Current issues of ACP Journal Club are published in Annals of Internal Medicine


Meta-analysis: Dietary strategies alone reduce weight best in NIDDM

ACP J Club. 1997 Jan-Feb;126:5. doi:10.7326/ACPJC-1997-126-1-005

Source Citation

Brown SA, Upchurch S, Anding R, Winter M, Ramírez G. Promoting weight loss in type II diabetes. Diabetes Care. 1996 Jun;19:613-24.



To evaluate the effectiveness of various strategies (behavioral therapies, exercise, diet, anorectic drugs, surgery, or a combination of strategies) used to promote weight loss in patients with non-insulin-dependent diabetes mellitus (NIDDM).

Data sources

Studies were identified by searching MEDLINE (1966 to 1994), the Combined Health Information Data Base (1978 to 1994), Psychological Abstracts (1967 to 1994), ERIC (1966 to 1994), and Dissertation Abstracts (1961 to 1994) using the keywords weight loss, diabetes, patient education, counseling, behavior modification, diet, exercise, anorectic drugs, and surgery. To identify unpublished studies, master's degree programs in nursing, public health, and dietetics were surveyed, and the Centers for Disease Control and Prevention was contacted.

Study selection

Studies were selected if they included patients who were obese and had NIDDM; used a behavioral, dietary, exercise, anorectic-drug, surgical, or combined strategy to promote weight loss; measured weight loss; and provided data sufficient to calculate effect sizes.

Data extraction

Data were extracted for 80 variables that characterized the patients and the study and for 23 outcome variables that included weight, metabolic control, lipids, and other physiologic parameters. Data were pooled, and effect sizes were calculated. Weighted mean postintervention changes for each strategy that compared the intervention group with the control group were calculated.

Main results

89 studies involving 1800 patients were included. Except for surgery, which had the greatest effect on weight loss (> 22.7 kg), dietary strategies led to the largest changes in body weight: a 9-kg reduction in the treatment group compared with the control group (P < 0.05). All interventions except anorectic drugs and behavioral therapies plus exercise led to reductions in mean body weight. Diet alone had the greatest effect on glycosylated hemoglobin levels (2.7%). The combination of behavioral therapies plus diet plus exercise led to reductions of 3.8 kg in mean body weight and 1.6% in glycosylated hemoglobin levels (P < 0.05 for both). Behavioral therapies alone and exercise alone led to the smallest reductions in mean body weight (2.9 kg and 1.5 kg, respectively). Exercise alone or together with behavioral therapy had the smallest effect on glycosylated hemoglobin levels (0.8% and 0%, respectively). Effects of weight reduction strategies were smaller for patients aged > 55 years.


Surgery and dietary strategies alone are most effective for promoting weight loss and metabolic control in patients with non-insulin-dependent diabetes mellitus.

Source of funding: National Institutes of Health.

For article reprint: Dr. S.A. Brown, University of Texas at Austin, 1700 Red River, Austin, TX 78701, USA. E-mail


The clinical heterogeneity of NIDDM leads to different therapeutic goals in different patients. Avoiding the symptoms that are related to diabetes may be sufficient for most patients. In older patients, unwanted effects on quality of life could result from strict dietary recommendations, rigorous exercise programs, and potentially dangerous drugs.

This meta-analysis by Brown and colleagues shows that after 30 years of intensive investigations and more than 900 published reports, only a small amount of valid scientific information is available on weight loss in NIDDM. For example, few studies have been extended beyond 6 months, and 72% have used a nonexperimental design. Despite these limitations, dietary strategies are clearly the most effective. Diet alone was associated with substantial weight loss and pronounced effects on metabolic control, but the long-term benefit of this approach is questionable. The combination of diet plus behavioral therapy and exercise had positive effects on glycosylated hemoglobin levels; but, again, the long-term sequelae are unknown. Anorectic drugs did not substantially influence mean body weight, and possible risks are associated with their prolonged use. Newer data (1) do not seem to show continuous benefits.

Does the fact that losing weight is so difficult warrant therapeutic nihilism? Promotion of weight loss needs to be integrated into a complex treatment program that consists of dietary counseling, control of glucosuria, information about the mode of action of insulin and oral antidiabetic drugs, and foot care (2). Self-measured glucosuria or blood sugar offers the opportunity to improve patients' motivation and compliance by providing feedback, with the added incentive that their use of hypoglycemic agents may be discontinued. A crucial question, however, remains: How can patients be motivated to modify their dietary and other behavior on a long-term basis?

Bernd Richter, MD
Heinrich-Heine-Universität Düsseldorf Düsseldorf, Germany


1. Pascale RW, Wing RR, Butler BA, Mullen M, Bononi P. Diabetes Care. 1995;18:1241-8.

2. Kronsbein P, Jörgens V, Mühlhauser I, et al. Lancet. 1988;II:1407-11.