Treatment of vascular risk factors in patients with a diagnosis of Alzheimer’s disease: a systematic review

One RCT examined the effectiveness of a vascular care package that included aspirin, folic acid and pyridoxine drug therapy as well as advice on smoking cessation, weight loss and exercise [9]. Within this trial, 123 patients with mild AD with neuroimaging evidence of co-existent cerebrovascular disease (white matter hyperintensities or infarcts) were randomised and there was a two-year follow-up. Eleven patients died during follow-up and 18 dropped out. Significant differences in homocysteine and cholesterol occurred, but there was no difference, not even a trend, in the primary endpoint of disability or in secondary endpoints of MMSE or the Revised Memory and Behavioural Problems Checklist (Additional file 1) [9].

We found no RCTs comparing antihypertensive treatment with placebo, although we identified two RCTs that compared the effectiveness of different antihypertensive agents in patients with AD.

One trial examined the hypothesis that brain-penetrating angiotensin converting enzyme inhibitors (ACE-I) would slow the rate of cognitive decline in patients with mild to moderate AD with hypertension [10]. This was on the rationale that certain components of the renin-angiotensin system (RAS) may have a role in learning and memory processes. In this trial, 162 patients were randomly assigned on an open basis to one of three treatment options: a brain-penetrating ACE-I (perindopril or captopril), a non-brain penetrating inhibitor (enalapril or imidapril), or a calcium antagonist (nifedipine or nilvadipine). Mean baseline MMSE values were 19.3, 20.7 and 20.5 in the three groups respectively. There was no difference in blood pressure between the three groups. The mean decline in the primary endpoint of the MMSE during the one-year follow-up was significantly lower in the groups treated with a brain-penetrating ACE-I (0.6, standard error (SE) = 0.1) than those in the other two groups: 4.6 (SE = 0.3) and 4.9 (SE = 0.3) respectively [10].

In a small study looking primarily at cerebral blood flow, 20 patients with AD were randomised to open therapy with telmisartan, an angiotensin receptor blocker (ARB), or amlodipine for six months. Cognition did not change in the telmisartan group but declined in the amlodipine group [11].

The only RCTs of diabetic treatment are of peroxisome proliferator-activated receptor gamma (PPARγ) agonists, and have examined whether this particular class of drug has a protective effect rather than whether tighter diabetic control itself improves outcome. PPARγ agonists increase glucose sensitivity, regulate lipid metabolism and promote mitochondrial biogenesis [12],[13]. They also exhibit robust anti-inflammatory actions via their ability to suppress NF-κB-dependent gene expression [14],[15]. AD is typified by impaired glucose utilization in the brain and a glial-mediated inflammatory response, suggesting the potential utility of these agents in the treatment of AD [14]-[16]. Studies in murine models of AD demonstrated that rosiglitazone lowers amyloid plaque burden, reduces vascular and plaque-associated inflammation, attenuates loss of synaptic connectivity, and improves memory and cognition [14],[17]-[19]. They have been tested both in patients with AD and diabetes, and patients with AD but without diabetes.

A pilot study suggested the PPARγ agonist rosiglitazone improved cognition in patients with mild-to-moderate AD [20]. Another small open study in 32 patients with both mild-to-moderate AD (or amnestic MCI) and diabetes mellitus not on insulin randomised patients between the PPARγ agonist pioglitazone or no additional treatment. A significant decrease in the AD Assessment Scale - Cognitive Subscale Japanese version (ADAS-JCog) and increase in the Wechsler Memory Scale Revised (WMS-R), but no change in MMSE, was found in the pioglitazone group [21]. Another small study in 42 patients with mild AD and type II DM primarily looked at regional cerebral blood flow but also included data on cognitive outcome, and suggested a protective effect of pioglitazone [22]. Patients were randomly assigned to open treatment with 15 or 30 mg daily pioglitazone in addition to their previous oral hypoglycaemic treatments. After six months, scores on the the MMSE, ADAS-JCog and WMS-R logical memory-I improved significantly in the pioglitazone group, while the ADAS-JCog worsened significantly in the control group. Neither group showed any significant change in the Frontal Assessment Battery and category fluency [22].

However, larger trials in patients with AD without diabetes have not replicated these positive results. A phase 2 trial in 511 non-diabetic patients with mild-to-moderate AD randomised participants between placebo or 2, 4 or 8 mg rosiglitazone [23]. There were no statistically significant differences in the primary endpoints (ADAS-Cog and Clinician’s Interview-Based Impression of Change Plus Caregiver Input (CIBIC+)) after 24 weeks’ treatment. Results were also stratified by ApoE genotype in a subgroup (n = 323); there was a significant interaction between ApoE epsilon4 allele status and ADAS-Cog (P = 0.014) with ApoE4-negative individuals appearing to show an improvement in response to rosiglitazone, whereas ApoE epsilon4 allele carriers showed no improvement and some decline was noted. But this interaction with ApoE status was not confirmed in a subsequent study. This double-blind phase 2 trial randomised 581 individuals with mild-to-moderate AD without diabetes to once-daily placebo, 2 mg or 8 mg rosiglitazone, placebo, or donepezil as a positive control [24]. At week 24, there was no significant difference from placebo in change from baseline in the ADAS-Cog score and CIBIC + with either rosiglitazone dose in the 50% of individuals who were ApoE epsilon4-negative, or overall [24].

Statin (3-hydroxy-3-methylglutaryl-coenzyme A reductase inhibitors) treatment might act via a general cardiovascular protective effect due to cholesterol lowering or via a more specific effect on AD pathology. Cholesterol modulates the processing of amyloid precursor protein (APP)-related Aβ production in vitro and in animal model studies [25],[26]. AD neuropathologic lesions are demonstrable in patients who have coronary artery disease and no dementia [27]. Reducing cholesterol through the use of statins appears to affect the processing of APP and the production of Aβ. A diet high in cholesterol increased Aβ accumulation and AD-related pathology in the transgenic mouse model, with relatively little change in brain cholesterol [28], and feeding a high-cholesterol diet to rabbits has also produced Aβ deposits in the hippocampus [29].

A single-centre study randomised 67 patients with mild-to-moderate AD (MMSE score of 12 TO 28) to either atorvastatin or placebo on a double-blind basis for one year [30]. A significant positive effect on ADAS-Cog performance occurred after six months of atorvastatin therapy compared with placebo, although the difference was not quite significant at one year [30]. Another small study of 44 patients with probable AD of mild-to-moderate severity (MMSE score 12 to 26), which was primarily looking at the effects of stains on cerebrospinal fluid markers, assessed cognition as a secondary outcome [31]. Patients were randomised to placebo or simvastatin and the simvastatin group had less decline in MMSE score during follow-up [31].

A post hoc analysis was conducted on data pooled from three double-blind, placebo-controlled, clinical trials of galantamine in patients with AD divided into four treatment groups: statin plus galantamine (n = 42), statin alone (n = 50), galantamine alone (n = 614), or neither galantamine nor statin (n = 619) [32]. Galantamine was associated with a significant beneficial effect on cognitive status, and although there was no significant benefit from statins (P = 0.083), there was a trend which led the authors to suggest further studies were needed [32].

By contrast, a subsequent larger international multi-centre, double-blind, randomised, parallel-group study failed to confirm these benefits [33]. The authors examined the use of atorvastatin in patients without an indication for statin, that is, no cardiovascular disease and low-density lipoprotein cholesterol levels on study entry between 95 and 195 mg/dL [33]. The 640 participants randomised had mild-to-moderate probable AD (MMSE 13 to 25), were aged 50 to 90 years, and were all taking 10 mg donepezil daily. They were randomised to 80 mg/day atorvastatin or placebo for 72 weeks. There were no significant differences in the co-primary endpoints of ADAS-Cog score or global function assessed by the Alzheimer’s Disease Cooperative Study Clinical Global Impression of Change. The proportion discontinuing treatment was 34.1% in the statin group and 24.5% in the placebo group, with those discontinuations felt to be related to therapy in 11.5% and 1.5% respectively. The authors suggest a limitation is that statin therapy may be more useful in those patients with AD patients and co-existing cardiovascular disease and/or high cholesterol level [33].

One study identified 280 patients from a memory clinic who had AD with no evidence of cerebrovascular disease and followed them up for six months [34]. Each VRF (high blood pressure, dyslipidaemia, DM, smoking) was considered treated if the patient received a specific medication. MMSE decline was slower for patients with all VRF treated compared with no VRF treated. There was a non-significant trend in patients with only some VRF treated. Although limited by the small numbers in each group, analysis of individual VRF found a significant effect only for the treatment of dyslipidaemia with either a statin or a fibrate [34].

In the Dementia Progression Study of the Cache County Study on Memory, Health and Aging, 216 individuals with incident AD were followed for a mean of three years [35]. The Clinical Dementia Rating Sum of Boxes (CDR-Sum) increased an average of 1.69 points annually, indicating a steady decline in functioning. After adjustment for demographic variables and the baseline presence of cardiovascular conditions, use of statins (P = 0.03) and beta-blockers (P = 0.04) were associated with a slower annual rate of increase in CDR-Sum of 0.75 and 0.68 points respectively, while diuretic use was associated with a faster rate of increase in CDR-Sum (P = 0.01; 0.96 points annually) [35].

A number of studies have examined whether ant-hypertensive therapy is associated with reduced cognitive decline in AD. Some studies have included only hypertensive patients [36] whereas others also included normotensive individuals [37],[38].

In the longitudinal OPTIMA study, among the 141 patients with AD in whom blood pressure was recorded, the rate of decline on Cambridge Cognitive Examination (CAMCOG) scores showed an inverted U-shape dependent on diastolic blood pressure. The use of any antihypertensive medication in those with AD was related to significantly better CAMCOG scores (P = 0.008) [36].

In a prospectively collected database of 321 patients with AD and hypertension with a mean follow-up of 34 months, cognitive function was assessed yearly by MMSE [37]. Fifty-four per cent of patients received at least one antihypertensive drug while 33% of those patients without antihypertensive treatment were hypertensive. Medication included different classes of drugs. MMSE was significantly higher among patients using antihypertensive drugs compared to those without antihypertensive treatment after adjustment for main confounders (19.0 versus 17.5, P <0.0001 at three years) [37].

A case-control study investigated the association between hypertension and cognitive decline in 719 patients diagnosed with AD who had been randomly assigned to the placebo arm of a clinical trial and followed-up for six months [38]. Eighty per cent had hypertension at baseline, defined as a past history, treatment or raised blood pressure. After controlling for baseline disease severity, patients with AD and hypertension were more likely to have increased cognitive decline with an odds ratio of 1.6. Secondary analysis suggested this effect was confined to younger patients (below 65 years). Treatment with antihypertensive medication appeared to have no effect on the rate of cognitive decline in those patients with AD and hypertension [38].

It has been suggested that drugs blocking the RAS, both ACE-I and ARB, might be particularly effective at preventing cognitive decline in AD; ARB may have had beneficial effects on cognition in some studies in patients without AD [39]. In a small study, 15 patients with both AD and hypertension treated with ACE-I were compared with 47 patients, of whom 43% were hypertensive, who were not treated [40]. Over a six-month follow-up, patients receiving ACE-I had a slower rate of decline in digit forward and Instrumental Activities of Daily Living scale, and an improved measure of caregiver burden after adjusting for other risks factors [40].

A further study evaluated 686 patients with AD, of whom 75% had hypertension [41]. Sixty-one were continuous users of ACE-I, 59 used ACE-I intermittently, 189 were users of other antihypertensive drugs, and 309 had never used antihypertensive drugs. The four-year decline in MMSE was 6.4, 7.9, 8.8 and 10.2 respectively. In a subgroup analysis, the 118 participants who had continuously or intermittently used ACE-I had a significantly lower decline compared with the 498 who had never used ACE-I (7.5 versus 9.7; P = 0.03) [41].

A study in 224 patients associated a wide variety of drugs with progression measured as a change in the Global Deterioration Scale. About half of the patients were on heart and antihypertensive drugs. As part of the analysis they found a protective effect of ACE-I [42].

A large study used the US Veterans database to examine the hypothesis that inhibition of the RAS might have a specific effect on dementia and that ARB treatment might be more effective than ACE-I [39]. The authors looked at progression of dementia in those with AD at baseline, with progression defined as death or admission to a nursing home. Patients on an ARB, on the ACE-I lisinopril, and on cardiovascular comparators (excluding an ARB, ACE-I or statin) were compared; the proportion with hypertension in each group was 93%, 91% and 80% respectively. Compared with the cardiovascular comparator, ARB in patients with pre-existing AD were associated with a significantly lower risk of admission to a nursing home (0.51; 95% confidence interval, 0.36 to 0.72) and death (0.83; 95% confidence interval, 0.71 to 0.97). ARB exhibited a dose-response as well as additive effects in combination with ACE-I [39].

In the large UK-based General Practice Research database it was hypothesised that the rates of progression to hospitalisation or death would be lower for patients with AD treated with an ARB compared to patients on other antihypertensive drugs through the reduction of angiotensin II signalling [43]. In 3,905 patients with AD, neither mortality or hospitalisation rates with ARB were different from those in patients treated with other antihypertensive drugs. Unexpectedly, ACE-I were associated with a significantly higher mortality, but not with any increase in hospitalisation [43].

We could find no studies addressing the effect of diabetic therapy or control on outcome in AD, but one looked at possible protective effects of insulin therapy. Cognitive decline was compared between patients with mild-to-moderate AD and DM treated with insulin (n = 55) and those on oral hypoglycaemic agents alone (n = 49) [44]. At 12 months, mean MMSE decreased in those treated with oral hypoglycaemic agents (20.4 ± 4.1 versus 18.2 ± 3.6; P = 0.001), but remained stable in insulin-treated patients (21.9 ± 5.1 versus 21.2 ± 3.9; P = 1.03). No analysis of glucose control was performed and therefore it is unclear whether this difference related to better diabetic control or other effects of insulin [44].

We could find no studies examining the effect of statins exclusively in patients with AD and hyperlipidaemia but found three studies that evaluated the effects of statin treatment in patients with AD in which the diagnosis of hypercholesterolaemia was not present in all patients. Interpretation is complex because some studies addressed both concerns as to whether short-term statin therapy may impair cognition, whereas others looked at protective effects over longer periods of follow-up.

A three-year observational study followed 342 patients with AD (MMSE 21.3 at entry) [45]. Patients were classified into those with dyslipaemia and treated with lipid-lowering agents (n = 129; 47% with statins), those who had untreated hyperlipidaemia (n = 105), and those who were normolipidaemic (n = 108) [45]. Patients treated with lipid-lowering agents had a slower decline on the MMSE (1.5 point/year, P = 0.01) than patients with untreated dyslipidaemia (2.4 points/year) and normolipidaemic patients (2.6 points/year) [45].

A study in 224 patients associated a wide variety of drugs with progression in AD. Only 12 (5%) were on statins, but this group had less decline on the Global Deterioration Scale [42].

The US Food and Drug Administration has added safety warnings to statins concerning confusion and memory loss [46]. Initial evidence of such adverse events came from case reports describing subjective and reversible worsening of cognition in individuals using statins, although none of those reports included objective cognitive measures [47]. Also case series are reported in which patients with MCI or dementia had a significant improvement in their MMSE score when statins were discontinued [48]. In 12 patients with AD, the short-term effects of statin withdrawal were studied in a 12-week prospective non-blinded study involving a six-week withdrawal phase and a six-week challenge phase [48]. A specific aim was to address concerns that statins might be associated with short-term memory impairment. There was an improvement in MMSE scores (+1.9 [3.0], P = 0.014) with discontinuation of statins and a decrease in MMSE scores (-1.9 [2.7], P = 0.007) after re-challenge [48]. Two prospective studies showed minor declines in cognition of uncertain significance in adults with hyperlipidaemia treated with statins [49],[50].