Prevalence and Risk of Mild Cognitive Impairment in Low and Middle-Income Countries: A Systematic Review

Abstract

Background:

Mild cognitive impairment (MCI) is a cognitive state associated with increased risk of dementia. Little research on MCI exists from low-and middle-income countries (LMICs), despite high prevalence of dementia in these settings.

Objective:

This systematic review aimed to review epidemiological reports to determine the prevalence of MCI and its associated risk factors in LMICs.

Methods:

Medline, Embase, and PsycINFO were searched from inception until November 2019. Eligible articles reported on MCI in population or community-based studies from LMICs and were included as long as MCI was clearly defined.

Results:

5,568 articles were screened, and 78 retained. In total, n = 23 different LMICs were represented; mostly from China (n = 55 studies). Few studies were from countries defined as lower-middle income (n = 14), low income (n = 4), or from population representative samples (n = 4). There was large heterogeneity in how MCI was diagnosed; with Petersen criteria the most commonly applied (n = 26). Prevalence of amnesic MCI (aMCI) (Petersen criteria) ranged from 0.6%to 22.3%. Similar variability existed across studies using the International Working Group Criteria for aMCI (range 4.5%to 18.3%) and all-MCI (range 6.1%to 30.4%). Risk of MCI was associated with demographic (e.g., age), health (e.g., cardio-metabolic disease), and lifestyle (e.g., social isolation, smoking, diet and physical activity) factors.

Conclusion:

Outside of China, few MCI studies have been conducted in LMIC settings. There is an urgent need for population representative epidemiological studies to determine MCI prevalence in LMICs. MCI diagnostic methodology also needs to be standardized. This will allow for cross-study comparison and future resource planning.

Keywords

Epidemiology low-and middle-income country,mild cognitive impairment prevalence risk factors systematic review

INTRODUCTION

Mild cognitive impairment (MCI) defines an intermediate state of cognitive function between normal aging and dementia. Numerous definitions for MCI exist and prevalence estimates vary (range <1%up to 56%across different studies and definitions) depending on population sampling (age, clinical versus population), MCI case definition, and operationalization of the component criterion for an MCI case diagnosis [1 –7]. However, the majority of MCI research has been undertaken in high-income countries (HICs), namely North America, Europe, and Australia. This raising questions of generalizability of findings to low-and middle-income countries (LMIC) which vary by wealth, culture, ethnicity, research capacity, and infrastructure to support aging populations.

Studies examining MCI prevalence in LMICs have produced conflicting results. For example, the 10/66 study reported a range of estimates (0.8 to 4.3%) of Petersen defined amnestic MCI (aMCI) [8] across sites in Cuba, the Dominican Republic, Peru, Venezuela, Mexico, China, India, and Puerto Rico [9]. Findings from the World Health Organization’s Study on Global Ageing and Adult Health reported an overall MCI prevalence of 15.3%(95%CI: 14.4–16.3) when applying the National Institute of Ageing-Alzheimer’s Association (NIA-AA) criteria [10] across sites in China, Ghana, India, Mexico, Russia, and South-Africa [11]; with the individual country prevalence estimates lower (e.g., 8.5%in South Africa [12]). It is not clear what is driving the differences. Within studies, the differences likely reflect variability in the profile of risk and protective factors across sites as well as cultural/ethnic perceptions of cognitive aging and symptom reporting. Across studies, differences are likely due to heterogeneity in methodology, e.g., differences in sample selection and the MCI criteria used for diagnosis.

While some have suggested that MCI as a mode of prodromal classification can have a limited role in clinical and epidemiological settings, others argue that MCI could be a pragmatic tool for identifying individuals who could benefit from risk reduction [13]. There is promising evidence to support dementia risk reduction interventions in HICs [14], with indications of similar opportunities in LMICs [15]. Determining how best to identify individuals with MCI and the prevalence of the condition in LMICs will have important implications for planning intervention trials, treatment strategies, budgeting and public health surveillance. While several reviews on MCI have been conducted [1 –7], to our knowledge none have focused specifically on LMIC settings. Therefore, the aim of this systematic review was to report on the population prevalence and risk factors for MCI in these settings.

MATERIALS AND METHODS

This review adhered to standard reporting guidelines [16] and full details of the MOOSE checklist [17] are provided as part of the Supplementary Material. The review protocol can be made available by a member of the research team upon request.

Search strategy

Medline, Embase, and PsycINFO were searched from inception to the 10 January 2018, with updated searches run from 10 January 2018–6 November 2018 and from 6 November 2018–30 November 2019 (CR; See Supplementary Table 1 for the list of search terms).

Inclusion/exclusion criteria

Studies were included if: 1) the sample was from a LMIC at the time of the study, as defined by The World Bank [18]; 2) the study reported population-level or community-based data; and, both cross-sectional and cohort study designs were included; 3) the study described how MCI had been mapped; 4) sample age was ≥50 years; and 5) MCI prevalence was reported. No restrictions were placed on the definition of MCI used (as long as it was clearly defined), language or publication date. Randomized controlled trials, case-control studies, unpublished studies, and conference abstracts were excluded. Studies were also excluded if, for analysis, cognitive groups (e.g., dementia and MCI groups) were combined or the sample restricted (e.g., studies investigating MCI in disease specific groups such as diabetics or in illiterate participants only). Reviews were also retained and the reference lists of these interrogated for any missed paper.

Data analysis

Titles/abstracts were first screened, followed by the full text of any identified articles (CR and CVA). Where multiple publications using the same study were identified, these were retained for full text review and kept if they presented original findings. Disagreements were resolved by consensus or a third party (BCMS). Data including study characteristics, operationalization of MCI criteria, MCI prevalence estimates and risk factors for MCI were independently extracted by four investigators (AMG [Chinese Studies], AS, BCMS, and CVA).

Study quality (bias) was assessed using the tool developed by Hoy et al. [19]. Nine items were selected related to representativeness of the study sample, methods for case definition, and the statistical calculation of MCI prevalence. Each risk of bias item was scored ‘0’ (low risk) or ‘1’ (high risk) of bias (total score range: 0 to 9).

Forest plots of the population prevalence estimates of MCI, defined using the most commonly applied criteria across the studies were created using Prism-GraphPad 8 for Windows (GraphPad Software, San Diego, USA). This included plots for MCI defined using Petersen criteria (including all-MCI and aMCI; n = 26 studies [9 , 20–44]), International Working Group (IWG) criteria (n = 14 studies [45 –58]), or study specific criteria for Cognitive Impairment no Dementia (CIND; n = 10 studies [59 –68]). A meta-analysis was not possible due to large heterogeneity in methodology across the studies and the lack of key statistical information (i.e., confidence intervals for the MCI prevalence estimates) in most studies.

Role of the funding source

The review was completed as part of the NIHR Global Health Group: DePEC (Grant number: 16/137/62). AMG and BCMS have full access to the data and final responsibility to submit for publication.

RESULTS

Search yields

The electronic search identified n = 4,548 studies, with duplicates removed (Fig. 1). Following title/abstract screening, 162 studies were selected for full text review. This included a systemic review on MCI prevalence in China [7], where an additional n = 48 studies in Mandarin were identified. From these, 73 studies were selected for inclusion. An updated electronic search in November 2019 yielded 972 studies, and an additional five were included (total n = 78 studies). Two studies [66, 67] used the same dataset, but these were retained as they provided MCI prevalence estimates for different age groups.

Fig. 1

Study selection.

Study characteristics

Table 1 shows the characteristics of each study. Sample size ranged from n = 120 [69] to n = 32,715 [11]. Most studies included participants aged ≥60 years (n = 46 studies [22 , 70–87]) or ≥65 years (n = 16 studies [9 , 88–90]). The remaining studies included participants aged ≥50 years (n = 6 studies [11 , 91–93]), ≥55 years (n = 7 studies [24 , 94–96]), ≥70 years (n = 1 study [49]), and ≥80 years (n = 2 studies [20, 66]). One study [92] included only women.

Table 1

Study characteristics and MCI prevalence arranged by definition (ordered by age [lowest to highest])

Author	Location	Income Class	Age (y)	Summary population characteristics	Mild cognitive impairment (MCI)
					MCI definition	Total sample size	Cases	Prevalence %(95%CI)
MCI ACCORDING TO PETERSEN CRITERIA [8 , 102] (n = 26 studies)
Henao-Arboleda, 2008 [21]	Columbia	UMIC	≥50	Residents of metropolitan area of Medellin	aMCI	848	82	9.7 (7.6–11.7)
Lao, 2011 [31]	China	UMIC	≥55	Non-demented residents of Hainan province	MCI	7,665	326	4.25 (3.80–4.68)
Li, 2013 [43]	China	UMIC	≥55	Non-demented residents of Beijing with an MMSE ≥24 (The Beijing Ageing Brain Rejuvenation Initiative: BABRI)	All MCI	1,020	160	15.7
					aMCI-SD	1,020	65	6.4
					aMCI-MD	1,020	38	3.7
					naMCI	1,020	57	5.6
Qin, 2014 [34]	China	UMIC	≥55	Non-demented residents of Shanghai	aMCI	4,086	612	14.98 (13.88–16.02)
Qiu, 2003 [24]	China	UMIC	≥55	Non-demented residents of Chengdu	aMCI	3,910	92	2.40 (1.88–2.80)
Gao, 2011 [29]	China	UMIC	≥60	Non-demented residents of a community of Suzhou	aMCI	1,773	243	13.71 (12.1–15.22)
Huang, 2008 [73]	China	UMIC	≥60	Non-demented residents of Guangzhou	aMCI	4,697	257	5.47(4.82–6.09)
Juarez-Cedillo, 2013 [22]	Mexico	UMIC	≥60	Residents, registered with family medicine units (IMSS), of Mexico City	All MCI	2,944	190	6.5 (5.6 –7.4)
					aMCI-SD	2,944	71	2.4 (1.9 –3.0)
					aMCI-MD	2,944	75	2.6 (2.0 –3.1)
					naMCI-SD	2,944	35	1.2 (0.8 –1.6)
					naMCI-MD	2,944	9	0.3 (0.1 –0.5)
Liao, 2012 [32]	China	UMIC	≥60	Non-demented residents of Yichun	aMCI	399	41	10.28 (7.30–13.10)
Pan, 2012 [28]	China	UMIC	≥60	Non-demented residents of Jinhua	aMCI	897	154	17.17 (14.7–19.51)
Song, 2012 [35]	China	UMIC	≥60	Non-demented residents of a district of Foshan	aMCI	2,279	167	7.33 (6.26–8.34)
Su, 2013 [36]	China	UMIC	≥60	Non-demented residents of Xian	aMCI	796	145	18.22 (15.53–20.76)
Tang, 2007 [37]	China	UMIC	≥60	Non-demented residents of Beijing	aMCI	1,865	217	11.60 (10.18–13.02)
Tiwari, 2013 [25]	India	LMIC	≥60	Residents of Malihabad and Bakshi Ka Talab of Lucknow district of the State of Uttar Pradesh	All MCI	2,146	98	4.6 (3.7 –5.5)
Tong, 2013 [38]	China	UMIC	≥60	Non-demented residents of Tangshan	aMCI	1,575	200	12.7
Wang, 2012 [42]	China	UMIC	≥60	Non-demented residents of Tianjin	aMCI	3,678	408	11.1
Wang, 2017 [26]	China	UMIC	≥60	Residents of rural and urban areas in Shanghai	aMCI	1,005	224	22.3
Xu, 2014 [27]	China	UMIC	≥60	Non-demented residents of the Hebei province	All MCI	2,426	526	21.3 (19.2–23.1)^‡
Yin, 2011 [39]	China	UMIC	≥60	Non-demented residents of Huzhou	aMCI	2,164	310	14.33 (12.85–15.73)
Zhang, 2011 [41]	China	UMIC	≥60	Non-demented residents of Suzhou	aMCI	5,388	691	12.8
Zhu and Li, 2015 [40]	China	UMIC	≥60	Non-demented residents of some communities of Xinyang	aMCI	1,755	245	13.96 (12.34–15.50)
Jia, 2014 [114]	China	UMIC	≥65	Non-demented residents of 5 city of China	All MCI	10,276	2,137	19.5 (18.8–20.3)^¶
					alzMCI	10,276	630	5.6 (5.2–6.0)^¶
					cvdMCI	10,276	392	3.4 (3.3–4.1)^¶
					vrfMCI	10,276	507	4.6 (4.2–5.0)^¶
					otherMCI	10,276	608	5.6 (5.2–6.1)^¶
M Guo, 2012 [33]	China	UMIC	≥65	Non-demented residents of Xian	aMCI	264	35	13.26 (9.17–17.14)
Sosa, 2012 [9]	Cuba	UMIC	≥65	Residents of urban and rural settings	aMCI	2,620	47	1.5 (1.0–1.9)^†
	DR	UMIC			aMCI	1,767	25	1.3 (0.7–1.8)^†
	Peru	UMIC			aMCI	1,767	55	2.6 (1.9–3.3)^†
	Venezuela	UMIC			aMCI	1,820	22	1.0 (0.7–1.4)^†
	Mexico	UMIC			aMCI	1,821	58	2.8 (2.0–3.6)^†
	China	UMIC			aMCI	2,014	16	0.6 (0.3–0.9)^†
	India	LMIC			aMCI	1,802	77	4.6 (3.7–5.4)^†
Ogunniyi, 2016 [23]	Nigeria	UMIC	≥65	Residents of the Lalupon community, Oyo State (The IDEA Study)	All MCI	6,613	111	18.4 (14.9–21.9) ^*
					aMCI-SD	613	47	n/a
					aMCI-MD	613	45	n/a
					naMCI-SD	613	18	n/a
					naMCI-MD	613	1	n/a
Hai, 2012 [20]	China	UMIC	≥80	Non-demented residents of Chengdu	aMCI	202	61	30.2
INTERNATIONAL WORKING GROUP (IWG) ON MCI CRITERIA 2004 [97 , 103] (n = 14 studies)
Ding, 2015 [45]	China	UMIC	≥60	Non-demented community residents of Shanghai	All MCI	2,985	601	20.1 (18.7 –21.5)
					aMCI	2,985	393	13.2 (12.0 –14.4)
					naMCI	2,985	208	7.0 (6.1 –7.9)
Fang, 2015 [56]	China	UMIC	≥60	Non-demented residents of a community of Shanghai	aMCI	1,059	137	12.90 (10.9–14.9)
Fang, 2009 [54]	China	UMIC	≥60	Non-demented residents of Hangzhou	MCI	925	195	21.10 (18.45–23.58)
Tsoy, 2019 [55]	Kazakhstan	UMIC	≥60	Residents from Almaty, Kazakhstan	MCI	662	201	30.4 (26.9–33.9)
Vanoh, 2016 [53]	Malaysia	LMIC	≥60	Residents of four different states: Perak, Selangor, Kelantan and Johor (Towards Useful Ageing Study)	aMCI	1,993	315	15.8
Zhang, 2013 [58]	China	UMIC	≥60	Non-demented residents of Taicang	aMCI	2,460	450	18.29 (16.76–19.74)
Su, 2013 [52]	China	UMIC	≥60	Elderly of rural and urban areas in Xi’an	aMCI	796	145	18.2
Lee, 2012 [47]	Malaysia	LMIC	≥60	Residents of Kuala Lumpur	All MCI	318	67	21.1
					aMCI	318	49	15.4
					naMCI	318	18	5.7
Janelidze, 2018 [46]	Georgia	LMIC	≥65	Residents, without moderate to severe cognitive impairment, from urban (Tbilisi) and two rural areas (Kakheti and Imereti)	All MCI^§	238	66	27.7
					aMCI	238	38	16.0
					naMCI	238	29	9.9
					aMCI-MD	238	8	3.3
					naMCI-MD	238	12	5.1
Ma, 2016 [48]	China	UMIC	≥65	Non-demented residents of sixteen districts within the Tianjin urban boundary	All MCI	5,067	574	11.3 (8.2–14.4)
					aMCI-SD	5,067	227	4.5 (2.2–6.7)
					aMCI-MD	5,067	106	2.1 (0.8–3.4)
					naMCI-SD	5,067	214	4.2 (1.4–7.1)
					naMCI-MD	5,067	27	0.5 (0.3–0.8)
Pilleron, 2015 [50]	CAR	LIC	≥65	Residents of Bangui and Nola	All MCI	9,860	62	7.2
	ROC	LMIC		Residents of Brazzaville and Gamboma	All MCI	1,913	56	6.1
Rao, 2018 [51]	China	UMIC	≥65	Residents of Guangzhou	All MCI	2,111	299	14.2
					aMCI	2,111	258	12.2
					naMCI	2,111	41	2.0
Sun, 2016 [57]	China	UMIC	≥65	Non-demented residents of Baotou	MCI	384	40	10.42 (7.36–13.32)
Paddick, 2015 [49]	Tanzania	LIC	≥70	Residents of the Hai district	All MCI	296	46	6.3 (2.9–9.7)^\|\|
EUROPEAN CONSORTIUM ON ALZHEIMER’S DISEASE CRITERIA 2006 [105] (n = 2 studies)
Mohan, 2019 [85]	India	LMIC	≥60	Residents from Thiruvananthapuram, Kerala	All MCI	426	111	26.1 (22.1–30.4)
Dimitrov, 2012 [88]	Bulgaria	UMIC	≥65	Residents from the city of Varna	All MCI	540	36	6.7 (4.6–8.8)
NIA -AA CRITERIA 2011[10] (n = 2 studies)
Vancampfort, 2017 [11]	South Africa	UMIC	≥50	Nationally representative sample of elderly without severe cognitive impairment (The WHO Study on Global Ageing and Adult Health)	All MCI	32,715	5,005	15.3 (14.4–16.3)^**
	China	UMIC
	Ghana	LMIC
	India	LMIC
	Mexico	UMIC
	Russia	UMIC
Koyanagi, 2019 [93]	South Africa	UMIC	≥50	Nationally representative sample from nine provinces across South Africa	All MCI	3,672	312	8.5 (6.9–10.3)
STUDY SPECIFIC CRITERIA FOR MCI (n = 13 studies)
Das, 2007 [91]	India	LMIC	≥50	Non-demented/non-depressed residents of Kolkata	All MCI	745	111	14.9 (12.2–18.0)
					aMCI (with/without complainers)	745	45	6.0 (4.4–8.1)
					aMCI (complainers only)	745	34	4.6 (3.1–6.3)
					mdMCI	745	66	8.9 (6.8–11.3)
Wang, 2013 [95]	China	UMIC	≥55	Non-demented residents of two cities of Ningxia province	aMCI	1,033	199	19.26 (16.86–21.55)
Wang, 2016 [94]	China	UMIC	≥55	Non-demented residents of Ningxia province	aMCI	2,168	457	21.08 (19.4–22.7)
Huang, 2007 [30]	China	UMIC	≥60	Non-demented residents of Shenzhen	MCI	410	88	21.46 (17.49–25.23)
Khedr, 2015 [70]	Egypt	LMIC	≥60	Residents of Qena governorates in Southern Egypt	aMCI	691	12	1.7 (0.8–2.7)
Li, 2013 [74]	China	UMIC	≥60	Non-demented residents of Jinan	aMCI	1,226	115	9.38 (7.75–10.93)
Lu, 2019 [86]	China	UMIC	≥60	Residents of Ji County	In 2010
					All MCI		1,278	22.9
					MCI-A	5,581	1,060	19.0
					MCI-VaD		128	2.3
					MCI-O		89	1.6
					In 2015
					All MCI		1,541	27.8
					MCI-A	5,542	1,019	18.4
					MCI-VaD		377	6.8
					MCI-O		149	2.7
Wang, 2014 [75]	China	UMIC	≥60	Non-demented residents of Zhoushan	MCI	1,906	318	16.68 (15.01–18.27)
Wu, 2012 [76]	China	UMIC	≥60	Non-demented residents of Xian	aMCI	1,583	396	25.02 (22.88–27.04)
Zhao, 2015 [79]	China	UMIC	≥60	Non-demented residents of Jilin province	aMCI	976	171	17.60 (15.14–19.78)
Zhou, 2016[80]	China	UMIC	≥60	Non-demented residents of Changji	MCI	804	223	27.74 (24.64–30.67)
Zhu, 2013 [84]	China	UMIC	≥60	Non-demented residents of Zhejiang province	MCI	1,211	251	20.70 (18.44–22.89)
Chu, 2015 [89]	China	UMIC	≥65	Non-demented residents of a community of Shanghai	MCI	842	180	21.4 (18.6–24.0)
STUDY SPECIFIC CRITERIA FOR CIND/COGNITIVE IMPAIRMENT (n = 10 studies)
Brucki, 2014 [59]	Brazil	UMIC	≥50	The Mamirauá and Amanã Sustainable Development Reserves, Amazonian Basin	CIND	163	10	6.1
Cesar, 2016 [60]	Brazil	UMIC	≥60	Residents of rural and urban areas in the Tremembe area, State of Sao Paulo	CIND	630	135	19.5 (16.6–22.8)
Dominguez, 2018 [61]	Philippines	LMIC	≥60	Residents of Marikina City	CIND	1,367	317	23.2
Lei, 2008 [68]	China	UMIC	≥60	Normal cognitive and MCI residents from urban and rural areas of Guizhou province	CIND	4,323	665	15.4
Mejia-Arango, 2011 [65]	Mexico	UMIC	≥60	Representative sample of elderly in Mexico (Mexican Health and Ageing Study)	CIND	6,847	1,719	28.7^‡‡
Zhang, 2014^§§ [67]	China	UMIC	≥60	Residents of rural Ji County	CIND	5,550	1,295	23.3
Fei, 2009 [62]	China	UMIC	≥65	Residents of Taiyuan city	CIND	6,192	600	9.7 (9.6–9.8)^††
Guerchet, 2009 [63]	Benin, West Africa	LIC	≥65	Residents in the rural commune of Djidja centre	CIND	502	34	6.8
Guerchet, 2010 [64]	CAR	LIC	≥65	Older age residents of Bangui	CIND	496	124	25.0 (21.2–29.0)
	ROC	LMIC		Older aged residents of Brazzaville	CIND	520	98	18.8 (15.6–22.5)
Shi, 2013^§§ [66]	China	UMIC	≥80	Residents of rural Ji County	CIND	626	297	47.4 (43.5–51.4)
CUT -OFF SCORE OF NEUROPSYCHOLOGICAL ASSESSMENT TOOL (n = 3 studies)
Saha, 2010 [92]	India	LMIC	≥50	Non-demented women, without history of neurological disorders, from a village of Singur block of Hoogly district, Kolkata	Cognitive impairment	179	76	42.4
Ruan, 2019 [87]	China	UMIC	≥60	Residents from 20 communities in the Zhoujiaqiao Primary Health Service Area in Changning district, Shanghai,	MCI	5,328	500	9.67
Inocian, 2016 [69]	Philippines	LMIC	≥65	Elderly, with no diagnosed mental health conditions, from Cebu City	All MCI	120	76	63.3
DSM -IV CRITERIA [104] (n = 8 studies)
Hu, 2012 [96]	China	UMIC	≥55	Non-demented residents of Nei Monggol	MCI	9,266	1,782	19.48 (18.43–19.99)
Guo, 2013 [72]	China	UMIC	≥60	Non-demented residents of six towns of rural area of Hunan	aMCI	1,367	139	10.17 (8.57–11.69)
Zhang, 2015 [77]	China	UMIC	≥60	Non-demented residents of Taian	MCI	1,971	651	33.03
Zhang and Zeng, 2014 [78]	China	UMIC	≥60	Non-demented residents of Changsha	aMCI	1,764	229	16.27 (11.41–14.47)
Zhou, 2009 [82]	China	UMIC	≥60	Non-demented residents of Xinjiang	aMCI	2,986	205	10.21 (5.96–7.73)
Zhou, 2011 [81]	China	UMIC	≥60	Non-demented residents of Ningbo	aMCI	1,227	107	10.68 (7.14–10.22)
Zhu, 2009 [83]	China	UMIC	≥60	Non-demented residents of Wulumuqi	aMCI	1,511	148	9.79 (8.3–11.22)
Xiong, 2013 [90]	China	UMIC	≥65	Non-demented residents of Tianjin	aMCI	2,798	339	11.38 (10.91–13.26)

^*Weighted prevalence for age using the World Health Organization world population estimates 2015. ^†Weighted prevalence for age, gender, and education level. ^‡Weighted prevalence calculated using reciprocal probability weighting based on the Sixth Nationwide Population Census in China, 2010. ^§Participants with ‘all MCI’ could be categorized in more than one MCI subtype. ^¶Weighted prevalence using the direct standardization method adjusted by age and sex to the total Chinese population (according to the census conducted in 2005). ^||Weighted prevalence for stratification and age according to the WHO standard population. ^**Weighted prevalence according to the population structure as reported by the United Nations Statistical Division. ^††Weighted prevalence to account for selection probabilities and controlling for age and sex. ^‡‡Weighted prevalence to represent the total Mexican population. ^§§The study from Shi, 2013 and Zhang, 2014 used data from the same cohort study. 95%CI, confidence interval; alzMCI, MCI caused by prodromal Alzheimer’s disease; aMCI, amnestic MCI; CAR, Central African Republic; CIND, cognitive impairment no dementia; cvdMCI, MCI resulting from cerebrovascular disease; DR, Dominican Republic; DSM-IV, Diagnostic and Statistical Manual of Mental Disorders, fourth edition; LIC, low income country; LMIC, lower middle income country; MCI, mild cognitive impairment; MCI-A, MCI with significant memory impairment; MCI-MD, multi domain MCI; MCI-O, non memory/nonvascular related types of mild cognitive impairment; MCI-SD, single domain MCI; MCI-VaD, significant executive function impairment and relationship with cerebral vascular disease; naMCI, non-amnestic MCI; NIA-AA National Institute on Aging and Alzheimer’s Association; otherMCI, MCI caused by other diseases; ROC, Republic of Congo; UMIC, upper middle income country; vrfMCI, MCI with vascular risk factors.

As shown in Table 1, four studies [9 , 64] analyzed MCI prevalence for multiple countries. The majority of studies have been conducted in China (n = 55 studies [9 , 94–96]), followed by India (n = 6 studies [9 , 92]), Mexico (n = 4 studies [9 , 65]), Brazil (n = 2 studies [59, 60]), Malaysia (n = 2 studies [47, 53]), the Philippines (n = 2 studies [61, 69]), Central African Republic (n = 2 studies [50, 64]), South Africa (n = 2 studies [11, 93]), Republic of Congo (n = 2 studies [50, 64]), and one each in Colombia [21], Nigeria [23], Cuba [9], Dominican Republic [9], Peru [9], Venezuela [9], Georgia [46], Kazakhstan [55], Tanzania [49], Bulgaria [88], Ghana [11], Russia [11], Egypt [70], and Benin West-Africa [63]. Therefore, most studies (n = 67 studies [9 , 94–96]) were from sites in upper middle-income countries, 14 studies [9 , 91–93] were from sites in lower middle-income countries, and four studies [49 , 64] were from sites in low-income countries. One study [9] included data collected during 2003–2007 from eight sites, one of which was Puerto Rico. This country was declared high-income by the World Bank in 2002. Therefore, the prevalence data for Puerto Rico has been excluded. Only four studies [11 , 93] selected participants from a representative country-wide sample. The remaining studies included a sample of community residents from a specific region, city or district(s).

Quality assessment

The detailed quality assessment is reported in Supplementary Table 2. Two studies [44, 93] obtained a low risk of bias score across all nine domains assessed, with the majority of studies (n = 66 studies) only having high risk scores in 1–3 domains [9 , 94–96]. These were mostly related to a lack of, or unclear use of, randomization procedures and that the study sample was unlikely to be representative of the national population.

MCI criteria

A shown in Table 1, numerous criteria were used to diagnose MCI including: 1) Petersen’s criteria [8 , 97–102] (n = 26 studies [9 , 20–44]); 2) IWG criteria [103] (n = 14 studies [45 –58]); 3) study specific criteria for CIND (n = 10 studies [59 –68]); 4) study specific criteria for MCI (n = 13 studies [30 , 95]); 5) Diagnostic and Statistical Manual of Mental Disorders, fourth edition (DSM-IV) criteria [104] (n = 8 studies [72 , 96]); 6) MCI based on a score from a neuropsychological assessment tool (n = 3 studies [69 , 92]); 7) NIA-AA criteria [10] (n = 2 studies [11, 93]); or 8) the European Consortium on Alzheimer’s Disease criteria [105] (n = 2 studies [85, 88]). A full description of the different MCI criteria applied across the studies is in Supplementary Tables 3 and 4.

Operationalizing MCI criteria

Full details of how MCI criteria were operationalized (and any modifications that were made to mapping the original diagnostic criteria) are detailed below and outlined in Supplementary Table 5. Overall, MCI was diagnosed using one or more of the following criteria: 1) subjective/informant cognitive or memory complaint; 2) global cognitive performance; 3) domain specific cognitive performance; 4) physical functioning; 5) no dementia; and 6) other factors (e.g., disease related co-mobility). Additional assessments were used in 7/78 studies [24 , 70]. In four studies, a Clinical Dementia Rating (CDR) score between 0 and 0.5 [27], or a score of 0.5 was needed for diagnosis of MCI [24 , 51]. In four studies [24 , 70], it was required that cognitive impairment was independent of other factors such as depression.

Domain specific MCI

Based on the cognitive domain test scores, 17 studies [9 , 91] stratified MCI into different subtypes. This included non-amnestic MCI (naMCI; n = 8 studies [22 , 51]), aMCI single domain (aMCI-SD; n = 4 studies [22 , 48]), multi-domain aMCI (aMCI-MD) (n = 5 studies [22 , 48]), multi-domain non-amnestic MCI (naMCI-MD; n = 4 studies) [22 , 48], and single domain naMCI (naMCI-SD; n = 3 studies [22 , 48]).

Subjective cognitive/memory complaint

Cognitive/memory complaints were included as part of the MCI diagnosis in 57/78 studies [9 , 93–95]. Complaint was typically required to be subjective and focused on memory (n = 43 studies [9 , 95]) or cognition in general (n = 11 studies [23 , 88]). In 25 studies, complaints could also be reported by an informant [22 , 95]. Thirty studies did not specify how cognitive/memory complaints were assessed or the information was not reported [20 , 96].

Global cognitive function

Global cognitive function was assessed in 61/78 studies [21–23 , 96]. In 33 studies [23 , 95], global cognitive function was required to be impaired, while in 28 studies [21 , 96], it was required to be preserved or within normal limits. In total, 10 different neuropsychological assessment tools were used to assess global cognitive function including the Mini-Mental State Examination (MMSE; n = 40 studies [20 , 96]), the Montreal Cognitive Assessment (MoCA; n = 10 studies [27 , 89]), the CDR (n = 5 studies [44 , 70]), the Community Screening Instrument for Dementia (CSI-D; n = 3 studies [50 , 64]), the Consortium to Establish a Registry for Alzheimer’s Disease battery (CERAD; n = 2 studies [21, 91]), the Five Word Test (n = 2 studies [63, 64]), the Cambridge Examination for Mental Disorders-Revised (CAMDEX-R; n = 1 study [20]), the Identification and Intervention for Dementia in Elderly Africans (IDEA) cognitive screen (n = 1 study [23]), The Memory Impairment Screen (MIS; n = 1 study [88]), and the Informant Questionnaire on Cognitive Decline in the Elderly (IQCODE; n = 1 study [65]). Of the studies using the MMSE, and with reported cut-off scores, 22 studies [26 , 94–96] used education specific cut-off scores, and 16 studies used a mixture of the following:

To demonstrate normal cognitive function: MMSE ≥24 (n = 4 studies [43 , 78]), MMSE one standard deviation (1SD) from norm (n = 1 study [68]), MMSE ≥19 (n = 2 studies [47, 53]), MMSE ≥23 (n = 1 study [22]) and MMSE range 24–26 (n = 1 study [36]); and,

To demonstrate cognitive impairment: MMSE≤24 (n = 2 studies [25, 92]), MMSE≤26 (n = 2 studies [34, 74]), MMSE≤25 (n = 1 study [88]), MMSE≤27 (n = 1 study [66]), and MMSE range 20–25 (n = 1 study [69]).

Domain specific cognitive function: Memory

Memory was individually assessed for impairment in 41 studies [9 , 93]. Some studies used neuropsychological assessment tools to assess memory impairment. The MMSE (n = 3 studies [20 , 52]), and the Wechsler Memory Scale (WMS; n = 3 studies [27 , 62]) were the most frequently used, followed by the MIS (n = 2 studies [22, 88]), the CSI-D (n = 1 study [9]), the Cross-Cultural Cognitive Examination (CCCE; n = 1 study [65]), and the CAMDEX (n = 1 study [20]). The remaining studies used individual memory tests, of which the Auditory-Verbal Learning Test (n = 6 studies [43–45 , 62]) was most often used followed by the CERAD 10 word learning test (n = 5 studies [9 , 93]), the digit span test (n = 4 studies [11 , 53]), the Brief Cognitive Screening Battery delayed recall task (n = 2 studies [59, 60]), the Fuld Object Memory Evaluation (n = 1 study [45]), the stick test (n = 1 study [45]), the Renminbi test (n = 1 study [45]), the IDEA 10 word learning test (n = 1 study [23]), the MoCA free delayed recall test (n = 1 study [46]), the Rey-Osterrieth complex figure test (n = 1 study [43]), Rey Auditory Verbal Learning Test (RAVLT) total learning (n = 1 study [47]), RAVLT delayed recall (n = 1 study [47]), the MMSE memory subtask (n = 1 study [48]), WMS-III local memory test (n = 1 study [70]), CDR memory score (n = 1 study [70]), CCCE verbal memory (n = 1 study [65]), and the free and cued selective reminding test (n = 1 study [50]). Nineteen studies describe cut-off scores for impairment including: <1.5 SDs adjusted for age and education (n = 8 studies [9 , 78]), <1.5 SDs below norms (n = 6 studies [27 , 91]), <1 SD below norms (n = 2 studies [47, 70]), < -1 SD adjusted for age, education, and country (n = 2 studies [11, 93]), and 1.5–2 SDs below the overall mean adjusted for age and education (n = 1 study [56]).

Domain specific cognitive function: Other domains

Non-memory cognitive test performance was assessed in 25 studies [11 , 93]. Nine different test batteries were used including: the Wechsler Adult Intelligence Scale (n = 3 studies [27 , 62]), CERAD (n = 1 study [88]), CSI-D (n = 1 study [49]), the Alzheimer’s Disease Assessment Scale Cognitive Subscale (n = 1 study [22]), IDEA cognitive screen (n = 1 study [23]), CAMCOG (n = 1 study [25]), MOCA (n = 1 study [46]), Malayalam version of Addenbrooke’s Cognitive examination (n = 1 study [85]), and the CCCE (n = 1 study [65]). In addition, domain-specific tests (e.g., attention and executive function) were used, with the four most common being the verbal fluency test (n = 10 studies [11 , 93]), the Trail Making Test (n = 7 studies [27 , 62]), the Clock Drawing Test (n = 5 studies [43 , 51]), and the Boston Naming Test (n = 2 studies [43, 62]). Two studies [56, 79] stated that non-memory domains were assessed, however they did not describe which tests were used.

Dementia

Criteria used to exclude dementia included the DSM-IV (n = 27 studies [20 , 92]), performance on a neuropsychological assessment battery (n = 8 studies [22 , 89]), a combination of cognitive test performance and evaluation of Activities of Daily Living (ADL)/Instrumental Activities of Daily Living (IADL; n = 4 studies [11 , 93]), National Institute of Neurological and Communicative Diseases and Stroke - Alzheimer's Disease and Related Disorders Association (NINCDS-ADRDA) criteria (n = 3 studies [23 , 62]), ICD-10 criteria (n = 1 study [25]), the 10/66 dementia algorithm (n = 1 study [9]), diagnoses by a doctor (n = 1 study [53]) or NIA-AA criteria (n = 1 study [60]).

Functional performance

Fifty-six studies reported that physical functioning, including ADL/IADL, were part of the assessment for MCI [9 , 93–95]. However, 26 of these studies did not report the method used to determine physical functioning status [24–26 , 95]. Thirty-six studies [9 , 95] exclusively assessed either ADL or IADL. The majority of studies required persevered ADL/IADL for a diagnosis of MCI, with only 13 studies [9 , 91] allowing for subtle changes/mild functional impairment. In 26 studies, impairment in ADL/IADL was assessed using previously developed tools, of which the Katz ADL scale was the most often used (n = 8 studies [11 , 93]), followed by the Lawton and Brody scale (n = 5 studies [21 , 53]), the Functional Activities Questionnaire (n = 4 studies [44 , 60]), the CDR (n = 2 studies [45, 51]), the CSI-D informant interview (n = 2 studies [9, 50]), the Everyday Ability Scale for India (n = 2 studies [85, 91]), the Barthel scale (n = 1 study [21]), the Clinician Home Based Interview to assess Function (n = 1 study [23]), and the IQCODE (n = 1 study) [60].

MCI prevalence

The prevalence estimates reported in this review were all determined at the time of the study. However, one study [86] reported MCI prevalence at two time points (2010 and 2015). Across the different definitions used, 31 studies [11 , 96] calculated overall MCI prevalence, 47 studies [9 , 95] calculated aMCI prevalence, eight studies [22 , 51] calculated naMCI prevalence, and 10 studies [59 –68] calculated CIND prevalence. Two studies [44, 86] subtyped MCI by etiology as defined by: MCI caused by prodromal Alzheimer’s disease [44], MCI resulting from cerebrovascular disease [44], MCI with vascular risk factors [44], MCI with significant memory impairment [86], MCI with significant executive function impairment and relationship with cerebral vascular disease [86], and MCI caused by other factors [86]. As shown in Table 1, MCI prevalence ranged from 0.3%(95%CI: 0.1–0.5) in a sample from Mexico (n = 2,944≥60 years; naMCI multiple domain, Petersen criteria) [22] to 63.3%in a sample from the Philippines (n = 120≥65 years; MCI defined as an MMSE score 20–25 out of 30) [69]. Specifically, for Petersen criteria, prevalence of aMCI ranged from 0.6%(95%CI: 0.3–0.9) [9] to 22.3%[26]. Similar variability was seen across studies using the IWG Criteria for aMCI (range 4.5%[48] to 18.3%[58]; n = 9 studies); IWG criteria for all-MCI (range 6.1%[50] to 30.4%[55]; n = 10 studies); studies using CIND criteria (range 6.1%[59] to 47.4%[66]; n = 10 studies), studies using study specific criteria to diagnose MCI (range 1.6%[86] to 27.7%[80]; n = 13 studies); DSM-IV criteria (range 9.8%[83] to 33.0%[77]; n = 8 studies); studies using neuropsychological tests (range 9.7%[87] to 63.3%[69]; n = 3 studies); NIA-AA criteria (range 8.5%[93] to 15.3%[106]; n = 2 studies) and European Consortium of AD criteria (range 6.7%[88] to 26.1%[85]; n = 2 studies).

The forest plots in Fig. 2 show the MCI prevalence estimates for Petersen defined all-MCI (Fig. 2A) and aMCI (Fig. 2B), the IWG criteria (Fig. 2C), and criteria for CIND (Fig. 2D). The plots show there is large variability in MCI prevalence across studies even when the same criteria are applied in the same country albeit in different samples. In contrast, prevalence estimates are generally, although not always, more consistent across countries in multi-site studies (n = 4 studies [9 , 64]) when the same methods are used (0.6%–4.6%[9]; 6.1%–7.2%[50]; 18.8%–25.0%[64]).

Fig. 2

A. Forest plot of MCI prevalence from studies using Petersen’s criteria for Amnestic MCI (ordered by age). Note: red dotted line indicates 10%prevalence. 95%CI, 95%confidence interval; aMCI, amnesic mild cognitive impairment; DR, Dominican Republic. B. Forest plot of MCI prevalence from studies using Petersen’s criteria for All MCI (ordered by age). Note: red dotted line indicates 10%prevalence. 95%CI, 95%confidence interval; MCI, mild cognitive impairment. C. Forest plot of MCI prevalence from studies using the International Working Group criteria (ordered by age). Note: red dotted line indicates 10%prevalence. 95%CI, 95%confidence interval; IWG, International Working Group; CAR, Central African Republic; ROC, Republic of Congo. D. Forest plot of MCI prevalence from studies using definition of Cognitive Impairment No Dementia criteria (ordered by age). Note: red dotted line indicates 10%prevalence. 95%CI, 95%Confidence Interval; CAR, Central African Republic; ROC, Republic of Congo; CIND, Cognitive Impairment No Dementia.

Associated risk factors

Risk factors for prevalent MCI were investigated in 64/78 studies [20–24 , 96]. Two studies [53, 67] did not report risk factor information in the original article; however, risk factor data for the same cohort were later published [107, 108]. In this scenario, we have added the risk factor information as documented in the most recent publication. One paper [27] reported additional vascular risk factor information in a separate article [109] and we have also included this data. Significant risk factors for MCI included increased age (n = 46/64 studies [22 , 96]), sex (n = 41/64 studies; in 37 studies [22–24 , 93–96] women had higher risk and in four studies [21 , 87] men had higher risk), and low level of education (n = 44/64 studies [20–23 , 94]). Other significant risk factors included the presence of disease related co-morbidities (e.g., hypertension, stroke, coronary heart disease) [12 , 94], low monthly income/low economic status [27–29 , 92], marital status (without spouse) [28 , 92], occupation (physical labor) [27 , 83], geographic area (rural location) [37 , 68], diabetes [34 , 91], alcohol consumption [20 , 94], high body mass index [22 , 94], living alone [28 , 81], Apolipoprotein E4 (APOE E4) carrier [32, 94], low physical activity [28 , 84], current or a history of smoking [26 , 91], sleep (poor) [26 , 84], depression [22 , 90], and an introverted personality [29 , 58]. Protective factors included maintaining social contact with others [29 , 78] and following a healthy diet/consuming healthy dietary components [35 , 81] or specifically drinking tea [26 , 89]. See Supplementary Table 6 for full details of all risk factors reported across the different studies.

DISCUSSION

This is the first systematic review, to our knowledge, that has focused on MCI prevalence and its risk factors specifically in LMICs. The results highlight that MCI research in LMICs is largely restricted to upper-middle income countries, namely China. Further, MCI research is characterized by wide variation in population sampling, the case definition used for an MCI diagnosis, operationalization of the component criterion and prevalence estimates. These differences make cross-study comparison extremely difficult and highlight the urgent need for consensus in how MCI is defined across different settings.

MCI prevalence ranged from 0.3%[22] to 63.3%[69]. This variability was not reduced when grouping prevalence estimates by case definition. However, as shown in Fig. 2, there was a general pattern. Similar to what is observed in HICs [110], we found that diagnostic criteria that are more restrictive and capture a single impairment (e.g., Petersen aMCI criteria; range 0.6%(95%CI: 0.3–0.9) [9] to 22.3%[26]) have generally lower prevalence estimates compared to more general criteria that capture broader dysfunction (e.g., CIND where the majority of studies n = 7/10 reported a prevalence >15%). Although it is important to note that estimated prevalence for specific criteria did vary considerably. In relation to age, across all definitions, MCI appears to be rare in the very young, i.e., people <50 years. Furthermore, aMCI (Petersen Criteria) and IWG generally have a lower prevalence in studies where people are aged ≥65 versus people aged ≥60 with the opposite trend observed for all MCI (Petersen Criteria) and CIND.

Regarding definition, across studies, the most widely applied criteria were Petersen defined aMCI (n = 26 studies [9 , 20–44]) requiring subjective/informant memory complaint, normal global cognitive function, impaired memory, preserved (or relatively preserved in later definitions) physical function and no dementia. Prevalence of aMCI ranged from 0.6%(95%CI: 0.3–0.9) [9] to 22.3%[26]. Similar variability was seen across studies using the IWG Criteria for aMCI (range 4.5%[48] to 18.3%[58]) and IWG criteria for all-MCI (range 30.4%[55] to 6.1%[50]). These results are in line with previous systematic reviews of MCI incorporating studies predominately from HICs [111]. Variability in prevalence is likely due to differences in sample characteristics (e.g., age, educational attainment, and distribution of risk and protective factors) and methodology (e.g., test batteries used to assess cognitive and physical function, cut-off scores for impairment and whether the analyses were adjusted for factors such as age, sex, and education) across studies. Indeed, all of the studies that included multiple sites demonstrated that when the same methods were used to diagnose MCI, prevalence estimates were generally (although not always) more comparable across countries [9 , 64].

Similar to findings in high-income countries, both modifiable and non-modifiable risks factors were identified for MCI. Key socio-demographic risk factors included increased age, sex (usually, but not always, female) and low educational attainment. Modifiable health and lifestyle risk factors included, but were not limited to, smoking, presence of cardiovascular related diseases, social contact, occupation, physical activity, and dietary related factors. These findings support the development of novel public health interventions to reduce risk of cognitive impairment targeting education, cardio-metabolic health, and lifestyle factors that are applicable to the specific context of LMIC settings. However, a key knowledge gap highlighted by the review is the lack of research into context specific risk factors. Indeed, compared to HICs factors such as lifelong disadvantage, food insecurity, poverty, and absence of robust health and social care services might also be important in increasing risk of MCI and dementia in these settings.

Of note, is the scarcity of studies on MCI from countries classified as low-income (only n = 4 studies [49 , 64]). Further, no studies were identified from LMICs in the Middle East, with the exception of one study from Egypt [70]. As shown in Fig. 3, most MCI research in LMICs has come from cohorts in the Far East (e.g., China and parts of Asia and South-Asia including Malaysia, Philippines, and India), South America and the Caribbean (including Cuba, Dominican Republic, Mexico) and an increase in research in Africa (Tanzania, Nigeria, Central African Republic, Republic of Congo) only in the last five years (i.e., from 2015 onwards). Few studies have also been conducted in European LMICs with the exception of Bulgaria [88], Russia [106], and Georgia [46]. This lack of research into MCI could reflect the more recent demographic transition and population ageing in LMICs, highlighted by an increase in dementia-specific research in the past 10 years [112]. Also, across different LMIC settings there are high levels of low educational attainment/illiteracy in older people and there are often no norms for cognitive testing making MCI diagnosis challenging. Furthermore, there are typically very few specialist clinicians able to supervise this type of work in LMIC settings, with the exception of some countries like China.

Fig. 3

World map showing each study site, number of studies in each site (n) and the reported MCI prevalence estimate(s).

Strengths and weaknesses

The study has a number of strengths. We undertook a wide literature search capturing many of the different definitions of MCI. This allowed for a more comprehensive synthesis of the types of criteria used to diagnose MCI across the many different LMICs. Some studies, however, could still have been missed if they defined MCI outside the scope of the search. There are some weaknesses. First, the electronic search was undertaken in English and therefore studies published in other languages, including those common in LMICs such as Spanish, Portuguese, and French could have been missed if they were not recorded in EMBASE, Pubmed, or PsycInfo. We minimized the risk of not capturing Chinese articles by including findings from a recent systematic review on MCI prevalence in China [7]. Structuring the search this way could possibly explain the large number of MCI studies captured from China compared to other LMICs. This difference could also be due to variability in research investment into aging and dementia. Second, we focused only on cross-sectional studies that reported MCI prevalence estimates. Therefore, we did not investigate whether MCI is predictive of future dementia in LMICs or what the risk factors for incident MCI are. As such, we are unable to make recommendations as to which criteria are the most “useful”. This was beyond the scope of the review. Last, given the paucity of research into aging and dementia in LMIC settings we included any population-based study in the review; and only four were population-representative [11 , 93]. MCI prevalence results in non-representative samples must be viewed with caution as they may be biased for example by sampling (e.g., difference in location such as urban versus rural) and differences in the profile of risk/protective factors (e.g., demographic, health, and socio-economic status).

Conclusion

Numerous definitions of MCI have been proposed [113]. Determining which, if any, are suitable for application in LMICs will require an in-depth evaluation of not only how well they capture people with cognitive impairment, but also whether the condition is predictive of future dementia in these settings. To achieve this, consensus on how MCI is defined will be required, particularly in settings with varying educational levels amongst older people, varying cultural milieu and expectations resulting in challenges in MCI case identification. Nevertheless, given the high burden of dementia now seen in LMICs, identification of these higher risk individuals at a stage where intervention could take place is likely to have a high impact on the burden of disease associated with cognitive impairment and dementia in these settings. Thus, to further understand MCI prevalence in these settings there is an urgent need for more high quality, population representative MCI prevalence studies, particularly in countries classified as low income.

Footnotes

ACKNOWLEDGMENTS

This project was funded by a Newcastle-Monash Universities SEED award and as part of NIHR Global Group: DePEC (Grant Number: 16/137/62).

Authors’ disclosures available online ().

The supplementary material is available in the electronic version of this article: .

References

Bischkopf

, Busse

, Angermeyer

(2002) Mild cognitive impairment–a review of prevalence, incidence and outcome according to current approaches. Acta Psychiatr Scand 106, 403–414.

Nie

, Xu

, Liu

, Zhang

, Lei

, Hui

, Zhang

, Wu

(2011) The prevalence of mild cognitive impairment about elderly population in China: A meta-analysis. Int J Geriatr Psychiatry 26, 558–563.

Roberts

, Knopman

(2013) Classification and epidemiology of MCI. Clin Geriatr Med 29, 753–772.

Alexander

, Perera

, Ford

, Arrighi

, Foskett

, Debove

, Novak

, Gordon

(2015) Age-stratified prevalence of mild cognitive impairment and dementia in European populations: A systematic review. J Alzheimers Dis 48, 355–359.

, Yu

, Sun

, Zhang

, Wang

, Qin

(2017) The prevalence and progression of mild cognitive impairment among clinic and community populations: A systematic review and meta-analysis. Int Psychogeriatr 29, 1595–1608.

Ward

, Arrighi

, Michels

, Cedarbaum

(2012) Mild cognitive impairment: Disparity of incidence and prevalence estimates. Alzheimers Dement 8, 14–21.

Xue

, Li

, Liang

, Chen

(2018) The prevalence of mild cognitive impairment in China: A systematic review. Aging Dis 9, 706–715.

Petersen

, Smith

, Waring

, Ivnik

, Tangalos

, Kokmen

(1999) Mild cognitive impairment: Clinical characterization and outcome. Arch Neurol 56, 303–308.

Sosa

, Albanese

, Stephan

, Dewey

, Acosta

, Ferri

, Guerra

, Huang

, Jacob

, Jimenez-Velazquez

, Rodriguez

, Salas

, Williams

, Acosta

, Gonzalez-Viruet

, Hernandez

, Shuran

, Prince

, Stewart

(2012) Prevalence, distribution, and impact of mild cognitive impairment in Latin America, China, and India: A 10/66 population-based study. PLoS Med 9, e1001170.

10.

Albert

, DeKosky

, Dickson

, Dubois

, Feldman

, Fox

, Gamst

, Holtzman

, Jagust

, Petersen

(2011) The diagnosis of mild cognitive impairment due to Alzheimer’s disease: Recommendations from the National Institute on Aging-Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease. Alzheimers Dement 7, 270–279.

11.

Vancampfort

, Stubbs

, Lara

, Vandenbulcke

, Swinnen

, Koyanagi

(2017) Mild cognitive impairment and physical activity in the general population: Findings from six low- and middle-income countries. Exp Gerontol 100, 100–105.

12.

Koyanagi

, Veronese

, Stubbs

, Vancampfort

, Stickley

, Oh

, Shin

, Jackson

, Smith

, Lara

(2019) Food insecurity is associated with mild cognitive impairment among middle-aged and older adults in South Africa: Findings from a nationally representative survey. Nutrients 11, 749.

13.

Saunders

, Ritchie

, Russ

, Muniz-Terrera

, Ritchie

(2018) Evolution and future directions for the concept of mild cognitive impairment. Int Psychogeriatr 30, 1431–1434.

14.

Livingston

, Sommerlad

, Orgeta

, Costafreda

, Huntley

, Ames

, Ballard

, Banerjee

, Burns

, Cohen-Mansfield

, Cooper

, Fox

, Gitlin

, Howard

, Kales

, Larson

, Ritchie

, Rockwood

, Sampson

, Samus

, Schneider

, Selbaek

, Teri

, Mukadam

(2017) Dementia prevention, intervention, and care. Lancet 390, 2673–2734.

15.

Mukadam

, Sommerlad

, Huntley

, Livingston

(2019) Population attributable fractions for risk factors for dementia in low-income and middle-income countries: An analysis using cross-sectional survey data. Lancet Glob Health 7, e596–e603.

16.

Moher

, Liberati

, Tetzlaff

, Altman

, medicine

PGJP

(2009) Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement. PLoS Med 6, e1000097.

17.

Stroup

, Berlin

, Morton

, Olkin

, Williamson

, Rennie

, Moher

, Becker

, Sipe

, Thacker

(2000) Meta-analysis of observational studies in epidemiology: A proposal for reporting. Meta-analysis Of Observational Studies in Epidemiology (MOOSE) group. JAMA 283, 2008–2012.

18.

WBG, Low and middle income - income level, TheWorld Bank Group, https://data.worldbank.org/income-level/low-and-middle-income, Accessed January 25, 2018..

19.

Hoy

, Brooks

, Woolf

, Blyth

, March

, Bain

, Baker

, Smith

, Buchbinder

(2012) Assessing risk of bias in prevalence studies: Modification of an existing tool and evidence of interrater agreement. J Clin Epidemiol 65, 934–939.

20.

Hai

, Dong

, Liu

, Zou

(2012) Occurrence and risk factors of mild cognitive impairment in the older Chinese population: A 3-year follow-up study. Int J Geriatr Psychiatry 27, 703–708.

21.

Henao-Arboleda

, Aguirre-Acevedo

, Munoz

, Pineda

, Lopera

(2008) Prevalence of mild cognitive impairment, amnestic-type, in a Colombian population. [Spanish]. Rev Neurol 46, 709–713.

22.

Juarez-Cedillo

, Sanchez-Arenas

, Sanchez-Garcia

, Garcia-Pena

, Hsiung

GYR

, Sepehry

, Beattie

, Jacova

(2013) Prevalence of mild cognitive impairment and its subtypes in the Mexican population. Dement Geriatr Cogn Disord 34, 5–6.

23.

Ogunniyi

, Adebiyi

, Adediran

, Olakehinde

, Siwoku

(2016) Prevalence estimates of major neurocognitive disorders in a rural Nigerian community. Brain Behav 6, e00481.

24.

Qiu

, Tang

, Zhang

, Han

, Dai

, Lu

, Wu

, Wang

, Chen

, Guo

, Ding

, Li

, Liu

(2003) The prevalence of mild cognitive impairment among residents aged 55 or over in Chengdu area. [Chinese]. Chin J Epidemiol 24, 1104–1107.

25.

Tiwari

, Srivastava

, Tripathi

, Pandey

, Agarwal

, Pandey

, Tiwari

(2013) Prevalence of psychiatric morbidity amongst the community dwelling rural older adults in northern India. Indian J Med Res 138, 504–514.

26.

Wang

, Xiao

, Chen

, Yang

, Dong

, Cheng

, Li

, Wang

, Zhu

, Yang

, Li

, Su

, Liu

, Dai

, Zhang

(2017) Prevalence, incidence, risk and protective factors of amnestic mild cognitive impairment in the elderly in Shanghai. Curr Alzheimer Res 14, 460–466.

27.

, Xie

, Song

, Yu

, Wang

, An

, Zhu

, Han

, Zhao

, Zhang

, Dong

, Chai

, Gao

, Zhang

, Wang

(2014) High prevalence of mild cognitive impairment in the elderly: A community-based study in four cities of the Hebei province, china. Neuroepidemiology 42, 123–130.

28.

Pan

, Wang

, Wu

, Cheng

(2012) Study on prevalence rate and quality of life of elderly patients with mild cognitive impairment in community. Chin J Prac Nurs 29, 6–9.

29.

Gao

, Jiang

, Gao

, Nie

, Xu

(2011) Prevalence of mild cognitive impairment and its risk factors among elderly people in Canglang District of Suzhou City. J Occup Health 27, 2676–2678.

30.

Huang

, Han

, Liu

(2007) The study and analysis on mild cognitive impairment of the elderly in communities. Mod Nurs 13, 2678–2680.

31.

Lao

, Zhang

, Luo

, Yi

, Huang

, Wu

(2011) Prevalence of mild cognitive impairment among 55-years old or over individuals in Hainan Island. Hainan Med J 22, 112–114.

32.

Liao

, Gao

, Xiong

, Yi

, Li

, Wan

, et al (2012) The early evaluation and intervention strategies of mild cognitive impairment in Yichun area. J Yichun Coll 24, 77–79.

33.

Guo

, Gao

, Zhang

, Li

, Xu

, Wang

, Qu

, Guo

(2012) Prevalence of dementia and mild cognitive impairment in the elderly living in nursing and veteran care homes in Xi’an, China. J Neurol Sci 312, 39–44.

34.

Qin

, Chen

, ZW

(2014) Investigation of mild cognitive impairment and its risk factors among 55 years old and above residents in Shanghai. J Clin Psychiatry 24, 155–158.

35.

Song

, Chen

, He

(2012) Investigation on correlation between prevalence of the mild cognitive impairment and eating habit in elderly in the communities of Shunde-city. Int Med Health Guid News 18, 1715–1718.

36.

, Hua

, Zhang

, LI

, Chen

, Zhang

(2013) Influencing factors of mild cognitive impairment of seniors in communities of Xi’an. J Nurs 20, 6–9.

37.

Tang

, Zhang

, Wu

, Liu

, Diao

, Guan

, et al. (2007) Prevalence of the mild cognitive impairment among elderly in Beijing. Chin Ment Health J 21, 116–118.

38.

Tong

, Guo

, Tao

, Guo

, Tian

, Xia

, et al. (2013) The elderly patients with mild cognitive impairment in the community of Tangshan. Chin J Helth Psychol 21, 1642–1644.

39.

Yin

, Nie

, Xu

(2011) The prevalence and risk factors of mild cognitive impairment among the aged in Huzhou. Chin Gen Pract 14, 4145–4147.

40.

Zhu

, Li

(2015) Analysis of influencing factors of mild cognitive impairment in elderly population. Chin J Pract Nerv Dis 18, 52–54.

41.

Zhang

(2011) The study on the present situation investigating risk factors and early intervention of mild cognitive impairment in the elderly, MSc Thesis. Soochow University.

42.

Wang

(2012) A study on the prevalence and risk factors of mild cognitive impairment among the elderly in Tianjin community, MSc Thesis. Tianjin Medical University.

43.

, Ma

, Zhang

, Liang

, Chen

, Wang

, Zhang

, Wang

, Beijing >Ageing Brain Rejuvenation Initiative (2013) Prevalence of and potential risk factors for mild cognitive impairment in community-dwelling residents of Beijing. J Am Geriatr Soc 61, 2111–2119.

44.

Jia

, Zhou

, Wei

, Jia

, Wang

, Li

, Wu

, Mok

, Gauthier

, Tang

, Chu

, Zhou

, Cui

, Wang

, Yin

, Hu

, Zuo

, Song

, Qin

, Wu

, Li

, Jia

, Song

, Han

, Xing

, Yang

, Li

, Qiao

, Tang

, Lv

, Dong

(2014) The prevalence of mild cognitive impairment and its etiological subtypes in elderly Chinese. Alzheimers Dement 10, 439–447.

45.

Ding

, Zhao

, Guo

, Meng

, Wang

, Luo

, Mortimer

, Borenstein

, Hong

(2015) Prevalence of mild cognitive impairment in an urban community in China: A cross-sectional analysis of the Shanghai Aging Study. Alzheimers Dement 11, 300–309 e302.

46.

Janelidze

, Mikeladze

, Bochorishvili

, Dzagnidze

, Kapianidze

, Mikava

, Khatiashvili

, Mirvelashvili

, Shiukashvili

, Lynch

, Nadareishvili

(2018) Mild cognitive impairment in Republic of Georgia. Gerontol Geriatr Med 4, 1–10.

47.

Lee

, Shahar

, Chin

, Mohd Yusoff

, Rajab

, Aziz

(2012) Prevalence of gender disparities and predictors affecting the occurrence of mild cognitive impairment (MCI). Arch Gerontol Geriatr 54, 185–191.

48.

, Wu

, Zhao

, Ji

, Song

, Zhang

, Huang

(2016) Prevalence of mild cognitive impairment and its subtypes among Chinese older adults: Role of vascular risk factors. Dement Geriatr Cogn Disord 41, 261–272.

49.

Paddick

, Kisoli

, Samuel

, Higginson

, Gray

, Dotchin

, Longdon

, Teodorczuk

, Chaote

, Walker

(2015) Mild cognitive impairment in rural Tanzania: Prevalence, profile, and outcomes at 4-year follow-up. Am J Geriatr Psychiatry 23, 950–959.

50.

Pilleron

, Jesus

, Desport

, Mbelesso

, Ndamba-Bandzouzi

, Clement

, Dartigues

, Preux

, Guerchet

(2015) Association between mild cognitive impairment and dementia and undernutrition among elderly people in Central Africa: Some results from the EPIDEMCA (Epidemiology of Dementia in Central Africa) programme. Br J Nutr 114, 306–315.

51.

Rao

, Luo

, Tang

, Shen

, Huang

, Yu

, Ren

, Cheng

, Lin

(2018) Prevalence of mild cognitive impairment and its subtypes in community-dwelling residents aged 65 years or older in Guangzhou, China. Arch Gerontol Geriatr 75, 70–75.

52.

, Shang

, Xu

, Li

, Chen

, Zhang

, Hua

(2014) Prevalence and predictors of mild cognitive impairment in Xi’an: A community-based study among the elders. Plos One 9, e83217.

53.

Vanoh

, Shahar

, Din

, Omar

, Vyrn

, Razali

, Ibrahim

, Hamid

(2016) Predictors of poor cognitive status among older Malaysian adults: Baseline findings from the LRGS TUA cohort study. Aging Clin Exp Res 29, 173–182.

54.

Fang

, Chen

, LJ

(2009) The prevalence rate of mild cognitive impairment and its related factors in community elderly in Hangzhou. Chin J Ger 28, 512–515.

55.

Tsoy

, Turuspekova

, Klipitskaya

, Mereke

, Cumming

(2019) Prevalence of mild cognitive impairment among older people in Kazakhstan and potential risk factors: A cross-sectional study. Alzheimer Dis Assoc Disord 33, 136–141.

56.

Fang

, Sheng

(2015) Investigation of the elderly with mild cognitive impairment in Shanghai Zhoujiaqiao Community. Med Health Care 23, 5–8.

57.

Sun

, Qu

, Liu

, Zhang

, Guo

, JIa

, et al. (2016) The epidemiological study of ethnic disparity in risk factors for mild cognitive impairment between Mongolia and Han population in Baotou, Inner Mongolia. J Apoplexy Nerv Dis 33, 454–456.

58.

Zhang

, Jiang

, Wang

, Gao

(2013) Investigation and Analysis on mild cognitive impaiment among the elderly in the communities of Taicang city. Pract Geriatr 27, 859–862.

59.

Brucki

SMD

, Nitrini

(2014) Cognitive impairment in individuals with low educational level and homogeneous sociocultural background. Dement Neuropsychol 8, 345–350.

60.

Cesar

, Brucki

SMD

, Takada

, Nascimento

LFC

, Gomes

CMS

, Almeida

MCS

, Oliveira

, Porto

FHG

, Senaha

MLH

, Bahia

, Silva

TBL

, Ianof

, Spindola

, Schmidt

, Jorge

, Vale

PHF

, Cecchini

, Cassimiro

, Soares

, Goncalves

, Martins

ACS

, Dare

, Smid

, Porto

, Carthery-Goulart

, Yassuda

, Mansur

, Nitrini

(2016) Prevalence of cognitive impairment without dementia and dementia in Tremembe, Brazil. Alzheimer Dis Assoc Disord 30, 264–271.

61.

Dominguez

, Fe de Guzman

, Reandelar

, Thi Phung

(2018) Prevalence of dementia and associated risk factors: A population-based study in the Philippines. J Alzheimers Dis 63, 1065–1073.

62.

Fei

, Qu

, Wang

, Yin

, Bai

, Ding

(2009) Prevalence and distribution of cognitive impairment no dementia (CIND) among the aged population and the analysis of socio-demographic characteristics: The community-based cross-sectional study. Alzheimer Dis Assoc Disord 23, 130–138.

63.

Guerchet

, Houinato

, Paraiso

, von Ahsen

, Nubukpo

, Otto

, Clement

, Preux

, Dartigues

(2009) Cognitive impairment and dementia in elderly people living in rural Benin, west Africa. Dement Geriatr Cogn Disord 27, 34–41.

64.

Guerchet

, M’Belesso

, Mouanga

, Bandzouzi

, Tabo

, Houinato

, Paraiso

, Cowppli-Bony

, Nubukpo

, Aboyans

, Clement

, Dartigues

, Preux

(2010) Prevalence of dementia in elderly living in two cities of Central Africa: The EDAC survey. Dement Geriatr Cogn Disord 30, 261–268.

65.

Mejia-Arango

, Gutierrez

(2011) Prevalence and incidence rates of dementia and cognitive impairment no dementia in the Mexican population: Data from the Mexican Health and Aging Study. J Aging Health 23, 1054–1074.

66.

Shi

, Zhang

, Yue

, Liu

, Huo

, Liu

, Xiang

, Liu

, Lu

, Wang

, Ji

(2013) Prevalence and clinical predictors of cognitive impairment in individuals aged 80 years and older in rural China. Dement Geriatr Cogn Disord 36, 171–178.

67.

Zhang

, Shi

, Liu

, Yue

, Liu

, Xiang

, Lu

, Liu

, Wisniewski

, Wang

, Ji

(2014) Prevalence of cognitive impairment no dementia in a rural area of Northern China. Neuroepidemiology 42, 197–203.

68.

Lei

, Huang

, Yang

, Gao

, Wei

, et al. (2008) Prevalence of mild cognitive impairment old people in urban and rural areas of Guizhou province. Chin Ment Health J 22, 387–391.

69.

Inocian

, Patalagsa

(2016) Cognitive impairment in older people living in the community. Nurs Older People 28, 25–30.

70.

Khedr

, Fawi

, Abbas

MAA

, Mohammed

, El-Fetoh

, Al Attar

, Noaman

, Zaki

(2015) Prevalence of mild cognitive impairment and dementia among the elderly population of qena governorate, upper Egypt: A community-based study. J Alzheimers Dis 45, 117–126.

71.

Ding

, Zhao

, Guo

, Meng

, Wang

, Luo

, Mortimer

, Borenstein

, Hong

(2015) Prevalence of mild cognitive impairment in an urban community in China: A cross-sectional analysis of the Shanghai Aging Study. Alzheimers Dement 11, 300–309.e302.

72.

Guo

, Zhao

, Li

, Q

(2013) Prevalence of mild cognitive impairment among rural Chinese elderly. Chin J Mult Organ Dis Elderly 12, 904–907.

73.

Huang

, Tang

, Ma

, Guo

, Han

, Huang

, et al. (2008) The prevalence of mild cognitive impairment of residents aged 60 years and over in the urban and rural areas in Guangdong. Chin J Nerv Ment Dis 34, 533–537.

74.

(2013) Survey of mild cognitive impairment in old people in community of Jinan city. Chin Nurs Res 27, 2196–2197.

75.

Wang

, Zhuang

, Lin

, Ding

, Wang

(2014) The study on present situation investigation of mild cognitive impairment among elderly people in Zhoushan Island city. Zhejiang Med J 8, 707–709.

76.

, Zhang

, Su

, Dang

, Hou

(2012) Investigation on mild cognitive impairment among elderly in urban community of Xi’an. Chin J Rehabil Theory Pract 18, 605–607.

77.

Zhang

, Li

, Tian

, Wang

, Liang

, Liu

(2015) The value of MoCA (Beijing) in screening mild cognitive impairment among old people in rural. Chin J Gerontol 35, 4016–4018.

78.

Zhang

, Zeng

(2014) Prevalence and factors associated with mild cognitive impairment among the elderly in Changsha Communities. Chin Gen Pract 17, 1031–1035.

79.

Zhao

, Gao

, Fang

(2015) Present situation and risk factors of mild cognitive impairment of 60-years old or over people in Jilin area. Chin Rural Health Serv Adm 35, 1434–1437.

80.

Zhou

, Chen

, Sui

, Yang

(2016) Investigation on the prevalence of mild cognitive impairment in the elderly in Changji city. Chin Community Doct 32, 178–183.

81.

Zhou

, Xu

, Chen

(2011) Prevalence of mild cognitive impairment among the elderly. Chin J Public Health 27, 1375–1377.

82.

Zhou

, Zhu

, Barhematy

, Yue

, Zhao

, Xing

, et al. (2009) Cross-sectional study of the mild cognitive impairment among elderly in Xinjiang Uygur and Han ethnic groups. Chin J Geriatr 28, 865–869.

83.

Zhu

, Zhou

, Barheimaty

, Yue

, Zhao

, Xing

, et al. (2009) Study of prevalence of the mild cognitive impairment among elderly in the communities of Urumqi city. J Xinjiang Med Univ 32, 578–584.

84.

Zhu

, Chen

, Zhong

(2013) A prevalence study on mild cognitive impairment among elderly population in Zhejiang providence. Chin J Epidemiol 34, 475–477.

85.

Mohan

, Iype

, Varghese

, Usha

, Mohan

(2019) A cross-sectional study to assess prevalence and factors associated with mild cognitive impairment among older adults in an urban area of Kerala, South India. BMJ Open 9, e025473.

86.

, Wang

X-D

, Shi

, Yue

, Zhang

, Liu

, Zhao

, Xiang

, Zhang

, Guan

, Su

, Li

, Wang

, Wisniewski

, Ji

(2019) Comparative analysis of cognitive impairment prevalence and its etiological subtypes in a rural area of northern China between 2010 and 2015. Sci Rep 9, 851.

87.

Ruan

, Xiao

, Gong

, Zhang

, Ruan

, Zhang

, Chen

, Yu

(2020) Prevalence of cognitive frailty phenotypes and associated factors in a community-dwelling elderly population. J Nutr Health Aging 24, 172–180.

88.

Dimitrov

, Tzourio

, Milanov

, Deleva

, Traykov

(2012) Prevalence of dementia and mild cognitive impairment in a Bulgarian urban population. Am J Alzheimers Dis Other Demen 27, 131–135.

89.

Chu

, Liang

, Chen

, Gu

, Qian

(2015) Prevalence and risk factors of mild cognitive impairment and dementia in community elderly. Chin J Clin Neuro 23, 673–677.

90.

Xiong

, Miao

, Wang

, Zhou

, Gao

, Ma

(2013) Prevalence and influencing factors of MCI among community elderly in Tianjin city. Chin J Public Health 29, 1–4.

91.

Das

, Bose

, Biswas

, Dutt

, Banerjee

, Hazra

, Raut

, Chaudhuri

, Roy

(2007) An epidemiologic study of mild cognitive impairment in Kolkata, India. Neurology 68, 2019–2026.

92.

Saha

, Sanyal

, Bhattacharyya

, Barman

, Mukherjee

(2010) A study on cognitive status of 50 years and above aged non-demented women in a rural area of West Bengal. J Indian Med Assoc 108, 726–729.

93.

Koyanagi

, Veronese

, Stubbs

, Vancampfort

, Stickley

, Oh

, Shin

, Jackson

, Smith

, Lara

(2019) Food insecurity is associated with mild cognitive impairment among middle-aged and older adults in South Africa: Findings from a nationally representative survey. Nutrients 11, 749.

94.

Wang

, Liu

, Huang

, Ding

, Ma

(2016) The status of mild cognitive impairment in the Hui and Han people aged 55 years and above in Ningxia. Chin J Gerantol 36, 4601–4603.

95.

Wang

, Ding

, Liu

, Li

, Ma

, Zhang

(2013) The present situation of mild cognitive impairment among 55-years old or over individuals of Hui and Han nationalities and its relationship with hormone. Chin J Nerv Ment Dis 39, 427–430.

96.

, Zhao

, Wang

, Wen

, Niu

, et al. (2012) A prevalence study on mild cognitive impairment among elderly population of Mongolian and Han nationalities in a pastrial area of Inner Mongolia. Chin J Epidemiol 33, 364–367.

97.

Petersen

(2004) Mild cognitive impairment as a diagnostic entity. J Intern Med 256, 183–194.

98.

Petersen

, Caracciolo

, Brayne

, Gauthier

, Jelic

, Fratiglioni

(2014) Mild cognitive impairment: A concept in evolution. J Intern Med 275, 214–228.

99.

Petersen

, Doody

, Kurz

, Mohs

, Morris

, Rabins

, Ritchie

, Rossor

, Thal

, Winblad

(2001) Current concepts in mild cognitive impairment. Arch Neurol 58, 1985–1992.

100.

Petersen

, Morris

(2005) Mild cognitive impairment as a clinical entity and treatment target. Arch Neurol 62, 1160–1163.

101.

Petersen

, Smith

, Waring

, Ivnik

, Kokmen

, Tangelos

(1997) Aging, memory, and mild cognitive impairment. Int Psychogeriatr 9, 65–69.

102.

Petersen

, Stevens

, Ganguli

, Tangalos

, Cummings

, DeKosky

(2001) Practice parameter: Early detection of dementia: Mild cognitive impairment (an evidence-based review). Report of the Quality Standards Subcommittee of the American Academy of Neurology. Neurology 56, 1133–1142.

103.

Winblad

, Palmer

, Kivipelto

, Jelic

, Fratiglioni

, Wahlund

, Nordberg

, Bäckman

, Albert

, Almkvist

(2004) Mild cognitive impairment-beyond controversies, towards a consensus: Report of the International Working Group on Mild Cognitive Impairment. J Intern Med 256, 240–246.

104.

Bell

(1994) DSM-IV: Diagnostic and Statistical Manual of Mental Disorders. JAMA 272, 828–829.

105.

Portet

, Ousset

, Visser

, Frisoni

, Nobili

, Scheltens

, Vellas

, Touchon

(2006) Mild cognitive impairment (MCI) in medical practice: A critical review of the concept and new diagnostic procedure. Report of the MCI Working Group of the European Consortium on Alzheimer’s Disease. J Neurol Neurosurg Psychiatry 77, 714–718.

106.

Vancampfort

, Stubbs

, Firth

, Smith

, Swinnen

, Koyanagi

(2019) Associations between handgrip strength and mild cognitive impairment in middle-aged and older adults in six low- and middle-income countries. Int J Geriatr Psychiatry 34, 609–616.

107.

Chong

, Shahar

, Haron

, Din

(2019) Habitual sugar intake and cognitive impairment among multi-ethnic Malaysian older adults. Clin Interv Aging 14, 1331–1342.

108.

Zhang

, Guan

, Shi

, Yue

, Liu

, Lu

, Zhao

, Zhang

, Su

, Ji

(2019) Sex differences in the prevalence of and risk factors for cognitive impairment no dementia among the elderly in a rural area of Northern China: A population-based cross-sectional study. Neuroepidemiology 52, 25–31.

109.

Wang

, Song

, Yu

, Wang

, An

, Xun

, Zhao

, Gao

, Wang

(2015) Mild cognitive impairment: Vascular risk factors in community elderly in four cities of Hebei Province, China. PLoS One 10, e0124566.

110.

Stephan

, Matthews

, McKeith

, Bond

, Brayne

, Medical Research Council Cognitive Function and Aging Study (2007) Early cognitive change in the general population: How do different definitions work? J Am Geriatr Soc 55, 1534–1540.

111.

Pessoa

RMP

, Bomfim

AJL

, Ferreira

BLC

, Chagas

MHN

(2019) Diagnostic criteria and prevalence of mild cognitive impairment in older adults living in the community: A systematic review and meta-analysis. Rev Psiquiatr Clí 46, 72–79.

112.

Patterson

, World Alzheimer Report 2018 - The state of the art of dementia research: New frontiers, https://www.alz.co.uk/research/WorldAlzheimerReport2018.pdf Accessed July 25, 2020.

113.

Matthews

, Stephan

, McKeith

, Bond

, Brayne

, Medical Research Council Cognitive Function and Ageing Study (2008) Two-year progression from mild cognitive impairment to dementia: To what extent do different definitions agree? J Am Geriatr Soc 56, 1424–1433.

114.

Jia

, Zhou

, Wei

, Jia

, Wang

, Li

, et al. (2014) The prevalence of mild cognitive impairment and its etiological subtypes in elderly Chinese. Alzheimers Dement 10, 439–447.