Table 2.

Extraction table for mechanism studies: angiotensin receptor blocker (ARB), angiotensin converting enzyme inhibitor (ACE-I), and diuretic

Author	Method: subjects	Methods: groups	Method: treatment	Method: treatment route	Methods: treatment time	Method: outcome	Method: statistic	Result
AbdAlla et al. [17]	-Aged Tg2576 mice	-N = 2	-Control -Captopril		-6 months	-Aβ plaque load in hippocampus	ANOVA	-Captopril treated Tg mice had significantly lower Aβ, upregulated genes associated with neuronal membrane and neuronal process, reduced oxidative stress markers, reduced amyloidogenic process of APP, and decreased ACE and angiotensin II levels -This effect is most likely direct effect of the centrally acting captorpil (not observed in enalapril which only acts peripherally)
Bild et al. [18]	-Wistar rats	-N = 5	-Control (saline) -Angiotenisn II -Captorpil -Losartan -PD-123177	-i.c.v.	-7 days	-Y-maze task -Step through passive avoidance task -Superoxide dismutase (SOD) activity in hippocampus -Glutathione peroxidase (GPX) activity in hippocampus -Malondialdehyde (MDA) levels in hippocampus	ANOVA	-On Y maze task angiotensin II-treated animals preformed worse and captoril, losartan and PD 123177 treated animals did significantly better when compared to control group -In avoidance task angiotensin II-treated animals preformed worse and losartantreated animals performed better when compared to control group -SOD, GPX, and MDA activity was decreased in angiotensin II-treated rats and increased in captopril-treated rats, when compared to control group
Dong et al. [13]	-WT mice (C57BL/6 J)	-N = 3	-Sham surgery -Bilateral carotid occlusion and reperfusion + vehicle -Bilateral carotid occlusion and reperfusion + aliskiren 2.5 mg/kg	-Mini pump	-35 days	-Blood pressure -Cerebral blood flow -Y-maze test -Superoxide dismutase (SOD) -Tumor necrosis factor α (TNF α) -Brain angiotensin and renin expression	ANOVA	-Aliskerin did not alter blood pressure or cerebral blood flow -Aliskerin-treated animals did better on Y-maze task -Aliskerin treatment did not alter SOD levels but did TNF α -Aliskerin treatment decreased renin and angiotensin expression -Aliskerin treatment decreased white matter lesions in brain
Goel et al. [26]	-Wistar rats -Spontaneously hypertensive rats (SHR)	-N = 8	-Wistar rats treated with vehicle -Wistar rats treated with lipopolysaccharide (LPS) 25 μg to induce inflammation -Wistar rats treated with LPS 50 μg -Wistar rats treated with LPS 50 μg + perindopril 0.1 mg/kg -Wistar rats treated with perindopril 0.1 mg/kg -SHR treated with venicle -SHR treated with treated with LPS 25 μg -SHR treated with treated with LPS 25 μg + perindopril 0.1 mg/kg -SHR treated with perindopril 0.1 mg/kg	-i.c.v. LPS -po perindopril	-15 days	-Blood pressure -Morris water maze test -Spontaneous locomotor activity (SLA) -Brain tissue (cortex, hippocampus) ACE activity, ACE gene expression, angiotensin II level, cytokine level, reactive oxidative species (ROS), nitrate, caspase 3 activity, TNF α, NF-kB, IL-10,	ANOVA	-LPS caused impaired memory in SHR and in normal rats which was reversed by perindopril -In SHR ACE activity and gene expression, angiotensin II level, oxidative stress, and inflammatory markers were increased -LPS caused further increase in ACE activity and gene expression, angiotensin II level, oxidative stress, and inflammatory markers, which was decreased by perindopril
Hajjar et al. [32]	-N = 890 participants with hypertension and autopsy available	-N = 3	-N = 133 have used ARB -N = 577 have used other antihypertensive medication -N = 180 have not used antihypertensive medication	-po		-AD pathology (CERAD, ADRDA/Khachaturian, Pathologic diagnosis) -Vascular pathology (large artery infarcts, microinfarcts, hemorrhage, atherosclerosis, arteriosclerosis)	Logistic regression analysis	-ARB users when compared to other and no antihypertensive medication users had significantly less AD pathology ( 50 %reduction) -ARB use was associated with higher vascular pathology
Hajjar and Levey 2015 [33]	-N = 319 participants with CSF available	-N = 2	-N = 26 have used ARB -N = 293 have used other antihypertensive medication	-po	-3 years	-ARB use and longitudinal tau decline in people with dementia, MCI, and normal controls	Mixed model	-ARB use was associated with decreased levels of tau and p-tau in people with MCI when compared to other antihypertensive medication users
Justin et al. [22]	-Sprague Dawley rats	-N = 5	-Sham surgery (no carotid occlusion) -Bilateral carotid occlusion and reperfusion + carboxymethyl cellulose (CMC) -Bilateral carotid occlusion and reperfusion + telmisartan 5 mg/kg -Bilateral carotid occlusion and reperfusion + telmisartan 10 mg/kg -Bilateral carotid occlusion and reperfusion + telmisartan 5 mg/kg + nimodipine 5 mg/kg	-po	-9 days	-Day 2 neurological assessment -Day 7 behavioral assessment -Day 9 histopathological studies -Day 9 oxidative stress and inflammation markers	Logistic regression model	-Telmisartan pretreatment resulted in less neurologic deficit and locomotor function -Telmisartan increased glutamate, aspartate, and ATP levels in brain and decreased glutathione, nitric oxide levels -Telmisartan decreased the pro-inflammatory cytokine (IL-1β, IL-6, TNF-α), lipid peroxide and nitric oxide levels; and increased anti-inflammatory cytokine IL-10 level -Best results in all test was when telmisartan was given in combination with nimodipine
Kishi et al. [15]	-Spontaneously hypertensive rats (SHR)	-N = 5	-Vehicle -Telmisartan 1 mg/kg -Telmisartan 1 mg/kg + GW9662 1 mg/kg (a PPAR gamma antagonist) -GW9662 1 mg/kg -Telmisartan 1 mg/kg + ANA-12 0.5 mg/kg)	-po	-4 weeks	-Blood pressure and heart rate -Morris water test -Hippocampus BDNF level and α-tubulin expression	ANOVA	-Blood pressure and heart rate was not changed in any of the treatment groups -Animals treated with telmisartan alone or with GW9662 or with ANA-12 had higher BDNF levels in hippocampus than the comparison groups -Animals treated with telmisartan alone or with GW9662 performed better on Morris water test than the comparison groups
Kume et al. [35]	-20 patients with hypertension and AD	N = 2	-N = 10 Telmisartan 40-80 mg daily -N = 10 Amlodipine 5-10 mg daily	-po	-6 months	-Blood pressure -Heart rate -Cognitive test (MMSE, ADAS-Jcog, logical memory) -Regional cerebral blood flow (rCBF) measured by SPECT	ANCOVA	-Systolic blood pressure was better in both groups -rCBF improved significantly in areas of frontal, parietal, occipital lobe in telmisartan users- no improvement in cognition
Liu et al. [23]	-Wild mice (C57BL/6J) -Humans with AD and normal controls					-ACE2 (which transforms angiotensin 1–10 to angiotensin 1–9) activity (mice) -Aβ levels (Aβ43, Aβ42, Aβ40) -Angiotensin II levels (humans)	Mann-Whitney U test	-ACE2 converted Aβ43 to Aβ42 -ACE converted Aβ43 to Aβ40 -ACE2 activity was decreased in human serum -Serum angiotensin levels were similar in people with AD and normal controls
Nation et al. [34]	-N = 871 stroke and dementia free people with available CSF	-N = 3	-N = 90 have used ARB -N = 343 have used other antihypertensive medication -N = 438 have used no antihypertensive medication	-po	-24 months	-CSF Aβ1–42, p-tau levels -progression to dementia	ANCOVA	-Older people had lower CSF Aβ1–42 -ARB users had higher levels of CSF Aβ1–42 and lower levels of CSF p-tau then the other groups -ARB users were less likely to progress to dementia
Omote et al. [24]	-Spontaneously hypertensive rats (SHR)	-N = 6	-Control group was treated with carboxymethyl cellulose (CMC) -Olmesartan low dose 2 mg/kg -Olmesartan high dose 10 mg/kg -Azelnidipne low dose 2 mg/kg -Azelnidipne high dose 10 mg/kg -Combination olmesartan 1 mg/kg and azelnidipine 1 mg/kg	-po	-14 days	-Blood pressure (BP) -Pulse -Body weight -Regional cerebral blood flow (rCBF) -Serum triglyceride, LDL, HDL -Oxidative stress markers -Inflammatory markers -Neurovascular units -Infarct volume -Motor coordination and balance	ANOVA	-BP and pulse decreased while body weight and rCBF remained stable in all treatment groups compared to control group -Infarct volume decreased in olmisartan and azelnidipine-treated group, and it was more effective than low-dose monotherapy; however, high-dose azelnidipine (better BP reduction) was more effective than high-dose olmesartan -All treatments reduced oxidative markers, inflammatory markers, and preserved neurovascular unit in a dose-reponse matter
Ongali et al. [25]	-TgAPP mice -Wild-type mice	-N = 4	-Wild-type treated with vehicle -Wild-type treated with losartan 1 and later 10 mg/kg -TgAPP treated with vehicle -TgAPP treated with losartan 1 and later 10 mg/kg	-po	-3 months (old) and 10 months (young)	-Morris water maze test -Cerebral blood flow CBF) -Glucose FDG-PET -Blood pressure -Vascular reactivity in brain tissue -AD neuropathology -SOD levels in brain -Angiotensin 1 and 4 receptors (ANG1R, ANG4R)	ANOVA	-In old TgAPP mice losartan did not improve learning but improved memory acquisition and recall -In old TgAPP SOD, ANG1R, and ANG4R levels were increased which was reduced with Losartan -Losartan increased CBF i cerebral glucose uptake, and cerebrovascular responsivenessin old TgAPP mice -Losartan did not decrease AD pathology (Aβ1–42)
Wang et al. [19]	-Tg2576 mice		-1600 FDA approved drugs were screened for Aβ regulating effect: 184 drugs lowered Aβ by >30 % and 26 drugs increased Aβ levels by >30 % -Short-term treatment: propranolol, carvedilol, nicardipine, losartan, amiloride, hydralazine, furosemide, trandalopril -Long-term treatment: nicardipine, propranolol	-Drinking water	-1 month -Selected group for chronic treatment: 6 months	-Blood pressure measurement -Total Aβ1–40 or Aβ1–42 in the brain and plasma after short-term treatment -Morris water maze test after long-term treatment	ANOVA	-Short-term use: propranolol, losartan significantly reduced blood pressure by 20 %; propranolol, nicardipine, carvedilol reduced significantly Aβ1–42 and 1–40 by 40 % in brain and plasma, losartan reduced Aβ1–42 but not 1–40 -Furosemide and trandalopril significantly reduced Aβ without affecting BP -Long-term treatment with propranolol and nicardipine did not improve cognition but decreased Aβ in brain but not plasma
Yamada et al. [27]	-Wild-type mouse treated i.c.v Aβ25–35 (AD mouse model)	-N = 3	-Perindopril (0.1, 0.3, 1.0 mg/kg) -Imidapril (0.3, 1.0, 3.0 mg/kg) -Enalapril (1.0, 3.0, 10 mg/kg)	-po	-5 days	-Spontaneous alteration test (SAT) -Object recognition test -Spontaneous locomotor activities -Anxiety-related behavior in elevated plus maze -ACE activity in brain and plasma	ANOVA	-Perindopril in all dosages improved working memory (measured by SAT), object recognition -Plasma ACE was inhibited perindopril 1 mg/kg, imidapril 3 mg/kg, 10 mg/kg enalapril by 90 % -Brain SACE was inhibited by 50 % by 1 mg/kg perindopril but was less in other ACE-Is
Zhai et al. [28]	-Wistar rats -Spontaneously hypertensive rats (SHR)	-N = 4	-Wistar rats -SHR treated with Vehicle -SHR treated with Telmisartan 0.3 mg/kg SHR -SHR treated with Telmisartan 3 mg/kg SHR	-po	-3, 9, and 15 months	-Blood pressure -LDL receptor, ApoE expression in the cortex and hippocampus	ANOVA	-In the cortex and hippocampus of SHR ApoE expression and LDL receptors was increased at all ages but was significantly reduced in both doses of telmisartan -At low dose, blood pressure remained unchanged
Zou et al. [29]	-Tg2576 mice -APP J20 mice	-N = 3	-Tg2576 mice treated with captopril 0.25 mg/kg -Tg2576 mice treated with vehicle -TgAPP J20 mice	-po	-11 months	-Aβ levels in brain (Aβ1–40, Aβ1–42, Aβ1–43) -ACE activity in brain	Student’s t test Spearman’s rank test	-In TgAPP mouse Aβ1–43 occurs before Aβ1–40 and Aβ1–42 -ACE converted Aβ1–43 to Aβ1–40 -Captopril pretreatment decreased ACE activity by 26 % and increased Aβ1–43 deposition -In people with AD serum, Aβ1–43 level is higher and CSF level is lower when compared to normal control