Table 1 Study characteristics

Cross-Sectional

 Alpass et al., 2004 [27]

232

1.29 (3/232)

53.8 to 95.2

Community- dwelling

MMSE

Hierarchical regression analysis

Age, education, income, social network, depression

 Bourgeois et al., 2020 [23]

359

15%

 > 50

Institutionalized (HIV clinic)

MoCA

Poisson regression analysis

Age, sex, education, income, marital status, ethnicity, tobacco use, employment

 Bourne et al., 2007 [28]

266

50 (134/266)

64

Community- dwelling

MHT (in 1947); Raven’s standard progressive matrices (at age 64)

Bivariate correlation

Sex

 Conroy et al., 2010 [29]

802

53% (423/802)

65 to 102

Community- dwelling

AMT

Multinomial odds ratio

Education, social network, marital status, loneliness, depression

 Deng & Liu, 2021 [30]

10,556

55.26%

 ≥ 65 years

Community- dwelling and institutionalized

Chinese-MMSE

Multivariate logistic regression

Age, sex, education, income, marital status, area of residence

 Ficker et al., 2002 [24]

194

71.60%

NR

Institutionalized

MDRS

Independent t-test

Race, education

 Frith & Loprinzi, 2017 [31]

1874

59.10%

60 to 85

Community- dwelling

WAIS: DSST

Weighted multivariable regression

Age, sex, race

 Ge et al., 2017 [32]

3159

58.90%

60 to 105

Community- dwelling

Chinese-MMSE, EBMT, SDMT, Digit Span Backwards Test

Linear regression analysis

Education, income, social network, marital status, ADLs, depression

 Gow et al., 2007 [33]

488

58%

NR

Community- dwelling

MHT—raw MHT scores converted into IQ scores

Regression analysis

Age, education, income, marital status, loneliness

 Gow et al., 2013 [34]

1091

NR

NR (mean age 70)

Community- dwelling

WAIS-III UK and WMS-III UK, tests of reaction and inspection time

Ancova

Social network, marital status, loneliness, depression

 Hamalainen et al., 2019 [35]

30, 029

50.90%

45 to 85

Community- dwelling

Mental Alternation Test, Animal Fluency test, Controlled Oral Word Association Test, Stroop test, RAVLT with immediate and 5-min recall

Multiple regression analysis

Age, sex, race, education, income

 Harling et al., 2020 [36]

5059

53.6%

 ≥ 40

Community- dwelling

Battery from Health and Retirement study (orientation in time, episodic memory, number patterns)

Poisson regression; linear regression

Age, sex, education, income, marital status, country of origin, self-reported literacy, self-rated childhood health, father’s occupation, household size, employment status

 Henderson et al., 1986 [37]

274

NR

70–79

80 + 

Community- dwelling

GMS; MMSE

Mancova

Age, sex, marital status

 Holtzman et al., 2004 [38]

354

68.60%

50 to 81

Community- dwelling

MMSE

Simultaneous linear or logistic

Regression

Age, sex, race, education, depression

 Jang et al., 2020 [39]

2061

66.8%

 ≥ 60

Community- dwelling

Korean-MMSE; self-rated cognitive health

Bivariate regression; hierarchical linear regression

Age, sex, education, social network, marital status, depression, chronic conditions (functional disability, chronic disease), tobacco use, alcohol use

Keller- Cohen et al., 2006 [40]

20

15/20

85–93

Independent living in retirement community

composite cognistat; BNT

Hierarchical multiple

Regression; bivariate correlation

Education

 Kim et al., 2019 [41]

410

252/410

60 + 

Community- dwelling

VMS; CERAD-TS; MMSE

One-way ANCOVA

Age, sex, education, depression

 Kotwal et al., 2016 [42]

3310

52%

62 to 90

Community- dwelling

MoCA-SA

Multiple linear regression

Age, sex, race, education, marital status, depression

 Krueger et al., 2009 [43]

838

75(NR/883)

NR

Subsidized housing facilities and continuous care retirement communities

Episodic: Word List Memory, Recall, and Recognition; WMS; Semantic: BNT, National Adult Reading Test; Working: Digit Span Forward and Backward, Digit Ordering

Linear regression analysis

Age, sex, education, depression

 La Fleur & Salthouse, 2017 [44]

2613

18–39: 66

40–59: 72

60–96: 63

NR

ND

Logical memory task; free recall task; paired associates’ task; Letter sets task; Shipley’s Abstraction; matrix reasoning

Multiple regression

Age, sex, education

 Lee & Waite, 2018 [45]

2260

52.05

57–85

Community–dwelling

MoCA-SA

Multivariate regression

Age, sex, race, education

 Mehrabi & Béland, 2021 [46]

1643

50.2%

 ≥ years

Community–dwelling

MoCA

Regression

Age, sex, education, income, smoking, alcohol consumption, sleeping disturbance

 Millán-Calenti et al., 2013 [47]

579

57.2

 ≥ 65 years

Community–dwelling residents

MMSE; The Geriatric Depression Scale-Short Form

Multinomial logistic regression

Age, sex, education, ADLs

 Murayama et al., 2019 [48]

897

50 (450/897)

 ≥ 65 years

Community- residents

MMSE-J

Multilevel logistic regression

Age, sex, education, income, social network, marital status

 Nakamura et al., 2019 [49]

331

100 (331/331)

 ≥ 65 years

ND

BOMC

Unadjusted bivariate analysis

Age, education, ADLs, depression

 Okabayashiet al., 2004 [50]

1976

NR

 ≥ 65 years

ND

Japanese-SPMSQ

Regression (unspecified)

Age, sex, education, depression

 Oremus et al., 2019 [6]

21,241

51%

45–85

ND

RAVLT, Animal Fluency Test, Mental Alternation Test

Rao-Scott chi square

Age, region of residence, urban / rural residence, education

 Oremus et al., 2020 [7]

21,241

51.3 (10,835/21241)

45–85

ND

RAVLT

Multiple linear regression

Age, sex, education, income, marital status, ADLs, depression

 Pillemer & Holtzer, 2016 [51]

355

55.2 (196/355)

65.00- 95.00

Community- dwelling

RBANS

Linear regression analysis

Age, sex, education, depression

 Poey et al., 2017 [52]

779

58 (452/779)

70–110 years

ND

Diagnosis of normal cognition, CIND, AD, and non-AD dementia

Multiple logistic regression

Age, sex, race, education, depression

 Rashid et al., 2016 [53]

2005

68 (1363/2005)

60–99

Community- dwelling

ECAG

Regression Analyses

Age, sex, race, education, social network, marital status

 Saenz et al., 2020 [54]

4,017 (married dyads)

50% (4017/8034)

50 + 

Community- dwelling

Cross-Cultural Cognitive Examination

Regression analysis

Age, sex, education, income, ADLs, depression

 Sims et al., 2014 [55]

175

45%

54–83

Community- dwelling

Stroop Color-Word Test, Judgment of Line Orientation; WAIS-R: The Block Design subscale, Digit Span Forward, Digit Span Backward; WMS: Logical Memory I and II Visual Reproductions I and II; The Grooved Pegboard, TMT

Multiple regression

Age, sex, race, education, depression

 Weng et al., 2020 [56]

1706

53.01%

 ≥ 45 years

Community- dwelling

Subjective cognitive decline

Univariate and multiple logistic regression

Age, sex, race, education, marital status, depression, chronic conditions (coronary heart disease, diabetes), exercise, employment status

 Yang et al., 2020 [57]

470

52.6%

 ≥ 65 years

Community- dwelling

CDR; MMSE

Multiple linear regression

Age, sex, education, income, ADLs, depression, functional assessment questionnaire, neuropsychiatric inventory questionnaire (nighttime behaviors, irritability, apathy, motor disturbances)

 Yeh & Liu, 2003 [58]

4993

46.67% ( 2330/4989)

65 + 

Community- dwelling

SPMSQ

Multiple regression analysis

Sex, education, marital status, loneliness, ADLs

 Zahodne et al., 2014 [59]

482

54.1

55–85

Community- dwelling

NIH Toolbox Cognition module: Dimensional Change Card Sort, Flanker, List Sorting, Pattern Comparison, Picture Sequence Memory

Regression analysis

Race, education, loneliness

 Zahodne et al., 2018 [60]

548

62.6

ND

Community- dwelling

NIH Toolbox Cognition module: Dimensional Change Card Sort, Flanker, List Sorting, Pattern Comparison, Picture Sequence Memory, Selective Reminding Test. Language scores, Benton Visual Retention Test, the Rosen Drawing Test, and the Identities and Oddities subtest of the DRW

Multiple regression

Race, education, income

 Zank & Leipold, 2001 [61]

63

76%

53–96

Geriatric day care units

MMSE

Hierarchical regression analysis

Education, marital status

 Zhaoyang et al., 2021 [62]

311

67%

70–90

Community- dwelling

5 cognitive domains (memory, executive function, attention, language, visual-spatial) with 10 neuropsychological instruments

Multilevel Poisson and logistic models

Age, sex, race, education, employment, marital status, living status

 Zhu et al., 2012 [63]

120

37.50%

60–86

Community- dwelling

MMSE

Multiple regression analysis

Age, sex, education, income, social network, marital status,

 Zuelsdorff et al., 2013 [64]

623

71%

40–73

Community- dwelling

RAVLT, Digits Forward, Digits Backward; WAIS-III: Letter-Number Sequence subtests; TMT, and Stroop Color-Word

Regression analysis

Age, sex, education, social network, marital status

 Zuelsdorff et al., 2019 [65]

1052

69%

40–78

Community- dwelling

RAVLT; BVMT-R; WAIS-R: Logical Memory immediate and delayed recall subtests; TMT, Stroop; Color-Word Interference condition; WAIS: Digit Span Forward, Digit Span Backward, and Letter-Number Sequencing

Regression analysis

Age, sex, race, education, social network, marital status, ADLs

 Zullo et al., 2021 [66]

1567

58.65%

 ≥ 65 years

Community- dwelling

Questionnaire de la Plaite Cognitive (QPC)

Binary logistic regression

Age, sex, depression, personality dimensions, quality of life, professional activity, interaction term between neuroticism and quality of life

Cohort

 Amieva et al., 2010 [67]

Study sample size = 3777, Analytic/included sample: 2089

59.9% (1251/2089)

ND

Community- dwelling

AD / Dementia diagnosis; MMSE; NINCDS-ADRDA criteria for AD

Multivariate analysis

Sex, education, social network, ADLs, Diabetes, CVD

 Andel et al., 2012 [68]

10,106

52%

ND

Community- dwelling

Dementia diagnosis using DSM-4 criteria

Regression analysis

Age, sex, education, vascular disease

 Bedard & Taler, 2020 [69]

11,152 (440 cases, 10,712 controls)

Controls: 55.3%

Cases: 42.1 – 44.9%

45–85

NR

Animal Fluency Test, controlled oral word association test, mental alternation test, and Victoria Stroop test, Ray auditory verbal learning test, Miami prospective memory test

Binary logistic regression

Age, sex, education, marital status, depression, testing language

 Bowling et al., 2016 [70]

9119

50.69% (4622/9119)

ND

Community- dwelling

Reading and comprehension test, arithmetic test, copying design test, general ability test

Multiple linear regression

Sex, education, social network, marital status

 Camozzato et al., 2015 [71]

220

70%

ND

Community- dwelling

DSM5 and NINCDS-ADRDA criteria

Multivariate cox proportional- hazards moel

Age, sex, education, income, marital status, ADLs

 Chen & Chang, 2016 [72]

2300

44.87%

65–93

Community- dwelling

SPMSQ; Chinese-MMSE

Multinomial logistic regression

Age, sex, education, ADLs, hypertension, diabetes, heart disease, stroke

 Chen & Zhou, 2020 [73]

16, 786

NR

 ≥ 65 years

Community- dwelling

Chinese-MMSE

Generalized structural equation modelling (GSEM)

Age, sex, education, marital status, cardiometabolic diseases (diabetes, cardiovascular, stroke, heart disease), residence

 Crooks et al., 2008 [25]

initial = 2249

100%

ND

Institutionalized

Telephone Interview for Cognitive Status; Telephone Dementia Questionnaire

Cox proportional hazards

Age, sex, education, social network, marital status, depression, stroke, myocardial infarction, diabetes, hypertension, PD

 Dickinson et al., 2011 [74]

213

63.85%

Community- dwelling

CERAD; WMS-R; Logical Memory subtest; TMT, SDMT; WAIS-R: Digit Span Forward; ascending Digit Span task modeled after the Digit Ordering Test

Linear regression models

Age, sex, education, social network

 Eisele et al., 2012 [75]

2367 (1869 = analytic sample)

65.90%

79–95

Community- dwelling

SIDAM

Multifactorial ANCOVA

Age, sex, education, marital status, ADLs, hypertension, CVD, coronary heart disease, alcohol use, BMI

 Ellwardt et al., 2013 [76]

2255

54.00%

55–85

Community- dwelling

MMSE; coding task, and Reven's Colored Progressive Matrices

Latent growth mediation model

Age, sex, education, loneliness, ADLs

 Heser et al., 2014 [77]

2300

ND

ND

Community- dwelling

SIDAM

Proportional hazard models, cox regression analysis

Age, sex, education, ADLs

 Holtzman et al., 2004 [38]

354

68.60%

50–81

Community–dwelling

MMSE

Simultaneous linear regression

Age, sex, race, education, social network

Howrey et al., 2015 [78]

2767

58.29%

ND

Community- dwelling

MMSE

Multivariate analyses by using simultaneous linear or logistic regression

Age, sex, education, income, marital status, ADLs, hypertension, heart attack, stroke, diabetes, vision, Nativity, BMI

 Hudetz et al., 2010 [26]

80

0%

55–85

Institutionalized

RBANS: Story Memory and Word List Memory subtests; BVMT-R

Stepwise multiple regression analysis

Age, sex, education, hypertension, hypercholesterolemia, angina, myocardial infarction, type 2 diabetes

 Hughes et al., 2008 [79]

at baseline = 417, analytic = 217

51.80%

ND

Community- dwelling

MMSE; Stroop test, TMT, Hopkins verbal learning tests

Random effects model

Age, sex, education, social network, marital status

 Kats et al., 2016 [80]

13,782

ND

48–64

Community- dwelling

DSST, DWRT, WFT

Generalized linear models

Age, sex, race, education, social network

 Khondoker et al., 2017 [81]

10,055

46%

ND

Community- dwelling

The short-form IQCODE questionnaire and physician

Proportional hazard regression models

Age, sex, education, income, diabetes, CVD, stroke, hypertension, cancer

 Khoo & Yang, 2020 [82]

1735

NR

40–70

NR

Brief Test of Adult Cognition by Telephone (BTACT)

Structural equation modelling

Age, sex, education, income, general health

 Liao et al., 2018 [83]

6,863

29.20%

ND

Community- dwelling

Alice Heim 4-I test (AH4-I), an inductive reasoning test, and two tests of verbal fluency

Bivariate dual change score model; goodness of fit

Age, sex, race, education, income, marital status, coronary heart disease, stroke, diabetes, cancer, depressive symptoms

 Liao & Scholes, 2017 [84]

10,241

53.30%

ND

Community- dwelling

Verbal fluency and letter cancellation task

Linear mixed model

Age, sex, education, income, ADLs

 Liu et al., 2020 [85]

13, 636

55%

 ≥ 65 years

Community- dwelling

Dementia Scale (Degree of Independence in Daily Living for Older Adults with Dementia)

Multivariate adjusted Cox proportional hazards model

Age, sex, education, history of disease (stroke, hypertension, myocardial infarction, diabetes, cancer), smoking, alcohol drinking, BMI, time spent walking per day, psychological distress score, motor function score, social participation

 Luo et al., 2021 [86]

497

48%

64–68

NR

Subtest of verbal comprehension index in German WAIS-R; verbal fluency and vocabulary; subtest of perceptual reasoning index in WAIS-R

Mplus8

NR

 Miyaguni et al., 2021 [87]

15, 313

51.80%

 ≥ 65 years

Community- dwelling

I to IV and Medical, I (= 22 on MMSE), II (= 16), III (= 13), IV (= 6)

Multilevel survival analyses with sensitivity analyses model

Age, sex, education, marital status, depression, living conditions, present illness, smoking status, alcohol consumption, individual social support

 Moreno et al., 2022 [88]

2242

100%

65–83

NR

Primary Mental Abilities Vocabulary Test; Category Fluency Test; Letter Fluency Test; Benton Visual Retention Test; California Verbal Learning Test; California Verbal Learning Test; Digit Span Test; Card Rotation Test

Linear mixed models with covariate adjustment

Age, race, education, income, region, job classification, major medical comorbidities

 Murata et al., 2019 [16]

14,088

50.97%

65–99

Community- dwelling

Incident dementia ascertained upon eligibility for Japan’s public

LTCI system, Level II or higher, on the index for the evaluation of care needs for people with dementia

Cox proportional hazard models

Age, sex, education, marital status, health behaviors (alcohol, smoking daily physical activity), cognitive complaints to predict dementia, depression

 Noguchi et al., 2019 [89]

121 (analytic sample)

47.10%

ND

Community- dwelling

Japanese MoCA

Multivariable

Linear regression analysis

Age, sex, income, ADLs, stroke, hypertension, dyslipidemia, diabetes, depression, living alone, BMI

 Okely et al., 2021 [90]

70–84

Community- dwelling

5 questions about current state of participants’ memory

Spearman’s rho

Age, sex, education, depression, diabetes, cardiovascular disease, occupational social class, personality, living situation, anxiety, older age fluid cognitive ability

 Pais et al., 2021 [91]

341

57.5%

60–85

Community- dwelling

MMSE

Multivariable Cox analysis of social support on cognitive impairment (hazard ratio)

Age, sex, social network, marital status,

 Pillemer et al., 2019 [17]

493

57.20%

65–95

Community- dwelling

RBANS

Cox proportional hazard ratio

Sex, race, education, diabetes, chronic heart failure, arthritis, hypertension, depression, stroke, Parkinson’s disease, chronic

obstructive lung disease, angina, myocardial infarction, depressive symptoms

 Riddle et al., 2015 [92]

299

normal = 59.43%, MCI = 57.89%, dementia = 70.83%

ND

Community–dwelling

Neuropsychological battery to detect incident dementia or cognitive impairment

Χ2 for categorical variables and ANOVA, logistic regression models

Age, sex, race, education, ADLs, depression

 Rote et al., 2021 [93]

2880

57.7%

 ≥ 65 years

Community- dwelling

MMSE

Logistic regression

Age, sex, country of birth (Mexico or USA), Medicaid (yes or no)

 Saito et al., 2018 [94]

13,984

50.90%

ND

Community- dwelling

Long-term Care Insurance, The Degree of Autonomy in the Daily Lives of Elderly

Individuals with Dementia Scale

Cox proportional hazard models

Age, sex, education, income, social network, marital status, ADLs, stroke, diabetes, depression, SCI, physical activity

 Salinas et al., 2017 [95]

1834 (for dementia analysis)

44%

Community- dwelling

DSM-IV

Cox proportional hazard models

Age, sex, education, social network, marital status, atrial fibrillation, diabetes, CVD, smoking status, depression, physical activity, antihypertensive treatment

 Seeman et al., 2001 [96]

1189

55.20%

70–79

Community- dwelling

BNT; WAIS-R

Multivariate linear regression

Age, sex, race, education, income, social network, marital status, physical activity

 Sörman et al., 2015 [97]

1715

No Dementia: 53.3%

all cause dementia: 65.1%

AD: 73.9%

Community- dwelling

DSM-IV

Multivariate linear regression

Age, sex, education, CVD, stroke, HBP, diabetes, alcohol use, smoking status, obesity, stress, depression

 Thomas & Umberson, 2018 [98]

2,788

64.70%

60–95

Community- dwelling

SPMSQ

Estimated growth curve models within a mixed-model framework Intercept (SE), Linear Slope

Age, sex, race, education, income, marital status, number of children, stressful life events

 Wilson et al., 2015 [99]

529

78.90%

Institutionalized and community- dwelling

Clinical classification of MCI

Proportional hazards model

Age, sex, education, social network, loneliness, depression, negative life events

 Yin et al., 2020 [100]

5897

51%

 ≥ 65 years

Community- dwelling

MMSE

Multivariable Cox regression (hazard ratio)

Age, sex, education, income / occupation, ADLS, residence, participation in physical activity, smoking, drinking, negative well-being, baseline MMSE, leisure activities, physical diseases

 Zahodne et al., 2019 [101]

8,538

56.24%

45–93

Community- dwelling

Consortium to Establish a Registry for Alzheimer’s Disease Word List; Tests of semantic and letter fluency

Multivariate-adjusted standardized estimates

Age, sex, race, education, income, social network, heart disease, dyslipidemia, diabetes, nonlife threatening cancer, kidney failure, number of adults and children in childhood home, prenatal education, systolic BP, systemic inflammation, depression symptoms, perceived stress, BMI

 Zahodne et al., 2021 [102]

578

663.5%

 ≥ 65 years

Community- dwelling

WHICAP neuropsychological battery (episodic memory, language, visuospatial functioning); NIH Toolbox cognition module (executive function, working memory)

Longitudinal models

Age, sex, race, education, depression, presence / absence of 15 chronic conditions, baseline cognition