Race- and Sex-Based Disparities in Alzheimer’s Disease Clinical Trial Enrollment in the United States and Canada: An Indigenous Perspective

US: Demographics

The US has three distinct indigenous populations: American Indian (AI) also called the Native American, the Alaska Native (AN), and the Native Hawaiian and other Pacific Islander (PI) group. In publications, census data, and other statistical references, the AI and AN are usually lumped together and called the “American Indian/Alaska Native” (AIAN). The Pacific Islander group includes Hawaiian, Samoan, and Tongan descendants, along with Asian populations. Little research has been done in these populations, and most literature reviewed in this paper focus on the AIAN group. Based on the US Department of HHS Office of Minority Health report from 2017 [68], the AIAN population was 5.6 million alone or in combination with one or more races, accounting for 1.7% of the US population. This number differs from other references including the Profile America Facts for Features article from October 2018, which states that 6.8 million people are of AIAN descent, including those of more than one race [69]. Whatever the current population number, the projected population estimate is the same: >10 million by 2060 [69, 70]. Overall, there are 326 reservations, with 573 federally recognized tribes and 100 state recognized tribes that speak more than 200 languages [68, 70, 71]. Alaska contains the highest proportion of native populations at 19.5%, followed by Oklahoma (13.6%), New Mexico (11.8%), South Dakota (10.3%), and Montana (8.3%) [70]. Seventy-eight percent of the AIAN population live off-reserve, with 22% on reserves. Sixty percent of the AIAN population reside in metropolitan areas, the lowest percentage among all racial minority groups [68]. Approximately, 569,000 AI and AN alone or in combination were≥ 65 years of age in 2015 and the oldest cohort, aged 85 and older, is expected to increase sevenfold by 2050 [72].

US: AD statistics

Only recently, the life expectancy of the AIAN has approached that of non-Latino whites [73]. The life expectancy for the AIAN in the 1940s was 51 years, but in the 1990s increased to 70 years. The number of AIAN elders will double by 2030 from 2016 levels, which increases their risk of AD by virtue of age [73]. Although the current prevalence of AD and related dementias (ADRD) in the AIAN is less than AA, Latino Americans, and non-Latino whites based on Medicare FFS claims [67], the AIAN have a 35% greater risk than non-Latino whites of acquiring AD over a 25-year period after age 65 [34]. This risk value is based on a publication by Mayeda et al. that examined dementia incidence between 2000–2013 along with 25-year cumulative risk in 274,283 healthcare members aged 64+ in California among six racial groups, stratified by age and sex [74]. However, the actual AD risk may be skewed based on the “certificate of degree of Indian blood” (CDIB) present in a particular tribe [75]. So-called pure blood natives rarely have the APOE e4 allele associated with increased risk of AD, and thus may have lower AD numbers than those without pure blood [75, 76]. Despite the presence of pure blood natives in the population, between 2014 and 2060, it is estimated that the numbers of AIAN≥ 65 years of age diagnosed with AD will increase fivefold [67, 72]. In the younger age group (≥45 years), one in every six AIAN reported “subjective cognitive decline” (SCD), which is a set of self-reported problems with memory [51, 72]. Whatever the current numbers for AD, those numbers may be underestimated in the US [71]. Reasons for this could include: lack of access to appropriate healthcare, competing risk factors for early mortality (accidents, suicides, infectious diseases), and the complexity of identifying AD in patients with vascular lesions from stroke [71].

US: AD risk factors

According to the WHO, “The social determinants of health are the conditions in which people are born, grow, live, work and age. These circumstances are shaped by the distribution of money, power and resources at global, national and local levels. The social determinants of health are mostly responsible for health inequities–the unfair and avoidable differences in health status seen within and between countries”.

There are specific risk factors in the AIAN population that may increase the prevalence of AD. The AIAN cannot control non-modifiable risk factors such as age, race, sex, and presence or absence of the APOE ɛ4 allele, but many of the risk factors they face are based on underlying socioeconomic conditions and lifestyle choices. These risk factors may include:

Lack of recognition of cognitive decline. Unless an AIAN is willing to recognize cognitive decline as a sign of disease and seek treatment, AD may go undiagnosed. The terms “Dementia” and “Alzheimer’s Disease” are not easily translated into a common language understood or appreciated by Natives Americans and many tribes have defined and explained cognitive decline and memory loss in their own language based on history and culture [71]. Some of these explanations include: 1) “Naturalization” in which cognitive decline is part of the Creator’s plan for ultimate learning; 2) An imbalance in a person’s spiritual, emotional, and social environment (the speaking of negative consequences of illness has broken a cultural taboo and caused the illness); 3) Communication with the spirit world (a “crossing over” into the next world); and 4) Dementia is a conflict between traditional and non-traditional ways of life. These explanations may delay a tribe member from seeking professional medical help and delay the diagnosis and treatment of AD.

Poverty, income, and lack of education. Natives have higher poverty levels and lower educational achievement than any other ethnic group in the US [71]. The median income of a single-race AI and AN household in 2015 was $38,530 compared to $57,775 for the nation as a whole [70]. 26.6% of single-race AI and AN were in poverty in 2015 [70]. The rural AI population is considerably poorer than the urban population [71]. Rural native elders have 70% or less of the per capita income of urban elders [71]. In terms of education, 19.6% of AIAN aged 25 and over had at least a bachelor’s degree, compared to 35.8% of non-Latino whites and only 6.8% of AIAN held an advanced graduate or professional degree, compared to 13.8% of non-Latino whites [68].

Cardiovascular disease (CVD). Native Americans are more likely to be diagnosed with CVD than whites, and the three contributing risk factors for heart disease are more common in AIAN than whites: mid-life obesity, higher blood pressure, and smoking [72, 77]. Mid-life obesity affects one in three AIAN over the age of 50; AIAN are 30% more likely to have high blood pressure than whites; and AIAN have the highest smoking rates of any racial minority in the US with one in three in their 50s and one in four in their 60s smoking cigarettes [72]. In a 2002 study by Weiner et al. on the Choctaw Indians of Oklahoma, modifiable cardiovascular risk factors that include hypertension, elevated cholesterol, diabetes, myocardial infarction, and obesity are higher in the Choctaw tribe than the white population [78].

Other risk factors. Tuberculosis rates were four times higher in AIAN than whites in 2017 (incidence rates of 3.9 versus 0.5, respectively) [68]. In fact, American Indians are 500 times more likely to die of tuberculosis than non-natives [79]. Other risk factors include cultural barriers, geographic isolation, and inadequate sewage disposal [68]. Of those AIAN over 5 years of age, 26.9% spoke a language other than English at home [80]. This complicates the diagnosis of AD, which often relies on cognitive testing formats presented in English. Diabetes also increases the risk of AD in the AIAN communities. Native Americans are 177 times more likely to die of diabetes than non-natives [79]. The Gerontological Society of America published a Special Issue in 2015 that looked at Depression and Dementia among the AIAN population; depression, linked to higher diabetes rates, is expected to affect 81,702 AIAN by 2050 [81].

US: AD clinical trial challenges

Indigenous populations in the US are poorly represented in RCTs. We searched the ClinicalTrials.gov website for trials specific to AD and specific to American Indigenous populations. Using the terms “Alzheimer’s Disease”, “Dementia”, and “Memory Loss” for the Condition or Disease title in combination with “Indigenous”, “American Indian”, “Indian”, “Native American”, “Alaska Native” and “Pacific Islander”, we found only one clinical trial that is not yet recruiting [43]. The RCT titled “Alzheimer’s Disease and Precision Medicine Research with American Indian and Alaska Native People” (NCT03448601) was first posted on February 28, 2018. This is a University of Colorado sponsored RCT (with collaboration from Washington State University and the University of Arizona) in which 3,000 AIAN participants, aged 18–100 years of both sexes will be recruited. This mixed-method, multi-state study will develop and evaluate culturally tailored recruitment materials and test their effectiveness for: a) engaging AIAN participants in a culturally tailored recruitment and engagement module focused on AD and Precision Medicine (AD-PM); b) enhancing enrollment in AD-PM clinical trials; and c) improving knowledge and attitudes about AD and PM. Because recruitment into RCTs is the greatest challenge for most studies, this trial will provide some much-needed information for the AD research community. Poor recruitment into RCTs by the Indigenous in America are partly caused by their inherent mistrust of the federal government and the institutions they support. There is a distinct history between natives and the government, starting with “destruction of entire native populations”, colonization, treaties, forced relocation, and attempts at assimilation through the removal of their children and the impact of missionaries [71]. Boarding school practices in the first half of this century are among recent events that have caused considerable trauma to the mental health of AIs [71, 82]. Alaska Natives that experienced boarding school described post-traumatic stress disorder, social phobias, contemplation of suicide, damage to sense of identity and well-being, alcohol abuse, emotional scars, difficulty integrating back into family and community, loss of culture, loss of language, feeling like strangers in their home, and failure to learn parenting skills [82]. Other past wrongs included the forced sterilization of indigenous women in the 1970s without consent and the misuse of blood samples from nearly every Havasupai tribe member of Arizona [83]. Other historical wrongs included the giving of “smallpox blankets” to the Haudenosaunee peoples of New York by Sir Jeffery Amherst and the use of “Indian” mascots by sports teams [65].

When AI individuals were asked why they don’t participate in a RCT, some of their responses included: 1) They don’t want to be “guinea pigs”; 2) The study findings are rarely shared with the community; 3) The study findings rarely improve local services; 4) The promised study benefits rarely reach the AI community; and 5) Access to resources which allow community members to partake (e.g., transportation) are lacking [65, 83]. Other barriers to research include distance from study sites [84, 85], patients not being informed of availability of trials, and the need for approvals from the Tribal/IHS (Indian Health Service) Institutional Review Board and Tribal Research Committee for a trial offered in a tribal clinic [84]. Also, universities that often run a RCT are viewed as elitist and not committed to the welfare of minority communities [86].

Canada: demographics

The Indigenous or Aboriginal populations in Canada consist of three main groups: First Nations (Indian), Metis, and Inuit (or Inuk). The First Nations are equivalent to the American Indian or Native American group in the US. The Metis are descendants of early 17th-century relationships between North American tribes and Europeans, often of French descent. Inuit is an Inuktitut word meaning “the people” and refer to those that live mainly in four polar and circumpolar regions of Canada called (from west to east); Inuvialuit, Nunavut, Nunavik, and Nunatsiavut [39, 87]. Neither the Metis or Inuit are protected under the Indian Act, created in 1867 to “colonize” the First Nations peoples [39, 88]. Reserves were created and held by the Crown (Canada is a British Commonwealth country) and administered based on treaties. Only those with registered Indian Status (i.e., Status Indian) may own land on a reserve [89].

The 2016 Canadian Census listed 1,673,780 Indigenous peoples in Canada, which then represented 4.9% of the population. Approximately 58.2% were First Nations (76.2% of these registered or treaty status); 35.1% were Metis; and 3.9% were Inuit [90]. Between 2006 and 2016, the Indigenous population increased by 42.6%, almost four times the rate of the non-Indigenous population, with Metis leading the increase at 51.2% compared to First Nations (39.3%) and Inuit (29.1%) [87]. There are more than 600 unique First Nation/Indian bands that represent more than 50 nations and speak more than 70 languages [87]. Like the rest of the world, the Indigenous population in Canada is aging. In 2006, 4.8% of the Indigenous were >65 years of age. By 2016, this number increased to 7.3% [87]. Within the First Nations, 44.2% lived on reserves in 2016 [89]. These reserves are geographically dispersed from coast to coast to coast in the second largest country (by area) in the world. This geographical isolation alone makes recruitment into an RCT a unique challenge in Canada.

Canada: AD statistics

Since Indigenous elders are among the most at risk and frail of the population, their AD numbers are expected to increase. One estimate indicates that ADRD cases are expected to increase by 3.3-fold in the Inuit and 4.2-fold (with increases to 4.8-fold on reserves) in First Nations by 2031 from 2006 levels [91, 92]. In a report on the Six Nations of Ontario in 2014 on ADRD, caregivers from the medical community suspected that 25% of their patients had AD; however, effective training of clinical staff in the diagnosis and treatment of AD and a definitive diagnosis from a doctor is lacking in many communities [93].

Canada: AD risk factors

Risk factors for AD in the Indigenous in Canada can be categorized as: 1) Modifiable risks; high cholesterol, increased blood pressure, diabetes, smoking, obesity, diet, and inactivity; 2) Partially modifiable risks; alcohol usage, reduced education, depression, and traumatic brain injuries; and 3) Non-modifiable risks; age, genetics, family history, and gender [91]. Indigenous populations cannot control their race, family history, sex or age, but many of the modifiable and partially modifiable risk factors occur in higher numbers and more severely impact the Indigenous populations than the non-Indigenous populations [39, 91, 92, 94, 95]. Modifiable risk factors may account for 75% of all AD in Indigenous people [96]. Some of these risk factors may include:

Memory loss. On reserves in particular, memory loss associated with AD is often considered a “normal part of aging”, a part of the “circle of life”, or a “return to childhood” [92, 93]. So long as there is no title attached to memory loss (such as “Dementia” or “Alzheimer’s Disease”), there is no stigma associated with it. This “childishness” and the presence of hallucinations can be seen as a return to the Spirit World and the Creator [93]. Often, elders will recommend “Red Whip Tea” (red willow) or a traditional “Eagle Dance” to help with forgetfulness. This view of memory loss is a risk factor, as Indigenous are less likely to seek help and get a proper cognitive screening test and diagnosis by a doctor. By the time they are properly diagnosed, their condition may be quite advanced and their long-term quality of life jeopardized.

Frailty. Frailty can be defined as having: exhaustion, shrinking, weakness, slowness, and low activity [49]. It is an accumulation of health problems and conditions referred to as “deficits”. In an Ontario First Nations Ageing Study, 25% of First Nations aged 75 and up had 5 or more chronic conditions [95]. Many Indigenous elders have had a difficult life due to their socio-economic conditions, including racial discrimination and the burdens of colonialization and thus experience more frailty and have a higher number of co-morbidities than non-Indigenous elders [95]. First Nations groups in Ontario experience frailty at a younger age than the general Canadian population [95]. Frail Indigenous are associated with poorer nutrition, lack of paid employment, use of illegal drugs, and are 1.7 times more likely to have parents that attended Residential Schools, and 2.4 times more likely to go to Traditional Healers for treatment rather than a licensed physician [95].

Diabetes. Diabetes increases the risk of AD by twofold [91] and First Nations populations are 3–5 times more likely to have diabetes than the general Canadian population [96]. Indigenous have higher rates of obesity, physical inactivity, and smoking than non-Indigenous, all of which are risk factors to acquire diabetes [96]. In 2008/2009, the age-standardized rates of diabetes were 17.2% in First Nations on reserves, 10.3% in First Nations off-reserve, 7.3% in the Metis, and 5% in the Inuit and general populations [91].

Lack of trust in the medical field. Most doctors, nurse practitioners, nurses, and caregivers on reserves are non-Indigenous and don’t speak the same language or have the cultural sensitivity training required to understand and appreciate the population they serve [96]. There is also a lack of geriatric resources, a high turnover among staff uncomfortable working on remote northern reserves, and a cultural insensitivity to elders in the population. Many nurses make the diagnosis in the absence of a doctor, and are not well trained to diagnose AD specifically when patients present with different chronic conditions or have normal memory loss caused by aging [91]. In addition, there is a high degree of mistrust with the federal government who run the health care systems on reserves. The Indian Act of 1867 that provided the “protection, assimilation and Christianization” of First Nations [88, 89], along with Residential Schools, the 60’s Scoop (in which Indigenous children were “scooped” from their homes and fostered or adopted into primarily middle class white families), and gender discrimination [88, 94] have all led to a mistrust in the Healthcare system especially by on-reserve Indigenous across Canada.

Other risk factors. Other AD risk factors affecting the First Nations, Metis, and Inuit in Canada include high cholesterol, CVD, hypertension (high blood pressure), lack of immunizations, poor dental health, depression, reduced education, post-traumatic stress, alcohol, poor diet, and exposure to toxins [91, 92, 95, 96].

Canada: AD clinical trial challenges

Previously we listed general challenges to recruitment and retention into an RCT, along with age-specific and AD-specific challenges. Add to this geographical isolation and separation of Indigenous peoples across a very large country and generalized mistrust of the healthcare system, as well as poor knowledge and understanding of AD, and recruitment into an RCT takes on a whole new level of challenge.

We searched the ClinicalTrials.gov website for trials specific to AD and specific to the Canadian Indigenous populations. We used the terms “Alzheimer’s Disease”, “Dementia”, and “Memory Loss” for the Condition or Disease title in combination with “Indigenous”, “Aboriginal”, “First Nation”, “Metis” and “Inuit”; here we found no ongoing clinical trials [43]. This lack of enrollment within Canada for AD specific RCTs reflects a deep-seated problem within the Indigenous communities as far as basic access to medical care and the proper knowledge of, diagnosis and treatment of AD, required before any enrollment into an RCT can occur [91–98]. According to the National Aboriginal Health Organization, “Aboriginal peoples have been studied to death, and many are disillusioned by research reports published by people who don’t know much about their lives and have little concern about the impact of the results on their community” [97]. This may be the case; however, most research studies from the last few decades have involved interviews, observations, and questionnaires, and were not highly regulated and controlled RCTs for new AD drugs or interventions.

This is not to say that RCTs with Indigenous populations in Canada have never taken place. In the excellent publication titled “A Systematic Review of Randomized Controlled Trials of Health-Related Issues Within an Aboriginal Context” from the National Collaborating Centre for Aboriginal Health (NCCAH), a review of all RCTs involving Aboriginal populations was undertaken. The review, based on the Cochrane and Campbell Collaborations method, found that only 35 such RCTs have been published worldwide in the past four decades, with only 6 in Canada [66]. These involved the Ojibway-Cree First Nation in Ontario, the Sagkeeng First Nation Ojibway in Manitoba, the First Nation People registered with the Battlefords Tribal Council in Saskatchewan, and an Inuit community in Nunavut [66]. None of these trials involved AD or dementia. This is not to say that AD specific or other RCTs have not occurred, as many trials fail to enroll participants or the results are never published.

Why the differences between the sexes?

Why do more women than men currently have AD? One obvious explanation is that women live longer than men, and those men that survive to older age have healthier hearts (no CVD) and are less likely to develop AD [123]. This may be true, but likely cannot explain the significant differences in AD numbers between men and women across all age categories. Some theories put forward include:

Loss of estrogen. The hormone estrogen is present in a high amount in premenopausal women, then drops drastically after menopause, right around the time the risk for AD increases in women. Estrogen has neuroprotective properties including: supporting growth and development of cholinergic neurons, increasing cholinergic activity, anti-oxidant properties through the upregulation of superoxide dismutase and glutathione peroxidase, alternative metabolism of amyloid resulting in reduced deposition of toxic Aβ, increased cerebral blood flow and glucose metabolism, and alterations to mitochondrial function [100, 131]. Whereas the drop in testosterone in men occurs slowly, the drop in estrogen occurs rapidly, which may explain the increased risk of AD in women [124]. The loss of estrogen may also be associated with reduced neurogenesis in the hippocampus that contains a high number of estrogen receptors and is important for the formation of new memories, along with a loss in plasticity and cognitive reserve [124, 132]. There is a risk to females of increased Aβ disposition, reduced FDG-glucose metabolism, and lower MRI grey and white matter, with menopause being the highest risk factor, followed by hormone replacement therapy (HRT), hysterectomy, and thyroid problems [133]. HRT may reduce the impact of estrogen loss and provide protection against the development of AD, but is only effective within 5 years of menopause [46]. This is the “window effect” or “timing effect” theory, in which protection from HRT in older women is outweighed by detrimental effects including enhanced clotting and inflammation [131], along with increased incidence of breast and uterine cancers and an increased incidence of CVD [46].

Caregiver effect. 75% of unpaid caregivers for people with chronic debilitating diseases such as dementia are women [123]. This caregiver burden is associated with lower rates of employment and an increased prevalence of psychological AD risk factors, such as sleep disorders and depression, in women caregivers as compared to male caregivers [92, 123]. Dementia care, compared to caregiving for other illnesses, is a particularly demanding form of caregiving. According to a NAC/AARP survey, female dementia caregivers provided an average of 24 hours of care per week, significantly more than other types of caregivers [131, 134]. More than half the dementia caregivers reported that they must maintain vigilance 24 hours per day. Compared to sons, daughters of AD patients provided more routine assistance, experienced more guilt, were less likely to receive support from their spouses, and were more likely to leave the labor market to care for their parent [102]. When the caregiver is a spouse, the risk is no less. Spousal caregivers may be at higher risk of cognitive impairment or dementia than non-caregiver spouses in response to psychosocial (e.g., depression, social isolation, and sleep problems), behavioral (e.g., exercise and diet), and physiological (e.g., metabolic syndrome and inflammatory) variables [102].

Current AD RCT challenges with women

Given the evidence that women are biologically different than men with regards to AD, what does this mean for future RCT enrollment? Historically, men, frequently of the Caucasian race, were considered the “normal” study population [103]. Researchers assumed women should have the same response to drugs as men, and women were considered confounding and more expensive test subjects because of fluctuating hormone levels. We now know that there are differences in the physiology of the sexes that may translate into differences in the pharmacokinetics and/or pharmacodynamics for specific drugs [103]. However, in a 2017 systematic review of sex and gender differences in 48 RCTs of three cholinesterase inhibitors (donepezil, galantamine, and rivastigmine) and memantine, including a total of 20,688 patients, only two studies investigated sex differences for safety and efficacy [123]. A second systematic study found an almost complete lack of data on sex differences in the adverse effects of cholinesterase inhibitors in clinical trials [123]. Therefore, even if we increase the numbers of women in AD RCTs, we must still strongly encourage researchers to extract critical sex-based results from their databases. In a review of nine medical journals in 2009 (not AD specific), most studies that were not sex-specific had an average enrollment of women of 37%; however, 64% did not specify their results by sex and did not explain why the influence of sex in their findings was ignored [103]. In a systematic analysis of 277 stroke specific RCT registered on the ClinicalTrials.gov website from 1990-2018, the enrollment rate for females did not change significantly (∼40%) and only 30% of the trials reported results according to sex [135]. Although Health Canada, the FDA, and the NIH have all developed multiple policies designed to increase women’s enrollment numbers in clinical trials and increase the presentation of RCT data stratified by sex, much work still needs to be done [103, 106]. Many publications have explored sex-based differences in AD along with sex-based disparities in AD-specific and other RCTs, and all have come to the same conclusion: for future studies to be relevant, the number of women enrolled in RCTs need to increase and the management of data must include sex-specific analyses [100, 103–106, 123, 136].