Z. Sternberg1, R. Podolsky1, J. Yu2, M. Tian2, D. Hojnacki1, B. Schaller3
1. Jacobs School of Medicine and Biomedical Sciences, Buffalo Medical Center, Buffalo, NY, USA; 2. Department of Biostatistics, University at Buffalo, Buffalo, NY, USA; 3. Institute of Physiopathology, Department of Pathology, University of Buenos Aires, Argentina
Corresponding Author: Zohara Sternberg, PhD, Clinical Associate Professor of Neurology, Buffalo Medical Center, Buffalo, NY, 14203, USA, Tel: 716-8597540, Fax: 716-8592430, 859-7573, Email: email@example.com
J Prev Alz Dis 2022;
Published online September 6, 2022, http://dx.doi.org/10.14283/jpad.2022.73
Background: Arterial hypertension is among factors with the potential for increasing the risk of cognitive impairment in elderly subjects. However, studies investigating the effects of antihypertensives on cognitive function have reported mixed results.
Methods: We have used the National Alzheimer’s Coordinating Center (NACC) Uniform Data Set (UDS) to investigate the effect of each class of antihypertensives, both as single and combined, in reducing the rate of conversion from normal to mild cognitive impairment (MCI).
Results: The use of antihypertensive drugs was associated with 21% (Hazard ratio: 0.79, p<01001) delay in the rate of conversion to MCI. This effect was modulated by age, gender, and genotypic APOE4 allele. Among different antihypertensive subclasses, calcium channel blockers (CCBs) (24%, HR: 0.76, P=0.004), diuretics (21%, HR: 0.79, P=0.006), and α1-adrenergic blockers (α1-ABs) (23%, HR: 0.77, P=0.034) significantly delayed the rate of MCI conversion. A significant effect was observed with the selective L-type voltage-gated CCBs, dihydropyridines, amlodipine (47%, HR=0.53, P<0.001) and nifedipine (49%, HR=0.51, P=0.012), whereas non-DHPs showed insignificant effect. Loop diuretics, potassium sparing diuretics, and thiazides all significantly reduced the rate of MCI conversion. Combination of α1-AB and diuretics led to synergistic effects; combination of vasodilators plus β-blockers (βBs), and α1-AB plus βBs led to additive effect in delaying the rate of MCI conversion, when compared to a single drug.
Conclusion: Our results could have implications for the more effective treatment of hypertensive elderly adults who are likely to be at high risk of cognitive decline and dementia. The choice of combination of antihypertensive therapy should also consider the combination which would lead to an optimum benefit on cognitive function.
Key words: Alzheimer’s disease, cognition, cohort Ssudy, dementia, genetic risk factors, hypertension, vascular disease.
Abbreviation: α1-AB: α1-adrenergic blockers; AD: Alzheimer disease; ADRC: Alzheimer`s Disease Research Centers; ACEI: Angiotensin-Converting Enzyme Inhibitors; ARB: Angiotensin Receptor Blockers; βB: β-blockers; Aβ: β-amyloid; CCB: Calcium Channel Blockers; CDR®: Dementia Staging Instrument; CDR-SOB: Clinical Dementia Rating Scale Sum of Boxes; MCI: Mild Cognitive Impairment; MMSE: Mini-Mental State Examination; MoCA: Montreal Cognitive Assessment; NACC: National Alzheimer’s Coordinating Center; NIA: National Institute on Aging; NFT: Neurofibrillary Tangles; UDS: Uniform Data Set.
Alzheimer disease (AD) is one of the most common age-related neurodegenerative disorders accounting for up to 80% of all dementia diagnoses, involving a progressive decline in cognitive domains, including memory, language, and executive function (1). AD is cellularly characterized by the abnormal aggregation and accumulation of β-amyloid (Aβ) in the form of extracellular plaques and aggregation of hyper-phosphorilated tau protein in the form of intracellular neurofibrillary tangles (NFT) (2).
The changes in memory and cognitive functions are also discernible in preclinical AD and in patients with mild cognitive impairment (MCI) (3, 4). However, the decline in cognitive function is often subtle, and not readily recognized in MCI patients due to an overlap between what could be due physiological aging and the underlying AD pathophysiology (5, 6). The combination of measures, including clinical presentation, neuropsychological testing, cerebrospinal fluid (CSF) biomarkers and neuroimaging modalities are used to enhance the complex MCI diagnostic accuracy (7).
MCI is heterogeneous in its clinical presentation; most present with the amnestic (aMCI) subtype which involves deficits in episodic memory as single or most prominent characteristic (8). Such patients are more likely to develop AD at follow up (9). The deficit in episodic memory is also accompanied by cortical Aβ deposition as measured by the PET scan and Pittsburgh Compound B (PiB)PET (10). In the non-amnestic (na) MCI patients, the memory remains intact, whereas other cognitive abilities such as executive function, attention, language, could be affected in a single or multiple domains (8).
Based on a number of international studies involving several thousand subjects, the prevalence of MCI in persons over the age of 60 years is reported to be between 12% to 18% range, with annual rate of 8% to 15% progression to dementia (11). Factors such as age, male gender, and the presence of APOE4 allele are all associated with a higher MCI prevalence (12, 13).
Vascular clinical risk factors such as hypertension (14), diabetes (15), hypercholesterolemia (16), as well as co-morbid conditions such as depression (17) are associated with higher risk for MCI. Nevertheless, studies investigating the use of antihypertensive drugs in aged populations are conflicting. Mossello et al. (18) reported steep decline in cognitive function in older MCI patients on antihypertensive agents with low BP. In contrast, Hanon et al. (19) reported higher cognitive function, indicated by Mini-Mental State Examination (MMSE) scores, in elderly subjects treated with antihypertensive drugs compared to untreated ones, independent of BP level.
These clinical results were corroborated in MRI studies of elderly population showing an association between larger hippocampal volume and brain parenchymal fraction, and a better performance on tests of episodic and verbal memory, language, and executive function in MCI patients on antihypertensive agents compared to those who were not on such regimen. These effects were more apparent in patients on angiotensin receptor blockers (ARBs) compared with those on angiotensin-converting enzyme inhibitors (ACEIs) (20). Nevertheless, a more recent systematic review of 16 studies reported no benefit of antihypertensive drugs on cognitive function in MCI patients (21), although this study has significant methodological problems (mix of different endpoints; lack of geographical representation). However, these findings together emerge to accumulating evidence for a working rationale for this investigation that pharmacological management of hypertension may reduce the risk of cognitive decline and/or dementia. These considerations let to the design of this study to determine the effect of each antihypertensive class of drug, as single and in combination, on the rate of progression from normal to MCI, and possible interaction between the use of antihypertensive drugs and genetic factors such as age, gender, and presence of the APOE4 allele. Although the relationship between the use of antihypertensive drugs and the incidence of dementia has been already studied by many prospective cohort studies, there is title evidence from an epidemiological perspective, regarding the relationship between use of these drugs and the incidence of mild cognitive impairment. This question has significant clinical implications as it is highly linked to preventive interventions. Therefore, the present study embodies an important and novel research topic in a large study sample.
The study was approved by the Internal Review Board of the State University of New York at Buffalo, Buffalo, NY, U.S.A., and all participants information was de-identified in the data set received from the National Alzheimer’s Coordinating Center (NACC). Written informed consent is obtained from all participants and co-participants.
All data were obtained only from the NACC. From September 2005 to the specific data freeze of March 2020 (containing data on to Feb 2020) for Alzheimer`s Disease Research Centers (ADRCs) across the USA have been contributing data to the UDS, using a prospective, standardized, and longitudinal clinical evaluation of the subjects in the National Institute on Aging’s (NIA`s) ADRC program. In each subject’s approximate annual UDS visit, data collection forms are completed by the clinician, covering topics from subject demographics to neurological examination findings, neuropsychological test results and psychiatric symptoms or other diagnosis on individuals with normal cognition, mild cognitive impairment, and dementia. For each ADRC visit, a multidisciplinary team or a single clinician determines a clinical diagnosis based on established guidelines.
The UDS reflects the total enrollment of the NIA’s ADRCs since 2005 up to the data freeze of March 2020 and includes participants with cognitive status ranging from normal cognition to MCI, and demented. Each Center enrolls its participants according to its own protocol — e.g., clinician referral, self-referral by participants or family members, active recruitment in the community organizations, etc. Most centers also enroll volunteers with normal cognition, and these tend to be highly educated. Overall, participants are enrolled using different methods and for different research purposes at the NIA`s ADRCs.
In this large prospective, standardized clinical case series rather than longitudinal study in a strict sense, the analysis is based on complete covariates cases under the assumption that the missing patterns in covariates are random and do not depend on observed or unobserved observations. Loss of follow-up is considered to be right-censored data and incorporated in the data analysis. In addition, the mortality was no outcome variable as it doesn’t have any relevance in the objectives of the current investigation.
Inclusion criteria were if there was more than one UDS visit.
Exclusion criteria were (i) mental disease signs at the initial visit, based on use of antipsychotic drugs, due to a relationship between psychiatric diseases and cognitive deficit (22) (ii) no records of medicines to be analysed, (iii) missing values in selected important covariates (age, gender, Parkinson`s disease (PD), history of traumatic brain injury (TBI), alcohol abuse, active depression in the last two years, heart attack/cardiac arrest, education history, smoked cigarettes in the last 30 days), (iv) current use of diabetes medication or lipid-lowering medication, and (v) missingness of APOE4 gene variables.
For this study, the UDS data, from the whole ADRCs data pool are collected using different standardized evaluation of participants enrolled in ADRC clinics. Data are recorded directly on UDS forms (hard copy or electronic) during the evaluation process. Information is collected during in-person office visits, home visits, and telephone calls. In addition, Milestone Forms are used to document participant death and drop-out. The UDS is longitudinal, and its protocol requires an approximate annual follow-up for as long as the participant can be involved. Late-stage participants forced to drop out due to health may continue to be followed strictly for autopsy purposes. Data are collected by trained clinicians and clinic personnel from participants and their co-participants (usually a close friend or family member). Depending on a given ADRC’s protocol, diagnosis is made by either a consensus team or clinicians.
Although the focus of the ADRCs is AD, the Centers also collect data on a variety of associated disorders, such as vascular dementia, Lewy body dementia, and frontotemporal lobar degeneration. Furthermore, the use of medications — e.g., antihypertensives, hypolipidemics, antidiabetics, antidepressants, and antipsychotics is documented at each visit and during follow up. However, the completion of the form assessing participant’s medications is optional, and therefore, the record of adherence to the treatment regimen may be incomplete. In addition, we analysed the relationship of systolic and diastolic BP on the rate of MCI conversion in participants who used antihypertensive agents compared to those who didn’t.
To ensure patient privacy, the stored and transmitted data are de-identified at the participants, and organization level. Structured data recorded in the electronic health records are assimilated into the database after mapping the data to standard and controlled clinical terms. A rigorous data quality assessment is done to exclude records that do not meet quality standards and basic formatting requirements for adequate data representation.
Further diagnostic evaluation
For some UDS subjects, CSF values are available for A-β, P-tau, and T-tau. In the here presented data set, genotypic data (i.e., APOE status) is available at NACC for 75 percent of UDS participants, as well as genetic information on whether the participant or their family has any known AD or FTLD mutations.
Psychiatric symptom data included (i) a history of depression (coded as consulting a clinician, (ii) being prescribed medication or receiving a diagnosis related to depressed mood), (iii) a depressive symptoms scale (Geriatric Depression Scale-Short Form), (iv) history of pseudobulbar affect, and (v) history of substance use disorder and the Neuropsychiatric Inventory Questionnaire (NPI-Q) assessing presence/absence of 12 neurobehavioral symptoms (23).
For this study, we have used a cohort of participants from NACC-UDS, limiting it to patients who enrolled having normal cognitive function as assessed by Dementia Staging Instrument CDR®=0 and who converted to MCI during follow up as assessed by CDR®= 0.5 for two consecutive visits and Clinical Dementia Rating Scale Sum of Boxes (CDR-SOB) of 0.5-2.5. The CDR® was used in combination with CDR-SOB to improve the identification of MCI patients, since CDR® has a relatively narrow range for differentiating normal from MCI subjects (24). The proportions of patients converted from normal to MCI was assessed for 15 years follow up. In addition, we have provided the values for the group differences in mini mental state examination (MMSE) scores and Montreal cognitive assessment (MoCA) scores.
We defined the following subgroup analyses to investigate heterogeneous results for the primary outcome. Subgroups based on different drug classes (“CCB”, “diuretics”) or combination of different drug classes (“combination drug”).
The SAS (Version 9.4) and R 4.0.2 software were used for all statistical analyses. We created the time to MCI progression data from participant’s repeated measure data, where MCI progression is defined under neurological examination. In the case of missing baseline demographics data, the first non-missing value from the repeated measure data was used. We differentiated between antihypertensive users and non-users with further stratifications based on age, gender and the existence of at least one copy of the genotypic APOE4 allele. For univariate data analysis, we used χ2 tests for categorical variables and independent sample t-test for continuous variable. For the multivariate categorical data analysis, we fitted a logistic model where MCI status was the binary outcome.
For the analysis of time to MCI progression, the date of initial visit was the starting point and progression to MCI was the event studied. Participants ending the follow-up period without progression to MCI produced censored observations. Unadjusted survival analyses were carried out using Kaplan-Meier (KM) curve. We performed the adjusted MCI conversion rate analysis using Cox proportional hazard regression incorporating covariates of interest and interactions. Several important covariates include age, status of PD, history of TBI, vitamin B12 deficiency, alcohol abuse, smoking, education, cardiovascular disease, active depression and usage of any antipsychotic drugs during the follow up. We adopted Cox models for any type of an antihypertensive drug use for investigating the effect of overall anti-hypertensive drugs. We also adapted Cox models for individual anti-hypertensive class of drugs, to examine whether a specific class of antihypertensives may show more efficacy than other classes. The hazard ratio is estimated based on Cox models. Also, due to many Also, due to many missing values in the genotypic APOE4 variable, we excluded this variable from the initial model and then added it to check whether there was any impact of this variable on the drug use effects.
To create the combination drug, we chose analyzing combination of antihypertensive drugs, some often used in clinical practice. We checked subject each visit, assign value 1 if they used both drugs in the same visit. We compared each combination drug with single drug for statistical significance. For each combination drug (drugA+drug B), we compared the group with the related single drug being often CCBs, β-blockers (βBs), angiotensin converting enzyme inhibitors (ACEIs) or angiotensin II receptor blockers (ARBs) as main drugs, and diuretics, vasodilators, or α1-adrenergic blockers (α1-AB) often prescribed as addition to the main antihypertensive regimen. However, in some cases, the combinations of CCBs+ARBs and α1-AB+diuretics were also available for analysis. The data was corrected for all-interested covariates as described above.
From 42`661 patients with 150`744 records (see Figure 1). We first excluded patients who had only one visit or patients who had a mental disease sign at the initial visit, including 13`859 patients with 75`686 records in the following steps. Secondly, we excluded patients who had no record of medications we wanted to analyse. This left us with a total number of 13`652 patients with 74`437 records. Then, we transformed the repeated measures dataset into a survival dataset, where one patient had only one record.
Patients were excluded if they had missing values for APOE4, and missing values for the selected important covariates (age, gender, PD, TBI, alcohol abuse, past documentation of a heart attack/cardiac arrest, education history, and smoking habit in the past 30 days, and active depression in the last two years), remaining with 12`725 patients for the next step. Subsequently, we excluded patients with CDR®>0.5, CDRSOB>2.5, and those who progressed directly from normal to CDR®>0.5, CDRSOB>2.5. In addition, patients on antidiabetic drugs (25) and those on statins (26) were excluded from the list of selected patients due to the modulating effects of these drugs on cognitive function. After these exclusions, we had a final sample size of 5`889 patients for analysis. Loss of follow-up was considered to be right-censored data and incorporated in the data analysis.
Conversion to MCI
Table 1A describes patient’s demographics in the included 5889 patients, 30% of whom were male. The table includes variables related to genetic risk factors, including age, gender, and the presence of genotypic APOE4 allele, all of which could hasten the conversion of normal to MCI state. Among included patients, 936 (15.9%) converted to MCI during 3.52±2.70 years follow up. The percentages of patients who converted from normal to MCI were significantly higher for males compared to females (18.8% vs. 14.7%, P <0.001). In addition, the rate of conversion from normal to MCI was accelerated for males compared to females (3.10±2.41 years vs. 3.74±2.81 years, P=0.001). Patients who converted to MCI were significantly older than those who did not (74.9±10.7 years vs. 68.4±12.1, P<0.001), with no differences between males and females. The division of the participants to two age subgroups (< 70 years and ≥ 70 years) was arbitrary to allow optimal sample sizes for adequate power analysis
A higher percentage of patients who converted to MCI had at least one copy of the genotypic APOE4 allele, compared to patients with no such allele (29.2% vs. 22.3%, HR=1.442, P=0.015), suggesting a positive association between genotypic APOE4 and conversion to MCI. Among males who converted to MCI, 30.3% had at least one copy of APOE4, compared to 21.6% who did not (P=0.031). A similar pattern was observed in females, though group differences did not reach statistical significance (28.6% vs. 22.5%, P=0.154) (Table 1A). Furthermore, the interaction between genotypic APOE4 and gender using Cox model was not statistically significant (p=0.326). The values of systolic, diastolic, and mean arterial BP did not influence the rate of conversion and the proportion of participants who converted to MCI (P>0.05) (data not shown).
Abbreviation: MCI: mild cognitive impairment. P≤0.05 is statistically significant.
Subgroup differences in covariates
Table 1B describes the differences in baseline covariates (p value, hazard ratio, confidence interval) between the subgroup that converted to MCI and the subgroup which did not. This table also presents subgroup differences in systolic, diastolic, and mean arterial BP. We observed higher percentages of participant who converted to MCI reporting depression in the past two years compared to those who did not convert (20.6% vs. 16%, P=0.001); However, the percentages of participants who used antipsychotic drugs were similar between the two groups. In addition, the percentages of participants with lower level of education were higher among the subgroup that converted to MCI compared to the group who didn’t (P=0.003).
Arterial BP has a non-normal distribution. Therefore, instead of the usual mean and standard deviation, we used the Kruskal test, providing the median and interquartile range (IQR). Participants who converted to MCI had higher systolic BP (P=0.001), but lower diastolic BP (P=0.036) compared to participants who did not convert to MCI. The group significance stems from a large sample size, and a relatively small standard deviation.; However, the mean arterial BP did not differ among the two subgroups (P= 0.668).
BP: blood pressure; CI: confidence interval; HR: hazard ratio; LB: lower border; MCI: mild cognitive impairment; ND: not determined; PD: Parkinson’s disease; TBI: traumatic brain injury; UP: upper border
Table 1C presents the neuropsychological tests. CDR® scores higher than 0.5 and CDRSOB scores of 0.5-2.5 were used to differentiate between normal and MCI converted subjects. The CDR® (0.03±0.13 vs. 0.05±0.15, P=0.009) and CDRSOB (0.07±0.25 vs. 0.14±0.32, P<0.001) were significantly higher for Participants who converted to MCI compared to those who did not. MoCA test scores (26.3±2.9 vs. 22.2±5.1, P<0.001) and modified MoCA test scores (26.4±2.8 vs. 22.4±5.1, P<0.001) significantly differed between normal subjects and MCI converted group. The MMSE scores, though only slightly different between normal and MCI subjects, these differences were statistically significant (28.9±1.4 vs. 28.56±1.6, P<0.001), suggesting that similar to CDR® and CDRSOB, MoCA, modified MoCA and MMSE, distinguished between normal and MCI subjects.
The results present mean±SD. P≤0.05 is statistically significant; Abbreviations: CDR: clinical dementia rating; MMSE: mini mental state examination; MoCA: Montreal cognitive assessment; SOB: Sum of boxes.
Table 2 presents the percentages of subjects who were on antihypertensive drugs, in both normal and MCI-converted group. A significantly higher percentages of MCI-converted Participants were on antihypertensive drugs compared to normal subjects (55.8% vs. 46.7%, P<0.001). Similar patterns were observed when stratifying Participants to females (54.9% vs. 44.7%, P<0.001) and males (57.4% vs. 51.8%, P=0.079). We did not observed differences between the conversion rate of amnestic and non-amnestic participants, in whom the underlying pathology differs (results not shown).
P≤0.05 is statistically significant; Abbreviation: MCI: mild cognitive impairment
Antihypertensive Therapy and MCI Conversion
We subsequently analyzed the effect of antihypertensive drugs on the proportion of Participants who converted to MCI as a function of the three genetic risk factors; age, gender, and the presence of genotypic APOE4 (Fig 2 A-C), as well as the possible interaction between these genetic risk factors and antihypertensive drugs.
Figure 2A shows Kaplan-Meier (KM) conversion rate estimate for the effect of antihypertensive drugs on the proportion of subjects who did not convert from normal to MCI stratified by two age categories (≤ 70 years and >70 years). The cut-off point is chosen by maximizing the hazard ratio between the two age groups. Participants above >70 years seemed to benefit from antihypertensive drugs’ use, as these Participants showed 18% (HR=0.81, P=0.012) less conversion rate to MCI compared to participants >70 years who were not on these drugs during the 15 years study follow up period. However, in the ≤70 years age subcategory, Participants on antihypertensive drugs showed 22% (HR=1.22, P=0.09) higher MCI conversion rate over the same follow up period compared to Participants of the same age category who were not on antihypertensive drugs.
Since age was a strong driver for MCI conversion, we measured the association between antihypertensive drug use and gender in participants stratified by two age subgroups (≤ 70 years and >70 years). KM 2B curve shows the effect of antihypertensive drugs on the proportion of participants who did not convert from normal to MCI, as a function of gender in the two age subgroups. The antihypertensive drugs’ effect was most effective among male aged >70 who showed 38% (HR=0.61, P<0.001) reduction in conversion rate to MCI, whereas the effect of the drug on the proportion of females >70 years was moderate and not statistically significant (HR=0.90, P=0.324).
KM 2C curve shows the effect of antihypertensive drugs on the proportion of participants who did not convert from normal to MCI as a function of one or more genotypic APOE4 allele in the two age- subgroups. Among the group of age >70 without genotypic APOE4 allele, the use of antihypertensive drugs reduced MCI conversion rate by 16% (HR=0.84, P=0.091) compared to the group without antihypertensive drug use; however, among the group of age >70 with genotypic APOE4 allele, the antihypertensive drug had less of a positive effect (HR=0.85, P=0.29). No significant difference in MCI conversion rate was observed in participants age <=70 without genotypic APOE4 whether on antihypertensive drugs or not (HR=1.05, P=0.74). Among the group of age <=70 with genotypic APOE4, the antihypertensive drugs increased MCI conversion rate by 53% (HR=1.51, P=0.042) compared to participants without antihypertensive drugs.
Furthermore, we have analyzed the overall effect of anti-diabetic drugs and demonstrated non significance on the rate of MCI conversion. Additionally, the overall effect of cholesterol-lowering drugs was significant, but when we included these treatments into the analyses, the profiles of significance of the antihypertensive drugs were maintained.
Hazard Ratios: 2A: Age>70: 0.816; Age<=70: 1.22; 2B: Age>70&Male: 0.61; Age>70&Female: 0.90; Age<=70&Male: 1.39; Age<=70&Female: 1.09; 2C: Age>70& without APOE4: 0.84; Age>70& with APOE4: 0.85; Age<=70& without APOE4: 1.05; Age<=70& with APOE4: 1.51
Table 3 presents the percent delay (and the associated hazard ratio and p-value) in the rate of normal to MCI conversion in relation to the use of antihypertensive drugs in general, and each specific class of drug, when compared to the absence of documentation of such drug regimen. The analysis was corrected for covariates as described above. The use of antihypertensive drugs in general significantly reduced the rate of MCI conversion by 21% (Hazard ratio: 0.79, p<0.001). It is noteworthy that the significant effect was observed only after correction for age.
Different class of drugs, including CCBs (24%, HR: 0.76, P=0.004), diuretics (21%, HR: 0.79, P=0.006), and α1-ABs (23%, HR: 0.77, P=0.034) significantly delayed the rate of MCI conversion. In addition, vasodilators (30%, HR: 0.70, P=0.193), βBs (14%, HR: 0.86, P=0.071) and ARBs (11%, HR: 0.89, P=0.283), also reduced the rate of MCI conversion, though statistically insignificant. ACEI had no effect on the rate of MCI conversion (1%, HR: 0.99, P= 0.95) (Table 3A-section 1).
We subsequently analyzed the two CCBs subgroups; the dihydropyridines (DHPs), which are highly selective for vascular smooth muscle L-type voltage-gated calcium channels (L-VGCC), and the non-DHPs. Table 3A, section 2 shows the percent delay (and the associated hazard ratio, P-value, and the sample size) in the rate of MCI conversion for DHPs, non-DHPs and for each specific DHP (as long as the sample size was adequate for statistical analysis). DHPs delayed the MCI conversion by 43%, (HR=0.57, P<0.001), whereas the delay in MCI conversion by non-DHPs was 20% (HR=0.82, P=0.221). Among DHPs, amlodipine (46%, HR=0.54, P<0.001) and nifedipine (44%, HR=0.56, P=0.029) were most effective in delaying the rate of MCI conversion, whereas the effect of felodipine (25%, HR=0.75, P=0.40), isradipine (28%, HR=0.72, P=0.65) and nisoldipine (18%, HR=0.82, P=0.708) remained statistically insignificant, in part, due to a relatively small sample size. There were no participants in the database treated with other DHPs such nitrendipine, nivaldipine or nicardipine.
Among diuretics, the loop diuretics (59%, HR=0.41, P<0.001) showed the most percent reduction in the rate of MCI conversion, although other diuretics including the potassium sparing diuretics (51%, HR=0.49, P=0.049) and thiazides (46%, HR=0.54, P<0.001) also showed significant reducing effect in the rate of MCI conversion. However, carbonic anhydrase inhibitors showed no effect on the rate of MCI conversion (Table 3A, section 3).
To demonstrate additional beneficial effects, we compared each combination of antihypertensive drug with each single antihypertensive drug for statistical significance on the rate of MCI conversion (Table 3B). This analysis led to mixed results since not all combinations had similar effect on the rate of MCI conversion. We assumed the drug combination to have additive effect if the percent delay in MCI conversion approached the sum of the delay observed by each drug. However, if the combination led to delay beyond what was expected from sum of the combination drug, then we assumed the combination having synergistic effect in delaying the conversion from normal to MCI. It is noteworthy that our study design does not have enough information to conclusively assess the mechanistic forms of interactions among antihypertensive drugs, whether additive or synergistic, as well as many factors including genetic or environmental, that could modulate such interactions.
The addition of α1-ABs to diuretics led to 50% reduction in the rate of MCI conversion (HR=0.50, P=0.01), whereas the addition of α1-ABs to βBs had an effect beyond the level of delay expected from the two drugs (35%, HR=0.65, P=0.074), suggesting a synergistic effect; Addition of vasodilators to βBs (36%, HR=0.64, P=0.287) led to additive effect in the rate of MCI conversion, when compared to a single drug. However, the addition of diuretics to either βBs, ARBs or ACEIs did not have a significant effect on the rate of MCI conversion compared to a single drug. Similar results were observed when CCBs were combined with ACEIs or ARBs.
Abbreviations: AB: adrenergic blocker; ACEI: angiotensin converting enzyme; ARB: angiotensin II receptor blocker; βB: beta blocker; CCB: calcium channel blocker; CAI: carbonic anhydrase inhibitors; DHP: dihydropyridine; MCI: mild cognitive impairment.
Abbreviations: as Table 3A
We have conducted exhaustive analysis of NACC subjects’ dataset showing that some antihypertensive drugs, used as single or in certain combinations, have the potential to both delay the rate of MCI conversion as well as reduce the proportion of individuals who convert to MCI. The effect of antihypertensive drugs on the proportion of participants converting to MCI was statistically significant in males compared to females, in older subjects compared to younger ones, and in the absence of genotypic APOE4 alleles compared to the presence of such allele. Furthermore, we corroborate the results of earlier studies (12, 13), showing an association between cognitive dysfunction and older age, male gender, and the presence of genotypic APOE4, with no significant interaction between these genetic risk factors and antihypertensives’ drug use. Our results have substantial and direct implications in the clinical treatment of elderly hypertensive individuals who are likely to be at high risk of cognitive decline and dementia. Nevertheless, one could not draw conclusion on the causes and mechanistic implications of the obtained results because of the study design, especially in the older subgroup of participants, where many additional physiological factors are at play.
Among the different classes of drugs, CCBs seem to be the most effective drug in reducing the rate of normal to MCI conversion, whereas ACEIs showed no effect. Our results agree with those of Hanon et al. (19) who reported significant efficacy of CCBs and no efficacy for ACEIs in preventing cognitive decline in both normal and MCI elderly subjects. Similarly, a systematic review and meta-analysis (27) reported significant effect of CCBs in reducing the risk of dementia in patients enrolled in randomized clinical trials and prospective cohort studies, whereas in PROGRESS trial conducted in stroke patients, the ACEI perindopril regimen did not reduce the risk of dementia in stroke patients despite reduction in BP . Our relatively large and prospective study sample might have clarified previous conflicting studies and gives rise to further research.
The absence of an association between arterial BP lowering effect and beneficial effect on cognitive function (19, 28) suggests that mechanisms other than antihypertensive activity contribute to the CCBs prevention of cognitive decline. The effect of antihypertensives may be due to preventing or minimizing one or more factors thought to play a direct or indirect role in cognitive decline in elderly patients, among them oxidative stress and A-β accumulation, blood brain barrier (BBB) damage/leakage, abnormal brain autoregulation and loss of neurovascular coupling (29). Such new knowledge shows that prevention of dementia might begin earlier as previously thought and current measures may might be less than effective.
Calcium (Ca2+) as a second messenger, regulates many signaling pathways including cell survival, proliferation, differentiation and apoptosis, through various mechanisms (30). Tight Ca2+ regulation is especially critical in brain neuronal cells. The dysregulation in the neuronal Ca2+ concentration can initiate signaling cascades leading to impaired balance between synaptic long-term potentiation (LTP) and long-term depression (LTD) in favor of LTD, hence increasing the threshold frequency for LTP induction. The imbalance in the activity of the two can lead to reduced synaptic strength and synaptic loss, leading to impairment in memory formation, storage, and neuronal atrophy (31). Ca2+ dysregulation thought to play a role in cognitive decline in normal aging (32), as well as in the pathophysiology of AD, including the MCI stage of the disease (33).
The perturbation in neuronal Ca2+ homeostasis extends also to enhanced Ca2+ influx through plasma membrane channels, including the L-type VGCCs (34). The increase in Ca2+ influx through these channels has been shown to be among mechanisms underlying Aβ neurotoxicity (35). Furthermore, studies in transgenic mouse model of AD show that some DHPs beneficially influence processes that contribute to Aβ accumulation; brain Aβ production and Aβ clearance across the BBB (36). Therefore, our work underlines evidence that blocking presynaptic calcium conducting ion channels avoids excitotoxic neurodegeneration and has future potential for more selective drugs.
We observed a significant effect of the selective L-type VGCC antagonists, DHPs, in delaying cognitive decline when compared to non-DHPs, further suggesting that biological, probably cellular mechanisms other than antihypertensive activity, contribute to DHP neuroprotective effects. Our results agree with studies showing that L-type VGCC antagonists improve learning and memory in aged animals (37) and in patients with cognitive dysfunction (38). Similarly, the double-blind placebo-controlled Syst-Eur (Systolic Hypertension in Europe) trial reported significant reduction in the incidence of dementia in older cognitively normal hypertensive subjects on DHP regimen (38).
The efficacy of DHPs were not unified across different DHPS: We observed significant effect with amlodipine and nifedipine, and lesser effect with other DHPs. DHPs are known to be mostly lipophilic able to cross the BBB, vasodilating cerebral vasculature, and increasing brain perfusion (39). However, amlodipine is hydrophilic and less likely to cross the BBB, and likely to need a transporter due to it positive charge (40), suggesting that peripheral mechanisms may play a role too, and that control of cognitive function might be more complex than previously thought.
Studies in transgenic mouse model of AD report no beneficial effects of nifedipine in reducing Aβ production and improving its clearance (36), suggesting that other mechanism, such as DHP improvements in cerebral blood flow by relaxing vascular smooth muscle cells, especially in brain capillaries, may account for the DHPs reduction in cognitive impairment in the MCI stages of the disease (41). However, it is also possible that the efficacy of DHPs is dependent on the type of cognitive impairment whether due to AD pathology or vascular pathology (42). Patients with chronic hypertension, as is the case with most participants in this database, are therefore likely to have cognitive impairment not only due to AD pathology, but also due to vascular pathology. Therefore, our work gives further evidence of the dual effect of CCBs in contrasting pathological processes that lead to MCI stages.
Like CCBs, we observed significant positive effects of diuretics in delaying the conversion to MCI. Our results agree with those of a systematic review of 15 studies showing the beneficial effects of diuretics in population of cognitively normal adults after mean of 6-year follow up (43). The inhibition of Aβ oligomerization as well as dissociation of pre-aggregated Aβ oligomers have been thought to contribute to the diuretic beneficial effects (44), though this effect may vary depending on the ability of these agents to cross the BBB (45). Nevertheless, our results contrast those of Walker et al. (46) who reported a 14% increase in dementia per every 1000 people treated with diuretics. The discrepancy between our results and those of Walker may stem from age differences in the subjects under study, as we clearly show the effect of antihypertensives on the rate of MCI conversion to be age dependent. Like us, Yasar et al. (47) also reported significant beneficial effects of diuretics on cognition in cognitively normal adults, 75 years and older, who transition to MCI, raising the possibility that the beneficial effects of these drugs may be more pronounced in the pre-MCI stage of the disease and be less efficacious when the dementia sets in as is the case with participants in the Walker et al. study (46). We see these findings as important to work towards more personalized and therefore more effective dementia prevention algorithms
We observed a synergistic effect of α1-ABs and diuretics in delaying the MCI conversion. Although like diuretics, α1-ABs are known to reduce Aβ toxicity and tau hyperphosphorylation (48), the synergistic effect of the combined regimen suggests that the two may act via differing mechanisms in reducing the rate of MCI conversion. For example, α1-ABs have been shown to exert anti-inflammatory effects and improve memory and cognitive function in animal models of AD (49). Inflammation has been identified in the preclinical stages of AD (50]), and its reduction by α1-ABs has the potential to delay cognitive decline.
The representativeness of the sample is always a question in larger study population like here as the sample must yield insights and observations that closely align with the sample population. Here we focused on strict exclusion criteria – excluding potential confounders – as a sampling method so that the internal validity of the research is maintained.
Although our results have implications for the treatment of elderly hypertensive subjects at risk of dementia, it is important to acknowledge the limitations of the study: Current results are related to the retrospective analysis of highly selective NACC dataset, which may or may not represent the general population. The large sample size and the strict inclusion criteria into the data records the risk of confounders to measure target variables is nearly at the level of a prospective study. Although we observed similar effect of antihypertensive drugs on amnestic and non-amnestic MCI participants, the mechanism behind such effect may differ between the two subgroups. Furthermore, the effect of antihypertensives may vary depending on the type of dysfunction and the number of dysfunctional domains.
We did not have the clinical and (molecular) neuroimaging data on these participants in order to assess the effect of antihypertensive drugs on preclinical biomarkers of the disease. Especially, PET shows altered glucose metabolism and accumulation of Aβ and tau (51). Imaging is the optimal method to follow the effect of treatments since changes in MRI scans are seen prior to those in cognition (51). Nevertheless, because of the strong effect size and the use of genotype data in our large study, neuroimaging would have not added new knowledge using the chosen study design.
The confirmation bias and overconfidence on clinical examination and diagnostic are always a problem in such highly selective patient population (52). Whereas a loss of follow-up bias is a known problem for observational studies like here (52), a commission bias (52) might have a special importance in the antihypertensive treatment in such patients at risk for decline of cognitive function. However, since in the current setting, the outcome status of participants is not known at the start of data sampling and as the subjects were enrolled using different methods and for different research purposes, the potential bias could be minimized for this specific study design. Finally, the lack of autopsy data prevents us from assessing the effect of antihypertensive drugs on cognitive impairment due to AD pathology vs. cognitive impairment due to the possible underlying vascular pathology.
Clinical impact of the findings and generalizability
Antihypertensive treatments are beneficial to cognitive function by lowering BP and/or by specific neuroprotective effects. The conflicting previous results may have their routes in insufficient follow-up (53), inconsistent study population (54), or less than adequate methodology (54). Our data give a unique opportunity to control the pathophysiological mechanism leading the cognitive decline in such patients. Antihypertensive subclasses, CCβs, diuretics, α1-A blockers, and vasodilators, have been associated with beneficial effects on cognitive function beyond arterial BP reduction. Nevertheless, cardio-geriatric use of some of these antihypertensives is currently limited: Alpha-blockers and vasodilators are not used as first line treatment for hypertension since they are less effective in reducing BP when used as monotherapy (55). For diuretic, the limitation in use is related to significant adverse effects, especially for the loop diuretics (56), and this class of drugs are not recommended in the treatment of hypertension in the absence of other comorbidities such as congestive heart failure. The design of more selective and potent antihypertensive subclasses may open the door for novel antidementive medications, which then will also require re-evaluation of the current clinical guidelines. Additionally, the complexity of effects suggests that the prevention of dementia may substantially profit from IT-based treatment algorithms.
Genotype data of this study give interesting insight into the role of genetic risk factors in the pathophysiology of cognitive decline. In addition, our data underline the importance of epigenetic effects. Future studies are needed to improve the understanding of the interplay between genetic and epigenetic factors and correlation with pathophysiological mechanisms, all of which will pave the way for novel therapeutic approaches.
The generalizability and therefore the external validity of this study is given as the patients are prospectively sampled according to specific characteristics (risk of cognitive decline) and the documented cause-and-effect relationship applies not across subdivisions of the target population. Additionally, the multi-site study design further eliminating the risk of bias and underlines the generalizability of the study. However, a low BP threshold that could be deleterious for cognitive function should still be determined.
On a large and representative sample size and with strict inclusion criteria, we found that antihypertensive treatments use was associated with delayed onset of mild cognitive impairment. This preventive model has direct effect on the daily treatment of patients as the choice of combination of antihypertensive therapy should consider the combination which would lead to an optimum benefit on cognitive function in patients treated with antihypertensive drugs. The challenge is to develop identification and treatment strategies applicable to early MCI that will reduce these adverse outcomes, making the treatment/prevention more effective.
Acknowledgement: The authors thank the National Alzheimer’s Coordinating Center (NACC) for provision of the Uniform Data Set, and Ms. Merilee Taylan and Mr. Zachary Miller for guidance and clarifications related to the data set. We are indebted to the study members and their families for their ongoing commitment and support of NACC research philosophy. The NACC database is funded by NIA/NIH Grant U01 AG016976. NACC data are contributed by the NIA-funded ADCs: P30 AG019610 (PI Eric Reiman, MD), P30 AG013846 (PI Neil Kowall, MD), P50 AG008702 (PI Scott Small, MD), P50 AG025688 (PI Allan Levey, MD, PhD), P50 AG047266 (PI Todd Golde, MD, PhD), P30 AG010133 (PI Andrew Saykin, PsyD), P50 AG005146 (PI Marilyn Albert, PhD), P50 AG005134 (PI Bradley Hyman, MD, PhD), P50 AG016574 (PI Ronald Petersen, MD, PhD), P50 AG005138 (PI Mary Sano, PhD), P30 AG008051 (PI Thomas Wisniewski, MD), P30 AG013854 (PI Robert Vassar, PhD), P30 AG008017 (PI Jeffrey Kaye, MD), P30 AG010161 (PI David Bennett, MD), P50 AG047366 (PI Victor Henderson, MD, MS), P30 AG010129 (PI Charles DeCarli, MD), P50 AG016573 (PI Frank LaFerla, PhD), P50 AG005131 (PI James Brewer, MD, PhD), P50 AG023501 (PI Bruce Miller, MD), P30 AG035982 (PI Russell Swerdlow, MD), P30 AG028383 (PI Linda Van Eldik, PhD), P30 AG053760 (PI Henry Paulson, MD,PhD), P30 AG010124 (PI John Trojanowski, MD, PhD), P50 AG005133 (PI Oscar Lopez, MD), P50 AG005142 (PI Helena Chui, MD), P30 AG012300 (PI Roger Rosenberg, MD), P30 AG049638 (PI Suzanne Craft, PhD), P50 AG005136 (PI Thomas Grabowski, MD), P50 AG033514 (PI Sanjay Asthana, MD, FRCP), P50 AG005681 (PI John Morris, MD), P50 AG047270 (PI Stephen Strittmatter, MD, PhD).
Ethical standards: The authors transfer, assign, and otherwise convey all copyright ownership worldwide, in all languages, to Serdi in the event the manuscript is published.
Conflict of interest: ZS, RP, JY, MT, DH, BS have no conflict of interest to report.
1. Crous-Bou M, Minguillon C, Gramunt N, Molinuevo JL: Alzheimer’s disease prevention: from risk factors to early intervention. Alzheimers Res Ther 2017, 9(1):71.
2. Selkoe DJ: Alzheimer’s disease: genes, proteins, and therapy. Physiological reviews 2001, 81(2):741-766.
3. Clement F, Gauthier S, Belleville S: Executive functions in mild cognitive impairment: emergence and breakdown of neural plasticity. Cortex 2013, 49(5):1268-1279.
4. Kochan NA, Breakspear M, Slavin MJ, Valenzuela M, McCraw S, Brodaty H, Sachdev PS: Functional alterations in brain activation and deactivation in mild cognitive impairment in response to a graded working memory challenge. Dement Geriatr Cogn Disord 2010, 30(6):553-568.
5. Schaller BJ: Strategies for molecular imaging dementia and neurodegenerative diseases. Neuropsychiatr Dis Treat 2008, 4(3):585-612.
6. Jacobs AH, Li H, Winkeler A, Hilker R, Knoess C, Ruger A, Galldiks N, Schaller B, Sobesky J, Kracht L et al: PET-based molecular imaging in neuroscience. Eur J Nucl Med Mol Imaging 2003, 30(7):1051-1065.
7. Gupta Y, Lama RK, Kwon GR, Alzheimer’s Disease Neuroimaging I: Prediction and Classification of Alzheimer’s Disease Based on Combined Features From Apolipoprotein-E Genotype, Cerebrospinal Fluid, MR, and FDG-PET Imaging Biomarkers. Front Comput Neurosci 2019, 13:72.
8. Petersen RC: Mild cognitive impairment as a diagnostic entity. J Intern Med 2004, 256(3):183-194.
9. Petersen RC, Doody R, Kurz A, Mohs RC, Morris JC, Rabins PV, Ritchie K, Rossor M, Thal L, Winblad B: Current concepts in mild cognitive impairment. Arch Neurol 2001, 58(12):1985-1992.
10. Smith EE, Muzikansky A, McCreary CR, Batool S, Viswanathan A, Dickerson BC, Johnson K, Greenberg SM, Blacker D: Impaired memory is more closely associated with brain beta-amyloid than leukoaraiosis in hypertensive patients with cognitive symptoms. PLoS One 2018, 13(1):e0191345.
11. Petersen RC: Mild Cognitive Impairment. Continuum (Minneap Minn) 2016, 22(2 Dementia):404-418.
12. Petersen RC, Roberts RO, Knopman DS, Geda YE, Cha RH, Pankratz VS, Boeve BF, Tangalos EG, Ivnik RJ, Rocca WA: Prevalence of mild cognitive impairment is higher in men. The Mayo Clinic Study of Aging. Neurology 2010, 75(10):889-897.
13. Burke SL, Hu T, Spadola CE, Li T, Naseh M, Burgess A, Cadet T: Mild cognitive impairment: associations with sleep disturbance, apolipoprotein e4, and sleep medications. Sleep Med 2018, 52:168-176.
14. Gorelick PB, Scuteri A, Black SE, Decarli C, Greenberg SM, Iadecola C, Launer LJ, Laurent S, Lopez OL, Nyenhuis D et al: Vascular contributions to cognitive impairment and dementia: a statement for healthcare professionals from the american heart association/american stroke association. Stroke 2011, 42(9):2672-2713.
15. Pal K, Mukadam N, Petersen I, Cooper C: Mild cognitive impairment and progression to dementia in people with diabetes, prediabetes and metabolic syndrome: a systematic review and meta-analysis. Soc Psychiatry Psychiatr Epidemiol 2018, 53(11):1149-1160.
16. Kalmijn S, Foley D, White L, Burchfiel CM, Curb JD, Petrovitch H, Ross GW, Havlik RJ, Launer LJ: Metabolic cardiovascular syndrome and risk of dementia in Japanese-American elderly men. The Honolulu-Asia aging study. Arterioscler Thromb Vasc Biol 2000, 20(10):2255-2260.
17. Dlugaj M, Winkler A, Dragano N, Moebus S, Jockel KH, Erbel R, Weimar C, Heinz Nixdorf Recall Study Investigative G: Depression and mild cognitive impairment in the general population: results of the Heinz Nixdorf recall study. J Alzheimers Dis 2015, 45(1):159-174.
18. Mossello E, Pieraccioli M, Nesti N, Bulgaresi M, Lorenzi C, Caleri V, Tonon E, Cavallini MC, Baroncini C, Di Bari M et al: Effects of low blood pressure in cognitively impaired elderly patients treated with antihypertensive drugs. JAMA Intern Med 2015, 175(4):578-585.
19. Hanon O, Pequignot R, Seux ML, Lenoir H, Bune A, Rigaud AS, Forette F, Girerd X: Relationship between antihypertensive drug therapy and cognitive function in elderly hypertensive patients with memory complaints. J Hypertens 2006, 24(10):2101-2107.
20. Edwards JD, Ramirez J, Callahan BL, Tobe SW, Oh P, Berezuk C, Lanctot K, Swardfager W, Nestor S, Kiss A et al: Antihypertensive Treatment is associated with MRI-Derived Markers of Neurodegeneration and Impaired Cognition: A Propensity-Weighted Cohort Study. J Alzheimers Dis 2017, 59(3):1113-1122.
21. Fink HA, Jutkowitz E, McCarten JR, Hemmy LS, Butler M, Davila H, Ratner E, Calvert C, Barclay TR, Brasure M et al: Pharmacologic Interventions to Prevent Cognitive Decline, Mild Cognitive Impairment, and Clinical Alzheimer-Type Dementia: A Systematic Review. Ann Intern Med 2018, 168(1):39-51.
22. Sheffield JM, Karcher NR, Barch DM: Cognitive Deficits in Psychotic Disorders: A Lifespan Perspective. Neuropsychol Rev 2018, 28(4):509-533.
23. Kaufer DI, Cummings JL, Ketchel P, Smith V, MacMillan A, Shelley T, Lopez OL, DeKosky ST: Validation of the NPI-Q, a brief clinical form of the Neuropsychiatric Inventory. J Neuropsychiatry Clin Neurosci 2000, 12(2):233-239.
24. Morris JC: Clinical dementia rating: a reliable and valid diagnostic and staging measure for dementia of the Alzheimer type. Int Psychogeriatr 1997, 9 Suppl 1:173-176; discussion 177-178.
25. Rizvi SM, Shaikh S, Waseem SM, Shakil S, Abuzenadah AM, Biswas D, Tabrez S, Ashraf GM, Kamal MA: Role of anti-diabetic drugs as therapeutic agents in Alzheimer’s disease. EXCLI J 2015, 14:684-696.
26. Geifman N, Brinton RD, Kennedy RE, Schneider LS, Butte AJ: Evidence for benefit of statins to modify cognitive decline and risk in Alzheimer’s disease. Alzheimers Res Ther 2017, 9(1):10.
27. den Brok M, van Dalen JW, Abdulrahman H, Larson EB, van Middelaar T, van Gool WA, van Charante EPM, Richard E: Antihypertensive Medication Classes and the Risk of Dementia: A Systematic Review and Network Meta-Analysis. J Am Med Dir Assoc 2021, 22(7):1386-1395 e1315.
28. Tzourio C, Anderson C, Chapman N, Woodward M, Neal B, MacMahon S, Chalmers J, Group PC: Effects of blood pressure lowering with perindopril and indapamide therapy on dementia and cognitive decline in patients with cerebrovascular disease. Arch Intern Med 2003, 163(9):1069-1075.
29. Yang W, Luo H, Ma Y, Si S, Zhao H: Effects of Antihypertensive Drugs on Cognitive Function in Elderly Patients with Hypertension: A Review. Aging Dis 2021, 12(3):841-851.
30. Sushma, Mondal AC: Role of GPCR signaling and calcium dysregulation in Alzheimer’s disease. Mol Cell Neurosci 2019, 101:103414.
31. Popugaeva E, Pchitskaya E, Bezprozvanny I: Dysregulation of neuronal calcium homeostasis in Alzheimer’s disease – A therapeutic opportunity? Biochem Biophys Res Commun 2017, 483(4):998-1004.
32. Kumar A, Bodhinathan K, Foster TC: Susceptibility to Calcium Dysregulation during Brain Aging. Front Aging Neurosci 2009, 1:2.
33. Jaworska A, Dzbek J, Styczynska M, Kuznicki J: Analysis of calcium homeostasis in fresh lymphocytes from patients with sporadic Alzheimer’s disease or mild cognitive impairment. Biochim Biophys Acta 2013, 1833(7):1692-1699.
34. Thibault O, Landfield PW: Increase in single L-type calcium channels in hippocampal neurons during aging. Science 1996, 272(5264):1017-1020.
35. Anekonda TS, Quinn JF, Harris C, Frahler K, Wadsworth TL, Woltjer RL: L-type voltage-gated calcium channel blockade with isradipine as a therapeutic strategy for Alzheimer’s disease. Neurobiol Dis 2011, 41(1):62-70.
36. Paris D, Bachmeier C, Patel N, Quadros A, Volmar CH, Laporte V, Ganey J, Beaulieu-Abdelahad D, Ait-Ghezala G, Crawford F et al: Selective antihypertensive dihydropyridines lower Abeta accumulation by targeting both the production and the clearance of Abeta across the blood-brain barrier. Mol Med 2011, 17(3-4):149-162.
37. Disterhoft JF, Oh MM: Pharmacological and molecular enhancement of learning in aging and Alzheimer’s disease. J Physiol Paris 2006, 99(2-3):180-192.
38. Forette F, Seux ML, Staessen JA, Thijs L, Babarskiene MR, Babeanu S, Bossini A, Fagard R, Gil-Extremera B, Laks T et al: The prevention of dementia with antihypertensive treatment: new evidence from the Systolic Hypertension in Europe (Syst-Eur) study. Arch Intern Med 2002, 162(18):2046-2052.
39. van Zwieten PA: The pharmacological properties of lipophilic calcium antagonists. Blood Press Suppl 1998, 2:5-9.
40. Yoshida H, Ayaori M, Suzukawa M, Hosoai H, Nishiwaki M, Ishikawa T, Nakamura H: Effects of Ca-antagonists on oxidative susceptibility of low density lipoprotein (LDL). Hypertens Res 1995, 18(1):47-53.
41. Wolters FJ, Zonneveld HI, Hofman A, van der Lugt A, Koudstaal PJ, Vernooij MW, Ikram MA, Heart-Brain Connection Collaborative Research G: Cerebral Perfusion and the Risk of Dementia: A Population-Based Study. Circulation 2017, 136(8):719-728.
42. Ogoh S: Relationship between cognitive function and regulation of cerebral blood flow. J Physiol Sci 2017, 67(3):345-351.
43. Tully PJ, Hanon O, Cosh S, Tzourio C: Diuretic antihypertensive drugs and incident dementia risk: a systematic review, meta-analysis and meta-regression of prospective studies. J Hypertens 2016, 34(6):1027-1035.
44. Zhao W, Wang J, Ho L, Ono K, Teplow DB, Pasinetti GM: Identification of antihypertensive drugs which inhibit amyloid-beta protein oligomerization. J Alzheimers Dis 2009, 16(1):49-57.
45. Hampel P, Romermann K, MacAulay N, Loscher W: Azosemide is more potent than bumetanide and various other loop diuretics to inhibit the sodium-potassium-chloride-cotransporter human variants hNKCC1A and hNKCC1B. Sci Rep 2018, 8(1):9877.
46. Walker VM, Davies NM, Martin RM, Kehoe PG: Comparison of Antihypertensive Drug Classes for Dementia Prevention. Epidemiology 2020, 31(6):852-859.
47. Yasar S, Lin FM, Fried LP, Kawas CH, Sink KM, DeKosky ST, Carlson MC, Ginkgo Evaluation of Memory Study I: Diuretic use is associated with better learning and memory in older adults in the Ginkgo Evaluation of Memory Study. Alzheimers Dement 2012, 8(3):188-195.
48. Coelho BP, Gaelzer MM, Dos Santos Petry F, Hoppe JB, Trindade VMT, Salbego CG, Guma F: Dual Effect of Doxazosin: Anticancer Activity on SH-SY5Y Neuroblastoma Cells and Neuroprotection on an In Vitro Model of Alzheimer’s Disease. Neuroscience 2019, 404:314-325.
49. Katsouri L, Vizcaychipi MP, McArthur S, Harrison I, Suarez-Calvet M, Lleo A, Lloyd DG, Ma D, Sastre M: Prazosin, an alpha(1)-adrenoceptor antagonist, prevents memory deterioration in the APP23 transgenic mouse model of Alzheimer’s disease. Neurobiol Aging 2013, 34(4):1105-1115.
50. Akiyama H, Barger S, Barnum S, Bradt B, Bauer J, Cole GM, Cooper NR, Eikelenboom P, Emmerling M, Fiebich BL et al: Inflammation and Alzheimer’s disease. Neurobiol Aging 2000, 21(3):383-421.
51. Heiss WD, Rosenberg GA, Thiel A, Berlot R, de Reuck J: Neuroimaging in vascular cognitive impairment: a state-of-the-art review. BMC Med 2016, 14(1):174.
52. Tripepi G, Jager KJ, Dekker FW, Wanner C, Zoccali C: Bias in clinical research. Kidney Int 2008, 73(2):148-153.
53. Rouch L, Cestac P, Hanon O, Cool C, Helmer C, Bouhanick B, Chamontin B, Dartigues JF, Vellas B, Andrieu S: Antihypertensive drugs, prevention of cognitive decline and dementia: a systematic review of observational studies, randomized controlled trials and meta-analyses, with discussion of potential mechanisms. CNS Drugs 2015, 29(2):113-130.
54. Oscanoa TJ, Amado J, Vidal X, Romero-Ortuno R: Angiotensin-Receptor Blockers and the Risk of Alzheimer s Disease: A Meta-analysis. Curr Rev Clin Exp Pharmacol 2021, 16(1):73-78.
55. Musso CG, Alfie J: Resistant hypertension in the elderly-second line treatments: aldosterone antagonists, central alpha-agonist agents, alpha-adrenergic receptor blockers, direct vasodilators, and exogenous nitric oxide donors. Cardiovasc Hematol Agents Med Chem 2015, 12(3):170-173.
56. Feig PU: Cellular mechanism of action of loop diuretics: implications for drug effectiveness and adverse effects. Am J Cardiol 1986, 57(2):14A-19A.