NAT2 2013.
| Methods | Nutritional AMD Treatment‐2 (NAT2) RCT, parallel, (EPA + DHA vs MUFA), 36 months Summary risk of bias: low |
|
| Participants | Patients with early age related macular degeneration N: 150 intervention, 150 control Level of risk for CVD: high (92.5% intervention and 79.8 controls had past CVD) Men: 31.3% intervention, 39.5% control Mean age in years (SD): 73.9 (6.6) intervention, 73.2 (6.8) control Age range: 55‐85 Smokers: 6.7% intervention, 8.5% control Hypertension: 58% total (not reported by study arm) Medications taken by at least 50% of those in the control group: lipid‐lowering medication Medications taken by 20%‐49% of those in the control group: agents acting on renin‐angiotensin system, anti‐inflammatory and anti‐rheumatic products Medications taken by some, but less than 20% of the control group: insulin or blood sugar lowering drugs Location: France Ethnicity: unclear |
|
| Interventions | Type: supplement (fish oil capsule) Comparison: EPA + DHA vs MUFA Intervention: 3 daily fish oil capsules containing 1110 total n‐3 FAs (EPA: 270 mg/day DHA: 840 mg/day) and vit E: 6 mg/day. Dose: 1.1 g/d EPA + DHA Control: 3 × 602 mg olive oil capsules a day containing 0.2 g total PUFA and vit E: 0.09 g/d Compliance: assessed during visits from unused capsules and serum PUFA levels. Overall compliance over the 3 years; 69.4% intervention, 70.5% control Length of intervention: 36 months |
|
| Outcomes | Main study outcome: time to occurrence of choroidal new vessels (CNV) in the study eye from prospective assessment of fluorescein angiography Dropouts: 29 intervention, 34 control Available outcomes: all cause mortality, plasma lipids, adverse events, serum FAs Response to contact: yes (no added data) |
|
| Notes | TG data not used as presented as median (5th‐95th percentile) Study funding: Laboratoire Chauvin, Bausch & Lomb Inc |
|
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Low risk | QL Ranclin software was used to generate the randomisation list before enrolment. The patients and the study personnel both were blinded to the treatment assignment |
| Allocation concealment (selection bias) | Low risk | As above |
| Blinding of participants and personnel (performance bias) All outcomes | Low risk | The capsules had the same appearance, the same size, and the same weight (602 mg) in both DHA and placebo groups. No masking flavour was added to the capsules, which were otherwise odourless |
| Blinding of outcome assessment (detection bias) All outcomes | Low risk | Author confirmed blinding of outcome assessors |
| Incomplete outcome data (attrition bias) All outcomes | Low risk | Any temporary discontinuation of the treatment was considered to be a deviation from the study protocol. Discontinuation for more than 5 months was considered to be a major deviation from the study protocol. Participants who dropped out were taken in account in the survival analysis and occurrence of CNV and were counted at last angiography performed. |
| Selective reporting (reporting bias) | Unclear risk | ISRCTN98246501. Retrospectively registered May 2007, recruitment started December 2003, completed November 2008, key publication 2013 |
| Attention | Low risk | Same amount of time spend with both study arms |
| Compliance | Low risk | Assessed during visits from unused capsules and serum PUFA levels. Overall compliance over the 3 years; 69.4% intervention, 70.5% control |
| Other bias | Low risk | None noted |