Table 1.

Epidemiologic research on dementia in sub-Saharan Africa—population-based studies.

Country	Authors	Study design/methodology	Prevalence rate	Incidence rate	Risk factors/associated conditions
Nigeria	Ogunniyi et al., 1992 [19]	(i) Community-based, cross-sectional survey in Idikan NW3 ward (ii) n = 932 (iii) Age ≥ 40 years; 293 subjects were ≥65 years (iv) Screening and diagnosis of dementia: modified ∗MMSE, ∗DSM-III-R	No prevalence data reported as none of the subjects met diagnostic criteria for dementia per DSM-III-R		Decline in cognitive function significantly correlated with age, female sex, and low level of education
	Kamboh et al., 1997 [20]	(i) Population-based study (ii) Aim: to determine prevalence of APOE-ɛ4 and association between 1-antichymotrypsin (ACT) and APOE-ɛ4 loci and risk of ∗AD in Nigerian Blacks versus Caucasians living in Pittsburgh, USA, and women versus men (iii) n = 1,533 (803 Caucasians and 730 Nigerian Blacks) (iv) Age: 40–89 (mean 57.5) for Caucasian and 19–70 (mean 40.8) for Nigerian Blacks	Authors note that despite much higher prevalence of APOE-ɛ4 and ACT∗A carriers in Blacks, prevalence of AD in Blacks was found to be comparable to, or even lower than in Caucasians in other studies		(i) Distributions of ACT and APOE alleles are significantly different between Caucasians and Nigerian Blacks (ii) There was a nonrandom association between the two polymorphisms in Caucasian and Nigerian women; this may in part explain higher prevalence of dementia in women versus men
	Indianapolis Ibadan Dementia Project (IIDP) Articles  Hendrie et al., 1995 [8], Osuntokun et al. 1995b [21], Ogunniyi et al., 1997 [22], Ogunniyi et al., 2000 [23], Hendrie et al., 2001 [24], Gureje et al., 2006a [25], Hall et al., 2006 [26], Ogunniyi et al., 2006 [27], Smith-Gamble et al., 2002 [28], Ogunniyi et al., 2011 [29–31], and Hendrie et al., 2013 [32]	Longitudinal prospective comparative study (survey conducted in two phases) Phase 1: initial selection criteria  (i) Age ≥ 65 years (ii) Community-dwelling elderly African Americans living in Indianapolis and Yoruba living in Ibadan, Nigeria (iii) Screening tool for dementia: harmonized ∗CSID Outcome  2494 Yoruba versus 2212 African Americans selected Phase 2: final selection criteria based on clinical assessment  Diagnostic tools  (i) Dementia: DMS-III-R and ∗ICD-10 (ii) Severity of dementia: DSM-III-R, ICD-10, and clinical dementia ratings (iii) Probable or possible AD: ∗NINCDS-ADRDA Outcome   423 Yoruba versus 351 African Americans included	1995: age-adjusted prevalence of dementia (2.29% versus 4.82%) and AD (1.41% versus 3.69%) were found to be lower in Yoruba than in ∗AA and, respectively, in community dwelling subjects (i) 1997: door to door survey in Idikan NW3 ward in subjects ≥65 years contested previous claims by Ogunniyi et al., 1992 [19], that dementia and AD were rare in this community (ii) 28 out of 2,494 subjects screened were found to have dementia. 18 subjects (64.3% of cases) were found to have AD (iii) ∗VaD was the 2nd most common subtype after AD	2001: Age-standardized annual incidence rates of dementia (1.35% versus 3.24%) and AD (1.15% versus 2.52%) found to be lower in Yoruba than AA	(i) 1995: in both AA and Yoruba, age was associated with increased dementia prevalence and AD was the most common dementia subtype (ii) Prevalence of APOE-ɛ4 allele high among elderly Yoruba but lack of association between the allele and AD among Yoruba subjects versus AA (iii) 1997: increasing age and female gender were found to be risk factors for AD. Living with others was protective against dementia (iv) 2000–2006: lack of association between AD and possession of APOE-ɛ4 allele in Nigerian sample, unlike AA (v) Cardiovascular risk factors (hypertension, diabetes, stroke, smoking, and mean BMI) were found to be higher in AA than in Yoruba suggesting that environmental factors play a major role in the development of dementia (vi) Increasing levels of cholesterol levels and LDL associated with increasing risk of AD in individuals without the APOE-ɛ4 allele but not in those with APOE-ɛ4 in Yoruba subjects (vii) Increasing age and female gender associated with increased risk of incident dementia (viii) 2011: personality change is a significant predictor of future dementia, independent of cognition and functional status, in both Yoruba and AA (ix) Hypertension and decline in BMI are associated with increased risk of incident dementia in elderly Yoruba (x) 2013: increased homocysteine levels associated with similar but nonsignificant increase in dementia risk for both Yoruba and AA subjects despite significant differences in folate levels between the two regions
	Ochayi and Thacher, 2006 [33]	(i) Cross-sectional survey Nigeria (Jos, Central Nigeria) Age ≥ 65 years (ii) n = 280 (iii) Central Nigeria (iv) Diagnostic tool: CSID	Slightly higher prevalence of dementia (6.4%) than reported in older studies from Ibadan in Southern Nigeria (2.29%)		(i) Female sex, low body mass index (BMI), lack of NSAID use, and increasing age are risk factors for dementia (ii) Lack of association between level of education and dementia
	Ibadan Study of Aging (ISA) articles  Gureje et al., 2006 [34], Gureje et al., 2011 [35]	(i) Community-based survey in 8 Yoruba speaking Nigerian states (ii) n = 2152 (iii) Age ≥ 65 years (iv) 3-year followup (v) Dementia assessment: 10-word delayed recall test and clinician home-based interview to assess function	2006: Prevalence of probable dementia was 10.1%	2011: Estimated incidence of dementia was 21.85 per 1000 person-years	(i) 2006: female gender and increasing age were risk factors (ii) Lifetime history of alcohol use doubles the risk (iii) 2011: increasing incidence of dementia in rural areas. Social isolation and low economic status were identified as risk factors for incident dementia
	Yusuf et al., 2011 [7, 36]	(i) Cross-sectional, community-based survey (ii) Nigeria (Zaria), Northern Nigeria (iii) Age ≥= 65 years (iv) n = 322, urban area (v) Screening: CSID, ∗CERAD, ∗SDT, Blessed Dementia Scale and a sociodemographic questionnaire (vi) Dementia diagnosis: ∗ICD-10 and DSM-IV	Prevalence of dementia estimated at 2.79% with AD accounting for most cases in the community studied (66.67%)		Age was identified as risk factor for dementia
Benin	Guerchet et al., 2009 [9]	(i) Cross-sectional community-based survey (ii) Rural community versus urban community (iii) Age ≥ 65 years (iv) n = 502 (v) Screening: CSID and five-word test (vi) Dementia diagnosis: DSM-IV (vii) Probable and possible AD: NINCDS-ADRDA	(i) Prevalence of dementia is 2.6% in this rural community (ii) Dementia prevalence appears to be lower in developed countries (iii) AD is the most common cause of dementia rural setting		(i) Low prevalence of dementia in this rural community despite high a high prevalence of the APOE-ɛ4 allele in study participants (ii) There was a significantly lower frequency of the APOE-ɛ2 allele in study participants with dementia
	Paraïso et al., 2011 [10]	(i) Cross-sectional community-based study (ii) Urban community (iii) Age ≥ 65 (iv) n = 1137 (v) Screening: CSID and five-word test (vi) Dementia diagnosis: DSM-IV (vii) Probable and possible AD: NINCDS-ADRDA	(i) Dementia prevalence was not significantly higher in urban versus rural community: 3.7% versus 2.6% (ii) Prevalence rate of dementia in this urban community was similar to that reported in other cities in developing countries (iii) AD is the most common cause of dementia in urban setting		(i) Age and female gender associate with dementia (ii) Differences in level of education did not affect dementia prevalence
Central African Republic and Republic of Congo	Guerchet et al., 2010 [11], Guerchet et al., 2012 [37], Mbelesso et al., 2012 [12], and Guerchet et al., 2013 [38]	(i) Multicenter cross-sectional community-based surveys (ii) Conducted in 2 urban communities (Bangui and Brazzaville) (iii) Age ≥ 65 (iv) n = 496 in Bangui and 520 in Brazzaville (Bangui in Central African Republic and Brazzaville in Republic of Congo) (v) Screening: CSID and five-word test (vi) Dementia diagnosis: DSM-IV (vii) Probable and possible AD: NINCDS-ADRDA	2010: prevalence rates of dementia in these 2 urban areas of central Africa is similar to high income countries: 8.1% in Bangui and 6.7% in Brazzaville		(i) 2012: increasing age, female gender, hypertension, a body mass index < 18.5, depressive symptoms, and the lack of a primary education were significantly associated with dementia (ii) Life events (death of one parent during childhood and recent move) were also associated with increased risk of dementia (iii) 2013: there is a link between atherosclerosis (represented by low ankle-brachial index) and cognitive disorders in Africans as previously reported in AA and other ethnic groups
Kenya	Chen et al., 2010 [39]	(i) Setting: rural community (ii) Cross-sectional study evaluating the usefulness of Kikuyu version of the CSID in diagnosis of AD (iii) Age = 70–96 years (mean age 70.7) (iv) n = 184 (84 controls versus 100 demented subjects) (v) Diagnosis of dementia: ICD-10 and DSM-III-R (vi) APOE genotyping was conducted (vii) Blood test: for HIV and syphilis			(i) There was no association between years of education or vascular factors (diabetes, stroke, lipid levels, and hypertension) and dementia status (ii) APOE-ε4 allele frequencies were high (~30%) and not different between normal subjects and those with probable AD
Tanzania	Longdon et al., 2013 [40] Tanzania (six rural communities)	(i) Two-phase cross-sectional survey (ii) Age ≥ 70 years (iii) n = 1198 (iv) 6 rural communities (villages) (v) Screening tool: CSID (vi) Dementia diagnosis: DSM-IV	(i) Age-standardized prevalence of dementia was 6.4% (ii) Prevalence of dementia in rural Tanzanian population is similar to that reported in high income countries		(i) Dementia prevalence rates increased with increasing age (ii) Education was not a significant predictor of dementia
	Paddick et al., 2013 [6]	(i) Two-phase cross-sectional survey (ii) Aim: to compare prevalence rates obtained in the study by Longdon et al., 2013 [40] using the DSM-IV criteria for diagnosis of dementia with those obtained using the 10/66 diagnostic criteria, which is specifically designed for use in low and middle income countries (iii) Age ≥ 70 years (iv) n = 1198 (v) 6 rural communities (villages) (vi) Screening tool: CSID (vii) Dementia diagnosis: 10/66 diagnostic criteria	Prevalence of dementia was found to be 21.6% (AD prevalence not reported) in the rural Hai district of Tanzania using the 10/66 diagnostic criteria for dementia		Education was a significant predictor of “10/66 dementia,” but not of DSM-IV dementia

AA: African Americans; AD: Alzheimer's dementia; AGECAT: automated geriatric examination for computer-assisted taxonomy; CERAD: the consortium to establish a registry for Alzheimer's disease; CSID: community screening instrument for dementia; DSM-III-R: diagnostic and statistical manual of mental disorders 3rd edition revised; DSM-IV: diagnostic and statistical manual of mental disorders 4th edition; ICD-10: international classification of diseases 10th revision; GMS: the geriatric mental status schedule; MMSE: minimental state examination; SDT: stick design test; NSAID: nonsteroidal anti-inflammatory drugs; NINCDS-AIREN: National Institute of Neurological Disorders and Stroke and the Association Internationale pour la Recherche et l'Enseignement en Neurosciences; NINCDS-ADRDA: National Institute of Neurological and Communicative Disorders and Stroke and the Alzheimer's Disease and Related Disorders Association; VaD: vascular dementia.