Introduction
We are experiencing an international rising tide of late-life dementias. This rising tide is the result of the general lengthening of lifespan worldwide, which in turn stems from improvements in public health, education, and medical care. The increase in late-life dementias accompanying a longer lifespan poses serious challenges for both personal and public health in all the nations of the world (Larson and Langa, Reference Langa2008).
Rates of dementia in the elderly have been found to vary widely among nations, but generally they are higher than had been thought previously (Llibre Rodriguez et al., Reference Llibre Rodriguez2008). As a rule of thumb, after age 65, rates of dementia tend to double every five years in developed countries, and every seven years in developing ones (Lobo et al., Reference Lobo2000). For people over age 85, the rate of dementia worldwide averages at about 30%, with a dramatic increase after age 80 in developed nations (Graves et al., Reference Graves1996).
These dementias, particularly Alzheimer's disease and “mixed dementias” with more than one etiology (Langa et al., Reference Langa, Foster and Larson2004), have profound effects on aging individuals and their caregivers. It is no wonder that so much research is being devoted to exploring varied approaches to preventing or delaying dementias.
In this review, I outline the prospects for prevention of dementia, including Alzheimer's disease, with an emphasis on the role of lifestyle and vascular risk factors. The review is admittedly selective and emphasizes our own work, but I have included up-to-date reviews which generally confirm my ideas. I present descriptive evidence from survey data that rates of dementia in the U.S.A. may be decreasing. I discuss the possibility of a “compression of cognitive morbidity” as an achievable goal along with an evolutionary phenomenon related to society becoming more “knowledge based” in the twentieth century (Galbraith, 1967; Drucker, Reference Drucker1983). And I emphasize the potential power of delaying the onset of dementia – providing we make such efforts a priority in the increasingly aging societies of today and tomorrow (Montine and Larson, Reference Montine and Larson2009).
Historical perspective
To understand how we got here, it is helpful to look backward. As recently as the early 1970s, experts considered cognitive decline, including extreme senility, to be part of the continuum of “normal aging;” and Alzheimer's disease (AD) was still lumped with Pick's disease as “pre-senile dementia.” Since then, “senile dementia” has been recognized as largely due to AD, indistinguishable from the classic form of the disease, which Alois Alzheimer first detected in a middle-aged patient.
In the 1970s and 1980s, research concentrated on distinguishing AD from multi-infarct or vascular dementias. Since then, a dramatic increase in the prevalence of AD and other dementias has been demonstrated in persons aged in their 80s and beyond. The dramatic increase is primarily explained by better detection and awareness of dementia in old age.
Research in the 1990s developed more precise definitions of AD and its prevalence and pathologic features. In 1993, a seminal paper established that vascular dementia is more common in late life than had previously been realized (Skoog et al., Reference Skoog, Nilsson, Palmertz, Andreasson and Svanborg1993). Mixed dementias became more commonly reported in clinical studies; and this was confirmed by reports of community-based neuropathologic studies (Langa et al., Reference Langa, Foster and Larson2004; Lim et al., Reference Lim1999).
The past decade has seen research establishing that before age 65, dementias are rare, while after age 65, their incidence rises exponentially, with an even steeper up-sloping curve after 80 years of age (Graves et al., Reference Graves1996; Kukull et al., Reference Kukull2002). In general, rates of dementia in higher-income countries somewhat exceed those in lower-income ones; however, rates are increasing around the world, particularly in urban Latin America and areas where overall longevity has improved (Llibre Rodriguez et al., Reference Llibre Rodriguez2008).
Survival differences between people with dementia and without it become narrower the later their age at onset (Larson et al., Reference Larson2004). However, dementia is not inevitable, as demonstrated by an individual living to 115 years who showed no discernable neurodegenerative changes (den Dunnen et al., Reference den Dunnen2008). Worldwide, reported rates of dementia vary widely.
As populations worldwide are increasingly aging, late-life dementias have far-reaching consequences for health. The rising tide of late-life dementia likely reflects gains in life expectancy. Dementia is likely made more evident as societies become increasingly “knowledge-based” (Drucker, Reference Drucker2001). Higher levels of education and the wealth generated by knowledge-based society are strongly associated with gains in life expectancy.
What influences the development of dementia?
Early-life socioeconomic status
Low socioeconomic status in early life is well known to affect growth and development, including that of the brain; and it has also been shown to affect the risks of other chronic diseases. The maturation of the brain continues into adolescence, and the areas that take the longest to mature – the intracortical association area and the hippocampal and reticular formations – are the same areas that show the earliest signs of AD (Conel, Reference Conel1939–1967; Yakovlev and Lecours, Reference Yakovlev, Lecours and Minkowski1967; Braak and Braak, Reference Braak and Braak1991; Arriagada et al., Reference Arriagada, Growdon, Hedley-Whyte and Hyman1992). Therefore, it is no surprise that early-life characteristics associated with a lower socioeconomic environment tend to promote less healthy brain development and raise the risk of late-life AD (Moceri et al., Reference Moceri, Kukull, Emanuel, van Belle and Larson2000; Reference Moceri2001).
Moceri et al. used birth records of persons enrolled in the Group Health/University of Washington Alzheimer Disease Patient Registry (ADPR) and the Genetic Differences Case-Control Study (Moceri, et al., 2000) to relate early socioeconomic status to later dementia. The results were consistent with the hypothesis that a healthier socioeconomic environment in childhood and adolescence leads to more “brain reserve” (the brain's ability to cope with increasing age- and disease-related changes while still functioning) and less risk of late-life dementia, including AD, later on. Factors associated with more favorable socioeconomic environments that appeared to protect against developing AD included living in a suburb, having a father whose work was skilled and not manual, and having fewer siblings – with odds ratios of 0.45, 1.8, and 1.08 per added sibling, respectively. For people with the Apolipoprotein E (ApoE) ɛ4 allele, reared in the countryside, with several siblings and a father doing unskilled or manual work, the increase in risk was multiplicative. These findings are consistent with the brain reserve hypothesis. The influence of early “lifestyle” factors on the risk of AD is noteworthy, particularly for those persons at higher genetic risk of developing the disease, an example of a gene-environment interaction.
Education, occupation, and leisure activities
In general, evidence suggests that increasing education may have protective effects against developing dementia (Albert et al., 2005; Schaie, Reference Schaie2005). This observation has a pathologic correlate: The dendritic field has been reported to be larger with increasing education levels (Gould et al., Reference Gould, Beylin, Tanapat, Reeves and Shors1999). This is also consistent with a link between education and brain reserve being preserved or enlarged.
The complexity of people's occupations also positively influences cognitive vitality, and this relationship becomes increasingly marked with age. Higher levels of cognitive function are also linked to an “engaged lifestyle,” with leisure activities that are cognitively stimulating. It is interesting that this association is stronger for blue-collar workers than for elite academics (Fratiglioni et al., Reference Fratiglioni, Paillard-Barg and Winblad2004).
Can a remedy prevent dementia?
The quest for a “fountain of youth,” as old as Ponce de Leon, continues unabated. Long before dementia became an issue, “youth” was the focus of those who were naturally yearning for a simple solution to aging. Now people seek a simple remedy to prevent AD and other dementias. Ever since the late 1970s when AD gained prominence, both scientists and laypersons have sought a golden pill to prevent or reduce the risk. In spite of claims and hopes, all of the simple preventive remedies have proven to be failures. Among the many examples of recent promising leads that resulted in dashed hopes are: antioxidant vitamins (Gray et al., Reference Gray2008); non-steroidal anti-inflammatory drugs (Breitner et al., Reference Breitner2009); and ginkgo biloba (Snitz et al., Reference Snitz2009). Cholinesterase inhibitors are also unlikely to provide a solution.
Many people have remarked that if only physical activity were a pill, it would be a popular medication indeed. Unlike the putative “magic bullets” listed above, physical activity, leisure activity, and social networks actually do seem promising as potential ways to help delay the onset of dementia. Their effects may be mediated by the plasticity of the brain, a phenomenon only recently realized.
Brain plasticity
Age-related declines in cognition and underlying brain function are not necessarily inevitable. It was long thought that damaged brain tissue could not regenerate or be restored. However, recent research suggests that lifestyle, education, occupation, expertise, and fitness can modify age-related declines in brain function, thanks to the plasticity of the brain (Kramer et al., Reference Kramer, Bherer, Colcombe, Dong and Greenough2004). For instance, training in reasoning appeared to slow functional decline in AD, with some cognitive abilities sustaining improvement for five years (Willis et al., Reference Willis2006).
Physical activity, leisure activity, and social networks
In 2004, Fratiglioni et al. analyzed 13 published longitudinal studies relating dementia, including AD, to social networks and leisure and physical activities (Fratiglioni et al., Reference Fratiglioni, Paillard-Barg and Winblad2004). All but one were conducted in Europe or North America; and all but one were embedded in large longitudinal population-based studies of aging starting with non-demented persons. Assessment was done at least three years after enrollment in 12 of the studies, and assessment ranged from one to seven years in the other study. Of the 13 studies, five focused on dementia, three on AD only, and five on both AD and dementia.
In seven of eight studies, physical activity – including exercise and daily physical activity – showed an inverse correlation to cognitive impairment. Beneficial effects of physical exercise have also been reported in diseases including hypertension, diabetes, obesity, osteoporosis, and depression. Questions of causation have been raised about the inverse correlation between physical activity and dementia, and evidence that exercise might prevent dementia published at the time of this 2004 review was more equivocal. However, at least one study has suggested that poor physical function may precede the onset of dementia and AD, with higher levels of physical function associated with a delayed onset (Wang et al., Reference Wang, Larson, Bowen and van Belle2006).
The studies of non-physical leisure time activity mostly involved cognitively stimulating activity. In six of seven of these studies, leisure activity was inversely correlated with cognitive decline or lower cognitive performance. Other studies have linked leisure activity – attending cultural events, reading books or magazines, solving crossword puzzles, and playing music or singing in a choir – with longer survival. This effect appears to be independent of the effect of fitness activity, which suggests involvement of mechanisms other than increased cardiopulmonary fitness.
The studies of social networks used disparate measures, but five of seven of them found the presence, quality, and size of a person's social network inversely related to the incidence of dementia. However, one concern about interpreting these findings is that declines in social network and physical and leisure activities might represent a manifestation of early dementia rather than a pre-morbid risk factor: People may become less active physically and socially as a result of their experiencing cognitive decline.
Work from Fratiglioni's own group has suggested that physical, mental, and social components in leisure activities contribute equally to decrease the risk of dementia (Fratiglioni et al., Reference Fratiglioni, Wang, Ericsson, Maytan and Winblad2000; Karp et al., Reference Karp, Paillard-Borg, Wang, Silverstein, Winblad and Fratiglioni2006). Interesting and likely related recent research has suggested that maintaining a strong purpose for one's life may be protective against the development of AD (Boyle et al., Reference Boyle, Buchman, Barnes and Bennett2010).
Growing evidence for vascular risk as a factor
Studies that rely on clinical criteria designed to detect AD may not adequately discern vascular brain injury as the sole or concomitant cause of dementia. Evidence is growing that much dementia stems from vascular brain injury – from both small- and large-vessel disease, but especially the former. Vascular brain injury, AD, and Lewy body disease occur through different disease processes to produce the phenotype of dementia. And the three processes can combine variably in individual patients to produce the dementia syndrome. For instance, neuropathologic studies have shown that not only “plaques” and “tangles” (Braak stage) but microvascular infarcts account for risk of dementia (Sonnen et al., Reference Sonnen2007).
Growing evidence also suggests that limiting vascular risk can help delay late-life dementia. Since the 2004 review by Fratiglioni et al, the preponderance of evidence from observational and community-based studies and early trials strongly supports the argument that among the vascular risk factors for dementia that can be modified, regular physical exercise has great promise. Descriptive studies, including those from our Adult Changes in Thought (ACT) project at Group Health, have associated habitual exercise with significant reductions in the risks of both AD and dementia from all causes (Abbott et al., Reference Abbott, White, Ross, Masaki, Curb and Petrovitch2004; Larson and Wang, Reference Larson and Wang2004; van Gelder et al., Reference van Gelder, Tijhuis, Kalmijn, Giampaoli, Nissinen and Kromhout2004, Weuve et al., Reference Weuve, Kang, Manson, Breteler, Ware and Grodstein2004; Podewils et al., Reference Podewils2005; Larson et al., Reference Larson2006; Larson and Langa, Reference Larson and Langa2008). Our community-based descriptive cohort study found that compared with sedentary participants, exercise at least three times per week was associated with an approximate 40% reduction in dementia risk. Critical, systematic reviews of exercise and physical activity programs to improve cognitive function in old persons or for persons with dementia have shown some promise but have not been definitive (Angevaren et al., Reference Angevaren, Aufdemkampe, Verhaar, Aleman and Vanhees2008, Forbes et al., 2008) because convincing evidence from randomized trials was not available.
In a recent randomized trial of exercise, Lautenschlager et al. (Reference Lautenschlager2008) showed that habitual physical exercise protected persons with subjective complaints about their memory (“mild cognitive impairment”) from cognitive decline (Lautenschlager et al., Reference Lautenschlager2008). At the 18-month follow-up, this trial showed a statistically significant difference of 0.69 point between people who walked regularly and a sedentary control group in changes on the primary outcome measure, the Alzheimer Disease Assessment Scale–Cognitive Subscale scale. This “effect size” was greater than that for cholinesterase inhibitors (Malouf and Birks, Reference Malouf and Birks2004). This result is very important. Like the seminal randomized trial of Willis et al. (Reference Willis2006) showing proof of concept that reasoning training can ameliorate and even reverse age-related cognitive decline, we now have evidence from a clinical trial supporting a similar role for habitual physical exercise.
Other vascular risk factors are also of great interest. While randomized trials of antihypertensive drug treatments have not proved that treatment prevents dementia (McGuinness et al., Reference McGuinness, Todd, Passmore and Bullock2009b), our research suggests an age-varying association of high systolic blood pressure with incident dementia in young elderly persons (<75 years) but not in older subjects (Li et al., Reference Li2007b). We also recently found that presence of cerebral microinfarcts was independently associated with high systolic blood pressure in younger participants (aged 65–80 years), especially those not taking antihypertensives (Wang et al., Reference Wang2009), but not in older participants. This suggests that adequate antihypertensive treatment may reduce dementia risk by minimizing microvascular injury to the cerebrum.
The picture for cholesterol as a vascular risk factor and statins as modifiers may be similar. Earlier reviews have not concluded that treatment with statins had a beneficial effect on dementia risk (McGuiness et al., 2009a). Descriptive studies have been plagued with methodological inconsistencies and problems (Li et al., Reference Li2004). In general, if there is an association with serum cholesterol and dementia risk it is likely to be restricted to mid and early late life cholesterol levels (possibly 15–20 years before onset), and not to late-life levels (Li et al., Reference Li2005). In our neuropathology study of 110 subjects with brain autopsies aged 65–79 years who were cognitively normal at enrollment in our Adult Changes in Thought study, findings demonstrated an association between antecedent statin use and reduced neurofibrillary tangle burden at autopsy. The risk for typical Alzhiemer's pathology was reduced in statin users (OR 0.20; 95% CI 0.05–0.86) (Li et al., Reference Li2007a).
Other vascular risk factors worth noting are smoking and diet. Although early reports emphasized that some nicotine exposure might be associated with reduced risk, more definitive research and reviews demonstrate smoking as a risk factor for cognitive decline and dementia (Anstey et al., Reference Anstey, von Sanden, Salim and O'Kearney2007). Recent research on diet include observations that a Mediterranean-style diet and physical activity have been associated with reduced risk of AD (Scarmeas et al., Reference Scarmeas2009) and a Mediterranean-style diet was associated with slower decline in Mini-mental State Examination (MMSE) scores (Féart et al., Reference Féart2009).
Clearly, vascular risk factors are of great interest for those interested in reducing late-life risk of dementia and age-related decline. Exercise may turn out to be as close as we can come to the long-sought fountain of youth, and exercise, particularly walking, has no significant adverse effects. Control of other vascular risk factors like smoking, hypertension, adverse lipid pattern and diets, also show great promise especially in mid-life and early old-life before brain degeneration may become harder to reverse or prevent.
Trends and prospects
Recent trends in medical, lifestyle, demographic, and social factors have been positive for the cognitive health of older Americans – especially higher levels of education and better control of cardiovascular risk factors. Between 1982 and 2004, the National Long-Term Care Surveys showed significant decline in the prevalence of chronic disability, especially “mixed” dementia, among the Medicare-enrolled population in the U.S.A. (Manton et al., Reference Manton, Gu and Ukraintseva2005). The decrease was associated with higher levels of education and fewer cerebrovascular disease events. Data from the longitudinal U.S. Health and Retirement Study suggest decreasing rates of cognitive impairment in people aged over 70 years from 12.2% in 1993 to 8.7% in 2002. There was also a decrease in the interval between when substantial cognitive decline started and when death occurred. About 40% of the fall in the rates of cognitive decline were linked to higher education and wealth – and probably to improved control of cardiovascular risk as well (Langa et al., Reference Langa2008). Although this study has many potential limitations, its findings suggest a “compression of cognitive morbidity,” which is a desirable outcome: Fewer persons reach a threshold of significant cognitive impairment – and those who do, have a more rapid decline to death (Fries, Reference Fries1980). This compression was accidental – not planned – but it is of instructive value as we approach the rising tide of late-life dementias. Delaying the onset of dementia in even part of the at-risk population can decrease disease prevalence significantly, since patients are likely to die from another disease before dementia occurs (Brookmeyer et al., Reference Brookmeyer, Gray and Kawas1998). If a one-year delay in onset could be achieved, it is estimated that it would prevent more than 9 million cases of AD worldwide (Brookmeyer et al., Reference Brookmeyer, Johnson, Ziegler-Graham and Arrighi2007).
Particularly in developed nations, society is now firmly knowledge based, and cognitive health – especially maintaining high function and independence in late life – is vital to successful development. Rapid acquisition, use, and exchange of information, including Internet use, may have profound effects on brain development. For instance, recent selective evolutionary forces appear to favor the survival of three genes causing increased brain size and higher microglial density (Balter Reference Balter2005, Evans et al., Reference Evans2005; Mekel-Bobrov et al., Reference Mekel-Bobrov2005; Hayakawa et al., Reference Hayakawa, Angata, Lewis, Mikkelsen, Varki and Varki2005). Building our brains throughout life (through education and purposeful activity) and efforts to reduce neurodegenerative processes throughout life will likely have profound effects on individual and public health in this century of the brain.
Conclusion
Absolute prevention of late-life dementia is unlikely, especially as more people live longer. However, “compression of cognitive morbidity” is achievable, with dementia starting later and lasting for a shorter period. Better health and well-being throughout the lifespan, along with control of vascular risk, has likely benefited late-life risk. These trends and a growing evidence base support further efforts to understand and ideally control vascular risk, which appears to offer great prospects for preventing late-life dementias.
Conflict of interest
None.
Acknowledgments
I thank and acknowledge the excellent editorial work of Rebecca Hughes and Rita Weikal in developing this paper. I also thank the many staff and faculty who have worked on the Group Health/University of Washington Alzheimer's Disease Patient Registry (ADPR)/Adult Changes in Thought (ACT) study and on our community-based Kame project, an epidemiologic study of dementia in Japanese Americans in King County, Washington. Without their efforts since this work began in 1986, the material and ideas in this paper would not have existed. The projects were supported by grants from the National Institutes of Health, AG 06781 and AG 09769.
A version of this paper was delivered in September 2009 as a Plenary Presentation to the International Psychogeriatric Association in Montreal, Quebec, Canada.
