Table 3.
Randomized controlled trials (RCTs) comparing water with other drinks or no drink.
| Author/y | Study Type | Population | Intervention | Comparison | Outcome | Results | Risk of Bias |
|---|---|---|---|---|---|---|---|
| Tate et al., 2012 [17] | RCT parallel 3 groups |
318 obese, BMI 36.3 kg/m2, 84% female, medium age 42 y, USA | Advice to replace more than 2 servings (each 350–500 mL) of caloric beverages per day with water for six months | CG1: Advice to replace more than 2 servings of caloric beverages per day with diet beverages CG2: no advice to change beverages, but general nutritional weight loss advices |
FPG (SO) Weight loss (PO) |
Water group had a sign. Increase of water consumption of 1 L after 3 months and 0.8 L after 6 months and a sign. Decrease of FPG compared to control group with no change of beverages after 3 and 6 months. | ITT, drop-outs 33/318 after 3 months and 46/318 after 6 months, no blinding |
| Madjd et al., 2015 [18] | RCT parallel 2 groups single-blind |
62 healthy overweight and obese women, BMI 27–40 kg/m2, 18–50 y, non-smokers, participating in a weight loss program, Iran | 250 mL tap water (not specified) per day after lunch for 24 weeks | 250 mL diet beverage | FPG Fasting insulin HOMA-IR GTT Body weight (all PO) |
Water group sign. more weight loss (8.8 kg vs. 7.6 kg p = 0.015) lower FPG (2.8 mU/L vs. 1.7 mU/L p < 0.001) better insulin sensitivity (HOMA-IR 1 vs. 0.8 p < 0.001) better GTT (1 mmol/L vs. 0.7 mmol/L, p < 0.001) compared to control group after 24 weeks. |
No ITT, 9/71 drop-outs, measures of blinding not reported |
| Madjd et al., 2017 [19] | RCT parallel 2 groups single-blind |
81 overweight and obese women, BMI 27–35 kg/m2 with type 2 diabetes HbA1c 6.5–7.2%, age 18–50 y, non-smokers, participating in a weight loss program with only metformin, Iran | 250 mL tap water (not specified) per day after lunch for 24 weeks | 250 mL diet beverage | FPG Fasting insulin HOMA-IR GTT Body weight (all PO) |
Water group sign. more weight loss (−6.40 vs. −5.25 kg p = 0.006); lower FPG (−1.63 vs. −1.29 mmol/L p = 0.06) lower fasting insulin (−5.71 vs. −4.16 mU/L p < 0.011) better insulin sensitivity (HOMA-IR −3.2 vs. −2.48 p < 0.003) better GTT (−1.67 vs. −1.35 mmol/L p = 0.027) Compared to the control group after 24 weeks. |
ITT, 16/81 drop-outs, measures of blinding not reported |
| Tonstad et al., 2006 [20] | RCT parallel 3 groups |
67 men and 27 post-menopausal women (total n = 94) with at least 1 CVRF, Norway |
1 L tap water (not specified) per day for 4 weeks | CG1: 1 L blueberry juice CG2: no change of habits |
FPG Fasting insulin C-peptide (all SO) Blood viscosity (PO) |
No sign. difference between groups after 4 weeks (exact data not presented) | No ITT, 5/99 drop-outs, no blinding, |
| Maersk et al., 2012 [21] | RCT parallel 2 groups |
47 healthy overweight adults, 20–50 y, BMI 26–40 kg/m2, RR < 160/100, Denmark | 1 L uncarbonated water per day for 6 months Aqua d’or HCO3− 71 mg/L Cl− 16.5 mg/L Na+ 14.9 mg/L Mg2+ 6 mg/L |
1 L regular Cola (sucrose-sweetened) 1 L diet Cola, aspartame-sweetened 1 L semi-skim milk |
FPG Fasting insulin HOMA-IR (all SO) |
No sign. difference for FPG Insulin and HOMA-IR between groups |
No ITT, 13/60 drop-outs, no blinding, sign. difference for sex at baseline (adjusted in analysis), randomization and allocation not reported |
| Mori et al., 2016 [22] | RCT crossover 3 groups |
24 well controlled type 2 diabetes HbA1c < 8.5%, 19 men, 5 women, 40–70 y, regular alcohol intake women 20–30 g/day, men 30–40 g/day, Australia | Tap water (not specified) women 230 mL per day men 300 mL per day for 4 weeks |
CG1: red wine women 230 mL/day (24 g alcohol/day) Men 300 mL/day (31 g alcohol/day), CG2: dealcoholized red wine women 230 mL/day Men 300 mL/day |
FPG Fasting Insulin HOMA-IR (all SO) CVRF (PO) |
No sign. difference for FPG Insulin and HOMA-IR between groups |
No ITT, 4/28 drop-outs, no blinding |
| Gepner et al., 2015 [23] | RCT parallel 3 groups | 224 alcohol-abstaining adults type 2 diabetes, HbA1c 6.4–10%, Israel | 150 mL mineral water) per day in the evening for 2 years Mey Eden, different European sources, all low mineralized |
CG1: 150 mL red wine CG2: 150 mL white wine |
FPG HOMA-IR (all PO) Fasting insulin HbA1c (all SO) HDL-Chol, Apolip-a (PO) |
Water group sign. Increase of FPG and HOMA-IR compared to white wine. No sign. difference for HbA1c, fasting insulin and outcomes compared to red wine | ITT, no blinding |
| Rafraf et al., 2015 [24] & Zemestani et al., 2016 [31] | RCT parallel 2 groups single-blind |
64 type 2 diabetes (males and females) aged 30–60 y Iran |
150 mL hot water three times per day immediately after meals for 8 weeks. | 150 mL chamomile tea (3 g/150 mL hot water) | FPG Fasting Insulin HOMA-IR HbA1c |
Chamomile tea group sign. Decrease for HbA1c (p = 0.023), serum insulin levels (p < 0.001), HOMA-IR (p < 0.001), and no sign. difference for FPG compared to hot water | ITT, no drop-outs, single-blind, sign. difference for FPG, insulin and HOMA-IR at baseline, no PO specified |
CG: control group; PO: primary outcome; SO: secondary outcome; sign.: p < 0.05; RCT: randomized controlled trial; FPG: fasting plasma glucose; GTT: glucose tolerance test; CVRF: cardiovascular risk factors; y: years.