Fiatarone Singh 2014

Methods	Design: international 4‐arm RCT with factorial design Quote study design: "randomized, fully‐factorial, double‐blind, double sham training‐controlled clinical trial" Recruitment period: not reported No. of centres involved: not reported Unit of randomisation: individuals No. randomised: 100 Number of arms considered in this review: 4 Maximum trial duration: 18 months Funding by non‐profit organisation: this study was funded by a National Health and Medical Research Council (NH&MRC) of Australia Dementia Research Grant, project grant ID No. 512672, from 2008 to 2011(https://www.nhmrc.gov.au). Additional funding for a research assistant position was sourced from the NHMRC Program Grant ID No. 568969, and the project was supported by the University of Sydney and the University of New South Wales Funding by commercial organisation: none reported Publication status: full‐text report
Participants	Type of MCI: MCI consistent with the Petersen 1999 criteria Patient flow: 24 randomised, 24 described at baseline in experimental group 1 (CT and sham physical exercise); 27 randomised, 27 described at baseline in experimental group 2 (CT and physical exercise); 27 randomised, 27 described at baseline in control group 1 (double sham); 22 randomised, 22 described at baseline in control group 2 (physical exercise and sham CT) Number of females overall: 68 of 100 (68%) Average age (SD) overall: 70 (6.7) years Average (SD) education: not reported Baseline cognitive function: instrument to measure baseline cognitive function not reported Selection criteria on cognition overall: Clinical Dementia Rating Algorithm (0 to 4): 0.14 (0.22); 71% rated 0, 29% rated 0.5; Mini Mental State Exam: 27 (1) (23 to 29) Ethnicity: not reported APOE: number of participants positive for APOE not reported
Interventions	Type of experimental intervention: computerised CT group, treatment duration 24 weeks; intervention provided in group format, under supervision Details of experimental intervention: "CT intervention involved computer‐based multimodal and multidomain exercises targeting memory, executive function, attention, and speed of information processing. The training used the COGPACK program". Participants also received progressive resistance training (PRT) performed with exercise or sham exercise (factorial design) Session duration: 75 minutes in experimental group Number of treatment sessions: 48 in experimental group Treatment frequency: 2/week in experimental group Maximum treatment duration: 24 in experimental group Type of control intervention: usual care, treatment duration 24 weeks; intervention provided in group format, under supervision Details of control intervention: sham cognitive consisted of watching 5 short National Geographic videos, followed by a set of 15 questions (3/video) regarding the presented material. Sham exercise consisted of stretching and seated callisthenics, designed so as not to notably increase heart rate or aerobic capacity, nor improve balance, enhance strength, or other physiological outcomes. PRT was performed with pneumatic resistance machines (Keiser Sports Health Equipment, Ltd., Gloucestershire, UK), which were used for training at high intensity, with 3 sets of 8 repetitions of each of 56 exercises/session for most major muscle groups (chest press, leg press, seated row, standing hip abduction, knee extension) Session duration: 60 minutes in control group Number of treatment sessions: 48 in control group Treatment frequency: 2/week in control group Maximum treatment duration: 24 in control group
Outcomes	Cognitive functioning outcomes considered Global cognitive functioning measured with ADAS‐Cog at 6 and 18 months, on a scale from not reported to not reported with lower values indicating benefit Episodic memory measured with Logical Memory II (delayed) at 6 and 18 months, on a scale from not reported to not reported with higher values indicating benefit* Executive functioning measured with WAIS‐III Similarities at 6 and 18 months, on a scale from not reported to not reported with higher values indicating benefit Speed of processing measured with SDMT at 6 and 18 months, on a scale from not reported to not reported with higher values indicating benefit Verbal fluency measured with COWAT at 6 and 18 months, on a scale from not reported to not reported with higher values indicating benefit Physical functioning outcome considered Daily function measured with BAYER‐ADL scale at 6 and 18 months, on a scale from not reported to not reported with lower values indicating benefit Quality of life outcome considered: none reported Safety outcome considered: none reported Depression outcome considered: none reported Available cognitive functioning outcomes not considered in this review Global cognitive functioning measured with Global Cognition Domain at 6 and 18 months, on a scale from not reported to not reported with higher values indicating benefit Episodic memory measured with BVRT at 6 and 18 months, on a scale from not reported to not reported with higher values indicating benefit Episodic memory measured with Logical Memory I (immediate) at 6 months, on a scale from not reported to not reported with higher values indicating benefit* Executive functioning measured with WAIS‐III Matrices at 6 and 18 months, on a scale from not reported to not reported with higher values indicating benefit Verbal fluency measured with Category Fluency at 6 and 18 months, on a scale from not reported to not reported with higher values indicating benefit *Our hierarchy did not indicate a preference for the delayed subscale over the immediate subscale. Whenever both immediate and delayed subscales were available, the delayed subscale was included in the meta‐analyses, as it was thought to be more clinically relevant
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Judgement: adequate method of random sequence generation Quote(s): "a concealed, computer‐generated sequence of randomly permuted blocks.. in a 1:1:1:1 ratio to each of the 4 intervention arms, stratified by sex and age (<75 and 75 þ years), was generated by a research assistant not otherwise involved in the study via a statistical website"
Allocation concealment (selection bias)	Low risk	Judgement: adequate method of concealment allocation Quote(s): "assignments were then placed in sealed opaque envelopes and delivered to participants by the recruitment officer"
Blinding of participants (performance bias)	Unclear risk	Judgement: study described as double‐blinded; however, it is not clear if patients were blinded Quote(s): "all training was fully supervised by research assistants from exercise physiology or physical therapy backgrounds"
Blinding of physicians / personnel	High risk	Judgement: researchers supervising training were not blinded Quote(s): "all training was fully supervised by research assistants from exercise physiology or physical therapy backgrounds"
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Judgement: blinded outcome assessment Quote(s): "blinded assessors administered all outcome measures at baseline, 6 and 18 months"
Incomplete outcome data (attrition bias) All outcomes	Low risk	Judgement: Comparison 1: 24 out of 24 (100%) randomised were analysed in experimental group 1, and 27 out of 27 (100%) randomised were analysed in control group 1 Comparison 2: 27 out of 27 (100%) randomised were analysed in experimental group 2, and 22 out of 22 (100%) randomised were analysed in control group 2. Statistical analyses were reported to be done according to the intent‐to‐treat principle Quote(s): "all patients randomised were included in the analysis"; "n = 100 for all outcomes"
Selective reporting (reporting bias)	Low risk	Judgement: all outcomes indicated in the methods section are reported in the results section
Other bias	Low risk	Judgement: no other sources of bias are apparent