The costs of publication of this article were defrayed, in part, by the payment of page charges. This article must, therefore, be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Food Science and Human Nutrition (FSHN) Department/Nutritional Sciences Division, University of Illinois, 445 Bevier Hall, 905 S. Goodwin Avenue, Urbana, IL 61801. E-mail: firstname.lastname@example.org
Objective: To determine the efficacy of a weight-loss diet using packaged portion-controlled entrees compared with a self-selected diet based on the U.S. Department of Agriculture Food Guide Pyramid (FGP) (United States Department of Agriculture, Center for Nutrition Policy and Promotion, Washington, DC; 1996).
Research Methods and Procedures: Sixty healthy women (BMI 26 to 40 kg/m2; 24 to 60 years old) were randomized into two intervention groups for an 8-week parallel arm study. The portion-controlled group consumed two frozen entrees daily, plus additional food servings from the FGP. The self-selected diet group consumed a recommended number of servings from the FGP. Diets were designed to be the same in composition (55% carbohydrate, 25% protein, 20% fat) and energy level (1365 kcal). Each group met weekly to monitor compliance and take measures. Outcomes included weight, body composition by DXA, hip and waist circumference, blood pressure, fasting blood lipids, glucose, insulin, and C-reactive protein. Significant differences were assessed using repeated measures ANOVA.
Results: The portion-controlled group (n = 26) experienced greater decreases in weight (5.6 ± 2.2 kg or 6.5% vs. 3.6 ± 2.5 kg or 4.2%), fat mass (3.6 ± 1.8 vs. 2.3 ± 1.4 kg), total cholesterol (24.4 ± 21.5 mg/dL or 12.4% vs. 13.0 ± 13.9 mg/dL or 6.7%), and fasting insulin (−1.8 ± 3.7 vs.+0.3 ± 3.8 μU/mL) than the self-selected diet group (n = 27) (p < 0.05).
Discussion: Consumption of portion-controlled entrees resulted in greater losses of weight and fat, thereby reducing cardiovascular disease risk. Accurate portion control is an important factor in weight loss success, and use of packaged entrees is an effective method of achieving this.
The incidence of obesity has been increasing worldwide over the past 40 years. Currently, in the United States, more than one-half of the population is overweight (1). Obesity is a major risk factor for cardiovascular disease (CVD)1 and is directly correlated with other risk factors such as hypertension, dyslipidemias, and diabetes (2, 3, 4, 5). As noted in a recent hospital-based study, well-controlled dietary interventions can lead to both weight loss and substantial reductions in CVD risk (6). However, this kind of success is difficult for individuals to achieve independently. The U.S. Department of Agriculture created the Food Guide Pyramid (FGP) as a visual representation of a healthful diet, with recommended amounts to consume from the various food groups (7) This is a helpful tool, but it still requires the user to make food choices and estimate portion sizes, which can be quite challenging.
In the late 1990s, a series of multicenter trials, conducted with subjects having one or more CVD risk factors, compared a self-selected diet to a prepared meal plan designed to be nutritionally equivalent. A 10-week trial was followed by a 52-week trial involving 2 cohorts of high-risk subjects: persons with hypertension and dyslipidemias and persons with type 2 diabetes. Compliant subjects in both groups experienced benefits. However, compliance was significantly better with the prepared meals, leading to greater reductions in weight, plasma lipids, lipoprotein, blood pressure (BP), glucose, and glycosylated hemoglobin (8, 9, 10). The researchers concluded that the prepared meals eliminated several of the obstacles faced by the self-selected diet group, including the time commitment required for food planning and preparation and the need to understand portion control.
Behavioral weight loss studies have determined that providing prepared meals leads to greater success, not because the food is provided free, but because the subjects know exactly what to eat (11, 12).
To our knowledge, the efficacy of self-selected diet vs. packaged prepared entrees for weight loss and concomitant improvements in CVD risk factors has not been studied in a healthy free-living population. In addition, we aimed to assess the impact of the weight loss protocols on related outcomes such as fat and fat-free mass and body fat distribution. We hypothesized that although all subjects would likely experience some degree of weight loss success, consumption of two frozen entrees per day (one at lunch and one at dinner) would significantly enhance weight loss in overweight women compared with a self-selected FGP plan by providing entrees with controlled portion size and a favorable nutrition profile. Weight losses were expected to lead to improvements in BP, blood lipids, glucose, and insulin levels and an improvement in body composition.
Research Methods and Procedures
Subjects were recruited from the Champaign-Urbana, Illinois area by means of a mailed flyer, using the following inclusion criteria: female, BMI of 26 to 40 kg/m2, 24 to 60 years old, access to microwave oven during the day, and willingness to consume foods from all food groups. Potential subjects were excluded for the following: use of cholesterol-lowering, hypertension, or weight loss drugs, or any other drugs that affect weight or alter body composition; use of herbal supplements; pregnancy or lactation; strict vegetarian diet; smoking; diabetes; severe hypertension defined as systolic BP >159 and diastolic BP >99 mm Hg; or other chronic disease states. The Institutional Review Board of the University of Illinois approved the protocol before subject recruitment.
Sixty women were recruited. A registered dietitian interviewed each subject before the study. Information collected in the interview included weight, height, BP, and health history including a physical activity questionnaire. Instructions were given for completing the baseline food and activity records. Because the study did not include an exercise component, the subjects were instructed to keep a consistent activity level throughout the 8-week intervention. The subjects were divided into the respective diet groups by means of a stratified randomization, based on BMI, age, and activity level. For the randomization, activity level was rated as low, medium, or high (designated by 1, 2, or 3, respectively) based on answers to the activity questionnaire during the interview.
Each group attended a separate orientation in which the details of their diet and study protocol were explained. The self-selected diet group (S) was instructed to consume a diet based on the FGP (7), consisting of two servings from the meat and alternatives group, two servings of nonfat dairy foods, two servings of fruit, three servings of vegetables, six servings from the grain group, and eight cups of water daily. The portion-controlled group (B) was instructed to consume two Uncle Ben's bowls (Masterfoods USA, Vernon, CA) per day (one at lunch and one at dinner), plus two cups of salad vegetables, two cups of nonfat milk or yogurt, two servings of fruit, three servings of whole grains, and eight cups of water daily. All serving sizes were as defined by the FGP. Both groups had the same optional choices from fats and oils, desserts, and alcohol. Diets were prescribed to have the same nutrient composition (55% carbohydrate, 25% protein, 20% fat) and the same approximate calorie level (1365 kcal). Each group was blinded regarding the protocol for the other group until the study was concluded.
Group B subjects were allowed to choose from 24 different Uncle Ben's bowls. Because there are some differences in composition among the bowls, they were divided into two categories (orange and blue), listed and described in Table 1. Subjects were instructed to choose one from each category every day to achieve a more equal daily diet composition. The average daily intake provided by two entrees was 733.4 kcal, 12.5 g of fat, 43.4 g of protein, and 113.4 g of carbohydrates, with the remainder of the daily intake coming from the additional foods mentioned above. The bowls contain pasta, rice, or noodles along with various vegetables, and meats including chicken, beef, turkey, or shrimp. Style of cuisine for the bowls includes American, Mexican, Chinese, Italian, or Thai. Thus, the bowls provided a great deal of variety in flavor and content.
Table 1. . Macronutrient composition of frozen entrees*
Subjects in the frozen entree group (B) were instructed to consume one bowl from the orange list and one from the blue list each day, in addition to two servings of non-fat dairy products, two servings of fruit, two servings of salad vegetables, three servings of whole grains, and eight cups of water daily. CHO, carbohydrate.
Blue bowl list
Chicken fried rice
Garlic chicken bowties
Homestyle chili with beans
Honey dijon chicken
Beef steak stir fry
Spicy peanut chicken
Three cheese ravioli
Tomato sausage rotini
Orange bowl list
Cajun style chicken and sausage
Chicken and vegetables
Spicy beef and broccoli
Sweet and sour chicken
Turkey, wild rice and cranberries
Chicken fettuccine alfredo
Four cheese lasagna with meat sauce
Garden vegetable lasagna
Parmesan shrimp penne
Honey ginger chicken
Spicy Thai style chicken
Santa Fe chicken
Baseline measurements including weight, height, BP, waist circumference, and body composition by DXA were taken. Subjects submitted a 3-day food record and a 2-day activity record representing their usual food intake and physical activity levels. Fasting blood samples were drawn on 2 consecutive days and analyzed for baseline levels of lipids, insulin, and basic metabolic panel measures.
To avoid investigator bias throughout the study, both groups were given clear, straightforward instructions but no individual behavioral and diet counseling. All subjects came in once per week for a breakfast meeting with the dietitians for monitoring. The study groups came in on different days, but both groups were offered the same breakfast foods to choose from, including fruits and juices, whole grain breads and fortified cereals, skim milk and yogurt, and coffee and tea. The purpose of the weekly visits was to assess compliance, take weight measurements, and promote subject camaraderie and involvement, and to provide frozen entrees to Group B.
All 3-day food records represented the intake from 2 week days and 1 weekend day. The baseline record represented typical daily intake before the study. The other 4 food records, submitted every 2 weeks during the intervention, represented intakes while following the study protocol. These four records were averaged to determine typical adherence to the diet regimen throughout the 8-week intervention. Diet composition was determined using Nutritionist Pro, Version 1.2 (First Data Bank, 2002, San Bruno, CA) supplemented with additional information on the product content of the Uncle Ben's bowls provided by Masterfoods USA (Vernon, CA).
Body Composition Measures
Weights were measured using a balance beam scale (physician's scale no. D439; Detecto, Webb City, MO). Subjects wore street clothes without shoes or heavy outerwear. Height was determined using a stadiometer. Whole body composition was determined by DXA (Hologic 4500A; Hologic, Bedford, MA). Waist, hip, arm, and thigh circumferences were assessed in triplicate using a retractable measuring tape (Gulick II; Country Technology Inc., Gays Mills, WI) by the same investigator.
BP was measured by standard sphygmomanometer (model nos. 79 and 682; Prestige Medical, Northridge, CA) on subjects who had been seated for 5 minutes. Two readings were taken, and the average was calculated.
Fasting blood samples were drawn on two consecutive days under identical conditions and results averaged. Laboratory Corporation of America (Elmhurst, IL) performed all blood analyses. Fasting blood samples were collected in serum separator tubes, allowed to clot at room temperature for 15 to 30 minutes, and centrifuged for 10 minutes. Serum was placed into transfer tubes within 60 minutes of venipuncture and refrigerated until collected by the laboratory courier. The following analyses were performed on the fresh serum: basic metabolic panel, insulin, lipid panel, and direct measurement of low-density lipoprotein-cholesterol (LDL-C). On completion of the study, high-sensitivity C-reactive protein (CRP) was analyzed on serum samples that had been kept frozen at −80 °C.
Activity levels were assessed by 2-day activity diary, based on a modification of the method of Bouchard et al. (13). Records from baseline and endpoint were compared for consistency of physical activity during the study.
Statistical Analysis and Power Calculation
The primary analysis was a per-protocol analysis. However, an intention-to-treat analysis was also conducted on the primary outcome, body weight change, using data from all randomized individuals, to determine both efficacy and acceptability of this weight loss regimen. For the intention-to-treat analysis, missing data due to lack of follow-up were determined using the “last value carried forward” approach. Effectiveness of randomization was determined by Student's t test. Significant differences in treatment compliance and dietary adherence were determined by Student's t test. Significant differences in changes in outcomes of interest between the two intervention groups were determined by repeated measures ANOVA. Significance for all tests was determined as α < 0.05.
The primary dependent variable in this study was body weight change. All other evaluated data were secondary outcomes of interest. It was anticipated that with an estimated effect size of 0.70 (difference between groups in weight loss in females of 2.0 ± 2.9 kg), based on work by Metz et al. (8) and McCarron et al. (9), an α (significance) level of 0.05 (one-tailed test), and a power of 80%, a sample size of 25 subjects per group would be required to find statistical differences in weight loss between Groups B and S, should it exist. With an estimated retention rate of 83%, it was planned to recruit 60 individuals into the study.
Sixty women were successfully recruited and randomized to Group S or Group B. As shown in Table 2, there were no differences between groups in BMI, age, waist circumference, and activity level. Two subjects were lost to follow-up: one because of a loss of interest (Group S) and one because of an unrelated illness (Group B). Five subjects were deemed protocol violators: one for lack of attendance (Group B), two for lack of record keeping (one from each group), and two for admitting lack of adherence to the diet (one from each group). As a consequence, of the 60 subjects, 53 women completed the study, with 26 and 27 participants in Groups B and S, respectively, and subsequently were included in the per-protocol analyses. The ethnicity of the finishing cohort was as follows: 41 whites, 9 African Americans, and 3 from other racial/ethnic groups. Subjects in Group B (N = 26) consisted of 19 whites and 7 minorities, whereas Group S (N = 27) contained 22 whites and 5 minorities.
Table 2. . Characteristics of subjects at enrollment (N = 60)
Group S (N = 30) mean ± SD (range)
Group B (N = 30) mean ± SD (range)
31.6 ± 3.2 (26.4–39.2)
31.5 ± 3.4 (26.3–39.1)
36.6 ± 9.4 (24–56)
37.5 ± 9.7 (24–55)
1.9 ± 0.9 (1–3)
2.0 ± 0.9 (1–3)
At each weekly visit, the subjects were asked the same direct questions by the researchers: “Did you consume everything on your diet plan? If not, describe. Did you consume anything that is not on your diet plan? If so, describe. Was the plan easy/hard to follow?” The researchers engaged the participants in conversation and noted any comments on how they felt and their opinions about the diet. The researchers also noted any observations on the appearance, mood, etc. of the subjects. Answers to the direct questions concerning dietary intakes were used to calculate a dietary adherence score, described below. The more subjective comments were noted but were not analyzed statistically.
Dietary adherence was based on answers to the questions noted above regarding inclusion or exclusion of dietary items. A percentage value was calculated based on how the intake differed from the prescribed diet. For example, everyone was required to eat two servings of fruit per day. If a subject skipped one serving of fruit during the week, that represented a 7% difference from the prescribed diet for that one food group. The score for all food groups provided a total exclusion score for that week. Inclusion of an item not on the plan was calculated by giving the extra food a weighted value, based on estimated kilocalories, and calculating the approximate percentage it added to the intake for that week. For each weekly visit, percentage scores for dietary exclusions and inclusions were compiled and averaged. The researchers were careful to remain nonjudgmental regardless of the responses, and weekly contacts allowed for the development of a close rapport with all the subjects.
Total compliance was measured based on attendance (percentage of all time points), completion of written records (percentage of all required records), and dietary adherence (calculated as described above). Subjects who fell below 70% compliance in any area were dropped. Table 3 lists the compliance scores for the 53 subjects who completed the study.
Table 3. . Compliance of subjects completing the study
In addition, the data from pre- and postintervention activity records indicate that the subjects complied with the instruction to maintain the same level of physical activity throughout the study. The percent change in activity score was −0.4% for Group S and +1.6% for Group B, neither of which was significant (p < 0.05).
The accuracy of food records is often questioned because of difficulty estimating portion size and a tendency toward underreporting, especially in overweight persons (14, 15, 16). However, keeping food records is a useful task for research subjects in that it helps them stay aware of what they are eating (17). Food record data are also useful to the researchers because they give an indication of the overall macronutrient composition of the subjects’ diets. As shown in Table 4, the food record data indicate that Group B was more successful at consuming the prescribed fat level of the diet. Reported caloric intakes during the intervention were not different for the two groups. However, Group B reported a significantly higher intake of carbohydrate and a significantly lower intake of fat than did Group S, both in grams consumed and in percentage of calories. Protein intakes were not different. Using “five a day” as a guideline for recommended number of servings of fruits and vegetables combined, both groups approached this goal, with Group S increasing intakes from 3.6 to 4.5 servings and Group B increasing from 3.3 to 4.8 servings per day. Differences between groups were not significant.
Table 4. . Reported dietary intakes: energy level and macronutrient content
Group S (N = 27) Mean ± SD
Group B (N = 26) Mean ± SD
Average intake before study
Average intake during study
Average intake before study
Average intake during study
Significant difference in change from baseline between groups in average intakes; p < 0.05.
As expected, both groups achieved significant weight losses over the 8 weeks of the study (see Table 5). Group B was superior to Group S in weight loss (−5.6 ± 2.2 vs. −3.6 ± 2.5 kg, p < 0.01, N = 53). Similarly, the intention-to-treat analysis involving all randomized individuals (N = 60) determined that Group B was superior to Group S in weight loss success (−5.3 ± 2.2 vs. −3.4 ± 2.6 kg, p < 0.01). This analysis is important because it demonstrates that portion-controlled weight loss regimens using prepackaged entrees are both effective and acceptable on a short-term basis. Based on the per-protocol analyses, Group B was more successful than the self-selecting Group S, losing 6.5% of initial weight compared with 4.2% (p < 0.05). The pattern of weight change over the 8-week intervention was consistent within the respective groups. Figure 1 illustrates weight changes at all time points. Concurrent with the differing success with weight loss, body composition changes (see Table 5) determined by DXA showed that Group B achieved a greater loss of body fat mass than Group S (3.6 ± 1.8 vs. 2.3 ± 1.4 kg, p = 0.05). However, there was no difference between groups in percentage body fat change due to parallel changes in lean body mass. With regard to regional fat, the DXA scans revealed that Group B had significantly greater losses of trunk fat compared with Group S, which was verified by a greater decrease in waist circumference. In addition, Group B had greater reductions in hip circumference (p < 0.05). Decreases in arm and thigh circumference were not significantly different between the groups.
Table 5. . Body weight and composition changes
Group S (N = 27) mean ± SD
Group B (N = 26) mean ± SD
Significant difference in change from baseline between groups; p < 0.01.
Significant difference in change from baseline between groups; p < 0.05.
Blood lipid measures (see Table 6) included serum total cholesterol, triglycerides (TGs), high-density lipoprotein cholesterol, LDL-C (calculated and direct), and very-low-density lipoprotein-cholesterol. Group B experienced a significantly greater decrease in total cholesterol than Group S (12.4% vs. 6.8%). All other changes in lipid values were not significantly different for the two groups.
Table 6. . Serum markers of CVD risk
Group S (N = 27) mean ± SD
Group B (N = 26) mean ± SD
Significant difference in change from baseline between groups; p < 0.05.
Fasting glucose and insulin levels are favorably affected by weight loss. Group B exhibited a decrease in serum insulin levels compared with a slight increase in Group S (−1.8 ± 3.7 vs. +0.3 ± 3.8 μU/mL, respectively, p < 0.05). There was a trend toward lower glucose levels in Group B and a slight increase in Group S, but the difference was not significant (−2.5 ± 5.2 mg/dL for B vs. +0.6 ± 6.4 mg/dL for S, p = 0.055).
There were no differences between the groups in CRP concentrations. When all subjects were pooled, CRP concentrations from baseline to 8 weeks decreased significantly (1.6 ± 4.6 mg/dL, p < 0.05), but the decrease over time for each individual group was not significant because of large individual variations. However, Group B approached significance with a decrease of 1.6 ± 4.0 mg/dL (p = 0.056).
Both groups of subjects were normotensive before initiating the intervention, with baseline systolic values of 109.5 ± 11.7 vs. 108.0 ± 11.1 mm Hg and diastolic values of 70.8 ± 10.8 vs. 70.8 ± 6.2 mm Hg for Groups B and S, respectively. No changes in BP were seen over time or between groups. At 8 weeks, systolic values were 107.0 ± 15.5 vs. 109.4 ± 8.8 mm Hg, and diastolic values were 69.7 ± 12.5 vs. 74.4 ± 7.6 mm Hg for Groups B and S, respectively.
The results of this study indicate that consumption of a diet using portion-controlled entrees enhances weight loss success. Appropriate portion control can be achieved by consuming nutritionally balanced frozen entrees. The U.S. Department of Agriculture FGP does offer reasonable guidelines on how much to consume from each of the various food groups. Choosing the lowest number of servings from each group provides a workable weight loss diet plan. However, a concerted effort is required to choose, prepare, measure, and consume foods in the recommended portion sizes. Although both groups in this study attempted to follow a 1365-kcal diet plan based on serving sizes used in the FGP, those using packaged portion-controlled entrees were more successful in achieving significant reductions in weight and body fat. For many people who want to lose weight, portion control may prove to be a key factor.
Meal replacement as part of a low-calorie diet has been shown to be an effective means of promoting weight loss (18, 19, 20). However, these kinds of diets typically use specially formulated liquid products that provide nutritional balance in a low calorie form. Meal replacement is convenient, provides consistent portion control, and is simple to use. The downside is the potential monotony of consuming the same or very similar products one or more times per day. In addition, meal replacement products lack some of the benefits of a balanced diet of real foods, such as the nontraditional nutrients and phytochemicals that are emerging as potentially important to health. Using a variety of frozen entrees, with the simple addition of a few foods, in the place of formulated meal replacement products would eliminate these two problems while maintaining the convenience and portion control.
A growing segment of our population chooses not to invest large amounts of time in meal preparation. Those who elect to eat out in restaurants on a regular basis are vulnerable to potential weight gain from the excessively large portions commonly served. Being served a large portion usually results in overeating because people tend to consume the amount of food that is presented to them (21). Packaged entrees provide a reasonable alternative to restaurant eating because preparation requires only a few minutes in a microwave oven.
The prescribed macronutrient composition of both diets in this study was 55% carbohydrate, 25% protein, and 20% fat, which would be considered a low-fat diet, moderate in protein and carbohydrate. Subjects were instructed to consume foods from all groups, including nonfat dairy foods and whole plant foods. Several of the subjects expressed surprise that they could eat real foods, feel satisfied, and still lose weight. They were delighted that no special diet foods, medications, or extreme measures were required for success. Overall comments indicated that the bowl diet was particularly easy to follow. The weekly interviews included questions on ease of following the diet. Of the 26 subjects consuming the frozen entrees (Group B), 24 found the diet easy to follow, whereas 2 found it hard. Of the 27 subjects in Group S, 14 found the diet easy to follow, 3 found it hard, and 10 found it moderate in difficulty.
We know that a permanent lifestyle change is required to maintain a weight loss. In an 8-week study, we can hope only to break old habits and provide training for a future of healthier eating habits. Certainly, the easier it is to follow an eating plan, the more likely it is to result in long-term success.
Although reported caloric intakes during the intervention were the same for both groups, Group B reported a higher intake of carbohydrate and a lower intake of fat. Protein intakes were not different. This is of interest with regard to the ongoing debate on the efficacy of high protein, moderate-fat diets vs. high-carbohydrate, low-fat diets for weight loss. As long as a diet provides the necessary nutrients and does not go to extremes in the manipulation of macronutrients, a variety of dietary patterns can provide healthy weight loss (22, 23, 24). According to the literature, diets too high in carbohydrates may lead to increased TG levels (23), whereas excessive intakes of protein may lead to losses of bone minerals or metabolic stress on the kidneys (24) and potentially may “crowd out” fruits and vegetables from the diet (25). An increase in fruit and vegetable consumption, as was achieved by both study groups, is beneficial for weight loss because these foods are high in fiber and nutrient density while low in calories. A low-calorie diet of moderate macronutrient composition, such as the one used in this study, may well be the best option for weight control. It is also important to take into account that individual food preferences will determine the dietary pattern that is easiest for a person to follow for the long term.
This study supports the frequent finding that small changes in weight can improve biomarkers of CVD risk. Total cholesterol levels were decreased in both groups, but significantly more so in Group B, which could be explained by the combined effects of decreased body fat and lower intake of dietary fat. Reductions were seen over time in all blood lipid measures, but there were no differences between the two diet groups. Group B experienced a significant decrease in serum insulin vs. a slight increase for Group S, despite the fact that Group B had a higher carbohydrate intake. This may be due to Group B's greater loss of weight and fat mass.
CRP has been identified as a potential biomarker for the inflammatory cascade involved in the atherosclerotic process (26). CRP concentrations have been shown to correlate with obesity and to fall with weight loss (27). In the current study, CRP values decreased over time for all subjects as a pooled group, but there were large individual variations and no significant differences between groups. Recent research has suggested that elevations in CRP are related to insulin resistance, independent of obesity (28). In the current study, fasting insulin and fasting glucose levels were in the normal range for both groups at all time points, so these subjects had no evidence of insulin resistance. Nevertheless, Group B experienced a significant decrease in fasting insulin level over the course of the study compared with Group S. However, we found no correlations between CRP and fasting insulin or glucose.
It is apparent that waist circumference is directly related to risk of CVD (29). In addition, BMI and central adiposity have been correlated with CVD risk (2). In all of these parameters, Group B had significantly better results than Group S.
In conclusion, consumption of portion-controlled entrees as part of a balanced low-calorie diet enhances losses of weight and fat mass in healthy overweight women and concomitant reductions in markers of CVD risk. Accurate portion control is an important factor in weight loss success and is easier to achieve with packaged entrees.
This research was supported by a grant from Masterfoods USA.