Background

Coronaviruses are enveloped, positive-stranded RNA viruses that mainly cause respiratory and gastrointestinal tract infections [1]. They are divided into four genera: alpha, beta, delta, and gamma. Alphacoronavirus and betacoronavirus cause human infections [1]. Betacoronaviruses are further divided into 4 clades, a–d [2]. SARS-CoV and MERS-CoV are betacoronaviruses which caused outbreaks in 2002 and 2012 respectively [3]. The likely reservoirs of SARS-CoV and MERS-CoV viruses were identified as bats [2]. SARS-CoV-2 is a coronavirus and is classified into the betacoronavirus 2b lineage; however, a distinct clade from the SARS-CoV and MERS-CoV [4, 5]. It has been postulated that reservoir of SARS-CoV-2 is also bats; however, more evidence is needed for proving the assumption [6]. The disease caused by SARS-CoV-2 is termed as COVID-19. COVID-19 outbreak started as a cluster of respiratory illnesses and the first case was reported from Wuhan, Hubei Province, China on 8th December [7, 8]. It was declared as a pandemic by WHO on March 11, 2020 [9].

The most common symptoms of COVID-19 are similar to other coronaviruses which include fever, fatigue, dry cough, anorexia, shortness of breath, myalgia, and headache [10,11,12]. Old age and co-morbidities are associated with higher mortality and morbidity as compared with younger patients and those without any co-morbidities [10, 12, 13].

The neuroinvasive and neurotropic potential of coronaviruses like SARS-CoV and MERS-CoV has been demonstrated in many previous studies [14, 15]. A similar mechanism is suggested for the SARS-CoV-2 also [16]. Neurological manifestations reported of SARS-CoV, MERS-CoV, and other coronaviruses include peripheral neuropathy [17], myopathies with elevated creatinine kinase [17], large vessel stroke [18], olfactory neuropathy/anosmia [19], meningoencephalitis [20, 21], post-infectious acute disseminated encephalomyelitis [22, 23], Bickerstaff’s encephalitis overlapping with Guillain-Barré syndrome [24], and Guillain-Barré syndrome [24]. This review is aimed at discussing various neurological manifestations in COVID-19, including the frequency of neurological symptoms, morbidity, mortality, laboratory parameters, and imaging findings associated with patients with neurological symptoms. In the meta-analysis, we estimated the proportion of COVID-19 patients developing neurological manifestations.

Methods

Selection criteria and search strategy

We searched the following electronic databases for articles published between 1st December 2019 to 25th June 2020; PubMed, MEDLINE, Scopus, EMBASE, Google Scholar, EBSCO, Web of Science, Cochrane Library, WHO database, and ClinicalTrials.gov. The MeSH terms and keywords used include: “COVID-19” OR “COVID 19” OR “SARS-CoV-2” OR “2019 novel coronavirus” OR “2019 nCoV” AND “Neurological” OR “Brain” OR “CNS features” OR “central nervous system features” OR “peripheral nervous system features” OR “neuropathy” OR “skeletal muscle” OR “myositis” OR “neuromuscular junction” OR “headache” OR “anosmia” OR “olfactory” OR “ageusia” OR “cranial neuropathy” OR “seizures” OR “encephalitis” OR “meningitis” OR “stroke” OR “cerebrovascular disease” OR “cerebral hemorrhage” OR “intracerebral hemorrhage” OR “cerebral infarct” OR “cortical venous thrombosis” OR “deep cerebral venous thrombosis” OR “impaired consciousness” OR “confusion” OR “weakness” OR “Guillain-Barre’ syndrome” OR “Miller Fisher syndrome” OR “ataxia” OR “myopathy” OR “myelitis” OR “myelopathy” with an additional filter of “studies in human subjects”. The search was done between 31st March 2020 and 25th June 2020. To ensure literature saturation, we inspected the references of all studies included in this review. The protocol of this review was registered at PROSPERO (ID-CRD42020185593) prospectively in May 2020.

Inclusion and exclusion criteria

All published randomized controlled trials, non-randomized controlled trials, case-control studies, cohort studies, cross-sectional studies, case series, and case reports, if they had sufficient data on neurological features, laboratory parameters, imaging findings were included in this review. Only those studies were included in which subjects were diagnosed with SARS-CoV-2 infection by real-time RT-PCR or high throughput sequencing analysis of swab specimens or serology or culture. We also included pre-prints and letters if they included data on neurological manifestations in COVID-19. Editorials, systematic reviews, meta-analysis, narrative reviews, conference abstracts, commentaries, animal studies, post-mortem studies, and where translation into English was not possible were excluded. The authors were contacted twice by email if any missing data in the articles.

Data extraction and study quality assessment

Databases selected were searched independently by two members (TF and AP) in the team, and, following duplicate removal, reviewed all the articles and selected articles based on inclusion and exclusion criteria. Reporting was done according to the recommendations of the PRISMA statement [25]. Quality of the non-randomized studies was evaluated using the Newcastle-Ottawa Scale [26, 27] and the quality of one randomized controlled trial was assessed using the CONSORT criteria [28]. Any disagreements between two main reviewers were discussed with a third evaluator. Data about the author’s name, publication date, study setting and design, time and duration of the study, follow-up, the total number of patients evaluated, study population, age, gender, co-morbidities, neurological features, laboratory parameters, imaging findings, morbidity, and mortality were extracted.

Outcome measures

Primary outcomes assessed were neurological manifestations in COVID-19 patients and its prevalence. For the categorical variables, simple and relative frequency and proportions were used. For continuous variables, central tendency (mean or median) and dispersion measures (standard error, standard deviation) were used. To measure association, risk ratios, odds ratios, and hazard ratios were used and 95% confidence intervals calculated. We also assessed secondary outcomes like the association of neurological manifestations with age, co-morbidities, lab parameters including CSF study, imaging features, length of hospital stay, ICU admission, time from onset of typical COVID-19 symptoms to neurological manifestations, and morbidity/mortality.

Strategy for data synthesis, statistical analysis for meta-analysis

Data synthesis and illustration were done in tables and figures. For the categorical variables, simple and relative frequency and proportions were used. For continuous variables, measures of central tendency (mean or median) and dispersion (standard error, standard deviation) were calculated. The primary aim of our study was to synthesize the findings from multiple studies that investigated the issues related to neurological manifestations in COVID-19 and thus provide a quantitative summary, to better direct future work. The data are available in the form of proportions, defined as the number of cases of interest in a sample with a particular characteristic divided by the size of the sample. To achieve the objective of obtaining a more precise estimate of the overall proportion for a certain event (neurological manifestations) related to COVID-19, the study employed meta-analysis of proportions using a random-effects model and by the DerSimonian-Laird method [29, 30]. We performed data analysis using meta-packages in R (version 3.5.0). Heterogeneity was assessed using the I2 value. I2 can take values from 0% to 100% and it is assumed that an I2 of 25%, 50%, and 75% indicate low, medium, and large heterogeneity respectively [31]. Forest plot was used to visualize the point estimates of study effects and their confidence intervals. Publication bias was evaluated using the funnel plot.

Results

Among the 6789 articles identified, 212 studies were included in the systematic review and 74 studies in the meta-analysis (PRISMA flow diagram (Fig. 1)). Out of them, most were retrospective studies, 18 were cohort studies, 11 were prospective studies, nine were cross-sectional studies, one was a randomized controlled trial, one was a case-control study and the rest were all case series and case reports. Among these studies, we found only 19 studies, which investigated specifically neurological features in COVID-19 patients. Other studies, evaluated parameters in general. Table 1 shows a summary of all the observational studies included.

Fig. 1
figure 1

Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) flow diagram

Full size image
Table 1 Characteristics of studies included and neurological manifestations
Full size table

Neurological manifestations

Neurological manifestations have been reported in patients with COVID-19 from all over the world. A multicentre, retrospective study by Mao et al. [32] was the first study to evaluate the neurological manifestations in COVID-19 and found that neurological manifestations were present in 36.4% of total 214 patients, out of which most common was CNS manifestations(24.8%) followed by peripheral nervous system manifestations(8.9%). Other large retrospective observational studies reported the incidence of neurological manifestations as 4.3% [45], 15% [47], and 57.4% [49]. The most common neurological manifestations reported in COVID-19 were smell disturbances, taste disturbances, headache, myalgia, and disturbances in consciousness/altered mental status. The prevalence of all the neurological manifestations assessed is given in Table 2. A summary estimate of pooled prevalence and heterogeneity of each neurological manifestation are given in Table 3. Forest plot and funnel plot is given in Figs. 2 and 3 respectively.

Table 2 Prevalence of neurological manifestations reported from systematic assessment
Full size table
Table 3 Meta-analysis, summary estimate of pooled prevalence and heterogeneity of each neurological manifestations
Full size table
Fig. 2
figure 2

Forest plot of each neurological manifestations

Full size image
Fig. 3
figure 3

Funnel plot for assessing publication bias of each neurological manifestations studied

Full size image

Smell and taste disturbances

The overall incidence of smell disturbances in the studies ranged from 4.9–85.6% [49, 54] and the most common type of smell disturbance was anosmia. Other smell disturbances noticed were hyposmia, phantosmia, and parosmia [54]. Similarly, the incidence of taste disturbances reported was 0.3–88.8% [47, 54] and the most commonly reported were dysgeusia and ageusia. In the meta-analysis, we found 17 and 14 studies, which assessed the prevalence of smell and taste disturbances respectively and disturbances of smell (35.8%; 95%CI 21.4–50.2) and taste (38.5%; 95%CI 24.0–53.0) sensation were the most common neurological manifestation followed by non-specific neurological manifestations.

A case-control study of 79 COVID-19 patients and 40 historical controls of influenza patients from Spain [52] revealed that new-onset smell and taste disorders were significantly higher in the COVID-19 group. Patients in COVID-19 were significantly younger. Another study reported olfactory and taste disturbances occur more frequently in females than males [53]. Lechien et al. [54], Gilani S et al. [140], and Rachel Kaye et al. [141] reported that anosmia can be the initial and early manifestations of COVID-19. Population surveys on new-onset olfactory dysfunction from Iran [142] and UK [143] have reported an increase in olfactory dysfunction during the COVID-19 pandemic.

Non-specific symptoms

The most common non-specific neurological symptoms reported in SARS-CoV-2 infection were myalgia, headache syncope, and dizziness. The overall pooled prevalence estimate of the proportion of cases are given in Table 3. Incidence of myalgia reported in various studies ranged from 1.8–62.5% [47, 111], headache from 0.6–70.3% [90, 111], and dizziness from 0.6–21% [47, 113]. In children, myalgia and dizziness were less common and rarely reported. In health care workers, the incidence of myalgia, headache, and dizziness was higher compared with the general population. Syncope was reported in three studies with incidence of 0.3% [45], 0.6% [49], and 4.8% [86]. Few studies showed an increase in creatine kinase, LDH, and myoglobin in COVID-19 patients [12, 62, 66].

Acute cerebrovascular disease

Acute cerebrovascular disease (CVD) was reported in 0.5–5.9% [33, 48] of COVID-19 patients. Out of them, the most common type was acute ischaemic stroke and severe COVID-19 patients were more at risk of developing the acute CVD [33]. From these studies, the incidence of acute CVD in severe/ICU patients reported were 0.8–9.8% [33, 41]. The incidence of ischaemic stroke, hemorrhagic stroke, and cerebral venous thrombosis reported from various studies ranged from 0.4–4.9% [33, 48], 0.2–0.9% [38, 48], and 0.3–0.5% [33, 47] respectively. A study by Mao et al. [32] reported that two patients presented with hemiplegia without any typical COVID-19 symptoms. The median time to onset of cerebrovascular disease was 9 days. Another study by Li Y et.al [33] showed that acute CVD was more likely to be present with severe COVID-19; however, they were older, and had cardiovascular risk factors. These findings were similar to the above study by Mao et al. [32]. In both these studies, the laboratory parameters in patients with CNS symptoms were different from the other COVID-19 patients, with a higher white cell and neutrophil counts, reduced lymphocyte and platelet counts, elevated CRP and D-dimer levels [32, 33].

We found two studies that specifically studied the thrombotic complications in COVID-19 patients and found acute ischaemic stroke in COVID-19 patients receiving thromboprophylaxis [35, 36]. A retrospective observational case series in COVID-19 patients from Italy [144] reported six cases of stroke, four were ischaemic and two were hemorrhagic. Five of them had pre-existing vascular risk factors. Three patients with ischaemic stroke and one patient with hemorrhagic stroke showed hypercoagulable blood parameters [144]. Two studies reported six cases of stroke in young(< 50 years) COVID-19 patients, out of which three patients did not have any risk factors [145, 146]. Also there are multiple case reports and case series of ischaemic stroke including large artery [147], aneurysmal [148, 149] and non-aneurysmal SAH [51], deep cerebral venous thrombosis [150,151,152,153,154,155,156,157], hemorrhagic stroke [38, 158, 159] and CNS vasculitis [160] from all over the world in COVID-19 patients [38, 51, 147,148,149,150,151,152,153,154,155,156,157,158,159,160,161,162,163,164,165,166,167,168,169,170,171,172,173].

Meningoencephalitis, encephalopathy, disturbances in consciousness

Several cases of meningoencephalitis and encephalopathy were reported in COVID-19 patients [39, 43, 49, 174,175,176,177,178,179,180,181,182,183]. The incidence of encephalitis reported in two retrospective studies was 0.03% [43] and 0.1% [49]. Only in four of the 15 reported cases of encephalitis, CSF RT-PCR test was positive for SARS-CoV-2 RNA, and surprisingly two cases among them had negative nasopharyngeal swab [50, 174,175,176]. Two reports showed elevated levels of cytokines like IL-6, IL-8, TNF-α, β2-microglobulin, IP-10, MCP-1 in CSF [177, 181]. Interestingly, fluid from the surgical evacuation of subdural hematoma was positive for SARS-CoV-2 RT-PCR in a COVID-19 patient [184]. Isolated meningoencephalitis without any respiratory involvement has also been reported [175, 185]. Another case of rhombencephalitis as a rare complication of COVID-19 patient has been reported [186]. Few retrospective studies [32, 47, 49] reported seizures with the incidence ranging from 0.5–1.4% [32, 47]. Cases of all types of seizures like febrile seizures [42], focal seizures [180, 187,188,189], generalized tonic-clonic seizures [183, 190,191,192], myoclonic status epilepticus [193], status epilepticus [188, 194] and non-convulsive status epilepticus [46] were reported in COVID-19 patients.

Generally, the SARS-CoV-2 virus causes mild disease in children. However, a study from Italy showed a total five patients with seizures, and out of them, two had febrile seizures (three children had a known history of epilepsy, one child had a history of febrile seizures, one child had a first episode of febrile seizures) [42]. Also, a case of a 6-week-old infant with SARS-CoV-2 in addition to rhinovirus, presenting with brief 10–15-s episodes of upward gaze and bilateral leg stiffening was reported with normal EEG and MRI brain [195]. Another case of an 11-year-old child with COVID-19 viral encephalitis has been reported, with CSF showing viral encephalitis picture [194].

PRES syndrome has also been reported in studies [47, 51]. Transient cortical blindness like presentation of PRES syndrome with MRI brain at admission revealing bilateral T2/FLAIR hyperintensities, especially left occipital, frontal cortical white matter and splenium of the corpus callosum and diffusion restriction in DWI revealing vasogenic edema has been reported [196]. Repeat MRI after 2 weeks showed a complete resolution of findings. Cases of acute necrotizing hemorrhagic encephalopathy [191, 197], hypoxic brain injury with encephalopathy [43, 47, 51, 65], delayed post-hypoxic leukoencephalopathy [198], mild encephalitis/encephalopathy with a reversible splenial lesion(MERS) [199], ADEM in elderly females [200, 201], MS plaque exacerbation [47] and CIS [176] were reported in SARS-CoV-2 infected patients.

Incidence of disturbances of consciousness/delirium ranged from 3.3–19.6% [49, 100] in retrospective studies. S.R. Beach, et al. [202] reported four cases of elderly COVID-19 patients, who presented to the hospital with altered mental status without any respiratory complaints, and only one among them developed respiratory complaints during the hospital stay. Similar cases have been reported in elderly patients from Saudi Arabia [203], Norway [204] and China [205]. An observational case series from France [39] in 58 COVID-19 patients with ARDS admitted in ICU reported agitation in 40(69%) patients, confusion in 26 of 40 patients, diffuse corticospinal tract signs in 39 patients (67%) and out of the 45 patients discharged, 15(33%) had a dysexecutive syndrome. MRI Brain showed enhancement of leptomeningeal spaces in eight patients, bilateral frontotemporal hypoperfusion in 11 patients who underwent perfusion imaging, two asymptomatic patients with small acute ischaemic stroke and one patient with subacute ischaemic stroke.

Guillain-Barré syndrome

There are multiple reports of GBS in patients with confirmed COVID-19. GBS has also been reported to be a presenting feature in one case report by Zhao H et al. [206] where the patient, later on, developed fever and other symptoms of COVID-19. All the variants of GBS like AIDP, AMAN, AMSAN has been reported in COVID-19 patients [47, 206,207,208,209,210,211,212,213,214,215,216,217,218,219] including both para [206,207,208,209,210,211,212, 220,221,222,223] and post-infectious pattern [210, 211, 214,215,216,217,218,219, 224,225,226]. Toscano et al. [227] reported a series of five patients of COVID-19 with GBS, with the interval between the onset of fever, cough and symptoms of GBS ranging from 5 to 10 days. Cases of MFS were also reported [47, 226, 228, 229]. One case of MFS was associated with a positive serum GD1b-IgG antibody [228]. Other rare variants reported were GBS/MF overlap syndrome [219], AMSAN variants with severe autonomic neuropathy [219], facial diplegia [222, 227] and post-infectious pattern of the demyelinating type of GBS with brainstem and cervical leptomeningeal enhancement [225]. Cranial neuropathies with abnormal perineural or cranial nerve findings [230], multiple cranial neuropathies [211, 219], peripheral motor neuropathy [231] and ataxia [32, 43, 51] are all reported as presentations of COVID-19.

Other neurological manifestations

The incidence of rhabdomyolysis has been reported between 0.2–2.6% in different studies [11, 49, 139]. A report illustrates a 38 year-old COVID-19 patient presenting with fever, dyspnea, and severe myalgia, with high creatine kinase (>42,670 U/L) and LDH (4301 U/La) and was diagnosed as viral myositis [232]. Another two cases of adult COVID-19 patients with lower extremity pain and weakness with rhabdomyolysis with high creatine kinase and LDH were reported [233, 234]. First case developed rhabdomyolysis on the 9th day of admission [233] and 2nd case presented to the hospital with rhabdomyolysis [234]. An isolated case of post-infectious myelitis has been reported from Germany in a COVID-19 patient [235].

Three cases of generalized brainstem type of myoclonus were reported from Spain, with normal CSF study in one patient (others not done) and normal imaging findings. However, nasopharyngeal RT-PCR for SARS-CoV-2 was positive in only one patient. In all these patients, EEG was showing mild diffuse slowing without any epileptic activity [236]. Paresthesias [51] and cutaneous hyperaesthesia [237] were reported as a presentation in COVID-19 patients. A case of COVID-19 patient with oropharyngeal dysphagia followed by aspiration pneumonia, taste impairment, impaired pharyngolaryngeal sensation, and nasopharyngeal contractile dysfunction with absent gag reflex was reported from Japan [238]. Visual symptoms were also reported in a few studies. Mao L et al. [32] reported visual impairment in 1.4% of the COVID-19 patients. Cases of optic neuritis [49], isolated central retinal artery occlusion [239], non-arteritic type of posterior ischaemic optic neuropathy (PION) [240] as a COVID-19 manifestation were also reported. The summary of all the neurological manifestations reported in COVID-19 is given in Table 4.

Table 4 Summary of all the neurological manifestations of COVID-19
Full size table

Heterogeneity

The heterogeneity was high in most of the neurological manifestations studied except for hemorrhagic stroke (medium), cerebral venous thrombosis (low), seizure (low), and ataxia (low). The funnel plots were symmetric in hemorrhagic stroke, ataxia, seizures, cerebral venous thrombosis and myalgia, which is pointing towards no bias in the selection of publications that are included in the study. However, the funnel plots were asymmetric in other neurological manifestations studied, which pointed towards the heterogeneity in the studies undertaken or bias in the selection of publications included in the study.

Discussion

In this systematic review and meta-analysis, we assessed the neurological manifestations, risk factors, mortality, laboratory parameters, and imaging findings in those patients with neurological features. Involving 30,159 patients, our meta-analysis is the first and most comprehensive study about the neurological manifestations of COVID-19.

The most common neurological manifestations reported were smell and taste disturbances. Another interesting finding is the geographical variations in the frequency of smell and taste disturbances. High incidence of smell and taste disturbances were noted in studies from most of the European countries [54] while studies from Asian countries showed a lower incidence [32]. However, most of the studies which reported a higher incidence of smell and taste disturbances evaluated mainly olfactory and taste symptoms only and studied mild to moderate cases and excluded severe/ICU patients compared with studies with lower incidence. This bias might have caused under-reporting of smell and taste disturbances in severe/ICU patients or could also be because of decreased awareness of investigator about these symptoms at the beginning of the pandemic. Supporting our assumption, a study from Spain which evaluated 841 COVID-19 patients with neurological manifestations reported only 4.9% of cases of smell disturbances and 6.2% cases of taste disturbances [49]. Other possibilities for these variations are, the difference in affinity of SARS-CoV-2 to tissues between populations, a different strain of mutated virus circulating in Europe compared with Asian countries. However, more studies are required to confirm these assumptions. Interestingly a study by Wan Y et al. [241] predicted that binding affinity between 2019-nCoV and human ACE2 may be enhanced by a single N501T mutation. Also, ACE2 receptors are highly expressed by sustentacular cells of the olfactory epithelium. Olfactory and taste disorders were more common in younger patients [52, 140] most occurs in the early stages as initial manifestations of the disease and even as the only manifestation of COVID-19. Hence, olfactory and gustatory disorders can be the initial and early manifestations of COVID-19 and early identification of these symptoms might lead to early diagnosis and disease containment.

Non-specific neurological manifestations could be just systemic features of a viral infection. Similar to olfactory disturbances, the incidence of myalgia, headache, and dizziness also shows geographical variations with the highest incidence reported from Europe, the USA, and Chile. The incidence of non-specific symptoms was lower in children. We noticed that non-specific symptoms were higher among the studies conducted in health care workers. This may be due to increased knowledge and awareness of the symptoms and disease.

The most common type of acute CVD reported was an ischaemic stroke. Hemorrhagic stroke, deep cerebral venous thrombosis, SAH (both non-aneurysmal and aneurysmal), and TIA were also reported; however, with much lesser prevalence. Severe infection or ICU requirement, older age, cardiovascular risk factors, prior co-morbidities, and hypercoagulable lab parameters were found to be a risk factor for developing acute CVD [32, 33]. The apparent association of COVID-19 and stroke is likely due to the sharing of similar risk factors. The severity of COVID-19 has been proved to be directly related to the presence of co-morbidities like hypertension and DM. An earlier meta-analysis by Yang J et al. [242] comprising [46, 243] COVID-19 patients reported the prevalence of risk factors, hypertension in 21.1%, DM in 9.7%, and cardiovascular diseases in 8.4%. Also, hypercoagulable blood parameters as shown by Li Y et.al [33], can lead to ischaemic stroke and cerebral venous thrombosis. Nervous system involvement in SARS-CoV-2 infection can be due to direct invasion of neural tissues, inflammatory response, or immune dysregulation. The SARS-CoV-2 virus uses the ACE2 and TMPRSS2 for entry to the host cell and it is one of the main determinants of infectivity [241, 244]. Susceptibility to infection correlated with ACE2 expression in previous studies [245].

Very few retrospective studies showed meningoencephalitis as a presentation of COVID-19; however, there are multiple case reports from all over the world. The probable mechanism can again be direct invasion via the hematogenous route or retrograde pathway via peripheral nerve terminals. Two studies even showed higher levels of inflammatory cytokines in the CSF analysis of these patients [177, 181]. SARS-CoV-2 could trigger a seizure in predisposing patients through neurotropic mechanisms as explained earlier [188]. However, more evaluation is required in this field to find a temporal factor. All types of seizures were reported like febrile seizures, focal seizures, generalized tonic-clonic seizures, status epilepticus and myoclonic status epilepticus, NCSE and also brainstem type of myoclonus. Demyelinating disorders like ADEM, exacerbation of MS plaque, and the clinically isolated syndrome were all reported in COVID-19 patients.

Cases of GBS and its variants were also reported in COVID-19. Both post-infectious and pre-infectious pattern of GBS were reported. The most common type of GBS reported was AIDP. Other variants like AMAN, AMSAN, Miller Fisher syndrome, and facial diplegic variant were also reported. Patients presenting as GBS without any other typical symptoms of COVID-19 were also reported. Possible pathogenesis of GBS in COVID-19 includes immune dysregulation secondary to systemic hyper inflammation and cytokines produced as described by McGonagle et al. [246] and Quin et al. [247]. Hence, it is important to suspect and test for COVID-19 in those patients presenting with GBS and MFS. However, more studies are required to conclude that these cases were not just coincidental and COVID-19 itself is a trigger for GBS and MFS. GBS was also reported in other recent important viral infections like MERS-CoV [248] and Zika virus [243].

Change in laboratory parameters was also reported in COVID-19 patients with neurological manifestations like higher white cell and neutrophil counts, reduced lymphocyte and platelet counts, elevated CRP and D-dimer levels, and higher levels of creatine kinase, LDH, and myoglobin [12, 32, 33, 62].

High heterogeneity in our study could be because of differences in the selection of patients and ethnicity, the severity of the disease, co-morbidities, only a few studies evaluated neurological symptoms specifically, variation in the number of patients in different studies, or due to publication bias and differences in the methodology among the studies.

Comparison with previous systematic reviews

Earlier meta-analyses addressing general clinical features in COVID-19 were published. One such study showed myalgia in (28.5%; 95%CI 21.2–36.2), headache (14.0%; 95%CI 9.9–18.6), and dizziness (7.6%; 95%CI 0.0–23.5) [249]. Our results also found similar results for myalgia, headache, and dizziness, i.e. (19.3%; 95%CI 15.1–23.6), (14.7%; 95%CI 10.4–18.9), and (6.1%; 95%CI 3.1–9.2) respectively. Another similar meta-analysis also showed myalgia in (21.9%; 95%CI 17.7–26.4) and headache in (11.3%; 95%CI 8.9–14.0) [250]. One more study reported the prevalence of headache as (8.0%; 95%CI 5.7–10.2) [251]. However, no meta-analyses are published on the specific neurological manifestations till now.

Strengths and limitations

The strength of our study is that we did a comprehensive search in all the electronic databases. Study limitations include high heterogeneity in the estimation of the prevalence of some neurological manifestations, the inclusion of studies with very small sample size, and lack of meta-regression analysis. We excluded studies in languages other than English where translation was not possible. Most of the included studies were of moderate quality. More good- quality prospective cohort studies are required to establish that the neurological manifestations reported in the studies were not just coincidental.

Conclusions

In conclusion, our study showed neurological manifestations are common in COVID-19 and are even present as the only symptom without any other manifestation of the respiratory system involvement. Hence it is important to suspect every COVID-19 patient with neurological manifestations. In this pandemic, a neurologist needs to take necessary precautions while examining the patients presenting to them. Also, some symptoms like smell and taste disturbance can be used as a screening tool for SARS-CoV-2 infection and can help isolate suspected patients earlier to avoid the spread of the disease.