Table 2 Examples of potential questions for comparative effectiveness research

Hypertensive patients at high risk of cognitive decline

Telmisartan versus other ARBs versus centrally acting ACEi versus non-centrally acting ACEi

Polymorphisms in ACE have been linked to Alzheimer’s disease [60] but whether central ACE inhibition will protect or harm is unclear. ACEi but not ARBs might accelerate beta-amyloid pathology by blunting ACE activity on non-angiotensin pathways and by inhibiting AT2 and AT4 receptors. One ARB in particular, telmisartan, has additional activity on PPAR gamma that might protect against neurodegeneration [61]

A network meta-analysis concluded that ARBs had more benefit on cognition than ACEi drugs (adjusted effect size 0.47 +/– 0.17, p = 0.04) [62]. Yet hypertension management with centrally acting versus non-centrally acting ACEi was associated with 25 % slower functional decline in Alzheimer’s patients [63]. A Phase II trial is underway to compare telmisartan to perindopril in patients with comorbid hypertension and Alzheimer’s (Table 1). Additional comparisons are needed

Hypertensive patients at risk of cognitive decline

Amlodipine or nifedipine versus other DHP CCBs

Most DHP CCBs are likely to penetrate the brain except for amlodipine. Nilvadipine and nitrendipine but not amlodipine decreased beta-amyloid accumulation and blunted apoptosis in a mouse model of Alzheimer’s. DHP CCBs varied in their capacity to increase amyloid clearance from the brain [64]. Effects on the brain may vary depending on their selectivity for different calcium channels [65]

In a small trial in hypertensive patients with MCI, nilvadipine versus amlodipine slowed cognitive decline and improved cerebral blood flow despite similar effects on blood pressure [34]. Nitrendipine and nimodipine, have clinical data to suggest utility for the prevention or treatment of dementia, respectively, while nifedipine was associated with an increased risk of cognitive decline (reviewed in [64]). A Phase III trial is underway to test nilvadipine as a treatment for Alzheimer’s (Table 1)

Diabetes patients at risk of cognitive decline

Centrally penetrant versus non-penetrant GLP-1 agonists

GLP-1 agonists have been shown to protect against hippocampal synapse loss, lower beta-amyloid pathology and related damage, reduce neuroinflammation, and promote neurogenesis. While exenatide, liraglutude, and lixisenatide cross the blood–brain barrier, albiglutide and dulaglutide are large proteins unlikely to reach the brain [18]

Treatment with liraglutide blocked decline in cerebral glucose metabolism over 6 months in Alzheimer’s patients in a Phase II trial [66]. Additional trials are underway to repurpose these drugs to treat cognitive impairment (Table 1) but not to compare cognitive outcomes of CNS-penetrant versus non-penetrant GLP-1 agonists for diabetes treatment

Diabetes patients with or without comorbid dementia

Choice of drugs to minimize the risk of severe hypoglycemia

Severe hypoglycemia can trigger acute cognitive impairment and possibly accelerate long-term cognitive decline [67]. The choice of drugs used to manage diabetes may alter the risk of severe hypoglycemia in some patients

Nursing home patients with both dementia and diabetes had up to 8× higher risk of severe hypoglycemia when treated with sulphonylurea instead of insulin analogs [68]. More research is needed on how the choice of drugs alters risk in diverse patient populations