Methods

Parallel group RCT, 2 x 2 factorial design

Both eyes included in the trial, both eyes received same treatment, adjustment made for within person correlation

Participants

Country: USA

Setting: community

Number of participants: 2080, 55% women

Average age: 74 years

Age range: 50 to 85 years

Inclusion criteria:

• bilateral large drusen or large drusen in 1 eye and advanced AMD in the fellow eye

• consent to follow‐up of at least 5 years

• took at least 75% of the run‐in supplements and agreed to stop the use of other supplements containing lutein, zeaxanthin, DHA, EPA, vitamin C, vitamin E, beta‐carotene, zinc, or copper

Exclusion criteria:

• other ocular diseases such as high myopia, glaucoma, clinically significant diabetic retinopathy (10 or more microaneurysms or retinal haemorrhages), and other diseases that might confound the assessment of the ocular outcome measurements

• eyes that had undergone intraocular (apart from cataract) surgeries

• systemic diseases, including oxalate kidney stones, Wilson disease, haemochromatosis, lung cancer, or other diseases associated with poor 5‐year survival

Approximately 90% of participants were taking an additional multivitamin supplement

Interventions

• Omega 3 fatty acids (n = 1068 people, 1753 eyes)

• Placebo (n = 1012 people, 1695 eyes)

Omega 3 fatty acids were DHA (350 mg per day) and EPA (650 mg per day). Composition of placebo not specified

All participants were asked to take the original AREDS formulation (vitamin C 500 mg, vitamin E 400 IU, beta‐carotene 15 mg, zinc oxide 80 mg, cupric oxide 2 mg). Those who agreed to take AREDS and consented to a second randomisation were assigned as follows

• Original AREDS formula: omega 3 fatty acids group n = 147 (13.8%); placebo group n = 168 (16.6%)

• No beta‐carotene: omega 3 fatty acids group n = 231 (21.6%); placebo group n = 201 (19.9%)

• Low‐dose zinc (25 mg): omega 3 fatty acids group n = 179 (16.8%); placebo group n = 184 (18.2%)

• No beta‐carotene and low‐dose zinc: omega 3 fatty acids group n = 201 (18.8%); placebo group n = 190 (18.8%)

The participants who did not agree to a secondary randomisation largely took the AREDS formula: omega 3 fatty acids group n = 305 (28.6%); placebo group n = 265 (26.2%)

Participants who were current smokers or former smokers who had stopped smoking within the year before enrolment were randomly assigned to 1 of the 2 arms without beta‐carotene

Duration: 5 years

Outcomes

Primary outcome:

• Development of advanced AMD, defined as central geographic atrophy or retinal features of choroidal neovascularization detected on central grading of the stereoscopic fundus photographs or a history of treatment for advanced AMD after study enrolment

Secondary outcomes:

• Progression to moderate vision loss (3 lines) from baseline or treatment for choroidal neovascularisation

• Serious adverse events

• Mortality

Follow‐up: annually

Dates participants recruited

10/2006 to 09/2008

Declaration of interest

Yes ‐ reported in paper. Including patent for AREDS formula

Sources of funding

This study was supported by the intramural program funds and contracts from the National Eye Institute (NEI), National Institutes of Health (NIH), Department of Health and Human Services, Bethesda, Maryland (contract HHS‐N‐260‐2005‐00007‐C; ADB contract N01‐EY‐5‐0007). Funds were contributed by the following NIH institutes: Office of Dietary Supplements; National Center for Complementary and Alternative Medicine; National Institute on Aging; National Heart, Lung, and Blood Institute; and National Institute of Neurological Disorders and Stroke. The study medications and raw materials were provided by Alcon, Bausch & Lomb, DSM, and Pfizer

Notes

In the primary randomisation 84% of participants took 75% of the study medications

http://clinicaltrials.gov/show/NCT00345176

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

“A random block design was implemented using the AREDS2 Advantage Electronic Data Capture system by the AREDS2 Coordinating Center (The EMMES Corporation, Rockville, MD) and stratified by clinical center and AMD status (large drusen both eyes or large drusen in 1 eye and advanced AMD in the fellow eye) to ensure approximate balance across centers over time.” Page 2285 of protocol paper

Allocation concealment (selection bias)

Low risk

Placebo‐controlled study

“Participants and study personnel were masked to treatment assignment in both randomizations.”

Page E2 of main study report

Blinding of participants and personnel (performance bias)
Visual acuity

Low risk

“Participants and study personnel were masked to treatment assignment in both randomizations.” Page E2 of main study report

Blinding of participants and personnel (performance bias)
Progression of AMD

Low risk

“Participants and study personnel were masked to treatment assignment in both randomizations.” Page E2 of main study report

Blinding of outcome assessment (detection bias)
Visual acuity

Low risk

“Participants and study personnel were masked to treatment assignment in both randomizations.”

Page E2 of main study report

Blinding of outcome assessment (detection bias)
Progression of AMD

Low risk

“Participants and study personnel were masked to treatment assignment in both randomizations.”

Page E2 of main study report

CNV was determined by masked readers from stereoscopic fundus photographs

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Follow‐up was high and balanced across groups

• DHA/EPA: 1062/1068 (99.4%)

• Placebo: 1007/1012 (99.5%)

Selective reporting (reporting bias)

Low risk

Not detected

Methods

Parallel‐group RCT

One eye only included, study eye was selected on the basis of early AMD with neovascular AMD (CNV) in the fellow eye

Participants

Country: France

Setting: community

Number of participants: 300, 65% women

Average age: 74 years

Age range: 55 to 85 years

Inclusion criteria:

• bilateral large drusen or large drusen in 1 eye and CNV in the fellow eye (grading performed using a validated classification grid http://www.ncbi.nlm.nih.gov/pubmed/16988630)

• visual acuity better than 0.4 logarithm of minimum angle of resolution units in the study eye

• patients likely to attend follow‐up visits during the study period and consent to follow‐up of at least 5 years

Exclusion criteria:

• CNV in both eyes or no CNV in either eye

• wide central subfoveal atrophy of the study eye

• progressive ocular diseases (severe glaucoma or other severe retinopathy)

• major corneal or lens opacities precluding retinal evaluation

• serious systemic disease (cancer, stroke, etc.) preventing long‐term participation

• known allergy to the substances used in the study (fish oil, fluorescein, indocyanine green)

• anticoagulant therapy (prohibited medication) or bleeding tendency

• current or recent treatment (< 6 months) with nutritional supplements (oral supplement containing long‐chain omega 3 fatty acids or alpha tocopherol acetate)

• any concomitant nutritional supplement

• participation in a clinical trial within the previous 30 days

• history of drug use or excessive use of medication

• patients likely to be lost to follow‐up or unlikely to comply with the study protocol

• monocular patients for reasons other than AMD

• patients not covered by the French National Health system or wards of the court

Interventions

• Omega 3 fatty acid (n = 150 people)

• Placebo (n = 150 people)

Omega 3 fatty acids were 3 fish oil capsules, each capsule contained: DHA (280 mg), EPA (90 mg) and vitamin E (2 mg) (Reti‐Nat, provided by Bausch & Lomb, Montpellier, France)

Placebo contained 602 mg of olive oil

Duration: 3 years

Outcomes

Primary outcome:

• time to occurrence of CNV in the study eye

Secondary outcome:

• percentage of patients in whom CNV developed

• changes in visual acuity from baseline (logMAR)

• visual acuity decrease of 15 letters on the Early Treatment Diabetic Retinopathy Study (ETDRS) chart

• drusen burden and progression, based on automatic detection of their number, size, and area on fundus photography

• changes in red blood cell membrane (RBCM) EPA plus DHA levels

• lens opacity

• blood lipids including fasting plasma lipoprotein profile

• signs of intolerance related to fish oil consumption

• occurrence of systemic adverse events

Follow‐up: annually

Dates participants recruited

12/2003 to 10/2005

Declaration of interest

Eric H Souied: Consultant and lecturer—Laboratoire Bausch & Lomb Chauvin
Pascale Benlian: Financial support and lecturer—Laboratoire Bausch & Lomb Chauvin

Cécile Delcourt: Consultant and financial support—Laboratoire Bausch & Lomb Chauvin; Consultant and financial support—Laboratoires Théa; Consultant—Novartis

Sources of funding

Sponsored by Laboratoire Bausch & Lomb Chauvin, Montpellier

Notes

http://www.controlled‐trials.com/ISRCTN98246501

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

"QL‐Ranclin software (Qualilab, Olivet, France) was used to generate the randomization list before enrolment."

Souied et al 2013 p3

Allocation concealment (selection bias)

Low risk

"The patients and the study personnel both were blinded to the treatment assignment."

Souied et al 2013 p3

Blinding of participants and personnel (performance bias)
Visual acuity

Low risk

Not well reported. However, Qualilab provides an independent trial auditing service (not clear if this was the case here). No information provided regarding the outcome assessors (?study personnel), however it is likely that they remained masked as to the allocation

Blinding of participants and personnel (performance bias)
Progression of AMD

Low risk

Not well reported. However, Qualilab provides an independent trial auditing service (not clear if this was the case here). No information provided regarding the outcome assessors (?study personal), however it is likely that they remained masked as to the allocation

Blinding of outcome assessment (detection bias)
Visual acuity

Low risk

Not well reported. However, Qualilab provides an independent trial auditing service (not clear if this was the case here). No information provided regarding the outcome assessors (?study personal), however it is likely that they remained masked as to the allocation

Blinding of outcome assessment (detection bias)
Progression of AMD

Low risk

Not well reported. However, Qualilab provides an independent trial auditing service (not clear if this was the case here). No information provided regarding the outcome assessors (?study personal), however it is likely that they remained masked as to the allocation

Incomplete outcome data (attrition bias)
All outcomes

Unclear risk

Used a per protocol analysis. Main reason for protocol deviation was premature withdrawal which occurred at a similar rate in DHA and placebo groups. Other protocol deviations included ‘non‐compliance with study medication or use of non‐permitted medication’; 263 of the original 300 patients randomised were included in the analysis

Selective reporting (reporting bias)

Low risk

All pre‐specified primary outcomes reported. All secondary outcomes (with the exception of mERG listed in trial protocol) were reported