Abstract
This study investigated Alzheimer’s disease (AD) mortality trends by urbanization level and geographical location in the U.S. The CDC’s WONDER database was used to investigate AD mortality from 1999–2019 stratified by urbanization level, census division, race, and sex. Data showed that while AD mortality increased across the U.S., rural areas, particularly in the South, had higher mortality compared to urban counterparts. AD mortality was higher among the female and White population. Data suggested that the urban-rural discrepancy is widening over time. Identifying health disparities underlying the urban-rural discrepancy in AD mortality is critical for allocating social and public health resources.
Keywords
INTRODUCTION
Lifespan has been increasing over the past several decades in the United States (U.S.), thus increasing prevalence of neurodegenerative diseases like Alzheimer’s disease (AD). An estimated 6.2 million Americans age 65 and older are living with AD in 2021 with an expected increase to 12.7 million by 2050 [1]. Two-thirds of Americans with AD are female, and Black Americans are about twice as likely to have AD as White Americans [1]. Mortality from AD is also rising, more than doubling from 2000 to 2019, while mortality from other major diseases such as coronary heart disease is decreasing [1, 2]. Given the substantial rise in AD prevalence and mortality, emphasis should be placed on creating approaches to mitigating the burden of AD on the population.
Although mortality rates in the U.S. have declined in the past several decades, there is a general discrepancy in all-cause mortality rates between urban and rural populations, with many major chronic diseases such as coronary heart disease, heart failure, and stroke disproportionately affecting rural areas more than urban areas [2–7]. However, AD mortality based on urbanization level and geographical location in the U.S. has not been studied in depth. Similarly, reports of AD mortality based on race and sex are limited.
In our study, we aim to investigate the pattern and magnitude of the variation in AD mortality between urban and rural areas, among geographical locations, and based on race and sex in the U.S. Given AD shares many risk factors with other chronic diseases, thus affecting similar populations, we expect AD to follow similar mortality trends.
MATERIALS AND METHODS
We accessed death certificate data via the National Center for Health Statistics (NCHS) to examine AD mortality rates from 1999–2019. The Centers for Disease Control and Prevention’s Wide-ranging Online Data for Epidemiologic Research (WONDER) website was used to process mortality and population statistics [8]. AD mortality was defined as those with an underlying cause of death listed on the death certificate as Alzheimer’s disease (G30) using the International Classification of Diseases, Tenth Revision (ICD-10) codes in persons 65 years and older. Deaths from early-onset AD were excluded to avoid the confounder of familial disease. To examine trends based on urbanization level, we used place of residence from death certificates and classified those based on the 2013 NCHS urban-rural classification scheme for counties [9]. Urban areas were classified as large and medium metros, and rural areas were classified as non-metros. Large metros were defined as counties in a metropolitan statistical area (MSA) of ≥1 million individuals; medium metros as counties in a MSA of 250,000–999,999 individuals; and non-metros as counties in a micropolitan statistical area of 10,000–50,000 individuals. Analyses were further stratified by census division as defined by the U.S. Census Bureau [10].
The main outcome measure was AD mortality rate defined as the number of deaths divided by the total number of people at risk. Rates were expressed as the number of deaths per 100,000 population. Crude and age-adjusted mortality rates from AD were calculated in the overall population, and then stratified by the three-level urbanization classification and by the nine census divisions. Age-adjusted AD mortality rates were plotted against year to examine trends from 1999 to 2019. Percentage change in mortality rates during the observation period were calculated by taking the difference between the 1999 and 2019 rates and dividing that result by the 1999 rate. Poisson regression modeling was used to describe trends in age-adjusted annual AD death rates and assess the significance of the region-urbanization interaction.
RESULTS
From 1999 to 2019, there was an overall 86.0%increase in age-adjusted AD mortality rates in individuals 65 years and older in the U.S., from 129 to 240 deaths per 100,000.
AD mortality rate by urbanization level
From 1999 to 2019, age-adjusted AD mortality rates increased across all levels of urbanization, but medium metros and non-metros were found to have consistently higher AD mortality rates compared to large metros (Fig. 1). The percentage increase in AD mortality rates was markedly higher in non-metros, with a 94%increase from 136 to 264 deaths per 100,000, compared to the 84%increase in AD mortality in both medium metros (137 to 252 deaths per 100,000) and large metros (120 to 221 deaths per 100,000). The divergence between AD mortality rates in large metros and non-metros also increased over the study period. From 1999 to 2019, AD mortality rates in non-metros increased from 13%higher to 20%higher than that in large metros. Age-adjusted AD mortality rates in rural areas were higher than in urban areas (p < 0.05).
Age-adjusted AD mortality rates per 100,000 population in the U.S. by urbanization level, 1999–2019. All population means over the study period fell within 95%confidence intervals.
AD mortality rate by census division
From 1999 to 2019, age-adjusted AD mortality rates increased across all census divisions in the U.S., but divisions in the south were found to have consistently higher AD mortality rates compared to other census divisions during this period (Table 1). In 2019, AD mortality rate was highest in the East South Central division at 322 deaths per 100,000, more than double that in the Middle Atlantic division in which AD mortality rate was the lowest at 132 deaths per 100,000. AD mortality rate increased by 126%in the East South Central division over the study period, markedly higher than the Middle Atlantic division in which AD mortality rate increased by 70%.
Age-adjusted AD mortality rates per 100,000 population in the U.S. by census division, 1999– 2019
AD mortality rate by urbanization level and census division
In most census divisions in the U.S. in 2019, medium and non-metros were shown to have higher AD mortality rates compared to large metros (Fig. 2). AD mortality rates were markedly higher in non-metros in the south, particularly in the East South Central division (355 deaths per 100,000), especially when compared to large metros of other divisions. From 1999 to 2019, the highest increase in AD mortality rates was in non-metros in the Pacific division at 164%while the lowest increase in AD mortality rates was in medium metros in the New England division at 10%.
Age-adjusted AD mortality rate per 100,000 population in the U.S. by census division and urbanization level, 2019.
AD mortality rate by race and sex
Within the study period, age-adjusted AD mortality rates were consistently higher in White populations than in Black or African American populations for every census division (Table 1). AD mortality rates were highest among Black or African American individuals living in the Pacific division at 247.7 deaths per 100,000, whereas mortality rates were highest among White individuals living in the East South Central division at 276.3 deaths per 100,000. AD mortality rates were also consistently higher in the female population compared to the male population for every census division (Table 1). Both females and males living in the East South Central division experienced the highest mortality rate, at 291.8 and 214.7 deaths per 100,000 respectively.
DISCUSSION
In our study, AD mortality rates were shown to be increasing in individuals age 65 years and older in the U.S. in the past two decades, with disproportionate trends in mortality demonstrated based on urbanization, census division, race, and sex. AD mortality rates were consistently higher in rural areas from 1999–2019, and divisions in the South had higher AD mortality rates than other census divisions. Rural areas within these southern divisions had the highest AD mortality. These urban-rural trends have also been demonstrated in mortality from other major chronic diseases in the U.S. such as coronary heart disease, heart failure, and stroke, prompting consideration of the reasons for such a discrepancy [2, 6, 7].
Interestingly, our study revealed that AD mortality was consistently higher in the White population than Black population, even though AD prevalence has been reported higher in Black Americans [1], This discrepancy could be attributed to Black individuals with AD having longer survival estimates than White individuals with AD [11]. Reasons for this survival advantage have not been fully explored, but possibilities include apolipoprotein E genotype; the lower rate of nursing home placement among Black Americans, or rather increased access to nursing homes among White Americans, when nursing home placement is associated with increased mortality; and differences in comorbidity profiles among Black and White AD patients, which may lead to more aggressive treatment of comorbidities at advanced stages of AD in the Black population [11].
AD mortality was also higher in females than males, which is consistent with AD prevalence trends. Possible reasons for this discrepancy include, to some extent, increased life expectancy in females, thus increasing AD prevalence and mortality; sex-dependent penetrance of APOE and other sex-specific genetic contributors to AD; increased severity of symptoms among females with AD; and higher likelihood of brain pathology related to AD in females [12].
The urban-rural discrepancy in AD mortality could be related to several other urban-rural health disparities. These include low socio-economic status, poor education, poor health status, longer time to diagnosis, difficulties with follow-up, limited internet availability, and differential access to health services including primary care [3, 13–16]. Education has been shown to improve cognitive health in older individuals thus aiding in AD prevention, but rural residents are less likely to have higher education degrees compared to urban residents [4, 13]. Rural regions also tend to have a relative lack of healthcare providers and resources, contributing to higher financial costs, longer travel time, and potentially delayed diagnosis and treatment of AD for rural residents [3, 15]. In a study examining urban-rural differences in risk for ambulatory care sensitive hospitalizations, veterans with dementia in rural counties were found to have more preventable hospitalizations, indicating that access to timely and appropriate primary care may be lacking in rural areas [17]. Urban-rural differences in societal norms and values are another potential reason for these AD mortality trends. Those largely held by rural inhabitants, such as the importance of self-reliance, hesitancy to seek medical assistance, and advocacy for less aggressive medical care to allow for more “natural death” may result in delayed medical attention relative to urban counterparts [17, 18]. Many health disparities disproportionately experienced by the South could also contribute to higher AD mortality, including relatively high poverty, rural isolation, and high prevalence of cigarette smoking [14].
The urban-rural divergence in AD mortality is also widening over time. In a study investigating all-cause mortality trends by urbanization level, overall mortality in rural areas compared to metropolitan areas increased from 2%higher in 1990–1992 to 13%higher in 2005–2009 [3]. Our study also revealed this discrepancy, where AD mortality in rural areas increased from 13%higher than urban areas in 1999 to 20%higher than urban areas in 2019. This discrepancy is expected to continue increasing in the future, suggesting that many social and public health interventions aren’t reaching more vulnerable rural areas [3].
This study does have limitations. Death certificate data was used given its abundance and ease of access. However, limitations regarding accuracy and completeness of this data do exist, include potentially inaccurate documented cause of death and potentially inaccurate or missing data of geographic location or urbanization level, which could result in misclassification bias. Mortality trends among the Asian/Pacific Islander population could also not be assessed given the limited data reported in that subgroup. This study also lacked data on any potential confounding variables or underlying risk factors for these urban-rural differences, such as lifestyle, aspects of healthcare, or health behavior.
In conclusion, our study demonstrates that urbanization level should be considered when studying health disparities, and AD mortality is a significant example of this discrepancy. Given the increasing prevalence and mortality of AD in the U.S., examining and addressing potential reasons for the urban-rural discrepancy in AD mortality is paramount. Future research should investigate underlying causes of this disparity, such as healthcare inequalities or differences in behavioral or lifestyle risk factors. Future research can also delve into the impact of COVID-19 on the urban-rural discrepancy, which has likely further increased the disparity between these populations. Such research can provide insight into the efficacy of our health interventions and allocation of health services and resources to more vulnerable communities.
