Introduction
Autism spectrum disorder (ASD) is a group of neurodevelopmental disorders characterized by social communication impairment and restricted interests/repetitive behaviors, which mainly occur in childhood and affect their whole life [
1]. A recent systematic analysis of the global status of mental disorders showed that the global age-standardized prevalence of ASD was 369.4 cases per 100,000 people in 2019, with a significant male predominance [
2]. ASD is often an overwhelming experience for parents, placing a heavy burden on families and society, which is reflected in both mental and economic burdens [
3,
4]. A Previous study have shown that the economic cost of ASD mainly stems from special education and parental productivity loss in childhood and supportive living accommodation and individual production loss in adults [
4]. In addition, medical costs are much higher for adults with ASD than for children. In the United States and the United Kingdom, the cost of supporting autistic patients without intellectual disabilities is approximately $ 1.4 million, while the cost of supporting autistic patients with intellectual disabilities is higher, reaching $ 2.4 million [
5]. Although the pathogenesis of ASD remains poorly understood, current research suggests that it may be related to both genetic inheritance [
6] and environmental factors [
7]. The cellular etiology of ASD includes abnormalities of one or more developmental events, including neurogenesis, neuronal migration, axonal projection, dendritic development, synaptogenesis and synaptic remodeling [
8]. Among them, the pioneering research of Rubenstein and Merzenich shows that the imbalance between excitatory and inhibitory neurosignaling (E/I imbalance) is a potential neuropathophysiology of autism, which is mainly due to the enhancement of glutamate (excitatory) signal pathway, or the reduction of inhibition caused by the weakening of GABAergic signal pathway [
9]. The migration of neurons in the cortex is mainly controlled by the paracrine action of neurotransmitters glutamate and GABA [
10], E/I imbalance can lead to abnormal differentiation and migration of neuron, delayed synapse maturation or abnormal myelination [
9]. The defect of neuronal migration also leads to the mis-localization of affected neurons [
8], and the consequent malformations and malfunctions of various brain circuits are considered to be the important causes of various neurodevelopmental disorders, including epilepsy [
11,
12], intellectual impairment [
13] and ASD [
12,
14,
15]. Abnormal changes in the levels of GABA receptors and GAD65/67 in brain samples of austic patients were also confirmed in animal models [
16‐
21]. Defects in either the production or migration of GABAergic neurons in the cortex led to the decrease of the number of GABAergic neurons in the cortex, resulting in the over-excitation of the cortex, thus affecting the synaptic plasticity process [
22]. In addition, neural plasticity can also be mediated by secondary neurotransmitters transmitted by subcortical nucleus, such as serotonin, acetylcholine, dopamine [
23‐
26]. In short, the important theory of E/I imbalance provides direction and basis for the later research on the treatment of ASD. Early psychological/behavioral intervention, complementary therapy, or drug treatment can improve some symptoms of autism, such as reducing anxiety and aggression, reducing the combined symptoms of attention deficit and hyperactivity, and promoting emotional communication [
1]. However, these treatments cannot completely correct the core symptoms of ASD, such as social communication impairment, which leads to many problems, such as poor patient tolerance, adverse reactions, and economic burdens [
27]. Therefore, novel therapeutic methods are required.
In the past decade, neuromodulation techniques have made impressive progress in the treatment of neuropsychiatric diseases [
28,
29]. Deep brain stimulation (DBS) has been widely clinically applied to effectively improve the symptoms of Parkinson's disease, such as tremor, slowed movement, and rigid muscles [
30]. Preclinical studies have shown that DBS in the prefrontal cortex, hypothalamic nucleus, and central thalamus can alleviate VPA (Valproate acid) -induced autism-like behaviors [
31‐
33]. Vagus nerve stimulation (VNS) is an FDA (Food and Drug Administration) -approved therapy to reduce the severity of intractable epilepsy and depression, which has become a potential adjuvant therapy for patients with autism [
34]. Dysregulated parasympathetic system and reduced vagal tone are frequently observed in ASDs, which is related to autistic behavioral and language disorders [
35,
36]. VNS treatment has been proven to overcome the problem of an insufficient vagal response [
37], suggesting that VNS treatment may be beneficial for neurodevelopmental disorders with altered parasympathetic activity. In addition, research on the use of VNS in children with epilepsy and ASD has achieved positive results [
38]. In recent years, promising results in adult neuropsychiatric diseases have promoted active research on noninvasive neuromodulation techniques in children and adolescents, particularly transcranial magnetic stimulation (TMS) and transcranial direct current stimulation (tDCS) [
39]. The stimulation frequency and duration of stimulation period determine the effects caused by rTMS [
40]. Low-frequency stimulation (< 1 Hz) has inhibitory effects, whereas high-frequency stimulation (> 5 Hz) leads to excitatory effects in the brain [
41]. In addition, low intensity (less than motor threshold) tends to reduce cortical excitability, while high intensity (greater than motor threshold) will increase cortical excitability [
42]. Early research of ASD mainly focused on the conventional low-frequency (1 Hz) repetitive TMS (rTMS) in the prefrontal cortex, and proved its positive impact on the behavior and electrophysiological results of children with ASD [
43,
44]. In recent years, high-frequency theta-burst stimulation protocols to the motor cortex was also tested in children and adults [
45,
46]. Previous studies have shown that rTMS can affect cortical excitability by altering the glutamatergic (Glx) or γ-aminobutyric acid (GABA) metabolite levels [
47], and reducing irritability, hyperactivity, and stereotyped behavior in autistic children [
43,
48,
49]. In addition, it also has a positive impact on children’s error monitoring and correction function [
50]. Transcranial direct current stimulation (tDCS) is considered to be a promising method for the rehabilitation of neurodevelopmental disorders [
51]. In humans, the fast oscillation of electroencephalogram (EEG) activity occurs in the frequency range of 30-120 Hz, which is called “gamma” frequency band [
52]. Several studies showed that gamma band response (30-80 Hz) is very dependent on E/I signal transduction [
53‐
55], mediating a series of basic neural functions, including sensorimotor integration, perceptual integration, working memory, network synchronization and higher-order cognition [
52,
56‐
58], which is obviously disturbed in multiple systems of ASD [
59‐
61]. A recent EEG study for adults found that GABA was coupled with gamma band response, and reported a positive correlation between GABA level and gamma band power [
62]. tDCS uses constant weak current to induce bidirectional and polar dependent changes in the cortex, which has been proved to regulate gamma oscillation and excitability in the cortex [
63,
64]. Recent clinical research shown that the tDCS in the prefrontal cortex, motor cortex and temporoparietal junction regions can effectively improve the social function and abilities of the theory of mind in patients with ASD [
65‐
68].
Neuromodulation techniques are rapidly changing the field of neuroscience, especially in recent years, and research on ASD has surged, showing that it is necessary to conduct a comprehensive analysis of the current situation and trends. Bibliometrics is widely used to analyze the published literature, and is an effective way to understand the development of the discipline. In recent years, bibliometric analysis has been conducted on the status and research trends of brain tumors [
69], neurodegenerative diseases [
70], and neuropsychiatric diseases [
71]. However, to the best of our knowledge, there has been no bibliometric analysis investigating the application of neuromodulation techniques in ASD. Therefore, in the present study, we analyzed the global research status of neuromodulation techniques for the treatment of ASD from 1992 to October 2022 using bibliometrics and literature visualization tools. The results are presented in the form of a visual map to further analyze the research hotspots, future trends, and application prospects of neuromodulation techniques in ASD.
Discussion
A previous bibliometric study found that research on ASD in the past 20 years mainly focused on its etiology, including brain structure, images, psychology, genes and animal models [
103]. From 1998 to 2009, the classification and research of ASD and its subtypes have made great progress [
104]. In particular, the Diagnostic and Statistical Manual of Mental Disorders (Fourth edition) (DSM-IV) use “pervasive developmental disorder” to name ASD, which including five subtypes: Rett Syndrome, Asperger Syndrome, Childhood Disintegrative Syndrome and Autistic Disorder. In addition, the etiology research of ASD is mainly focused on developmental neuroimaging and genetics during this period [
101]. Many key brain regions have been proved to be related to the occurrence and development of autism, such as the medial prefrontal cortex, temporal-parietal junction [
103], which provide potential options for the future treatment of ASD. From 2010 to 2018, with the promotion of the fifth edition of ASD diagnostic criteria in 2013, the research on ASD subtypes has been reduced to a certain extent, and chromosome and gene detection technologies have been widely used in ASD, which promoted the development of cytogenetic diagnostic tests for developmental delay/intellectual disability, ASD, or multiple congenital anomalies [
101]. In addition, a large number of autism-related genes or risk genes have been found [
104‐
106], such as PTCHD1 gene mutation on Xp22.11 [
104]. However, some researchers believed that these gene mutations are lack of specificity and may widely exist in many mental diseases, so further exploration is needed [
107]. In any case, the results of gene research provide a certain direction for the diagnosis and treatment of ASD or other mental diseases. In recent years, related research gradually turned to the comorbidities of ASD. ASD patients are often accompanied by comorbid symptoms such as anxiety, depression, attention deficit hyperactivity disorder and self-injurious behavior, etc., which bring serious burden to families and society. During the COVID-19 epidemic, home isolation may bring more negative effects to ASD children and their parents, such as more serious and frequent destructive behaviors in children and the depression and anxiety of parents. A large number of studies have confirmed that patients with ASD also show many digestive system symptoms, such as abdominal distension, abdominal pain and diarrhea, which may be related to disorders of intestinal flora [
108‐
110]. Abnormal intestinal metabolites interfere with brain function and behavior through the microbiota, including metabonomic pathway [
111] and vagal pathway [
112]. The research on the treatment of ASD has been continuous in the past 20 years. The treatment of ASD is still based on education and behavior training, such as the applied behaviour analysis (ABA). However, in recent years, candidate drugs for the treatment of ASD have also brought hope. For example, the application of bumetanide, cannabis diphenol and oxytocin improves the severity of ASD symptoms and combined symptoms, including repeated patterns of behavior, interest or activity [
113‐
115]. In addition, complementary and alternative treatments (CAT) has also received a lot of attention and has shown a positive role, such as dietary supplement and music therapy [
116‐
118]. In the past 20 years, with the development of medical physics and medical physiology, based on the early research of etiology in ASD patients, neuromodulation techniques has come to the fore and has been widely used in preclinical and clinical research of treatment for ASD.
To the best of our knowledge, this is the first study to use VOSviewer and CiteSpace to review the progress of research on neuromodulation techniques as a treatment for ASD. To the best of our knowledge, this is the first study to use VOSviewer and CiteSpace to review the progress of research on neuromodulation techniques as a treatment for ASD. As such, we explore the characteristics of the publications from 1992 to October 2022, and our results reveal the current research focus and future prospects. Since the first study was published in 1992, the number of publications in this field has continued to grow. In particular, since 2014, research achievements in this field have surged, which may be due to the following factors: first, the diagnostic criteria of “Autism Spectrum Disorder” in the Diagnostic and Statistical Manual of Mental Disorders (fifth edition) (DSM-5) released in 2013 were more extensive, which may have increased the diagnostic rate of ASD [
72,
73]. Second, this may be related to the release of the European Expert Group’s guidelines on the use of repeated TMS treatment in 2014, which has promoted research cooperation in this field [
119]. The number of publications in 2020 declined slightly, which may be related to the reduction in international academic cooperation caused by the COVID-19 pandemic [
120]. However, according to the current trend analysis, research on the treatment of autism using neuromodulation techniques will continue to show an overall growth trend in the next few years.
The authors of the publications published in the past 31 years covering Asia, Europe, North America, South America, and Oceania. The USA and Canada have become central members of a worldwide collaborative network, focusing on the role and mechanism of repetitive transcranial magnetic stimulation rTMS and tDCS in adult patients with autism in recent years [
47,
121]. It is worth noting that, despite the existence of international cooperation, research in this field is mainly concentrated in developed or high-income countries, and African countries are not involved. Recent research has shown that the age-standardized prevalence of ASD in Africa is higher than the global average [
2]. However, there are few studies in this field in Africa. The top three countries contributed 271 papers, accounting for 69.1% of all papers. The USA, which ranked first in the number of papers, contributed 170 papers. The top three institutions in terms of the number of papers contributed 64 papers, accounting for 16.8% of all papers. These results show that research in this field is unevenly distributed among countries or regions, and further promotion of participation and cooperation of institutions and countries is needed. The number of papers published and cited by the top 10 authors was 172 (43.9% of all papers) and 4720 (15.5% of the total papers cited), respectively. This means that most publications and influences are concentrated among a few co-authors, and most authors still lack a cooperative relationship. This may limit research progress on the treatment of ASD using neuromodulation techniques. The top 10 journals in terms of number of publications and citations were mainly involved in the fields of neuromodulation, neuroscience, and mental cognition, which showed that neuromodulation techniques have received extensive attention in neuropsychiatric disorders, especially in improving the psychological and cognitive impairment of patients with ASD [
47,
48,
76,
83,
91].
In bibliometrics, keywords represent a high generalization of an article, whereas high-frequency keywords are often used to identify hot spots and frontiers in the research field. In this study, the top three keywords in terms of co-occurrence and centrality were “Children,” “Autism spectrum disorder” and “Transcranial magnetic stimulation,” which shows the importance and popularity of TMS in the treatment of autistic children. In the past 20 years, “children” has been a keyword with high co-occurrence frequency and representativeness, indicating that children are the key population in ASD research [
1,
72,
122]. In recent years, the incidence of ASD in children has been rising, which may be related to the popularization of family health education concepts and the progress of overall medical diagnosis, early screening, and intervention measures [
123,
124]. It is worth noting that in the past 40 years, our understanding of autism has significantly progressed, but the services of adults with autism still lags far behind those for children, and diagnosis and treatment may face greater challenges [
125]. Peter G Enticott, McLeod Frampton Gwynette, Pushpal Desarkar, Stephanie H Ameis, and Douglas Teixeira Leffa. etal. have conducted significant research on adult autistic patients, suggesting that TMS and tDCS will bring clear benefits to their social functions and emotions, which may help promote the development of clinical treatment for adult autistic patients [
76,
77,
92,
126,
127]. “Modulation” is the keyword with the highest burst intensity, and is also the core content of treatment in ASD. Previous studies have shown that mental disorders such as ASD may arise from an imbalance of E/I in the neural microcircuit, whereas the compensatory increase in the excitability of inhibitory cells partially alleviates the social defects caused by the increase in E/I balance [
55,
128,
129]. Recent studies have shown that non-invasive neuromodulation techniques can regulate the distribution of neurotransmitters and metabolites in the brain [
47,
48,
55], and improve the social and cognitive functions of patients with autism by altering the excitability of the cortex and neural circuits [
39,
51,
79], which provides a theoretical basis for follow-up research in the treatment of ASD. The burst keywords in the most recent five years mainly focus on “obsessive–compulsive disorder,” “transcranial direct current stimulation,” “working memory,” “double blind” and “adolescent”. Although in our analysis, these were five independent high-frequency keywords, in the actual publications, we found that these keywords often appeared together. Previous studies have shown that ASD and obsessive–compulsive disorder (OCD) are highly comorbid, and that there may be a pathophysiological basis for comorbidity between them [
130,
131]. A prior study showed that 25% of young people with OCD were diagnosed with ASD, and 5% of young people with ASD were diagnosed with OCD [
132]. ASD combined with OCD is associated with more overall dysfunction, which is far more difficult to treat than a single disease [
133,
134]. A recent study showed that DBS in the ventral forelimb of the internal capsule or medial forebrain bundle could reduce the symptoms of obsessive–compulsive behavior and depression, which indicates that neuromodulation techniques may be an effective intervention in the comorbidity of ASD, but further exploration is needed [
135]. In addition, we found that some recent clinical studies predominantly focused on the positive effects of tDCS in adolescents with ASD, including social cognition and emotional behavior [
66,
68,
102]. Working memory is often impaired in ASD, which may be the basis of the core defects of cognitive and social functions [
136]. These findings suggest that tDCS is a simple and well-tolerated adjuvant therapy that can improve the quality of life and autonomy of patients with ASD. Based on our analysis and the above literature review, the five keywords “obsessive–compulsive disorder,” “transcranial direct current stimulation,” “working memory,” “double blind” and “adolescent” have gradually come to reflect hot spots and fronts in the field of neuromodulation techniques in the treatment of ASD.
Limitations
This study has several limitations. First, as the data are only from the WoS database, the analysis may not be sufficiently comprehensive. Second, our search was limited to literature published in English, which makes the analysis incomplete to a certain extent. Although only 11 articles written in non-English languages were excluded, these still had reference values and significance. In addition, although VOSviewer and CiteSpace are professional bibliometric analysis software tools that allow objective analysis, different researchers may have different perspectives on the same content; therefore, their bias is unavoidable. Finally, the text content of some pictures displayed using these software programs is incomplete; although this does not affect the understanding of the literature, these software programs need to be further improved.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.