Very-low-calorie diet for obese type 2 diabetic patients
ACP J Club. 1991 Nov-Dec;115:76. doi:10.7326/ACPJC-1991-115-3-076
Wing RR, Marcus MD, Salata R, et al. Effects of a very-low-calorie diet on long-term glycemic control in obese type 2 diabetic subjects. Arch Intern Med. 1991 Jul;151:1334-40.
To determine whether a very-low-calorie diet promotes short- and long-term glycemic control when added to a behavior modification program for obese, type 2 diabetic patients.
Randomized, controlled trial of 20 weeks duration with follow-up 1 year later.
10 men and 26 women, age 35 to 70 years, who were 30% or more above ideal body weight and who had type 2 diabetes but no evidence of liver, renal, or heart disease were enrolled. 33 patients (92%) completed all assessments.
Patients were randomly assigned to either behavior modification (BT, n = 19), which served as the control, or to behavior modification plus very-low-calorie diet (VLCD, n = 17). BT was conducted at weekly meetings with groups of 8 to 10 patients and included instruction on diet, exercise, and behavior modification. Both groups started and ended the 20-week intervention period with a 4200 to 6300 J/d (1000 to 1500 cal/d) diet, but the VLCD group switched during weeks 4 to 12 to 1680 J/d (400 cal/d) of lean meat, fish, fowl, and occasionally a commercial dietary supplement. Both groups were assessed at baseline, at the end of the 20-week intervention period, and 1 year after its completion.
Main outcome measures
Fasting blood glucose, glycosylated hemoglobin, lipids, and weight.
Fasting blood glucose fell in the VLCD group from 14.2 mmol/L at baseline to 7.7 mmol/L at 20 weeks and to 10.4 mmol/L at 72 weeks compared with values of 12.8, 9.3, and 13.5 mmol/L for the BT group. Decreases in fasting blood glucose and glycosylated hemoglobin were greater for the VLCD group at 72 weeks (P = 0.002 and 0.001, respectively). Average weight loss over the 20-week intervention period was 18.6 kg in the VLCD group compared with 10.1 kg in the BT group; the average weight loss (8.6 kg vs. 6.8 kg, respectively) and lipid values did not differ from baseline to follow-up 1 year later.
A very-low-calorie diet for 8 weeks added to a behavior modification program did not lead to additional long-term weight loss but produced an improvement in glycemic control that was sustained for 1 year after treatment.
Sources of funding: The American Diabetes Association and the National Institutes of Health.
Address for article reprint: Dr. R. Wing, The Western Psychiatric Institute and Clinic, 3811 O'Hara Street, Pittsburgh, PA 15213.
The high rate of recidivism following weight-losing programs was predictably confirmed in this study of considerably overweight type 2 diabetic patients. This study also showed the immediate improvement in glycemic control that follows calorie-restricted diets. A third clinically important effect was the sustained improvement in glycemic control in the VLCD group a year after the intervention. This relative improvement remained despite a 10-kg average weight gain during the same period.
The authors offer a number of explanations for this phenomenon, including improved insulin action or secretion, or both, which were supported by other studies and by their observation of improved insulin secretion in the VLCD group. One possible explanation discounted by the authors is that the caloric intake of the VLCD group may have been less than that of the BT group (data not collected in the study). An enhanced energy-efficiency period (lasting for months to years) has been shown in the post-obese state that could contribute to this phenomenon (1-3).
This study is generally methodologically sound. The reader is not, however, told the method of patient selection, which means that the outcome observed may not apply to diabetic patients other than those who fit the (largely unknown) study-subject profile.
VLCDs can be safely recommended for selected overweight type 2 diabetic patients as a means to achieve both weight loss and improved glycemic control. Medical supervision is recommended because of the likelihood of need for changes in pharmacotherapy.
Eugene C. Corbett, Jr., MD
University of Virginia Charlottesville, Virginia