Abstract
Understanding the lifetime risk of dementia can inform public health planning and improve patient engagement in prevention. Using data from a community-based, prospective cohort study (n = 15,043; 26.9% Black race, 55.1% women and 30.8% with at least one apolipoprotein E4 (APOE ε4) allele), we estimated the lifetime risk of dementia (from age 55 years to 95 years), with mortality treated as a competing event. We applied lifetime risk estimates to US Census projections to evaluate the annual number of incident dementia cases from 2020 to 2060. The lifetime risk of dementia after age 55 years was 42% (95% confidence interval: 41–43). Rates were substantially higher in women, Black adults and APOE ε4 carriers, with lifetime risks ranging from approximately 45% to 60% in these populations. The number of US adults who will develop dementia each year was projected to increase from approximately 514,000 in 2020 to approximately 1 million in 2060. The relative growth in new dementia cases was especially pronounced for Black adults. These results highlight the urgent need for policies that enhance healthy aging, with a focus on health equity.
The United States has experienced substantial population aging over the past century, resulting in a rise in late-life diseases1. Dementia, once an uncommon condition, now affects more than 6 million Americans2. It is a leading cause of disability among older adults and accounts for more than 100,000 deaths each year3. Dementia also carries a substantial economic burden, with total costs exceeding $600 billion annually in the United States4.
The lifetime risk of dementia is a critical public health measure that can raise awareness, enhance engagement in prevention and inform policymaking5,6. Major health organizations, including the American Heart Association and the Alzheimer’s Association, report this metric to educate patients and clinicians about dementia risk across the lifespan5,7. Lifetime risk estimates can also be used to generate and refine projections of dementia, which can help optimize public health planning6.
Previous studies suggest that 11–14% of men and 19–23% of women in the United States will develop dementia during their lives6,8,9. However, these estimates were based on older data with limited dementia ascertainment, potentially resulting in underestimation. Racial disparities in the lifetime risk of dementia are also poorly characterized, as population-based analyses have typically been limited to Non-Hispanic White populations.
In the present study, our primary objective was to generate contemporary estimates of the lifetime risk of dementia, overall and across different population subgroups. We also characterized differences in the age of diagnosis and projected the number of new dementia cases that will develop in the United States over the next four decades. To achieve these aims, we analyzed over three decades of longitudinal data (1987–2020) from participants in the Atherosclerosis Risk in Communities (ARIC) study.
The study population included 15,043 participants free of dementia at age 55 years (26.9% Black race; 55.1% women). Approximately 31% of participants carried at least one apolipoprotein E4 (APOE ε4) allele (28.1% one copy; 2.7% two copies). Additional baseline characteristics are available in Extended Data Table 1.
Over a median follow-up of 23 years (interquartile range, 16–27 years), there were 3,252 incident cases of dementia, 5,803 deaths without dementia and 2,131 participants lost to follow-up. A total of 783 (24%) dementia cases were diagnosed at study visits with cognitive testing, 1,589 (49%) through phone interviews and 880 (27%) through review of hospital and death records (Extended Data Table 2). The crude incidence of dementia was higher in APOE ε4 carriers and Black adults but was similar across sex (Extended Data Table 3). In contrast, the incidence of death without dementia did not differ by APOE ε4 status but was higher in Black adults and men.
At age 55 years, the lifetime risk of dementia (up to age 95) was 42% (95% confidence interval (CI): 41–43) (Fig. 1a). The cumulative incidence of dementia remained low from age 55 to age 75 (3.9%) but increased substantially after age 75. There was a higher lifetime risk of dementia in women versus men (48% (95% CI: 46–50) versus 35% (95% CI: 33–36)) and Black versus White adults (44% (95% CI: 41–46) versus 41% (95% CI: 40–43)) (Fig. 1b,c). Differences in lifetime risk across race emerged by age approximately 75 years, whereas sex differences occurred starting at approximately 85 years. Adults with two copies of the APOE ε4 alle had a substantially higher lifetime risk of dementia (59% (95% CI: 53–65)) compared to those with one copy (48% (95% CI: 45–50)) and those with no copies (39% (95% CI: 37–40)), with differences beginning at age approximately 70 years (Fig. 1d). The cumulative incidence of dementia at each year of age corresponding to Fig. 1 are provided in Supplementary Table 1.
Fig. 1 ∣. Lifetime risk of dementia (from age 55 years to 95 years), the ARIC study, overall and by sex, race and APOE ε4 status (n = 15,043).
The cumulative incidence of dementia at a given age in the overall study population (a) and by sex (b), race (c) and APOE ε4 status (d). Estimates are expressed as percentages and account for the competing risk of death. The 95% CIs are indicated in the shaded area. The numbers on the right margins indicate the cumulative incidence at 95 years of age (94.2 years for Black adults and 93.3 years for those with two APOE ε4 alleles).
APOE ε4 carriage was associated with a marked increase in the lifetime risk of dementia for White men and women and for Black women but a modest increase for Black men (Extended Data Table 4). Lifetime risk of dementia increased progressively with older index ages (Extended Data Table 5). For example, among individuals alive and dementia free at age 75 years, lifetime risk exceeded 50% in the overall population and for nearly every population subgroup. Among participants with dementia identified through phone interviews, 50.4% also had dementia indicated on hospital records or death certificates. After excluding dementia cases identified by phone without further corroboration by medical or death records, the lifetime risk of dementia was 36%, with differences across subgroups that were similar to the primary analysis (Supplementary Table 2).
The median age of dementia diagnosis was 81 years (interquartile range: 77–86 years) (Extended Data Table 6). Dementia occurred at earlier ages for APOE ε4 carriers (median age: 79 years for two copies, 81 years for one copy and 82 years for no copies) and Black versus White adults (median age: 79 years versus 82 years). Overall, 17% of participants with dementia were diagnosed before the age of 75 years, with higher rates for men, Black adults and APOE ε4 carriers.
Applying our lifetime risk estimates to US Census population projections, the annual number of incident dementia cases is projected to increase from approximately 514,000 (2020) to approximately 1 million (2060) (Fig. 2). The largest absolute increase is expected to occur in individuals aged 85–95 years (increase of ~232,000) and 75–84 years (increase of ~238,000). The number of individuals who develop dementia each year is expected to nearly double in White adults and triple in Black adults.
Fig. 2 ∣. Projected number of incident dementia cases in US adults, 2020–2060, overall and by age, sex and race.
Estimated number of US adults who will develop dementia each year from 2020 to 2060, overall (a) and by sex (b), race (c) and age (d). The projected number of total incident dementia cases in the overall population (a) may differ slightly from the total in stratified analyses (b–d) due to rounding and because incidence rates were calculated separately within each subgroup.
In this large community-based cohort study, 42% of participants developed dementia after midlife. Lifetime dementia risk was highest in APOE ε4 carriers, women and Black adults, with lifetime risks in these groups ranging from approximately 45% to 60%. The annual number of incident dementia cases is expected to double over the next four decades, reaching approximately 1 million by 2060.
Our lifetime risk estimates are higher than those in previous population-based studies. In the Framingham Heart Study, 14% of men and 23% of women developed dementia from age 45–105 years8. The lifetime risk of dementia (age 45–95 years) in the Rotterdam Study was 19% and 31% for men and women, respectively10. The higher lifetime risk estimates in our analyses may reflect differences in dementia ascertainment. Previous research primarily relied on cognitive testing at study visits and a review of medical and death records to identify dementia. However, this may result in under-detection, because participants with cognitive impairment are less likely to attend in-person assessments, and administrative records lack sensitivity11. In contrast, the ARIC study combined cognitive evaluations at study visits with intensive surveillance (phone interviews and review of hospital and death records) to maximize dementia ascertainment. Ongoing phone interviews with participants and informants were especially important, as they identified approximately half of all dementia cases in our study.
The ARIC cohort is also more geographically, racially and socioeconomically diverse than cohorts in previous studies. Racial and ethnic minority adults and individuals from lower economic backgrounds have a higher burden of important risk factors, potentially contributing to differences in long-term dementia risk.
Our results suggest that the current lifetime risk of dementia may be substantially higher than previously thought, emphasizing the importance of prevention throughout the life course. Policies focused on optimizing cardiovascular health and preserving hearing may be particularly important12. Accumulating data from clinical trials have linked healthy lifestyle behaviors, the absence of vascular risk factors and hearing rehabilitation with improved cognitive outcomes13-17. However, only approximately 20% of US adults are meeting recommended lifestyle and cardiovascular health targets18, and only approximately 30% of older adults with hearing loss are using a hearing aid19. These trends highlight broad opportunities for dementia risk reduction in the population.
We found that APOE ε4 carriers had a very high absolute risk of dementia. Roughly half of those with one APOE ε4 copy and approximately 60% with two copies developed dementia after midlife. Our estimates were similar to previous population-based research, which reported cumulative incidence rates ranging from 16% to 38% by age 80–85 years20. Consistent with previous studies21,22, we also found that APOE ε4 carriers developed dementia substantially earlier than non-carriers. As a result, differences in the cumulative incidence of dementia between APOE ε4 carriers and non-carriers were evident by age 70 years. These findings align with clinical and basic research, which suggests that APOE ε4 primarily increases Alzheimer’s disease risk by causing earlier amyloid pathology in the brain21,22.
We observed striking differences in dementia risk across race. Black adults had earlier dementia onset and a higher lifetime risk as compared to White adults. These findings extend previous studies, which have predominantly focused on racial/ethnic differences in prevalence or relative risk2,23,24. Racial disparities in dementia may reflect the cumulative effects of structural racism and inequality throughout the life course. For instance, poor access to education and nutrition may contribute to earlier differences in cognitive reserve, and socioeconomic disparities and limited access to care may lead to a higher burden of vascular risk factors at midlife. Reducing racial disparities will, therefore, require interventions that target high-risk individuals, along with reforms that address social determinants of health.
Consistent with previous studies, women had a substantially higher lifetime risk of dementia6,8,10. This pattern is thought to reflect women’s higher life expectancy25. In our analyses, the incidence of death without dementia was nearly two times higher in men compared to women. Thus, men were less likely to survive to older ages, reducing their overall lifetime risk of dementia. These findings corroborate previous population-based research, which similarly found that sex differences in survival are a major contributor to the higher lifetime risk of dementia in women8,26.
Similar to our results, the Global Burden of Disease Study estimated that approximately 516,000 US adults developed dementia in 2017 (ref. 27). We extended existing research by quantifying the growth of incident dementia over the next four decades. This trend will likely be driven by the large ‘Baby Boom’ generation reaching older age. The shifting racial composition of the United States may also contribute. Adults belonging to racial and ethnic minority groups have an increased long-term risk of dementia and are expected to make up the majority of the population by 2045 (ref. 28). In our projections, the relative increase in incident dementia cases was notably higher in Black versus White adults.
This study has several strengths. The ARIC study comprehensively captured dementia in all participants through cognitive testing, active surveillance and adjudication of dementia cases. ARIC is a diverse, contemporary, population-based cohort of adults followed from middle age.
There are also several study limitations. First, there may be some misclassification of dementia for participants identified through surveillance, because these were not adjudicated. Nonetheless, phone interviews were conducted using validated instruments, and diagnostic codes for dementia are highly specific29. Additionally, approximately half of dementia cases identified with phone interviews also had dementia documented in hospital or death records, indicating high reliability. Second, despite the comprehensive ascertainment of dementia, there still may be cases not captured in the ARIC study. This suggests that our analyses may be conservative and underestimate the true lifetime risk. Related to this, medical and death records may underestimate dementia, particularly in racial and ethnic minority adults30. However, differential ascertainment was mitigated by the use of multiple methods to identify dementia. Fourth, before study visit 5 (2011–2013, mean participant age 75 years), dementia was ascertained retrospectively with phone interviews and review of hospital and death records. As a result, there may be some underestimation of dementia before age 75 years. However, this bias may be minimal because most dementia cases develop at older ages. Fifth, we did not externally validate our results. Future population-based studies with long follow-up and comprehensive dementia ascertainment are needed to validate our lifetime risk estimates.
Sixth, our projections may not be generalizable to the entire US population. Nonetheless, our estimates of incident dementia cases were very similar to the Global Burden of Disease Study, which generated estimates by combining a broad set of data sources in the United States27. This suggests that our projections may provide a useful approximation of the current and future burden of incident dementia in the United States. Seventh, we compared lifetime dementia risk only for Black and White adults because ARIC did not include a large number of participants from other racial and ethnic groups. More diverse studies are needed to understand how dementia risk differs across populations over the life course. Eighth, our projections assumed that the incidence of dementia will remain stable over the next four decades.
In conclusion, more than four in 10 adults developed dementia in this large, longstanding, community-based cohort study, with higher rates in APOE ε4 carriers, women and Black adults. Approximately 1 million US adults will develop dementia annually by 2060. Policies that enhance prevention and healthy aging are urgent public health priorities for reducing the substantial and growing burden of dementia31.
Methods
Study population
The ARIC study is a community-based cohort of 15,792 adults from four US communities (Forsyth County, North Carolina; Jackson, Mississippi; Minneapolis, Minnesota; and Washington County, Maryland). Participants aged 45–64 years were originally recruited from 1987 to 1989 and have undergone clinical examinations (including cognitive testing), laboratory testing and medical interviews (in-person and by phone) for the past three decades. Beginning at study visit 5 (2011–2013), ARIC investigators began conducting comprehensive neuropsychological testing (with expert adjudication) at study visits and semi-annual phone-based cognitive assessments. Phone-based assessments were administered to participants and informants (for example, relatives). All participants provided written informed consent, and the study was approved by institutional review boards at all research sites. Further details about the ARIC study are available elsewhere32.
We excluded participants with prevalent dementia before age 55 years (n = 4), missing information for covariates (n = 543) and those who died or were lost to follow-up before the age of 55 years (the index age for our primary analyses, n = 202). These restrictions yielded an analytic sample of 15,043 participants.
Dementia ascertainment
Dementia ascertainment in ARIC is summarized in Supplementary Fig. 1 and was described in detail previously33,34. In brief, dementia status was determined in all participants, including those who did not return for clinical study visits or died during follow-up, using three broad approaches.
The first approach was based on cognitive testing at clinical study visits. A three-test neurocognitive battery was administered at visit 2 (1990–1992) and visit 4 (1996–1998), and an expanded 10-test neuropsychological battery was administered at visit 5 (2011–2013), visit 6 (2016–2017) and visit 7 (2018–2019)35. Due to the coronavirus disease 2019 (COVID-19) pandemic, a shortened six-battery test was administered at visit 8 (2020). The Clinical Dementia Rating scale, the Functional Activities Questionnaire, the Mini-Mental State Examination and the Blessed scale were used at visits 5–8, with the former two in only a subset of participants36-39. Dementia was initially diagnosed using an algorithm based on criteria recommended by the National Institute on Aging–Alzheimer’s Association diagnostic guidelines workgroups and the fifth edition of the Diagnostic and Statistical Manual of Mental Disorders40. An expert panel subsequently reviewed and adjudicated cases of dementia identified by the algorithm.
For those who did not return for in-person study visits, dementia was identified through annual or semi-annual telephone interviews with participants and informants using validated instruments, including the Telephone Interview for Cognitive Status, the Clinical Dementia Rating scale, the Functional Activities Questionnaire, the Six-Item Screener and the Ascertain Dementia 8-item screener36-38,41,42.
Dementia cases were identified by using diagnostic codes in hospital records or death certificates for all participants throughout the study (see Supplementary Table 3 for International Classification of Diseases (ICD) 9/10 codes used for ascertainment)32. Where possible, hospital or death certificate–based diagnoses were supported by interview informants using the Ascertain Dementia 8-item screener. Response rates for all three methods of dementia ascertainment (study visits, phone-interviews and medical and death records) are provided in Supplementary Table 4.
We used all available data (in-person testing, phone interviews and medical and death records) to classify participants. When multiple data sources were available, we used the following priority order to ascertain dementia: (1) expert diagnosis of dementia based on neuropsychological testing at study visits; (2) computer algorithm diagnosis based on neuropsychological testing at study visits; (3) low score on the Telephone Interview for Cognitive Status (adjusted for education level); (4) elevated score on the Clinical Dementia Rating scale and the Functional Activities Questionnaire; (5) low score on the Six-Item Screener and elevated score on the Ascertain Dementia 8-item screener; (6) two low scores on the Six-Item Screener; (7) one low score on the Six-Item Screener if participant was lost to follow-up or deceased; (8) dementia on hospital discharge codes; and (9) dementia on death certificate codes. The priority list was designed to classify participants with the most reliable data source available.
The date of dementia onset was defined as the earliest date dementia was diagnosed in any data source (study visit, phone interview or hospital or death records). For cases identified through informant interviews (for participants who were deceased), hospital records or death certificates, we subtracted 180 days from the date of diagnosis to account for a potential lag in reporting.
In sensitivity analyses, we re-estimated lifetime risk using a more conservative definition of incident dementia. In particular, participants were classified with dementia through phone interviews only if the diagnosis was also indicated in hospital records or death certificates. Participants with a phone-based diagnosis of dementia that was not ‘confirmed’ with medical records were classified as having no dementia and censored at the date of the phone interview. This conservative definition of dementia thus included cases identified through study visits, phone interviews that were subsequently confirmed through hospital or death records, or hospital or death records.
Sociodemographic characteristics and APOE ε4 status
Demographic characteristics (age, sex and race) were self-reported. Participants reported race from a list generated by the researchers (Asian, Black, American Indian/Alaskan Indian or White). APOE ε4 genotyping was performed using the TaqMan assay, and participants were classified by APOE ε4 status (zero, one or two alleles).
Statistical analyses
We examined baseline characteristics and the proportion of dementia cases identified through different ascertainment methods (study visits, telephone interviews or diagnostic codes). We estimated crude rates of dementia and death without dementia, overall and across subgroups.
We estimated non-parametric cumulative incidence function curves to calculate the lifetime risk of dementia (from age 55 years to age 95 years), overall and by sex, race and APOE ε4 status43. We treated death without dementia as a competing event to avoid overestimating lifetime dementia risk. We used age as the timescale and calculated lifetime risk for those who were alive and free of dementia at age 55 years (index age). Participants older than 55 years were included as late entries. We calculated person-time from age 55 years until dementia diagnosis, death free of dementia, loss to follow-up, age 95 years or administrative censoring (31 December 2020), whichever occurred first.
In secondary analyses, we re-estimated the lifetime risk of dementia with cumulative incidence function curves using (1) a more conservative definition of dementia and (2) older index ages (ages 65, 75 and 85). We also examined interactions between race and sex with the presence of APOE ε4.
Among individuals with incident dementia, we calculated the median age at diagnosis and the percentage of adults diagnosed between the ages of 55–74 years, 75–84 years and 85–95 years.
To project the burden of incident dementia, we generated smoothed cumulative incidence function curves for dementia using interpolation, overall and across subgroups. We then estimated age-specific incidence rates by calculating the increase in dementia incidence with each additional year of age. We multiplied these age-specific rates with corresponding population counts from the US Census to estimate the number of new dementia cases expected in the United States28 from 2020 to 2060, overall and by age categories (55–74 years, 75–84 years and 84–95 years), sex (men, women) and race (White, Black).
All analyses were conducted in Stata/SE 18.0 (StataCorp)44.
Extended Data
Extended Data Table 1 ∣.
Participant characteristics at age 55 years, the ARIC study
| Overall (N=15,043) |
|
|---|---|
| Sex, % | |
| Men | 44.9 |
| Women | 55.1 |
| Race, % | |
| White | 72.8 |
| Black | 26.9 |
| Other | 0.3 |
| Study site | |
| Forsyth County, North Carolina | 25.6 |
| Jackson, Mississippi | 23.4 |
| Minneapolis, Minnesota | 25.2 |
| Washington County, Maryland | 25.7 |
| APOE ε4 status, % | |
| 0 alleles | 69.2 |
| 1 allele | 28.1 |
| 2 alleles | 2.7 |
| Diabetes, % | |
| No | 90.7 |
| Yes | 9.3 |
| Smoking status, % | |
| Never | 40.0 |
| Former | 35.8 |
| Current | 23.8 |
| BMI, mean (SD), kg/m2 | 28.2 (6) |
| BMI categories, % | |
| Normal (BMI <25 kg/m2) | 30.0 |
| Overweight (BMI 25-<30 kg/m2) | 39.4 |
| Obesity (BMI≥30 kg/m2) | 30.6 |
| Missing | 0.1 |
| Hypertension, % | |
| No | 61.1 |
| Yes | 38.7 |
| Missing | 0.2 |
Diabetes, hypertension, smoking and body mass index (BMI) were based on available measurements from the clinical study visit closest to when participants were 55 years of age. Diabetes was defined as fasting glucose ≥ 126 mg dl−1, non-fasting glucose ≥ 200 mg dl−1, self-report of a diagnosis of diabetes by a physician or use of glucose-lowering medication. BMI was calculated as weight in kilograms divided by height in meters squared. Hypertension was defined as blood pressure ≥ 140/90 mmHg or self-reported use of blood pressure–lowering medication. Smoking status was self-reported.
Extended Data Table 2 ∣.
Number of participants with dementia by different diagnostic criteria, the ARIC study (n = 3,252)
| Number of cases | |
|---|---|
| Dementia identified at study visits | 783 (24%) |
| 1. Expert diagnosis of dementia based on neuropsychological testing | 783 |
| 2. Computer algorithm diagnosis based on neuropsychological testing | |
| Dementia identified by phone interviews | 1,589 (49%) |
| 3. Low score on the TICS (adjusted for education level) | 779 |
| 4. Elevated score on the CDR and FAQ | |
| 5. Low score on the SIS and elevated score on the AD8 | 689 |
| 6. Two low scores on the SIS | 82 |
| 7. One low score on the SIS if participant was loss to follow up or deceased | 39 |
| Dementia identified by administrative records | 880 (27%) |
| 8. Dementia on hospital discharge codes | 685 |
| 9. Dementia on death certificate codes | 195 |
AD8, Ascertain Dementia 8; CDR, Clinical Dementia Rating; Functional Activities Questionnaire, FAQ; SIS, Six-Item Screener; TICS, Telephone Interview for Cognitive Status.
Extended Data Table 3 ∣.
Crude incidence for dementia and death free of dementia from age 55 years to age 95 years, the ARIC study, overall and by sex, race and APOE ε4 status (n = 15,043)
| Number of participants | Number of incident dementia cases |
Crude incidence for dementia, per 1,000 person-years |
Number of deaths without dementia |
Crude incidence for mortality without dementia, per 1,000 person-years |
|
|---|---|---|---|---|---|
| Overall | 15,043 | 3,252 | 10.3 | 5,803 | 18.4 |
| Sex | |||||
| Men | 6,751 | 1,327 | 9.9 | 3,145 | 23.4 |
| Women | 8,292 | 1,925 | 10.7 | 2,658 | 14.8 |
| Race | |||||
| White | 10,958 | 2,233 | 9.5 | 4,103 | 17.4 |
| Black | 4,041 | 1,012 | 13.0 | 1,690 | 21.8 |
| APOE ε4 status | |||||
| 0 alleles | 10,407 | 1,916 | 8.7 | 4,075 | 18.4 |
| 1 allele | 4,234 | 1,164 | 13.5 | 1,597 | 18.5 |
| 2 alleles | 402 | 172 | 22.7 | 131 | 17.3 |
Extended Data Table 4 ∣.
Lifetime risk of dementia (from age 55 years to age 95 years), by APOE ε4 status, sex and race
| APOE ε4 status | Difference (2 alleles – 0 alleles) |
|||
|---|---|---|---|---|
| 0 alleles | 1 allele | 2 alleles | ||
| White Women | 45 (42, 48) | 56 (51, 60) | 64 (52, 73) | 19 |
| White Men | 31 (29, 34) | 40 (37, 44) | 60 (49, 70) | 29 |
| Black Women | 44 (39, 48) | 53 (48, 58) | 67 (56, 76) | 23 |
| Black Men | 34 (29, 38) | 38 (33, 43) | 39 (26, 51) | 5 |
Estimates are reported as percentages and indicate the cumulative incidence at the age of last observation (up to age 95 years) after accounting for the competing risk of death. The 95% CIs are reported in parentheses.
Extended Data Table 5 ∣.
Lifetime risk of dementia after select index ages (ages 55, 65, 75 and 85) to 95 years, the ARIC study, overall and by sex, race and APOE ε4 status
| Index age | ||||
|---|---|---|---|---|
| 55 years (n=15,043) |
65 years (n=13,830) |
75 years (n=10,599) |
85 years (n=2,940) |
|
| Overall | 42 (41, 43) | 45 (44, 47) | 52 (50, 54) | 56 (53, 59) |
| Sex | ||||
| Men | 35 (33, 36) | 38 (36, 40) | 45 (43, 48) | 51 (46, 55) |
| Women | 48 (46, 50) | 51 (49, 53) | 57 (54, 59) | 60 (56, 64) |
| Race | ||||
| White | 41 (40, 43) | 44 (42, 46) | 50 (48, 52) | 55 (51, 58) |
| Black | 44 (41, 46) | 49 (46, 51) | 58 (55, 62) | 63 (56, 69) |
| APOE ε4 status | ||||
| 0 alleles | 39 (37, 40) | 42 (40, 43) | 48 (46, 50) | 53 (50, 57) |
| 1 allele | 48 (45, 50) | 51 (49, 54) | 60 (57, 63) | 64 (58, 69) |
| 2 alleles | 59 (53, 65) | 65 (59, 71) | 74 (66, 81) | 67 (47, 81) |
Estimates are reported as percentages and indicate the cumulative incidence at the age of last observation (up to age 95 years) after accounting for the competing risk of death. The 95% CIs are reported in parentheses.
Extended Data Table 6 ∣.
Median age at dementia diagnosis and distribution of diagnosis age, the ARIC study, overall and by sex, race and APOE ε4 status (n = 3,252)
| Median age (IQR) at diagnosis, years |
Percentage diagnosed between ages 55 and 74 |
Percentage diagnosed between ages 75 and 84 |
Percentage diagnosed between ages 85 and 95 |
|
|---|---|---|---|---|
| Overall | 81 (77, 86) | 17% | 53% | 30% |
| Sex | ||||
| Men | 81 (76, 85) | 20% | 53% | 28% |
| Women | 82 (77, 86) | 15% | 54% | 31% |
| Race | ||||
| White | 82 (78, 86) | 14% | 53% | 34% |
| Black | 79 (75, 84) | 25% | 55% | 20% |
| APOE ε4 status | ||||
| 0 alleles | 82 (77, 86) | 16% | 49% | 35% |
| 1 allele | 81 (77, 85) | 17% | 59% | 24% |
| 2 alleles | 79 (75, 82) | 28% | 59% | 13% |
Percentages may not add up to 100% due to rounding. IQR, interquartile range.
Supplementary Material
Supplementary information The online version contains supplementary material available at https://doi.org/10.1038/s41591-024-03340-9.
Acknowledgements
This study was supported by National Institutes of Health (NIH)/National Heart, Lung, and Blood Institute (NHLBI) grant K24 HL152440 (to E.S.); NIH/National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) grant K01 DK138273 (to M.F.); NIH/National Institute on Aging (NIA) grant R01AG054787 (to B.G.W.); the National Institute of Neurological Disorders and Stroke (NINDS) Intramural Research Program (to R.F.G.); and the NIA Intramural Research Program (to K.A.W.). The ARIC study is carried out as a collaborative study supported by NHLBI contracts 75N92022D00001, 75N92022D00002, 75N92022D00003, 75N92022D00004 and 75N92022D00005. The ARIC Neurocognitive Study is supported by U01HL096812, U01HL096814, U01HL096899, U01HL096902 and U01HL096917 from the NIH. The funders had no role in the design and conduct of the study; in collection, management, analysis and interpretation of the data; in preparation, review or approval of the manuscript; and in the decision to submit the manuscript for publication. We thank the staff and participants of the ARIC study for their important contributions. We also thank K. Da Silva and J. Pike for their valuable suggestions and help during revisions.
Competing interests
D.S.K. has no disclosures relevant to the current work. D.S.K. serves on a Data Safety Monitoring Board for the Dominantly Inherited Alzheimer Network Treatment Unit study. He was an investigator in Alzheimer clinical trials sponsored by Biogen, Eli Lilly Pharmaceuticals and the University of Southern California and is currently an investigator in a trial in frontotemporal degeneration with Alector. He has served as a consultant for Roche, AriBio, Linus Health, Biovie and Alzeca Biosciences but receives no personal compensation. He receives funding from the NIH. The other authors declare no competing interests.
Footnotes
Online content
Any methods, additional references, Nature Portfolio reporting summaries, source data, extended data, supplementary information, acknowledgements, peer review information; details of author contributions and competing interests; and statements of data and code availability are available at https://doi.org/10.1038/s41591-024-03340-9.
Extended data is available for this paper at https://doi.org/10.1038/s41591-024-03340-9.
Reporting summary
Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.
Data availability
ARIC data access procedures are in accordance with participant informed consent and NIH data-sharing policy. Anonymized data from the ARIC study are available at the NHLBI Biologic Specimen and Data Repository Information Coordinating Center and can be accessed through the website (https://biolincc.nhlbi.nih.gov/studies/aric/). Requests for access of ARIC data may also be submitted to the ARIC Publications Committee according to established study procedures, which include submission of a completed ARIC Manuscript Proposal Form (available at https://aric.cscc.unc.edu/aric9/publications/policies_forms_and_guidelines) to the ARIC Publications Committee at aricpub@unc.edu. Review and approval of data access requests typically takes approximately 1 month.
Code availability
The analytic code used for analyses in this study is available from the corresponding author upon reasonable request.
References
- 1.Caplan Z. US older population grew from 2010 to 2020 at fastest rate since 1880 to 1890. United States Census Bureau. https://www.census.gov/library/stories/2023/05/2020-census-united-states-older-population-grew.html (2023). [Google Scholar]
- 2.Rajan KB et al. Population estimate of people with clinical Alzheimer’s disease and mild cognitive impairment in the United States (2020–2060). Alzheimers Dement. 17, 1966–1975 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Murphy S, Kochanek K, Xu J & Arias E Mortality in the United States, 2020. NCHS Data Brief, No. 427 National Center for Health Statistics. https://www.cdc.gov/nchs/data/databriefs/db427.pdf (2021). [PubMed] [Google Scholar]
- 4.Nandi A. et al. Global and regional projections of the economic burden of Alzheimer’s disease and related dementias from 2019 to 2050: a value of statistical life approach. EClinicalMedicine 51, 101580 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.2023 Alzheimer’s disease facts and figures. Alzheimers Dement. 19, 1598–1695 (2023). [DOI] [PubMed] [Google Scholar]
- 6.Seshadri S & Wolf PA Lifetime risk of stroke and dementia: current concepts, and estimates from the Framingham Study. Lancet Neurol. 6, 1106–1114 (2007). [DOI] [PubMed] [Google Scholar]
- 7.Martin SS et al. 2024 heart disease and stroke statistics: a report of US and global data from the American Heart Association. Circulation 149, e347–e913 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Chêne G. et al. Gender and incidence of dementia in the Framingham Heart Study from mid-adult life. Alzheimers Dement. 11, 310–320 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Beiser A, Au R, McNulty K, White R & D’agostino R Lifetime risk of dementia and Alzheimer’s disease: the impact of mortality on risk estimates in the Framingham Study. Neurology 49, 1498–1504 (1997). [DOI] [PubMed] [Google Scholar]
- 10.Licher S. et al. Lifetime risk of common neurological diseases in the elderly population. J. Neurol. Neurosurg. Psychiatry 90, 148–156 (2019). [DOI] [PubMed] [Google Scholar]
- 11.Chatfield MD, Brayne CE & Matthews FE A systematic literature review of attrition between waves in longitudinal studies in the elderly shows a consistent pattern of dropout between differing studies. J. Clin. Epidemiol 58, 13–19 (2005). [DOI] [PubMed] [Google Scholar]
- 12.Livingston G. et al. Dementia prevention, intervention, and care: 2020 report of the Lancet Commission. Lancet 396, 413–446 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Ngandu T. et al. A 2 year multidomain intervention of diet, exercise, cognitive training, and vascular risk monitoring versus control to prevent cognitive decline in at-risk elderly people (FINGER): a randomised controlled trial. Lancet 385, 2255–2263 (2015). [DOI] [PubMed] [Google Scholar]
- 14.van Charante EPM et al. Effectiveness of a 6-year multidomain vascular care intervention to prevent dementia (preDIVA): a cluster-randomised controlled trial. Lancet 388, 797–805 (2016). [DOI] [PubMed] [Google Scholar]
- 15.Lin FR et al. Hearing intervention versus health education control to reduce cognitive decline in older adults with hearing loss in the USA (ACHIEVE): a multicentre, randomised controlled trial. Lancet 402, 786–797 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Coley N, Giulioli C, Aisen PS, Vellas B & Andrieu S Randomised controlled trials for the prevention of cognitive decline or dementia: a systematic review. Ageing Res. Rev 82, 101777 (2022). [DOI] [PubMed] [Google Scholar]
- 17.Williamson JD et al. Effect of intensive vs standard blood pressure control on probable dementia: a randomized clinical trial. JAMA 321, 553–561 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Lloyd-Jones DM et al. Status of cardiovascular health in US adults and children using the American Heart Association’s new ‘Life’s Essential 8’ metrics: prevalence estimates from the National Health and Nutrition Examination Survey (NHANES), 2013 through 2018. Circulation 146, 822–835 (2022). [DOI] [PubMed] [Google Scholar]
- 19.Reed NS et al. Prevalence of hearing loss and hearing aid use among US Medicare beneficiaries aged 71 years and older. JAMA Netw. Open 6, e2326320 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Qian J. et al. APOE-related risk of mild cognitive impairment and dementia for prevention trials: an analysis of four cohorts. PLoS Med. 14, e1002254 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Yamazaki Y, Zhao N, Caulfield TR, Liu C-C & Bu G Apolipoprotein E and Alzheimer disease: pathobiology and targeting strategies. Nat. Rev. Neurol 15, 501–518 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Fortea J. et al. APOE4 homozygozity represents a distinct genetic form of Alzheimer’s disease. Nat. Med 30, 1284–1291 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Matthews KA et al. Racial and ethnic estimates of Alzheimer’s disease and related dementias in the United States (2015–2060) in adults aged ≥65 years. Alzheimers Dement. 15, 17–24 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Weuve J. et al. Cognitive aging in black and white Americans: cognition, cognitive decline, and incidence of Alzheimer disease dementia. Epidemiology 29, 151 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Ferretti MT et al. Sex differences in Alzheimer disease—the gateway to precision medicine. Nat. Rev. Neurol 14, 457–469 (2018). [DOI] [PubMed] [Google Scholar]
- 26.Ott A, Breteler MM, Harskamp FV, Stijnen T & Hofman A Incidence and risk of dementia: the Rotterdam Study. Am. J. Epidemiol 147, 574–580 (1998). [DOI] [PubMed] [Google Scholar]
- 27.Feigin VL et al. Burden of neurological disorders across the US from 1990–2017: a global burden of disease study. JAMA Neurol. 78, 165–176 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Vespa J, Armstrong DM & Medina L Demographic Turning Points for the United States: Population Projections for 2020 to 2060 (U.S. Department of Commerce, Economics and Statistics Administration, U.S. Census Bureau, 2018). [Google Scholar]
- 29.Wilkinson T. et al. Identifying dementia cases with routinely collected health data: a systematic review. Alzheimers Dement. 14, 1038–1051 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Lin PJ et al. Dementia diagnosis disparities by race and ethnicity: health services research: cost of care and implications for intervention. Alzheimers Dement. 16, e043183 (2020). [Google Scholar]
- 31.Office of the Assistant Secretary for Planning and Evaluation. US Department of Health and Human Services. National Plan to Address Alzheimer’s Disease. https://aspe.hhs.gov/collaborations-committees-advisory-groups/napa/napa-documents/napa-national-plan (2023).
- 32.Wright JD et al. The ARIC (Atherosclerosis Risk in Communities) Study: JACC Focus Seminar 3/8. J. Am. Coll. Cardiol 77, 2939–2959 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Knopman DS et al. Mild cognitive impairment and dementia prevalence: the atherosclerosis risk in communities neurocognitive study. Alzheimers Dement. (Amst.) 2, 1–11 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Knopman DS et al. Patterns of cognitive domain abnormalities enhance discrimination of dementia risk prediction: the ARIC study. Alzheimers Dement. 20, 4559–4571 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Schneider AL et al. Normative data for 8 neuropsychological tests in older blacks and whites from the Atherosclerosis Risk in Communities (ARIC) study. Alzheimer Dis. Assoc. Disord 29, 32–44 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Morris JC The Clinical Dementia Rating (CDR): current version and scoring rules. Neurology 43, 2412–2414 (1993). [DOI] [PubMed] [Google Scholar]
- 37.Folstein MF, Folstein SE & McHugh PR ‘Mini-mental state’: a practical method for grading the cognitive state of patients for the clinician. J. Psychiatr. Res 12, 189–198 (1975). [DOI] [PubMed] [Google Scholar]
- 38.Pfeffer RI, Kurosaki TT, Harrah C Jr, Chance JM & Filos S Measurement of functional activities in older adults in the community. J. Gerontol 37, 323–329 (1982). [DOI] [PubMed] [Google Scholar]
- 39.Blessed G, Tomlinson BE & Roth M The association between quantitative measures of dementia and of senile change in the cerebral grey matter of elderly subjects. Br. J. Psychiatry 114, 797–811 (1968). [DOI] [PubMed] [Google Scholar]
- 40.McKhann GM et al. The diagnosis of dementia 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, 263–269 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Brandt J, Spencer M & Folstein M The telephone interview for cognitive status. Neuropsychiatry Neuropsychol. Behav. Neurol 1, 111–117 (1988). [Google Scholar]
- 42.Callahan CM, Unverzagt FW, Hui SL, Perkins AJ & Hendrie HC Six-item screener to identify cognitive impairment among potential subjects for clinical research. Med. Care 40, 771–781 (2002). [DOI] [PubMed] [Google Scholar]
- 43.Lin D. Non-parametric inference for cumulative incidence functions in competing risks studies. Stat. Med 16, 901–910 (1997). [DOI] [PubMed] [Google Scholar]
- 44.Stata 18 documentation. https://www.stata-press.com/manuals/documentation-set/
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
ARIC data access procedures are in accordance with participant informed consent and NIH data-sharing policy. Anonymized data from the ARIC study are available at the NHLBI Biologic Specimen and Data Repository Information Coordinating Center and can be accessed through the website (https://biolincc.nhlbi.nih.gov/studies/aric/). Requests for access of ARIC data may also be submitted to the ARIC Publications Committee according to established study procedures, which include submission of a completed ARIC Manuscript Proposal Form (available at https://aric.cscc.unc.edu/aric9/publications/policies_forms_and_guidelines) to the ARIC Publications Committee at aricpub@unc.edu. Review and approval of data access requests typically takes approximately 1 month.
The analytic code used for analyses in this study is available from the corresponding author upon reasonable request.