Epidemiology

In 2013, an estimated 44.35 million individuals worldwide were living with dementia. This number is expected to double every 20 years, with between 115 and 135 million individuals living with dementia by 2050 (Prince et al., 2013; Alzheimer Disease International, 2013). The prevalence of dementia increases with advancing age, affecting an estimated 25% of individuals by age 85 (Ferri et al., 2005).

The sex difference in dementia prevalence also increases with advancing age; roughly 5% of men and an almost equal percentage of women between ages 71–80 are diagnosed with dementia; after age 80, almost 28% of women and over 17% of men develop dementia (Plassman, et al., 2007).

The type of dementia appears to account for the sex difference in prevalence rates. Among individuals older than 65 years of age, the most common cause of dementia is Alzheimer disease (AD), accounting for about 70% of cases (Alzheimer’s Association, 2011). Defined neuropathologically by the presence of intraneuronal neurofibrillary tangles, extracellular amyloid plaques, and neuronal cell death, AD disproportionately affects women. Two-thirds of patients with AD are women (www.alz.org/downloads/facts_figures_2012.pdf) and after age 80, the proportion of women with the disease (21%) is almost twice that of men (12%; Plassman et al., 2007). Although women generally live longer than men, women’s increased longevity is not sufficient to explain the 1.5 times greater likelihood of developing AD (Gao et al., 1998). Although the reasons why women are at greater risk than men for developing AD are not well understood, one possible explanation is that the disease appears to cause a more rapid global cognitive decline in women than it does in men (Ito et al., 2011). Sex differences in disease progression are supported by the fact that men are more likely than women to be diagnosed with mild cognitive impairment (MCI; Katz et al., 2012; Petersen et al., 2010), a condition that is often considered a prodrome of AD.

Vascular dementia, resulting from small vessel disease, a series of strokes, or a single stroke, is widely considered the second most common cause of late-life dementia, and estimated to account for around 20% of cases (e.g. Jellinger, 2013; Plassman et al., 2007). In contrast to the sex differences in the prevalence of AD, vascular dementia is equally prevalent in women and men, affecting 2–3% of each sex over age 70 years (Plassman, et al., 2007). Of studies that do find a sex difference in the prevalence of vascular dementia, a greater proportion of men than women are reported to be affected (e.g., Kalaria et al., 2008).

The hallmark pathologic feature of Lewy body dementia is the presence of Lewy bodies in the cortical and subcortical regions of the brain. Lewy body dementia is the third-most common type of dementia and has been estimated to account for 10–20% of cases of late-life dementia (Weisman et al., 2007; Oda et al., 2009). Because of the relatively recent development of diagnostic criteria for Lewy body dementia, estimates of its prevalence are not as consistent as those for other dementias. In general, Lewy body disease is more common in men than women, with a four-fold incidence rate excess in men after age 70 in one North American study (Savica et al., 2013), and a smaller sex difference (with an incidence of 121/100,000 in men compared to 107/100,000 in women) noted in a French cohort (Perez et al., 2010). In a large autopsy series of patients seen in dementia research centers in the United States, patients with Lewy bodies were significantly more likely to be male, with an odds ratio of roughly 2.9 (Nelson et al., 2010).

Parkinson’s disease (PD) is characterized by loss of dopaminergic neurons in the substantia nigra and the presence of Lewy bodies in the cytoplasm of neurons in subcortical regions with accompanying gliosis. The prevalence of PD increases with advancing age, ranging from 40.5 per 100,000 individuals between ages 40–49 years to 1,903 per 100,000 individuals aged 80 years and older (Pringsheim et al., 2014). There is a consistently greater prevalence of PD in men (1,057 per 100,000) than in women (794 per 100,000) across all geographic regions (Pringsheim et al., 2014). Similarly, many studies find a higher incidence of PD in men, with a male:female ratio of 1.5–2.0:1 (Twelves et al., 2003). The mechanisms underlying sex differences in PD are unknown. Women’s higher estrogen levels, which lead to increased striatal dopamine, have been proposed as potentially related to sex differences in this disease (Haaxma et al., 2007).

Frontotemporal dementia (FTD), defined by neurodegeneration in the frontal and temporal lobes, typically has an earlier age of onset than other forms of dementia, often becoming manifest in the 50s or 60s. Consequently, roughly 60% of individuals with FTD are between ages 56 and 64 years, and the disease is as common as AD in those younger than 65 years. The estimated point prevalence of FTD is 15–22/100,000, with an annual incidence of 2.7–4.1/100,000 (Onyike & Diehl-Schmid, 2013). Because FTD is a heterogeneous disorder with several known genetic factors and other as-yet-unidentified causes, reports of the sex distribution of this disease vary widely. In one study, for example, the ratio of cases in men compared to women was 14:3 (Ratnavalli et al., 2002), whereas other studies found a females excess, with a ratio of 1:3 (Bernardi et al., 2012; Gilberti et al., 2012).

Taken together, epidemiological studies reveal that age is a risk factor for all leading causes of dementia in late life and that AD accounts for the majority of cases of dementia. Most types of dementia are either equally prevalent in women and men, or are slightly more prevalent in men, with the notable exception of AD. Not only does AD affect many more women than men, the risk of developing this disease is greater in women than in men.

Clinical presentation

Because dementia caused by neurodegenerative disease is characterized by insidious onset, it often goes undetected until well into the course of the disease. This is particularly true in patients with AD, who are often able to maintain social skills, typically have little insight into their cognitive decline and can appear unimpaired in casual conversation. Unless cognition is specifically assessed, their disease is likely to be overlooked in its early stages. Indeed, studies consistently find that many cases of dementia are not recognized in primary care settings (e.g., Valcour et al., 2000). One meta-analysis of physician accuracy in diagnosing cognitive impairment indicated that approximately one in four cases of dementia, and approximately half of cases of mild cognitive impairment or mild dementia, remain undetected in primary care (Mitchell et al., 2011). Of particular concern are older individuals who live alone, in whom both physicians and knowledgeable informants are less likely to detect dementia compared to persons with dementia living with others (Lehmann et al., 2010). Thus, being aware of the prevalence of dementia among older people and of some of the ways in which it may present may aid earlier detection and diagnosis as well as life planning.

Impairment of memory for recent events is among the earliest symptoms of dementia. Although individuals with dementia may be able to converse about recent events in general, their memory impairment may cause them to repeat questions or stories, forget appointments, or lose their belongings. Because normal aging is associated with cognitive changes, including general slowing of information processing, increased difficulty with multitasking, and mild impairment of episodic memory (Glisky, 2007), it can be difficult to determine whether a memory complaint indicates disease or simply an awareness of the gradual inefficiency in encoding new information that normally accompanies advancing age. For this reason, family members are often better informants about a person’s functioning than is the individual him- or herself, as he or she may lack insight or be overly sensitive to cognitive failures due to excessive worry.

Deficient access to semantic information, manifest as word-finding difficulties, is also a common early symptom of dementia, and of AD in particular. As with memory difficulties, occasional difficulty thinking of a particular word becomes more common with normal aging. Pathological word-finding difficulties, however, occur with greater frequency in individuals with dementia and can lead to vague speech (words devoid of content) and interfere with the ability to communicate meaningful information effectively.

Personality changes in the absence of frank cognitive impairment can be an early symptom of dementia. In patients with FTD, for example, family members sometimes report marked changes affecting food preferences, judgment or interest in others. Individuals with FTD frequently exhibit disinhibited and impulsive behaviors and may curse or demonstrate uncharacteristic sexual impropriety in social situations. The patient, in contrast, is not likely to be aware of these changes.

Neuropsychiatric disturbances commonly occur in the course of a dementing illness (Steinberg et al., 2006). They include depression and anxiety, psychotic symptoms such as hallucinations, paranoia, or delusions, irritability and sleep disturbances. In general, the severity of the behavioral disturbance is related to the severity of dementia (Majic et al., 2012). Although symptoms vary widely, some aspects of the clinical presentation of dementia differ in women compared to men. In patients with dementia living in residential care facilities, staff ratings indicated that women and men differed on 7 of 39 behaviors and symptoms (Lövheim et al., 2009). Women were more often depressed, whereas men were more often aggressive and engaged in inappropriate (“regressive”) behaviours. No sex differences in passiveness (lack of spontaneous speech, initiative and/or cooperation with staff) or hallucinations were found.

Specific dementing illnesses can manifest different symptoms in women compared to men. Among patients with AD seen in outpatient clinics, caregivers’ ratings indicated that women were more reclusive, were more emotionally labile, and tended to hoard and refuse help, as well as display inappropriate laughter or crying more frequently than men. Men, in contrast, exhibited behaviors more indicative of psychomotor changes (apathy, pacing) and vegetative changes such as excessive eating and sleeping (Ott et al., 1996).

Very few studies have explored sex differences in the clinical presentation of vascular dementia. In a Chinese cohort of 467 patients with vascular dementia, Xing and colleagues (2012) found that among patients with mild vascular dementia, women were rated by caregivers to be more likely than men to have delusions, hallucinations and depression. In moderate to severe vascular dementia, men were more likely than women to manifest apathy. These findings suggest that, compared to men, women have more neuropsychiatric symptoms in mild vascular dementia, whereas in more severe cases, only apathy distinguishes women from men.

Women with PD have been shown to be diagnosed later than men with PD (51.3 vs. 53.4 years; Haaxma et al., 2007). At disease onset, women are reported to have more dyskinesias, whilst men have more rigidity, possibly related to medication use (Accolla et al., 2007). Abnormal sleep behaviors may be less common in women than in men with PD (Ozekmekçi et al., 2005), although results from one study found no sex difference in the prevalence of sleep disturbance in patients with PD, but reported that men had more violent behaviors during sleep than women did (Bjørnarå et al., 2013).

Studies of Lewy body dementia typically control for sex rather than compare women and men. For this reason, there are not sufficient findings to comment on whether sex differences exist in the clinical presentation of this disease. Similarly, research on FTD has not yielded reports of sex differences in its clinical presentation.

Risk factors for dementia

Because treatments for dementia are limited, efforts aimed at identifying risk factors have received increasing research attention. This line of work has implicated numerous risk factors, many of which pertain to general health status (e.g., obesity, high blood pressure at midlife, diabetes mellitus all increase risk), lifestyle (e.g., smoking increases risk; moderate alcohol consumption decreases risk), or demographic factors (e.g., higher socioeconomic status reduces risk) and have been found to be salient for both women and men. Several factors, however, have been shown to pose differential risks in women and men.

Education. The protective effect of higher education on dementia risk has been hypothesized to be related to “cognitive reserve” (Satz et al., 1993). That is, the brains of individuals with higher cognitive abilities prior to onset of dementia can withstand more disease before overall functioning is affected. The association between education and dementia-related death appears to be particularly relevant for women. In a meta-analysis of 11 prospective studies that included over 85,000 individuals in the United Kingdom, Russ and colleagues (2013) found that leaving full-time education at an earlier age increased the risk of dementia death in women but not in men, even after controlling for occupational social class and other common risk behaviors and comorbidities. Similarly, an Italian study found education to reduce the risk of dementia in women but not in men (Noale et al., 2013).

Diet. Lifestyle choices, including diet, have received much research attention as a potentially modifiable factor in dementia risk. Although most studies in this area have not revealed differential effects in women and men, findings from one study suggest that fruit and vegetable consumption is protective in women only. In a cohort of Swedish twins discordant for dementia in (in particular AD), greater consumption of fruit and vegetables during mid-life reduced the risk of dementia and AD 30 years later in women, but not in men (Hughes et al., 2010). This risk was reduced even further in women with angina, supporting the idea that improved vascular health is the mechanism by which a healthier diet is protective against the development of dementia in women.

Sex hormones. The relationship between sex hormones and dementia is complex and different studies have produced conflicting findings. A number of studies provided support for the notion that increased exposure to estrogens (endogenous estradiol) reduced the risk of dementia in women. Studies comparing women with higher versus lower lifetime exposures to endogenous or hormone supplements make up a large portion of the findings.

Parity. Several studies have shown that having children increases the likelihood of developing AD in women but not men (Colucci et al., 2006), and is positively correlated with AD neuropathology in women but not men (Beeri et al., 2009). Similarly, a series of studies revealed that women with a greater number of pregnancies have a higher risk of developing AD and/or a younger age of onset (Colucci et al., 2006; Sobow et al., 2004). The association between parity and age of AD onset appears confined to women without the APOE4 allele as it was not observed in women with the APOE4 allele in one study (Corbo et al., 2007), suggesting fertility is an independent risk factor for AD in women.

Hysterectomy. Women who have undergone hysterectomy have an increased risk of early-onset dementia (Phung et al., 2010), and this risk is increased with younger age of hysterectomy (Rocca et al., 2012; Phung et al., 2010). Furthermore, the combined results from two cohort studies (in the United States and Denmark) indicated that extent of gynecological surgery was associated with a stepwise increase in the risk of dementia. Specifically, women who had hysterectomy were at increased risk of cognitive impairment or dementia. The risk was further increased in women who had undergone hysterectomy with unilateral oophorectomy, and even further increased in women who had undergone hysterectomy with bilateral oophorectomy (Rocca et al., 2012).

Hormone replacement therapy. Whether hormone replacement therapy should be offered to women who are experiencing symptoms of menopause is a topic of controversy. Many early observational studies found a protective association between “ever-use” hormone therapy in women and dementia in later life (Hogervorst et al., 2000; LeBlanc et al, 2001). Results from a large, randomized, placebo-controlled trial of hormone replacement therapy in postmenopausal women (age 65–79 years) with no history of hormone replacement therapy prior to age 65 showed that the risk of dementia (not just AD) was doubled in women receiving combined hormone (estrogen + progesterone) therapy, but hormone supplementation did not reduce the risk for developing mild cognitive impairment (Shumaker et al., 2003). Moreover, hormone replacement therapy also increased risk for breast cancer, stroke and cardiovascular disease (Grady et al., 2002; Writing Group, 2002). Since initial results of that study were published, a meta-analysis of all double-blind, placebo-controlled trials studying the effects of estrogen replacement therapy or hormone therapy on cognitive functioning confirmed that neither estrogen alone nor estrogen combined with a progestagen prevented cognitive decline in postmenopausal woman (Lethaby et al., 2008).

One study reported smaller hippocampal volumes in women who received hormonal therapy (Shumaker et al., 2003), leading the researchers to postulate that the cognitive impairment seen in hormonally treated women may be mediated by brain atrophy, rather than ischemic brain changes (Espeland et al., 2009). However, studies focused on exposure to HRT of younger, perimenopausal women converge in support of the “critical window” hypothesis such that hormone replacement therapy confined to perimenopause, rather than later in life, actually decreases the risk of dementia later in life (Whitmer et al., 2011; Shao et al., 2012).

Stress. Environmental stress can impair cognition, particularly memory (see Lupien et al., 2005, for a review). While the mechanisms underlying these effects are not well understood, the extent to which stressors impair cognition appears to be related to the degree of an individual’s cortisol response to stress, rather than the experience of stress itself (Takahashi et al., 2004; Wolf et al., 2001).

Cortisol response to stress. One meta-analysis of 45 studies found that the effect of age on the cortisol response to a pharmacologic or psychological challenge was 3-fold higher in women than in men (Otte et al., 2005). Importantly, effect sizes did not differ in studies that controlled for sex hormone variations in women (e.g., standardizing menstrual cycles, excluding women on oral contraceptives or hormone replacement therapy) compared to those that did not, suggesting that sex hormones do not alter the effect of aging on the stress response in women. In line with this finding, studies have shown that acute psychosocial stressors are associated with memory impairment in elderly women but not in elderly men (Almela et al., 2011; Wolf et al., 1998).

Neuroticism. The propensity to perceive situations as negative or stressful, termed neuroticism, has been shown to increase the risk for dementia in several studies. In a 35-year longitudinal study of women, Johansson and colleagues (2010) found that reports of “frequent/chronic” stressors during mid-life increased the risk of AD at follow-up. Similarly, a recent meta-analysis of five prospective studies (Terraccino et al., 2014) reported that individuals in the top quartile of “distress proneness” (high scores on neuroticism) had a three-fold increased risk of AD. Sex differences in neuroticism are consistently reported across the world, with women scoring higher than men (Lynn & Martin, 1997), and the magnitude of the difference remains constant across the adult age span (Chapman et al., 2007; Costa et al., 2001). These findings suggest that, even if neuroticism increases the risk equally in women and men, the fact that women tend to score higher on neuroticism may place them at differential risk.

Stressful life events. Given that stress can impair cognition and increase risk for AD, it is notable that one of life’s most stressful events – death of a spouse – is much more common in older women than in older men. From ages 75–84, 55% of women and 18% of men have lost a spouse to death. After age 85, 72% of women and 35% of men have been widowed (U.S. Census Bureau, 2006). Widowhood during midlife increases the risk of cognitive impairment and AD in those who never remarry, and the risk is even greater among those with APOE4 allele (Håkansson, 2009). Of note, a recent 25-year follow-up study found that death of a spouse did not increase the risk of dementia, but women who lost a partner had a temporary decline in executive functioning during the first 2 years after their husbands’ deaths compared with women who were still married (Vidarsdottir et al., 2014). To the extent that elderly individuals are vulnerable to stress-induced cognitive decline or AD, women might be expected to be at particular risk for dementia compared to men simply by virtue of being much more likely to experience death of a spouse (see Munro, 2014, for a review).

Mild Cognitive Impairment (MCI). MCI is a term used to describe a heterogeneous condition in individuals who have cognitive deficits at an intermediate stage between normal aging and dementia. Individuals with MCI have subjective memory complaints and score lower on objective memory testing than cognitively normal older adults, but the cognitive changes of MCI do not cause impairment in overall functioning. MCI may also be considered a risk factor for dementia; roughly 6–10% of individuals with MCI progress to dementia each year, but some will show improvement in cognitive functioning or will remain stable without progression to dementia (Petersen et al., 2010). Some studies have reported differences in incidence rates of MCI between women and men and risk factors for MCI may differ in women and men as well. In one community study, higher rates of MCI in younger men were associated with higher rates of cardiovascular disease and higher homocysteine levels compared with women, whereas lower rates of MCI in women correlated with better physical fitness and lower homocysteine levels (Sachdev, 2012).

Depression. Late-life depression can be mistaken for dementia. Many older patients with depression develop reversible impairment in executive functioning with cognitive slowing, and often these changes are more prominent than changes in mood (Butters et al., 2004). In addition, depression is frequently an early symptom of dementia, in both AD and vascular dementia (Alexopoulos, et al., 1997; Ownby et al., 2006). A large retrospective cohort study of over 13,000 patients found that individuals with late-life depressive symptoms showed a two-fold increase in risk of AD, whereas those with midlife and late-life depression had a greater than three-fold increase in risk of vascular dementia (Barnes et al., 2012). A longitudinal study of 436 healthy older women reported that those with significant depressive symptoms at baseline were more likely to become impaired on cognitive testing during a 7-year follow-up period (Rosenberg et al., 2010). Depression is twice as common in women than it is in men from puberty through mid-life, and continues to be more prevalent among women through early old age. These findings suggest that not only is recognition and treatment of late-life depression important in improving quality of life and functioning, it may also help prevent cognitive decline, especially for women.

Treatment approaches

Management of dementia requires a multifaceted approach. To date, there are no disease-modifying medications that prevent or cure dementia. The mainstay of pharmacologic treatment for individuals who have dementia involves cholinesterase inhibitor medications, which include donepezil, galantamine and rivastigmine, as well as memantine, which is an N-methyl-D-aspartate receptor antagonist. Individuals with dementia are particularly vulnerable to the further cognition-impairing effects of anticholinergic medications and such medications should be minimized in all patients with dementia (Bishara and Harwood, 2014). There are no known sex differences in response to these pharmacologic agents.

Reduction of risk factors is still the key approach to prevention of dementia and management of cognitive impairment. Among potential factors associated with reduced dementia risk, physical exercise has the most consistent empirical support (Ahlskog et al., 2011), and may be more beneficial for women than for men. A study of women who had vascular disease and coronary risk factors found that those who engaged in regular physical activity, including walking, had lower rates of cognitive decline over a 5-year follow-up period (Vercambre et al., 2011). Results from a randomized, controlled trial of aerobic exercise in patients with MCI found that 6 months of aerobic exercise resulted in improvements on multiple tests of executive functioning in women but not in men (Baker et al., 2010). These findings suggest that the potential cognitive benefits of aerobic exercise should be considered, particularly in women who are at risk for dementia.