Aging - extent and prevalence of cognitive decline

Most older adults do not suffer cognitive impairment. Around 30-40% of adults over 65 have the type of cognitive loss we regard as a normal consequence of age — a measurable (but slight) decline on memory tests; a feeling that you're not quite as sharp or as good at remembering, as you used to be (age-related cognitive impairment). Around 10% of adults over 65 develop mild cognitive impairment (MCI), which does impact everyday living, and is a precursor of Alzheimer's.

There are significant differences in prevalence as a function of age. For example, in the U.S., a large sample found MCI in 9% of those aged 70 to 79 and nearly 18% of those 80 to 89. Prevalence decreased with years of education: 25% in those with up to eight years of education, 14% in those with nine to 12 years, 9% in those with 13 to 16 years, and 8.5% in those with greater than 16 years.

Large-scale population surveys of mild cognitive impairment in the elderly have produced large differences in national levels, ranging from 10% to 26%.

Although women may decline at a faster rate than men, prevalence of decline may be greater among men. For example, a large Dutch survey of those aged 85 and older found more women than men had good memory (41% vs 29%) and mental speed (33% vs 28%), despite the fact that more women than men had a limited education.

However, severe memory problems in the elderly have become more rare. The main reasons seem to be better physical fitness (partly due to better healthcare), higher levels of education, and greater personal wealth.

As we all know, people are living longer and obesity is at appalling levels. For both these (completely separate!) reasons, we expect to see growing rates of dementia. A new analysis using data from the long-running Framingham Heart Study offers some hope to individuals, however.

Looking at the rate of dementia during four distinct periods in the late 1970s, late 1980s, 1990s, and 2000s, using data from 5205 older adults (60+), the researchers found that there was a progressive decline in the incidence of dementia at a given age, with an average reduction of 20% per decade since the 1970s (22%, 38%, and 44% during the second, third, and fourth epochs, respectively).

There are two important things to note about this finding:

  • the decline occurred only in people with a high school education and above
  • the decline was more pronounced with dementia caused by vascular diseases, such as stroke.

The cumulative risk over five years, adjusted for age and gender, were:

  • 3.6 per 100 persons during the first period (late 1970s and early 1980s)
  • 2.8 per 100 persons during the second period (late 1980s and early 1990s)
  • 2.2 per 100 persons during the third period (late 1990s and early 2000s)
  • 2.0 per 100 persons during the fourth period (late 2000s and early 2010s).

Part of the reason for the decline is put down to the decrease in vascular risk factors other than obesity and diabetes, and better management of cardiovascular diseases and stroke. But this doesn't completely explain the decrease. I would speculate that other reasons might include:

  • increased mental stimulation
  • improvements in lifestyle factors such as diet and exercise
  • better health care for infectious and inflammatory conditions.

The finding is not completely unexpected. Recent epidemiological studies in the U.S., Canada, England, the Netherlands, Sweden and Denmark have all suggested that “a 75- to 85-year-old has a lower risk of having Alzheimer’s today than 15 or 20 years ago.” Which actually cuts to the heart of the issue: individual risk of dementia has gone down (for those taking care of their brain and body), but because more and more people are living longer, the numbers of people with dementia are increasing.

A study involving 14 years of health records from more than 274,000 Northern Californians has assessed the relative dementia risk of six different ethnicities.

The average annual rate of dementia was:

  • 26.6 cases per 1,000 for blacks
  • 22.2 cases per 1,000 for American Indians/Alaskan Natives
  • 19.6 cases per 1,000 for Latinos and Pacific Islanders
  • 19.3 cases per 1,000 for whites
  • 15.2 cases per 1,000 for Asian Americans.

But this is an annual rate, not particularly useful at a practical level. How do these numbers convert to lifetime risk? Statistical calculations estimate that among those who reach age 65 dementia-free, the following percentages of each ethnicity will develop dementia in the next 25 years:

  • 38% of blacks
  • 35% of American Indians/Alaskan Natives
  • 32% of Latinos
  • 30% of whites
  • 28% of Asian Americans
  • 25% of Pacific Islanders (this is probably the least reliable number, given the small number of Pacific Islanders in the sample).

The study population included 18,778 African-Americans, 4543 American Indians/Alaskan Natives, 21,000 Latinos, 206,490 white Americans, 23,032 Asian-Americans, and 440 Pacific Islanders.

A long-running study comparing African-Americans and Nigerians has found the incidence of dementia has fallen significantly over two decades among the African-Americans, but remained the same for the Nigerians (for whom it was lower anyway).

The study enrolled two cohorts, one in 1992 and one in 2001, who were evaluated every 2–3 years until 2009. The 1992 cohort included 1440 older African-Americans (70+) and 1774 Nigerian Yoruba; the 2001 cohort included 1835 African-Americans and 1895 Yoruba. None of the participants had dementia at study beginning.

The overall standardized annual incidence rate was 3.6% for the 1992 African-American cohort, and 1.4% for the 2001 cohort. For the Yoruba, it was 1.7% and 1.4%, respectively.

It's suggested that one reason for the improvement among African-Americans may be medications for cardiovascular conditions. Although both groups had similar rates of high blood pressure, this was recognized and treated in the American group but not in the Nigerian.

As you can see, African-Americans in the earlier cohort were more than twice as likely as Africans to develop dementia. Their decrease has brought them into line with the African rate.

Although the rate of new cases of dementia decreased, the African-Americans enrolling in 2001 had significantly higher rates of diabetes, hypertension and stroke, but also higher treatment rates, than the African-Americans who enrolled in 1992.

The finding offers hope that treatment can offset the expected increase in dementia resulting from the rise in lifestyle diseases.

New research supports the classification system for preclinical Alzheimer’s proposed two years ago. The classification system divides preclinical Alzheimer's into three stages:

Stage 1: Levels of amyloid beta begin to decrease in the spinal fluid. This indicates that the substance is beginning to form plaques in the brain.

Stage 2: Levels of tau protein start to increase in the spinal fluid, indicating that brain cells are beginning to die. Amyloid beta levels are still abnormal and may continue to fall.

Stage 3: In the presence of abnormal amyloid and tau biomarker levels, subtle cognitive changes can be detected by neuropsychological testing.

Long-term evaluation of 311 cognitively healthy older adults (65+) found 31% with preclinical Alzheimer’s, of whom 15% were at stage 1, 12% at stage 2, and 4% at stage 3. This is consistent with autopsy studies, which have shown that around 30% of cognitively normal older adults die with some preclinical Alzheimer's pathology in their brain. Additionally, 23% were diagnosed with suspected non-Alzheimer pathophysiology (SNAP), 41% as cognitively normal, and 5% as unclassified.

Five years later, 2% of the cognitively normal, 5% of those with SNAP, 11% of the stage 1 group, 26% of the stage 2 group, and 56% of the stage 3 group had been diagnosed with symptomatic Alzheimer's.

[3614] Vos SJB, Xiong C, Visser PJ, Jasielec MS, Hassenstab J, Grant EA, Cairns NJ, Morris JC, Holtzman DM, Fagan AM. Preclinical Alzheimer's disease and its outcome: a longitudinal cohort study. The Lancet Neurology [Internet]. 2013 ;12(10):957 - 965. Available from:

A survey of 7,072 older adults in six provinces across China, with one rural and one urban community in each province, has identified 359 older adults with dementia and 328 with depression. There were only 26 participants who had doctor-diagnosed dementia reported and 26 who had doctor-diagnosed depression. Overall, 93% of dementia cases and 93% of depression were not detected.

Undetected dementia was strongly associated with low socioeconomic status such as a low educational and occupational class, and living in a rural area.

In comparison, research in high income countries has found that about 60% of older adults with dementia are not diagnosed, and generally there has not been a strong association between low socioeconomic status and undetected dementia. One factor in China’s high rate may be that most older Chinese live with their families, who may be inclined to see dementia as a normal part of aging.


New estimates of dementia in China

A new review of 89 studies, involving more than 340,000 participants in total, has estimated that 9.19 million people in China had dementia in 2010, of whom 5.69 million had Alzheimer’s disease. Previous studies appear to have considerably underestimated the true burden of dementia in China, largely due to limited data availability. However, this study examined a much wider range of data sources than earlier studies, including many Chinese-language reports.

Of the 340 247 participants, 6357 had Alzheimer's disease (1.87%). Of 254,367 assessed for other forms of dementia, 3543 (1.4%) had vascular dementia, frontotemporal dementia, or Lewy body dementia.

Total dementia prevalence in 1990 was 1·8% at 65-69 years, and 42·1% at age 95-99 years; in 2010, prevalence had increased to 2·6% and 60·5%, respectively.

Prevalence was higher for women than men, but didn't differ significantly between urban and rural residents.

[3561] Chen R, Hu Z, Chen R-L, Ma Y, Zhang D, Wilson K. Determinants for undetected dementia and late-life depression. The British Journal of Psychiatry [Internet]. 2013 ;203(3):203 - 208. Available from:

[3559] Chan KY, Wang W, Wu JJ, Liu L, Theodoratou E, Car J, Middleton L, Russ TC, Deary IJ, Campbell H, et al. Epidemiology of Alzheimer's disease and other forms of dementia in China, 1990–2010: a systematic review and analysis. The Lancet [Internet]. 2013 ;381(9882):2016 - 2023. Available from:

A survey of 7796 older adults (65+) living in three geographic areas in England has allowed us to compare dementia rates over time, with an identical survey having been taken between 1989 and 1994. The overall prevalence of dementia fell significantly, from 8.3% to 6.5%.

The finding provides further evidence that a cohort effect exists in dementia prevalence.

[3589] Matthews FE, Arthur A, Barnes LE, Bond J, Jagger C, Robinson L, Brayne C. A two-decade comparison of prevalence of dementia in individuals aged 65 years and older from three geographical areas of England: results of the Cognitive Function and Ageing Study I and II. The Lancet [Internet]. 2013 ;382(9902):1405 - 1412. Available from:

Functional impairment good indicator of mild cognitive impairment

Evaluation of 816 older adults, of whom 229 had no cognitive problems, 394 had a diagnosis of amnestic mild cognitive impairment, and 193 had a diagnosis of mild Alzheimer’s, has revealed that most of those with aMCI (72%) or AD (97%) had trouble with at least one type of function on the Pfeffer Functional Activities Questionnaire. Only 8% of controls had any difficulty. In both impaired groups, those who had the most difficulty functioning also tended to score worse on cognition tests, have smaller hippocampal volumes, and carry the APOe4 gene.

Two of the ten items in the questionnaire were specific in differentiating the control group from the impaired groups. Those items concerned "remembering appointments, family occasions, holidays, and medications” and "assembling tax records, business affairs, or other papers." Only 34% of those with aMCI and 3.6% of those with AD had no difficulty with these items.

The findings suggest that even mild disruptions in daily functioning may be an important clinical indicator of disease.

Early-onset Alzheimer’s poorly diagnosed when initial symptoms aren’t memory related

Post-mortem analysis of 40 people diagnosed  with early-onset Alzheimer’s has revealed that about 38% experienced initial symptoms other than memory problems, such as behavior, vision or language problems and a decline in executive function, or the ability to carry out tasks. Of these, 53% were incorrectly diagnosed when first seen by a doctor, compared to 4% of those who had memory problems. Of those with unusual initial symptoms, 47% were still incorrectly diagnosed at the time of their death.

The mean age at onset was 54.5 years (range 46-60). The average duration of the disease was 11 years, with an average diagnostic delay of 3 years.

GPs misidentify and fail to identify early dementia and MCI

A review of 30 studies involving 15,277 people seen in primary care for cognitive disorders, has found that while GPs managed to identify eight out of ten people with moderate to severe dementia, they only identified 45% of those with early dementia and mild cognitive impairment. Moreover, they were very poor at recording such diagnoses. Thus, though they recognized 45% of the MCI cases, they only recorded 11% of these cases in their medical notes. Although they identified 73% of people with dementia, they made correct annotations in medical records in only 38% of cases.

But the problem is not simply one of failing to diagnose — they were even more likely to misidentify dementia, and this was particularly true for those with depression or hearing problems.

The findings point to the need for more widespread use of simple cognitive screening tests.

Prevalence of dementia & MCI in 'oldest old' women

Data from 1,299 women enrolled in the Women Cognitive Impairment Study of Exceptional Aging suggests that the incidence of dementia almost doubles with every 5 years of age and prevalence rises from approximately 2-3% in those 65 to 75 years to 35% in those 85+.

Among those with mild cognitive impairment, amnestic multiple domain was most common (34%), followed by non-amnestic single domain (29%). Amnestic single domain (affecting only one type of cognitive function, including memory difficulty) affected 22%.

Alzheimer's disease and mixed dementia accounted for nearly 80% of dementia cases, and vascular dementia for 12.1%.

Those with dementia tended to be older, less likely to have completed high school, more likely to have reported depression, a history of stroke, and to have the APOEe4 gene.

The women in the study had an average age of 88.2 years and 27% were older than 90. 41% had clinical cognitive impairment (17.8% with dementia and 23.2% with mild cognitive impairment).

The high prevalence of cognitive impairment in this age group points to the importance of screening for cognitive disorders, particularly among high-risk groups.

Beginning in 1971, healthy older adults in Gothenburg, Sweden, have been participating in a longitudinal study of their cognitive health. The first H70 study started in 1971 with 381 residents of Gothenburg who were 70 years old; a new one began in 2000 with 551 residents and is still ongoing. For the first cohort (born in 1901-02), low scores on non-memory tests turned out to be a good predictor of dementia; however, these tests were not predictive for the generation born in 1930. Those from the later cohort also performed better in the intelligence tests at age 70 than their predecessors had.

It’s suggested that the higher intelligence is down to the later cohort’s better pre and postnatal care, better nutrition, higher quality education, and better treatment of high blood pressure and cholesterol. And possibly the cognitive demands of modern life.

Nevertheless, the researchers reported that the incidence of dementia at age 75 was little different (5% in the first cohort and 4.4% in the later). However, since a substantially greater proportion of the first cohort were dead by that age (15.7% compared to 4.4% of the 2nd cohort), it seems quite probable that there really was a higher incidence of dementia in the earlier cohort.

The fact that low scores on non-memory cognitive tests were predictive in the first cohort of both dementia and death by age 75 supports this argument.

The fact that low scores on non-memory cognitive tests were not predictive of dementia or death in the later cohort is in keeping with the evidence that higher levels of education help delay dementia. We will need to wait for later findings from this study to see whether that is what is happening.

The findings are not inconsistent with those from a very large U.S. national study that found older adults (70+) are now less likely to be cognitively impaired (see below). It was suggested then also that better healthcare and more education were factors behind this decline in the rate of cognitive impairment.

Previous study:

A new nationally representative study involving 11,000 people shows a downward trend in the rate of cognitive impairment among people aged 70 and older, from 12.2% to 8.7% between 1993 and 2002. It’s speculated that factors behind this decline may be that today’s older people are much likelier to have had more formal education, higher economic status, and better care for risk factors such as high blood pressure, high cholesterol and smoking that can jeopardize their brains. In fact the data suggest that about 40% of the decrease in cognitive impairment over the decade was likely due to the increase in education levels and personal wealth between the two groups of seniors studied at the two time points. The trend is consistent with a dramatic decline in chronic disability among older Americans over the past two decades.

A review of brain imaging and occupation data from 588 patients diagnosed with frontotemporal dementia has found that among the dementias affecting those 65 years and younger, FTD is as common as Alzheimer's disease. The study also found that the side of the brain first attacked (unlike Alzheimer’s, FTD typically begins with tissue loss in one hemisphere) is influenced by the person’s occupation.

Using occupation scores that reflect the type of skills emphasized, they found that patients with professions rated highly for verbal skills, such as school principals, had greater tissue loss on the right side of the brain, whereas those rated low for verbal skills, such as flight engineers, had greater tissue loss on the left side of the brain. This effect was expressed most clearly in the temporal lobes of the brain. In other words, the side of the brain least used in the patient's professional life was apparently the first attacked.

These findings are in keeping with the theory of cognitive reserve, but may be due to some asymmetry in the brain that both inclines them to a particular occupational path and renders the relatively deficient hemisphere more vulnerable in later life.

Data from 330 participants in The 90+ Study, of whom 70% were women, has revealed an overall annual incidence rate of 18.2% for dementia, rising from 12.7% per year in the 90-94 age group, to 21.2% in the 95-99 age group and 40.7% per year in the 100+ age group. 60% of the cases were attributed to Alzheimer's disease, 22% vascular dementia, 9% mixed Alzheimer’s and vascular dementia and 9% other/unknown. Unlike previous findings, rates were very similar for men and women.

Older news items (pre-2010) brought over from the old website

Most older people with mild cognitive impairment have Alzheimer's or cerebral vascular disease

Another finding from the Religious Orders Study. It seems that mild cognitive impairment is often the earliest clinical manifestation of Alzheimer’s or vascular dementia. By studying the brains of study participants after death, researchers could ascertain that, of the 37 individuals with mild cognitive impairment, 23 met pathologic criteria for Alzheimer's disease, and 12 had cerebral infarcts (5 had both). Only 9 did not have either pathology. The researchers conclude that even mild loss of cognitive function in older people should not, therefore, be viewed as normal, but as an indication of a disease process.

[893] Bennett DA, Schneider JA, Bienias JL, Evans DA, Wilson RS. Mild cognitive impairment is related to Alzheimer disease pathology and cerebral infarctions. Neurology [Internet]. 2005 ;64(5):834 - 841. Available from:

Population level of frontotemporal dementia

A large-scale epidemiological study in the Netherlands has found an incidence of frontotemporal dementia that equates to a population level of 1.1 per 100,000. The prevalence was highest among those ages 60 to 69, at 9.4 per 100,000. The prevalence among people ages 45 to 64 was estimated to be 6.7 per 100,000. Symptoms began after age 65 in 22% of patients. Whites accounted for 99% of all cases despite an ample nonwhite population. A family history of dementia was present in 43% of patients.

[586] Ravid R, Niermeijer MF, Verheij F, Kremer HP, Scheltens P, van Duijn CM, Heutink P, van Swieten JC, Rosso SM, Kaat LD, et al. Frontotemporal dementia in The Netherlands: patient characteristics and prevalence estimates from a population-based study. Brain [Internet]. 2003 ;126(9):2016 - 2022. Available from:

Cognitive impairment high among older people

In the first population-based study of cognitive impairment in the United States, nearly one in four older African Americans in Indianapolis were found to have measurable cognitive problems (short of dementia or Alzheimer's). The prevalence of cognitive impairment grew significantly with age, with rates increasing by about 10 percent for every 10 years of age after age 65. Of those aged 85 and older, 38% had some degree of cognitive impairment. Surveys in other countries (which cannot be directly compared due to differences in methodology, diagnostic criteria, etc) have reported results ranging from 10.7% in Italy to 26.6% in Finland.

[992] Musick B, Hall KS, Hui SL, Hendrie HC, Unverzagt FW, Gao S, Baiyewu O, Ogunniyi AO, Gureje O, Perkins A, et al. Prevalence of cognitive impairment: Data from the Indianapolis Study of Health and Aging. Neurology [Internet]. 2001 ;57(9):1655 - 1662. Available from:

More women than men do well on memory tests in old age

Researchers from Leiden University tested the mental functioning of 599 Dutch men and women aged 85 years. Good mental speed on word and number recognition tests was found in 33% of the women and 28% of the men. Forty one per cent of the women and 29% of the men had a good memory. This despite the fact that significantly more of the women had limited formal education compared to the men (not surprising given the time in which they grew up). The authors suggested that biological differences - such as the relative absence of cardiovascular disease in elderly women compared with men of the same age - could account for these sex differences in mental decline.

[2615] van Exel E, Gussekloo J, de Craen AJM, Bootsma-van der Wiel A, Houx P, Knook DL, Westendorp RGJ. Cognitive function in the oldest old: women perform better than men. Journal of Neurology, Neurosurgery & Psychiatry [Internet]. 2001 ;71(1):29 - 32. Available from:

Severe memory problems in older adults have become more rare

Severe memory problems in older adults have become more rare, probably because of better treatments for dementia, depression and strokes. Researchers from the University of Michigan interviewed more than 10,000 people ages 70 and older from 1993 to 1998. People tested in 1998 did significantly better on the memory tests than those tested earlier. In 1998 less than 4% of those 70 and older showed severe memory problems, and only 8% of those 85 and older. Surprisingly, the greatest improvement was seen among those in their 80s and those with less than a high school education. The decline in memory problems is believed to be associated with the improvement in physical fitness seen among the elderly. It is speculated that the increase in number of women on hormone replacement therapy may also play a part.

[2616] Freedman VA, Aykan H, Martin LG. Aggregate Changes in Severe Cognitive Impairment Among Older Americans. The Journals of Gerontology Series B: Psychological Sciences and Social Sciences [Internet]. 2001 ;56(2):S100 -S111 - S100 -S111. Available from:

Add comment

Filtered HTML

  • Web page addresses and e-mail addresses turn into links automatically.
  • Allowed HTML tags: <a> <em> <strong> <cite> <blockquote> <code> <ul> <ol> <li> <dl> <dt> <dd>
  • Lines and paragraphs break automatically.

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.
By submitting this form, you accept the Mollom privacy policy.