Table 3.
The methodical properties of the reviewed studies.
| Study | Population | Research methods | Outcome measures | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| N | Mean age (SD) | Mean MMSE (SD) | Study design | Time points of cognitive evaluation | Diagnostic criteria | Randomization | Blinding | Allocation concealment | Interval scaling | Practice effect | Missing data and drop-outs | Other statistical practices | Cognitive domain (tests) | Results | |
| Studies involving patients with AD | |||||||||||||||
| Ahmed et al. (2012) | 45 | 68.4 | 14.84 (5.5) | Double-blind, randomized, sham-controlled, parallel-group study | Baseline, day 5 (post-intervention), 1 month later, 3 months later (follow-up) | NINCDS-ADRDA | Method not specified | Patients and assessor blinded to group assignment | Using closed envelopes | Global cognitive performance (MMSE), daily activity (IADL) and depression (GDS) | Improvement in global cognitive performance and daily activity in HF-rTMS group compared to LF and sham groups | ||||
| Alcalá-Lozano et al. (2018) | 19 | 72.15 (5.15) | Group 1: 19.5 Group 2: 18.2 | Single-blind, randomized, parallel-group study | Baseline, week 3 (post-intervention), week 7 (follow-up) | DSM-5, MMSE score ≧15, GDS-Reisberg level 2–4 | Method not specified | Patients blinded to the type of stimulation | Explicitly reported no drop-outs | A priori sample size calculation, predefined cutoff scores indicating clinically significant change | Global cognitive performance (MMSE, ADAS-Cog) | Improvement in global cognitive performance immediately after 4 weeks of treatment, which remained 7 weeks later as well in both groups | |||
| Cotelli et al. (2006) | 15 | 76.6 (6.0) | 17.8 (3.7) | Randomized, sham-controlled, crossover study | Baseline, during stimulation | NINCDS-ADRDA | Method not specified | Language (picture [action and object] naming) | Improvement of action naming speed during the stimulation of both LDLPFC and RDLPFC | ||||||
| Cotelli et al. (2008) | 24 | 76.3 (6) | Randomized, sham-controlled, crossover study | Baseline, during stimulation | NINCDS-ADRDA | Method not specified | Language (picture [action and object] naming) | Improved action naming performance in the mild AD group and improved picture naming in the severe AD group after active stimulation | |||||||
| Cotelli et al. (2011) | 10 | 72.8 (4.95) | Double-blind, sham-controlled, parallel-group study | Baseline, week 2, seek 4 week 12 (follow-up) | NINCDS-ADRDA | Patients and assessor blinded to the type of stimulation | Global cognition (MMSE), (IADL), language (picture [object, action] naming, Battery for Analysis of Aphasic Deficits), auditory sentence comprehension subtest(SC-BADA) | Improvement in the active group in auditory sentence comprehension compared to baseline or placebo (even after 8 weeks) | |||||||
| Devi et al. (2014) | 10 | 73.1 (7.9) | 25.1 (5.8) | Single-arm, open-label study | Baseline, week 2 (post-intervention), week 4 (follow-up) | NINCDS-ADRDA | Allocation based on the order of recruitment | Global cognition (MMSE), language (BDAE) | Immediate improvement in verbal agility and delayed improvement in nonverbal agility | ||||||
| Eliasova et al. (2014) | 10 | 72 (8) | 23 (3.56) | Randomized, sham-controlled, crossover study | Baseline, retest within 30 min | Not defined | Method not specified | Tasks practiced before trial commencement | Global cognitive performance (ACE-R, MMSE), memory (RCFT, WMS-III), attention, psychomotor speed, working memory (Stroop task, TMT-A), executive functions (TMT B, verbal fluency tasks) | Enhancement of attention and psychomotor speed after right IFG stimulation after active stimulation | |||||
| Haffen et al. (2012) | 1 | 75 | 20 | Case study | 4 months before intervention (baseline), 1 month after stimulation period, 5 months after stimulation period (follow-up) | NINCDS-ADRDA | Baseline 4 months prior the commencement of stimulation period | Executive function (Isaacs Set Test), episodic memory (Memory Impairment Screen, Free and Cued Recall Test, Isaacs Set Test), information processing (TMT-A), visuospatial skills (copying geometric figure), naming | Improved performance on 8 of the 10 measures with maintained cognitive functioning at follow-up | ||||||
| Koch et al. (2018) | 14 | 70.0 (5.1) | 26.1 (1.8) | Double-blind, randomized, sham-controlled, crossover study | Baseline, week 2 (post-intervention) | Revised NINCDS-ADRDA criteria by Dubois et al. (2016) | Method not specified | Patients and assessor blinded to condition | Global cognition (ADCS-PACC, MMSE), attention and psychomotor speed (TMT) auditory verbal learning (RAVL-T), episodic memory (DSST) executive function (Modified Card Sorting test, Verbal fluency, FAB) | Improvement in active group in episodic memory, but not in global cognition and executive function | |||||
| Rutherford et al. (2015) | 11 | 57–87 | Double-blind, randomized, sham-controlled, crossover + open-label study | Stage 1: baseline, week 4 (post-intervention) | Diagnosed by neuropsychiatrist or neurologist or MOCA score between 5 and 26 | Method not specified | Patients and assessor blinded, the effectiveness to blinding was measured, when assessor was not blinded it got reported | Alternate versions of tasks used | Mean imputation used and reasons of drop-out reported | Calculating observed power of tests, average test-retest improvement calculated | Global cognitive performance and associative memory (ADAS-Cog, RMBC, spatial awareness, word–image association) | Improvement in global cognitive performance in the active group compared to sham, especially during the early stage of the treatment | |||
| Wu et al. (2015) | 54 | 15.25 (3.1) | 15.25 (3.1) | Double-blind, randomized, sham-controlled, parallel-group study | Baseline, week 4 (post-intervention) | NINCDS-ADRDA | Standard table of random numbers | Patients and assessor blinded to group assignment | Patients and assessor blinded to the group assignment before starting the trial, method not specified | Using cutoff scores based on the findings of other studies | Behavioral pathology (BPSD) and global cognitive performance (ADAS-Cog) | Improvement of behavioral and global cognitive symptoms | |||
| Zhao et al. (2017) | 30 | 70.8 (5.6) | 22.5 (2.7) | Prospective, double-blind, randomized, sham-controlled, parallel-group study | baseline, week 6 (post-intervention), week 12 (follow-up) | DSM IV | Method Not specified | Patients and assessors blinded to group assignment | Global cognition (MMSE, MoCA), verbal memory (WHO-UCLA AVLT) | Improvement in global cognitive performance in the active group, especially in mild AD regarding memory and language | |||||
| Bentwich et al. (2011) | 8 | 75.4 (4.4) | 22.9 (1.7) | Single-arm open-label study | Baseline 3 weeks prior treatment, after week 6 (post-treatment), 4.5 months later (follow-up) | DSM-IV criteria, MMSE score of 18–24, CDR score of 1 | Baseline 3 weeks prior the commencement of stimulation period | Drop-outs reported and reasoned, managing is not reported | Global cognitive performance (ADAS-cog) | Improvement in global cognitive performance after 6 weeks and 18 weeks | |||||
| Lee et al. (2016) | 27 | 71.6 (6.8) | 22.5 (2.7) | Prospective, double-blind, randomized, sham-controlled, parallel-group study | Baseline, week 3 (post-intervention), week 9 (follow-up) | DSM IV | Method not specified | Patients and assessor blinded to group assignment | Drop-outs reported and reasoned, managing is not reported | Global cognitive performance (MMSE, ADAS-Cog) depression (GDS), global function (CGIC) | Improvement in global cognitive performance and global functioning after 6 weeks compared to sham, especially regarding language and episodic memory in mild AD | ||||
| Nguyen et al. (2017) | 10 | 73.0 (7.2) | 18.8 (1.9) | Prospective, single-arm, open-label study | Baseline, week 6 (post-intervention), 6 months later (follow-up) | Not defined | Alternate versions of ADAS-Cog used | Global cognitive performance (ADAS-Cog, MMSE, Dubois score), executive functions (FAB, Stroop color test) | Improvement in global cognitive performance, but it diminished after 6 months and remained detectable only in good responders (with high baseline MMSE) | ||||||
| Rabey et al. (2013) | 15 | 74 (8.99) | 22 (1.52) | Double-blind, randomized, sham-controlled, parallel-group study | Baseline, week 6, biweekly follow-up for 3 months | DSM-IV, MMSE score of 18–24, Clinical Dementia Rating score of 1 | Method not specified | Patients and assessor blinded to group assignment | Drop-outs reported and reasoned, principal investigator decided about the randomness of dropouts; last observation was carried forward method | Global cognitive performance (ADAS-cog) and daily activity (IADL) | Improvement in global cognitive performance and daily activity in HF, compared to sham | ||||
| Rabey and Dobronevsky (2016) | 30 | 22.2 (0.5) | Single-arm, open-label study | Baseline, week 6 (post-intervention) | Not defined | Alternate versions of ADAS-Cog used | Multiple imputation used on missing values with sensitivity analyses for observed data only and for worst-case analysis, reported the results of both analyses | Global cognitive performance (ADAS-Cog, MMSE) | Improvement in global cognitive performance on both scales | ||||||
| Avirame et al. (2016) | 11 | 76 (7) | Single-arm open-label study | Baseline, 2–3 weeks later (post-intervention) | Diagnosed by an expert neurologist and confirmed by a psychiatrist | Different stimuli within the tasks | Missing data reported and reasoned, managing is not reported | Global cognitive performance (Mindstreams, ACE) | Improvement of global cognition compared to baseline | ||||||
| Penolazzi et al. (2015) | 1 | 60 | 23.2 | Case study | Two cycles of baseline, week 4 (post-intervention), week 8 (follow-up), 2 months apart | Based on neuropsycholgical evaluation and neuroimaging | Patient blind to the stimulation, method not specified | Comparison to a normative score | Memory (Brief Neuropsychological Examination-2), psychomotor speed and executive function (TMT A and B, clock drawing) | Improvement on the trained tasks whith more enhancement when training was combined with active stimulation | |||||
| Andrade et al. (2016) | 1 | 73 | Case study | Baseline (1 week prior), 1 week after the intervention | NINCDS-ADRDA | Baseline 1 week prior to the commencement of the stimulation period | Global cognitive performance (ADAS-Cog), neuropsychiatric and behavioral symptoms (NPI, DAD, Blessed Dementia Scale) | Improvement of global cognitive performance, executive function and behavioral symptoms compared to baseline | |||||||
| Boggio et al. (2009) | 10 | 79.1 (8.8) | 17.0 (4.9) | Single-blind, randomized, sham-controlled, crossover study | During stimulation | NINCDS-ADRDA | Method not specified | Patients blinded to the type of stimulation | Randomized use of alternate versions | Selective attention (Stroop test, Victoria version), working memory (Digit span test backward and forward), recognition memory (visual memory task using IBV software) | Improvement of visual recognition memory after LDLPFC and temporoparietal stimulation compared to sham | ||||
| Boggio et al. (2012) | 15 | 79.05 (8.2) | 20 (3) | Double-blind, randomized, sham-controlled, crossover study | Baseline, day 5 (post-treatment), week 2, week 4 (follow-up) | NINCDS-ADRDA and DSM-IV | Method not specified | Patients and assessor blinded to group assignment | Randomized use of alternate versions of tasks | Global cognition (MMSE, ADAS-Cog), visual recognition (VRT), visual attention (VAT) | Improvement of memory performance in active stimulation group | ||||
| Bystad et al. (2016) | 25 | 72.5 (8.35) | 20.6 (3.35) | Double-blind, randomized, sham-controlled, parallel-group study | Baseline, day 6 (post-intervention) | Revised NINCDS-ADRDA | Computer randomized list containing 5-digit codes provided by the manufacturer of the tDCS device | Patients and assessor blinded to the type of stimulation | Assignment disclosed until the end of the intervention | Scaling according to standardized norm tables, transformation to z-scores | Two versions of CVLT-II used | Explicitly reported no drop-outs | Sample size based on other studies | Global cognitive performance (MMSE), Verbal learning (CVLT-II), Attention and executive function (TMT, clock-drawing test | No changes in either cognitive function |
| Bystad et al. (2017) | 1 | 60 | 20 | Case study | Baseline, 5 months later (during stimulation period), 8 months later (post-intervention) | Revised NINCDS-ADRDA | Alternate versions used | Global cognition (RBANS) | Stabilized cognitive decline of patient with minor impairment of visuospatial function | ||||||
| Suemoto et al. (2014) | 40 | 80.5 (7.5) | 15.2 (2.85) | Double-blind, randomized, sham-controlled, parallel-group study | Baseline, week 2 (post-intervention), week 3 (follow-up) | NINCDS-ADRDA | Computer-generated list of random numbers | Patients and assessor blinded to condition, numbered | Opaque and sealed envelopes | Reasons of missing data not reported, intention to treat analyses conducted using the method of last observation carried forward | A priori sample size calculation, using the method of minimal clinically relevant difference, planned pairwise comparisons | Apathy (Apathy Scale), global cognitive performance (MMSE, ADAS-Cog) | No change in active and sham group | ||
| Ferrucci et al. (2008) | 10 | 75.2 (7.3) | 22.7 (1.8) | Double-blind, randomized, sham-controlled, crossover study | baseline, 30 min after (post-intervention) | NINCDS-ADRDA and DSM IV | Method not specified | Patients and assessor blinded to condition | Alternate versions used | Recognition memory (WRT), visual attention (modified Posner task) | Anodal stimulation improved, while cathodal stimulation decreased word recognition comparing to sham | ||||
| Marceglia et al. (2016) | 7 | 75.4 (7.2) | 22.4 (1.39) | Double-blind, randomized, crossover study | baseline, 30 min later (post-intervention) | NINCDS-ADRDA | Method not specified | Alternate versions used | Recognition memory (WRT) | Improvement on WRT after anodal stimulation that correlated with increased delta and theta power measured by EEG | |||||
| Khedr et al. (2014) | 34 | 69.7 (4.8) | mild: 23–19, moderate: 18–11 | Double-blind, randomized, sham-controlled, parallel-group study | baseline, end of 10th session (post-intervention), 1 month and 2 months later (follow-up) | NINCDS-ADRDA | Computer generated randomization table | Patients and assessor blinded to group assignment, the effectiveness of blinding was measured | Serials numbered opaque closed envelopes | Reportedly no drop-outs | Global cognitive performance (MMSE and WAIS-III) | Improvement in MMSE after both anodal and cathodal tDCS in contrast to sham, improvement in performance IQ after cathodal stimulation | |||
| Studies involving patients with MCI | |||||||||||||||
| Turriziani et al. (2012) | 8 | 66.4 (5.7) | 26.9 (2) | Sham-controlled, crossover study | Criteria of Petersen et al. (1999) | Non-verbal recognition memory (faces and buildings recognition) | Improvement in non-verbal recognition memory compared to sham condition | ||||||||
| Drumond Marra et al. (2015) | 34 | 65.15 (3.8) | 24.35 (2.05) | Double-blind, randomized, sham-controlled, parallel-group study | Baseline, week 2 (post-intervention), 1 months later (follow-up) | Not specified, MoCA <26 | Computer generated randomization | Patients and assessors blinded to group assignment, the effectiveness of blinding was measured | Different staff members responsible for the allocation | Scores adjusted according to age, gender and education level | Everyday memory (RBMT), global cognitive function (MMSE), logical memory (WMS I, II), memory (RAVLT), working memory (WAIS III), psychomotor speed, executive function (TMT, verbal fluency tasks, Victoria Stroop Test) | Selective improvement in everyday memory compared to sham group | |||
| Padala et al. (2018) | 6 | 66(9) | Double-blind, randomized, sham-controlled, crossover study | Baseline, week 2 (post- intervention), week 6 (end of treatment-free period), and week 8 (post -intervention), week 12 (end of treatment-free period) | Criteria of Petersen et al. (1999) | Randomized block design | Patients and assessors blinded to condition | Independent staff member responsible for the allocation | Random subject effect calculated | Drop-outs reported and reasoned | Baseline measurements set as covariates | Apathy (AES-C), global cognitive performance (MMSE, 3MS), executive function (TMT, EXIT-25), global clinical evaluation (CGI), daily activity (IADL, ADL) | Improvement in apathy symptoms, global cognition, processing speed and clinical improvement compared to sham condition | ||
| Sole-Padulles et al. (2006) | 40 | 67.82 (8.6) | 26.33 (2.0) | Double-blind, randomized, sham-controlled, parallel-group study | Baseline, immediately after stimulation | MMSE ≥ 24, low performance in at least one predefined memory test | Patients and assessors blinded to condition | New stimuli used | Drop-outs reported and reasoned | Associative memory (face-name task) | Improvement in associative memory compared to sham group | ||||
| Cotelli et al. (2012) | 1 | 61 | 27 | Case study | Two baselines, week 2 (post-intervention), week 24 (follow-up) | Criteria of Petersen et al. (1999) | Repeated baseline evaluation, comparisons to a healthy control group | Global cognitive performance (MMSE), non-verbal reasoning (RCPM), memory (FNAT. story recall, AVLT, RCFT, spatial span, digit span), language (Token Test, verbal fluency tasks), praxia (De Renzi Imitation test), executive function (TMT, WCST) | Improvement in associative memory and encoding performance which was maintained for 24 weeks | ||||||
| Cruz Gonzalez et al. (2018) | 5 | 72.8 (6.65) | Single-blind, sham-controlled, crossover study | Screening, week 1 (baseline), week 2 (post-intervention), week 3 (post-intervention), week 4 (baseline) | Criteria of Portet et al. (2006) | Order of conditions has been kept the same across patients | Patients blinded to condition | Reducing the number of administrations of MoCA | Missing data reported, managing is not reported | Sensitivity of the measures is mentioned | Planning ability, processing speed, short-term memory, working memory (“Neuron Up” tablet-based tasks, digit span), processing speed, attention, executive function (TMT), global cognitive functioning (MoCA) | Improvement in processing speed, selective attention, planning ability and working memory compared to sham stimulation or cognitive stimulation alone | |||
| Meinzer et al. (2015) | 18 | 67.44 (7.27) | Double-blind, randomized, sham-controlled, crossover study | During online stimulation | Criteria of Albert et al. (2011) | Method not specified | Patients and assessors blinded to condition, the effectiveness of blinding was measured | z-transformation of scores | Semantic-word-retrieval (Boston naming test) | Improvement of semantic word-retrieval compared to sham condition | |||||
| Murugaraja et al. (2017) | 26 | 59.6 | Single-arm, open label study | Baseline, day 5 (post-intervention), 1 month later (follow-up) | Criteria of Albert et al. (2011) | Alternate versions used | Memory (PMIT) | Improvement in picture memory that persisted for 1 months | |||||||
AD, Alzheimer's disease; MCI, mild cognitive impairment; HF-rTMS, high frequency repetitive transcranial magnetic stimulation; LF-rTMS, low frequency repetitive transcranial magnetic stimulation; DSM IV, Diagnostic and Statistical Manual of Mental Disorders 4th Edition; NINCDS-ADRDA, National Institute of Neurological and Communicative Diseases and Stroke/Alzheimer's Disease and Related Related Disorders Association; MMSE, Mini-Mental State Examination; ADAS-cog, Alzheimer Disease Assessment Scale-cognitive subsection; ACE, Addenbrooke's Cognitive Examination; ACE-R, Addenbrooke's Cognitive Examination – Revised; WHO-UCLA, World Health Organization University of California-Los Angeles; AVLT, Auditory Verbal Learning Test; RCPM, Raven Colored Progressive Matrices; RAVLT, Rey Auditory Verbal Learning Test; ADCS-PACC, Alzheimer's Disease Cooperative Study Preclinical Alzheimer Cognitive Composite; WRT, Word Recognition Test; BDAE, Boston Diagnostic Aphasia Examination; SC-BADA, Battery for Analysis of Aphasic Deficits; IADL, Instrumental daily activity scale; CDR, Clinical Dementia Rating; DAD, Disability Assessment for Dementia; NPI, Neuropsychiatric Inventory; CGIC, Clinical Global Impression of Change; TMT, Trail Making Test; MOCA, Montreal Cognitive Assessment Test; CVSET, complex visual scene encoding task; CVLT-II, California Verbal Learning Test-II; FAB, Frontal Assessment Battery; RMBC, Revised Memory and Behavior Checklist; RAVL-T, Rey Auditory Verbal Learning Test; WMS, Wechsler Memory Scale; RCFT, Rey Complex Figure Test; BPSD, behavioral and psychological symptoms of dementia; TMT, Trail Making Test; WCST, Wisconsin Card Sorting Test; FNAT, Face-Name Association Task; RBMT, Rivermead Behavioral Memory Test; WMS, Wechsler Memory Scale; PANAS, Positive and Negative Affect Schedule; AES-C, Apathy Evaluation Scale-Clinician version; 3MS, Modified Mini Mental State Exam; EXIT-25, Executive Interview; MoCA, Montreal Cognitive Assessment; IADL, Instrumental Activities of Daily Living; ADL, Activities of Daily Living.