Pioneers in this field were Rappaport and Young in 1990. They analyzed the effect of PEMF with a 27.1 MHz carrier modulated by 400 Hz pulse with 65 μs burst width and 585 W of peak power, on rats submitted to permanent MCAO (Rappaport and Young, 1990). No effect of PEMF on brain concentrations of sodium, potassium, and calcium could be detected. Brain water content was also evaluated, but even though it was significantly reduced in the treated group, changes were no consistent with the electrolyte content and water reduction was not consistent in all brain regions.

The research of Grant et al. (1994) is the most cited study in this field. These authors studied the effect of PEMF exposure in a rabbit model of ischemic stroke. Animals were submitted to 2 h occlusion of the left internal carotid, proximal left anterior cerebral, and proximal left middle cerebral arteries followed by 4 h of reperfusion (transient focal ischemia). The EMF stimulation was started 10 min after the onset of ischemia and maintained during 4 h of reperfusion. They use an EMF with 2.8 mT, 75 Hz, and pulse wide of 1.3 ms (Cadossi and Marazzi, 1988). Magnetic resonance imaging studies and histological examination were performed to assess the extent of cortical edema and ischemic neuronal damage, respectively. PEMF treatment resulted in a reduction of edema of approximately 65%. Histological examination showed a decrease of ischemic neuronal damage in the neocortical and striatum regions, but in general, the differences were not statistically significant.

A modulated PEMF (27.12 MHz carrier modulated by a 2 ms burst with 2 Hz frequency and signal amplitude of 3 ± 0.6 V/m) affected cytokine/chemokine gene expression in mice submitted to dMCAO (Pena-philippides et al., 2016). PEMF downregulated three proinflammatory and one dual action cytokine/chemokine but did not have clearly anti-inflammatory downregulation at 4 days of stroke. After 7 days, a significant downregulation of 19 genes was observed, most of them encoding proinflammatory factors. IL-1α and members of the TNF superfamily were downregulated, while IL-10 and IL-11 were upregulated; this shows that PEMFs provide a stronger suppression of inflammation at later stages of stroke. MRI imaging and TTC staining showed that this type of PEMF also reduced infarct volume and edema.

In a model of focal ischemia obtained by photothrombotic occlusion, PEMF showed an effect not only in gene expression and protein levels of inflammatory mediators but also in several apoptosis indicators (Urnukhsaikhan et al., 2017). Mice were exposed to 10 mT PEMF with a pulse frequency of 4 kHz and a burst frequency of 60 Hz and 15 ms burst wide. Treatment started within 30–40 min after ischemia induction and was maintained for 6 h/day, for up to 3 and 14 days. Infarction volume was significantly reduced by PEMF 14 days after occlusion. Treated mice had significantly longer Rotarod times compared to sham-exposed mice, indicating better motoric functions. At the molecular level BDNF, TrkB, phospho-Akt, phospho-Bad, and Bcl-xL in the peri-infarcted brain area were significantly increased compared to ischemic stroke mice without LF-PEMF treatment at 3rd and 14th day. Also, Bad, Bax, and Caspase-3 decreased in treated groups on the 3rd and 14th days. These data show that PEMF decreased apoptosis. Concerning inflammation, EMF significantly decreased levels of IL-1β and MMP9 in the peri-infarct area at 24 h and 3rd day of the experiment while IL-6β levels were not altered.

In a spontaneously hypertensive rat model of focal ischemia (induced by 90 min of MCAO) on the other hand, PEMF did not improve functional or histological outcome (Bates et al., 2012). Three different groups were submitted to diverse magnetic field application strategies: all treated groups were first submitted to 10 min of 8 mT/20 Hz PEMF with 0.3 ms pulse width, 1 h, and 24 h after reperfusion. Two groups had also an additional chronic exposure (6 days), of either 1 Hz or 6–9 Hz. The authors reported no significant effect on the behavioral tests of treated groups compared to controls and that all groups showed improvement in locomotor activity as a result of time. No significant differences were found in the ischemic volume, glial scarring, cell death, and microglial activation either. However, PEMF increased macrophage infiltration restricted to the ischemic lesion site at frequencies and intensity chosen in this research. The authors claimed that the non-beneficial or detrimental effects of this therapy may be associated with the low intensity of the PMF, its frequency, or treatment period. However, other studies used lower magnetic field intensities with beneficial effects on the ischemic model (Grant et al., 1994; Rauš et al., 2012; Rauš et al., 2013; Rauš et al., 2014).

In vitro, the effect of PEMF on glutamate-mediated excitotoxicity was assessed in mouse hippocampal cell line HT22 (Luo and Xu, 2017). Cells were exposed to 4 h of a PEMF of 0.92 mT, burst frequency of 15 Hz, burst width of 5 ms, and a 0.2 ms pulse width with 0.02 ms pulse wait (Wang et al., 2014). The authors used three different protocols: PEMF followed by 24 h glutamate exposition, PEMF simultaneously with 20 h glutamate, and, finally, PEMF followed by 24 h glutamate treatment with a second PEMF session 10 h after starting glutamate exposure. Results showed that PEMF reduced glutamate-induced excitotoxicity in all protocols by increasing cell viability and decreasing lactate dehydrogenase (LDH) release. The third protocol presented the best results. A deeper analysis demonstrated that PEMF exerts regulation of the glutamate-induced excitotoxicity through the endocannabinoid (eCB) system, as the beneficial effects of PEMF were suppressed by inhibition of the cannabinoid receptor CB1R. PEMF also increased the production of two types of eCBs (N-arachidonylethanolamide and 2-arachidonoylglycerol) with and without glutamate stimulation. Activation of extracellular signal-regulated kinases (ERK) signaling was an essential downstream mechanism for PEMF-mediated regulation of the eCB/CB1R. Besides, PEMF affected the eCB-related enzymatic system by increasing the enzymes involved in eCBs production while decreasing the enzymes involved in eCBs degradation. In presence of glutamate, this effect was diminished since only degrading enzymes were significantly downregulated by the PEMF. This indicates that the presence of ischemic stimuli (glutamate) changes the intracellular signaling cascades and thus the effects resulting from PEMF stimulation. This study presents a novel analysis of the mechanisms of actions related to PEMF; it also presents evidence of a prophylactic effect of PEMF. Nevertheless, future studies aiming to analyze the effect of PEMF after the glutamate exposure should be considered, in order to mimic the traditional stroke-therapy time course.

A series of in vitro studies have been done using a PEMF of 1.5 mT, 75 Hz (Cadossi et al., 1992), focusing on different cell types and ischemic models (Vincenzi et al., 2017, 2020; Gessi et al., 2019; Merighi et al., 2020). PEMF attenuated inflammation and hypoxia-induced injury using different cell lines (human neuroblastoma-derived SH-SY5Y cells, rat pheochromocytoma PC12 cells, and N9 microglial cells) (Vincenzi et al., 2017). PEMF decreased cell death after 6, 12, 24, and 48 h of incubation in hypoxic conditions (2% O₂). It also decreased apoptosis after hypoxia and the active caspase-3 levels after 12, 24, and 48 h of incubation in PC12 and SH-SY5Y cells. PEMF treatment also significantly reduced HIF-1α expression (25% with respect to hypoxia) in neuron-like cells and ROS production in both cell types. Microglial cells (in hypoxic conditions) also showed a decrease in ROS production after 2 and 48 h of PEMF exposure. In N9 microglial cells exposed to lipopolysaccharide (LPS, 0.1, 0.5, or 1 µg/ml), PEMF exposure (24 and 48 h) reduced the production of very important proinflammatory cytokines (TNF-α, IL-1β, IL-6, and IL-8). These results suggest a direct anti-inflammatory effect of PEMF, consistent with other in vivo experiments (Pena-Philippides et al., 2014; Urnukhsaikhan et al., 2017).

In a subsequent study, the protective effect of PEMF was studied in “nerve growth factor-” (NGF-) differentiated pheochromocytoma PC12 cells, injured with 48 h of hypoxia (Gessi et al., 2019). Similar to the study of Vincenzi et al., PEMF reduced cell death by 13%. These authors also found that PEMF exerted its protective effect by triggering p38 kinase phosphorylation and stimulating cytoprotective chaperone molecule HSP70. PEMF also increased CREB phosphorylation and restored BDNF basal levels. Finally, PEMF treatment decreased BAD proapoptotic regulation and increased antiapoptotic Bcl-2 molecule. The findings that the CREB/BDNF and BAD/Bcl-2 pathways are involved are in agreement with the in vivo studies of Urnukhsaikhan et al. (2017). Interestingly, Caspase-3 apoptotic pathways were not activated by PEMF in the PC12 cells while Urnukhsaikhan described its reduction in ischemic PEMF treated models. The study of Gessi et al. represents one of the most complete analyses of the PEMF mechanism of action in the apoptotic pathway. Similar studies should be performed in other cell types to know whether this mechanism can be generalized.

PEMF showed a protective effect on N9 microglia cells exposed to LPS-induced inflammatory conditions (Merighi et al., 2020). PEMF significantly reduced proinflammatory TNF-α, IL-1β, and IL-6 levels after LPS (1 µg/ml). Using a plethora of inhibitors, Merighi et al. found that inhibition of the c-Jun N-terminal kinase JNK1/2 reversed the protective effect mediated by PEMF on TNF-α and IL-1β production, suggesting that the protective effect of PEMF is mediated by the recruitment of JNK1/2 MAPK kinases. In the presence of LPS, PEMF significantly increased JNK1/2 phosphorylation starting from 20 min with a maximum effect at 30 min, thus confirming the role of JNK1/2 in the protective effect of PEMF. In addition, PEMF significantly reduced crucial cell functions specific to activated microglia like ROS generation (basal and LPS-stimulated), cell invasion, and phagocytosis, all essential processes involved in inflammation. It is worth mentioning that PEMF treatment by itself did not significantly affect the cell viability of microglia.

The study of Merighi et al. is in agreement with other research regarding EMF effect on inflammation and ROS production (Pena-Philippides et al., 2014; Urnukhsaikhan et al., 2017). However, some studies show that JNK1/2 activation induces an increase in proinflammatory cytokines secretion (Resnick and Fennell, 2004; Yamasaki et al., 2012). The exact role of JNK1/2 phosphorylation in ischemic conditions should be further studied to better understand its effect on cytokine production inhibition. These authors also speculated that the PEMF effect on IL-1β levels could be related to the blocking of diverse pathways, among them p38. Interestingly, exposure to the same type of EMF, in neuron-like cells under hypoxia conditions leaded to an increase of p38 kinase phosphorylation (Gessi et al., 2019). This suggests a possible differentiated effect of EMF depending on cell type. Further research of the effect of EMF in glial cells in hypoxia conditions and using primary cells instead of cell lines should be performed. Also, the use of other inflammation stimulators like IFN-γ should be considered because these proinflammatory cytokines rather than LPS are associated with central nervous system damage and disease (Gresa-Arribas et al., 2012; Liu et al., 2017; Lively and Schlichter, 2018).

Complementing the beneficial effects of PEMF induced in microglia, PEMF triggered protective actions in astrocytes (Vincenzi et al., 2020). PEMF exposure increased vascular endothelial growth factor (VEGF) released in 1321N1 astrocytes in a time dependent manner, with a maximum effect at 24 h (3.2-fold vs. normoxia). Furthermore, it was observed that the PEMF-mediated increase of VEGF was independent of the activation of the transcription factor HIF-1α (using a HIF-1α inhibitor chetomin). On the other hand, no effect was found on erythropoietin and “transforming grow factor-β1” (TGF-β1), which are important proteins secreted by astrocytes and which are interesting targets for stroke treatment (Dhandapani et al., 2003; Garzón et al., 2018). An interesting result from this study is that conditioned media derived from astrocytes subjected to PMEF increased viability and proliferation of SH-SY5Y cells in 48 h OGD conditions, thus suggesting the protective effect of PEMF-astrocyte culture media.

Many open questions concerning EMF stimulation and astrocytes remain. Some studies have been done showing that EMF affects viability and morphological properties (Khodarahmi et al., 2010), microvesicle motility (Golfert et al., 2001), and apoptosis (Liu et al., 2012) of normal astrocytes cell cultures subject to different types of magnetic field treatments. Therefore, this study of Vincenzi’s group represents the first one linking astrocytes, stroke, and EMF. However, it should be noticed that PEMF-stimulated astrocytes were not subjected to ischemic conditions. It seems logical to postulate that if hypoxia and EMF separately increase VEGF (Vincenzi et al., 2020), then their combined actions should lead to a greater increase in VEGF. However, previous results have demonstrated different effects of EMF in the presence or absence of an ischemic scenario (Rauš et al., 2013). In addition, the induction of VEGF by PEMF in astrocyte was HIF-1α-independent. It is important to point out that PEMF exposure did not modulate HIF-1α expression in 1321N1 astrocytes, while in SH-SY5Y neuron-like cells, it reduced the ischemia-induced HIF-1α (Vincenzi et al., 2017) suggesting a different effect, depending on the cell type. Therefore, more research on PEMF on astrocytes in ischemic conditions should be further studied.

Even though in Vincenzi’s study astrocyte cultured media did not affect viability in neuron cells under OGD conditions (Vincenzi et al., 2020), other authors have proven that OGD-astrocyte conditioned media alone could augment neuronal death (Jin et al., 2017) and decrease microglia cell death (Redzic et al., 2015). Thus, EMF experiments where culture media is collected in cell cultures subjected to OGD or other stimuli mimicking the ischemic processes should also be performed because all cell types of the neurovascular units are affected after a stroke.

Boland et al. (2002) designed a study to assess the effect of 50 Hz SEMF on rat hippocampal cells in the presence or absence of two substances that induce oxidative stress: Fe²⁺ and sodium nitroprusside (SNP) which is a NO donor. Different types of SEMF were used: 0–5 mT/50 Hz (called continuous) and an intermittent form using the same signal but with a 1 min on/off cycle. In both cases, a 48 h exposure time was used. Neither type of SEMF modified basal cell mortality. Furthermore, 0–5 mT/50 Hz had no significant effect on cell mortality in the presence of SNP or Fe²⁺. Intermittent EMF did not significantly modulate cell mortality in Fe²⁺. However, intensities of 2.5 and 5 mT potentiated cell mortality triggered by SNP. This suggested that EMF at these intensities affected NO-induced oxidative stress and may increase the ability of various toxic conditions to induce cell damage. Since SEMF treatment was carried out before and after oxidative stress induction, it is unclear whether the SEMF effect is due to cell preconditioning mechanisms.

Rauš et al. (2012), Rauš et al. (2013), and Rauš et al. (2014) developed a series of studies analyzing the effects of 0.5 mT/50 Hz sinusoidal extremely low-frequency magnetic field (ELF-MF) in the behavior, survival of neuronal cells, and oxidative stress in gerbils submitted to bilateral common carotid arteries occlusion. The locomotion activity of the ischemic and treated group was higher compared to the control group; however, 7-day exposure to ELF-MF increased motor activity in healthy gerbils but decreased ischemic induced motor activity (Rauš et al., 2012). These results are similar to those obtained by Bertolino et al. (2013) using the same animal model but SMF.

In another study using the same models and treatment, ELF-MF decreased the density of damaged neurons and increased the response of glial cells (astrocytes and microglia) in the hippocampus of ischemic gerbils (Rauš et al., 2013). Activated microglia are able to release both pro- as well as anti-inflammatory mediators during ischemic stroke, depending on the time and microenvironment. During the acute phase after stroke, microglia have a negative effect on stroke outcome, while these cells are essential for tissue regeneration in the post-acute phases (Guruswamy and ElAli, 2017). Rauš et al. (2013) stated that activated microglia were observed on the 7th day of reperfusion using Iba-1 marker. In addition, a decrease in neuronal death was observed, which may be an indicator of a possible effect of ELF-MF in microglia phenotype and/or its mediators. To understand the link between the neuroprotective effects of ELF-MF and astrocytes or microglia, the effect of EMF in glial cells submitted to ischemic conditions requires further research.

Ischemic process and ELF-MF exposure can both, independently, increase oxidative stress, by means of nitric oxide (NO) and superoxide (O₂ ⁻) production, lipidperoxidation (ILP) and superoxide dismutase (SOD) activity (Rauš et al., 2014). When measured 7 days after stroke, values reaching ELF-MF alone (in non-stroke animals) are even higher than those obtained in the ischemic group. However, combined with ischemia, ELF-MF showed an oxidative reducing effect. On the 14th day after reperfusion, oxidative stress parameters in the brain of these gerbils were mostly at the control levels. Other authors have shown that ELF-MF increased oxidative stress (Boland et al., 2002; Jelenković et al., 2006; Selaković et al., 2013); however, they were mostly associated with long term exposures.

Very low intensity and very low-frequency alternating electromagnetic fields were used in rats submitted to 90 min transient MCAO by Segal et al. (2016). The frequencies used in these studies were chosen according to theta and beta waves (3.93 and 15.72 Hz, respectively) and the magnetic field intensity was ±0.05 mT. In contrast to previous animal experiments, the first EMF exposure was performed 2 days after the onset of ischemia and persisted for 30 days with 2 min sessions daily. This is the largest therapeutic window used in the literature and the largest chronic evaluation of the effect of EMF. EMF had beneficial effects in increasing the recovery of sensory-motor functional deficits in treated groups as evaluated with neurological scores and forelimb placement. MRI studies demonstrated that EMF decreased edema and right ventricle volume. Fiber tracking studies showed an increase of fibers in the white matter system of fornix-fimbria and internal capsule when compared with the untreated group. In addition, immunohistochemistry showed an increase in neural generation by the presence of Dcx+ and nestin+ neural progenitor cells in treated animals. Although only three animals per group were used for MRI, histology, and fiber tracking, this study is a pioneer analyzing the effect of EMF on the white matter. However, because of the small sample size, these results should be interpreted very carefully.

ELF-MF attenuated OGD-induced cell death, calcium, and ROS intracellular levels in microglia HMO6 cell cultures (Duong and Kim, 2016). In this study, different combinations of frequencies and intensities were used (10, 50, or 100 Hz/1 mT and 50 Hz/0.01, 0.1, or 1 mT). The best results were obtained for the combination 50 Hz/1mT. A significant but less potent protective effect was also found at 10 Hz/1 mT. These authors also found that a specific inhibitor of xanthine oxidase inhibited OGD-induced ROS production and reversed the reduction in cell survival. However, no experiments were performed to know if the beneficial actions of ELF-MF are associated with this enzyme.

In a similar experiment, ELF-MF (50 Hz/1 mT) increased cell viability and decreased intracellular ROS/RNS in mesenchymal stem cells submitted to OGD conditions and 3 h ELF-MF exposure (Jung and Kim, 2017). The other combinations (10 Hz/0.01 mT, 10 Hz/0.1 mT, and 100 Hz/1 mT) did not show any protective effect. Exposure to N-acetylcysteine (NAC) and ethylenediaminetetraacetic acid (EDTA) increased the effects of ELF-MF in ROS/RNS scavenging and intracellular calcium decrease, respectively.

From both studies, we can conclude that the effect of SEMF depends on intensity and frequency. In these studies, relatively higher intensities and frequencies below 100 Hz have better effects. Also depending on the type of cell, some frequencies produce better results. Microglia respond better to 50 Hz and mesenchymal cells to 10 Hz, which could be an indicator for the possible use of combined frequencies therapies. Segal et al. (2016) also used two different frequencies targeting functional neural network by frequency rather than location, hence achieving clinical recovery and neuronal plasticity promotion.

ELF-MF improved survival, neurological output, and locomotor activity in rats with permanent bilateral common carotid artery occlusion at 24 h and 7 days after surgery (Font et al., 2019). Animals were divided into three groups: ischemic control treated with SEMF of 13.5 mT/60 Hz and treated with the same magnetic field intensity but 10 Hz. All ELF-MF regimens were given in 20 min sessions for 4 days (n = 13/group). On the 7th day after surgery, neurological impairment was significantly reduced in both 10 and 60 Hz compared to the control group. Infarction volume was also reduced by more than 50% in the 60 Hz treated groups. These authors also found that injections of the general NOS blocker L-NAME reversed the increased survival rates after ELF-MF. This suggests that the beneficial effects of ELF-MF are related to NO production. 60 Hz ELF-MF treatment also significantly increased NO levels in Human immortalized microvascular endothelial cells (HMEC- 1) 24 h after the treatment. NOS inhibitor L-NMMA was able to dose-dependently inhibit the NO production resulting from the ELF-MF exposure. This study confirms that NO production is one of the key mechanisms of the beneficial effect of ELF-MF after ischemic stroke. This is in line with previous findings that showed that L-NAME reduced the angiogenic effects of endothelial cells induced by pulsed ELF-MF (Li et al., 2015). Rauš et al. (2014) also found an increase in NO levels in their experiments but associated it with oxidative stress. Specification about the isoform of the NOS which is responsible for the NO production is needed in order to clarify the real role of the NO resulting for the EMF.

At in vitro and in vivo models, only Bates et al. (2012) have reported no beneficial effects of EMF (in this case PMF) after stroke. Of course, the lack of negative results in the literature can also be caused by the fact that negative data are hardly published (so-called negative publication bias). The study of Bates et al. is the only one reporting the use of animals with a comorbidity factor (hypertension). Investigating the effect of EMF in animals with other comorbidity factors present in patients (such as age or diabetes) is highly needed to understand the clinical potential of this treatment. In addition, the effect of EMF in hemorrhagic strokes is not described in the literature. Once the cause of hemorrhage is solved, the cellular and neuronal damage is similar to ischemic stroke (Dreier et al., 2006; MacLellan et al., 2010; Naranjo et al., 2013). Therefore, EMF as a treatment strategy should also be studied in this type of stroke.

A diversity of EMF therapies is proposed in the literature, using acute and chronic exposure and the results have been most beneficial. The acute effect of EMF is the most studied in the literature, showing an interest in the research community of using the EMF as a neuroprotective agent. However, some studies use a prolonged time of exposure (24 h for 7 days) that may be difficult to propose as therapy for clinical use.

Clinical studies reported in this field, such as preclinical studies, present a diversity of doses and treatments, but only PEMF has reached the clinical phase.

Capone et al. (2017) performed a dose-escalation exploratory study to assess the effect of PEMF in acute ischemic stroke patients (within 48 h after the onset of the stroke). Eligible patients for the study were those older than 18 years, with first mono-hemispheric ischemic stroke, and National Institutes of Health Stroke Scale (NIHSS) score greater than 4. Exclusion criteria were acute intracranial hemorrhage; previous ischemic or hemorrhagic stroke; the history of seizure; contraindications to magnetic fields exposure life expectancy of fewer than 3 months, and other serious illness or complex disease that may confound treatment assessment.

The same PEMF’s characteristics were used as in the studies of Grant et al. (1994) and Vincenzi et al. (2017) but with a magnetic field intensity of 1.8 mT. This small clinical trial consisted of 5 EMF sessions and a 12-month follow-up. Six patients were divided into two groups (n = 3) according to the dose-escalation scheme: 45 and 120 min. Clinical evaluation was assessed immediately after EMF treatment and reported at 30, 90, and 365 days later. Improvement of the clinical conditions was observed in all patients obtaining the best results for the longer exposure time. Also, for 120 min of exposure time, the lesion volume was reduced. Another important finding of this research was the no correlation between the intensity of the field and the volume of the lesion. No adverse events were observed in any group during PEMF exposure. Parameters like respiratory and heart rate, blood pressure, pulse oximetry, and ECG signal were stable during EMF stimulation. It is important to mention that Capone et al. (2017) wanted to include a treatment of 240 min; however, no patient accepted this exposure time because it was considered too long and potentially interfering with the standard of care. This proves our previously mentioned concern on elevated exposure times in some treatments proposed in preclinical studies.

The group of Cichón developed several studies to explore the effect of PEMF regarding brain plasticity and oxidative stress in patients (Cichoń et al., 2017a; Cichoń et al., 2017b; Cichoń et al., 2018a; Cichoń et al., 2018b; Cichoń et al., 2018c; Cichoń et al., 2019). PEMF was not applied on the brain but on the pelvic girdle. 40 Hz PEMF with a rectangular and bipolar waveform was applied. This type of waveform is different from those used in the aforementioned studies where pulses are thinner and without a negative phase.

One of the studies (Cichoń et al., 2017a) used 57 post-ischemic stroke patients with moderate stroke severity. The patients were divided into two groups: EMF (n = 23) and shame PEMF exposure (n = 34) with ages between 50 and 90 years. PEMF was applied at 7mT/40 Hz for 15 min per 10 days. Both groups underwent the same 4-week rehabilitation program. Patients with a medical history of pre-stroke dementia, hemorrhagic stroke, or inflammatory factors were left out of the experiment. No drug with anti-oxidative properties was supplied to the patients. PEMF increased catalase (CAT) and superoxide dismutase (SOD) activity in hemolysate. However, low molecular weight antioxidants in plasma showed no significant difference between groups. Functional status, depression, and cognitive impairment were assessed by Bartel Index of Activities of Daily Living (ADL), Geriatric Depression Scale (GDS), and Mini-Mental State Examination (MMSE), respectively. PEMF improved functional status in about 20% and cognitive function in about 40% of the patients. PEMF also decreased depression syndrome in 60%.

In another study (Cichoń et al., 2017b) with 7 mT/40 Hz PEMF, similar clinical results were obtained. In this trial, 48 patients were enrolled; 25 of them were submitted to 15 min of PEMF during 10 sessions. Besides the functional status, cognitive impairment, and depression severity, 3-nitrotyrosine (3-NT), nitrate/nitrite estimation, and tumor necrosis factor (TNFα) were measured in plasma. Nitric oxide synthase 2 (NOS2/iNOS) was measured in whole blood samples. An increase of 3-NT in plasma was observed in 68% of the patients in the treated group versus 17% presented by the sham-exposed group. Nitrate/nitrite concentration was also elevated in the treated group. On the other hand, iNOS and TNFα did not show a significant difference between groups. The authors demonstrated that EMF increased NO generation and its metabolites improving post-strokes treatment.

The effect of PEMF (5 mT/40 Hz) in plasma oxidative stress markers was also assessed (Cichoń et al., 2018c). A statistically significant decreased level of oxidative stress parameters such as thiobarbituric acid reactive substances (TBARS), thiol groups, and carbonyl groups was found in the treated patients after 10 and 20 PEMF treatments. The reduction of oxidative stress markers was significantly greater when the number of treatments was increased. Along with these biochemical markers, the clinical outcome of the patients was evaluated using the same criteria as in Cichoń et al. (2017a). PEMF-treated patients showed an improvement in their functional status. A correlation between the change in these clinical tests and the change in carbonyl groups showed a linear relationship between these parameters. Pointing this to a reduction in plasma oxidative damage markers related to the improvement in independence, cognitive functions, and degree of depression of stroke patients (Cichoń et al., 2018c).

PEMF therapy (5 mT/40 Hz) after 10 sessions also increased the expression of genes involved with the antioxidant defense system (Cichoń et al., 2018b). CAT, SOD1, SOD2, and glutathione peroxidase GPx1 and GPx4 mRNA levels were increased by more than 100% in the treated group (n = 25) with respect to the control group (n = 23) in whole blood samples. This is in accordance with the previous work of Cichoń et al. (2017a) where EMF increased levels of SOD and CAT activity in erythrocytes.

PEMF has an effect on the growth factors involved in the neuroplasticity process during the stroke (Cichoń et al., 2018a). Treatment consisting of 10 sessions of EMF increased the plasma level of BDNF and VEGF by about 200 and 50%, respectively. BDNF mRNA expression increased by 195%. Plasma level of five different cytokines was measured obtaining an increase in hepatocyte growth factor (HGF) and stem cell factor (SCF) by 35 and 25%, respectively. Stromal cell-derived factor 1 (SDF-1α), β-nerve growth factor (β-NGF), and leukemia inhibitory factor (LIF) plasma level were not significantly different between the treated and control group. In this study, the clinical status was also evaluated. As in previous studies (Cichoń et al., 2017a; Cichoń et al., 2017b; Cichoń et al., 2018c), an increase in the functional status and cognitive function and a decrease in depression syndrome and also stroke-related neurologic deficits were observed.

In a more recent study, PEMF treatment increased plasma levels of IL-1β and IL-2, as well as IL-1β mRNA expression in whole blood samples (Cichoń et al., 2019). However, INF-γ and TGF-β plasma levels remained the same. Experimental groups and treatment were the same used by Cichón et al. in previous studies (Cichoń et al., 2017b; Cichoń et al., 2018a; Cichoń et al., 2018b).

Cichón et al. suggested that the increase of NO levels obtained in their research can be associated with nNOS and/or eNOS activities, but not with iNOS expression. They also claimed that their results are consistent with evidence shown by Cho et al. (2012) who proved that EMF increased NO in rat brains. However, Cichón et al. evaluated NO in plasma, while Cho et al. (2012) measured it in the brain. Also, time exposures of EMF were different and Cho et al. did not use stroke rodents. Moreover, Rauš et al. (2014) proved that ELF-MF has a different effect on healthy and ischemic rats with an immediate EMF exposure. Exposure time, a moment of exposure, besides and EMF characteristics are important parameters that must be evaluated comparatively. However, these results are in line with the findings of Font et al. (2019) in which NO-inhibition abrogated the beneficial effects of EMF in ischemic rats.

Clinical increase of BDNF levels in plasma agreed to the findings of Urnukhsaikhan et al. (2017) in mice. BDNF is an activator of various signaling pathways involved in the regulation of neurogenesis and the survival of neurons. Lower levels of BDNF are correlated with an increased risk of stroke, worse functional outcomes, and higher mortality (Kotlega et al., 2017). This protein upregulation can improve post-stroke outcomes and the efficacy of the rehabilitation process within a physical exercise, as can be seen in Cichón et al.’s clinical assessment results. As with CAT and SOD, an increase in BDNF levels is related to gene overexpression.

On the other hand, VEGF protein not only has a pro-angiogenic action but also has a neurotrophic and neuroprotective effect, on both the central and peripheral nervous systems. VEGF generated by ependymal cells activates and enhances neuronal precursor proliferation and growth. These results are compatible with the results of Delle Monache et al. (2008). They found an increase in VEGF receptor 2 (KDR/Flk-1) in cultured umbilical human vein endothelial cells using 1 mT/50 Hz suggesting that the EMF has an angiogenic effect. This is in contrast with another publication of the same authors showing the anti-angiogenic effect of EMF at 2 mT and the same frequency (Delle Monache et al., 2013). The magnetic field intensity used by Cichoń et al. (2017a) is 5 mT, close to the inhibitory induction; however, the conditions of the experiment are not the same, cell lines instead of stroke patients and no ischemic environment in the study of Della Monache et al.

IL-2 is a proinflammatory cytokine (Nayak et al., 2012), but a dual role has been recognized by some authors (Shachar and Karin, 2013; Pena-Philippides et al., 2014). High plasma levels of this cytokine were found in patients in the first 24 h after stroke, which decreased during the next 6 days (Nayak et al., 2012). However, Cichón et al. study reported an increase of IL-2 in the treated group, coinciding with this a good clinical outcome; this result agrees with the EMF effect reported by Pena-Philippides et al. (2014) in a mouse model. On the other hand, the increase of IL-1β reported by Cichón et al. is in contrast with several other studies where a decrease was reported (Urnukhsaikhan et al., 2017; Vincenzi et al., 2017; Merighi et al., 2020). The authors refer to that a neuroprotective role of IL-1β might be attributable to the IL-1β-dependent regulation of neurotrophic factors (Cichoń et al., 2019), specifically a positive correlation between IL-1β and BDNF levels. However, in Urnukhsaikhan et al. (2017), both of these indicators’ levels were inversely proportional. It is worthy to highlight that all in vitro and preclinical cytokines assessments were performed in the acute, or early sub-acute (3–4 days) stroke phases, whereas the clinical studies of Cichón et al. were made 3 weeks after stroke (Pena-Philippides et al., 2014; Urnukhsaikhan et al., 2017; Varani et al., 2017; Merighi et al., 2020). Inflammation can be both harmful and protective depending on which particular stage after a stroke. While it can contribute to the expansion of the infarct, it is also responsible for infarct resolution and influences remodeling and repair (Lambertsen et al., 2019), thus further research in both acute and chronic stages of stroke is highly necessary.

The findings of Cichón et al. are a result of a PEMF applied with a pelvic griddle contrary to the study performed by Capone et al. where ELF-MF was applied on the head. In this case, Cichón et al. explained that exposure to the head of ELF-MF can affect the activation of the epileptic focus, but no reference to any study proving this is given by the authors. Moreover, Capone et al. performed an experiment to prove the safety of the ELF-MF showing that the treatment does not produce any side effect in humans when placed onto the skull (Capone et al., 2017).

The studies of Cichón et al. are the only ones that used a localized EMF exposure far from the brain, with a biochemical and molecular mechanism elucidated. The improvement obtained may be an indicator of a more systemic effect of EMF, in a rehabilitation phase, but as a neuroprotective agent, this localization may not be effective.

In conclusion, at a clinical level, only a small number of studies were found in the literature. Similar to the in vitro and in vivo experiments, researchers use different EMF characteristics. Most studies used EMF as a therapeutic agent in a chronic phase of the disease. Only one study reported a 48 h therapeutic intervention (Capone et al., 2017). No side effects are described in any experiments and a 12-month follow-up shows a long-term effect of EMF.