NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Database of Abstracts of Reviews of Effects (DARE): Quality-assessed Reviews [Internet]. York (UK): Centre for Reviews and Dissemination (UK); 1995-.

Cover of Database of Abstracts of Reviews of Effects (DARE): Quality-assessed Reviews

Database of Abstracts of Reviews of Effects (DARE): Quality-assessed Reviews [Internet].

Show details

Weight management using a meal replacement strategy: meta and pooling analysis from six studies

, , , , and .

Review published: .

CRD summary

The review assessed partial meal replacement plans for long-term weight management. The authors concluded that these types of plans can safely and effectively produce significant sustainable weight loss and improve weight-related risk factors for disease. Significant differences between the included studies suggest that these conclusions should be interpreted with caution.

Authors' objectives

To review existing literature on the safety and effectiveness of a partial meal replacement (PMR) plan using one or two vitamin or mineral fortified meal replacements, as well as regular foods, for long-term weight management.

Searching

MEDLINE, EMBASE and the Cochrane Controlled Trials Register were searched from 1960 to January 2001; the search terms were not reported. In addition, bibliographies were handsearched. The authors did not state whether any language restrictions were applied.

Study selection

Study designs of evaluations included in the review

Randomised controlled trials (RCTs) of at least 3 months duration were eligible for inclusion.

Specific interventions included in the review

For their review, the authors defined PMR as a plan or programme that included one or more meals replaced by a commercially available, calorie-reduced product(s), fortified with vitamins and minerals, and at least one daily meal consisting of regular foods. The calorie content of the plan had to be between 800 and 1,600 kcal/day. Studies had to compare PMR with a traditional reduced calorie diet (RCD) plan.

Two of the included studies used the Slim-Fast plan with two meal replacements per day. One study used a 1,200 to 1,500 kcal/day diet with two meal replacements per day and two snack replacements, dropping to one meal replacement per day after 3 months. One study used a 1,200 kcal/day diet with one or more meal replacements per day. One study used the 10-day Jump Start plan; participants who had up to 10% weight loss had two meal replacements per day, after 10% weight loss this dropped to one meal replacement per day. One study used a traditional lifestyle group with two meal replacements per day.

The treatments in the control groups consisted of an isoenergetic control diet, an isoenergetic traditional low-fat diet, an isoenergetic diabetic diet, the American Diabetic Association diet, a 1,500 kcal/day control diet and a traditional lifestyle-group.

Within each study, the participants received the same behavioural modification programme, dietary instruction regarding calorie intake, and exercise prescription.

Participants included in the review

Studies with participants over 18 years of age, with a body mass index of at least 25 kg/m2, were eligible for inclusion in the review. The majority of studies included participants over the age of 30, and the majority of participants were female. The baseline body mass index ranged from 25 to 40 kg/m2 and one study specified that participants had to have gained weight in the previous year. In 3 studies the participants were all healthy, in one study co-morbidities were allowed, and in 2 studies the participants had non insulin-dependent diabetes. The only risk factor that differed significantly between the patients in the intervention and control groups was the baseline glucose.

Outcomes assessed in the review

Studies that did not report self-reported data were eligible for inclusion. The primary outcomes reported in the review were weight loss at 3 months and one year.

How were decisions on the relevance of primary studies made?

The authors did not state how the papers were selected for the review, or how many reviewers performed the selection.

Assessment of study quality

The Jadad tool was used to assess the quality of the included studies. This assesses the randomisation procedure, blinding of the patient and outcome assessor, and description of drop-outs and missing values. The authors did not state how the papers were assessed for quality, or how many reviewers performed the quality assessment.

Data extraction

The authors did not state how the data were extracted for the review, or how many reviewers performed the data extraction. Data were extracted on the participants, intervention and control diets, study duration, the effects on biomarkers, and weight loss at 3 months and one year. It was also stated that the authors of the identified publications were contacted and asked to supply primary data for analysis.

To estimate the effect of the interventions on weight loss, the difference in weight at 3 months from baseline and at one year from baseline was calculated. The relative effectiveness of the PMR intervention was measured by calculating the difference in weight loss between the PMR group and the control group. To estimate the effects of the interventions on the biomarkers, the changes observed at 3 months and one year were measured. The biomarkers evaluated were fasting plasma glucose, insulin, cholesterol, high-density lipoprotein cholesterol, low-density lipoprotein cholesterol, triacylglycerols, and systolic and diastolic blood-pressure. All analyses were conducted on an intention-to-treat basis.

Methods of synthesis

How were the studies combined?

The published results of the 6 studies were combined to estimate weight loss in the control group, weight loss in the PMR group and between-group weight loss. Both fixed-effect and random-effects models were used.

Individual patient data (IPD) from the 6 studies were pooled into a single data set. Estimates of weight loss in the control group, weight loss in the PMR group and between-group weight loss were then calculated using a random-effects model. Data using the last-observation-carried-forward (LOCF) and multiple imputation (MI) methods for missing imputations were also pooled.

A general linear model was applied to the pooled data for each risk factor to evaluate whether risk factor improvements were associated with their baseline values, weight loss and treatment assignment.

Publication bias was assessed by calculating the fail-safe N and Kendall's tau statistic.

How were differences between studies investigated?

Heterogeneity was assessed for statistical significance using Student's t-test and the chi-squared test.

When combining the published results of the 6 studies, the authors used a linear regression model to adjust for the covariates of gender, baseline age and baseline body mass index. Heterogeneity of the estimates was assessed for statistical significance using the chi-squared test.

When pooling IPD from the 6 studies, the authors adjusted for gender, baseline age, baseline body mass index, study effects, and interaction between study and group effects. An analysis stratified by diabetic status was also conducted. Heterogeneity of the estimates was assessed by testing the significance of interaction between study and group effects.

Results of the review

Six RCTs with a total of 487 participants were included in the review; 249 patients were randomised to the PMR group and 238 to the RCD group. The duration of the included studies ranged from 3 to 51 months, with the majority of studies lasting 12 months. Due to the nature of the interventions, blinding of the patients was not possible in any of the studies.

All of the studies achieved a moderate quality assessment rating.

Weight loss at 3 months.

Each individual study showed a statistically significant weight loss at 3 months in both the PMR and RCD groups (P<0.001). Whilst each study reported greater weight loss in the PMR group than the RCD group, this was only statistically significant in 3 studies. However, when the studies were combined using summary data and using IPD, the difference in weight loss between the PMR and RCD groups was statistically significant.

Based on the fixed-effect meta-analysis, the random-effects meta-analysis and the pooling of IPD from participants who completed 3 months of the studies, the PMR group lost between 6.19 and 6.5 kg and the RCD group lost between 3.34 and 3.96 kg. The weight loss in the PMR group was 2.54 to 3.01 kg more than that in the RCD group.

The authors stated that when using Hedges and Olkin's Q statistic, there was significant heterogeneity between studies for the estimates of weight loss and the effectiveness of the PMR programme. Heterogeneity was also statistically significant when testing the significance of interactions between study and group effects.

Data pooled using the LOCF and MI methods for missing imputations produced the same results as the completer analysis.

At 3 months, 34% of the participants in the RCD group and 72% of the participants in the PMR group had lost 5% or more of their initial body weight. The difference was statistically significant.

The drop-out rate between the PMR and RCD groups was not statistically significantly different at 3 months: 16% and 19%, respectively. None of the participants who dropped out reported any programme- or product-related adverse events.

Weight loss at one year.

Four studies for the RCD and five for the PMR groups reported one year follow-up data. Three of the 4 studies reported statistically significant weight loss in the RCD group, while four of the 5 studies reported statistically significant weight loss in the PMR group. Three of 4 studies reported greater weight loss in the PMR group than the RCD group; this was statistically significant in 2 studies.

When summary data were combined using a fixed-effect model, and when IPD were pooled, the difference in weight loss between the PMR group and the RCD group was statistically significant. However, when combining summary data using the random-effects model, the difference was not statistically significant (P=0.142).

Based on the fixed-effect meta-analysis, the random-effects meta-analysis and the pooling of IPD from participants who completed one year of the studies, the PMR group lost between 6.97 and 7.31 kg and the RCD group lost between 2.61 and 4.35 kg. The weight loss in the PMR group was 2.43 to 3.39 kg more than that in the RCD group.

The authors stated that when using Hedges and Olkin's Q statistic, there was significant heterogeneity between studies for the estimates of weight loss and the effectiveness of the PMR programme. Heterogeneity was also statistically significant when testing the significance of interactions between study and group effects.

Data pooled using the LOCF method for missing imputations produced the same results as the completer analysis. However, when using the MI method, the statistical significance of the difference in weight loss between the PMR and RCD groups was 0.119.

At one year, 33% of the participants in the RCD group and 74% of the participants in the PMR group had lost 5% or more of their initial body weight.

The drop-out rate was statistically significantly higher in the RCD group than in the PMR group at one year: 64% and 47%, respectively. None of the participants who dropped out reported any programme- or product-related adverse events.

Biomarker analyses.

At both 3 months and one year, all eight of the risk factors improved and were positively associated with their baseline values (P<0.001). Improvements in glucose, triglyceride, low-density lipoprotein cholesterol and systolic blood-pressure were statistically, significantly positively associated with weight loss after 3 months. There was a statistically significant additional effect of PMR on improvements in plasma insulin levels after 3 months and one year. Total cholesterol was statistically, significantly, positively associated with weight loss at one year, in addition to glucose, triglyceride, low-density lipoprotein cholesterol and systolic blood-pressure.

Results of analysis stratified by diabetes status.

The weight loss of diabetic participants was lower at one year than at 3 months for both the RCD group (1.74 kg at one year versus 4.88 kg at 3 months) and the PMR group (4.5 kg at one year versus 7.34 kg at 3 months). However, nondiabetic patients had lost more weight at one year than at 3 months in both groups.

At one year, 77% of diabetic participants had dropped out of the RCD group and 79% out of the PMR group, compared with 60% and 35% of nondiabetic participants, respectively.

Publication bias.

Kendall's tau correlation did not support publication bias, except for weight loss at one year in the RCD group.

The authors also presented narrative results of 2 studies, identified during the course of their study, which evaluated a PMR programme. Both studies found that participants in the PMR groups lost more weight than participants in the control groups at 12 weeks; this result was statistically significant in one study.

Authors' conclusions

PMR weight management plans can safely and effectively produce significant sustainable weight loss and improve weight-related risk factors for disease.

CRD commentary

The review question was clear in terms of the study designs, participants, interventions and outcomes of interest. The authors searched three relevant electronic databases. However, no attempts were made to identify unpublished studies and the authors did not state whether any language restrictions were applied; therefore, the possibility of publication and language biases cannot be excluded. Publication bias was assessed for statistical significance, which the authors stated was not supported, except in the assessment of weight loss at one year. A flow chart representing the study selection process was presented, but this suggested that no RCTs were included in the meta-analysis, which is contradictory to their results. The authors did not report details of the review process pertaining to the study selection, quality assessment and data extraction processes; the potential for reviewer error or bias cannot, therefore, be assessed. The studies were assessed for quality using an appropriate, validated, quality assessment tool. Adequate details of the included studies were reported. Heterogeneity was assessed and was found to be statistically significant. However, the studies were still pooled using both summary data and IPD. The authors' conclusions follow from the results presented but, owing to the possibility of publication bias and the significant heterogeneity between the included studies, these conclusions should be interpreted with caution.

Two authors were from Unilever Research and Development and one was from the SlimFast Foods Corporation. The lead author was a member of the Slim-Fast Nutrition Institute, a non profit organisation that reviews and supports nutrition-related investigator-initiated research. SlimFast is a brand of PMR.

Implications of the review for practice and research

The authors did not state any implications for practice or further research.

Bibliographic details

Heymsfield S B, van Mierlo C A, van der Knaap H C, Heo M, Frier H I. Weight management using a meal replacement strategy: meta and pooling analysis from six studies. International Journal of Obesity 2003; 27(5): 537-549. [PubMed: 12704397]

Indexing Status

Subject indexing assigned by NLM

MeSH

Adolescent; Adult; Aged; Biological Markers; Body Mass Index; Diet, Reducing; Female; Follow-Up Studies; Humans; Male; Middle Aged; Obesity /diet therapy; Randomized Controlled Trials as Topic; Weight Loss

AccessionNumber

12003000919

Database entry date

31/01/2005

Record Status

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.

Copyright © 2014 University of York.
Bookshelf ID: NBK69881

Views

  • PubReader
  • Print View
  • Cite this Page

Similar articles in PubMed

See reviews...See all...

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...