Efficacy and Safety of Cholinesterase Inhibitors for Mild Cognitive Impairment:A Systematic Review and Meta-Analysis

Abstract

Background:

The clinical benefit of cholinesterase inhibitors (ChEIs) for mild cognitive impairment (MCI) remains inconclusive.

Objective:

We performed a systematic review and meta-analysis of the efficacy/safety of ChEIs on subjects with MCI.

Methods:

We included randomized controlled trials (RCTs) of ChEIs in subjects with MCI, using cognitive function scores as a primary outcome measure.

Results:

Fourteen RCTs (six using donepezil, four using galantamine, and four using rivastigmine) with 5,278 subjects were included. We found no significant difference in cognitive function scores between the ChEIs and placebo groups [standardized mean difference (SMD) = –0.06, p = 0.38, I² = 76% ]. However, in the secondary outcomes, ChEIs were associated with a lower incidence of progression to dementia compared with placebo (risk ratio = 0.76, the number needed to treat = 20). For safety outcomes, ChEIs were associated with a lower prevalence of fall than placebo. On the other hand, compared with placebo, ChEIs were associated with a higher incidence of discontinuation due to all causes, discontinuation due to adverse events, at least one adverse event, abnormal dreams, diarrhea, dizziness, headache, insomnia, loose stools, muscle cramps, nausea, vomiting, and weight loss.

Conclusions:

Although ChEIs have a slight efficacy in the treatment of MCI, there are many safety issues. Therefore, ChEIs are difficult to recommend for MCI. However, the efficacy and safety of ChEIs on MCI with a biomarker-based diagnosis is unclear. Further RCTs are needed to confirm the efficacy and safety of ChEIs when used for individual neuropathological classifications of MCI.

Keywords

Cholinesterase inhibitors donepezil galantamine meta-analysis mild cognitive impairment rivastigmine

INTRODUCTION

Mild cognitive impairment (MCI) describes older subjects with a small demonstrable loss of cognitive ability who have not yet crossed the threshold for dementia [1]. Recent meta-analysis estimates of MCI incidence per 1,000 person-years were 22.5 for ages 75–79 years, 40.9 for ages 80–84 years, and 60.1 for ages 85+ years [2]. Individuals with MCI have an increased risk of developing Alzheimer’s disease (AD) or other dementia, which has long made the development of effective treatments for MCI a critical issue [1, 2].

Cholinesterase inhibitors (ChEIs), including donepezil, galantamine, and rivastigmine, are currently approved treatments for AD, and indeed, the efficacy of these medications on the progression of mild and moderate AD has been well established [3]. Furthermore, recent meta-analyses have suggested that ChEIs also alleviate cognitive impairment against Lewy body disease and vascular cognitive impairment [4, 5]. On the other hand, the efficacy of ChEIs for MCI has not been shown. Previous meta-analyses suggested that ChEIs had minimal impact on the progression to dementia and no impact on cognitive function scores for MCI, even while producing a higher frequency of adverse events [6, 7]. However, the results of previous meta-analyses regarding MCI may have been affected by small sample sizes. The rate of conversion from MCI to dementia is about 10% each year, and the pace of cognitive decline for individuals with MCI is slow [8], so larger sample sizes are needed to determine the overall net benefits of ChEIs. Therefore, we conducted an updated systematic review and meta-analysis to achieve more robust evidence regarding the consequences of using ChEIs to treat MCI.

MATERIALS AND METHODS

This systematic review and meta-analysis was performed according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines (PRISMA checklist, Supplementary Material) [9] and was registered with the International Prospective Register of Systematic Reviews (PROSPERO, available in the public domain at http://www.crd.york.ac.uk/PROSPERO/. CRD42019126818).

Search strategy and inclusion criteria

We included only RCTs on the ChEIs treatment of subjects with MCI. To identify the relevant studies, two authors (S.M. and H.T.) independently searched the MEDLINE, Cochrane Library, and Scopus databases, without language restrictions, from the date of these databases’ inception to April 16, 2019, using the following search strategy: (“cognitive dysfunction” [Mesh] OR “mild cognitive impairment” OR “MCI”) AND (“cholinesterase Inhibitors” [Mesh] OR “cholinesterase Inhibitors” OR “donepezil” [Mesh] OR “donepezil” OR “E2020” OR “Rivastigmine” [Mesh] OR “Rivastigmine” OR “ENA713” OR “Galantamine” [Mesh] OR “Galantamine”) AND (“randomized” OR “randomly” OR “random”). They also searched ClinicalTrials.gov (http://clinicaltrials.gov/), the ISRCTN registry (https://www.isrctn.com/), and the International Clinical Trials Registry Platform (http://www.who.int/ictrp/en/) websites to ensure comprehensive inclusion of RCTs and to minimize the possibility of publication bias. They independently assessed the inclusion/exclusion criteria and selected the relevant studies. References to included articles and reviews were also searched for citations of additional relevant published and unpublished studies, including conference abstracts.

Data synthesis and outcome measures

The primary outcome was the change in cognitive function scores from baseline to the last available follow-up. Cognitive function scores were derived from the Alzheimer’s Disease Assessment Scale-cognitive subscale (ADAS-cog) [10], the Mini-Mental State Examination (MMSE) [11], and the Montreal Cognitive Assessment (MoCA) [12]. The following secondary outcomes were also assessed: progression to dementia, Clinical Dementia Rating Sum of Boxes (CDR-SB) [13], global scores [Clinical Dementia Rating (CDR) [13], clinical global impression change (CGI-C) and CGI-C MCI version [14, 15], activities of daily living [the Alzheimer’s Disease Cooperative Study-Activities of Daily Living (ADCS-ADL) and ADCS-ADL MCI version [16], the Neuropsychiatric Inventory 12 (NPI12) [17], the Symbol Digit Modalities Test (SDMT) [18], the Digit Span Backwards Test (DSBT) [19], and the Global Deterioration Scale (GDS) [20]. Also included were data on the discontinuation due to all causes, discontinuation due to adverse events, and the incidence of individual adverse events. The methodologic quality of each trial we selected for inclusion was assessed according to the risk-of-bias criteria in the Cochrane Handbook for Systematic Reviews of Interventions (version 5.2.0; Cochrane Collaboration, available in the public domain at http://training.cochrane.org/handbook).

Data extraction

Two authors (S.M. and H.T.) independently extracted data from the included studies. Where possible, an intention-to-treat or a full analysis set population was used. When such data were unavailable, the per-protocol analysis results were extracted from each study. When data required for the meta-analysis were missing, investigators or the pharma sponsors of the relevant research were contacted and asked to provide unpublished data.

Meta-analysis methods

The meta-analysis was conducted using the Review Manager software (version 5.3 for Windows; Cochrane Collaboration, available in the public domain at http://tech.cochrane.org/revman). A random effects model was selected for this meta-analysis because of its potential heterogeneity across studies. Dichotomous outcomes were presented as risk ratios (RRs) with 95% confidence intervals (CIs). When the random effects model showed significant differences between groups, the number needed to treat (or harm) (NNT or NNH) was calculated. The NNT (or NNH) values were derived from the risk difference (RD) using the following formula: NNT (or NNH) = 1/RD. For continuous data, we used the mean differences (MD) and 95% CIs when data that were measured using the same scale were being combined. To combine the data measured using different scales, we used the standardized mean difference (SMD), combining effect size (Hedges’ g) data, and 95% CIs. Lower MMSE, MoCA, ADCS-ADL, SDMT, and DSBT scores indicated more impairment or more severe symptoms, and therefore, the algebraic sign of the numerical scores was reversed for these scales. Study heterogeneity was assessed using the I² statistic, with I²≥50% reflecting considerable heterogeneity [21]. In cases with I²≥50% for the primary outcome measure, we conducted sensitivity analyses to determine the reasons for heterogeneity. We examined the following confounding factors: neuropathological classification, sponsorship, study duration (102 weeks≤or 102 weeks>), and type of drug. A meta-regression analysis was performed to evaluate the association between the results of the meta-analysis on cognitive function scores and specific modulators (patient age and study duration) using the Comprehensive Meta-Analysis software, version 3 (Biostat, Inc., Englewood, NJ, USA). We also performed subgroup analyses to examine the efficacy and safety of individual ChEIs. Funnel plots were examined visually to assess publication bias in primary outcomes when≥10 studies were included [21]. In addition, Egger’s regression test was used to detect publication bias in meta-analyses.

RESULTS

Study characteristics

Of the 576 articles obtained from our literature search, 184 were excluded because they were duplicates, 365 were excluded after a review of the abstract or title, and 15 were excluded after a review of the full text (one non-RCT, ten review articles, and four same studies that were included in our meta-analysis). In the hand search, we added one RCT [22] from the review article [6]. No further studies were added from the clinical trials registration. In the end, thirteen studies, which included 14 RCTs that tested ChEIs for MCI, were accepted for the current meta-analysis [15 , 22–33] (Supplementary Figure 1).

The characteristics of the included trials are shown in Table 1. Six studies assessed the effects of donepezil on MCI [15 , 23–27]), four studies assessed the effect of galantamine on MCI [28 –30], and four studies assessed the effect of rivastigmine on MCI [22 , 31–33]. All studies were double-blind RCTs. The mean duration of the included studies was 67.9 weeks, and the mean patient age was 70.3 years. Two studies were supported by funding from Eisai and Pfizer [15, 24], one study was supported by funding from Eisai, Teaneck NJ, and Pfizer [27], one study was supported by funding from Eisai [25], one study was supported by funding from Pfizer [26], one study was supported by funding from Janssen [28], two studies were supported by funding from Janssen and Johnson & Johnson [30], and four studies were supported by funding from Novartis [22 , 31–33].

Table 1

Characteristics of included randomized controlled trials

Study, country, sponsorship	Total	Methods	Patients	Intervention, dose (mg/day)	n	Age	Male	Baseline MMSE	Main outcomes
	n	(1) Study design	(1) Diagnosis			(mean±SD), y	(%)	(mean±SD)
		(2) Duration	(2) Inclusion criteria
		(3) Analyzed population
Devanand 2018 [23], USA, non-industry	61	(1) DBRCT	(1) MCI, ADNI criteria	DON, 5–10 mg	30	67.1±7.7	46.7	27.5±2.2	ADAS-cog, BDI-II, BNT-15, FAQ, HAM-D
		(2) 62 weeks	(2) age 55–95, CDR = 0.5, MMSE≥21, HAM-D≥14, major depression or dysthymic disorder (DSM-IV)	PLA	31	72.4±8.9	51.6	28.1±1.7	LAN, SCWT, SRT, TMT-A, TMT-B, WAIS-III (BD), WAIS-III (DS), WMS-III (VR), AEs
		(3) ITT
Doody 2009 [24], USA, industry	821	(1) DBRCT	(1) MCI, Petersen criteria	DON, 10 mg	409	70.2±9.7	51.7	27.5±1.9	ADAS-cog, CDR-SB, CGIC-MCI, DSB, MMSE, NPI, PDQ, PDQ-R, PGA, SDMT, AEs
		(2) 48 weeks	(2) age 45–90, CDR = 0.5, MMSE 24–28	PLA	412	69.8±10.3	57.4	27.4±1.9
		(3) ITT
Dubois 2015 [25], France, industry	216	(1) DBRCT	(1) MCI, NR	DON, 10 mg	113	74.1±6.4	47.8	26.1±2.2	ADAS-cog-MCI, BT, CVLT IVLFT, MMSE, MRI, TMT-A, TMT-B, AEs
		(2) 52 weeks	(2) age≥50, CDR = 0.5	PLA	103	73.7±6.6	47.6	25.8±2.6
		(3) ITT
Montero-Odasso 2018 [26], Canada, industry	60	(1) DBRCT	(1) MCI, Petersen criteria	DON, 10 mg	31	73.5±5.7	52	27.4±2.2	BNT, DSB, DSF, DTC, GS, GV, LNS, MMSE, MoCA, RAVLT, TMT-A, TMT-B, AEs
		(2) 26 weeks	(2) age≥65, CDR = 0.5	PLA	29	77.2±8.1	59	27.5±1.7
		(3) ITT
Petersen 2005 [15], USA and Canada, industry	512	(1) DBRCT	(1) MCI, Petersen criteria	DON, 10 mg	253	73.1±7.1	56	27.3±1.8	ADAS-cog, ADCS- ADL-MCI, BNT, CDR, CDR-SB, CDT, CFT, DSB, GDS, MMSE, MTT, NC, NYUPR, SDMT, AEs
		(2) 156 weeks	(2) age 55–90, CDR = 0.5, MMSE 24–30	PLA	259	72.9±7.6	53	27.4±1.8
		(3) ITT
Salloway 2004 [27], USA, industry	270	(1) DBRCT	(1) MCI, NR	DON, 10 mg	133	72.1±8.0	57.6	27.5±2.2	ADAS-cog, BNT, CGIC-MCI, DSB, MT, NC, NYUPDR, NYUPIR, PGA, SDMT, VF, AEs
		(2) 24 weeks	PLA	(2) age 55–90, CDR = 0.5, MMSE≥24	137	72.7±8.0	57.7	27.4±2.0
		(3) ITT
Koontz 2005 [28], USA, industry	19	(1) DBRCT	(1) MCI, Petersen criteria	GAL, 24 mg	8	71±9.3	100	NR	CANTAB, CVLT, FAQ, AEs
		(2) 16 weeks	(2) MMSE≥26	PLA	11
		(3) NR
Peters 2012 [29], Germany, non-industry	154	(1) DBRCT	(1) MCI, NR	GAL, 16 mg	75	67.9±8.3	NR	27.4±2.2	ADAS-cog, CDT, CERAD, TMT-A, TMT-B, AEs
		(2) 104 weeks	(2) CDR = 0.5	PLA	79	67.2±7.6	26.9±2.1
		(3) ITT
Winblad 2008 study 1 [30], USA and Canada, industry	995	(1) DBRCT	(1) MCI, NR	GAL, 16–24 mg	497	69.2±9.1	48	NR	ADAS-cog-MCI, ADCS- ADL-MCI, CDR, CDR-SB, DSST, AEs
		(2) 104 weeks	(2) age≥50, CDR = 0.5	PLA	498	70.1±9.1	45	NR
		(3) ITT
Winblad 2008 study 2 [30], USA and Canada, industry	1062	(1) DBRCT	(1) MCI, NR	GAL, 16–24 mg	532	70.6±8.7	45	NR	ADAS-cog-MCI, ADCS- ADL-MCI, CDR, CDR-SB, DSST, AEs
		(2) 104 weeks	(2) age≥50, CDR = 0.5	PLA	530	70.9±8.7	41	NR
		(3) ITT
Bokde 2016 [31], Germany, industry	12	(1) DBRCT	(1) MCI, Petersen criteria	RIV, 9 mg	8	68±8.5	60	25.2±3.6	CERAD, MRI
		(2) 12 weeks	(2) NR	PLA	4	50±2.6	100
		(3) NR
Feldman 2007 [32], multiple countries, industry	1018	(1) DBRCT	(1) MCI, NR	RIV, 3–12 mg	508	70.3±7.4	46.9	27.0±2.6	ADAS-cog, ADCS-ADL, BDI, BFCSR, CDR, CDT, DCT, DWLR, GDS, LNS, MMSE, MRI, MT, NPI, NYUPR, QOL, SDMT, VF, AEs
		(2) 208 weeks	(2) age 55–85, CDR = 0.5	PLA	510	70.6±7.6	48.6	26.9±2.8
		(3) ITT
Mamikonyan 2015 [33], USA, industry	28	DBRCT (cross over)	(1) MCI, Winblad criteria	RIV-P, 4.6–9.5 mg	14	62.6±10.1	71.4	NR	CGIC, ECB, DRS-2, GDSy, GDS-15, MoCA, NCTS, PDAQ, PDQ-8, PPS, STAI (anxiety subscale), UPDRS-III, AEs
		(2) 10 weeks	(2) age 40–85, CDR = 0.5, PD for more than 2 years	PLA	14	66.1±5.5	85.7
		(3) ITT
Narasimhalu 2010 [22], Singapore, industry	50	(1) DBRCT	(1) MCI, NR	RIV, 3-9mg	25	68.1±6.8	20	23.7±3.4	ADAS-cog, ADCS- ADL-MCI, CB, CDT, CTT, FAB, GD, MMSE, NPI, AEs
		(2) 24 weeks	(2) age 55–85	PLA	25	69.4±9.1	48	23.9±3.2
		(3) ITT

ADAS-cog, Alzheimer’s Disease Assessment Scale-cognitive subscale; ADAS-cog-MCI, Alzheimer’s Disease Assessment Scale-cognitive subscale, Mild Cognitive Impairment version; ADCS- ADL-MCI, Alzheimer’s Disease Cooperative Study Activities of Daily Living scale Mild Cognitive Impairment version; ADNI, Alzheimer’s Disease Neuroimaging Initiative; AEs, adverse events; BDI, Beck depression inventory; BFCSR, Buschke free and cued selective reminding test; BNT(-15), Boston Naming Test (15-item); BT, Benton Test; CANTAB, Cambridge Automated Neuropsychiatric Test Assessment Battery; CB, Cognitive Battery; CDR, Clinical Dementia Rating; CDR-SB, Clinical Dementia Rating Sum of Boxes; CDT, clock-drawing test; CERAD, Consortium to Establish a Registry for Alzheimer’s Disease test; CFT, category-fluency test; CGIC(-MCI), Clinical Global Impression of Change(-Mild Cognitive Impairment); CVLT, California Verbal Learning Test; DBRCT, double-blind randomized controlled trial; DCT, digit cancellation task; DON, donepezil; DRS-2, Dementia Rating Scale-2; DSB, Digit Span Backwards test; DSF, Digit Span Forward test; DSM-IV, Diagnostic and Statistical Manual of Mental Disorders Fourth Edition; DSST, Digit Symbol Substitution Test; DTC, dual-task gait cost; DWLR, delayed word list recall; ECB, Everyday Cognition Battery; FAB, Frontal Assessment Battery; FAQ, Pfeffer Functional Activities Questionnaire; GAL, galantamine; GD, Geriatric Depression Scale; GDS, Global Deterioration Scale; GDSy, Gordon Diagnostics System; GDS-15, 15-item Geriatric Depression Scale; GS, gait speed; CTT, Color Trails Test; GV, gait variability; HAM-D, Hamilton Rating Scale for Depression; ITT, intention to treat; IVLFT, Isaacs verbal fluency and lexical fluency test; LAN, letter and animal naming; LNS, letter–number sequence; MCI, mild cognitive impairment; MMSE, Mini-Mental State Examination; MoCA, Montreal Cognitive Assessment; MRI, magnetic resonance imaging; MT, Maze test; MTT, maze-tracing task; n, number of patients; NC, number-cancellation test; NCTS, Neurotrax Comprehensive Testing Suite; NPI, Neuropsychiatric Inventory; NR, not reported; NYUPDR, New York University paragraph test delayed recall; NYUPIR, New York University paragraph test immediate recall; NYUPR, New York University paragraph-recall test; PD, Parkinson’s disease; PDAQ, Penn Daily Activities Questionnaire; PDQ, Perceived Deficits Questionnaire; PDQ-R, Perceived Deficits Questionnaire for Relatives; PDQ-8, Parkinson’s Disease Questionnaire-8; PGA, Patient Global Assessment; PLA, placebo; PPS, Parkinson Psychosis Scale; QOL, quality of life; RAVLT, Rey Auditory Verbal Learning Test; RIV, rivastigmine; RIV-P, rivastigmine patch; SCWT, Stroop Color and Word Test; SD, standard deviation; SDMT, Symbol Digit Modalities Test; SRT, Selective Reminding Test; STAI, State Trait Anxiety Inventory; TMT, Trail Making Test; UPDRS, Unified Parkinson’s Disease Rating Scale; VF, Verbal Fluency test; WAIS-III (BD), Wechsler Adult Intelligence Scale-Third Edition Block Design Subtest; WAIS-III (DS), Wechsler Adult Intelligence Scale-Third Edition Digit Symbol Subtest; WMS-R, Wechsler Memory Scale Revised; WMS-III (VR), Wechsler Memory Scale-Third Edition Visual Reproduction Subtest.

Fig.1

Forest plot of cognitive function scores.

Table 2

Sensitivity analysis (cholinesterase inhibitors for cognitive function scores)

Subgroup		N	n	I²	SMD [95% CIs]	p	Test for subgroup differences
Neuropathological classification	Parkinson disease	1	27	na	–0.66 [–1.44, 0.12]	0.10	p = 0.13, I² = 56.3%
	No classification	9	4605	77	–0.05 [–0.19, 0.09]	0.51
Sponsorship	Industry	9	4571	78%	–0.06 [–0.21, 0.09]	0.42	p = 0.89, I² = 0%
	Non-industry	1	61	na	–0.10 [–0.60, 0.40]	0.70
Study duration	Long study duration (102 weeks≤)	4	3431	0	–0.03 [–0.09, 0.04]	0.41	p = 0.90, I² = 0%
	Short study duration (102 weeks>)	6	1207	85	–0.05 [–0.45, 0.34]	0.79
Type of drug	Donepezil	5	1637	21%	–0.16 [–0.28, –0.04]	0.009	p = 0.20, I² = 38.4%
	Galantamine	2	1901	9%	–0.03 [–0.12, 0.06]	0.53
	Rivastigmine	3	1094	92%	0.31 [–0.74, 1.36]	0.57

CIs, confidence intervals: N, number of studies: n, number of patients: na, not applicable: SMD, standardized mean difference.

The methodologic quality of the included studies was evaluated according to the Cochrane risk-of-bias criteria (Supplementary Figure 2). Six studies did not mention the randomization method [23 , 31]. Six studies did not mention the method of allocation concealment [23 , 31]. Six studies did not mention the method of blinding [15 , 31]. Two studies did not mention how to address incomplete outcome data [28, 31]. One study did not report some of the preplanned outcomes [23].

Results of the meta-analysis

Efficacy outcomes

No significant difference between the ChEIs and placebo groups were found regarding cognitive function scores (SMD = –0.06, 95% CI = –0.20–0.08, p = 0.38, I² = 76%; N = 10, n = 4,632; Fig. 1). Regarding the individual ChEIs, donepezil significantly improved cognitive function scores compared with placebo (SMD, –0.16; 95% CI, –0.28 to –0.04; p = 0.009, I² = 21%; N = 5, n = 1,637). In contrast, galantamine and rivastigmine produced no significant differences in the cognitive function scores compared with placebo. The data for cognitive function scores in each treatment group were simulated with no publication bias (Funnel plot: Supplementary Figure 3, Egger’s test p-value = 0.99103).

We detected considerable heterogeneity in the cognitive function scores (I² = 76%) and therefore performed four sensitivity analyses to identify the confounding factors that might be affecting these scores (Table 2). When divided according to the type of drug (donepezil, galantamine, or rivastigmine subgroups), the significant heterogeneity disappeared in the donepezil and galantamine subgroups; moreover, there was a significant effect of the donepezil subgroup (donepezil subgroup: SMD, –0.16; 95% CI, –0.28 to –0.04; p = 0.009, I² = 21%; N = 5, n = 1,637; galantamine subgroup: SMD, –0.03.; 95% CI, –0.12–0.06; p = 0.53, I² = 9%; N = 2, n = 1,901; rivastigmine subgroup: SMD, 0.31.; 95% CI, –0.74–1.36; p = 0.57, I² = 92%; N = 3, n = 1,094; test for subgroup differences, I² = 38.4%, p = 0.2; Table 2). We also performed two meta-regression analyses to identify the confounding factors affecting these scores (Table 3; Supplementary Figure 4). However, we did not find any correlation between the cognitive function scores and the confounding factors.

Table 3

Meta-regression analysis (cholinesterase inhibitors for cognitive function scores)

Covariate	Coefficient	Standard error	95% lower	95% upper	p
Patient age	–0.0465	0.0396	–0.1242	0.0312	0.2407
Study duration	0.0005	0.0012	–0.0019	0.0029	0.6916

Fig.2

Forest plot of progression to dementia.

In the secondary outcomes, ChEIs were associated with a lower incidence of progression to dementia compared with placebo (RR = 0.76, NNT = 20, mean duration of included studies = 126.8 weeks; Fig. 2). For individual ChEIs, galantamine was associated with a lower incidence of progression to dementia compared with placebo (RR = 0.68, NNT = 17). In contrast, donepezil and rivastigmine demonstrated no significant differences in this regard. There were no significant differences in global scores between the ChEIs and placebo groups (Supplementary Figure 5). However, the rivastigmine subgroup significantly improved global scores compared with placebo (SMD, –0.25; 95% CI, –0.37 to –0.13; p = 0.0001, I² = 0%; N = 2, n = 1,045).

No significant differences in other secondary outcomes were found between the ChEIs and the placebo groups (Supplementary Figure 5).

Safety outcomes

There was a significantly higher rate of discontinuation due to all causes in the ChEIs compared with the placebo group (RR = 1.25, NNH = 11; Supplementary Figure 5). Regarding the individual ChEIs, donepezil, galantamine, and rivastigmine were associated with a higher rate of discontinuation due to all causes compared with placebo (donepezil, RR = 1.50, NNH = 8; galantamine, RR = 1.14, NNH = 14; rivastigmine, RR = 1.14, NNH = 14). Moreover, there was also a significantly higher rate of discontinuation due to adverse events in the ChEIs compared with the placebo group (RR = 2.14, NNH = 11; Supplementary Figure 5). For the individual ChEIs, donepezil, galantamine, and rivastigmine were associated with a higher rate of discontinuation due to adverse events compared with placebo (donepezil, RR = 2.48, NNH = 11; galantamine, RR = 2.21, NNH = 8; rivastigmine, RR = 1.28, NNH = 20).

There was also a significantly higher rate of at least one adverse event in the ChEIs group compared with the placebo group (RR = 1.10, NNH = 13; Supplementary Figure 5). When these drugs were considered individually, there was a significantly higher rate of at least one adverse event for donepezil and galantamine compared with placebo (donepezil, RR = 1.19, NNH = 8; galantamine, RR = 1.04, NNH = 33). The incidence of serious adverse events was similar between the ChEIs and placebo groups (Supplementary Figure 5). For individual ChEIs, the rates of serious adverse events were also similar between groups.

With respect to individual adverse events, ChEIs overall were associated with a lower incidence of falls than placebo (RR = 0.69, NNH = 33; Supplementary Figure 5). For individual ChEIs, we found a significantly lower incidence of falls for galantamine compared with placebo (RR = 0.71, NNH = 50).

In addition, the ChEIs overall had a higher incidence of abnormal dreams (RR = 3.98, NNH = 13), diarrhea (RR = 2.59, NNH = 10), dizziness (RR = 1.61, NNH = 17), headache (RR = 1.36, NNH = 33), insomnia (RR = 1.81, NNH = 20), loose stools (RR = 3.16, NNH = 25), muscle cramps (RR = 7.78, NNH = 8), nausea (RR = 2.97, NNH = 8), vomiting (RR = 4.35, NNH = not significant), and weight loss (RR = 2.01, NNH = 33), compared with placebo (Supplementary Figure 5). For individual ChEIs, compared with placebo, we found that donepezil was associated with a higher incidence of abnormal dreams (RR = 3.98, NNH = 13), diarrhea (RR = 3.07, NNH = 8), headache (RR = 2.33, NNH = 25), insomnia (RR = 2.58, NNH = 17), loose stools (RR = 3.16, NNH = 25), muscle cramps (RR = 7.78, NNH = 8), nausea (RR = 2.59, NNH = 17), and vomiting (RR = 3.89, NNH = 17. Galantamine was associated with a higher incidence of diarrhea (RR = 1.75, NNH = 14), dizziness (RR = 1.59, NNH = 20), insomnia (RR = 1.52, NNH = 25), nausea (RR = 2.76, NNH = 5), and weight loss (RR = 1.99, NNH = 33) than placebo. Rivastigmine was associated with a higher incidence of dizziness (RR = 1.62, NNH = 11), headache (RR = 1.39, NNH = not significant), insomnia (RR = 1.45, NNH = 25), nausea (RR = 3.42, NNH = not significant), and vomiting (RR = 4.48, NNH = 5) than placebo. No significant differences were found in the incidences of death between the ChEIs and the placebo groups (Supplementary Figure 5).

DISCUSSION

Our meta-analysis results showed that ChEIs did not improve cognitive function scores for subjects with MCI. However, this overall finding masked significant underlying heterogeneity. Therefore, we performed the sensitivity analyses to identify the confounding factors affecting these scores. According to the sensitivity analysis, one of the reasons for heterogeneity may be the type of ChEIs that were investigated. Our sensitivity analyses detected that the donepezil subgroup had less deterioration in cognitive function scores than the placebo groups without considerable heterogeneity. However, the clinical benefit may be limited because we found that the effect size of donepezil for cognitive function scores on subjects with MCI was very small (SMD = –0.16).

From the secondary outcomes, our meta-analysis results suggested that ChEIs reduced the progression from MCI to dementia. However, looking at the ChEIs individually, we found that only galantamine achieved this effect. But although donepezil and rivastigmine had no statistically significant effect on the progression of MCI to dementia compared with placebo, these ChEIs did produce a lower rate of transition from MCI to dementia than placebo. From these results, ChEIs may reduce the progression from MCI to dementia. However, the clinical benefit may be limited because the effect size of ChEIs for the progression from MCI to dementia is very small (RR = 0.76, NNT = 20).

Regarding other secondary outcomes, there were no significant differences in either the ChEIs or the placebo groups. For the subgroup analysis of secondary outcomes, only rivastigmine significantly improved global scores compared with placebo; however, again, this effect size was small (SMD = –0.25).

Although our results indicated that ChEIs had slight effects for treating MCI, there were also significantly higher rates of discontinuation due to all causes, discontinuation due to adverse events, and at least one adverse event in the ChEIs group compared with placebo. Moreover, treatment with ChEIs was associated with a higher incidence of gastrointestinal adverse events, abnormal dreams, dizziness, headache, insomnia, muscle cramps, and weight loss compared with placebo. The results of these safety outcomes were similar to the effects of previous meta-analyses for MCI or dementia [6 , 35]. On the other hand, our meta-analysis showed that ChEIs had a significantly lower incidence of falls compared with placebo. A recent meta-analysis suggests that ChEIs may improve gait performance in AD [36], and this implies they might also have similar effects on MCI. However, our results regarding the incidence of falls may include the possibility of statistical errors because we have included only three studies addressing this area in our meta-analysis.

Several study limitations should be addressed. First, we could not investigate the relationship between neuropathological features of MCI and the efficacy of ChEIs. Although we performed sensitivity analyses for the pathological classification of MCI, there was only one RCT that examined for neuropathological features in the studies included in our meta-analysis (Table 2). Since MCI has neuropathological heterogeneity, recent guidelines regarding MCI recommend the inclusion of patient cohorts with specific biomarker data in treatment studies targeted at specific pathologies (e.g., MCI due to AD) [37, 38]. Therefore, it is necessary to conduct RCTs in the future for neuropathologically classified MCI. Second, there is the paucity of studies included. For individual ChEIs, there were only six studies for donepezil, four for galantamine, and four addressing rivastigmine. Because the individual ChEIs have different pharmacological mechanisms of action [39], further studies regarding the effectiveness of individual ChEIs for MCI are needed. Third, the RCTs included in our meta-analysis had a short study duration (mean duration = 67.9 weeks). Our subgroup analysis and meta-regression analysis did not show a correlation between study duration and cognitive function scores. However, the pace of cognitive decline for individuals with MCI is quite slow, and studies with longer durations may be needed to determine the overall benefits of ChEIs. Finally, twelve of the included studies were industry sponsored, which tend to show more favorable efficacy results than those sponsored by other sources [40]; therefore, a sponsorship bias may exist in our results.

In conclusions, we found that although ChEIs have a slight efficacy in the treatment of MCI, there are many significant safety issues, and this makes them difficult to recommend for MCI. However, the efficacy and safety of ChEIs on MCI with a biomarker-based diagnosis remain unclear. Therefore, further RCTs are needed to confirm the efficacy and safety of ChEIs when used for individual neuropathological classifications of MCI.

DISCLOSURE STATEMENT

Authors’ disclosures available online (https://www.j-alz.com/manuscript-disclosures/19-0546r1).

Footnotes

Appendix

References

Lopez

(2013) Mild cognitive impairment. Continuum (Minneap Minn) 19, 411–424.

Gillis

, Mirzaei

, Potashman

, Ikram

, Maserejian

(2019) The incidence of mild cognitive impairment: A systematic review and data synthesis. Alzheimers Dement (Amst) 11, 248–256.

Tsoi

, Chan

, Hirai

, Kwok

, Wong

(2019) Monotherapy is good enough for patients with mild-to-moderate Alzheimer’s disease: A network meta-analysis of 76 randomized controlled trials. Clin Pharmacol Ther 105, 121–130.

Jin

, Liu

(2019) Comparative efficacy and safety of cognitive enhancers for treating vascular cognitive impairment: Systematic review and Bayesian network meta-analysis. Neural Regen Res 14, 805–816.

Matsunaga

, Kishi

, Yasue

, Iwata

(2015) Cholinesterase inhibitors for Lewy body disorders: A meta-analysis. Int J Neuropsychopharmacol 19, pyv086.

Russ

, Morling JR (2012) Cholinesterase inhibitors for mild cognitive impairment. Cochrane Database Syst Rev CD009132.

Tricco

, Soobiah

, Berliner

, Ho

, Ng

, Ashoor

, Chen

, Hemmelgarn

, Straus

(2013) Efficacy and safety of cognitive enhancers for patients with mild cognitive impairment: A systematic review and meta-analysis. CMAJ 185, 1393–1401.

Behrman

, Valkanova

, Allan

(2017) Diagnosing and managing mild cognitive impairment. Practitioner 261, 17–20.

Moher

, Liberati

, Tetzlaff

, Altman

, Group

(2009) Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement. BMJ 339, b2535.

10.

Rosen

, Mohs

, Davis

(1984) A new rating scale for Alzheimer’s disease. Am J Psychiatry 141, 1356–1364.

11.

Folstein

, Folstein

, McHugh

(1975) “Mini-mental state”. A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 12, 189–198.

12.

Nasreddine

, Phillips

, Bedirian

, Charbonneau

, Whitehead

, Collin

, Cummings

, Chertkow

(2005) The Montreal Cognitive Assessment, MoCA: A brief screening tool for mild cognitive impairment. J Am Geriatr Soc 53, 695–699.

13.

Morris

(1993) The Clinical Dementia Rating (CDR): Current version and scoring rules. Neurology 43, 2412–2414.

14.

Schneider

, Olin

, Doody

, Clark

, Morris

, Reisberg

, Schmitt

, Grundman

, Thomas

, Ferris

(1997) Validity and reliability of the Alzheimer’s Disease Cooperative Study-Clinical Global Impression of Change. The Alzheimer’s Disease Cooperative Study.S. Alzheimer Dis Assoc Disord 11(Suppl 2), 22–32.

15.

Petersen

, Thomas

, Grundman

, Bennett

, Doody

, Ferris

, Galasko

, Jin

, Kaye

, Levey

, Pfeiffer

, Sano

, van

Dyck CH

, Thal

, Alzheimer’s Disease Cooperative Study G (2005) Vitamin E and donepezil for the treatment of mild cognitive impairment. N Engl J Med 352, 2379–2388.

16.

Galasko

, Bennett

, Sano

, Ernesto

, Thomas

, Grundman

, Ferris

(1997) An inventory to assess activities of daily living for clinical trials in Alzheimer’s disease. The Alzheimer’s Disease Cooperative Study.. Alzheimer Dis Assoc Disord 11(Suppl 2), S33–39.

17.

Cummings

, Mega

, Gray

, Rosenberg-Thompson

, Carusi

, Gornbein

(1994) The Neuropsychiatric Inventory: Comprehensive assessment of psychopathology in dementia. Neurology 44, 2308–2314.

18.

Smith

(1982) Western Psychological Services, Los Angeles. Symbol digit modalities test (SDMT). Manual (revised)..

19.

Lezak

(1995), Neuropsychological assessment (3rd ed.) New York Oxford University Press.

20.

Reisberg

, Ferris

, de Leon

, Crook

(1982) The Global Deterioration Scale for assessment of primary degenerative dementia. Am J Psychiatry 139, 1136–1139.

21.

Higgins

, Thompson

, Deeks

, Altman

(2003) Measuring inconsistency in meta-analyses. BMJ 327, 557–560.

22.

Narasimhalu

, Effendy

, Sim

, Lee

, Chen

, Hia

, Xue

, Corrales

, Chang

, Wong

, Chen

, Tan

(2010) A randomized controlled trial of rivastigmine in patients with cognitive impairment no dementia because of cerebrovascular disease. Acta Neurol Scand 121, 217–224.

23.

Devanand

, Pelton

, D’Antonio

, Ciarleglio

, Scodes

, Andrews

, Lunsford

, Beyer

, Petrella

, Sneed

, Ciovacco

, Doraiswamy

(2018) Donepezil treatment in patients with depression and cognitive impairment on stable antidepressant treatment: A randomized controlled trial. Am J Geriatr Psychiatry 26, 1050–1060.

24.

Doody

, Ferris

, Salloway

, Sun

, Goldman

, Watkins

, Xu

, Murthy

(2009) Donepezil treatment of patients with MCI: A 48-week randomized, placebo-controlled trial. Neurology 72, 1555–1561.

25.

Dubois

, Chupin

, Hampel

, Lista

, Cavedo

, Croisile

, Louis Tisserand

, Touchon

, Bonafe

, Ousset

, Ait Ameur

, Rouaud

, Ricolfi

, Vighetto

, Pasquier

, Delmaire

, Ceccaldi

, Girard

, Dufouil

, Lehericy

, Tonelli

, Duveau

, Colliot

, Garnero

, Sarazin

, Dormont

, Hippocampus Study Group, Hippocampus Study Group (2015) Donepezil decreases annual rate of hippocampal atrophy in suspected prodromal Alzheimer’s disease. Alzheimers Dement 11, 1041–1049.

26.

Montero-Odasso

, Speechley

, Chertkow

, Sarquis-Adamson

, Wells

, Borrie

, Vanderhaeghe

, Zou

, Fraser

, Bherer

, Muir-Hunter

(2019) Donepezil for gait and falls in mild cognitive impairment: A randomized controlled trial. Eur J Neurol 26, 651–659.

27.

Salloway

, Ferris

, Kluger

, Goldman

, Griesing

, Kumar

, Richardson

, Donepezil 401 Study Group (2004) Efficacy of donepezil in mild cognitive impairment: A randomized placebo-controlled trial. Neurology 63, 651–657.

28.

Koontz

, Baskys

(2005) Effects of galantamine on working memory and global functioning in patients with mild cognitive impairment: A double-blind placebo-controlled study. Am J Alzheimers Dis Other Demen 20, 295–302.

29.

Peters

, Lorenz

, Fesche

, Schmidtke

, Hull

, Perneczky

, Ruther

, Moller

, Jessen

, Maier

, Kornhuber

, Jahn

, Luckhaus

, Gertz

, Schroder

, Pantel

, Teipel

, Wellek

, Frolich

, Heuser

(2012) A combination of galantamine and memantine modifies cognitive function in subjects with amnestic MCI. J Nutr Health Aging 16, 544–548.

30.

Winblad

, Gauthier

, Scinto

, Feldman

, Wilcock

, Truyen

, Mayorga

, Wang

, Brashear

, Nye

, GAL-INT-11/18 Study Group (2008) Safety and efficacy of galantamine in subjects with mild cognitive impairment. Neurology 70, 2024–2035.

31.

Bokde

, Cavedo

, Lopez-Bayo

, Lista

, Meindl

, Born

, Galluzzi

, Faltraco

, Dubois

, Teipel

, Reiser

, Moller

, Hampel

(2016) Effects of rivastigmine on visual attention in subjects with amnestic mild cognitive impairment: A serial functional MRI activation pilot-study. Psychiatry Res Neuroimaging 249, 84–90.

32.

Feldman

, Ferris

, Winblad

, Sfikas

, Mancione

, He

, Tekin

, Burns

, Cummings

, del

Ser T

, Inzitari

, Orgogozo

, Sauer

, Scheltens

, Scarpini

, Herrmann

, Farlow

, Potkin

, Charles

, Fox

, Lane

(2007) Effect of rivastigmine on delay to diagnosis of Alzheimer’s disease from mild cognitive impairment: The InDDEx study. Lancet Neurol 6, 501–512.

33.

Mamikonyan

, Xie

, Melvin

, Weintraub

(2015) Rivastigmine for mild cognitive impairment in Parkinson disease: A placebo-controlled study. Mov Disord 30, 912–918.

34.

Soysal

, Isik

, Stubbs

, Solmi

, Volpe

, Luchini

, D’Onofrio

, Pilotto

, Manzato

, Sergi

, Schofield

, Veronese

(2016) Acetylcholinesterase inhibitors are associated with weight loss in older people with dementia: A systematic review and meta-analysis. J Neurol Neurosurg Psychiatry 87, 1368–1374.

35.

Tricco

, Ashoor

, Soobiah

, Rios

, Veroniki

, Hamid

, Ivory

, Khan

, Yazdi

, Ghassemi

, Blondal

, Ho

, Ng

, Hemmelgarn

, Majumdar

, Perrier

, Straus

(2018) Comparative effectiveness and safety of cognitive enhancers for treating Alzheimer’s disease: Systematic review and network metaanalysis. J Am Geriatr Soc 66, 170–178.

36.

Beauchet

, Launay

, Montero-Odasso

, Annweiler

, Allali

(2015) Anti-dementia drugs-related changes in gait performance while single and dual tasking in patients with Alzheimer disease: A meta-analysis. Curr Alzheimer Res 12, 761–771.

37.

Hanfelt

, Peng

, Goldstein

, Lah

(2018) Latent classes of mild cognitive impairment are associated with clinical outcomes and neuropathology: Analysis of data from the National Alzheimer’s Coordinating Center. Neurobiol Dis 117, 62–71.

38.

Petersen

, Lopez

, Armstrong

, Getchius

TSD

, Ganguli

, Gloss

, Gronseth

, Marson

, Pringsheim

, Day

, Sager

, Stevens

, Rae-Grant

(2018) Practice guideline update summary: Mild cognitive impairment: Report of the Guideline Development, Dissemination, and Implementation Subcommittee of the American Academy of Neurology. Neurology 90, 126–135.

39.

Hogan

, Patterson

(2002) Progress in clinical neurosciences: Treatment of Alzheimer’s disease and other dementias–review and comparison of the cholinesterase inhibitors. Can J Neurol Sci 29, 306–314.

40.

Lundh

, Lexchin

, Mintzes

, Schroll

, Bero

(2017) Industry sponsorship and research outcome. Cochrane Database Syst Rev 2, MR000033.