Sage Journals: Discover world-class research

Abstract

Background:

There is increasing evidence that children with autism spectrum disorder are accompanied by specific anatomical alterations. However, the anatomical abnormalities in adults with autism spectrum disorder are poorly understood. This study was aimed to identify the neuroanatomical substrates underlying the pathophysiology of adults with autism spectrum disorder. We also investigated the relationship between neuroanatomical alterations and clinical and demographic characteristics.

Methods:

A total of 13 datasets were enrolled, of which 12 studies compared whole-brain differences of 382 adult patients with autism and 393 healthy control subjects. We conducted a meta-analysis to quantitatively estimate regional gray matter volume abnormalities in individuals with autism using the effect-size signed differential mapping.

Results:

The voxel-wise meta-analysis revealed that relative to controls, adults with autism spectrum disorder had significantly increased gray matter volume in the middle temporal gyrus, superior temporal gyrus, postcentral gyrus and parahippocampal gyrus, and reduced gray matter volume in the anterior cingulate cortex and cerebellum. Variations in gray matter volume were significantly associated with the mean age and mean total IQ score of the patients, as well as with the percentage of male patients with autism.

Conclusion:

These findings confirmed that the neuroanatomical alterations in the fronto-temporal cortices, limbic system and cerebellum in adult individuals with autism were different from the children and young adolescent’s autism. The effects of demographic characteristics on the brain morphological changes allow us to further clarify the neurobiological mechanisms and developmental trajectory in adult population with autism spectrum disorder.

Keywords

Autism spectrum disorder adult voxel-based morphometry meta-analysis

Introduction

Autism spectrum disorder (ASD) is considered as a lifelong neurodevelopmental disorder and is characterized by deficits in communication and social interaction as well as repetitive behavior or restricted interest (Diagnostic and Statistical Manual of Mental Disorders–Fifth Edition [DSM-5], American Psychiatric Association, 2013).Since first behavioral symptoms appear early in life and the diagnosis of autism is commonly made by 2–4 years of age, the literature on the autism has largely focused on children and young adolescents. For those diagnosed with ASD in childhood, most symptoms would tone down or even disappear in several typical childhood autistic traits when they become adults, but leaving core autism features, such as limited reciprocity in communication and social interaction and a very restricted range of self-chosen behaviors (Billstedt et al., 2007). In contrast to children with autism, adults with autism often show significant disability, as well as psychiatric and behavioral comorbidities (Marriage et al., 2009). Moreover, comorbid conditions, medication and other factors can complicate and change the ASD symptom profile, and thus bring difficulty in obtaining an accurate diagnosis of adult with autism.

Although the etiopathogenesis of autism has yet to be determined, it has been suggested that brain structure abnormalities play an important role in the development of autism. Studies have shown age-specific anatomical alterations in brain size and volume in ASD (Courchesne et al., 2011; Redcay and Courchesne, 2005). Both cortical gray matter (GM) and white matter (Abell et al., 1999; Carper et al., 2002), as well as several subcortical structures, such as amygdala (Sparks et al., 2002) and hippocampus (Sparks et al., 2002), are observed to be expanded in autism. However, these gross anatomical differences are most prominent during early postnatal life and childhood and may be less apparent during adolescence and adulthood. A variety of postmortem and meta-analyses also validated that such enlargements are present only in children and adolescents with ASD (Stanfield et al., 2008). Although a great percentage of the literature on autism anatomy has addressed the anatomical abnormalities that are found in the beginning stages, the older and mature autistic neural structural abnormalities that underlie the emergence of autistic behavior in adults have not yet been addressed.

Although some neuroimaging studies recruited older adolescents or adults with autism, the results of these adult studies were less consistent than findings for children. For example, volume increases and decreases have been reported in temporal lobe, limbic system, as well as cerebellum (Abell et al., 1999; Craig et al., 2007), and those regions are hypothesized to be critically involved in ASD. The heterogeneity of subjects, variability of outcome measures used and other fundamental problems of research design have led to discrepancies in describing the etiology of the disorder.

There are many aspects of the medical, environmental and socio-demographic domains that are associated with or even cause neuroanatomical abnormalities in the brain. ASD is a kind of neurodevelopmental disorder, and therefore, related changes in age and IQ are crucial to understanding the neurobiological mechanisms associated with autistic behavior. It appears that the adult autistic populations included in past adult population studies were far from homogeneous with respect to age range and cognitive development (Murphy et al., 2011). Therefore, knowledge of the influence of confounding variables such as age, gender and IQ, as well as clinical characteristics on brain anatomical changes of ASD, is needed. To delineate the characteristic abnormalities associated with ASD in the context of normative longitudinal changes in neuroanatomy, demographic and clinical domains properly require large and complex studies.

Currently, a novel coordinate-based meta-analytic technique called effect-size signed differential mapping (ES-SDM) has enabled the identification of the most spatially consistent brain changes within the literature through the use of the coordinate information reported in each study. ES-SDM can also allow us to investigate the potential effect of several relevant socio-demographic and clinical variables by means of simple linear regression (Radua et al., 2012). This method has been successfully applied to identify the most consistent structural brain changes in autism (Aoki et al., 2013; Cauda et al., 2011, 2014; Redcay and Courchesne, 2005; Stanfield et al., 2008), but to our knowledge, no study has focused on brain changes in adults with autism. Only one meta-analysis study has split categorically into pediatric and adolescents samples and adults samples, but in these 10 adult datasets, 3 were not pure adult sample (Hyde et al., 2010; Kosaka et al., 2010; Rojas et al., 2006); for example, in Rojas’ study, the mean age was 20.78 years, but the age range was 7.8–44 years (Rojas et al., 2006). The different age stage included in this mixed sample which included both adolescents and adults will work as a confounding factor, which makes it difficult to identify the neurobiological mechanisms underlying adults with autism.

Here, we use ES-SDM to identify the most consistent brain structural changes found in voxel-based morphometry (VBM) studies involving individuals with adult ASD and typically developing individuals. Furthermore, by combining these techniques with meta-regression methods, we were able to characterize the impact of key clinical variables on brain structure in adults with ASD.

Materials and methods

Study selection

A systematic search strategy involving two steps was performed to identify relevant studies. First, a literature search of articles published from 1 January 1999 through 14 January 2015 was conducted using the PubMed, ScienceDirect, Medline and Web of Science database. We used the following search terms: ‘autism or Asperger, or adult autisms plus VBM, morphometry, voxel based or voxel wise’. Review articles and reference lists of included studies were checked to identify further studies for inclusion.

Studies were selected for the meta-analysis using the following inclusion criteria: (1) an adult autistic disorder or Asperger’s disorder diagnosed based on the Diagnostic and Statistical Manual of Mental Disorders–Third Edition, Revised (DSM-III-R), Diagnostic and Statistical Manual of Mental Disorders–Fourth Edition (DSM-IV) or International Statistical Classification of Diseases and Related Health Problems–10th Revision (ICD-10) diagnostic criteria; (2) use of the whole-brain VBM method to investigate differences in whole brain; (3) comparisons of individuals with autistic disorder or Asperger’s disorder with a healthy comparison group; (4) age range of patient group and control group is 18–60 years; (5) clearly reported the three-dimensional coordinates of changes in stereotactic space. The exclusion criteria for the final meta-analysis were as follows: (1) literature reviews or any other meta-analysis results; (2) subjects were children or adolescent, or subjects’ ages ranged between adolescent and adulthoods; (3) patients group or control group comprising less than 10 individuals; (4) studies that re-analyzed previously published data; (5) non-English or unavailable full-text record studies. Meta-analysis of Observational Studies in Epidemiology guidelines were followed in the study (Stroup et al., 2000).

Broad searches yielded 951 hits, 831 of which did not meet the inclusion criteria based on inspections of their titles and/or abstracts. These searches initially identified 35 candidate autism-related articles for possible inclusion after careful assessment for methodology consideration (Figure 1). Recent studies have identified age-specific brain morphological changes in autism (Courchesne et al., 2007, 2011), which means that the brain impairments may differ at different ages (e.g. overgrowth of the amygdala and temporal and frontal cortex emerges at young ages but not at older ages) (Courchesne et al., 2011; Stanfield et al., 2008) and that defects present at younger ages may not be present at older ages and vice versa. Because of these age-specific changes in anatomical pathology, 16 studies that enrolled patients who were younger than 18 years were discarded (Boddaert et al., 2004; Brieber et al., 2007; Cheng et al., 2011; Kaufmann et al., 2013; Ke et al., 2008; Kurth et al., 2011; Kwon et al., 2004; McAlonan et al., 2005, 2008; Mengotti et al., 2011; Riva et al., 2011, 2013; Salmond et al., 2005, 2007; Uddin et al., 2011; Waiter et al., 2004). Eight studies also had to be excluded because subjects’ age ranges are between adolescents and adults (Bonilha et al., 2008; Freitag et al., 2008; Greimel et al., 2013; Hyde et al., 2010; Kosaka et al., 2010; Rojas et al., 2006; Spencer et al., 2006; Waiter et al., 2004). One article was identified in one review article (Riedel et al., 2014). Therefore, a total of 12 high-quality datasets were finally included in this meta-analysis (Abell et al., 1999; Craig et al., 2007; Ecker et al., 2012; Itahashi et al., 2015; Lai et al., 2013; McAlonan et al., 2002; Mueller et al., 2013; Riedel et al., 2014; Schmitz et al., 2006, 2008; Toal et al., 2010; Wilson et al., 2009).

Figure 1.

Flow diagram of inclusion and exclusion process of selected articles of VBM studies in patients with ASD.

Voxel-based meta-analysis

Prior to conducting the voxel-based meta-analysis, the reported peak coordinates were used to recreate a statistical map of the differences between patients and controls for each study. For each article included in the meta-analysis, we recorded the following variables: sample size, sex, mean age of the participants, diagnosis, IQ scores and performance in behavioral tests.

The ES-SDM (www.sdmproject.com/software), version 4.13, for meta-analysis was conducted. As an intrinsically whole-brain technique, this method does not restrict analyses to a priori regions of interest, thereby providing a regionally unbiased assessment of neuroanatomical differences on a point-by-point basis across the entire brain. This new meta-analytic technique uses the reported peak coordinates to recreate maps of the effect size of group differences. It has been widely applied in neuropsychiatric disorders including schizophrenia and attention-deficit/hyperactivity disorder (ADHD) (Bora et al., 2012; Hart et al., 2013; McCarthy et al., 2014), described in detail elsewhere (Radua and Mataix-Cols, 2009) and only briefly summarized here. First, we ensured that the same threshold throughout the whole brain was used within each included study, in order to avoid biases toward liberally thresholded brain regions (Radua and Mataix-Cols, 2009). Second, a map of the differences in GM is separately recreated for each study. Third, the mean map is obtained by voxel-based calculation of the mean of the study maps, weighted by the squared root of the sample size of each study so that studies with large sample sizes contribute proportionally more to the final map. The standard ES-SDM thresholds (uncorrected p < 0.005, extent threshold of clusters >10 voxels) which were proposed to optimally balance the sensitivity and specificity and to be an approximate equivalent to corrected p-value < 0.05 in ES-SDM for effect size were used.

Sensitivity analysis and heterogeneity analysis

A systemic whole-brain voxel-based jackknife sensitivity analysis was conducted to test the replicability of results in different studies. The analysis was repeated 20 times, with a different study being excluded each time. If the findings of previous SDM significant results could be replicated in all or most of the included studies, the findings might be very replicable and conclusive.

The statistical heterogeneity of individual clusters was examined using a random effects model with Q statistics (χ² distribution converted to z values and tested with a permutation approach [p < 0.005, uncorrected for false discovery rate [FDR]; peak height z = 1; cluster extent = 10 voxels]); the same approach was used in our previous study (Chen et al., 2015). To control for any possible methodological differences between the studies, the subgroup analysis was repeated several times and included only those studies that were methodologically homogeneous (see Supplementary Appendix).

Meta-regression of confounding biases

We enrolled all the possible confounding factors that might influence the analytic results, such as the mean age of the patients and controls, the gender ratios of both groups, percentage of medicated patients, scores of clinical rating scales, onset age, family history, duration of illness and comorbidities. Meta-regression steps were performed for these factors to determine whether they were correlated with meta-analytic GM results (see Supplementary Appendix for details of meta-regression method).

Results

As shown in Figure 1, the search strategy identified 951 studies; of these, 12 studies compared whole-brain differences of 382 adult patients with autism and 393 healthy control subjects (Abell et al., 1999; Craig et al., 2007; Ecker et al., 2012; Itahashi et al., 2015; Lai et al., 2013; McAlonan et al., 2002; Mueller et al., 2013; Riedel et al., 2014; Schmitz et al., 2006, 2008; Toal et al., 2010; Wilson et al., 2009). The percentage of male subjects in autism (306 patients, 80.1%) and control (316 control subjects, 80.4%) samples was not statistically different (χ² = 6.286, p = 0.615). The mean age of the autism (32.40 years) and the control subjects (32.24 years) was not significantly different (t = 0.098, p = 0.923). However, the mean IQ of the healthy control was marginally significantly higher than that of the patient group (t = −2.023, p = 0.054). The clinical and demographic data of the participants from all recruited studies are presented in Table 1.

Table 1.

Demographic characteristics of voxel-based morphometry datasets included in the meta-analysis.

Author	Patients inclusion		Patients			Controls			Methodological characteristics
	Diagnosis	Diagnostic tool	Age (M ± SD)	IQ (M ± SD)	N (F/M)	Age (M ± SD)	IQ (M ± SD)	N (F/M)	Smooth kernel	Significance level
Abell et al. (1999)	Autism	DSM-IV	28.75 ± 6.6	91.4 ± 15.5	15 (3/12)	25.33 ± 3.1	97.4 ± 8.6	15 (3/12)	12 mm	p < 0.001, uncorrected
Craig et al. (2007)	ASD	ICD-10	37.9 ± 11.4	103.4 ± 17	14 (14/0)	35 ± 14	111.2 ± 14.5	19 (19/0)	5 mm	p < 0.004
Ecker et al. (2012)	ASD	ICD-10	26 ± 7	110 ± 15	89 (0/89)	28 ± 6	113 ± 12	89 (0/89)	4 mm	p < 0.05
Itahashi et al. (2015)	HFA	DSM-IV	30.21 ± 7.48	106 ± 14.22	46 (0/46)	30.54 ± 6.39	109.22 ± 7.47	46 (0/46)	4 mm	p < 0.05, corrected
Lai et al. (2013)	ASD	ICD-10	27.5 ± 7.45	115.15 ± 13.95	60 (30/30)	27.85 ± 6.05	118.85 ± 9.35	60 (30/30)	4 mm	p < 0.001
McAlonan et al. (2002)	ASP	ICD-10	32 ± 10	96 ± 15	21 (2/19)	33 ± 17	114 ± 14	24 (2/22)	NA	p < 0.001
Mueller et al. (2013)	HFA	ICD-10	35.5 ± 11.4	111.3 ± 13.4	12 (4/8)	33.3 ± 9	110.8 ± 14.4	12 (3/9)	4 mm	p < 0.001, uncorrected
Riedel et al. (2014)	ASD	ICD-10 and DSM-IV	35.4 ± 9.1	124.5 ± 12.3	30 (11/19)	35.5 ± 8.3	123.6 ± 13.8	30 (11/19)	8 mm	FWE = 0.07
Schmitz et al. (2006)	ASD	ICD-10	38 ± 9	105 ± 14	10 (0/10)	39 ± 6	106 ± 13	12 (0/12)	10 mm	p < 0.001, uncorrected
Schmitz et al. (2008)	Autism	ICD-10	37.8 ± 7	107 ± 9	10 (0/10)	38.2 ± 6	106 ± 13	10 (0/10)	10 mm	p < 0.05, corrected
Toal et al. (2010)	ASP	ICD-10	32 ± 12	106 ± 15	39 (4/35)	32 ± 9	105 ± 12	33 (3/30)	8 mm	cluster p = 0.003
	Autism	ICD-10	30 ± 8	84 ± 23	26 (5/21)	32 ± 9	105 ± 12	33 (3/30)	8 mm	cluster p = 0.003
Wilson et al. (2009)	Autism	DSM-IV	30.10 ± 9.18	91.5 ± 19.67	10 (3/7)	29.40 ± 7.91	127.2 ± 9	10 (3/7)	12 mm	p < 0.05

DSM-IV: Diagnostic and Statistical Manual of Mental Disorders–Fourth Edition; ICD-10: International Statistical Classification of Diseases and Related Health Problems–10th Revision; ASD: Autism spectrum disorder; HFA: high-functioning autism spectrum disorder; ASP: Asperger’s disorder, FWE: family wise error.

Regional differences in GM

In the pooled meta-analysis, patients with adult autism had increased GM mainly in left middle temporal gyrus (MTG, extend to superior temporal gyrus [STG] and parahippocampal gyrus), right postcentral gyrus and left parahippocampal gyrus when compared with healthy controls. We also observed decreased GM primarily in the right cerebellum and left anterior cingulate cortex (ACC) (Table 2; Figure 2).

Table 2.

Regional differences in gray matter between adult autism patients and HC subjects identified by the present meta-analyses (voxel-wise p < 0.005).

MNI coordinates	SDM-Z	p	Voxels	Region	Cluster breakdown (voxels, n)
Autism > HC
−58, −20, −6	3.051	~0	987	Left middle temporal gyrus, BA 22	Middle temporal gyrus, BA21 (184); middle temporal gyrus, BA22 (192); superior temporal gyrus, BA22 (143), BA48 (96); Middle temporal gyrus, BA20 (63); inferior temporal gyrus, BA20 (15); insula, BA48 (11); postcentral gyrus (19); temporal pole, superior temporal gyrus, BA38 (24)
58, −6, 22	2.347	0.000071219	523	Right postcentral gyrus, BA43	Postcentral gyrus, BA43, BA48, BA3 (139); superior temporal gyrus, BA22 (62); precentral gyrus, BA6(50); middle temporal gyrus, BA21 (41); superior temporal gyrus, BA21, BA42 (49); right superior temporal gyrus, BA42, BA48 (17); middle frontal gyrus, BA6 (19)
−14, −6, −24	2.027	0.000968168	69	Left parahippocampal gyrus, BA28	Parahippocampal gyrus, BA28 (34); inferior temporal gyrus, BA20 (17); fusiform gyrus, BA20 (9)
Autism < HC
22, −72, −18	−2.474	0.000079476	884	Right cerebellum, hemispheric lobule VI, BA18	Inferior temporal gyrus, BA37, BA20 (176); cerebellum, hemispheric lobule VI, BA18 (95); cerebellum, crus I (75); fusiform gyrus, BA18, BA19, BA37 (110); middle temporal gyrus, BA37 (21)
0, 46, 10	−1.961	0.00204781	99	Left anterior cingulated cortex/paracingulate gyri	Anterior cingulate/paracingulate gyri, BA32, BA24, BA10 (98)

SDM: signed differential mapping; HC: healthy control.

Figure 2.

Regions showing increased and reduced gray matter in patients with adult autism compared with healthy controls.

Reliability analysis and heterogeneity analysis

As shown in Table 3, a whole-brain jackknife sensitivity analysis of the pooled meta-analysis showed that the findings of GM increases in patients in left middle temporal gyrus and right postcentral gyrus were preserved throughout all 13 combinations of datasets. The results of increased GM in the left parahippocampus gyrus and decreased GM in right cerebellum remained significant in all but one combination, and those with decreased GM in the left ACC in all but three combinations.

Table 3.

Subgroups and sensitivity analyses.

	Increased GM			Decreased GM
	L MTG	R PTG	L PHG	R cerebellum	L ACC
Subgroup analysis
Studies that acquired slices with a thickness of 1.2–1.5 mm (n = 9)	N	Y	N	Y	Y
Studies using a 1.5 T MR scanner (n = 7)	N	Y	N	Y	Y
Studies with a modulation step (n = 10)	Y	Y	Y	Y	N
Jackknife sensitivity analysis, discarded study
Abell et al. (1999)	Y	Y	Y	Y	Y
Craig et al. (2007)	Y	Y	Y	Y	N
Ecker et al. (2012)	Y	Y	N	Y	Y
Itahashi et al. (2015)	Y	Y	Y	Y	Y
Lai et al. (2013)	Y	Y	Y	Y	N
McAlonan et al. (2002)	Y	Y	Y	Y	N
Mueller et al. (2013)	Y	Y	Y	Y	Y
Riedel et al. (2014)	Y	Y	Y	Y	Y
Schmitz et al. (2006)	Y	Y	Y	Y	Y
Schmitz et al. (2008)	Y	Y	Y	Y	Y
Toal et al. (2010) (ASP)	Y	Y	Y	N	Y
Toal et al. (2010) (Autism)	Y	Y	Y	Y	Y
Wilson et al. (2009)	Y	Y	Y	Y	Y

GM: gray matter; L: left; R: right; MTG: middle temporal gyrus; PTG: postcentral gyrus; PHG: parahippocampal gyrus; ACC: anterior cingulate cortex; MR: magnetic resonance; ASP: Asperger’s disorder.

Y: Yes, brain region remains significantly increased/decreased in the subgroup analysis or after exclusion of the study in the jackknife analysis.

N: No, brain region is no longer significantly increased/decreased in the subgroup analysis or after exclusion of the study in the jackknife analysis.

Analysis of heterogeneity revealed that a number of regions (right amygdala, right postcentral gyrus, left superior frontal gyrus and left precentral gyrus) with altered GM exhibited significant statistical heterogeneity among the studies (p < 0.005).

Subgroup analysis

The results described above remained largely unchanged when the analyses were limited to methodologically homogeneous groups of studies. Only one additional significant cluster in the right amygdala (Montreal Neurological Institute [MNI] coordinates x = 32, y = 0, z = −26, SDM Z = −1.541, p = 0.0018) emerged in the subgroup analysis of studies that acquired slices with a thickness of 1.2–1.5 mm and of studies that used a 1.5-T magnetic resonance (MR) scanner.

Meta-regression analysis

We explored the mean age, mean total IQ score and the percentage of male patients with autism, which were available for all 382 patients in the 13 datasets, using meta-regression. The mean age of the patients group was negatively associated with GM in the left MTG (r = −0.661, permutation-derived p < 0.0001; Figure 3). The mean total IQ score of the patients was positively correlated with GM in the left parahippocampal gyrus (r = 0.607, permutation-derived p < 0.0001; Figure 4). The percentage of male patients with autism was significantly associated with GM changes in the right ACC (r = 0.890, permutation-derived p < 0.0001) and right cerebellum (r = 0.521, permutation-derived p < 0.0001) (see Figure 5). The following variables could not be studied because data were available for fewer than 10 studies: mean age at onset, illness duration and percentage of medicated participants. We were also unable to assess whether autism symptom severity was associated with the reported structural changes because the included studies used a wide range of different measures.

Figure 3.

Results of the meta-regression analysis showing a significant inverse association between age and gray matter volumes in the left middle temporal gyrus.

Figure 4.

Results of the meta-regression analysis showing a significant association between total IQ score and gray matter volumes in the left parahippocampal gyrus.

Figure 5.

Results of the meta-regression analysis showing a significant association between the percentages of male patients with autism and gray matter volumes in the right anterior cingulate and right cerebellum.

Discussion

The primary objective of this study was to identify distinct GM abnormalities in adult populations with autism. We used quantitative ES-SDM meta-analytic methods to synthesize findings of 12 structural neuroimaging studies of adults with autism. The study is timely given that a sufficient number of high-quality studies have only recently become available. The most robust GM increases were identified in MTG, STG, parahippocampal gyrus and postcentral gyrus, while the most evident reductions were located in the cerebellum and ACC. Furthermore, the variations in GM volume (GMV) were significantly correlated with the mean age and mean total IQ score of patients, as well as the percentage of male patients with autism; these correlations may reflect the effects of demographical variation on the brain morphological changes of ASD.

Regional increases in GM regions

It is well known that lesions in the STG, MTG and parahippocampal gyrus are associated with marked impairment in social perception (Zilbovicius et al., 2006), emotional processing and theory of mind (Redcay, 2008). Many neuroimaging studies have revealed brain abnormalities that are consistent with the primary pathological findings that have been identified in the temporal region in autism. Magnetic resonance imaging (MRI) volumetric studies have revealed significant enlargement of GM in the frontal and temporal lobes in autism when compared to controls (Abell et al., 1999; Ecker et al., 2012; Waiter et al., 2004). Additionally, increases in GM in this region were correlated with increased social symptom severity in autism (Ecker et al., 2012). However, adult autistic patients are characterized by hypoactivation within the temporal lobe during sustained attention (Murphy et al., 2014) and decreased functional connectivity when processing emotional body language (Libero et al., 2014). These findings are in line with resting functional magnetic resonance imaging (fMRI) studies that have shown reduced spontaneous brain activity and connectivity in temporal regions (Itahashi et al., 2014, 2015). Considering that social perception and empty function have been associated with this region and that improved social communication skill has been reported in adult autistic patients (Howlin et al., 2004; Tobin et al., 2014), we tentatively assume that increased temporal GMV or density in adult autism may be a compensatory mechanism to alleviate communication-related symptoms in ASD. In contrast to adults with autism, autistic children show significant reduction GM concentration (Boddaert et al., 2004) and functional hypoperfusion (Zilbovicius et al., 2000) in the STG, which has been implicated in the social communication skills that are characteristically impaired in children and adolescents with autism. This finding may contribute to understanding of the growth trajectory in social perception impairment that is characteristic of autism.

We also observed regional increases in GMV in the postcentral gyrus. As part of the somatosensory cortex, the postcentral gyrus receives its afferents mainly from the ventral posterior nucleus and from the ventral posteromedial nucleus of the thalamus (Young et al., 2003). A previous fMRI study revealed that the pre- and post-central sulci were activated by observation of human movement only in autistic individuals but not in controls, demonstrating atypical impairment in autism (Martineau et al., 2010). This area has also been implicated in cognitive control, particularly in executive function (Koechlin et al., 2003), which is known to be specifically impaired in ASD. Our finding of increased GMV in the fronto-temporal cortices and limbic system lends support to theories implicating these regions in the pathophysiology of autism.

Regional decreases in GM regions

One of most consistently reported regions of GM deficits in ASD is the ACC, which serves as an integration hub for many different types of executive function, affect and socio-emotional behavior. There is compelling evidence that altered metabolism (Levitt et al., 2003) and reduced volume of the ACC were found in autism (Haznedar et al., 1997). Likewise, microscopic study revealed that increased cell packing density and decreased cell size in several sub-regions within the ACC were observed in many autistic individuals (Simms et al., 2009). There are also evidence of abnormal ACC activation and functional connectivity in autism (Agam et al., 2010; Assaf et al., 2010; Thakkar et al., 2008). Importantly, these deficiencies in function have been associated with inflexible and repetitive symptoms in ASD (Thakkar et al., 2008). Thus, impairment in the development of this system may constitute a neural substrate for the socio-cognitive deficits present in autism.

Individuals with ASD had a significant decrease in GMV located in cerebellum (mainly in posterior lobules VI, crus I) in this study, which is consistent with some previous meta-analyses of VBM in ASD (Cauda et al., 2011; DeRamus and Kana, 2014). GM abnormalities within the cerebellum are well documented in the ASD literature (Abell et al., 1999; Riva et al., 2013; Rojas et al., 2006). Traditional evidence indicates cerebellar dysfunction is associated with motor impairments in autism (Mostofsky et al., 2009). Recent studies indicate that the cerebellum plays an important role in non-motor functions as well. Townsend et al. (1999) reported smaller cerebellar vermal lobules VI–VII were associated with greater attention-orienting deficits. The presence of motor and cognitive deficits in autism may be related to the cerebellar dysfunction in autism.

In contrast, structural neuroimaging studies in pediatric and adolescent patients with ASD have reported more widely spread volumetric impairments, which include not only the ACC and cerebellum (McAlonan et al., 2008; Riva et al., 2011, 2013) but also the amygdala–hippocampal complex and medial parietal cortex (Brieber et al., 2007; Ke et al., 2008; McAlonan et al., 2005; Riva et al., 2013). It is clear that the amygdala and its adjacent structures are implicated in social cognition and emotional processing (Adolphs, 2010; Emery et al., 2001) and that the medial parietal cortex is involved in self-awareness and self-representation (Lombardo and Baron-Cohen, 2010); thus, the much more widely distributed brain impairment may reflect the more severe autistic symptoms in children and adolescents. Further longitudinal and cross-sectional studies with a larger number of different age cohorts of patients and comparison subjects are needed to observe the effect of age on autism more directly.

Effects of demographical variation on the structural changes

The mean age of the patient group was negatively associated with GM in the left MTG, which indicated that the older adult autistic individuals tended to have smaller MTG. Prior studies have observed thinner cortices in the STG and adjacent areas with increasing age in participants with ASD relative to typically developing controls (Doyle-Thomas et al., 2013; Scheel et al., 2011; Wallace et al., 2010). Maturational brain changes in this region have been implicated in multimodal sensory integration and cognitive functions are believed to be associated with improvement in these functions (Chao et al., 1999; Onitsuka et al., 2004). Thus, deviation from the typical developmental trajectory of GM in temporal brain regions may underlie age-related impairments in social and communication functions in ASD.

Past researches suggest that autistic symptoms varied during cognitive development (Mayes and Calhoun, 2004; Miller and Ozonoff, 2000), that is, as IQ increases, autistic symptoms decrease. Consistent with this, we found disturbed GM in the parahippocampal gyrus was positively related to the mean total IQ of this study population. The increased GMV in the parahippocampal gyrus with higher IQ level in autism patients may indicate higher mind theory ability and other related function implicated in this area. Furthermore, some researches hold that a discrepancy between verbal and nonverbal intellectual abilities may be a particularly informative phenotype of ASD (Black et al., 2009). However, few of the included articles reported the verbal and nonverbal IQ subscale scores, which limited our ability to perform further regression analysis. Therefore, further studies taking IQ and subscale IQ variables into account are needed for a more complete understanding of the brain abnormalities in patients with autism.

The observed sex-related increase in the ACC and cerebellum was also well documented by Lai (Lai et al., 2013) and Takahashi (Takahashi et al., 2011) among adult males on the autism spectrum, which suggested the male autism tends to increase GMV. The sex difference in the ACC may contribute to superior rational cognitive functions in males. As it is consistent with what is known as the ‘extreme male brain’ theory, it posits autism represents an extreme of the male pattern with enhanced systemizing ability (Baron-Cohen, 2002; Baron-Cohen et al., 2005). Sex differences are also seen in epidemiological and clinical aspects of autism. Male individuals show a greater incidence of autism (4.3 for a mean male/female sex ratio) than females (Fombonne, 2003). Males are found to show more externalizing behavioral problems than females, such as aggressive behavior, hyperactivity, reduced prosocial behavior and increased repetitive/restricted behaviors and interests (Werling and Geschwind, 2013). These evidences could explain the biological basis of the male bias in ASD.

Although MRI studies have shown some evidence for brain enlargement in early ages of children that may be related to degeneration and decline during adolescence and young adulthood, it is still unclear how anatomical pathology alters with age after these early years. Our study suggests that in adult autism, abnormal neural and anatomy processes may be different from children: brain enlargement in adult autism seems to be most prominent in the frontal, temporal and limbic area, but brain atrophy is mainly located in the cerebellum and anterior cingulate area. Consistent with this, Raznahan et al. (2010) reported that the ratio of cerebral lobar volume to overall total brain volume follows an abnormal trajectory across adolescence and adulthood, suggesting that in autism, abnormal neural processes continue to occur during adulthood. In older samples, clinical studies have observed part of reduced symptom severity but keep some core autism symptom with increasing age in participants with ASD relative to typically developing controls (Howlin et al., 2004; Tobin et al., 2014). Thus, evidence of current meta-analysis in the autistic person may reflect complex change with age and secondary remodeling processes (possibly both brain degeneration and enlargement) after the neuroinflammatory processes (Morgan et al., 2010). These evidences suggest that cortical dysmaturation in ASD is not restricted to childhood but extends across the lifespan, which led to the theory of age-specific anatomical abnormalities in autism.

The finding of associations between these brain structural changes and several demographical variations provides further evidence for different neurobiological processes in autistic samples. Both the group analysis and regression technique revealed that people with ASD had significant age-related differences in brain anatomy compared with the typically developing individuals, which reflect the different progresses of disease phases of autism. Although relatively little is known of the factors that shape age-related changes in autism during aging, individual differences in maturation of cortical thickness have been related to IQ (Shaw et al., 2006). This neuroanatomical correlation with IQ in adult autism was also found in the present meta-regression analysis. In other words, differences could be accounted for by preexisting differences in IQ. Furthermore, gender differences associated with GMV changes may help explain the increased vulnerability of males to ASD. These findings will help us to understand the neurobiology and developmental trajectory in adult autism. However, further longitudinal studies are needed to fully elucidate the influences of demographic profiles on brain morphometry and the processes underpinning the age-related brain differences in adults with autism.

Limitations

Despite our success in demonstrating consistencies in findings across studies, specific methodological limitations should be considered with respect to the current structural neuroimaging literature examining adult subjects with autism. First, peak-based meta-analyses are based on summarized (i.e. coordinates from published studies) rather than full statistical image from the original studies, and this may result in inaccurate results (Salimi-Khorshidi et al., 2009). Also, different studies used different statistical thresholds which would bias the results. Future studies of larger and more diverse samples are needed to determine whether this approach is justifiable. Third, the heterogeneity analysis showed that a number of regions with altered GM exhibited significant statistical heterogeneity among studies. To examine the moderator variables that contributed to the heterogeneity of the findings, we conducted meta-regression analyses and subgroup analyses (Radua and Mataix-Cols, 2009). Fourth, the included studies varied in terms of the data acquisition, analysis techniques, patient characteristics and clinical variables. For example, the patients with autism or Asperger’s syndrome are both included in our study. As different subtypes of the classical autistic spectrum, autism and Asperger’s syndrome may have different etiologies, which could influence the results of our findings. It would also be valuable to analyze the effect of subtypes or comorbidities; however, there were no sufficient studies to carry out these subgroup analyses. The differences between GM volume and GM concentration in the included studies might also influence our results. Further longitudinal follow-up studies are necessary to understand the natural progression of autism, as well as its different anatomy and cognitive markers that characterize the differential stages of autism.

Conclusion

In summary, we used the ES-SDM method to identify regions that reliably exhibited anatomical deficits in adults with autism. Our findings indicate that adult autistic patients have significant and robust GM changes in the fronto-temporal cortices, limbic system and cerebellum, which is different from childhood autism. These brain structural changes, together with the results of correlation with demographic characteristics, may clarify the neurobiology and developmental trajectory in adult population with ASD.

Footnotes

Acknowledgements

Xun Yang and Lihua Qiu contributed equally to this work.

Declaration of interest

The author(s) declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.

Funding

This study was supported by the National Natural Science Foundation (Grant Nos. 81371536 and 91132720, and 81271532), the Fundamental Research Funds for the Central Universities, Southwest University for Nationalities (2014NZYQN39), and the China Postdoctoral Science Foundation (Grant No. 2015M572479).

References

American Psychiatric Association (2013) Diagnostic and Statistical Manual of Mental Disorders (DSM-5). American Psychiatric Association, Washington, DC.

Abell

Krams

Ashburner

. (1999) The neuroanatomy of autism: A voxel-based whole brain analysis of structural scans. NeuroReport 10: 1647–1651.

Adolphs

(2010) What does the amygdala contribute to social cognition? Year in Cognitive Neuroscience 1191: 42–61.

Agam

Joseph

Barton

. (2010) Reduced cognitive control of response inhibition by the anterior cingulate cortex in autism spectrum disorders. NeuroImage 52: 336–347.

American Psychiatric Association (2013) Diagnostic and Statistical Manual of Mental Disorders (DSM-5). Washington, DC: American Psychiatric Association.

Aoki

Abe

Nippashi

. (2013) Comparison of white matter integrity between autism spectrum disorder subjects and typically developing individuals: A meta-analysis of diffusion tensor imaging tractography studies. Molecular Autism 4: 25.

Assaf

Jagannathan

Calhoun

. (2010) Abnormal functional connectivity of default mode sub-networks in autism spectrum disorder patients. NeuroImage 53: 247–256.

Baron-Cohen

(2002) The extreme male brain theory of autism. Trends in Cognitive Sciences 6: 248–254.

Baron-Cohen

Knickmeyer

Belmonte

(2005) Sex differences in the brain: Implications for explaining autism. Science 310: 819–823.

10.

Billstedt

Gillberg

(2007) Autism in adults: Symptom patterns and early childhood predictors. Use of the DISCO in a community sample followed from childhood. Journal of Child Psychology and Psychiatry 48: 1102–1110.

11.

Black

Wallace

Sokoloff

. (2009) Brief report: IQ split predicts social symptoms and communication abilities in high-functioning children with autism spectrum disorders. Journal of Autism and Developmental Disorders 39: 1613–1619.

12.

Boddaert

Chabane

Gervais

. (2004) Superior temporal sulcus anatomical abnormalities in childhood autism: A voxel-based morphometry MRI study. NeuroImage 23: 364–369.

13.

Bonilha

Cendes

Rorden

. (2008) Gray and white matter imbalance—Typical structural abnormality underlying classic autism? Brain Development 30: 396–401.

14.

Bora

Fornito

Yucel

. (2012) The effects of gender on grey matter abnormalities in major psychoses: A comparative voxelwise meta-analysis of schizophrenia and bipolar disorder. Psychological Medicine 42: 295–307.

15.

Brieber

Neufang

Bruning

. (2007) Structural brain abnormalities in adolescents with autism spectrum disorder and patients with attention deficit/hyperactivity disorder. Journal of Child Psychology and Psychiatry 48: 1251–1258.

16.

Carper

Moses

Tigue

. (2002) Cerebral lobes in autism: Early hyperplasia and abnormal age effects. NeuroImage 16: 1038–1051.

17.

Cauda

Costa

Palermo

. (2014) Concordance of white matter and gray matter abnormalities in autism spectrum disorders: A voxel-based meta-analysis study. Human Brain Mapping 35: 2073–2098.

18.

Cauda

Geda

Sacco

. (2011) Grey matter abnormality in autism spectrum disorder: An activation likelihood estimation meta-analysis study. Journal of Neurology, Neurosurgery & Psychiatry 82: 1304–1313.

19.

Chao

Haxby

Martin

(1999) Attribute-based neural substrates in temporal cortex for perceiving and knowing about objects. Nature Neurosciences 2: 913–919.

20.

Chen

Zhao

. (2015) Voxel-wise meta-analyses of brain blood flow and local synchrony abnormalities in medication-free patients with major depressive disorder. Journal of Psychiatry & Neuroscience 40: 140119.

21.

Cheng

Chou

Fan

. (2011) ANS: Aberrant neurodevelopment of the social cognition network in adolescents with autism spectrum disorders. PLoS ONE 6: e18905.

22.

Courchesne

Campbell

Solso

(2011) Brain growth across the life span in autism: Age-specific changes in anatomical pathology. Brain Research 1380: 138–145.

23.

Courchesne

Pierce

Schumann

. (2007) Mapping early brain development in autism. Neuron 56: 399–413.

24.

Craig

Zaman

Daly

. (2007) Women with autistic-spectrum disorder: Magnetic resonance imaging study of brain anatomy. British Journal of Psychiatry 191: 224–228.

25.

DeRamus

Kana

(2014) Anatomical likelihood estimation meta-analysis of grey and white matter anomalies in autism spectrum disorders. NeuroImage: Clinical 7: 525–536.

26.

Doyle-Thomas

Duerden

Taylor

. (2013) Effects of age and symptomatology on cortical thickness in autism spectrum disorders. Research in Autism Spectrum Disorders 7: 141–150.

27.

Ecker

Suckling

Deoni

. (2012) Brain anatomy and its relationship to behavior in adults with autism spectrum disorder: A multicenter magnetic resonance imaging study. Archives of General Psychiatry 69: 195–209.

28.

Emery

Capitanio

Mason

. (2001) The effects of bilateral lesions of the amygdala on dyadic social interactions in rhesus monkeys (Macaca mulatta). Behavioral Neuroscience 115: 515–544.

29.

Fombonne

(2003) Epidemiological surveys of autism and other pervasive developmental disorders: An update. Journal of Autism and Developmental Disorders 33: 365–382.

30.

Freitag

Konrad

Haeberlen

. (2008) Perception of biological motion in autism spectrum disorders. Neuropsychologia 46: 1480–1494.

31.

Greimel

Nehrkorn

Schulte-Ruther

. (2013) Changes in grey matter development in autism spectrum disorder. Brain Structure & Function 218: 929–942.

32.

Hart

Radua

Nakao

. (2013) Meta-analysis of functional magnetic resonance imaging studies of inhibition and attention in attention-deficit/hyperactivity disorder: Exploring task-specific, stimulant medication, and age effects. JAMA Psychiatry 70: 185–198.

33.

Haznedar

Buchsbaum

Metzger

. (1997) Anterior cingulate gyrus volume and glucose metabolism in autistic disorder. American Journal of Psychiatry 154: 1047–1050.

34.

Howlin

Goode

Hutton

. (2004) Adult outcome for children with autism. Journal of Child Psychology and Psychiatry 45: 212–229.

35.

Hyde

Samson

Evans

. (2010) Neuroanatomical differences in brain areas implicated in perceptual and other core features of autism revealed by cortical thickness analysis and voxel-based morphometry. Human Brain Mapping 31: 556–566.

36.

Itahashi

Yamada

Nakamura

. (2015) Linked alterations in gray and white matter morphology in adults with high-functioning autism spectrum disorder: A multimodal brain imaging study. NeuroImage: Clinical 7: 155–169.

37.

Itahashi

Yamada

Watanabe

. (2014) Altered network topologies and hub organization in adults with autism: A resting-state fMRI study. PLoS ONE 9: e94115.

38.

Kaufmann

Zotter

Pixner

. (2013) Brief report: CANTAB performance and brain structure in pediatric patients with Asperger syndrome. Journal of Autism and Developmental Disorders 43: 1483–1490.

39.

Hong

Tang

. (2008) Voxel-based morphometry study on brain structure in children with high-functioning autism. NeuroReport 19: 921–925.

40.

Koechlin

Ody

Kouneiher

(2003) The architecture of cognitive control in the human prefrontal cortex. Science 302: 1181–1185.

41.

Kosaka

Omori

Munesue

. (2010) Smaller insula and inferior frontal volumes in young adults with pervasive developmental disorders. NeuroImage 50: 1357–1363.

42.

Kurth

Narr

Woods

. (2011) Diminished gray matter within the hypothalamus in autism disorder: A potential link to hormonal effects? Biological Psychiatry 70: 278–282.

43.

Kwon

Pedatella

. (2004) Voxel-based morphometry elucidates structural neuroanatomy of high-functioning autism and Asperger syndrome. Developmental Medicine & Child Neurology 46: 760–764.

44.

Lai

Lombardo

Suckling

. (2013) Biological sex affects the neurobiology of autism. Brain 136: 2799–2815.

45.

Levitt

O’Neill

Blanton

. (2003) Proton magnetic resonance spectroscopic imaging of the brain in childhood autism. Biological Psychiatry 54: 1355–1366.

46.

Libero

Stevens

Jr Kana

(2014) Attribution of emotions to body postures: An independent component analysis study of functional connectivity in autism. Human Brain Mapping 35: 5204–5218.

47.

Lombardo

Baron-Cohen

(2010) Unraveling the paradox of the autistic self. Wiley Interdisciplinary Reviews-Cognitive Science 1: 393–403.

48.

Marriage

Wolverton

Marriage

(2009) Autism spectrum disorder grown up: A chart review of adult functioning. Journal of the Canadian Academy of Child and Adolescent Psychiatry 18: 322–328.

49.

Martineau

Andersson

Barthelemy

. (2010) Atypical activation of the mirror neuron system during perception of hand motion in autism. Brain Research 1320: 168–175.

50.

Mayes

Calhoun

(2004) Influence of IQ and age in childhood autism: Lack of support for DSM-IV Asperger’s disorder. Journal of Developmental and Physical Disabilities 16: 257–272.

51.

McAlonan

Cheung

. (2005) Mapping the brain in autism. A voxel-based MRI study of volumetric differences and intercorrelations in autism. Brain 128: 268–276.

52.

McAlonan

Daly

Kumari

. (2002) Brain anatomy and sensorimotor gating in Asperger’s syndrome. Brain 125: 1594–1606.

53.

McAlonan

Suckling

Wong

. (2008) Distinct patterns of grey matter abnormality in high-functioning autism and Asperger’s syndrome. Journal of Child Psychology and Psychiatry 49: 1287–1295.

54.

McCarthy

Skokauskas

Frodl

(2014) Identifying a consistent pattern of neural function in attention deficit hyperactivity disorder: A meta-analysis. Psychological Medicine 44: 869–880.

55.

Mengotti

D’Agostini

Terlevic

. (2011) Altered white matter integrity and development in children with autism: A combined voxel-based morphometry and diffusion imaging study. Brain Research Bulletin 84: 189–195.

56.

Miller

Ozonoff

(2000) The external validity of Asperger disorder: Lack of evidence from the domain of neuropsychology. Journal of Abnormal Psychology 109: 227–238.

57.

Morgan

Chana

Pardo

. (2010) Microglial activation and increased microglial density observed in the dorsolateral prefrontal cortex in autism. Biological Psychiatry 68: 368–376.

58.

Mostofsky

Powell

Simmonds

. (2009) Decreased connectivity and cerebellar activity in autism during motor task performance. Brain 132: 2413–2425.

59.

Mueller

Keeser

Samson

. (2013) Convergent findings of altered functional and structural brain connectivity in individuals with high functioning autism: A multimodal MRI study. PLoS ONE 8: e67329.

60.

Murphy

Christakou

Daly

. (2014) Abnormal functional activation and maturation of fronto-striato-temporal and cerebellar regions during sustained attention in autism spectrum disorder. American Journal of Psychiatry 171: 1107–1116.

61.

Murphy

Beecham

Craig

. (2011) Autism in adults. New biological findings and their translational implications to the cost of clinical services. Brain Research 1380: 22–33.

62.

Onitsuka

Shenton

Salisbury

. (2004) Middle and inferior temporal gyrus gray matter volume abnormalities in chronic schizophrenia: An MRI study. American Journal of Psychiatry 161: 1603–1611.

63.

Radua

Mataix-Cols

(2009) Voxel-wise meta-analysis of grey matter changes in obsessive-compulsive disorder. British Journal of Psychiatry 195: 393–402.

64.

Radua

Mataix-Cols

Phillips

. (2012) A new meta-analytic method for neuroimaging studies that combines reported peak coordinates and statistical parametric maps. European Psychiatry 27: 605–611.

65.

Raznahan

Toro

Daly

. (2010) Cortical anatomy in autism spectrum disorder: An in vivo MRI study on the effect of age. Cerebral Cortex 20: 1332–1340.

66.

Redcay

(2008) The superior temporal sulcus performs a common function for social and speech perception: Implications for the emergence of autism. Neuroscience & Biobehavioral Reviews 32: 123–142.

67.

Redcay

Courchesne

(2005) When is the brain enlarged in autism? A meta-analysis of all brain size reports. Biological Psychiatry 58: 1–9.

68.

Riedel

Maier

Ulbrich

. (2014) No significant brain volume decreases or increases in adults with high-functioning autism spectrum disorder and above average intelligence: A voxel-based morphometric study. Psychiatry Research: Neuroimaging 223: 67–74.

69.

Riva

Annunziata

Contarino

. (2013) Gray matter reduction in the vermis and CRUS-II is associated with social and interaction deficits in low-functioning children with autistic spectrum disorders: A VBM-DARTEL Study. The Cerebellum 12: 676–685.

70.

Riva

Bulgheroni

Aquino

. (2011) Basal forebrain involvement in low-functioning autistic children: A voxel-based morphometry study. American Journal of Neuroradiology 32: 1430–1435.

71.

Rojas

Peterson

Winterrowd

. (2006) Regional gray matter volumetric changes in autism associated with social and repetitive behavior symptoms. BMC Psychiatry 6: 56.

72.

Salimi-Khorshidi

Smith

Keltner

. (2009) Meta-analysis of neuroimaging data: A comparison of image-based and coordinate-based pooling of studies. NeuroImage 45: 810–823.

73.

Salmond

Ashburner

Connelly

. (2005) The role of the medial temporal lobe in autistic spectrum disorders. European Journal of Neuroscience 22: 764–772.

74.

Salmond

Vargha-Khademl

Gadian

. (2007) Heterogeneity in the patterns of neural abnormality in autistic spectrum disorders: Evidence from ERP and MRI. Cortex 43: 686–699.

75.

Scheel

Rotarska-Jagiela

Schilbach

. (2011) Imaging derived cortical thickness reduction in high-functioning autism: Key regions and temporal slope. NeuroImage 58: 391–400.

76.

Schmitz

Rubia

Daly

. (2006) Neural correlates of executive function in autistic spectrum disorders. Biological Psychiatry 59: 7–16.

77.

Schmitz

Rubia

van Amelsvoort

. (2008) Neural correlates of reward in autism. British Journal of Psychiatry 192: 19–24.

78.

Shaw

Greenstein

Lerch

. (2006) Intellectual ability and cortical development in children and adolescents. Nature 440: 676–679.

79.

Simms

Kemper

Timbie

. (2009) The anterior cingulate cortex in autism: Heterogeneity of qualitative and quantitative cytoarchitectonic features suggests possible subgroups. Acta Neuropathologica 118: 673–684.

80.

Sparks

Friedman

Shaw

. (2002) Brain structural abnormalities in young children with autism spectrum disorder. Neurology 59: 184–192.

81.

Spencer

Moorhead

TWJ

Lymer

GKS

. (2006) Structural correlates of intellectual impairment and autistic features in adolescents. NeuroImage 33: 1136–1144.

82.

Stanfield

McIntosh

Spencer

. (2008) Towards a neuroanatomy of autism: A systematic review and meta-analysis of structural magnetic resonance imaging studies. European Psychiatry 23: 289–299.

83.

Stroup

Berlin

Morton

. (2000) Meta-analysis of observational studies in epidemiology: A proposal for reporting. Meta-analysis Of Observational Studies in Epidemiology (MOOSE) group. Journal of the American Medical Association 283: 2008–2012.

84.

Takahashi

Ishii

Kakigi

. (2011) Gender and age differences in normal adult human brain: Voxel-based morphometric study. Human Brain Mapping 32: 1050–1058.

85.

Thakkar

Polli

Joseph

. (2008) Response monitoring, repetitive behaviour and anterior cingulate abnormalities in autism spectrum disorders (ASD). Brain 131: 2464–2478.

86.

Toal

Daly

Page

. (2010) Clinical and anatomical heterogeneity in autistic spectrum disorder: A structural MRI study. Psychological Medicine 40: 1171–1181.

87.

Tobin

Drager

KDR

Richardson

(2014) A systematic review of social participation for adults with autism spectrum disorders: Support, social functioning, and quality of life. Research in Autism Spectrum Disorders 8: 214–229.

88.

Townsend

Courchesne

Covington

. (1999) Spatial attention deficits in patients with acquired or developmental cerebellar abnormality. Journal of Neuroscience 19: 5632–5643.

89.

Uddin

Menon

Young

. (2011) Multivariate searchlight classification of structural magnetic resonance imaging in children and adolescents with autism. Biological Psychiatry 70: 833–841.

90.

Waiter

Williams

Murray

. (2004) A voxel-based investigation of brain structure in male adolescents with autistic spectrum disorder. NeuroImage 22: 619–625.

91.

Wallace

Dankner

Kenworthy

. (2010) Age-related temporal and parietal cortical thinning in autism spectrum disorders. Brain 133: 3745–3754.

92.

Werling

Geschwind

(2013) Sex differences in autism spectrum disorders. Current Opinion in Neurology 26: 146.

93.

Wilson

Tregellas

Hagerman

. (2009) A voxel-based morphometry comparison of regional gray matter between fragile X syndrome and autism. Psychiatry Research: Neuroimaging 174: 138–145.

94.

Young

Geyer

Grefkes

. (2003) Regional cerebral blood flow correlations of somatosensory areas 3a, 3b, 1, and 2 in humans during rest: A PET and cytoarchitectural study. Human Brain Mapping 19: 183–196.

95.

Zilbovicius

Boddaert

Belin

. (2000) Temporal lobe dysfunction in childhood autism: A PET study. American Journal of Psychiatry 157: 1988–1993.

96.

Zilbovicius

Meresse

Chabane

. (2006) Autism, the superior temporal sulcus and social perception. Trends in Neurosciences 29: 359–366.

Brain gray matter alterations and associated demographic profiles in adults with autism spectrum disorder: A meta-analysis of voxel-based morphometry studies

Abstract

Background:

Methods:

Results:

Conclusion:

Keywords

Introduction

Materials and methods

Study selection

Voxel-based meta-analysis

Sensitivity analysis and heterogeneity analysis

Meta-regression of confounding biases

Results

Regional differences in GM

Reliability analysis and heterogeneity analysis

Subgroup analysis

Meta-regression analysis

Discussion

Regional increases in GM regions

Regional decreases in GM regions

Effects of demographical variation on the structural changes

Limitations

Conclusion

Footnotes

Acknowledgements

Declaration of interest

Funding

References