Wing 1991.

Methods	Trial design: RCT Randomisation procedure: "...subjects were randomly assigned to either..." Allocation concealment: Unclear Blinding: Recipients: No Providers of Care: No Outcome Assessors: Unclear Quality Assessment Category: C (two points awarded) ‐ high risk of bias.
Participants	Country: USA Setting: Unclear ‐ possibly university based. Number: 36 participants were randomised, 19 to CG and 17 to IG. (Note that the data that is presented is for 33 participants, 16 in CG and 17 in IG.) Age: CG: 51.9 ± 9.9 years IG: 50.6 ± 7.7 years Sex: CG: 25% male, 75% female. IG: 24% male, 76% female. Diagnostic criteria: Participants had type 2 diabetes as defined by the NDDG criteria. Inclusion criteria: 1. Participants must be between 35 and 70 years old. 2. Must be 30% or more above ideal body weight. 3. Having type 2 diabetes as defined by the NDDG criteria. 4. Having no evidence of liver, renal or heart disease that would contraindicate the use of VLCD's.
Interventions	Trial intervention(s): Participants were given a calorie goal of 4200 to 6300 joules per day (depending on their initial body weight) . Information was presented regarding the different caloric content of protein, fat and carbohydrate. Subjects were encouraged to increase their complex carbohydrate intake especially fibre, while decreasing dietary fat. Participants allocated to the IG followed this diet for one month, and then switched to a very‐low‐calorie diet (VCLD) for the second and third months of this program. While following the VCLD, participants were instructed to consume 1680 joules per day of lean fish, meat or fowl and had the option of using a liquid meal for occasional meals. After the two months, other foods were gradually re‐introduced so that by week 17 all subjects were back onto a 4200 to 6300 joules per day diet. Comparison intervention(s): Participants were given a calorie goal of 4200 to 6300 joules per day (depending on their initial body weight) . Information was presented regarding the different caloric content of protein, fat and carbohydrate. Subjects were encouraged to increase their complex carbohydrate intake especially fibre, while decreasing dietary fat. The control and intervention lasted for twenty weeks in total.
Outcomes	Mean body weight for the participants in the low calorie diet group at baseline was 107.7 ± 18.7kg, which rose to 118.2 ± 11.6kg at twelve months. Participants in the very low calorie diet group had a mean baseline weight of 105.8 ± 19.4kg, which decreased to 91.6 ± 10.3kg at twelve months. Glycated haemoglobin for the low calorie diet group was 10.5 ± 2.0% at baseline, decreasing to 8.8 ± 1.8% at six months, which rose to 9.2 ± 2.0% at twelve months and then rose again to 10.7 ± 2.4% at twenty‐four months. Glycated haemoglobin in the very low calorie diet group, was measured at 10.4 ± 2.0% at baseline, decreased to 8.4 ± 2.2% at six months, and then slightly rose to 8.9 ± 2.5% at twelve months and a greater rise to 10.4 ± 2.2% at twenty‐four months for the very low calorie diet group. 1. Weight 2. BMI 3. Fasting plasma glucose 4. Fasting insulin 5. Lipids were all measured at one year.
Notes	Source of funding of trial: Supported by grants from Western Pennsylvania Affiliate of the American Diabetes Association and the National Institutes of Health (NIDDK 29757), Drop‐outs: Three participants dropped out of the CG.
Risk of bias
Bias	Authors' judgement	Support for judgement
Allocation concealment?	Unclear risk	B ‐ Unclear