Bartlett 2007.

Methods	Parallel group RCT Method of allocation: sponsor prepared coded tablets Masking: participant ‐ yes; provider ‐ yes; outcome ‐ yes Losses to follow‐up: 5 (2 treatment, 3 control)
Participants	Country: UK Number of people randomised: 30 (30 eyes) Number (%) of people followed up: 25 (83%) (25 eyes) Average age (range): 69 years (55 to 82) Percentage women: 53% Ethnic group: 100% white Baseline visual acuity: average visual acuity in intervention group was 0.20 logMAR and in control group was 0.08 logMAR Comorbidities affecting the eye: unknown Percentage current smokers: unknown Inclusion criteria: provide written informed consent be available to attend one of the research centres present with no ocular pathology in at least 1 eye, or no ocular pathology other than soft or hard drusen, and areas of increased or decreased pigment associated with drusen. Fundus examination was used to determine the presence of AMD. Exclusion criteria: type I and II diabetes prescribed antiplatelet or anticoagulant medication concurrent use of nutritional supplements advanced AMD in 1 or both eyes
Interventions	Intervention: lutein esters 6 mg, retinol 750 mg, vitamin C 250 mg, vitamin E 34 mg, zinc 10 mg, copper 0.5 mg (daily) 17 people randomised (17 eyes) 15 (88%) people followed up (15 eyes) Comparator: placebo tablets containing cellulose (daily) 13 people randomised (13 eyes) 10 (77%) people followed up (10 eyes) Duration: 9 months Similarity between intervention and comparator: Quote: "The study formulation and placebo tablets were produced by Quest Vitamins Ltd, and were identical in external and internal appearance, and taste."
Outcomes	Primary: unknown Secondary: unknown Outcome measures specified on trial registration entry: Distance and near visual acuity (VA) measured using Bailey‐Lovie logMAR charts Contrast sensitivity (CS) measured using a Pelli‐Robson chart (Clement Clarke International, Harlow Essex, UK) Colour vision measured using the PV‐16 quantitative colour vision test Macular Mapping (MM) test Eger Macular Stressometer (EMS) used to assess glare recovery Fundus photographs of the macular will be assessed using colour and edge analysis software Trial publication provided data on contrast sensitivity at 9‐month follow‐up. Protocol listed more outcomes (see below under selective reporting) and specified 9 and 18 months follow‐up. Follow‐up: 9 months (reported) and 18 months (not reported) Eyes: Trial eye selected (initial visit only). If both eyes were eligible for inclusion, the right eye was used
Notes	Sample size calculations reported in trial report: "A group size of nine was calculated to be sufficient to provide 80% power at the 5% significance level for CS based on an effect size of 0.3 log units, and mean and standard deviation (SD) values taken from a sample of 50 ARM and atrophic AMD patients of the University optometry clinic (1.3970.22 log CS)." Sample size calculations reported in protocol paper: "From initial data collection we have calculated the treatment group sizes required in order to have 80% power at the 5% significance level for VA, CS, MM test, and the EMS. These values suggest that a total of 63 normal, and 96 age‐related macular disease participants are required." Source of funding: Quote: "The project was sponsored by the UK College of Optometrists. Intervention and placebo tablets were provided by Quest Vitamins Ltd UK." Declaration of interest: unknown Date study conducted: March 2003 and December 2004 Trial registration number: ISRCTN78467674 (registered retrospectively)
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	"The random number generator function in Microsoft Excel is being used to allocate participants to μ and λ groups. Odd numbers allocate to the μ group." "Only one investigator (HB) was involved in the randomization process, which employed the random number generator in Microsoft Excel for Windows XP. Odd and even numbers were used to identify group."
Allocation concealment (selection bias)	Low risk	"Enrolment was carried out by HB, who, along with FE, was masked to group assignment." "Only one investigator (HB) was involved in the randomization process, which employed the random number generator in Microsoft Excel for Windows XP. Odd and even numbers were used to identify group." "Investigators and participants do not know which symbol represents the placebo tablets, and which represents the active formulation."
Blinding of participants and personnel (performance bias) Visual acuity	Low risk	"The study formulation and placebo tablets have been produced by Quest Vitamins Ltd, Aston Science Park, Birmingham, B7 4AP, and are identical in external and internal appearance, and taste. The manufacturer has allocated distinguishing symbols, μ and λ. The tablets are packaged in identical, sealed, white containers; the only difference being the symbol on the label. Investigators and participants do not know which symbol represents the placebo tablets, and which represents the active formulation."
Blinding of participants and personnel (performance bias) Progression AMD	Low risk	Not reported
Blinding of outcome assessment (detection bias) Visual acuity	Unclear risk	"The study formulation and placebo tablets have been produced by Quest Vitamins Ltd, Aston Science Park, Birmingham, B7 4AP, and are identical in external and internal appearance, and taste. The manufacturer has allocated distinguishing symbols, μ and λ. The tablets are packaged in identical, sealed, white containers; the only difference being the symbol on the label. Investigators and participants do not know which symbol represents the placebo tablets, and which represents the active formulation." "End of trial assessment using questionnaires indicated`masking success. Out of those participants taking the placebo tablet, 10% correctly guessed which tablet they were taking, and 10% incorrectly guessed. Out of those taking nutritional supplement, 13% guessed correctly which tablet they were taking, and 7% incorrectly guessed. The remaining participants did not know which group they were randomised to."
Blinding of outcome assessment (detection bias) Progression AMD	Unclear risk	Not reported
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	"Statistical analysis was carried out on a per protocol basis."
Selective reporting (reporting bias)	High risk	Protocol report: following outcomes listed: visual acuity, contrast sensitivity, colour vision, macular mapping test, glare recovery, fundus photographs analysed by colour and edge analysis software. Trial report only reported contrast sensitivity (CS): Quote: "Outcome measure CS was measured using a Pelli‐Robson chart (Clement Clarke International, Edinburgh Way, Harlow, Essex, CM20 2TT, UK) and scored per letter."