What was once just an idea now seems more certain: In developed countries, the number of new cases of dementia is on the decline. In the February 11 New England Journal of Medicine, researchers led by Sudha Seshadri at the Boston University School of Medicine report findings from the Framingham Heart Study that reveal a steadily falling dementia incidence, with rates cut almost in half over four decades. The decline correlated with improvements in cardiovascular health, but this alone did not explain the whole effect. The findings add weight to other recent studies from the United States and Europe that reported fewer new cases and a higher average age at disease onset.

“There’s an emerging consensus that the rates are going down. People are beginning to believe it,” said Eric Larson at Group Health Research Institute, Seattle. He was not involved in the study.

At the same time, many questions remain, including how widespread this trend is and how much it could affect the worldwide prevalence of the disease, predicted to hit 130 million by 2050 and driven mostly by cases in lower and middle-income countries (see Aug 2015 news). Because of the aging of the global population, most researchers still predict a boom in overall dementia, even with fewer new cases in some nations. Researchers also want to know what factors underlie the trend, and whether the findings will hold in different ethnic groups.

Studies of dementia incidence have tended to follow people of European ancestry, but new data from the U.S. population suggests that risk varies with ethnicity. For example, researchers led by Rachel Whitmer at the University of California, San Francisco, calculated that dementia risk runs higher for African-Americans than for whites or Asian-Americans. The data appeared in the February 11 Alzheimer’s & Dementia. Scientists agree more studies need to investigate whether incidence is also changing in these groups.

While the drops in incidence seen to date may not budge the worldwide disease burden, the real importance of the findings, researchers stressed, is that they suggest public health improvements can delay or even prevent dementia. “If we can elucidate the changes that have contributed to these improvements, perhaps we can extend them. … Primary and secondary prevention might diminish the magnitude of the long-feared dementia epidemic,” wrote David Jones at Harvard University and Jeremy Greene at Johns Hopkins, Baltimore, in an accompanying NEJM editorial.

Firming Up the Evidence
The evidence for declining rates of dementia has grown over the last decade, although much of it was indirect or statistically weak. Some studies inferred a drop in incidence from prevalence data, while other short-term studies of 10 years or so reported trends that just missed statistical significance (see May 2013 newsJul 2013 conference newsWu et al., 2016). The Framingham study now strengthens the evidence by covering a period of nearly 40 years, finding consistent, statistically significant declines in each decade. 

First author Claudia Satizabal analyzed data collected from 5,205 participants during four five-year periods from the late 1970s to the early 2010s. The volunteers were 60 years or older. In each of the three most recent periods, new dementia cases fell by about 20 percent relative to the previous period. This is comparable to the decline of 20-30 percent per decade reported by other recent studies. By the last time period, new dementia cases in the Framingham cohort had dropped by a total of 44 percent. In addition, the average age at diagnosis climbed from 80 to 85 years. Satizabal presented preliminary findings from this data set in 2014 (see Jul 2014 conference news). 

These findings jibe with a recently published study from Germany. Researchers led by Gabriele Doblhammer at the German Center for Neurodegenerative Disease, Bonn, used health insurance claim data to compare dementia incidence in 2006-2007 to incidence in 2009-2010. Using separate cohorts of about 140,000 people in each time period, they found roughly 10 percent fewer new dementia cases in the latter group. Moreover, they reported that the time spent with dementia was compressed, with people in the later cohort on average enjoying three to five more months of dementia-free life than those in the earlier group (see Doblhammer et al., 2015). This would be consistent with a later average age at onset. It is a counterintuitive finding, as growing awareness of early dementia symptoms among both physicians and the general public during those years might suggest a trend toward more diagnoses, not fewer.

Dissecting Out Contributing Factors
What explains the lower risk of dementia? As in previous studies, the Framingham researchers attribute some of it to better cardiovascular health. Numerous vascular health measures, such as blood pressure, smoking, and rates of atherosclerosis, improved over those decades. This had a measurable effect: For example, the risk of dementia after a stroke dropped by three-fourths from the beginning to end of the test period, suggesting better disease management. Nonetheless, adjusting the analysis for all vascular risk factors did not significantly change the overall finding, suggesting that other factors were at work as well. Satizabal noted that the study did not include data on diet or exercise, which affect AD risk. “We need to study the impact of all these factors and find the best combination to translate into public health policies,” she told Alzforum. Notably, the decline in incidence was only seen in people who had at least a high school diploma, suggesting that education and socioeconomic status could play a role as well.

Researchers would also like to know whether the decrease reflects lower rates of vascular dementia or Alzheimer’s. Of 371 total new dementia cases in the Framingham study, 264 were clinically diagnosed as AD and 84 as vascular dementia. The incidence declined for both but missed significance for AD, hinting that the drop might be driven more by vascular disease. The authors had too few autopsy samples to examine brain pathology. However, they noted that a recent autopsy study from Switzerland reported a 25 percent drop in amyloid pathology over three decades in 1,599 people tested postmortem, indicating that Alzheimer’s, too, might be on the decline (see Kövari et al., 2014). 

Factors other than lifestyle may play a role as well. In both the Framingham and German study, incidence declined more in women than men. “Attention to sex-related factors (chromosomal, endocrine, and sex biology) or to gender issues (social and cultural factors) may help us to interpret the trends. We and others have recommended looking at risk and protective factors for dementia separately in men and women,” Walter Rocca at the Mayo Clinic in Rochester, Minnesota, wrote to Alzforum (see Rocca et al., 2014Mielke et al., 2014). 

Meanwhile, what role ethnicity might play remains largely unexamined. Most previous studies, including Framingham, enrolled primarily people of European ancestry. At least one study hints that similar trends may hold in other ethnic groups in the United States. Researchers at Indiana University, Indianapolis, reported last year that dementia incidence has fallen by more than half in African-Americans over age 70 living in Indianapolis. The researchers enrolled a cohort of 1,440 African-Americans in 1992 and 1,835 in 2001, and followed both groups until 2009. The annual incidence rate in the 1992 cohort was 3.6 percent, but fell to 1.4 percent in the 2001 cohort. By contrast, incidence did not change among Africans living in Nigeria over the same time period (see Gao et al., 2015). 

Although incidence may be falling among African-Americans, Whitmer’s new data indicates they remain at higher risk than other ethnic groups. First author Elizabeth Rose Mayeda analyzed data from 274,283 people covered by Kaiser Permanente, a large insurer in Northern California, between 2000 and 2013. The authors found that dementia risk ran highest in African-Americans and people of Native American/Alaska Native descent, with 27 and 22 new cases per 1,000 person-years, respectively. Latinos, Pacific Islanders, and whites had an intermediate risk of around 19 new cases per 1,000 person-years, while Asian-Americans enjoyed the lowest risk, at 15 cases per 1,000. Altogether, the risk for African-Americans was about two-thirds higher than for Asian-Americans. The disparities remained after adjusting for age, sex, and access to health care, and are as yet unexplained.

Future Estimates Uncertain
While the drop in incidence is good news, researchers warn that other factors could wipe out the gains. Many studies, including Framingham, report rising rates of obesity and diabetes, both of which are associated with dementia. In their editorial, Jones and Greene cautioned, “Even if a dementia decline has begun, it might not last: The outcome depends on the balance of diverging trends.”

In addition, the situation is not the same worldwide. Some studies suggest dementia cases are on the rise in low and middle-income countries (see Dec 2012 news; Jun 2013 news). 

“We will need to know whether such patterns are confirmed or not in further, larger comparative studies which look across geographies (including low and middle-income countries),” Carol Brayne at the University of Cambridge, U.K., wrote to Alzforum (see full comment below). “It certainly points to the fact that we should not assume that current dementia is the same as dementia in the past, nor will it be the same in the future. This is a shift in thinking compared to a decade ago.”

Researchers said the findings should encourage more research into prevention. “The 22 to 44 percent reduction in risk or incidence over short, five- to 10-year periods is remarkably large and of the same magnitude as that expected in Alzheimer disease prevention studies. If a 25 percent risk reduction were to be observed for a drug tested in a prevention clinical trial, then we would be talking about a clinically meaningful effect as if it were a ‘cure,’” Lon Schneider at the University of Southern California wrote to Alzforum (see full comment below).

The next problem will be how to motivate people to make positive lifestyle changes, Larson noted. He believes that the fear of Alzheimer’s may influence people more than other disease risks do. “I have a lot of patients who are retired faculty, and they would never listen to me when I recommended exercise to prevent heart attacks. When they hear they can save their mind by exercising, however, a lot of them start,” Larson said.—Madolyn Bowman Rogers


  1. The accumulating evidence for decreases in the age-specific incidence of Alzheimer disease or dementia, over fairly short periods of five to 10 years, encourages speculating that this may be due to environmental and public health improvements over the past half-century, and that the Flynn effect and Fries’ predictions of compression of morbidity may be in play. People may be getting smarter, stronger, bigger, and healthier; surviving longer, and pushing back illness and age-related conditions closer to the end of life.

    It is tempting to consider that the cohorts who are benefiting were born in the early 20th century and suffered as children through a major flu epidemic, the effects of two world wars, the Great Depression, and periods of starvation and poverty. Then as survivors and adults—into the latter half of the century—they had relatively healthy and prosperous lives with improvements in diet, nutrition, hygiene, education, environmental pollution, automation, and overall quality of life. One might expect this trend to be further enhanced for the baby boomers born into relative and continuing prosperity.

    The 22 percent to 44 percent reduction in risk or incidence over short, five- to 10-year periods is remarkably large and of the same magnitude as that expected in Alzheimer’s disease-prevention studies. If a 25 percent risk reduction were to be observed for a drug tested in a prevention clinical trial then we would be talking about a clinically meaningful effect as if it were a “cure.” Such drugs for prevention or for mild Alzheimer disease—if proven effective—will be very costly. For example, solanezumab may command an average wholesale price of about $4,500 per month or $54,000 per year, for a planned effect that is relatively small. Given the epidemiological data, one might wonder if CMS or Medicare would be willing to pay this amount to middle-aged and older beneficiaries for them to engage in further health-promotion activities. Although much more than an ounce of prevention, this public health intervention still might outweigh several thousand pounds of cure.

  2. We did a comprehensive analysis of the studies that have been comparable that have been published in Europe, with indications of stabilization or reduction in both prevalence and incidence (Wu et al., 2016). The Rotterdam study within that analysis is, like Framingham, a dynamic cohort with refreshment of younger cohorts, and it suggests incidence reduction. All incidence studies to date have been single-site comparisons across time within single countries. The Framingham and those emerging from France have also looked at the profile of risk across time.

    There is growing evidence from the United States and Western Europe that dementia itself is changing as health and other life experience profiles change across generations.

    There are more studies of incidence to come. We will need to know whether or not such patterns are confirmed in further, larger comparative studies that look across geographies (including low- and middle-income countries). Japan has not reported such changes for example, although it is possible that diagnostic drift could account for the observed increase in prevalence.

    It certainly points to the fact that we should not assume that current dementia is the same as dementia in the past, nor will it be the same in the future. This is a shift in thinking compared to a decade ago, even though vascular factors were beginning to be re-recognized at that stage. This has implications for the nature of future dementia policy and research investment.


    . Dementia in western Europe: epidemiological evidence and implications for policy making. Lancet Neurol. 2016 Jan;15(1):116-24. Epub 2015 Aug 21 PubMed.

Make a Comment

To make a comment you must login or register.


News Citations

  1. World Alzheimer Report 2015: Revised Estimates Hint at Larger Epidemic
  2. Dementia Incidence Said to Drop as Public Health Improves
  3. Dementia Prevalence Falls in England
  4. Falling Dementia Rates in U.S., Europe Hint at Prevention Benefit
  5. Research Brief: Around the Globe, More People Die of Dementia
  6. Prevalence of Dementia, AD, in China Eclipses Predictions

Paper Citations

  1. . Dementia in western Europe: epidemiological evidence and implications for policy making. Lancet Neurol. 2016 Jan;15(1):116-24. Epub 2015 Aug 21 PubMed.
  2. . Compression or expansion of dementia in Germany? An observational study of short-term trends in incidence and death rates of dementia between 2006/07 and 2009/10 based on German health insurance data. Alzheimers Res Ther. 2015 Nov 5;7(1):66. PubMed.
  3. . Amyloid deposition is decreasing in aging brains: an autopsy study of 1,599 older people. Neurology. 2014 Jan 28;82(4):326-31. Epub 2013 Dec 20 PubMed.
  4. . Sex and gender differences in the causes of dementia: a narrative review. Maturitas. 2014 Oct;79(2):196-201. Epub 2014 May 27 PubMed.
  5. . Clinical epidemiology of Alzheimer's disease: assessing sex and gender differences. Clin Epidemiol. 2014;6:37-48. Epub 2014 Jan 8 PubMed.
  6. . Dementia incidence declined in African-Americans but not in Yoruba. Alzheimers Dement. 2015 Jul 26; PubMed.

External Citations

  1. exercise
  2. obesity
  3. diabetes

Further Reading


  1. . A comparison of health expectancies over two decades in England: results of the Cognitive Function and Ageing Study I and II. Lancet. 2016 Feb 20;387(10020):779-86. Epub 2015 Dec 9 PubMed.

Primary Papers

  1. . Incidence of Dementia over Three Decades in the Framingham Heart Study. N Engl J Med. 2016 Feb 11;374(6):523-32. PubMed.
  2. . Is Dementia in Decline? Historical Trends and Future Trajectories. N Engl J Med. 2016 Feb 11;374(6):507-9. PubMed.
  3. . Inequalities in dementia incidence between six racial and ethnic groups over 14 years. Alzheimers Dement. 2016 Mar;12(3):216-24. Epub 2016 Feb 11 PubMed.