S‐Connect study 2013.

Study characteristics
Methods	Randomised, double‐blind, placebo‐controlled trial Trial duration from March 2009 to March 2011
Participants	Country: USA Diagnosis: mild‐to‐moderate Alzheimer’s disease (mean MMSE 20) Follow‐up: 24 weeks Inclusion criteria: age ≥ 50 years; diagnosis of probable AD (NINCDS‐ADRA); MMSE score between 14 and 24 inclusive; use of US Food and Drug Administration‐approved AD medication on a stable dose for at least 4 months prior to baseline; and availability of a responsible study partner Exclusion criteria: diagnosis of a neurological/psychiatric disease significantly contributing to cognitive difficulties other than AD; depression; recent use of potent anticholinergic agents, antipsychotics, omega‐3 fatty acid‐containing supplements and/or oily fish consumption more than twice a week, high‐energy or high‐protein nutritional supplements or medical foods, vitamins B, C and/or E containing supplements at > 100% of daily value, or other investigational products; recent change in lipid‐lowering medications, antidepressants, or antihypertensives; alcohol or drug abuse in the opinion of the investigator; or institutionalisation in a nursing home Setting: participants were recruited from community or clinic setting but not institutionalised in a nursing home Total number of participants: n = 527 (n = 265 Souvenaid group / n = 262 control group) Per protocol population: n = 451 (n = 228 Souvenaid group / n = 223 control group) Baseline characteristics: age, mean (SD): Souvenaid group 76.6 (8.2) , control group 76.9 (8.2) women in percentage: Souvenaid group 52%, control group 52%; years of education beyond primary school, mean (SD): Souvenaid group 6.7 (3.6), control group 6.4 (3.5); MMSE, mean (SD): Souvenaid group 19.5 (3.2), placebo group 19.4 (3.0); body mass index, kg/m2 (SD): Souvenaid group 26.2 (4.5), placebo group 26.6 (4.6); apolipoprotein e4 positive carrier, n (%): Souvenaid group 135 (60.8), control group 116 (58.0); participants used acetylcholinesterase inhibitor (%): Souvenaid group 87 (32.8), control group 92 (35.1); participants used memantine (%): Souvenaid group 13 (4.9), control 19 (7.3) ; participants used acetylcholinesterase inhibitor and memantine combined (%): Souvenaid group 164 (61.9), control group 151 (57.6)
Interventions	Intervention 1: Souvenaid 125 mL drink/once daily for 24 weeks which contains Fortasyn Connect™ (DHA, EPA, phospholipids, choline, UMP, vitamin B12, B6, and folate, vitamins C and E, and selenium) (see Table 4). Participants chose vanilla or strawberry flavours based on personal taste preferences Intervention 2: Iso‐caloric control product 125 ml/once daily for 24 weeks. Product lacked Fortasyn Connect™, but was otherwise identical to the active product with identical taste profile and appearance. Participants chose vanilla or strawberry flavours based on personal taste preferences
Outcomes	Primary (24 weeks): The Alzheimer’s Disease Assessment Scale ADAS‐cognitive subscale (ADAS‐Cog) (Rosen 1984) Secondary (24 weeks): The Alzheimer’s Disease Cooperative Study‐Activities of Daily Living (ADCS‐ADL) (Galasko 1997) Cognitive test battery named Global cognitive function composite z‐score; 4 components: Digit Span from the Wechsler Memory Scale – Third Edition, the Concept Shifting Test, the Letter Digit Substitution Test, and Category Fluency The Clinical dementia rating scale‐Sum of Boxes (CDR‐SOB) (O'Bryant 2008) Safety and tolerability Nutritional blood parameters assessed are not included in this review
Methods of AE Assessment	Safety parameters: examination of participant medical history, (serious) adverse events, concomitant medication, nutritional supplement, study product compliance, vital signs and safety laboratories for liver function, renal function, and coagulation Assessment not specified Monitoring of the parameters over a period of 24 weeks
Notes	Funding and methods used to control bias resulting from conflict of interest: see Table 5
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	S‐Connect study 2013: Quote: "Allocation to active or control product was performed through a central randomization procedure in the Electronic Data Capture system using four different randomization codes (A, B, C, and D)." p. 2
Allocation concealment (selection bias)	Low risk	S‐Connect study 2013: Quote: "Allocation to active or control product was performed through a central randomization procedure in the Electronic Data Capture system using four different randomization codes (A, B, C, and D)." p. 2
Blinding of participants and personnel (performance bias) All outcomes	Low risk	S‐Connect study 2013: Quote: "Participants, study partners, and study staff were masked to study group assignment during the trial. Unmasking did not occur until initial statistical modeling of the primary outcome was complete." p. 2
Blinding of outcome assessment (detection bias) All outcomes	Low risk	S‐Connect study 2013: Quote: "Participants, study partners, and study staff were masked to study group assignment during the trial. Unmasking did not occur until initial statistical modeling of the primary outcome was complete." p. 2
Incomplete outcome data (attrition bias) All outcomes	Low risk	Provided ITT The reasons for discontinuation from 70 participants are unclear apart from SEA but drop outs are balanced in numbers and overall drop out 14% (IG 14%, CG 15%).
Selective reporting (reporting bias)	Low risk	All outcomes reported as described in protocol, study conclusion congruent,with results and relevant AEs/SAEs are named with distribution to group
Other bias	Low risk	The model specifications for the outcomes measured with "trajectories over time" could be resolved by personal communication with the sponsor