The Association of Beta-Blocker Use to Cognitive Impairment among Adults with Hypertension or Cardiovascular Diseases in the United States

Background: Some studies have shown that beta-blocker use is associated with better cognitive impairment. However, these studies did not control for pain. The relationship between pain and cognitive impairment has been exhaustively investigated. The association of beta blockers to cognitive impairment in the presence of chronic pain is still unknown. Objective: To examine the independent association of beta-blocker use to cognitive impairment among adults with hypertension or Cardiovascular Diseases (CVDs). Methods: We used a cross-sectional study design. We derived data on 8,279 adults from the 2015 Medical Expenditure Panel Survey (MEPS). Study participants were adults (age > 21 years), with hypertension or CVDs and without intracranial injury, Parkinson, Alzheimer’s disease and Related Dementia. Cognitive impairment was measured based on 1) confusion or memory loss; 2) problems making decisions, or 3) supervision for participant’s safety. Anti-hypertensive medications were categorized into 1) beta-blockers; 2) other anti-hypertensives; and 3) no antihypertensive medication. We used multivariable survey logistic regressions to examine the association between beta-blockers and cognitive impairment after controlling for biological factors, pain, chronic conditions, socioeconomic status, access to healthcare services, behavioral, socio-cultural and external environmental factors. Results: Overall, 24.2%, 41.9%, and 33.9% reported using beta-blockers, other antihypertensives, and no antihypertensive medications, respectively; 18.1% participants reported cognitive impairment. After controlling for pain, beta-blocker use was not significantly associated with cognitive impairment (AOR= 1.22, 95%CI= 1.00-1.49). In fully adjusted models, the AOR for beta-blockers use was 1.05 (95%CI = 0.84-1.31). Conclusion: In this first large cross-sectional study, we found that the use of beta-blockers was not associated with cognitive impairment. Future prospective studies that include pain management and blood pressure control are needed to confirm the findings.


Introduction
Cognitive impairment defined as difficulties in learning new things, remembering, concentrating, and making rational decisions affects nearly 16 million adults in the United States (US) [1]. Cognitive impairment has profound negative effects on patients, caregivers, healthcare providers, payers, and society [1][2][3]. Mild Cognitive Impairment (MCI) is a risk factor for Alzheimer disease and other dementias (ADRD) [4,5] and places excessive humanistic and financial burdens on both adults and their caregivers [6,7].
There are several plausible explanations of why the use of beta-blockers may favorably affect cognition. Beta-blockers inhibit the sympathetic nervous system by binding to betaadrenergic receptors which lowers the hearts' contractile force, dilates arterial vasculature, and lowers blood pressure [16]. As high blood pressure [17,18] and pathological vascular systems [19] have been shown to have harmful effects on cognitive function, it is plausible that beta-blockers may improve blood pressure and indirectly improve cognitive functioning. Gelber [20] and White [21] reported that beta-blocker use was associated with lower risks of cognitive impairment due to the effect of beta blockers on improved brain tissue functions.
Other studies have hypothesized that the use of adrenergic receptor antagonists may affect cognitive function through the pain pathway [15,22]. It is important to note that the association between pain and cognitive impairment is complex and involves changes to brain morphology, electrophysiology, neurotransmitters and receptors, glial cells, cytokines, enzymes, and neurotrophic factors [15]. Several pre-clinical and clinical studies have documented the association of chronic pain to cognitive impairment [14,15,23]. Whitlock and colleagues reported that respondents with persistent pain had 15% higher rates of memory decline compared to participants without pain [24]. In addition, higher levels of pain have been correlated with increased depressive symptoms and reduced neuropsychological test scores [25].
To date, no study has evaluated the association of beta-blockers to cognitive impairment after controlling for chronic pain. Therefore, the primary objective of the current study is to explore the association of beta-blocker medication use to cognitive impairment after controlling for pain among adults with Cardiovascular Diseases (CVDs) or hypertension using a nationally representative sample of adults in the United States. Adults with hypertension or CVDs were chosen because they are more likely to receive beta-blockers for treatment of these conditions [16].

Study design
This study adopted a cross-sectional design.

Data source
We used data from the 2015 Medical Expenditure Panel Survey (MEPS), a household survey of non-institutionalized adults in the US [26]. The survey collects a variety of health-related information including individual socio-demographic characteristics, health conditions, healthcare use and expenditures, sources of payment and health insurance coverage. We combined data from MEPS Household Survey, Prescribed Medicines, and Medical Condition Files to create our analytical dataset.

Sample
Study participants were adults over 21 years with CVDs or hypertension. The cardiovascular conditions included were pulmonary heart disease, acute myocardial infarction, cardiomyopathy, coronary atherosclerosis, conduction disorders, cardiac dysrhythmias, congestive heart failure, cardiac arrest, ventricular fibrillation, and other ill-defined heart diseases. These conditions were measured based on Clinical Classification Code, International Classification of Diseases, Ninth Revision (ICD-9), and self-reported data (especially for hypertension) on MEPS. We excluded adults with missing information on cognitive impairment, (n = 158) individuals with intracranial injury, (n = 59) Parkinson's disease or ADRD (n = 199) and individuals who died during the calendar year. The final sample size consisted of 8,279 adults with CVDs or hypertension.

Measures
Dependent variable: Cognitive Impairment-We identified the presence or absence of cognitive impairment from the household (full year consolidated) file based on the questions asking if the person (1) experienced confusion or memory loss, (2) had problems making decisions, or (3) required supervision for their own safety. We used codes COGLIM31, COGLIM53 -representing cognitive limitations (round 3/1 and 5/3), and Code DFCOG42 -representing serious cognitive difficulties (round 4/2). We categorized this into a binary variable Yes (cognitive limitation) and No (no cognitive limitation)

Key Independent Variables
Use of Beta-Blockers-We identified Beta-blockers and other antihypertensive medications using the prescribed medicines file. This file contains information on drug names, national drug code, drugs-supplied and other medication related information. Prescription drugs were classified into various therapeutic classes based on the proprietary "Multum Lexicon" files. Multum Lexicon provides a 3-level nested category system for each drug based on the ingredients [27]. Therapeutic classification code 47 represented betablockers and codes 42, 43, 48, 49, 53, 55, 56 represented other antihypertensives (ACEinhibitor, alpha-adrenergic blockers, calcium channel blockers, diuretics, vasodilators, angiotensin receptor blockers, antihypertensive combinations, and angiotensin II inhibitors). This variable was categorized into three groups: (1.) Beta-Blockers (2.) other antihypertensives (3.) No antihypertensive medication.
Pain-A self-administered questionnaire was used to collect information on pain and other domains of health-related quality of life. Pain was measured using one item from the Short Form (SF)-12, asking how much pain "interfered with normal work (including both work outside the home and housework)" during the past 4 weeks prior to the interview. Responses were recorded using a 5-item Likert scale: 1) 'Not at all'; 2) 'A little bit'; 3) 'Moderate'; 4) 'Severe'; and 5) 'Extreme'. We used this pain interference with activities as a proxy for chronic pain [28,29].

Other Independent Variables
Other variables that may influence cognitive impairment were selected based on the determinants of health model initially proposed by Park in the Textbook of Preventive and Social Medicine (2015) and published studies that examined factors associated with cognitive impairment. These included: biological characteristics like sex (female vs. male), age, categorized into 4 groups (21-44, 45-54, 55-64, 65 and older) and race (White, African American, Latino, and other), socio-economic factors like education (Less than high school, high school, and above high school) and poverty status (poor, near poor, middle income, and high income measured in terms of percentages of federal poverty line), socio-cultural factors like marital status (married, widowed, divorced/separated, and not married)], and behavioral factors like body mass index (underweight, normal, overweight, obese), physical activity (exercising five times/week and others), and smoking status (yes/ no).We also included variables of access to healthcare services, measured by health insurance (public, private, and uninsured) and prescription coverage (yes/no), medical conditions including the presence of asthma, chronic obstructive pulmonary disease (COPD), cancer, diabetes mellitus, mental health conditions (depression and anxiety), hypercholesterolemia, and arthritis (yes/no), and region of residence (Northeast, Midwest, South and West).

Statistical analysis
Statistically significant differences in cognitive impairment by type of antihypertensive medications were determined by Rao-Scott chi-square tests. Multivariable logistic regression models were used to assess the relationship between type of antihypertensive medications and cognitive impairment after controlling for sex, age, race, pain, education, poverty status, marital status, body mass index, physical activity, smoking status, health insurance, prescription coverage, asthma, Chronic Obstructive Pulmonary Disease (COPD), cancer, diabetes mellitus, mental health conditions (depression and anxiety), hypercholesterolemia, arthritis, region of residence. Since MEPS uses as complex survey design with clustering, stratification, and weights, we conducted all the analyses using the survey procedure in SAS version 9.4. Because pain information was assessed from self-administered questionnaires, we applied appropriate weights for the responses derived from those self-administered questionnaires.

Results
Our study participants consisted of 8,279 adults (Wt. N= 94,815,011) with hypertension or cardiovascular disease. Majority of the respondents were female (50.8%), white (67.6%) and married (58.8%). Thirty-seven percent of the participants were 65 years or older and 54.3% had greater than high school education. About 52.9% of the sample had prescription medicine coverage and only 5.9% were uninsured (Table 1).
Overall, 24.2% of the participants reported using beta-blockers and 41.9% used other antihypertensives (Appendix 1). A higher percentage of elderly participants (65 years or older) used beta-blockers compared to adults aged 21-44 years (36.3% vs 9.2%). A similar trend was also observed among individuals with chronic conditions. Compared to individuals without chronic conditions, respondents with chronic conditions such as COPD (31.9% vs 23.3%), cancer (32.8% vs 22.9%), diabetes mellitus (32.7% vs 21.4%), hypercholesterolemia (30.1% vs 16.0%), and arthritis (28.8% vs 17.6%) were more likely to use beta blockers. We found statistically significant differences in reported pain and cognitive impairment between groups of individuals who used beta-blockers, other antihypertensive medications, compared to those who did not use antihypertensive medications.
Overall, 18.1% of the participants reported some cognitive impairment. Individuals who reported cognitive impairment differed significantly from those who did not with respect to all the independent variables included in the study except race/ethnicity, cancer, and region ( Table 2). For example, compared to adults with no antihypertensive medications, a higher percentage of those with beta-blocker use had cognitive impairment (13.2% vs 20.2%). The Unadjusted Odds Ratios (UOR) indicated that adults who used beta-blockers had significantly higher odds of cognitive impairment (OR = 1.67; 95%CI= 1.39, 2.01, p <.0001) compared to individuals who did not use any anti-hypertensive medication. We did not find significantly higher odds of cognitive impairment among individuals who used other antihypertensive medications (OR=1.17; 95%CI= 0.97, 1.41, p = 0.0920). Pain was significantly associated with cognitive impairment; adults who reported extreme pain (UOR = 13.52; 95%CI = 9.99, 18.30, p= <0.001), severe pain (UOR = 8.70; 95%CI = 6.85, 11.06, p= <0.001), and moderate pain (UOR = 4.47; 95%CI = 3.49, 5.73, p= <0.001) all had higher odds of cognitive impairment compared to those with no pain.
After controlling for pain in our model, beta-blocker use was not significantly associated with cognitive impairment (AOR= 1.22, 95%CI= 1.00-1.49). Adults who reported extreme pain were 13.2 times more likely to report impaired cognition (95%CI= 9.71, 17.92, p <.001) compared to those with no pain. In the fully adjusted model, the use of beta-blockers was not associated with cognitive impairment. (OR=1.05, 95% CI = 0.84-1.31, p=0.6901) ( Table 3).

Discussion
This paper examined the association of beta-blockers with cognitive impairment among adults with hypertension or CVDs. In this study, 18.1% of all participants and 19.0% of the elderly reported cognitive impairment. This finding is broadly consistent with previous studies in adults over 65 years that report rates of cognitive impairment between 5% and 36.7% [30].
After controlling for pain, the association between beta blocker use and cognitive impairment was not statistically significant. In the fully adjusted model, Individuals with extreme pain were 5.4 times more likely to have cognitive impairment compared to those without pain. The prevalence of moderate and extreme pain was comparable within the antihypertensive medication categories (Appendix 1) Taken together, these findings suggest that pain rather than antihypertensive medication use may account for the observed association to cognitive impairment.
Our study findings are not consistent with previous studies by Gelber, et al. and White, et al. which reported that individuals who used beta blockers had lower risks of developing cognitive impairment compared to those who did not use antihypertensive medications [20,21,31]. Differences in study design, study population, cognitive impairment scales, and controlling for pain may partly explain the inconsistent findings. Gelber, et al. used a prospective cohort of Japanese American men living in Hawaii and measured cognitive impairment using a combination of the Hasegawa Dementia Screening Scale, the Folstein Mini-Mental State Examination, and the Modified Mini-Mental State Examination but did not control for pain. Several reports postulated that the mechanism behind this protective effect of beta-blocker on cognition involved the action of beta-blockers on vasculature to improve and maintain the ability of the heart and blood vessels to adjust to systemic or local physiological disturbances in the cardiovascular system [16,17]. Conversely, Gliebus and Lippa posited that beta-blockers could potentially block the activation of norepinephrine, the neurotransmitter which plays an essential role in the retrieval of memories from the hippocampus [32]. On the other hand, our main study finding that there is no significant association between beta blocker use and cognitive impairment is consistent with an analysis of published clinical trial data from 19 trials on beta-blockers [33] and a randomized, double-blind, controlled crossover trial [34].
Several factors that may account for the different findings in this study include extent of blood pressure control, which has been shown to be correlated with medication adherence [35,36], and other residual factors which were not measured in this study [37]. The cognitive impairment scale used in MEPS was relatively crude and we expect that more specific measures of cognition (such as attention, learning and memory, speed of processing, psychomotor ability and executive function) could improve the robustness the study design.
Our study findings are consistent with previous reports on the association of pain to cognitive impairment which has been exhaustively investigated [14,15]. Almost all pain categories were associated with cognitive impairment with adjusted odds ratios ranging from 1.28 for mild pain, 2.51 for moderate pain, 3.68 for severe pain, and 5.41 for extreme pain. This is consistent with a review of pre-clinical and clinical studies by Moriarty and colleagues [15]. \Our study has many strengths such as the use of a representative sample of adults in the US, availability of prescription drug information, and adjustment for a comprehensive list of variables, especially pain, which could confound the relationship between antihypertensive medications and cognitive impairment. However, our study has some limitations including the use of cross-sectional data, which precludes the assumption of any causal relationship. We did not adjust for blood pressure control due to the fact that MEPS does not include clinical parameters. Other limitations are that this study utilized self-reported overall cognitive impairment as a crude measure of cognitive function and did not include data on duration of beta-blockers use. Future studies that employ a prospective study design, with robust measures of cognitive impairment and duration of beta-blocker use and medication adherence measures are needed to confirm our study findings.  Chronic Pain Manag. Author manuscript; available in PMC 2020 July 13.