AD is the most common cause of dementia overall, accounting for more than half of all dementia cases. The lifetime risk of developing AD is estimated to be between 12% and 17% (32,60). The single greatest risk factor for AD is advancing age. Epidemiologic age-specific estimates of incidence and prevalence vary somewhat by region and study due to differences in diagnostic criteria and population demographics, but all show an exponential increase in incidence and prevalence with advancing age at least through the ninth and probably through the tenth decade. The prevalence of severe dementia over the age of 60 years is estimated at 5%, and over the age of 85, it is estimated to be between 20% and 50%.

Diabetes, the metabolic syndrome [defined as three or more of the following: abdominal obesity, hypertriglyceridemia, low high-density lipoprotein cholesterol, hypertension, and hyperglycemia (31)] (74,94), and hyperinsulinemia (23,66) are associated with an increased risk of AD. In the Rotterdam study, diabetes mellitus doubled the risk of incident AD over a follow-up period that averaged slightly over 2 years (74). Prevalence of AD in a Finnish population aged 69 to 78 years was 7.2% in those with metabolic syndrome compared with 2.8% in those without metabolic syndrome, but the difference seemed confined to women (94). In addition to diabetes, other vascular risk factors such as hypercholesterolemia (88), hypertension, and hyperhomocysteinemia are correlated with a greater risk for AD (45,65,98). Less clear, however, is the relationship between stroke and AD. Vascular risk factors correlate with increased susceptibility to both AD and vascular dementia (45,47,49,90,98), and some studies have shown a correlation between cerebral atherosclerosis and AD (48,81). Cerebral infarction, whether related to atherosclerosis or amyloid angiopathy, further impairs cognition in patients with AD (46,99) but does not appear to predispose to AD itself (83).

Other purported risk factors have weaker associations, and some are more controversial because not all well-controlled epidemiologic studies have found them to be significant. Some of these factors include limited education (15), depression (75), gender (27,97), estrogen replacement therapy (30,102), and head trauma (34,54). Similarities between dementia pugilistica and AD include abeta deposition (40), neurofibrillary pathology (70), and similar genetic predisposition factors [apolipoprotein E (ApoE) ε4] (58). Possible protective effects have been claimed and disputed regarding the use of vitamin E (61,78) and the use of anti-inflammatory drugs (2,91). Neither can be strongly encouraged for AD prevention at this time.

Chronic aluminum exposure was once thought to play a role in AD mainly due to encephalopathy occurring in dialysis patients who were exposed to toxic levels of aluminum. The possible mechanisms by which aluminum could lead to AD histopathology include promoting hyperphosphorylation of tau and subsequent formation of neurofibrillary tangles, altering processing of amyloid precursor protein (APP) and then leading to the formation of neuritic plaques, and initiating a local inflammatory response (4). However, at present, aluminum exposure is no longer thought to be a major risk factor for AD. More recently, Sparks et al. (86,87) have proposed that copper exposure from drinking water may play a role in predisposition or progression of AD.

There are three levels of genetic susceptibility to AD. The first level is that conferred by autosomal dominant mutations with high penetrance that cause early-onset familial AD (EOFAD). EOFAD strikes patients when they are young, typically ranging from midthirties to mid-fifties. To date, more than 100 mutations of three genes have been identified that may cause EOFAD, and all are inherited in an autosomal dominant pattern. The largest number of EOFAD cases are caused by mutations of the presenilin-1 (PS1) gene, and thus, PS1 is thought to account for the majority of autosomal dominant kindreds (84). The PS1 mutation is currently the only autosomal dominant mutation for which there is a commercially available genetic test. A smaller number of mutations have been localized to the APP gene on chromosome 21 (100) and the presenilin-2 (PS2) gene on chromosome 1 (64). All result in elevated levels of abetaamyloid (Aβ), underscoring the pathogenetic importance of Aβ in the evolution of AD (43). Genetic testing of patients with possible EOFAD may confirm the genetic cause in a family, thus allowing preclinical testing of unaffected family members who may be at risk.