Antihypertensive Use and the Risk of Alzheimer’s Disease and Related Dementias Among Older Adults in the United States

Abstract

Background

Epidemiological evidence on different classes of antihypertensives and risks of Alzheimer’s disease and related dementias (ADRD) is inconclusive and limited. This study examined the association between antihypertensive use (including therapy type and antihypertensive class) and ADRD diagnoses among older adults with hypertension.

Methods

A retrospective, cross-sectional study was conducted, involving 539 individuals aged 65 and older who used antihypertensives and had ADRD diagnosis selected from 2013–2018 Medical Expenditure Panel Survey (MEPS) data. The predictors were therapy type (monotherapy or polytherapy) and class of antihypertensives defined using Multum Lexicon therapeutic classification (with calcium channel blockers [CCBs] as the reference group). Weighted logistic regression was used to assess the relationships of therapy type and class of antihypertensives use with ADRD diagnosis, adjusting for sociodemographic characteristics and health status.

Results

We found no significant difference between monotherapy and polytherapy on the odds of ADRD diagnosis. As to monotherapy, those who used angiotensin-converting enzyme inhibitors (ACEIs) had significantly lower odds of developing AD compared to those who used CCBs (OR=0.36, 95% CI = 0.13 – 0.99).

Conclusions

Findings of the study suggest the need for evidence-based drug therapy to manage hypertension in later adulthood and warrant further investigation into the mechanism underlying the protective effect of antihypertensives, particularly ACEIs, against the development of AD among older adults with hypertension.

1. Introduction

Alzheimer’s disease and related dementias (ADRD) describe a group of disorders characterized by pathological brain changes that lead to cognitive decline and significantly interfere with individuals’ activities of daily living and quality of life [1–3]. Decreased quality of life is found among people living with ADRD, and is associated with ADRD symptoms, loss of independence, depression, use of antipsychotic or anxiolytic medication, and caregiver characteristics [4]. An estimated 50 million people worldwide are currently affected by dementia, with nearly 10 million new cases diagnosed each year [5]. Alzheimer’s disease (AD) is the most common form of dementia accounting for nearly 70% of cases [1, 6,7]. In the United States alone, over 6.2 million Americans aged 65 and older are living with AD, a figure estimated to triple within the next 30 years [8]. This is a major public health problem because ADRD is among the leading causes of mortality for older adults.

Cardiovascular risk factors have been found to be associated with the risk of cognitive decline in individuals with no dementia, and more than 50% clinically diagnosed AD cases present with vascular changes and disrupted homeostasis of the renin-angiotensin system (RAS), which is shown to play an important role in AD pathophysiology [9]. Hypertension is one of the major modifiable risk factors for cardiovascular disease, late-life cognitive impairment, and development of ADRD [6, 10–12]. This is not surprising, given that chronic hypertension exerts numerous damaging systemic effects that increase the brain’s vulnerability to ADRD and other neurodegenerative disease. These effects are not limited to but include reduced cerebral blood flow [13], impaired neurovascular regulation [14], enhanced systemic inflammation [15], increased oxidative stress of the vascular-glial blood brain barrier [16,17] and compromised clearance of pathological amyloid beta and neurofibrillary tau proteins—the hallmark neuropathology of AD [18].

One promising strategy for preventing or delaying the onset of ADRD involves therapeutic management of hypertension symptoms using antihypertensives. Current literature indicates that untreated long-standing hypertension deteriorates arteriolosclerosis, causes blood pressure variation in late life, and fails to sustain cerebral metabolic demands, which increases the risk of cognitive decline and dementia [10,19,20]. The use of antihypertensives can protect blood pressure against variation in late life, thus may reduce the risk of ADRD. The most commonly prescribed antihypertensives include angiotensin-converting enzyme inhibitors (ACEIs), angiotensin receptor blockers (ARBs), calcium channel blockers (CCBs), β-blockers (BBs), and thiazide diuretics (TDs) [6,21–23]. ACEIs and ARBs are RAS-acting drugs that both block and mitigate the deleterious effects of the principal effector peptide of the RAS, angiotensin II (although each drug works to do this through different mechanisms of action). Angiotensin II causes elevated blood pressure, retention of water and sodium, increased aldosterone and vasopressin secretion [9, 24]. The CCBs have a different mode of action from the RAS-acting drugs. They block L-channels along with direct interaction with ligand-gated ion channels that mediate a Ca²⁺–permeable component of excitatory neurotransmission in the central nervous system [25]. However, previous epidemiological studies on whether these antihypertensives have similar efficacy in hypertension control and cognitive function when deployed at older ages are inconsistent [20,22,26], and this may be explained by variations in study design (clinical vs. cohort studies), study duration, data acquisition methods, definitions of hypertension and antihypertensives variables, and measures of cognitive function [22]. Preliminary reports show that specific class of antihypertensives may be more effective in reducing the risk for cognitive decline in older patients with hypertension. A meta-analysis of 27 international observations, pre-clinical, and clinical studies found that, among several classes of antihypertensives, only diuretic use was associated with reduction in cognitive decline or dementia among people aged 65 or older [21]. Another meta-analysis of 15 observational studies on people aged 58 and older indicated that ARBs but not ACEIs use significantly reduced the risk of any dementia and AD compared with other antihypertensives [1]. By contrast, a recent systematic review indicated that ARBs (telmisartan in particular) and CCBs (nilvadipine in particular) may protect older individuals with hypertension from developing AD [6]. Additionally, one of the latest observational studies using prospective longitudinal data found that the use of antihypertensive monotherapy (either BBs, CCBs, or ACEIs alone) was not associated with changes in cognitive function among adults aged 50 or older [22].

Given the complexity of the relationship between antihypertensives and dementia, the question of whether specific class of antihypertensives is associated with risks of having ADRD remains unanswered. Therefore, the current study aims to explore how class of antihypertensives is associated with risks of having ADRD and further investigate whether therapy type (monotherapy vs. combinational therapy) differentiate odds of having ADRD diagnosis among hypertensive older adults in the United States. Although age is identified as an essential risk factor for development of late-life ADRD, there is limited evidence on relationship between antihypertensive pharmacology and ADRD risk, leaving the status of this association inconclusive. Therefore, this study was not designed to assess any hypothesizes regarding the modifying effect of age on the relationship between antihypertensive use and ADRD.

2. Methods

2.1. Data Source and Study Design

The current study used retrospective cross-sectional design. Data used for the analysis were selected from the Medical Expenditure Panel Survey (MEPS, 2013–2018). MEPS, which began in 1996 and was conducted on an annual basis, is a nationally representative survey of the civilian non-institutionalized population of the United States [27]. MEPS uses a probability weighted complex multistage survey design with primary sampling units, strata, and person level sampling weights [28]. The Household component of MEPS consists of the household full-year core files and multiple sub-data files. Each file collects detailed information for each person in the household on sociodemographic characteristics, health conditions, health status, use of medical services, medical expenses and source of payments, health insurance coverage [27]. We pooled six years of individual-level household data (2013–2018). We did not include a longer period for this study because antihypertensive prescription and treatment patterns might be renewed over time [28], and the 2018 wave was the most recently available MEPS dataset we could obtain. For each year, we linked the household core file to two sub-data files, Prescribed Medicine data and Medical Conditions file, as these files collected information on hypertension and ADRD diagnoses and class of antihypertensives. Prescribed Medicine data and Medical Conditions files are public data files that provide information on household-reported prescribed medicine and medical conditions, confirmed with medical records from respondents’ medical providers (e.g., doctors, hospitals, pharmacies, etc.) [27]. These data were collected based on respondents’ verbatim text and converted into Internal Classification of Diseases (ICD). The ICD-9 was used to code and classify disease diagnosis until 2016, when use of ICD-10 for disease coding started [27, 29]. In the Prescribed Medicine data, each record represents a unique prescribed medicine event. The data include an identifier for each unique prescribed medicine; detailed characteristics associated with the medicine (e.g., medicine name, etc.); selected Multum Lexicon variables (therapeutic medication classification system); the date on which the person first used the medicine; total expenditure and sources of payments; types of pharmacies that filled the household’s prescriptions; and a full-year person level weight [27].

We included respondents aged 65 and older in our study because ADRD incidence increases with age and a majority of patients develop ADRD after the age 65. Studies have reported that mutations in high-risk genes (i.e., APP, PSEN1, or PSEN2 genes) may cause early-onset AD; patients with these mutations have 100% chance of developing AD before age 65, but this rare form of disease applies to less than 5% of all AD cases and usually occurs independently of risk factors such as hypertension [30–32].

The total number of respondents to the Household component from 2013 to 2018 was 195,513 including 27,149 adults at the age of 65 and above, of those, 16,949 reported their age when they started to have physician diagnosed hypertension and whether they used any antihypertensives to control for their hypertensive symptoms. There were 16,114 (95.1%) adults aged 65 or above who reported that they had hypertension diagnosis in their midlife and used one or multiple antihypertensives, of these, 539 had ADRD diagnoses between 2013 and 2018. We restricted our analysis to the 539 individuals. Given the minimum missing data on the key variables (approximately 1%), we deleted them from the analyses.

2.2. Measures

2.2.1. Outcome Variables

Three outcome variables in the study included ADRD diagnosis, AD diagnosis, and diagnosis of other types of dementia including vascular dementia and unspecified dementia. Each variable was identified using the ICD-9-CM codes (290 and 291) for ADRD; ICD-9-CM (331) and ICD-10-CM (G30) for AD; and ICD-10-CM (F01, 02, 03, 04, and 10) for other types of dementia. Older adults without these diagnosis codes were considered to have no ADRD.

2.2.2. Explanatory Variables

The explanatory variables in the study were class of antihypertensives and therapy type for controlling hypertension symptoms. The class of antihypertensives was defined using Multum Lexicon therapeutic classification (code 40 for cardiovascular agents) in the MEPS. According to Ding et al. [26], medication class was categorized into six groups: ACEIs, ARBs, CCBs, TDs, BBs, and others. The current study started with Multum therapeutic class of 40 (cardiovascular agents) and used the Multum therapeutic subclasses in the data to identify specific medication class: 42 (ACEIs), 43 (antiadrenergic agents peripherally acting – tamsulosin was excluded), 44 (antiadrenergic agents centrally acting), 46 (antiarrhythmic agents - only diltiazem and verapamil were included), 47 (BBs), 48(CCBs), 49 (diuretics), 53 (vasodilators), 54 (vasopressors), 55 (combination), and 56 (ARBs).

Following similar studies on antihypertensive utilization [28,33], therapy type was categorized into monotherapy and polytherapy. Monotherapy refers to patients utilizing a single active ingredient of antihypertensives in a given year. Polytherapy refers to those who used two or more active ingredients of antihypertensives in a given year, including single-pill combination, multiple-pills combination, and switching between active ingredients within the same medication class and between different medication classes [33].

2.2.3. Covariates

We also controlled for the impact of respondents’ age, gender, educational attainment, race, marital status, family income, geographic region, health insurance status, smoking status, and physical and mental health on ADRD diagnoses. Age was defined as chronological age. Gender included males and females. Educational attainment was indicated with three categories: less than high school, high school, and some college. Race comprised non-Hispanic Whites, non-Hispanic Blacks, Hispanics, and others. Marital status was defined as whether the respondents were married or not at the time of interview. Family income was defined as annual household income and indicated with three categories using the federal poverty level (FPL) of each survey year (low income: <200% FPL, middle income: 200%≤FPL<400%, and high income: FPL≥400%). Inflation change in family income over time was adjusted. Geographic region referred to four regions in the nation: Northeast, Midwest, South, and West. Health insurance status was defined as whether the respondents were covered by any type of insurance in a given year. Given that the study population were 65 years old and above, they were automatically enrolled in Medicare. Thus, health insurance was categorized into Medicare and private health insurance. Additionally, smoking status was defined as whether the respondents were currently smoking cigarettes by the time of interview. The impacts of whether the respondents had coronary heart disease, diabetes, stroke, or depression on ADRD diagnosis were also controlled.

2.3. Statistical Analysis

Descriptive analyses were conducted to estimate the mean, frequency, and attrition. T-test and Chi-square tests were used to determine the statistically significant differences between individuals with ADRD diagnosis and those without ADRD diagnosis. We used weighted logistic regression (WLR) models to examine the relationships of ADRD diagnosis and therapy modality (monotherapy vs. polytherapy, with polytherapy as the reference group) or class of antihypertensives. WLR is an effective and powerful method to produce the accurate estimates and their standard errors with diminished bias and variance for large-scale imbalanced data or rare events and complex survey designed with stratification, cluster, and unequal weighting [34,35].

In addition, among respondents on monotherapy, to explore the potentially different protection effects against ADRD in relation to class of antihypertensives, we examined the association between each specific class of antihypertensive drug and each ADRD outcome. We set individuals on CCBs as the reference group because the existing evidence has shown no beneficial effect of CCBs on cognitive health among individuals aged 80 and above [36], while some study has identified null effects of CCBs on cognitive function among a younger cohort [37]. MEPS sampling weights were used in the analyses. All analyses and procedures were conducted in SAS, version 9.4 (SAS Institute, Inc., Cary, North Carolina).

3. Results

Characteristics of the study population was summarized in Table 1. Approximately 3% of older adults who used antihypertensives (N=16,114) were diagnosed with ADRD (N=539 including 31% of AD and 69% of other dementias) between 2013 and 2018. Given that our sample consisted of hypertensive patients, the proportion of other dementias was higher due to a greater percentage of individuals with vascular dementia. The average age of adults with ADRD diagnoses was 80 years old, which was older than those without ADRD diagnoses (74 years old). There were higher proportions of women, Non-Hispanic Whites, individuals with less than high school education, Medicare users, current smokers, and those with comorbidities in individuals with ADRD as compared to those without ADRD. Table 2 demonstrated the characteristics of therapy type and class of antihypertensives by ADRD diagnosis. Regardless of ADRD diagnoses, proportions of polytherapy users were higher compared to monotherapy users (55% vs. 45%). As for class of antihypertensives, the proportion of ACEIs utilization was lower but the proportions of CCBs and BBs utilization were higher in those with ADRD diagnoses as compared to those without (approximately 9% vs. 15% for ACEIs; 11% vs. 7% for CCBs; 15% vs. 12% for BBs). The results of Chi-square tests showed that the differences in distributions of therapy type and class of antihypertensives were not statistically significant between the ADRD and non-ADRD groups.

Table 1.

Sample Characteristics of Antihypertensive Users by ADRD Status, Medical Expenditure Panel Survey (MEPS, 2013–2018)

Category	Overall		ADRD		Alzheimer’s Disease		Other Dementias		No Dementia		p b
	N	Weighted %	N	Weighted %	N	Weighted %	N	Weighted %	N	Weighted %
Total	16,114	100%	539	100%	169	100%	406	100%	15,575	100%
Age (mean (SD))		74.5(0.1)		80.7 (0.3)		81.1 (0.2)		80.6 (0.2)		74.3 (0.1)	0.00
Gender											0.01
Female (ref.)	9,050	54.4	354	63.0	110	66.0	272	63.2	8,696	54.1
Male	7,064	45.6	185	37.0	59	34.0	134	36.8	6,879	45.9
Race											0.12
Non-Hispanic White (ref.)	9,247	74.9	292	75.3	87	73.4	221	75.4	8,955	74.9
Hispanic	2,296	7.7	85	7.7	30	9.4	60	7.1	2,211	7.7
Non-Hispanic Black	3,198	10.9	123	12.8	35	11.0	97	13.6	3,075	10.8
Other	1,373	6.5	39	4.2	17	6.1	28	3.8	1,334	6.6
Region											0.67
Northeast (ref.)	2,804	18.9	107	22.0	51	35.5	67	17.6	2,697	18.7
Midwest	3,241	22.4	96	20.6	21	13.4	80	23.8	3,145	22.4
South	6,401	37.9	200	36.9	64	31.9	151	38.8	6,201	38.0
West	3,668	20.8	136	20.6	33	19.2	108	19.9	3,532	20.9
Education											<0.01
Less than high school (ref.)	3,126	18.7	147	32.7	40	28.7	115	34.9	2,979	18.2
High school	3,742	31.1	112	28.0	35	32.5	92	28.6	3,630	31.2
Some college and above	5,496	50.2	137	39.3	45	38.8	94	36.5	5,359	50.6
Family Income											<0.01
Poor/Low income (ref.)	6,522	32.8	274	45.3	88	51.0	206	45.7	6,248	32.3
Middle income	4,618	28.7	163	30.6	46	25.6	124	30.5	4,455	28.7
High income	4,974	38.5	102	24.0	35	23.4	76	23.8	4,872	39.0
Health insurance coverage											<0.01
Private insurance (ref.)	7,362	53.2	180	39.1	43	29.1	144	40.8	7,182	53.7
Medicare	8,752	46.8	359	60.9	126	70.9	262	59.2	8,393	46.3
Currently married	7,885	53.2	181	39.0	54	29.8	135	41.2	7,704	53.7	<0.01
Current smokers	5,285	30.7	183	35.1	53	34.4	145	36.9	5102	30.5	0.10
Comorbidities
Coronary heart disease	2,915	19.3	137	24.9	37	18.9	110	28.0	2778	19.1	0.02
Diabetes	4,280	24.6	168	28.5	55	25.3	120	29.4	4112	24.5	0.17
Stroke	1,979	12.3	187	36.1	45	27.6	153	39.3	1792	11.4	<0.01
Depression	2,153	14.1	127	25.2	34	23.3	98	25.2	2026	13.7	<0.001

Abbreviations: ADRD, Alzheimer’s disease and related dementia.

^a.Family income level was calculated based on the annual federal poverty level (FPL) which was adjusted the inflation between 2013–2018: low income: FPL<200%, middle income: 200%≤FPL<400%, high income: FPL≥400%.

^b.p value for age was calculated by t-test and other p values were calculated by Chi-square test, showing the difference in variable distribution between the dementia and no-dementia groups.

Table 2.

Utilization of Antihypertensives by ADRD Status in Adults (65 and above) in the United States (MEPS, 2013–2018)

Category	Overall		ADRD		Alzheimer’s Disease		Other Dementias		No Dementia		p c
	N	Weighted %	N	Weighted %	N	Weighted %	N	Weighted %	N	Weighted %
Overall	16,114	100%	539	100%	169	100%	406	100%	15,575	100%
Therapy type											>0.05
Monotherapy a	7,128	44.6	236	44.7	81	48.7	174	43.5	6,892	44.6
Polytherapy b	8,986	55.4	303	55.3	88	51.3	232	56.5	8,683	55.4
Class of Antihypertensives											>0.05
ACEIs	2,269	14.4	66	9.4	23	9.2	48	9.3	2,203	14.6
ARBs	1,226	7.8	30	7.3	12	9.5	21	7.4	1,196	7.8
CCBs	1,302	7.4	57	10.5	18	12.6	44	9.6	1,245	7.3
TDs	434	2.6	11	2.3	3	3.1	8	1.7	423	2.6
BBs	1,897	12.3	72	15.2	25	14.3	53	15.4	1,825	12.2

Abbreviations: ADRD, Alzheimer’s disease and related dementia; ACEIs, angiotensin-converting enzyme inhibitors; ARBs, angiotensin receptor blockers; CCBs, calcium channel blockers; TDs, Thiazide diuretics; BBs, β-blockers.

^a.Patients on single pill with single active antihypertensive agent only.

^b.Patients on multiple pills with active antihypertensive agent or multiple pills with multiple active antihypertensive agents.

^c.p value was calculated by Chi-square test to compare the difference in variable distribution between the ADRD and non-ADRD groups.

Table 3 presented results from a series of WLR models assessing the relationships of therapy type and class of antihypertensives with ADRD diagnosis after fully adjusting the impacts of covariates. We found that neither therapy type nor any class of antihypertensives was significantly associated with ADRD diagnosis. However, when we limited our analyses to those on monotherapy and compared every drug class with the reference category of CCBs users (Table 4), those on ACEIs had lower odds of having AD diagnosis (OR= 0.36, 95% CI: 0.13 – 0.99) rather than the diagnoses of ADRD or other types of dementia. Other drug classes showed no significant associations with any of the ADRD outcomes compared to CCBs in Table 4.

Table 3.

Odds Ratios of Weighted Logistic Regression on Therapy Type and Medication Class and ADRD Diagnoses (MEPS, 2013–2018)

	ADRD	Alzheimer’s Disease	Other Dementia	ADRD	Alzheimer’s Disease	Other Dementia
	OR (95% CI)	OR (95% CI)	OR (95% CI)	OR (95% CI)	OR (95% CI)	OR (95% CI)
Monotherapy vs. polytherapy	1.15 (0.86– 1.54)	1.34 (0.87 – 2.07)	1.09 (0.78 – 1.53)
Class of antihypertensives a.
ACEIs				0.78 (0.51 – 1.19)	0.78 (0.42 – 1.45)	0.78 (0.47 – 1.28)
ARBs				1.15 (0.63 – 2.10)	1.36 (0.64 – 2.90)	1.21 (0.59 – 2.48)
CCBs				1.54 (0.95 – 2.52)	1.89 (0.84 – 4.28)	1.37 (0.81 – 2.30)
TDs				0.84 (0.30 – 2.35)	1.16 (0.21 – 6.28)	0.64 (0.24 – 1.71)
BBs				1.17 (0.78 – 1.76)	1.11 (0.61 – 2.02)	1.16 (0.72 – 1.87)
Age (year)	1.17** (1.14 – 1.20)	1.17** (1.12 – 1.23)	1.17** (1.14 – 1.20)	1.17**(1.14 – 1.20)	1.17** (1.12 – 1.23)	1.17** (1.14 – 1.20)
Male b.	0.86 (0.63 – 1.18)	0.94 (0.54–1.65)	0.80 (0.57 – 1.12)	0.87 (0.63 – 1.19)	0.94 (0.53–1.65)	0.81 (0.56 – 1.13)
Race c.
Hispanic	0.92 (0.62 – 1.39)	1.08 (0.54 – 2.18)	0.86 (0.54 – 1.37)	0.92 (0.62 – 1.38)	1.08 (0.54 – 2.16)	0.86 (0.54 – 1.37)
Non-Hispanic Black	1.39 (0.95 – 2.03)	1.08 (0.52 – 2.25)	1.47 (0.97 – 2.22)	1.38 (0.94 – 2.01)	1.07 (0.51 – 2.22)	1.45 (0.96 – 2.20)
Other	0.73 (0.46 – 1.14)	1.08 (0.53 – 2.21)	0.69 (0.41 – 1.16)	0.72 (0.46 – 1.13)	1.07 (0.52 – 2.19)	0.68 (0.40 – 1.15)
Region d.
Midwest	0.89 (0.56 – 1.41)	0.37* (0.16 – 0.86)	1.31 (0.76 – 2.24)	0.89 (0.55 – 1.42)	0.37* (0.16 – 0.86)	1.31 (0.76 – 2.25)
South	0.97 (0.68 – 1.38)	0.50* (0.27 – 0.93)	1.30 (0.86 – 1.99)	0.97 (0.68 – 1.39)	0.51*(0.27 – 0.94)	1.31 (0.86 – 2.00)
West	1.02 (0.65 – 1.58)	0.52 (0.25 – 1.11)	1.29 (0.80 – 2.10)	1.03 (0.67 – 1.60)	0.54 (0.25 – 1.14)	1.31 (0.81 – 2.12)
Education e.
High school	0.77 (0.54 – 1.11)	0.97 (0.53 – 1.79)	0.77 (0.52 – 1.15)	0.77 (0.54 – 1.11)	0.99 (0.53 – 1.83)	0.77 (0.52 – 1.15)
Some college	0.94 (0.67 – 1.33)	1.07 (0.64 – 1.80)	0.84 (0.56 – 1.26)	0.94 (0.67 – 1.33)	1.08 (0.64 – 1.82)	0.84 (0.56 – 1.26)
Family Income f.
Middle income	1.06 (0.79 – 1.43)	0.82 (0.50 – 1.35)	1.05 (0.74 – 1.47)	1.06 (0.78 – 1.43)	0.82 (0.50 – 1.35)	1.04 (0.74 – 1.47)
High income	0.83 (0.58 – 1.20)	0.85 (0.45 – 1.61)	0.81 (0.53 – 1.22)	1.90 (0.83 – 1.72)	0.87 (0.46 – 1.64)	0.81 (0.54 – 1.23)
Medicare g.	1.48* (1.10 – 2.00)	2.20** (1.31 – 3.70)	1.36 (0.98 – 1.88)	1.48 (1.10 – 2.00)	2.20** (1.31 – 3.71)	1.36 (0.98 – 1.89)
Currently married h.	1.25 (0.89 – 1.75)	0.81 (0.48 – 1.36)	1.43 (0.96 – 2.12)	1.25 (0.89 – 1.75)	0.82 (0.49 – 1.36)	1.43 (0.97 – 2.12)
Current smokers i.	1.55** (1.18 – 2.04)	1.27 (1.78 – 2.06)	1.76** (1.30 – 2.39)	1.56** (1.19 – 2.05)	1.28 (0.79 – 2.08)	1.77** (1.30 – 2.41)
Comorbidities j.
Coronary heart disease	1.03 (0.77 – 1.38)	0.78 (0.45 – 1.34)	1.22 (0.87 – 1.71)	1.02 (0.76 – 1.36)	0.77 (0.45 – 1.33)	1.20 (0.85 – 1.68)
Diabetes	1.25 (0.91 – 1.70)	1.10 (0.66 – 1.84)	1.26 (0.89–1.77)	1.24 (0.91 – 1.69)	1.08 (0.64 – 1.80)	1.25 (0.89 – 1.77)
Stroke	3.21** (2.36 – 4.39)	2.03* (1.21 – 3.40)	3.54** (2.51–4.99)	3.20** (2.34 – 4.37)	1.98* (1.18 – 3.34)	3.54** (2.51 – 4.98)
Depression	2.30** (1.71 – 3.08)	1.95* (1.08 – 3.51)	2.25** (1.59–3.21)	2.29** (1.70 – 3.07)	1.95* (1.07 – 3.53)	2.24** (1.58 – 3.18)

Abbreviations: ADRD, Alzheimer’s disease and related dementia; ACEIs, angiotensin-converting enzyme inhibitors; ARBs, angiotensin receptor blockers; CCBs, calcium channel blockers; TDs, Thiazide diuretics; BBs, β-blockers; OR, odds ratio; CI, confidence interval.

^*p <.05,

^**p <.01

^a.Reference groups are other medication classes.

^b.Female is the reference group.

^c.Non-Hispanic Whites is the reference group.

^d.Northeast is the reference group.

^e.Less than high school is the reference group.

^f.Low income (FPL<200%) is the reference group.

^g.Private health insurance users is the reference group.

^h.The non-married including divorced, widowed, separated, and never married is the reference group.

^i.Non-smokers including abstainers and never smokers are the reference group.

^j.Reference groups are individuals without the diagnosis of coronary heart disease, diabetes, or depression.

Table 4.

Odd Ratios of Weighted Logistic Regression on Class of Antihypertensives and ADRD Diagnoses in Monotherapy Users (MEPS, 2013–2018)

	ADRD	Alzheimer’s Disease	Other Dementias
	OR (95% CI)	OR (95% CI)	OR (95% CI)
CCBs (ref.)
ACEIs	0.66 (0.32 – 1.35)	0.36 * (0.13 – 0.99)	0.92 (0.40 – 2.15)
ARBs	0.86 (0.34 – 2.14)	0.63 (0.21 – 1.89)	1.23 (0.43 – 3.54)
TDs	0.82 (0.25 – 2.76)	0.81 (0.13 – 5.03)	0.73 (0.23 – 2.33)
BBs	0.69 (0.33 – 1.47)	0.45 (0.17 – 1.20)	0.86 (0.36 – 2.04)
Age (year)	1.21** (1.16– 1.26)	1.24** (1.14 – 1.34)	1.19** (1.14 – 1.25)
Male a.	1.10 (0.64 – 1.89)	1.01 (0.40 – 2.52)	1.03 (0.60 – 1.75)
Race b.
Hispanic	1.10 (0.54 – 2.20)	1.31 (0.44 – 3.89)	1.32 (0.56 – 3.09)
Non-Hispanic Black	1.28 (0.67 – 2.47)	1.26 (0.37 – 4.30)	1.64 (0.80 – 3.37)
Other	0.46 (0.19 – 1.11)	0.38 (0.09 – 1.56)	0.69 (0.25 – 1.92)
Region c.
Midwest	0.81 (0.41 – 1.59)	0.17** (0.05 – 0.52)	1.27 (0.58 – 2.78)
South	0.65 (0.36 – 1.21)	0.32* (0.12 – 0.84)	0.77 (0.39 – 1.51)
West	0.82 (0.43 – 1.57)	0.66 (0.26 – 1.68)	0.73 (0.34 – 1.58)
Education d.
High school	0.80 (0.42 – 1.52)	1.43 (0.49 – 4.16)	0.81 (0.39 – 1.70)
Some college	0.83 (0.43 – 1.59)	1.78 (0.73 – 4.32)	0.59 (0.26 – 1.36)
Family Income e.
Middle income	1.02 (0.59 – 1.76)	0.79 (0.32 – 1.92)	1.05 (0.57 – 1.94)
High income	0.84 (0.44 – 1.60)	1.25 (0.51 – 3.03)	0.75 (0.38 – 1.51)
Medicare f.	1.53 (0.92 −2.54)	2.25* (1.11 – 4.56)	1.26 (0.66 – 2.43)
Currently married g.	1.45 (0.85 −2.48)	1.00 (0.46 – 2.19)	1.71 (0.93 – 3.15)
Current smokers h.	1.17 (0.75 −1.82)	1.02 (0.50 – 2.07)	1.31 (0.78 – 2.21)
Comorbidities i.
Coronary heart disease	1.19 (0.72 −1.98)	0.80 (0.30 – 2.14)	1.70 (0.98 – 2.95)
Diabetes	1.02 (0.59 −1.78)	1.58 (0.69 – 3.60)	0.92 (0.50 – 1.67)
Stroke	2.71**(1.59–4.60)	1.43 (0.57 – 3.64)	3.24** (1.77 – 5.94)
Depression	3.13**(1.87 −5.24)	2.44 * (0.98 – 6.07)	2.99** (1.67 – 5.35)

Abbreviations: ADRD, Alzheimer’s disease and related dementia; ACEIs, angiotensin-converting enzyme inhibitors; ARBs, angiotensin receptor blockers; CCBs, calcium channel blockers; TDs, Thiazide diuretics; BBs, β-blockers; OR, odds ratio; CI, confidence interval.

^*p <.05,

^**p <.01

^a.Female is the reference group.

^b.Non-Hispanic Whites is the reference group.

^c.Northeast is the reference group.

^d.Less than high school is the reference group.

^e.Low income (FPL<200%) is the reference group.

^f.Private health insurance users is the reference group.

^g.The non-married including divorced, widowed, separated, and never married is the reference group.

^h.Non-smokers including abstainers and never smokers are the reference group.

^i.Reference groups are individuals without the diagnosis of coronary heart disease, diabetes, or depression.

4. Discussion

The current study is one of the first to explore ADRD diagnosis in relation to class of antihypertensives and/or therapy modality (monotherapy versus polytherapy) using a nationally representative sample of people aged 65 and above. The MEPS data enabled the study to explore detailed pharmaceutical information about antihypertensives and the risk of ADRD in the older U.S. population. Our results showed that CCB utilization was higher among respondents with dementia diagnosis than those without dementia diagnosis. Prior clinical studies have shown that CCBs are commonly used for hypertension treatment and provide effective blood pressure control independently (such as nitrendipine [25]) or in combination with inhibitors of the RAS pathway (such as ACEIs and ARBs) [38]. Recent clinical studies also suggest that CCBs have a beneficial effect on preventing ischemic heart disease and are more effective in reducing the risk of stroke than other first-line medicines [25,39]. Further, CCBs (such as amlodipine) are found to play a significant role in reversing disruptions in calcium homeostasis, which could be a common pathophysiological contributor to Alzheimer’s disease and vascular dementia [25,39,40]. As shown in Table 1, the prevalence of coronary heart disease and stroke was higher among respondents with dementia than those without dementia, which could also explain the higher frequency of CCB utilization among that population.

With demographic and lifestyle factors and health status covariates being fully adjusted, we did not find statistically significant association between ADRD risk and any specific class of antihypertensives or therapy modality. This finding was not surprising as it was consistent with some recent findings from prospective cohort studies and meta-analyses that tested similar relationships [21,22,26]. Among hypertensive patients who were only on monotherapy, we found that those treated with ACEIs had a significantly lower risk of AD diagnosis compared to patients treated with CCBs. While we cannot interpret this pattern as ACEIs being superior to CCBs in terms of reducing AD risk, ACEIs use has shown to inhibit the unregulated angiotensin converting enzyme 1/angiotensin II/ angiotensin type 1 receptor (ACE 1/Ang II/AT1R) axis of the RAS in clinical studies, which is responsible for deleterious effects such as increased oxidative stress, neuroinflammation, and a decrease in cerebral blood flow, and these effects could lead to AD [9]. ACEIs use is also shown to reduce AD risk even among non-hypertensive individuals [9, 41,42], therefore ACEIs may afford neuroprotection to a larger population [43]. The fact that ACEIs benefit cognitive health even among the normotensive individual [42] suggests that these drugs may affect other aspects of the renin-angiotensin pathway [44] unrelated to blood pressure regulation. An additional issue to consider is differences in action of peripherally acting (PACEI) versus centrally acting (CACEI) subclasses of ACEIs, as there are reports that they may differentially impact AD risk [45]. However, the current dataset did not include sufficient information to assess these effects, thus investigations of clinical trials that can clarify the molecular mechanisms evoked by ACEIs may provide beneficial information to understand the pathology of AD and finally achieve a new therapy for AD [9].

Several limitations to this study should be noted. First, variables on blood pressure levels, antihypertensive dose, treatment duration, and severity and duration of hypertension were unavailable in the MEPS dataset. These variables could influence the extent to which neurovascular regulatory mechanisms are impaired, and thus, how this impacts cognitive function and risk for ADRD diagnosis [22]. Therefore, this study could not examine the control of blood pressure in different degrees of hypertension affected risk for ADRD diagnosis. Possible ways to estimate this in our study is through the inference that polytherapy (i.e., multiple medication) use is typically recommended to patients with more severe, chronic, and treatment resistant forms of hypertension [33]. Future studies that incorporate measures on blood pressure, hypertension severity and onset of ADRD into a longitudinal design are needed to investigate if controlled hypertension is the only pathway between antihypertensive use and ADRD risk. Moreover, although we selected respondents who had hypertension diagnosis since midlife, the age of ADRD diagnosis was not available in the MEPS dataset. Therefore, study results should be interpreted with the caveat that the temporal relationship between hypertension diagnosis and ADRD diagnosis is not definitively established. Another limitation of this study is that data from the MEPS are self-reported, thus they are subject to recall bias [28,33], which could result in the underestimation or overestimation of antihypertensive utilization, thus influencing this variable’s ability to predict dementia risk. Last, the cross-sectional design of the MEPS prevents the study from capturing longitudinal changes in therapy modality, medication class, and/or compliance to therapy, which are likely to change over time, particularly in patients with more severe, treatment-resistant presentations of hypertension.

Despite these limitations, findings of the study improved our understanding of the relationship between therapy modality and/or class of antihypertensives and ADRD diagnosis in the older U.S. population. Results from this study provide the foundation for future prospective, clinic-based, longitudinal studies that investigate whether therapy modality or specific classes of antihypertensives utilization would affect the change in cognitive function over time, and the pathways of how they affect cognitive function through blood pressure. In particular, the finding on the negative association between ACEIs and the risk of having AD suggests the need to investigate the cellular mechanism of how ACEIs affects brain through alternative pathways such as the ACE 1/Ang II/AT2R; ACE 2/Ang (1− 7)/MasR; Ang IV/ AT4R (IRAP) axis, which is believed to counterbalance the negative effects produce by the ACE 1/Ang II/AT1R axis and exert beneficial effects on memory and cognition [9], and eventually reduces brain inflammation and accumulation of β-amyloid, a marker of neuronal dysfunction. Given the increase in global aging and high incidence of untreated or poorly controlled hypertension [26,46], cardiovascular disease is the leading cause of death among older adults in the United States. It is urgent to develop evidence-based and cost-efficient interventions to manage hypertension and its related health consequences. Findings of the study could inform healthcare providers to consider using ACEIs as a routine drug therapy for older adults with hypertension to optimize their blood pressure management and minimize the risks of developing ADRD. Healthcare institutes, community leaders, and healthcare professionals can take the responsibility of educating older adults about differences in class of antihypertensives and their impact on cognitive function, while facilitating medication selection and utilization.

Further research also needs to identify the cause of utilization of specific therapy and/or class of antihypertensives among older Americans. Gaining a better understanding of resources inherent in hypertension management and beyond would allow us to gauge the strength and weaknesses in their hypertension management to help older adults preserve their cognitive function and well-being. Considering significant healthcare costs involved with dementia care, this study also calls for improvement in public health policy concerning hypertension and cognitive function for the older population in the United States. In particular, federal and state governments need to create welfare to support affordable care for financially disadvantaged hypertensive older adults and encourage and support more health care providers with a specialization in neurology to help prevent or delay the onset of ADRD among older adults.

5. Conclusions

ADRD is a major public health problem because AD and other forms of dementia are among the leading causes of mortality for older adults. Hypertension is one of modifiable risk factors for late-life cognitive impairment and ADRD. Results of the study indicate that older adults who used ACEIs had a significantly lower risk of AD diagnosis compared to those used CCBs for hypertension control. Findings of the study could be used to develop evidence-based drug therapy to optimize blood pressure control while preserving cognitive function and well-being for older adults with hypertension.

Key Points.

• While the risk of having Alzheimer’s disease and related dementias (ADRD) diagnosis does not vary by therapy modality, utilization of angiotensin-converting enzyme inhibitors (ACEIs) might lower the risk of ADRD diagnosis, especially Alzheimer’s disease (AD) in older adults with hypertension.

• Further investigation on the biological mechanisms underlying specific classes of antihypertensives on cognitive health via blood circulation using longitudinal data is needed.

• Given population aging and the epidemics of ADRD, it is urgent to develop evidence-based interventions to optimize blood pressure control and improve public health policy concerning hypertension and cognitive function for the older population.

Availability of data and material

Anonymized data will be shared at the reasonable request of any bona fide investigator.