Background

The disease burden of dementia in Australia has been estimated to total 121,737 disability adjusted life years (DALYs), making dementia the fourth leading cause of burden from disease in this context.1 Alzheimer's disease is the most common form of dementia, representing 50% to 75% of cases.1 Due to the aging population, this increasing prevalence of dementia is not expected to decline without a major breakthrough in prevention. Another disease with a troublingly high prevalence is diabetes, which affects around 5% of the population.2 Diabetes has been strongly linked to the onset of dementia and Alzheimer's disease.3 A meta-analysis of 28 longitudinal studies demonstrated that people with diabetes had a 73% increased risk of developing dementia and a 56% increased risk of developing Alzheimer's disease compared to the general population.3 There is evidence supporting a number of different mechanisms that may underlie this association including: inflammation, oxidative stress, vascular effects (influencing the circulation of blood to the brain), increased cerebral [beta]-amyloid peptides, hyperinsulinemia, brain insulin resistance and the formation of advanced glycation end-products.4-6 Ultimately, it is likely that the cause is multifactorial, however, a clearly influential factor is elevated blood glucose, which occurs as a direct consequence of diabetes and has been shown to cause impaired episodic memory even in people who do not have diabetes.7

The biguanide metformin, a first-line antidiabetic drug for type 2 diabetes, acts as an insulin sensitizer and reduces blood glucose by increasing glucose uptake into muscles while reducing liver gluconeogenesis through the activation of AMP-activated protein kinase (AMPK).8 The seminal study supporting the contemporary popularity of metformin as an antidiabetic therapy came from the UK Prospective Diabetes Study Group (UKPDS) which carried out a multi-center, randomized controlled trial (RCT) with a median follow-up of over 10 years in overweight patients with type 2 diabetes.9 This study revealed that intensive blood glucose control with metformin resulted in clinically significant reductions in the incidence of diabetes-related endpoints (including both macro and microvascular complications such as myocardial infarction, stroke, amputation and blindness), diabetes and all-cause mortality. Since this study, the rate of metformin prescription and utilization by patients with diabetes has been steadily increasing.10,11 Although older people with newly treated incident type 2 diabetes are now prescribed metformin at a high rate, most older people with diabetes have had it for a long period of time. As such their medication usage is more likely to reflect past practices and consequently they are less likely to be receiving metformin than younger people.12,13

This difference in prescribing patterns may be unfortunate as metformin use has been associated with reduced rates of dementia14,15 and improved cognitive function16-18 in older people with diabetes. Interestingly, the association between metformin use and cognition was found to persist when success in diabetes management was controlled for, as well as when users of metformin were compared to people on other medications for diabetes management. This is suggestive that metformin could play a role in preventing dementia and Alzheimer's disease beyond its influence on blood glucose in patients with type 2 diabetes. Metformin has become known as a caloric restriction mimetic (CRM) as its effect activating AMPK is similar to that produced by caloric restriction.19 Both caloric restriction and metformin have been found to slow the aging process in animals and extend lifespan.20,21 Additionally, caloric restriction has been shown to improve memory in older people.22 Two clinical trials (MILES NCT02432287 and TAME) are underway to investigate the effect of metformin on aging in humans. If clinical studies uphold that metformin is an effective prophylactic or early intervention for Alzheimer's disease in older people who do not have diabetes, this finding could be rapidly translated into clinical practice due to metformin being a well characterized, commonly used drug with a wide safety margin. However, some evidence exists that metformin use could be detrimental to the cognitive health of older people with diabetes, with one study finding an increased rate of Alzheimer's disease23 and another finding lower cognitive function.24 Although controversy exists with respect to the relationship between metformin and vitamin B₁₂ levels,25 the study that identified lower cognitive function suggested that this could be due to vitamin B₁₂ deficiency which can result in cognitive impairment26 and has been hypothesized to be potentiated by metformin.27

The unresolved question as to whether treatment with metformin safeguards or harms cognitive health in older people impedes clinical practice in this area. Data that exists comes primarily from observational studies conducted over several years in multiple countries. One meta-analysis has been performed that investigated insulin sensitizers as a class compared to other forms of diabetes treatment, which included a subgroup analysis for metformin, and found a statistical trend (P = 0.06) which suggested that it did reduce the incidence of dementia.28 This meta-analysis by Ye et al.28 only examined incidence of dementia omitting continuous measures of cognitive ability. It also restricted eligibility to longitudinal studies, thereby excluding numerous potentially relevant studies. Importantly, this included Moore et al.24 and data from Imfeld et al.23 which suggested metformin could have a deleterious effect (Imfeld et al. was included, but for a different data-set). Searches of the Cochrane database, the JBI Database of Systematic Reviews and Implementation Reports, and PROSPERO did not find any further completed or pending reviews on this topic. As such, we propose to systematically review best available evidence on the effect of metformin on the risk, progression and severity of Alzheimer's disease and any other form of dementia, as well as its effect on any measures of cognitive ability or impairment.

Inclusion criteria

Types of participants

Any study on human participants will be eligible for inclusion. Although it is expected that the majority, of studies found will be on older people with diabetes, these are not essential elements to answer the review question.

Intervention/exposure

This review will include studies in which participants are receiving metformin at any dose for any duration. No restrictions will be applied based on what, if any, conditions the participants have which metformin is being administered to treat.

Control

Any study that includes a control group not being treated with metformin will be eligible for inclusion. This includes people not receiving any therapy as well as those receiving other treatments for the condition which metformin is being applied to treat (most likely other drugs or therapeutic strategies for glycemic control in people with diabetes). Where participants serve as their own controls (i.e. before-after study designs), they will be eligible for inclusion. If the control group are participants receiving one intervention for diabetes (i.e. insulin) and the intervention participants are receiving that intervention as well as metformin (i.e. insulin + metformin), the study will be eligible for inclusion.

Outcomes

This review will include studies that investigate risk (or incidence), severity or progression of dementia (including subtypes of dementia such as Alzheimer's disease or vascular dementia) as well as cognitive impairment or performance (whether investigated as a continuous measure or dichotomized using threshold values). Any criteria for the diagnosis of dementia will be accepted (including "self" [more likely carer] reports) as well as any measures of cognition. Examples include: the Mini-Mental State Examination (MMSE), Auditory Verbal Learning Test (AVLT), Clinical Dementia Rating (CDR) or Stroop Color Word Test (SCWT).

Types of studies

Any controlled experimental studies (i.e. randomized controlled trials, quasi-randomized trials, before-and-after studies) or observational study designs (prospective or retrospective, including cohort, case-control or cross-sectional studies) will be eligible for inclusion in this systematic review. Case reports and case series will be not be included.

Search strategy

A three-step search strategy will be utilized in this review. An initial limited search of MEDLINE (PubMed) and Embase will be undertaken followed by an analysis of the text words contained in the title and abstract, and of the index terms used to describe the article. A second search using all identified keywords and index terms will then be undertaken across all included databases. For published studies, this will include PubMed, Embase and PsycInfo; for gray literature, this will include the International Clinical Trials Registry Platform, ProQuest Dissertations and Theses Global, and OpenThesis. No date restrictions will be applied due to the relatively limited nature of the field and the high specificity that will be achievable in the search strategy. Only studies published in English will be included. Thirdly, the reference lists of all included reports and articles will be reviewed for additional studies.

Initial keywords will be: metformin, biguanide, dementia, Alzheimer's, cognitive, cognition, memory, MMSE, mini mental state examination. A full, example search string which will be applied in PubMed is provided in Appendix I.

Assessment of methodological quality

Papers selected for retrieval will be assessed by two independent reviewers for methodological validity prior to inclusion in the review using standardized critical appraisal instruments from the Joanna Briggs Institute.29 Any disagreements that arise between the reviewers will be resolved through discussion, or with a third reviewer.

Data extraction

Data will be extracted from papers included in the review using a preconstructed spreadsheet based on the standardized JBI data extraction tool.29 The data extracted will include specific details about the interventions, populations, study methods and outcomes of significance (including the relevant tools and units) to the review question and specific objectives. The authors of the included studies will be contacted if important data that is relevant to the review is missing in the papers.

Data synthesis

Quantitative data will, where possible be pooled in statistical meta-analysis using RevMan (v5.3.5) (Copenhagen: The Nordic Cochrane Centre, Cochrane). All results will be subject to double checking to ensure fidelity. Effect sizes expressed as odds ratio (for categorical data) and weighted mean differences (for continuous data) and their 95% confidence intervals will be calculated for analysis. Heterogeneity will be assessed statistically using the standard Chi-square and also explored using sensitivity analysis and subgroup analyses based any systematic differences between cohorts. Subgroups which will be considered for analysis as appropriate will include: metformin dose, metformin duration, gender, ethnicity/geographical region, diabetes type or severity, treatment or diabetic status of the control group, and age. Vitamin B₁₂ levels have been hypothesized to significantly interact with the effect of metformin on dementia and will be investigated where possible.

Where statistical pooling is not possible the findings will be presented in narrative form including tables and figures to aid in data presentation where appropriate.

Acknowledgements

This work is supported by a grant from the JJ Mason and H S Williams Foundation.

Appendix I: Search strategy (PubMed)

Intervention

Keywords

Metformin OR Biguanide OR Biguanides

MESH

Metformin OR Biguanides

Outcome

Keywords

Dementia OR Alzheimer's OR Alzheimer OR Cognitive OR Cognition OR Memory OR Recall OR MMSE OR "Mini mental state examination"

MESH

Cognition OR Dementia OR Cognition Disorders OR Cognitive dysfunction OR Alzheimer disease OR Mental recall

References