See Q&A with Walter Rocca below.
17 May 2013. Are fewer people getting Alzheimer’s disease now than they did 20 years ago? In the face of dire predictions of a coming avalanche of dementia, the idea seems surprising. Nonetheless, studies suggest that new dementia cases in the Western world have fallen by a quarter to a third over the last 10 to 20 years. Most recently, in the May 14 Neurology, researchers led by Chengxuan Qiu and Laura Fratiglioni at Karolinska Institute-Stockholm University, Sweden, presented indirect evidence that the incidence of dementia was declining over two decades in one district of Stockholm. Other studies from the U.S. and the Netherlands report similar findings. What explains such a trend? While no one knows for sure, researchers suggest it could be due to environmental and social changes such as better management of cardiovascular risk factors and a more highly educated population. If the findings hold up, they will support the idea that public health initiatives reduce dementia risk, and could stimulate more research into prevention, scientists said.
Many researchers believe that the data reflect a real phenomenon. “We can be cautiously optimistic that the rates of cognitive impairment and dementia may be going down,” Walter Rocca at the Mayo Clinic in Rochester, Minnesota, wrote to Alzforum (see full Q&A below). However, he added that new studies are needed to confirm the decline and explore the underlying mechanisms.
While the findings are encouraging, researchers warned that improvements could be wiped out by negative health trends such as rising obesity and diabetes, two important risk factors for AD. In a May 12 presentation at the European Congress on Obesity in Liverpool, U.K., researchers led by Laura Webber and Tim Marsh at the U.K. Health Forum, London, predicted that growing obesity rates in England, if left unchecked, could contribute to a 40 percent increase in dementia cases by 2050. Moreover, even if incidence wiggles a bit, the overall prevalence of dementia will continue to rise due to aging populations, scientists said. Researchers led by Liesi Hebert and Denis Evans at Rush University, Chicago, Illinois, estimated from U.S. Census data that AD cases will nearly triple in the next 40 years. Their report appeared in the May 7 Neurology.
Costs associated with dementia will mushroom as well. Researchers led by Michael Hurd at the RAND Corporation, a not-for-profit policy think tank based in Santa Monica, California, and Kenneth Langa at the University of Michigan, Ann Arbor, put the current costs of dementia in the U.S. between $159 billion to $215 billion a year and predicted that it will more than double by 2040. Described in the April 4 New England Journal of Medicine, the finding made The New York Times. Even with a possible decline in the rate of new cases, dementia will remain a colossal health problem, researchers stressed.
Real Trend or Artifact?
Researchers first need to verify that incidence is truly nudging down. Evidence to date is suggestive but mixed. While data from the Chicago Health and Aging Project showed no change in new cases of dementia from 1997 to 2008 (see Hebert et al., 2010), researchers at the Rochester Mayo Clinic saw a statistically significant decline of about 3 percent per year in new dementia cases from 1985 through 1994, for a total drop of 30 percent (see Rocca et al., 2011). Both of these studies sampled people in one small geographic area, perhaps limiting the applicability of the results to the general population.
By contrast, the Health and Retirement Study, directed by David Weir at the University of Michigan, Ann Arbor, surveys a representative sample of people over 70 years of age across the U.S. every two years. It encompasses cognitive tests that measure memory, processing speed, and language. From 1993 to 2002, the prevalence of cognitive impairment among participants dropped from 12.2 percent to 8.7 percent, roughly a 30 percent decrease (see Rocca et al., 2011; also Langa et al., 2008, and Sheffield and Peek, 2011). The National Long-Term Care Survey corroborated this, finding that severe cognitive impairment in older U.S. adults fell by about half from 1982 to 1999 (see Manton et al., 2005).
The same trend occurs in Europe. In the Rotterdam Study in the Netherlands, new dementia cases in people 60 to 90 years old fell by a quarter in 2000 compared to 1990, although the finding just missed statistical significance (see Schrijvers et al., 2012). Strengthening the data, structural MRI scans found significantly less brain atrophy and cerebrovascular disease in the more recent cohort, suggesting better brain health.
For their Stockholm study, Qiu and Fratiglioni used data from two cross-sectional surveys of people 75 years or older living in the Kungsholmen district of the city—one conducted from 1987-1989, the second from 2001-2004. The researchers diagnosed dementia in the same way in each cohort, using criteria from the Diagnostic and Statistical Manual of Mental Disorders, revision III-R (DSM-III-R). They followed participants for six years, recording death rates. While people in the more recent cohort lived longer, the prevalence of dementia was unchanged compared to the earlier cohort. Because prevalence stayed constant while survival time increased, the authors inferred that incidence must have fallen. Qiu told Alzforum they plan to directly compare incidence once the data become available from the newer cohort. Study participants were mostly white and had high socioeconomic status. It is unclear whether the findings would generalize to other populations.
Overall, researchers agreed that these papers suggest but do not prove that the number of new dementia cases is dropping in the U.S. and Europe. “I think the evidence is pretty strong, but it’s a complex process to pin this down 100 percent,” Langa told Alzforum. Researchers said that more incidence studies over longer periods of time would help settle the matter. Ideally, these studies would be done in diverse populations and in different countries, and would include nursing home populations, Langa said. Methodology challenges include making sure diagnostic methods are consistent over time and accounting for people who refuse to participate or who drop out so that they do not bias the results.
Public Health Improvements Pay Off
What might underlie the trend toward less dementia? Over the last few decades, better management of blood pressure and cholesterol, less smoking, and more exercise (see Risk Factor Overview) have steadily curtailed heart disease in the U.S. and Europe. Cardiovascular risk factors are strongly linked to dementia (see, e.g., Qiu et al., 2010; Whalley, 2012). Hypertension in midlife heightens the risk for AD, though the evidence remains inconsistent (see AlzRisk). Altogether, the data imply that better heart health could translate into fewer AD cases.
“The most plausible explanation is not that the pathology of AD itself is going down, but that the burden of cerebrovascular disease is going down and therefore is contributing to the overall reduction in dementia rate,” said David Knopman at the Rochester Mayo Clinic. He is the deputy editor at Neurology and handled the paper by Qiu et al. Future incidence studies should include biomarkers to help researchers shed light on underlying causes of dementia, Knopman suggested. “Had more people in recent studies had MRI or amyloid imaging, we might be able to disentangle how much of the trend is due to cerebrovascular versus degenerative disease,” Knopman said. Qiu agreed, noting that a subset of the Kungsholmen cohort received structural MRI scans. Qiu and colleagues plan to examine these data to look for links among vascular risk factors, brain pathology, and dementia.
More recent cohorts are also more educated than earlier groups. Education and lifetime cognitive activity seem to help people resist Alzheimer’s disease (see ARF related news story; ARF
news story; ARF
news story; and ARF
news story). In the Health and Retirement Study, higher levels of education in people born later explained about 40 percent of the drop in incidence.
Other factors may play a role as well. As Lon Schneider at the University of Southern California, Los Angeles, pointed out on his blog, since the 1970s Americans have enjoyed many improved health options, including better food choices, gyms for working out, smoking restrictions, and cleaner air. “The idea that public health interventions may not only work but have marked effects, that we might grow our way out of a dementia 'tsunami' and substantially reduce risk, is intriguing and wholly consistent with the range of hypotheses about Alzheimer's disease and the amyloid theory,” he wrote. Diet, exercise, smoking, and air quality have all been tied to dementia risk (see, e.g., ARF related news story; ARF news story; ARF news story; and ARF news story).
Previous research supports the idea that better health can delay dementia. A 2011 study by Deborah Barnes and Kristine Yaffe at the University of California, San Francisco, predicted that improving seven measures of health—including hypertension, smoking, exercise, cognitive activity, obesity, and diabetes—could prevent about 10 percent of AD cases worldwide (see ARF related news story). Three large prevention trials in Europe are testing the benefits of lifestyle interventions (see ARF related news story). In the U.S., the Alzheimer’s Disease Cooperative Study is running a study to determine whether exercise can slow cognitive decline in people with mild cognitive impairment (see ARF related news story).
But Wait—Negative Trends Could Gut the Gains
Notwithstanding these improvements in health, obesity and diabetes are on the march in the U.S. and Europe. Having either of these risk factors in midlife appears to predispose people to Alzheimer’s, suggesting that cases could rise again as today’s middle-aged population ages. “The numbers in the Stockholm study reflect health behaviors of probably 20 or 30 years ago,” Knopman pointed out. Jennifer Weuve at Rush University agreed, “It may be that the incidence goes down for a while, and then it goes back up again.” In a Neurology editorial on the paper by Qiu et al., Lawrence Whalley at the Institute of Applied Health Sciences, Foresterhill, Aberdeen, U.K., and Kathleen Smyth at Case Western Reserve University, Cleveland, Ohio, wrote, “Failure to address the growing obesity epidemic in the United States and Europe has serious implications for our ability to reduce dementia incidence.”
At the European Congress on Obesity, Webber and Marsh's data revealed how burgeoning waistlines could increase the prevalence of dementia in England. Currently, about 25 percent of the population falls into the obese classification, but researchers predict that by 2050, this will expand to 46 percent of men and 31 percent of women. In an already aging population, this could lead to a 40 percent increase in dementia, from 5 percent to 7 percent of people over 65, the authors said. If, instead, obesity holds constant at current levels, new dementia cases in 2050 would drop by about 10 percent. This would save about £940 million in healthcare costs, the authors project.
Similar trends are afoot across the globe. Because of rising longevity, AD cases are expected to swell worldwide over the next decades. In the U.S., the aging of the large Baby Boomer population will increase prevalence even if incidence dips. “Any slight change, as indicated by this decreasing incidence, would be dwarfed by the demographic changes,” Knopman said. Other researchers agreed. “Dementia will still be very common, and it has a terrible impact on families. It will still loom large as a population burden,” said Robert Wallace at the University of Iowa, Iowa City.
Different groups arrive at slightly different numbers for future prevalence, but all agree it will be immense. Hebert and colleagues’ projection of a tripling of cases, from 4.7 million today to 13.8 million by 2050, was based simply on the aging of the population. They took mortality and education into account, but assumed the risk of developing AD remained constant. Incidence data for the study came from the Chicago Health and Aging Project (CHAPS), which estimates higher than other studies.
What will be the economic impact? Hurd and Langa used data from the Health and Retirement Study to arrive at their cost estimates. They focused on a subset of participants who underwent detailed dementia testing using DSM criteria as part of the Aging, Demographics, and Memory Study (ADAMS). From these data, the authors estimated that about 4.2 million people in the U.S. had dementia in 2010. This is lower than the CHAPS numbers, possibly because the Chicago study uses less stringent criteria for dementia and may count people who have mild cognitive impairment, Langa told Alzforum.
To determine total costs, the authors added up nursing home care, formal home care, Medicare, and out-of-pocket spending. They estimated the value of informal home care based either on the lost wages of the caregiver or on the cost of equivalent paid home care. The former method led to a total dementia cost estimate of $157 billion per year, the latter to $215 billion per year. Both are greater than the financial burden of heart disease and cancer. The authors calculated that these costs will more than double by 2040, based on the aging of the population and assuming no change in other factors.
Researchers agreed that even if incidence truly falls, dementia has not been solved. There will still be millions of people with the disease, and finding an effective treatment will remain urgent. “If we can confirm the decline in incidence, this should increase research, rather than decrease it. We need to demonstrate which factors have determined the decline, so that additional interventions can be introduced in the population,” Rocca wrote to Alzforum.—Madolyn Bowman Rogers.
Qiu C, von Strauss E, Bäckman L, Winblad B, Fratiglioni L. Twenty-year changes in dementia occurrence suggest decreasing incidence in central Stockholm, Sweden. Neurology. 2013 May 14;80(20):1888-94. Abstract
Whalley LJ, Smyth KA. Human culture and the future dementia epidemic: Crisis or crossroads? Neurology. 2013 May 14;80(20):1824-5. Abstract
Hebert LE, Weuve J, Scherr PA, Evans DA. Alzheimer disease in the United States (2010-2050) estimated using the 2010 census. Neurology. 2013 May 7;80(19):1778-83. Abstract
Hurd MD, Martorell P, Delavande A, Mullen KJ, Langa KM. Monetary costs of dementia in the United States. N Engl J Med. 2013 Apr 4;368(14):1326-34. Abstract
Q&A With Walter Rocca. Questions by Madolyn Bowman Rogers.
Q: Several recent studies seem to show a decline in the incidence of cognitive impairment in Europe and the U.S. How good are the data? Can we say with confidence yet that incidence is truly going down?
A: Combining the initial 2011 report of a decline in incidence or prevalence from the U.S. (Rocca et al., 2011) with the 2012 report from the Netherlands (Schrijvers et al., 2012), and more recently from Sweden (Qiu et al., 2013), we can be cautiously optimistic. The rates of cognitive impairment and dementia may be going down. However, new studies are needed to confirm the decline and to explore the underlying mechanisms. The studies so far are limited in the way they measured prevalence or incidence and by the short time window (10 or 20 years maximum).
Q: What kind of studies would be needed to really nail down this finding? About how long would such studies have to be, involving how many participants, over how large a geographic area?
A: In general, the time frame should be longer than 20 years to be able to reveal long-term secular trends. It may be important to study the concurrent trends in dementia, cardiovascular diseases, and other neurodegenerative diseases such as Parkinson's disease in the same population to separate the contribution to the trend of underlying vascular or degenerative mechanisms. The findings from Rochester can be extended for almost 20 more years if funding were available (1995-2013). Unfortunately, we do not have funding to extend our study.
Q: If incidence is declining by, for example, 3 percent per year as your Rochester study suggested, how much would that lower estimates for AD prevalence in 2050? (For example, the Alzheimer's Association estimates about 16 million U.S. cases in 2050. How much could that drop?)
A: It is premature to make alternative projections for the year 2050. In our paper, we suggested a possible 3 percent decline per year that would translate into a 30 percent decline over 10 years. However, our estimate was based only on the 10-year period from 1985 to 1994. Did the trend continue after 1994? Did it accelerate or slow down?
Q: If incidence is truly declining, what will that mean for AD research? Do you think it will be harder to get funding, for example? Will it increase interest in public health interventions such as exercise and education, since these seem to be behind the decline?
A: If we can confirm the decline in incidence, this should increase research, rather than decrease research. We need to demonstrate which factors have determined the decline so that additional interventions can be introduced in the population. You mentioned education and exercise; however, other factors such as control of hypertension, diabetes, obesity, reduction of smoking, and improvement in diet may be very important. The use of aspirin, statins, beta blockers, and other drugs may have also contributed.