WELCOME 2015.
| Methods | RCT, parallel, (n3 EPA + DHA vs MUFA), 15‐18 months Summary risk of bias: low |
|
| Participants | Patients with NAFLD N: 51 intervention, 52 control (analysed, 47 intervention, 48 control) Level of risk for CVD: moderate Male: 49% intervention, 67% control Mean age (SD): 48.6 (11.1) intervention, 54 (9.6) control Age range: not reported (18‐75 inclusion criteria) Smokers: 14.3% intervention, 11.8% control Hypertension: not reported Medications taken by ≥ 50% of those in the control group: lipid‐lowering drugs Medications taken by 20%‐49% of those in the control group: antihypertensives, metformin (data not provided by group) Medications taken by some, but < 20% of the control group: none reported Location: UK Ethenicity: not reported |
|
| Interventions | Type: supplement (Omacor capsules) Comparison: DHA + EPA vs MUFA Intervention: 4 g Omacor/d (providing 1.84 g EPA, 1.52 g DHA as ethyl esters)), 3.36 g/d EPA + DHA Control: 4 g olive oil capsules/d (providing; ALA 1%, oleic acid 67%, palmitic acid 15%, stearic acid 2%, n‐6 fat: 15%), 0.64 g/d PUFA Dose aim: increase 2.72 g/d PUFA, 1.22% E PUFA Baseline PUFA unclear Compliance by biomarkers: erythrocyte EPA + DHA both increased in intervention, not in control (EPA% 1.0%, SD 0.2% in control vs 2.4% SD 1.8% in intervention at latest point, DHA% 5.0 SD 1.0 in control, 7.1% SD 1.3% in intervention), no other fatty acids reported. TC remained 4.8 mmol/L in control but fell by 0.2 mmol/L to 4.7 mmol/L in intervention at 15‐18 months. Compliance by dietary intake: not reported
Compliance, other methods: assessed by recording the returned unused capsules, but results not reported Inclusion basis: no intention to increase total PUFA stated. Planned total PUFA increase 2.72 g/d PUFA, 1.22% E PUFA, > 10% higher than assumed 6% E from total PUFA at baseline. Confirmed by TC fall in intervention, no other biomarker or intake data PUFA dose: 1.22%E PUFA Duration of intervention: 15‐18 months |
|
| Outcomes | Main trial outcome: changes in mean liver fat %, changes in 2 liver fibrosis scores, change in serum biomarkers Dropouts: 4 intervention, 4 control Available outcomes: weight, BMI, lipids, BP, glucose, insulin sensitivity, body fat measures, liver enzymes, HbA1c, serum n‐3 fatty acids, trial authors provided details of diabetes diagnoses, % body fat, BP and carotid intima media thickness. Response to contact: yes |
|
| Notes | Trial funding: Omacor and placebo were provided by Pronova Biopharma through Abbott Laboratories, Southampton, UK. This work was supported by a National Institute for Health Research (NIHR) Southampton Biomedical Research Unit grant and by a Diabetes UK Allied Health Research training fellowship awarded to KGM (Diabetes UK. BDA 09/ 0003937). CDB, PCC and ES were supported in part by the NIHR Southampton Biomedical Research Centre (McCormick‐2015, p9; see WELCOME 2015) | |
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Low risk | Participants were block randomised by an independent clinical trials pharmacist ....randomised according to standardised procedures (computerised block randomisation) by a research pharmacist at University Hospital Southampton NHS Foundation Trust. Simple randomisation in blocks of 4... |
| Allocation concealment (selection bias) | Low risk | Only the clinical trials pharmacist was unblinded, and randomisation group allocation was concealed from all trial members throughout the trial. (McCormick‐2015, p2). |
| Blinding of participants and personnel (performance bias) All outcomes | Low risk | Paper states that only the clinical trials pharmacist was unblinded, and randomisation group allocation was concealed from all trial members throughout the trial. However, the trial register record states "single blind (investigator)". Although the capsules were identical, no information provided as to their smell and taste |
| Blinding of outcome assessment (detection bias) All outcomes | Low risk | As above |
| Incomplete outcome data (attrition bias) All outcomes | Low risk | The ITT included all participants randomised who had complete data (baseline and end‐of‐trial measurements), regardless of whether they were later found to be ineligible, a protocol violator, given the wrong treatment allocation, or never treated). (Scorletti 2014, p4; see WELCOME 2015) |
| Selective reporting (reporting bias) | Unclear risk | Prospectively registered September 2008, trial start September 2009, end February 2017. Outcome data for cardiac function not yet published (may be ongoing as trial only recently completed), though other CV measures reported |
| Attention bias | Low risk | Both groups had the same attention |
| Compliance | Low risk | Erythrocyte EPA + DHA both increased in intervention, not in control (EPA% 1.0%, SD 0.2% in control vs 2.4% SD 1.8% in intervention at latest point, DHA% 5.0 SD 1.0 in control, 7.1% SD 1.3% in intervention), no other fatty acids reported. TC remained 4.8 mmol/L in control but fell by 0.2 mmol/L to 4.7 mmol/L in intervention at 15‐18 months. |
| Other bias | Low risk | None noted |