Table 1.

Study characteristics and findings.

References	Study Location/ Recruitment Sites/ Age Range	Design/Population/ Sample Size	Ethnicity/Race/ Language	Key Findings
Clark et al. (2005) [28]	U.S. Community-based clinics (urban and suburban)	Cohort study Patients with probable AD and primary family caregivers n = 79 units	African Americans	Time from noticing first AD signs to recognition No difference urban vs. suburban (Median 9 months (range 1–84) vs. 6 months (0–72)) Time from recognition to physician consultation: No difference urban vs. suburban (Median 3 months (0.2–84) vs. 2 months (0.1–48)) Longer delay in recognition associated with longer delay in physician consultation
Holsinger et al. (2011) [29]	U.S. 2 sites: Clinics and community-based clinics (urban) ≥50 years	Cross-sectional study Patients presenting for primary care appointments n = 345	White vs. minority Site 1 (n = 152): White 73% Site 2 (n = 193): White 57%	Majority accepted screening After exposing various potential risks and benefits, more accepted screening. No difference between white and minority
Fowler et al. (2012) [30]	U.S. Community-based clinics (urban) ≥65 years	Cross-sectional study Patients with no dementia receiving primary care n = 554	White 41.5%, African Americans 56.5% Other 1.4%	Majority willing to screen; 12.7% screened positive Refusal rates did not vary with ethnicity, education, other SES. Odds of refusal higher in older age groups
Fowler et al. (2015) [31]	U.S. 2 sites Clinic (urban) and community-based clinics (urban and suburban) ≥65 years	Cross-sectional study Patients with no dementia receiving primary care n = 400	Site 1 (n = 278): White 78.1%, African American 20.9%, Other 1.1% Site 2 (n = 122): White 96.7%, African American 2.5%, Other 0.8% English	Site 1: No difference in acceptance and refusal between White and African Americas Site 2: only one African American participant No differences in refusal between two sites Perceptions about the benefits of screening associated with acceptance of screening No effect of sociodemographic data except education predicted acceptance
Savva et al. (2015) [32]	U.S. Nationwide ≥71 years	Cross-sectional study Patients with dementia Data source: ADAMS substudy from HRS 2000–2002 waves n = 307	White 73%, Non-White 27% English or Spanish	121 informants reported prior diagnosis Grater CDC rate associated with prior diagnosis Race or nursing home residency no link with prior diagnosis Aged <90 years or married women associated with prior diagnosis. ¾ undiagnosed have mild dementia
Casado et al. (2017) [33]	U.S. Community (community outreach, local business sites, flyers, newsletters, social media) ≥40 years	Cross-sectional survey Adults n = 234	Korean Americans English or Korean	20.7% reported having experience with caring for someone with AD. Attitude scores were slightly more positive toward AD specialists (mean = 55.92 ± 7.40) than toward PCPs (mean = 54.24 ± 9.82).
Amjad et al. (2018) [34]	U.S. Nationwide ≥65 years	Cross-sectional observational study Patients with probable dementia or proxy Data source: NHATS n = 585	Non-Hispanic White, non-Hispanic Black, Hispanic or other non-Hispanic (Asian, Pacific Islander, and Native American)	39.5% undiagnosed Among diagnosed, 31% of those persons or their proxies were unaware of diagnosis. Undiagnosed persons likely to be non-White and lower education. But OR was statistically significant only for Hispanic/other non-White race Majority of older adults with dementia either undiagnosed or unaware of the diagnosis
Harrawood et al. (2018) [35]	U.S. 3 sites: clinic (urban) and community-based clinics (urban and suburban) ≥65 years	Cross-sectional study Patients with no dementia receiving primary care n = 954	African American 42.4% (n = 317): Site 1 (n = 280), Site 2 (n = 35), Site 3 (n = 2)	21.6% refused screening 78.4% agreed to be screened 10.2% screened positive: 11.7% African American; 9.0% White and other Older age (>75 years) low education, and perceived problem with memory associated with screening positive but no effect from race and research sites.
Gianattasio et al. (2019) [7]	U.S. Nationwide ≥70 years	Longitudinal study Patients with dementia Data source: HRS biannual interviews with participants or proxy linked with Medicare claims n = 4647–5201 (2000 to 2010, 6 observations)	Non-Hispanic White: 91–93%, Non-Hispanic Black: 7–9% English or Spanish	Whites were “correctly diagnosed” Blacks were “underdiagnosed” Black had double the risk of underdiagnosed compare with White at all 6 waves Risk of over diagnosed increased over time in both groups
Park et al. (2020) [36]	U.S. Community (urban, suburban, and rural) >50 years	Cross-sectional study Individuals with no dementia n = 1043	White 82.7%, Black/African American 11.6%, Hispanic 1.2%, Asian 0.6%, Native Hawaiian/Pacific Islander 0.3%, American Indian 0.8% English	In terms of demographic difference, female and participants with long-term care insurance have greater intention to screen but no mention about the effect of race. Younger age, higher level of perceived barriers, perceived benefit, higher social support and self-efficacy associated with increased intention
Lin et al. (2021) [37]	U.S. Nationwide ≥70 years	Prospective cohort study Patients with probable dementia Data source: HRS 2000–2014 linked with Medicare and Medicaid n = 3966	Non-Hispanic White 80.8%, non-Hispanic Black 11.9%, Hispanic 7.3% English or Spanish	A higher proportion of Blacks and Hispanics had a missed/delayed clinical dementia diagnosis compared with White (46%, s = 54% vs. 41%) Blacks and Hispanics had a poorer cognitive function and more functional limitations than White when received dementia diagnosis. Estimated mean delay: Blacks: 34.6 months; Hispanic 43.8 months; White: 31.2 months
Tsoy et al. (2021) [38]	U.S. Statewide	Retrospective cross-sectional study Patients with no prior dementia or MCI Data source: California CMS claims 2013–2015 n = 10,472	White 74.6%, Black 3.9%, Hispanic 12.0%, Asian 9.5%	Incident MCI diagnosis 23.3% White, 18.28% Black, 12.3% Asian, 15.8% Hispanic Timeliness of diagnosis Asian, Blacks, and Hispanic less likely to receive an incident diagnosis of MCI vs. dementia than White Estimated mean marginal effects of race/ethnicity on incident diagnosis of MCI were −11.0% for Asian, −6.6% for Hispanic, and −5.6% for Black
Wiese et al. (2019) [39]	U.S. Local service organizations, physician offices, church councils, senior center (rural)	Mixed methods Stakeholders -Social workers, healthcare administrators, nurses, nurses’ aides, physician, ministers, clerical worker, kitchen aids, farmworkers, auto mechanic, church worker n = 21	Non-Hispanic White (n = 5): professionals 4, layperson 1 African American (n = 11): professionals 9, laypersons 2 Afro-Caribbean (n = 2): professional 1, layperson 1 Hispanic American (n = 2): professional 1, layperson 1 English	81%: willing to screening annually if they developed memory problems or AD 85% of those previously screened would want to know if they were at higher risk of AD.
Wiese et al. (2021) [40]	U.S. Local city hall, senior centers, healthcare clinics, faith-based organizations (rural)	Mixed methods Stakeholders -Senior center administrators, senior center volunteer staffs, health clinic administrators, law enforcement officers, emergency medical technicians, physicians, nurse practitioners, nurses, paid caregivers, family caregivers, residents n = 22	White (n = 21), African American (n = 1) English	100%: willing to screening 82%: agreeable to blood testing 86%: agreeable to pictures of head or brain to detect dementia All would want their provider to screen them annually for memory problems
Williams et al. (2010) [41]	U.S. Churches, senior centers, health fair (announcements and flyers)	Mixed method Open-ended questions A part memory screening study of 793 community dwelling older adults n = 119	African American (n = 26) Afro-Caribbean (n = 31) European American (n = 29) Hispanic American (n = 33) English or Spanish	More African Americans recruited from churches than Hispanic and European American 89% valued the screening 92% would recommend screening to others 39% would seek professional help if they screened positive. More Hispanic Americans (70%) planned to seek help than did than European Americans (35%), African Americans (31%), or Afro-Caribbean (16%).
Hinton et al. (2004) [42]	U.S. Community (urban) (Physician referrals, Alzheimer’s Association, newspaper advertisements, etc.) Caregiver to patient ≥50 years	Qualitative Study In-depth interview A part of Survey Study: 33% of 117 family caregivers to community dwelling dementia patients -Wife, daughters, sons, others n = 39	African American (n = 10) Chinese American (n = 14) Anglo European-American (n = 15) English or three Chinese dialects (Mandarin, Cantonese, and Toisanese)	Help-seeking was most often initiated by family members or formal care providers Lack of a final diagnosis: more commonly reported by Chinese Americans compared with Anglos and African Americans Fragmentation in the referral process was common across all groups. Four general types of pathways to diagnosis: Smooth pathways/fragmented pathways/crisis pathways/dead-ended pathways
Hugh et al. (2009) [43]	U.S. Community (urban and rural) (A dementia outreach partnership)	Qualitative study Face-to-face semi-structured interview Health belief model Family caregivers of dementia patients -Daughters, spouses, sons, siblings n = 17	African American	Not knowledgeable about AD prior to their family diagnosed Knew that there is no known cure and expected a continued decline Almost half attributed a change in cognition was normal, age-related memory loss Some caregivers received support or resistance from other family member A supportive social network facilitated a diagnosis. Perplexing behavior and an increasing loss of ability are seen as cues to action
Leung et al. (2011) [44]	Canada Community and clinic (urban) (Alzheimer’s Society, posters) Patients: >55 years	Qualitative study Semi-structured interview Dyads of patients with dementia and family caregivers -Caregivers: wives, daughter, son-in-law, husband n = 6 dyads (7 caregivers)	Anglo-Canadian English	Symptom recognition to a dementia diagnosis 2–4 years Demented patients noticing memory difficulties earlier than careers but perceived as ambiguous and normalized or attributed to current health problem Diagnosis process was multiple visits and interactions with health professionals, obtained as more severe cognitive deficit emerged
Koehn et al. (2012) [45]	Canada Community (urban) (Chinese Resource Center of Alzheimer’s Society)	Qualitative study Semi-structured interview A Help-seeking Model Dyads of patients with probable dementia and their careers -Caregivers: wives, husband, daughter n = 10 dyads	Chinese Canadian Cantonese or Mandarin	The average pre-diagnosis interval: 1.5 years Caregivers and patients reported a diversity of experiences regarding the early symptoms of the patients’ cognitive deficit. Normalized of early symptoms Decision to seek care was made by family member, either spouse or consulted with adult children Two diagnosed done during acute care admission The role of family caregivers was more influenced by structural factors than by traditional Chinese cultural norms about family responsibilities and filial piety. 60% of the dyads experienced delays in diagnosis because Chinese family doctors dismissed the caregivers’ appraisals of the patients’ symptoms. Gender-based power imbalance between female family caregivers and male Chinese Canadian physicians
McCleary et al. (2012) [46]	Canada Community (urban) (Adult daycare center and flyers to community health center, local Alzheimer’s Society) Patient: >70 years	Descriptive qualitative study Semi-structured interview Dyads of patients with dementia and either one or two of their family careers -Caregivers: wives, daughters, daughter-in-law, husband, son, son-in-law n = 6 dyads	South Asian-Canadian English, Hindi, or Tamil	Early signs of dementia were seen as normal that are related to the aging process or patients’ personality characteristics. Seek attention when dementia symptoms were worsened after episodes Health seeking was delayed up to four years, even with significant dementia symptoms Safety concerns, new symptoms, treatment for other health problem influenced the recognition of a health problem
Garcia et al. (2013) [37]	Canada Clinic setting (a memory disorder clinic) Patients: >60 years	Qualitative study Semi-structured interview Dyads of patients with dementia and family or friends -Caregivers: spouses, daughters n = 7 dyads	French-speaking Canadian French	Estimated first suspicion of a problem to an official diagnosis: 1–7 years Not easy to identify signs and symptoms Lack of knowledge about the importance of the changes they were experiencing. No single symptoms sufficient to alert participants Preferentially sought from francophone Recognition to consultation with family physician from 4 months to 6 years. All final diagnoses were made by specialists, but family physicians clearly suspected dementia Variety of reasons for the delay.

Abbreviations: AD, Alzheimer’s disease; ADAMS, Aging, Demographics and Memory Study; HRS, Health and Retirement Study; CDC, Centers for Disease Control and Prevention; PCP, primary care provider; NHATS, National Health and Aging Trends Study; OR, odds ratio; MCI, mild cognitive impairment; CMS, Centers for Medicare & Medicaid.