Samaha 2003.

Study characteristics
Methods	Review comparison(s) addressed by this study: 1, 3 Study design: RCT, parallel, single‐centre (1) Trial registry number: NR Total number of trial arms: 2 Year trial started: 2001 Sample size calculation: Yes Outcome(s) used for sample size calculation: Weight loss Duration of run‐in period (weeks): NA What was the duration of the weight loss phase: 12 months What was the duration of the weight maintenance phase: 24 months Other notes about methods: NA
Participants	Country and setting: USA, veteran affairs medical centre in Philadelphia Eligibility criteria: Participants aged 18 years or older with a BMI of at least 35 kg/m2. Participants were excluded if they had a serum creatinine level of more than 1.5 mg/dL (132.6 µmol/L); hepatic disease; severe, life‐limiting medical illness; an inability of diabetic subjects to monitor their own glucose levels; or if they had active participation in a dietary programme, or use of weight‐loss medications. Type 2 diabetes at baseline: Mixed; at baseline 64% in treatment and 56% in control group had diabetes ‐ stratified with T2DM for change in weight at 3 to < 12 months; change in BMI at at 3 to < 12 months was reported separately for participants without and with diabetes ‐ these data were included in the appropriate comparisons where possible.  Impaired glucose tolerance at baseline: Unclear Cardiovascular conditions/risk factors/events at baseline: Unclear Gender: Mixed Total number randomised: 132 Total attrition in trial: 53 Treatment diet Participants randomised: 64 Participants withdrawn (voluntary): NR Total attrition: 21 Control diet: Participants randomised: 68 Participants withdrawn (voluntary): NR Total attrition: 32 Baseline data treatment diet: Randomised participants not included: None Age (years): mean (SD) 53 (9) Gender distribution (as reported): female 20.0%, male 80.0% Weight (kg): mean (SD) 130.0 (22.7) BMI (kg/m2): mean (SD) 42.9 (6.6) DBP (mmHg): mean (SD) 78 (11) SBP (mmHg): mean (SD) 133 (15) HbA1c (%): mean (SD) 7.8 (1.2) LDL (mmol/L): mean (SD) 2.95 (0.93) HDL (mmol/L): mean (SD) 1.06 (0.28) Non‐HDL (mmol/L): NR TC (mmol/L): mean (SD) 4.69 (1.35) TG (mmol/L): mean (SD) 2.12 (1.99) Baseline data control diet: Randomised participants not included: None Age (years): mean (SD) 54 (9) Gender distribution (as reported): female 15%, male 85% Weight (kg): mean (SD) 131.8 (27.3) BMI (kg/m2): mean (SD) 42.9 (7.7) DBP (mmHg): mean (SD) 80 (9) SBP (mmHg): mean (SD) 135 (16) HbA1c (%): mean (SD) 7.4 (1.5) LDL (mmol/L): mean (SD) 3.06 (0.75) HDL (mmol/L): mean (SD) 1.06 (0.26) Non‐HDL (mmol/L): NR TC (mmol/L): mean (SD) 4.97 (0.78) TG (mmol/L): mean (SD) 1.99 (1.36) Group differences at baseline: No Characteristic(s) with significant group difference and relevant statistic: NA Other notes about participants: HbA1c and blood lipid baseline equivalence NR; could not assess statistical significance
Interventions	Energy (E) comparison of treatment vs control diets: Different ‐ ad libitum in treatment diet and restricted in control diet Treatment diet: Name (as reported) and brief description: Low‐carbohydrate ad libitum diet, subjects were instructed to restrict carbohydrates to 30 g per day or less. No instructions on reducing total fat intake were provided. Vegetables and fruits with high ratios of fibre to carbohydrates were recommended. Treatment diet type (carbohydrate‐fat‐protein): Very low‐unclear‐unclear Exercise component? No Recipients: Subjects were 20% female and aged mean (SD) 53 (9) years, with BMI mean (SD) 42.9 (6.6) and weight 130.0 (22.7) kg. Forty‐one percent had diabetes mellitus, 45% had metabolic syndrome without diabetes, 72% had hypertension, 51% had hyperlipidemia and 16% had coronary artery disease. Why? Severely obese subjects with a high prevalence of diabetes or the metabolic syndrome would have a greater weight loss, without detrimental effects on risk factors for atherosclerosis, while on a low‐carbohydrate diets. There are concerns that carbohydrate restriction might have adverse effects on serum lipids. The intervention was administered to determine weight loss and changes in cardiovascular risk factors resulting from a low‐carbohydrate diet. What (materials)? Diet overview handout, instructional nutrition labels, sample menus and recipes, and a book on counting calories and carbohydrates. No specific exercise programme was recommended What (procedures)? Participants were prescribed carbohydrate restriction of 30 g or less per day. They received two hours of group‐teaching sessions each week for four weeks, thereafter one hour per month for five additional months. All sessions were led by a an expert in nutritional counselling. Who provided? Experts in nutritional counselling How and where? Face‐to‐face group sessions, location NR When and how much? Four weekly sessions of two hours each, followed by five monthly sessions of one hour; over the course of six months Strategies to improve or maintain fidelity; tailoring and modification: Sessions were held in groups, weighing and waist measurement at six months and determination of dietary compliance using a validated instrument for 24‐hour recall of consumption Extent of intervention fidelity: Subjects decreased their caloric intake from 2153 (1060) to 1343 (731) kcal (P < 0.001). These subjects also significantly increased their intake of fat to 44 (16)% and protein to 25 (10)%, and decreased their carbohydrate intake to 31 (19)%. Concomitant interventions: Of intervention participants; 11% were treated with sulfonylurea, 17% with metformin, 2% with peroxisome proliferator‐activated receptor gamma antagonist and 9% with insulin. Sixty‐four percent received antihypertensive therapy, 42% received statins and 3% received gemfibrozil. Control diet: Name (as reported) and brief description: Low‐fat diet, subjects were instructed to follow diet according to the obesity‐management guidelines of the National Heart, Lung and Blood Institute, including caloric restriction sufficient to create a deficit of 500 calories per day, with 30% or less of total calories from fat. From the guideline: 55% or more of total calories; protein approximately 15% of total calories and 30% or less of total fat. Control diet type (carbohydrate‐fat‐protein): Balanced‐balanced‐balanced Exercise component? No Recipients: Subjects were 15% female and aged mean (SD) 54 (9) years, with BMI mean (SD) 42.9 (7.7) and weight 131.8 (27.3) kg. Thirty‐eight percent had diabetes mellitus, 41% had the metabolic syndrome without diabetes, 57% had hypertension, 50% had hyperlipidemia and 16% had coronary artery disease. Why? NR What (materials)? Diet overview handouts, instructional nutritional labels, sample menus and recipes as well as a book on how to count calories and carbohydrates (National Heart, Lung, and Blood Institute 1998).  What (procedures)? Participants were prescribed a low‐fat diet. They received two hours of group‐teaching sessions each week for four weeks, thereafter one hour per month for five additional months. All sessions were led by a an expert in nutritional counselling. Who provided? Experts in nutritional counselling How and where? Face‐to‐face group sessions, location NR When and how much? Four weekly sessions of two hours each, followed by five monthly sessions of one hour; over the course of six months Strategies to improve or maintain fidelity; tailoring and modification: Sessions were held in groups, weighing and waist measurement at six months and determination of dietary compliance using a validated instrument for 24‐hour recall of consumption Extent of intervention fidelity: [From 25 Seshadri 2005] Subjects decreased their caloric intake from 1882 (820) to 1590 (679) kcal (P = 0.057). There was no significant change in macronutrient intake, with intake at 16 (5)% for protein, 51 (15)% for carbohydrate and 32 (15)% for fat at six months. Concomitant interventions: Of control participants; 16% were treated with sulfonylurea, 13% with metformin, 2% with peroxisome proliferator‐activated receptor gamma antagonist and 4% with insulin. Fifty‐seven percent received antihypertensive therapy, 37% received statins and 2% received niacin.
Outcomes	Change in body weight (kg) at 3 to < 12 months: Yes, in participants with type 2 diabetes mellitus Change in body weight (kg) at ≥ 12 months: No Number of participants with 5% weight loss from baseline at 3 to < 12 months: No Number of participants with 5% weight loss from baseline at ≥ 12 months: No Change in BMI (kg/m2) at 3 to 12 months: No Change in BMI (kg/m2) at ≥ 12 months: No Number of participants with 5% BMI reduction from baseline at 3 to 12 months: No Number of participants with 5% BMI reduction from baseline at ≥ 12 months: No Change in DBP (mmHg) at ≥ 12 months: No Change in SBP (mmHg) at ≥ 12 months: No All‐cause mortality at ≥ 12 months: No Cardiovascular mortality at ≥ 12 months: No Non‐fatal myocardial infarction at ≥ 12 months: No Non‐fatal stroke at ≥ 12 months: No Diagnosis of type 2 diabetes mellitus at ≥ 12 months: No Change in HbA1c (%) at ≥ 12 months: No Change in LDL (mmol/L) at ≥ 12 months: No Change in HDL (mmol/L) at ≥ 12 months: No Change in non‐HDL (mmom/L) at ≥ 12 months: No Change in total cholesterol (TC) (mmol/L) at ≥ 12 months: No Change in triglycerides (or triacylglycerides) (TG) (mmol/L) at ≥ 12 months: No Participant‐reported adverse effects: No
Notes	Number and type of records(s): journal article Trial acronym/name: None Trial funded by: Veterans Affairs Healthcare Network Competitive Pilot Project Grant Declaration of interest: NR for primary reference. Stern et al 2004: "None disclosed."