Description of the condition

Dementia is the most common neurodegenerative disease causing disability and mortality in older individuals (Jia 2014;McCarney 2003). The prevalence is 4% to 8% in people aged 65 and older in high‐income countries (Berr 2005; Lobo 2000), and is 3% to 7% in low‐ and middle‐income countries (Dong 2007; Jia 2014; Kalaria 2008; Llibre Rodriguez 2008; Wang 2000). Usually, mild cognitive impairment (MCI) is believed to be the prodromal stage of dementia and the early interventional stage.

Dementia affects memory, thinking, behaviour, emotion and social function. It comprises various types, such as Alzheimer's disease, vascular dementia, dementia with Lewy bodies, etc. Dementia impairs an individual's functional and social independence and response to participation in treatment programmes and rehabilitation (Shaji 2009). People with cognitive impairment often require continuous care and support, which places strains on caregivers and society. People living with dementia may have symptoms such as memory loss, difficulties in expressing themselves and understanding others, executive dysfunctions, emotional instability and personality changes (Chen 2015; Karkou 2014; Li 2014; Orimo 2008; Whitwell 2007).

So far, the available pharmacological treatments for dementia are unsatisfactory with limited effectiveness. Therefore, methods of non‐invasive electrical brain stimulation may be considered for dementia and MCI (Nardone 2012).

Description of the intervention

Modalities of this nature include repetitive transcranial magnetic stimulation (rTMS), theta burst stimulation (TBS), transcranial direct current stimulation (tDCS), cranial electrotherapy stimulation (CES, also known as transcranial alternating current stimulation, i.e. transcranial alternating current stimulation (tACS)) and reduced impedance non‐invasive cortical electrostimulation (RINCE) (O'Connell 2014). These modalities are increasingly used in clinical practice because they are safe and may prolong functional changes especially rTMS and tDCS (Ahmed 2012; Elsner 2016; Ferrucci 2008).

Non‐invasive brain stimulation techniques aim to induce an electrical stimulation in the brain in an attempt to treat neurological disease by directly altering brain activity. They require no surgical procedure and are therefore easier and safer to apply than invasive procedures (O'Connell 2014). There are various types of non‐invasive brain stimulation techniques, such as cranial electrotherapy stimulation, reduced impedance non‐invasive cortical electrostimulation, repetitive transcranial magnetic stimulation, theta‐burst stimulation, transcranial direct current stimulation, etc; each of them has distinct but different mechanisms (O'Connell 2014).

Cranial electrotherapy stimulation (CES) was initially developed as a treatment for anxiety and depression in the 1950s and later used as a treatment for pain and depression (Kavirajan 2014; Kirsch 2000). The electrical current in CES is commonly pulsed and applied via clip electrodes that are attached to the patient's earlobes. A small, pulsed electric current is applied across a patient's head to treat anxiety, depression, insomnia and chronic pain (O'Connell 2014).

Reduced impedance non‐invasive cortical electrostimulation (RINCE) has been tested as a treatment for chronic pain. An electrical current is applied via scalp electrodes but utilises specific stimulation frequencies, which is hypothesised to reduce electrical impedance from the tissues of the skin and skull, allowing deeper cortical penetration and modulation of lower‐frequency cortical activity (Hargrove 2012; O'Connell 2014).

Repetitive transcranial magnetic stimulation (rTMS) is a painless method used to stimulate small regions of the brain (Hao 2013). An electromagnetic coil is held against the forehead and short electromagnetic pulses are administered through the coil. The magnetic field thus is created and its pulse easily passes through the skull, causing small electrical currents that depolarise superficial axons and activate neural networks in the cortex (Lefaucheur 2014; Fregni 2007). Trains of these stimuli are applied to the target region of the cortex to induce alterations in brain activity both locally and in remote brain regions (Leo 2007). Nowadays, it can be applied in the treatment of various neurological and psychiatric disorders including migraine, stroke, Parkinson's disease, dystonia, tinnitus and depression (Cao 2014; Forogh 2014; Hoekstra 2013; O'Connell 2014; Pirio Richardson 2015; Serafini 2015; Zanjani 2015; ).

Theta‐burst stimulation (TBS) has been used to investigate long‐term potentiation (LTP) and long‐term depression (LTD) ‐like plasticity non‐invasively in the primary motor cortex (M1) in healthy humans and in people with various types of movement disorders (Huang 2005). It is patterned rTMS that modulates human M1 excitability using low intensity, repetitive bursts of magnetic stimuli that are similar to the mechanism of rTMS. However, TBS was shown to produce a more controllable, consistent, long‐lasting, and powerful effect on motor cortex in physiology and behaviour after a shorter application period compared with rTMS (Antal 2010; Huang 2005).

Transcranial direct current stimulation (tDCS) was originally developed to improve brain injuries such as stroke and recently has been further developed as a clinical tool for the modulation of brain activity. Administration of tDCS typically involves applying two large (25 cm² to 35 cm²) saline‐soaked sponge electrodes, consisting of an anode and a cathode, to the scalp. A weak constant current in the range of 1 to 2 mA then passes through the electrodes for several minutes, resulting in either facilitation or inhibition of spontaneous neuronal activity within the underlying cortex (Elsner 2013; Lefaucheur 2008). It is widely used currently as a therapeutic and rehabilitative tool for epilepsy, chronic pain, stroke, depression and Parkinson's disease (Hill 2015; Kang 2015; Summers 2016).

How the intervention might work

The aim of brain stimulation in the treatment of dementia is to modulate neuronal activity by altering activity in the areas of the brain involved in cognitive processing.

Both tDCS and rTMS have been shown to modulate brain activity specific to the site of application and the stimulation parameters (O'Connell 2014). As a general rule, low frequency rTMS (≤ 1 Hz) results in lowered cortical excitability at the site of stimulation, On the other hand, high frequency rTMS (≥ 5 Hz) results in raised cortical excitation (O'Connell 2014; Hallett 2007). Similarly, A‐tDCS, defined as positive (V+) stimulation, may increase the neuronal excitability of the area of stimulation, whereas C‐tDCS, defined as negative (V‐) stimulation, may decrease the neuronal excitability of the area of stimulation (Nitsche 2003). Meanwhile, the prolonged post‐treatment effects of rTMS and tDCS are probably due to synaptic and nonsynaptic mechanisms (Ardolino 2005; Lefaucheur 2014; Liebetanz 2002). Taking rTMS for instance, the duration of after‐effects increases with the number of stimuli delivered, and may persist for minutes to hours, or even days, after the end of an rTMS session (Lefaucheur 2014). From therapeutic and rehabilitative perspectives, the main interest of rTMS and tDCS resides in the persistence of clinical changes well beyond the time of stimulation (Cogiamanian 2007; Lefaucheur 2014; Nardone 2012). Used in this way, rTMS has confirmed the involvement of the dorsolateral prefrontal cortex (DLPFC) in various aspects of cognitive control. In some cases (e.g. memory), rTMS can interfere with task performance, inducing an increase in the number of errors (Miniussi 2008). If applied over an area that is causally engaged in that task being executed, short bursts of rTMS can interfere with function for brief periods and transiently change behavioural performance. This is sometimes referred to as on‐line interference (Ahmed 2012; Miniussi 2008). A possible effect on sentence comprehension is predicted on the basis of a study in young participants, who provide direct evidence of DLPFC involvement (Manenti 2008). In dementia cases (e.g. naming), even with comparable stimulation parameters, performance can be enhanced (Cappa 2002; Cotelli 2006; Cotelli 2008; Töpper 1998). Similarly, tDCS may also improve the cognitive performances in people with dementia, possibly by a similar mechanism of increased excitability (Fregni 2005; Sparing 2008). In people without dementia, an equivalent effect may be induced for more sophisticated cortical functions, such as memory or language (Boggio 2006; Ferrucci 2008; Fregni 2005; Sparing 2008).

Modulation of activity in brain networks is also proposed as the mechanism of action of TBS, CES and RINCE therapy, the therapeutic effects of which are primarily achieved by direct action upon the hypothalamus, limbic system or the reticular activating system (O'Connell 2014), or a combination of these. Although brain stimulation of this nature may be useful for several neuropsychiatric disorders, such as pain and post‐traumatic stress disorder (O'Connell 2014; Novakovic 2011), no information is available on cognitive changes induced by them in people with dementia. Further investigation is needed on this topic as well as on clarifying whether TBS, CES and RINCE may be associated with therapeutic change in people with dementia.

Why it is important to do this review

Dementia affects a sizable proportion of populations in low‐ and middle‐income countries, as well as high‐income countries, and there is a pressing need to find techniques to address the problem. The techniques may be new treatment options, since (at this early phase of research) they appear to be cost‐effective and low risk. This approach for dementia and MCI treatment is relatively novel. Data from small randomised controlled trials (RCTs) suggest that non‐invasive brain stimulation techniques are efficacious for many kinds of neurological diseases including dementia (Ahmed 2012; Cotelli 2006; Cotelli 2008; Ferrucci 2008). However, most studies with this observation are single‐centre RCTs with small samples. Another complicating factor in deriving inferences from the extant RCTs is that there are five kinds of stimulations and each type has different subtypes (Antal 2010; Hallett 2007). Most of these RCTs compare not only real and sham stimulations within the same types but also different subtypes.

We aim to extract data from RCTs and conduct a rigorous and comprehensive meta‐analysis, so that the evidence may be more accurately ascertained than in individual RCTs. Ultimately, such a meta‐analysis may be clinically useful when treating people with dementia and MCI.