Table 1 Characteristics of eligible studies of migration and cognitive status

Adelman, S., et al. (2011) (28)

436 aged ≥60 living in UK (218 African-Caribbean immigrants, 218 Whites UK-born)

Community and institutional setting

Cross-sectional study

Cognitive status:

1. CAMDEX

2. MMSE

Dementia:

ICD-10

DSM-IV

NINCDS-ADRDA

NINDS-AIREN

Country of birth

Age, Education, Gender, Marital status

Compared to native Whites, the prevalence of dementia was significantly higher in the African-Caribbean immigrant group

Kave, G., et al. (2008) (29)

814 older Jewish population in Israel aged ≥75 (138 Asian/African immigrants, 276 European/American immigrants, 400 nonimmigrants)

Community and institutional setting

Longitudinal study follow-up period: up to 12 years

Cognitive status:

1. The Katzman et al. test

2. MMSE

Country of birth, Age at immigration

Age, Education, Occupation, Income, Number of language speak

No significant differences were found in cognitive function among Asian/African immigrants, European/American immigrants, and nonimmigrants

Age at immigration was not significantly associated with cognitive function.

Multilingualism was significantly associated with better cognitive status

Chertkow, H., et al. (2010) (30)

632 patients with probable AD from a Memory Clinic in Montreal, Canada (158 immigrants and 474 nonimmigrants)

Clinical setting

Longitudinal study follow-up period: up to 1 year

Cognitive status:

1. MMSE

Dementia:

AD: NINCDS-ADRDA

MCI diagnosis: clinical evaluation

First language

Born in Canada

Age, Education, Gender, Occupation

No significant differences were found regarding age of dementia diagnosis between nonimmigrants and immigrants.

No significant differences were found regarding age of symptom onset between nonimmigrants and immigrants

Compared to immigrant unilinguals, native English unilinguals were likely to be diagnosed at a later age.

No significant differences were found regarding rate of cognitive decline between nonimmigrants and immigrants

Zahodne, L. B., et al. (2014) (31)

1067 Spanish speaking immigrants in the U.S. aged ≥65 from Dominican Republic, Puerto Rico, or other Caribbean countries

Community setting

Longitudinal study, follow-up period: up to 23 years

Cognitive status:

1. SRT

2. 15 item version of the Boston Naming Test

3. WAIS-R

4. Identities and Oddities subtest of the Mattis Dementia Rating Scales

5. the total number of words beginning with P, S, or V

6. Color Trails Test

Dementia:

DSM-III

Country of birth

Age at immigration

Age, Education, Gender, Cohort, Bilingualism, Fluency in English

Immigrants from Dominican Republic/Puerto Rico showed poorer initial performance on memory and language tests

Longer stay in the US was associated with better language function at baseline

Country of origin and time in the US were not associated with rate of change in any cognitive domain

Hill, T. D., et al. (2012) (32)

2734 aged ≥65 Mexican Americans in the U.S, (first and second generation)

Community setting

Longitudinal study, follow-up period: up to 14 years

Cognitive status:

1. MMSE

Country of birth

Age at immigration

Age, Education, Gender, Financial strain, English proficiency, Depression, Smoking and drinking behaviors, Physical functioning, Chronic diseases (diabetes, stroke, hypertension, heart attack)

Compared to U.S born Mexican American, immigrants who migrated to the U.S in their middle age showed better cognitive function at baseline.

Compared to U.S born Mexican American, male immigrants who migrated to the U.S in their middle age showed slower rate of cognitive decline.

Black, S. A., et al. (1999) (33)

2853 aged ≥65 Mexican Americans (first and second generation) in the U.S.

Community setting

Cross-sectional study

Cognitive status:

1. MMSE

U.S.-born/foreign born

Age, Education, Gender, Literacy, Depression, Marital status, Physical functioning, Chronic diseases (diabetes, stroke, hypertension)

Immigrants were more likely to experience severe cognitive impairment (MMSE < 18) compared to U.S-born Mexican Americans

Individuals that were interviewed in Spanish were less likely to be cognitively impaired or severe cognitively impaired.

Nguyen, H. T., et al. (2002) (34)

1759 aged ≥65 Mexican Americans in the U.S. (first and second generation)

Community setting

Longitudinal study follow-up period: 5 years

Cognitive status:

1. MMSE

U.S. born/foreign born

Age, Education, Gender, Marital status, Household composition, Chronic diseases (diabetes, stroke, hypertension), Depression, Hearing and vision difficulties

No significant differences in cognitive decline (Decline to MMSE ≤17, Decline at least three points) were found between U.S. born Mexican Americans and Mexican immigrants to the U.S.

Sheffield, K. M. and M. K. Peek (2009) (35)

3050 aged ≥65 Mexican Americans in the U.S (first and second generation)

Community setting

Longitudinal study follow-up period: up to 9 years

Cognitive status:

1. MMSE

U.S. born/foreign born

Age, Education, Gender, Marital status, Chronic diseases (diabetes, stroke, hypertension), Depression, Neighborhood characteristics, Socioeconomic disadvantages

No significant differences in both baseline cognitive function and cognitive decline were found between U.S. born Mexican Americans and Mexican immigrants to the U.S.

Respondents living in economically disadvantaged neighborhood were more likely to have faster cognitive decline than those living in advantaged neighborhood

Lawton, D. M., et al. (2015) (36)

81 (55 AD, 26 VaD) Mexican Americans in the U.S. (first and second generation) aged ≥60

Community setting

Cross-sectional study

Cognitive status:

1. 3MSE

2. SEVLT

3. The Spanish English Neuropsychological Assessment Scale

4. The Informant Questionnaire on Cognitive Decline in the Elderly

Dementia:

AD: NINCDS-ADRDA

VaD: NINDS-AIREN

Age of dementia diagnosis

U.S. born/foreign born

Bilingualism

No significant differences were found regarding the proportion of types of dementia among bilingual U.S born Mexicans, bilingual Mexican immigrants, monolingual U.S born Mexican, and monolingual Mexican immigrants.

No significant associations were found between immigrant status and age of dementia onset

No significant interactions were found between bi/monolinguals and immigrant status on age of dementia diagnosis

Mejia, S., et al. (2006) (37)

12,008 aged ≥65 (3398 Spanish-speaking U.S. immigrants, 8610 Mexicans)

Community setting

Cross-sectional study using several large datasets

Cognitive status:

1. Reduced Mini-Mental State Examination

2. Cross-cultural cognitive examination

3. Minineuropsi

4. MMSE

5. Modified Telephone Interview Cognitive Scale

Mexicans:

1. The Cognitive Factors Survey in Mexico City’s Elderly Population,

2. Salud, Bienestary Envejcimiento,

3. The Mexican Health and Aging Study

Immigrants:

1. Hispanic Established Population for the Epidemiological Study of the Elderly

2. Health and Retirement Survey and Assets and Health Dynamics Among the Oldest Old Survey

Age, Education, Gender, Marital status, Physical functioning,

The prevalence estimates of cognitive impairment were similar among immigrants and Mexicans

The prevalence of cognitive and functional impairment for immigrants were as least twice as high as for Mexicans

Raina, S. K., et al. (2014) (38)

2000 aged ≥60 older adults in India (500 urban residents, 500 rural residents, 500 migrants, 500 tribal population)

Community setting

Cross-sectional study

Cognitive status:

1. Hindi Mental State Examination (HMSE)

Dementia:

Clinical evaluation

N/A

N/A

Urban residents had a higher prevalence of cognitive impairment than rural and migrant population.

No significant differences were found in the prevalence of cognitive impairment between rural and migrant population.

No case of dementia was reported from tribal population.

Lindesay, J., et al. (1997) (39)

297 aged ≥65 older adults living in UK (149 Gujarati and 148 native whites)

Community setting

Cross-sectional study

Cognitive status:

1. CAMDEX

2. MMSE

Country of birth

Age, Education, Visual impairment, Ethnic group

Compared to native whites, Gujarati immigrants showed poorer orientation in time (day, year) and place (street), measured by MMSE.

The proportion of cognitive impaired participants at all severities in the Gujarati immigrants were higher than in the native whites

Segers, K., et al. (2013) (40)

1058 new patients from the Memory Clinic in Brussels, Belgium (861 nonimmigrants, 106 European immigrants, and 91 non-European immigrants)

Clinical setting

Cross-sectional study

Cognitive status:

1. MMSE

Dementia:

Medical record

Country of birth

N/A

Compared to nonimmigrants and non-European immigrants, European immigrants were more likely to have Parkinson related cognitive disorders.

Compared to nonimmigrants and European immigrants, non-European immigrants showed poorer cognitive function measured by MMSE

Raina, S. K., et al. (2010) (41)

1856 aged ≥60 living in North India (Dogra population and migrant Kashmiri Pandit population)

Community setting

Cross-sectional study

Cognitive status:

1. MMSE

Dementia:

Clinical evaluation

N/A

N/A

Compared to migrant Kashmiri Pandit population, the prevalence of dementia was significantly lower in the Dogra population

Kahana, E., et al. (2003) (42)

1720 aged ≥75 living in Ashkelon, Israel (659 Afro-Asian origin, 841 Europe-American origin)

Community setting

Cross-sectional study

Dementia:

DSM-III

Country of birth

Age, Education, Gender,

The prevalence of dementia was significantly higher in people of Afro-Asian origin than in people of Europe-American origin

Ethnic origin was not a significant risk factor for dementia while controlling for all other factors

Livingston, G., et al. (2001) (43)

1085 aged ≥65 older adults living in UK (667 White British, 139 Irish, 71 Cypriot, 98 Black African/Caribbean, 60 immigrants from European countries, 50 immigrants from other countries)

Community setting

Cross-sectional study

Dementia:

The shortened version of the Comprehensive Assessment and Referral Evaluation (Short-CARE)

Country of birth

Ethnicity

Age, Education, Gender, Income, Occupation, Drinking behavior, Chronic diseases (diabetes, stroke, CVD)

Compared to natives, the prevalence of dementia was higher in African-Caribbeans, after controlling for other factors

Compared to natives, the prevalence of dementia was lower in individuals born in Ireland, after controlling for other factors

Touradj, P., et al. (2001) (45)

193 non-Hispanic Whites in the U.S. aged ≥65 (106 nonimmigrants and 87 immigrants)

Community setting

Cross-sectional study

Cognitive status:

1. SRT

2. Benton Visual Retention Test (BVRT)

3. MMSE

4. WAIS-R

5. Identities and Oddities subtest of the Mattis Dementia Rating Scales (SRS Identities and Oddities subtest)

6. Boston Naming Test

7. Letter fluency:

8. Boston Diagnostic Aphasia Examination

9. Rosen Drawing Test

10. BVRT matching

U.S. born/foreign born

Age at immigration

Age, Education, Gender

Compared to immigrants, nonimmigrants showed significantly higher scores on measure of abstract reasoning (WAIS-R), naming (Boston Naming Test), and fluency (letter and category fluency)

Longer stay in the U.S. was associated with better score in WAIS-R.

Haan, M. N., et al. (2011) (46)

1789 aged ≥60 Mexican American in the U.S (827 first generation immigrants, 645 s generation, 67 third generation, 93 fourth generation)

Community setting

Longitudinal study, follow-up period: 9 years

Cognitive status:

1. 3MSE

2. SEVLT

Country of birth

Nativity of their parents and grandparents

Age, Education, Gender, Income, Occupation, Insurance, Childhood and adulthood SES, Chronic diseases (diabetes, stroke, hypertension)

Compared to first generation immigrants, second and third generation demonstrated better global cognitive function

Compared to first generation immigrants, second generation demonstrated better short term memory

Compared to first generation immigrants, the effect of life-time SES disadvantages on short term memory was greater in second generation immigrants.

Plitas, A., et al. (2009) (47)

142 aged ≥55 (76 Greek older adults, 66 Greek Australian)

Clinical setting

Cross-sectional study

Cognitive status:

1. CAMDEX

2. MMSE

Country of recruitment

Age, Education, Gender

Greek Australians showed a significantly poorer performance on CAMDEX compared to Greek elderly

Greek Australians showed a significantly poorer performance on MMSE compared to Greek elderly

A significant Group × Gender interaction on both CAMCOG and MMSE were observed

Nielsen, T. R., et al. (2012) (48)

109 aged ≥50 (73 Turkish immigrants, 36 Danish)

Community setting

Cross-sectional study

Cognitive status:

1. Recall of Pictures Test (RPT)

2. Clock Reading Test (CRT)

3. Supermarket Fluency (SF)

Country of birth

Age, Education, Acculturation

Turkish immigrants showed a significantly pooper performance on CRT and SF compared to Danish.

No significant differences were found regarding RPT test

Age, acculturation level were significantly association with test performances only among Turkish immigrants

Wilson, R. S., et al. (2005) (49)

6158 aged ≥65 living Chicago, United States (non-migrants and migrants)

Community setting

Longitudinal study follow-up period: up to 6 years

Cognitive status:

1. MMSE

2. The East Boston Story: immediate and delayed recall

3. The Symbol Digit Modalities Test

County of birth

Age, Education, Gender, Childhood SES, Occupation, Birth county characteristics

Compared to residents born in Cook county, residents born elsewhere in the U.S showed lower global cognition score at baseline.

Born in a higher socioeconomic level county was associated with a higher cognitive function.

No significant associations were found regarding county of birth and rate of cognitive decline.

Graves, A. B., et al. (1999) (50)

1836 Japanese Americans: Issei (Japanese immigrants) Kibei (American-born but educated in Japan), and Nisei (American-born and educated) aged ≥65 in the U.S.

Community setting

Longitudinal study, follow-up period: 2 years

Cognitive status:

1. CASI

Country of birth

U.S./Japan educated

Years living in Japan

Age, Education, Gender, Income, Acculturation, APOE ε 4,

Compared to Nisei, Issei and Kibei were less likely to experience cognitive decline during the 2-year follow-up period

Longer years lived in Japan before age 18 was associated with less likelihood of cognitive decline

Yano, K., et al. (2000) (51)

3734 Japanese American men in the U.S. aged 71–93 (first and second generation)

Community setting

Cross-sectional study

Cognitive status:

1. CASI

Country of birth

Childhood years in Japan

Age, Education, Gender, Income, Occupation, Fluency in Japanese, Tofu and fish consumption, APOE ε 4

Longer childhood years in Japan was significantly associated with lower CASI score.

Among participants who preferred to be testing in Japanese, longer childhood years in Japan was significantly associated with higher CASI score.

Among participants who preferred to be tested in English, longer childhood years in Japan was significantly associated with lower CASI score

Stouten, L. H., et al. (2013) (52)

407 schizophrenia patients in the Netherland (157 Dutch, 138 1st generation immigrants, 112 2nd generation immigrants)

Clinical setting

Cross-sectional study

Cognitive status:

1. Rey’s Auditory Verbal Learning task (RAVLT) immediate/delayed recall

2. Continuous Performance Task (CPT)

Country of birth

Age, Education, Gender, Physical functioning, Medication

Immigrants showed larger cognitive deficits in RAVLT (both immediate and delayed) and CPT comparing to the Dutch

First generation Immigrants showed larger cognitive deficits in RAVLT (both immediate and delayed) and CPT comparing to the second generation

Immigrants from Morocco, Turkey and other Non-Western countries demonstrated larger cognitive deficits in RAVLT immediate recall and CPT comparing to immigrants from other countries

Al Hazzouri, A. Z., et al. (2011) (53)

7042 aged ≥60 Mexicans and Mexican Americans (4687 Mexicans, 562 Mexican-return migrants, 908 Mexican-immigrants to the U.S., 871 Mexican-U.S. born)

Community setting

Cross-sectional study using two large datasets

Cognitive status:

1. SEVLT

2. Three 8-word memory trails

Country of birth

Current place of residency

History of immigrating to the U.S.

Age, Education, Gender, Income, Insurance, Parents’ education, Chronic diseases (diabetes, stroke, hypertension, heart attack)

Compared to Mexicans, return migrants showed better cognitive function while both immigrants to the U.S. and U.S. born Mexicans showed poorer cognitive function

Compared to Mexicans, the positive effect of parents’ education on cognitive function is greater in U.S born Mexicans