Review Article

A Systematic Review of Intervention Studies Examining Nutritional and Herbal Therapies for Mild Cognitive Impairment and Dementia Using Neuroimaging Methods: Study Characteristics and Intervention Efficacy

Table 3

Summary of study characteristics, intervention, neuroimaging, and neuropsychological methodologies, efficacy, and study design.

Author (reference)Aim, recruitmentStudy populationIntervention and durationNeuroimaging and neuropsychological measuresEfficacy on cognition, neuroimaging measures, any associations, adherence, retention, and adverse eventsStudy design

Bi et al. (2011) [13],
China
Aim: to test the efficacy of fuzhisan (FZS) in people with AD.
Recruitment location: hospital patients.
Mild to moderate AD (): randomised to FZS (7M : 5F, yrs, MMSE 19.58 ± 3.23, ADAS-Cog 33.58 ± 6.24, NPI 19.50 ± 6.24) or placebo (6M : 4F, 68.60 ± 6.35 yrs, MMSE 19.70 ± 3.30, ADAS-Cog 34.50 ± 5.80, NPI 19.80 ± 6.11).Intervention: 1 × 10 g FZS (or placebo) per day, orally.
Duration: 12 weeks.
Free from over-the-counter medications for at least 2 weeks prior to study entry; free from psychotropic drugs for at least 4 weeks prior.
PET (30 min)
NTB: MMSE, ADAS-cog, NPI.
Cognition: significant but small improvements in ADAS-Cog and NPI scores (versus no change in placebo).
Neuroimaging: increased or stabilised rCMRglc across frontal, parietal, and temporal cortices, posterior cingulate gyrus, hippocampus, thalamus, cerebellum (versus decreased rCMRglc in placebo).
Retention: total 3 withdrew: 1x FZS, 2x placebo (i.e., 88% completion rate).
Adverse effects: 2x mild, transient side-effects (1x nausea, 1x constipation).
Randomised, double-blind, placebo-controlled pilot study.

Cohen et al. (2003) [14],
USA
Aim: to investigate whether citicoline improves neurocognition & neuroimaging in people with VaD.
Recruitment location: specialty care centre (memory & cognitive disorders clinic).
VaD (): randomised to citicoline (7M : 8F, 78.1 ± 5.8 yrs, MMSE 19.0 ± 4.7) or placebo (5M : 10F, 78.0 ± 5.1 yrs, MMSE 21.3 ± 2.9).Intervention: 2 × 500 mg citicoline (or placebo) per day, orally.
Duration: 12 months.
Follow-up: 6 & 12 months.
MRI total brain volume
MRI SH volume
NTB: BNT, COWAT, CVLT, DRS, Grooved Pegboard Test, MMSE, RCFT, TMT, WAIS-R [block design, digit span & symbol, similarities, vocabulary], WMS-R [logical memory, visual reproduction].
Cognition: no difference between citicoline and placebo in neuropsychological performance (i.e., both groups significantly declined from baseline to both 6- and 12-month follow-up).
Neuroimaging: no difference between citicoline and placebo in structural brain measures (i.e., both groups had significantly decreased total brain volume & increased SH from baseline to 12-month follow-up).
Retention: total 9 withdrew prior to follow-up: 4x citicoline, 5x placebo.
Randomised, double-blind, placebo-controlled trial.

de Waal et al. (2014) [15],
Belgium, France, Germany, Italy, The Netherlands, Spain
Aim: to investigate the effect of Souvenaid on brain-activity based networks in people with AD.
Recruitment location: AD medical centres.
Mild AD (; drug-naïve): randomised to Souvenaid (45M : 41F, 74.1 ± 6.8 yrs, MMSE 25.1 ± 2.9) or placebo (47M : 46F, 72.5 ± 8.0 yrs, MMSE 25.4 ± 2.7).Intervention: 1 × 125 mL drink (active Souvenaid or isocaloric control) per day.
Duration: 24 weeks.
Follow-up: 12 & 24 weeks.
EEG functional connectivity networks
NTB: MMSE, COWAT, RAVLT, TMT, WMS-R [digit span, verbal paired associates].
Cognition: no relationship between EEG network and NTB memory performance. Association between beta activity and memory performance at midpoint in the treatment group only.
Neuroimaging: decreased beta network EEG in the placebo but not Souvenaid group.
Suggests improved synaptic integrity and function and counteraction of network decline.
Retention: total 12 withdrew prior to follow-up; 14 excluded due to protocol deviations &/or <80% compliance: 4x Souvenaid, 10x placebo.
Randomised, double-blind, placebo-controlled, multisite trial.

Douaud et al. (2013) [16], Jernerén et al. (2015) [17] and Smith et al. (2010) [18],
UK
Aim: (1) to investigate the effect of B-vitamin treatment on brain atrophy in people with MCI. (2) To investigate the effect of (a) plasma ω-3 fatty acid concentrations, (b) plasma tHCy concentrations, on B-vitamin treatment of brain atrophy in people with MCI.
Recruitment location: community.
MCI () [15]: randomised to B-vitamin (33M : 47F, 77 ± 5 yrs, MMSE 28.5 ± 1.5) or placebo (27M : 49F, 76 ± 4 yrs, MMSE 28.5 ± 1.5).
MCI () [19, 20]: randomised to B-vitamin (35M : 50F, 77.0 ± 5.2 yrs, MMSE 28.3 ± 1.8) or placebo (31M : 52F, 76.2 ± 4.5 yrs, MMSE 28.3 ± 1.5).
Intervention: high dose B-vitamin treatment (folic acid 0.8 mg/d, vit. B6 (pyridoxine HCl) 20 mg/d, vit. B12 (cyanocobalamin) 0.5 mg/d).
1x active or placebo tablet per day.
Duration: 24 months.
MRI regional grey matter volume [15]
MRI whole brain atrophy [19, 20]
NTB: CDR-SOB, MMSE, HVLT-R (delayed recall), category fluency (animals).
Neuroimaging: significantly reduced brain atrophy (0.5% versus 3.7%) in posterior brain regions (bilateral hippocampus, parahippocampal gyrus, retrosplenial precuneus, lingual and fusiform gyrus, cerebellum). Rate of brain atrophy was significantly slower (by 29.6%) with B-vitamin treatment than placebo.
This effect was even greater (53% lower atrophy) for individuals with high baseline tHcy, a risk factor for brain atrophy.
This effect was also even greater (40% lower atrophy) for individuals with high baseline ω-3 fatty acid concentrations, a risk factor for brain atrophy.
Adherence: >78% took at least 75% of the tablets; 81.4% (70/84 active, 66/83 placebo) determined biologically compliant via blood samples.
Retention: total 20 withdrew (11x B-vit, 9x placebo). Total 15 lost to death or cancer (7x B-vit, 8x placebo). Further 8 lost or excluded for “miscellaneous” reasons (5x B-vit, 3x placebo).
Adverse effects: no significant safety issues or group differences in adverse events.
Randomised, double-blind, placebo-controlled trial (VITACOG study).

Heo et al. (2016) [21],
Korea
Aim: To investigate the effect of Korean red ginseng (KRG) on brain activity in people with AD.
Recruitment location: medical centre.
AD (; 3M : 11F, 74.93 ± 7.63 yrs, K-MMSE 19.93 ± 4.80).Intervention: 4.5 g/d KRG, orally (total powder capsule, 6-yr-old root; KT&G Corporation, Daedeok District, Korea; 8.54% Ginsenosides).
Duration: 12 weeks.
qEEG (resting, eyes closed)
NTB: MMSE, FAB
Cognition: significant improvement in cognitive function (FAB).
Neuroimaging: relative alpha power increased in temporal regions for responders versus nonresponders.
Suggests increased frontal lobe function.
Open, case series study.

Hofferberth (1994) [19],
Germany
Aim: to investigate the effect of Ginkgo biloba special extract (GBE; EGb 761) treatment on brain activity in people with AD.AD (): randomised to GBE (14M : 7F, 63.6 yrs) or placebo (14M : 7F, 63.6 yrs).Intervention: 80 mg/d GBE (or placebo), orally.
Duration: 3 months.
Follow-up: 1, 2, and 3 months.
EEG (theta/alpha quotient)
NTB: memory, attention, choice reaction time.
Cognition: significant improvement in cognitive function following 1 month and 2 months and maintained following 3 months of GBE.
Significant improvement in choice reaction time following 1 month and maintained following 2 and 3 months of GBE.
Neuroimaging: significant reduction in theta wave component of theta/alpha quotient following 1 month and maintained following 2 and 3 months of GBE.
Randomised, double-blind, placebo-controlled trial.

Köbe et al. (2015) [22],
Germany
Aim: to investigate the combined effects of omega-3 fatty acids (FA), aerobic exercise, and cognitive stimulation on brain atrophy in people with MCI.
Recruitment location: memory clinics.
MCI (): randomised to target intervention (9M : 4F, 70.0 ± 7.2 yrs, MMSE 28.5 ± 1.1) or control intervention (5M : 4F, 70.0 ± 5.2 yrs, MMSE 27.9 ± 1.7).Intervention: both target and control groups received omega-3 FA (2.2 g/d; 4x oral capsules daily).
Target intervention: aerobic training (cycle ergometer; 2 × 45 min/wk) + cognitive stimulation (AKTIVA: Aktive Kognitive Stimulation-Vorbeugung im Alter (active cognitive stimulation-prevention in the elderly); 1x individual + 12x group sessions, 90 min duration, plus daily home practice, beginning week 4).
Control intervention: nonaerobic training (stretching & toning; 2 × 45 min/wk).
Duration: 6 months.
MRI brain volume
NTB: digit span, TMT, Stroop, AVLT, verbal fluency.
Cognition: no change in executive function, memory, sensorimotor speed, or attention.
Neuroimaging: increased or reduced atrophy in GM volume for target versus control intervention (middle and superior frontal cortices, frontal pole, angular cortex, posterior cingulate cortex).
Significant decrease in total homocysteine concentration.
Retention/adherence: original ; discontinued due to time constraints or poor compliance.
Adverse effects: None.
Randomised controlled trial

Matsuoka et al. (2012) [23],
Japan
Aim: to investigate the effect of toki-shakuyaku-san (TSS) on rCBF in people with MCI or AD.
Recruitment location: university clinic.
MCI/AD (; 3M : 5F, 77.8 ± 4.9 yrs, MMSE 23.4 ± 3.6).Intervention: 7.5 g TSS, orally (powder).
Duration: daily for 8 wks.
rCBF (SPECT)
NTB: MMSE, NPI, PSMS.
Cognition: no change in MMSE scores (trend toward improved orientation to place).
Neuroimaging: significant increase in rCBF in posterior cingulate.
Retention/adherence: original ; discontinued due to poor compliance, change in location, or withdrawal.
Adverse effects: none.
Open, case series study.

Muresanu et al. (2010) [24],
Romania
Aim: to investigate persistence of the effects of cerebrolysin on cognition & qEEG in people with VaD.VaD (): randomised (2 : 3 : 3) to cerebrolysin 10 mL (4M : 9F, 72.46 ± 2.80 yrs, MMSE 18.92 ± 1.32), cerebrolysin 30 mL (7M : 4F, 70.36 ± 3.62 yrs, MMSE 20.27 ± 1.92), or placebo (5M : 4F, 71.89 ± 3.52 yrs, MMSE 18.89 ± 1.81).Intervention: 50 mL i.v. infusions of cerebrolysin (10 mL + 40 mL saline or 30 mL + 20 mL saline) or placebo (saline) 5 days/wk.
Duration: 4 weeks.
Follow-up: 12 weeks (±1).
qEEG, eyes closed resting state
NTB: MMSE, ADAS-cog.
Cognition: significant improvement in cognitive performance maintained at follow-up.
Neuroimaging: significant (dose-dependent) reduction in qEEG power ratio maintained at follow-up.
Retention/adherence: original study ; lost to follow-up or due to poor compliance (i.e., receiving new drug treatment).
Open-label extension of a randomised, double-blind, placebo-controlled trial.

Nilsson et al. (2000) [25],
Sweden
Aim: to investigate the effect of cobalamin (vitamin B12) treatment on brain function in people with a medical history of cognitive deterioration.
Recruitment location: hospital outpatients.
Mild to severe dementia ( (VaD: ; AD: , mixed ); 15M : 14F, 78.9 ± 6.8 yrs; MMSE 9–23).Intervention: intramuscular injection of hydroxycobalamin (vit. B12); 1 mg every second day, total 10x.
Duration: 1 month.
rCBF (xenon 133 inhalation and cortexplorer with 254 scintillation detectors)
NTB: MMSE, OBS.
Cognition: classified as “clinically improved” (orientation to time and space, recent memory), though dementia severity did not change. did not show any clinical improvements.
Neuroimaging: “clinically improved” had significant increase in general blood flow level. slight trend towards a decrease in blood flow.
Open, case series study.

Oishi et al. (1998) [26],
Japan
Aim: to investigate the effect of traditional Chinese medicine treatment on brain function in people with AD.AD (; 65 ± 8 yrs; MMSE 16.0 ± 5.1).Intervention: traditional Chinese medicine (astragalus root 8 g, Prunella vulgaris 3 g, pueraria root 9 g, Lycii fructus 8 g, cnidium rhizome 5 g, rhubarb 1 g, alisma rhizome 6 g, peach kernel 6 g, ginseng 3 g, oyster shell 8 g).
Duration: 3 months.
ERP (auditory oddball P300)
rCBF (stable xenon CT method)
NTB: MMSE.
Cognition: significant improvement in MMSE scores (though still below normal range).
Neuroimaging: significantly improved (shortened) P300 latency. Increased white matter CBF.
Adverse effects: none.
Open, case series study.

Onofrj et al. (2002) [27],
Italy
Aim: to test the effects of donepezil (DPZ) versus vitamin E on brain function in people with varying severities of AD.Mild to severe AD ( completed): first divided into mild AD (group I) versus moderate-severe AD (group II) then randomised to treatment: group I DPZ (6M : 9F, 65.2 ± 1.8 yrs, MMSE 21.5 ± 0.4, ADAS-Cog 22.3 ± 1.0), Group I vit E (7M : 8F, 65.5 ± 1.7 yrs, MMSE 21.5 ± 0.6, ADAS-Cog 22.5 ± 0.9), Group II DPZ (6M : 9F, 66.7 ± 1.5 yrs, MMSE 11.5 ± 0.6, ADAS-Cog 44.5 ± 1.2), Group II vit E (8M : 7F, 66.5 ± 1.6 yrs, MMSE 11.6 ± 0.4, ADAS-Cog 43.5 ± 1.4).Titration: 14 days, 5 mg/day DPZ or 1000 IU/day vit E, orally.
Intervention: 10 mg/day DPZ or 2000 IU/day vit E, orally.
Duration: 6 months.
ERP (P300 auditory oddball)
NTB: MMSE, ADAS-cog, WAIS.
Cognition: DPZ: significant improvement in neuropsychological test performance, regardless of AD severity, though more pronounced for moderate-severe than mild AD.
Vitamin E: severe deterioration of neuropsychological test performance, regardless of AD severity.
Neuroimaging: DPZ: significantly reduced P300 latency, regardless of AD severity, though more pronounced for moderate-severe than mild AD.
Vitamin E: significantly increased P300 latency, regardless of AD severity.
Retention/adherence: total 7 withdrew during initial titration phase (6x DPZ (adverse effects), 1x Vit E (noncompliance)).
Adverse effects: 6x DPZ (3x nausea or abdominal discomfort, 3x confused agitation).
Pseudo-randomised, double-blind, controlled trial.

Saletu et al. (1995) [20],
Austria
Aim: to test the efficacy of nicergoline (NIC) in people with unspecified dementia (all-cause).
Recruitment location: nursing home for seniors.
Mild to moderate dementia (, MMSE 13–25; equal distribution SDAT : MID (), equally randomised () to placebo control (PLAC) or treatment (NIC)): SDAT/NIC: 5M : 23F, 78 ± 7 yrs; SDAT/PLAC: 7M : 21F, 77 ± 10 yrs; MID/NIC: 6M : 22F, 81 ± 7 yrs; MID/PLAC: 9M : 19F, 79 ± 7 yrs.Intervention: 2 × 30 mg NIC (or placebo) per day, orally.
Duration: 8 weeks NIC or placebo, 2-week washout period.
EEG mapping (3 min V-EEG)
ERP (P300 auditory oddball)
NTB: MMSE, SCAG, CGI.
Cognition: significant improvements in CGI, MMSE, and SCAG (versus pretreatment and placebo group).
Neuroimaging: decreased relative power delta/theta and alpha-1, increased relative power alpha-2 and beta (right temporal to frontotemporal and left parietal and temporo-occipital regions) (versus opposite effects in placebo).
Suggests improved vigilance.
Acceleration of total centroid power spectrum (versus pre- and placebo).
Shortened latency of P300 (versus pre- and placebo).
Suggests improved information processing.
Retention: total 14 withdrew: 4x SDAT/NIC, 4x SDAT/PLAC, 4x MID/NIC, 2x MID/PLAC.
Responder to nonresponder ratio: SDAT/NIC 16 : 8 (i.e., 66.6% responders), SDAT/PLAC 8 : 16, MID/NIC 17 : 7 (i.e., 70.83% responders), MID/PLAC 7 : 19.
Adverse effects: 9 of 48 NIC reported adverse effects (mild or marked insomnia (×2), mild itching, blocked nose, sweating, dry mouth, diarrhea, weight loss, constipation, moderate rigor).
Randomised, double-blind, placebo-controlled crossover trial.

Thomas et al. (2001) [28],
Italy
Aim: to test the effects of donepezil (DPZ) versus vitamin E versus rivastigmine (Riv) on brain function in people with AD.Mild to moderately severe AD (): randomised to double-blind treatment (DPZ (9M : 11F, 66.50 ± 9.19 yrs, MMSE 16.0 ± 0.5, ADAS-Cog 33.34 ± 2.70) or vit E (10M : 10F, 65.50 ± 10.61 yrs, MMSE 16.0 ± 0.5, ADAS-Cog 33.45 ± 2.60)), or open trial Riv (9M : 11F, 65.00 ± 8.49 yrs, MMSE 16.0 ± 0.5, ADAS-Cog 33.39 ± 2.70).
Age-matched control group (): 25M : 35F, 67.50 ± 14.85 yrs, MMSE 29.0 ± 0.4, ADAS-Cog 14.25 ± 0.50.
Titration: 1 month, 5 mg/d DPZ, 2000 IU/d vit E, or 1.5 mg/d Riv, orally.
Intervention: 10 mg/d DPZ or 2000 IU/d vit E, orally. Riv: 1x capsule 2x/d; month 2: total 3 mg/d, month 3: total 6 mg/d, month 4: total 9 mg/d, months 5 & 6: total 12 mg/d.
Duration: 6 months.
Follow-up: each month.
ERP (P300 auditory oddball)
NTB: MMSE, ADAS-Cog, WAIS subscales.
Cognition: DPZ and Riv: significant improvement in neuropsychological test performance.
Vitamin E: severe deterioration of neuropsychological test performance.
Neuroimaging: DPZ and Riv: significant reduction in P300 latency (no difference between).
Vitamin E: significantly increased P300 latency.
P300 latency changes were significantly correlated with neuropsychological test scores.
Retention/adherence: total 4 withdrew from Riv (3x nausea, 1x noncompliance). Total 2 excluded from vit E (no detectable P300).
Adverse effects: 3x nausea.
Randomised three-arm trial with one open-label arm and two double-blind arms.

Tsolaki et al. (in press) [29],
Greece
Aim: to test the effects of Crocus sativus L. (saffron) on brain function in people with aMCI.
Recruitment location: outpatient memory & dementia clinic.
aMCI (): Crocus (5M : 12F, 71.47 ± 6.73 yrs, MMSE 27.41 ± 1.70, MoCA 22.91 ± 3.00) or wait-list control (4M : 14F, 69.72 ± 7.33 yrs, MMSE 27.89 ± 1.84, MoCA 22.81 ± 2.64).Intervention: Crocus (no further information available).
Follow-up: 12 months.
MRI (global maxima of case “a”; subgroup)
ERP (P300 novelty auditory oddball using HD-EEG (256 channel); subgroup)
NTB: MMSE, MoCA, NPI, activities of daily living.
Cognition: significant improvement in MMSE (versus nonsignificant decline in the wait-list group).
Neuroimaging: significantly greater left temporal inferior gyrus volume.
Significantly reduced P300 latency.
Single-blind, nonrandomised, waitlist-controlled pilot trial.

Yamaguchi et al. (2004) [30],
Japan
Aim: to test the effects of choto-san on brain function in people with VaD/MCI.
Recruitment location: hospital outpatients.
VaD/MCI (): choto-san (8M : 2F, 71.3 ± 9.8 yrs, MMSE 23.8 ± 3.6) or control (9M : 1F, 68.0 ± 8.6 yrs, MMSE 26.8 ± 2.8).Intervention: 3x choto-san extract (TJ-47, Tsumura, 7.5 g/day) orally, daily [contains 4.5 g of extract of 11 kinds of dried medical herbs: Uncariae Uncis Cum Ramulus (3 g hooks and branch of Uncaria sinensis Oliver), Aurantii Nobilis pericarpium (3 g peel of Citrus unshiu Markovich), Pinelliae tuber (3 g tuber of Pinellia ternate Breitenbach), Ophiopgonis tuber (3 g root of Ophiopogon japonicus Ker-Gawler), Hoelen (3 g fungus of Poria cocos Wolf), Ginseng radix (2 g root of Panax ginseng C.A. Meyer), Chrysanthemi flow (2 g flower of Chrysanthemum morifolium Ramatulle), Saphoshnikoviae radix (2 g root and rhizome of Saposhnikovia divaricata Schischkin), Glycyrrhizae radix (1 g root of Glycyrrhiza uralensis Fisher), Gypsum Fibrosum (5 g CaSO4 2H2O) and Zingiberis rhizoma (1 g, rhizome of Zingiber officinale Roscoe)].
Duration: 12 weeks.
ERP (P300 novelty auditory oddball)
NTB: MMSE, verbal fluency test.
Cognition: significantly faster RT and increased accuracy on auditory oddball task. Significant improvement in MMSE and verbal fluency.
Neuroimaging: Significant reduction in P300 latency.
Trend towards increased novelty P300 amplitude.
Open, cohort study.

Zhang et al. (2015) [31],
China
Aim: to investigate the effect of Bushen capsule (BSC) on brain function in people with aMCI.
Recruitment location: hospital and community.
aMCI (): randomised to BSC (12M : 10F, 65.05 ± 6.67 yrs, MMSE 26.27 ± 1.58) or placebo (11M : 11F, 62.41 ± 5.76 yrs, MMSE 26.45 ± 1.40).Intervention: 4 × 300 mg BSC [main components Zexie (Alismatis rhizoma) and Roucongrong (Cistanches Herba)] or 4x placebo tablet, 3x/day.
Duration: 3 months.
fMRI (episodic memory encoding task).
NTB: MMSE, AVLT, CVLT, Stroop, digit symbol, clock drawing.
Cognition: significant improvement in MMSE, stroop, and AVLT.
Neuroimaging: increased brain activation in right putamen; this was significantly associated with stroop performance.
Reduced brain deactivation in right middle temporal gyrus; this was significantly associated with AVLT performance.
Adverse effects: 1x decreased appetite for 3 days (BSC), 1x mild nausea for 1 week (placebo). Neither discontinued use or withdrew from study.
Randomised, double-blind, placebo-controlled trial

Zhang et al. (2014) [32],
China
Aim: to investigate the effect of Congrongyizhi capsule (CCRC) on brain function in people with aMCI.
Recruitment location: hospital and community.
aMCI (): randomised to CCRC (8M : 8F, 64.25 ± 7.10 yrs, MMSE 26.38 ± 1.50), placebo (4M : 6F, 60.20 ± 3.52 yrs, MMSE 26.70 ± 1.64), or control (6M : 7F, 60.08 ± 6.53 yrs, MMSE 26.77 ± 1.30).Intervention: 4x CCRC [main components Cistanche and Polygonum multiflorum thunb] or 4x placebo tablet, 3x/day or nothing (control).
Duration: 3 months.
fMRI (n-back task).
NTB: MMSE, AVLT, CVLT, stroop, digit span, clock drawing.
Cognition: significant improvement in MMSE and digit span, which were significantly associated with increased brain deactivation in posterior cingulate cortex.
Neuroimaging: increased brain deactivation in posterior cingulate cortex, inferior frontal gyrus, and lingual gyrus.
Adverse effects: 1x decreased appetite for 3 days (CCRC), 1x mild nausea for 1 week (placebo). Neither discontinued use or withdrew from study.
Randomised, double-blind, placebo-controlled trial.

Zhang et al. (2012) [33],
China
Aim: to investigate the effect of Ginkgo biloba (EGb761) on brain activity in people with VaD.
Recruitment location: hospital.
VaD (: 46M : 34F, 66.5 ± 5.6 yrs) randomised to GBT () or aspirin ().Intervention: 19.2 mg GBT + 75 mg aspirin or 75 mg aspirin tablet, 3x/day.
Duration: 3 months.
rCBF (transcranial Doppler)
NBT: Montreal cognitive assessment (MoCA).
Cognition: significant improvement in global score MoCA, as well as MoCA score indices for executive function, attention, delayed memory, and orientation.
Neuroimaging: significant increase in blood flow velocity in middle and anterior cerebral arteries.
Randomised, controlled trial.

Note. ACC = anterior cingulate cortex; AD = Alzheimer’s disease; ADAS-Cog = Alzheimer’s Disease Assessment Scale–cognitive subscale; aMCI = amnestic mild cognitive impairment; AVLT = Auditory Verbal Learning Test; BNT = Boston Naming Test; CGI = clinical global impression; COWAT = Controlled Oral Word Association Test; CVLT = California Verbal Learning Test; d = day (i.e., /d = per day); DMN = default mode networks; DRS = Dementia Rating Scale; EEG = electroencephalography; ERP = event-related potential; FAB = frontal assessment battery; F : M = females to males; fMRI = functional magnetic resonance imaging; GM = grey matter; HVLT-R = Hopkins Verbal Learning Test-Revised; MCI = mild cognitive impairment; MFG = medial frontal gyrus; MID = multi-infarct dementia; MMSE = Mini-Mental State Examination; MoCA = Montreal cognitive assessment; MRI = magnetic resonance imaging; MTG = medial tegmental gyrus; NPI = Neuropsychiatric Inventory; NTB = neuropsychological test battery; OBS = Organic Brain Syndrome Scale; PCC = posterior cingulate cortex; PET = positron emission tomography; PSMS = Physical Self-Maintenance Scale; RAVLT = Rey Auditory Verbal Learning Test; rCBF = regional cerebral blood flow; RCFT = Rey Complex Figure Test; rCMRglc = regional cerebral metabolic rate of glucose consumption; ReHo = regional homogeneity; RT = reaction time; SCAG = Sandoz Clinical Assessment-Geriatric; SDAT = senile dementia of the Alzheimer type; SH = subcortical/periventricular hypertensity; SPECT = single-photon emission computed tomography; tHcy = total homocysteine; TMT = Trail Making Test; VaD = vascular dementia; V-EEG = vigilance-controlled EEG; yrs = years (i.e., age in years); WAIS-R = Wechsler Adult Intelligence Scale–Revised; WMS-R = Wechsler Memory Scale–Revised.