Summary of findings and GRADE evidence profile
| Quality assessment | No. of patients | Weight loss information | Effect | Quality | |||||||||
| Studies with data | Study limitations | Inconsistency | Indirectness | Imprecision | Publication bias | Other | Study | Weight loss/No weight loss | Weight loss, kg (% of body weight) | Weight loss pr. month (% of body weight) | Effect on outcome | Overall conclusion | GRADE rating |
| Serum uric acid (μmol/L) | |||||||||||||
| 7 | Serious (−1) | Not serious | Not serious | Not serious | Not serious | Dose response (+1) | Nguyen | NA/NA | – | – | NA | Weight loss results in a decrease of sUA after medium/long follow-up. | Low ⨁ ⨁ ◯ ◯ |
| Dalbeth (part 1) | 12/0 | 5.5 kg (3.9) | 0.9 kg (0.7) | 30* (−48 to 108) | |||||||||
| Dalbeth (part 2) | 12/0 | 34 kg (24.3) | 2.8 kg (2.0) | −110* (−188 to −32) | |||||||||
| Romero-Talamás | 99/56 | 29 kg† (23) | 2.4 kg (1.8) | −168 (−213 to −123) | |||||||||
| Zeng | 30‡/31 | 3.6 kg (4.8) | 0.6 kg (0.8) | −47 (−66 to −27) | |||||||||
| Perez-Ruiz | 25§/167§ | 5.6 kg (6.9) | 0.2 kg (0.2) | 17 (-40 to 73) | |||||||||
| Zhu | NA/NA | – | – | Dose–response relationship between weight change and sUA change¶ | |||||||||
| Barskova | 23‡/0 | 3 kg† (3.2) | 0.3 kg (0.3) | −127* (−183 to −71) | |||||||||
| Friedman | 21/0 | NA | NA | NA | |||||||||
| Dessein | 13/0 | 7.7 kg** (8.4) | 2.1 kg (2.3) | −100*†† (−174 to −26) | |||||||||
| Brandstetter | 11‡/11 | 5.7 kg (6.9) | 1.0 kg (1.1) | −2 (−116 to 113) | |||||||||
| Achieving serum uric acid target that is, sUA<360 μmol/L | |||||||||||||
| 4 | Serious (−1) | Not serious | Not serious | Not serious | Not serious | Large effect (+1) Dose response (+1) | Nguyen | NA/NA | – | – | NA | Weight loss results in a higher chance of achieving sUA target after medium/long follow-up. | Moderate ⨁ ⨁ ⨁ ◯ |
| Dalbeth (part 1) | 12/0 | 5.5 kg (3.9) | 0.9 kg (0.7) | 0% (0/10) reduction in pts with raised sUA* | |||||||||
| Dalbeth (part 2) | 12/0 | 34 kg (24.3) | 2.8 kg (2.0) | 60% (6/10) reduction in pts with raised sUA* | |||||||||
| Romero-Talamás | 99/56 | 29 kg† (23) | 2.4 kg (1.8) | NA | |||||||||
| Zeng | 30‡/31 | 3.6 kg (4.8) | 0.6 kg (0.8) | NA | |||||||||
| Perez-Ruiz | 25§/167§ | 5.6 kg (6.9) | 0.2 kg (0.2) | 1% reduction in pts with raised sUA | |||||||||
| Zhu | NA/NA | – | – | Dose–response relationship between weight change achieving serum uric acid target** | |||||||||
| Barskova | 23‡/0 | 3 kg† (3.2) | 0.3 kg (0.3) | 48% (11/23) reduction in pts with raised sUA* | |||||||||
| Friedman | 21/0 | NA | NA | NA | |||||||||
| Dessein | 13/0 | 7.7 kg **(8.4) | 2.1 kg (2.3) | 50% (6/12) reduction in pts with raised sUA*‡‡ | |||||||||
| Brandstetter | 11‡/11 | 5.7 kg (6.9) | 1.0 kg (1.1) | NA | |||||||||
| Gout attacks | |||||||||||||
| 8 | Serious (−1) | Not serious | Not serious | Not serious | Not serious | Dose response (+1) | Nguyen | NA/NA | – | – | Dose–response relationship between BMI change and recurrent gout attacks** | Weight loss results in fewer gout attacks after medium/long follow-up. | Low ⨁ ⨁ ◯ ◯ |
| Dalbeth (part 1) | 12/0 | 5.5 kg (3.9) | 0.9 kg (0.7) | 0 pts had ≥1 attack 6 months (At baseline two pts had ≥1 attack in 3 months)* | |||||||||
| Dalbeth (part 2) | 12/0 | 34 kg (24.3) | 2.8 kg (2.0) | Three pts had ≥1 attack in 12 months (At baseline for part 2, 0 pts had ≥1 attack in 6 months)* | |||||||||
| Romero-Talamás | 99/56 | 29 kg† (23) | 2.4 kg (1.8) | RR of 0.72 for ≥1 attack at follow-up | |||||||||
| Zeng | 30‡/31 | 3.6 kg (4.8) | 0.6 kg (0.8) | 39% fewer attacks at follow-up | |||||||||
| Perez-Ruiz | 25§/167§ | 5.6 kg (6.9) | 0.2 kg (0.2) | RR of 0.35 for ≥1 attack at follow-up | |||||||||
| Zhu | NA/NA | – | – | NA | |||||||||
| Barskova | 23‡/0 | 3 kg† (3.2) | 0.3 kg (0.3) | 33% fewer attacks*†† | |||||||||
| Friedman | 21/0 | NA | NA | 33% (7/21) pts had attacks*§§ | |||||||||
| Dessein | 13/0 | 7.7 kg** (8.4) | 2.1 kg (2.3) | 71% fewer attacks*†† | |||||||||
| Brandstetter | 11‡/11 | 5.7 kg (6.9) | 1.0 kg (1.1) | NA | |||||||||
Modified from table made with GRADEpro (computer program on www.gradepro.org), McMaster University, 2014. Effect on outcome was calculated as change from baseline to latest follow-up for NRS with one group, difference in changes for NRS with two groups and difference between groups at follow-up for RCTs, unless otherwise indicated. Effects on outcomes are presented as means (95% CI), numbers or RRs.
*Studies with only one group, hence the effect is not a contrast between groups.
† Weight loss estimated from BMI.
‡ On average a weight loss was seen for these pts as a group. Hence, some individuals may not have lost weight.
§Due to loss of data in the study, the number of pts in the groups were 25 and 167 at baseline, and 29 and 163 at last follow-up, respectively. For calculation of 95% CI, a mean, that is, 26 and 165, was used.
¶It should be noted that non-overweight and non-gout patients were included as well.
**It should be noted that non-overweight gout patients were included as well.
††Based on medians.
‡‡Dessein28 used a cut-off of ≤510 μmol/L.
§§For this study, no comparison (such as a comparison group or a measurement before the intervention) was provided. Hence, the number is the absolute number.
BMI, body mass index; GRADE, The Grading of Recommendations Assessment, Development and Evaluation; NRS, non-randomised study; pts, patients; RCT, randomised controlled trial; RR, risk ratio; sUA, serum uric acid.