study so that overall withdrawal rate did not differ significantly; 21 of 51 subjects (41.2%) did not complete the Soy MR intervention while 17 of 39 subjects (43.6%) did not complete the Milk MR intervention.
Ten subjects withdrew early related to side effects; 3 subjects reported increased gas or more frequent bowel movements with the Soy MR. Seven subjects withdrew early because they could not tolerate the supplement (6 Soy MR and 1 Milk MR). Eleven subjects withdrew reporting they were too busy to continue participation (3 Soy MR and 8 Milk MR). Seventeen subjects did not return for follow-up and did not return phone calls (9 Soy MR and 8 Milk MR).
Supplement and Lifestyle Diary Compliance
Subjects maintained lifestyle diaries in an excellent (defined as >90%) fashion during the first six weeks. The completion rates of diaries 6- and 12-weeks were as follows: Soy MR, 91.4% and 80.0%, respectively; and Milk MR, 90.6% and 86.4%, respectively. With the Soy MR intervention, subjects reported a lower percentage of recommended intake of shakes than with Milk MR. The recommended use of Soy MR was 5 shakes daily; subjects consumed an average of 3.7 shakes daily at 5 and 6 weeks and an average of 3.7 shakes daily at 11 and 12 weeks. The recommended use of Milk MR was 2 shakes daily; subjects consumed 1.9 shakes at 5 and 6 weeks and 1.9 shakes at 11 and 12 weeks.
Weight Loss
Weight losses by week are summarized in Table 2. Subjects using Soy MR lost more weight at each week than with Milk MR use but these differences were not significant. With the Soy MR, the weight losses as percentage of initial body weight were as follows: 2 weeks, 2.6%; 4 weeks, 4.0%; 6 weeks, 5.8%; 8 weeks, 6.7%; 10 weeks, 7.5%; and 12 weeks, 9.0% (95% confidence intervals (CI), 7.3 to 10.6%). With the Milk MR, the weight losses were as follows: 2 weeks, 2.5%; 4 weeks, 3.7%; 6 weeks, 5.1%; 8 weeks, 5.9%; 10 weeks, 6.8%; and 12 weeks, 7.9% (95% CI, 5.9 to 9.9%).
|
Waist Circumference Changes
Both the Soy MR and the Milk MR groups had significant reductions in waist circumferences (P < 0.0001). The Soy MR group had larger reductions in waist circumferences (–10.6%) than the Milk MR group (–8.0%) but these differences did not differ significantly (P = 0.21).
Serum Cholesterol Changes
Cholesterol changes were larger at 6 weeks than 12 weeks with both diets (Table 3). Both diets significantly decreased values at 6 weeks (P < 0.001). The Soy MR was associated with consistently lower serum cholesterol values than the Milk MR. At 6 weeks cholesterol values were –15.2% (95% CI, –11.3 to –19.2%) with the Soy MR and –7.9% (95% CI, –4.7 to –11.1%) with Milk MR. This difference between Soy MR and Milk MR was statistically significant (P = 0.0069).
|
Serum LDL-Cholesterol Changes
LDL-cholesterol values followed the same pattern as cholesterol values. Values for both interventions were significantly reduced from baseline at 6 weeks (P < 0.001) but only Soy MR values were significantly reduced at 12 weeks. The values for Soy MR were as follows: 6 weeks, –17.4% (95% CI, –12.1 to –22.7%); and 12 weeks, –7.5% (–1.5 to –13.5%). The values for Milk MR were as follows: 6 weeks, –7.7% (95% CI, –2.3 to –13.0%); and 12 weeks, –2.4% (6.3 to –11.0%). This difference between Soy MR and Milk MR at 6 weeks was statistically significant (P = 0.0136).
Serum Triglyceride Changes
Triglyceride values were significantly reduced at weeks 6 and 12 with the Soy MR but not with Milk MR. At 6 weeks, values for the Soy MR were –13.4% (95% CI, –4.1 to –22.7%) while values for Milk MR were 0.8% (95% CI, –11.4 to 12.9%). At 12 weeks values for the Soy MR were –20.1% (95% CI, –8.9 to –31.3%) while values for Milk MR were 0.4% (95% CI, 17.2 to –21.1%). The analysis of log-transformed triglyceride values yielded similar results. Milk MR intake was not associated with significant changes in serum triglyceride values whereas Soy MR intake was associated with significant reductions at 6 weeks (P = 0.0006) and 12 weeks (P < 0.0001).
Serum HDL-Cholesterol Changes
At 6 weeks reductions in HDL-cholesterol values with Milk MR were significant while reductions with Soy MR were not. At 6 weeks, values for the Soy MR were –2.9% (95% CI, +2.6 to –8.3) while values for Milk MR were –6.0% (95% CI, –1.4 to –10.6%). At 12 weeks values for the Soy MR were –1.3% (95% CI, +4.5 to –7.2%) while values for Milk MR were –1.3% (95% CI, +5.6 to –8.2%).
Serum Glucose Changes
At 12 weeks reductions in glucose were significant with the Soy MR (–3.6%, 95% CI, –1.1 to –6.1%) but not with Milk MR (–2.9%, 95% CI, +0.2 to –6.0%); these changes between Soy MR and Milk MR did not differ significantly.
Blood Pressure Changes
Both diet interventions were accompanied by small reductions in average blood pressure that were maximal at 6 weeks. While reductions in systolic blood pressure were significant at 6 weeks with the Soy MR, differences between diets were not significant. At 6 weeks the following systolic and diastolic blood changes were noted: Soy MR, –6.2 (95% CI, –2.5 to –9.9) and –1.7 mmHg, not significant (ns); Milk MR, –2.7 (ns) and –1.1 mmHg (ns). At 12 weeks the following systolic and diastolic blood changes were noted: Soy MR, –1.5 (ns) and –3.1 mmHg (ns); Milk MR, –4.4 (ns) and –2.4 mmHg (ns).
Side Effects
No serious adverse events or side effects were reported. Three subjects reported increased belching or flatulence and more frequent bowel movements on the Soy MR.
|
DISCUSSION |
|---|
This study confirms the efficacy and safety of meal replacements for weight loss for obese individuals. This comparator trial evaluated the efficacy of the Soy MR intervention with the Milk MR intervention. A limitation of the study is that blinding was not possible because the Soy MR group was instructed to use 5 packets daily while the Milk MR group was instructed to use 2 packets daily. Furthermore, the Soy MR powder was mixed in water while the Milk MR powder was mixed in skim milk. Nevertheless, we compared the two interventions as consumers would be instructed to use them based on the product information.
Overall dropout rates were higher than usually seen in our clinical trials based on published [16] and unpublished data (Anderson JW, unpublished observations). For example, in a similar trial we enrolled 71 subjects and 56 (79%) completed the 12-week intervention (Anderson JW, unpublished observations). By design, the behavioral intervention was less intense than we usually employ because consumers will ordinarily use these products without nutrition counseling. For example, we did not employ our usual practice of providing pedometers and monitoring daily steps or miles walked [16]. However, dropout rates did not differ significantly between the two groups.
Compliance to lifestyle diary use was excellent in both diet groups at 6 weeks and did not differ significantly. At six weeks, over 90% of subjects had maintained diary entries and reported number of shakes used daily. At 12 weeks, 77.8% of subjects using the Soy MR and 86.4% of subjects using the Milk MR were compliant with lifestyle diary use. Overall, compliance to lifestyle diary use was 85.5% with Soy MR and 88.9% with Milk MR. Most subjects did not use 5 packets of Soy MR daily and the average use was 3.7 packets daily (74% of recommended). However, most subjects used 2 packets of Milk MR daily and the average use was 1.9 packets daily (95% of recommended). The Soy MR shakes were considered less acceptable than the Milk MR shakes. However, dropout rates after 6 weeks were lower with the Soy MR (8 of 35 subjects, 22.9%) than with Milk MR (10 of 32 subjects, 31.3%).
The weight loss of 
8.5% for the combined groups compares very favorably with values reported in the literature. Allison reported weight losses of 7.6% of initial body weight using a similar protocol with Soy MR [24]. In our recent meta-analysis of weight loss from four studies including 405 non-diabetic subjects consuming two Milk MR daily the average weight loss at 12 weeks was 6.9% (95% CI, 6.0–7.8%) [16]. Heymsfield and colleagues [8] analyzed data from six studies including 403 non-diabetic and diabetic subjects consuming two Milk MR daily and reported weight loss at 12 weeks of 7.1%.
Serum cholesterol and LDL-cholesterol values were consistently lower with Soy MR than with Milk MR. This is consistent with the many reports reviewed in our meta-analysis [13] related to the hypocholesterolemic effects of soy protein and the earlier reports with weight-reducing diets employing soy protein [25,26]. Allison [24] also reported that serum cholesterol and LDL-cholesterol reductions were larger with the Soy MR intervention than with the control diet; the LDL-cholesterol differences were persistent after adjusting for weight loss.
Serum triglyceride values were favorably affected by soy protein intake [13]. The Soy MR was associated with significant reductions in serum triglycerides at 6 and 12 weeks while the Milk MR diet, even with significant weight reduction, was not accompanied by significant reductions in serum triglycerides. With weight loss in a predominately female group, reductions in serum HDL-cholesterol values are often seen [27]. However, changes with the Soy MR were small and insignificant while significant reductions were observed with Milk MR at 6 weeks. Small but significant reductions in fasting serum glucose values were seen with the Soy MR but not with Milk MR at 12 weeks. These observations are consistent with the emerging evidence that soy protein intake increases insulin sensitivity [14] and may even act to slow the development of diabetes [15].
Subjects taking four to five shakes daily had more modification of their eating behavior than those taking two shakes daily; this may have contributed to differences in weight loss. In addition to differences in source of protein, there may have been other differences in the nutrient intake of the two study groups. Subjects in the Soy MR group consumed more protein, calcium and fiber with their product than did those in the Milk MR group. These differences in protein [28], calcium [29,30] or fiber [31,32] may have contributed to weight loss differences.
The weight loss of 8.5% of initial body weight at 12 weeks using combined data from both MR groups is equivalent to or greater than would be expected with intensive behavioral interventions. In our meta-analysis of 10 behavioral programs using low-energy diets (
1500 kcal/day) the average weight loss at 12 weeks was 7.9% [9]. Furthermore, the observed weight loss of 8–9% of initial body weight with the meal replacements is approximately twice that reported with pharmacotherapy [33]. Haddock [33] asserts that net weight loss (i.e., treatment minus control weight loss) never exceeds 4% of initial body with pharmacotherapy.
|
FOOTNOTES |
|---|
|
TOPFOOTNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSREFERENCES |
|---|
Received December 7, 2004. Accepted March 20, 2005.
|
REFERENCES |
|---|
|
TOPFOOTNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  M.-P. St-Onge and A. Bosarge Weight-loss diet that includes consumption of medium-chain triacylglycerol oil leads to a greater rate of weight and fat mass loss than does olive oil Am. J. Clinical Nutrition, March 1, 2008; 87(3): 621 - 626. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
