Meta-analysis of brain activation in depressive patients with emotional perception disorders***☆

Medical Imaging Center, First Hospital of Xi’an Jiaotong University, Xi’an 710061, Shaanxi Province, China

INTRODUCTION

Dysfunctional emotional perception constitutes a key symptom of depression, but the neurobiological mechanisms remain unclear.Functional magnetic resonance imaging (fMRI) studies of emotion processing have revealed a functional neural circuit involving the prefrontal cortex,basal ganglia, and limbic lobe[1].However,these fMRI studies have typically investigated a small number of subjects, and there has been marked differences between studies in terms of study design,subject characteristics, imaging methods,and analysis methods, leading to inconsistent results[2].Due to the lack of a gold standard, fMRI studies have typically used figures to summarize previous literature,with no objective, quantitative means for data collection and analysis[3].Activation likelihood estimation (ALE) is a method based on the random effects model, which seeks to resolve this bottleneck[4].ALE has been used in meta-analysis of brain function in the fields of cognition and neuropsychiatric disorders[3,5].The current study used ALE to quantitatively examine the distribution of abnormally activated brain regions in depressive patients during emotional perception, as part of a broader attempt to clarify the neuropathological basis of depression.

MATERIALS AND METHODS

Data retrieval

We examined the PscyINFO, Pubmed,and CNKI databases, as well as electronic journals of the Chinese Medical Association, from the establishment of each database until June 2010.We searched for the key words "depressive,depression, fMRI, and emotion" in English, and the words "depression and brain function" in Chinese.In addition, online and manual retrievals were also performed to find additional references mentioned in the literature.This search was conducted independently by two postgraduates majoring in medical imaging with a broad knowledge of literature regarding the effects of emotional stimuli on brain function in depression.Discussion was undertaken if inconsistencies emerged in the retrieval.

Inclusion and exclusion criteria

Inclusion criteria

(1) Articles were written in Chinese or English.(2) Differences in brain activation intensity were reported between the case and control groups.(3) Results were presented as MNI or Talairach coordinates.(4) The studies were fMRI experiments on emotion processing in patients with major depression.(5) Different papers published by the same research group included different data sets.

Exclusion criteria

(1) Repeated studies and studies on bipolar depression.(2) Studies where data or sample descriptions were unclear.(3) Studies without a control group.(4) Animal studies.(5) Overviews, letters, reviews, editorials and other non-original research.

Literature evaluation and extraction

The sample size of included subjects and central coordinates of active areas were extracted from each paper,MNI coordinates were converted to Talairach coordinates using Icbm2tal.According to the nature of emotional stimuli (positive or negative) and the effect of stimulus presentation (increase or decrease in activation intensity), the extracted coordinates were divided into four groups: (1) decreased activation of brain regions elicited by negative emotional stimuli; (2) increased activation of brain regions elicited by negative emotional stimuli; (3) increased activation of brain regions elicited by positive emotional stimuli.(4) decreased activation of brain regions elicited by positive emotional stimuli.The anatomical position of pixel coordinates was independently determined by two researchers manually according to Talairach and Tournoux’s neuroanatomical stereotactic atlas[6], and all inconsistencies were resolved by discussion between researchers.

Outcome indices

The activation of brain regions in depressive patients with emotional perception disorders.

Data analysis

ALE is a method for the analysis of functional brain areas based on voxel coordinates, proposed by Turkeltaubet al[7]and modified by Eickhoffet al[4]as previously described[8].Briefly, all coordinates were converted into Talairach coordinates using Icbm2tal software(http://www.brainmap.org/icbm2tal).Data were analyzed with Ginger-ALE 2.0 software(http://www.brainmap.org/ale), and models were established with the peak coordinates of maximum activation and three-dimensional Gaussian distribution, using the following specific parameter settings: Gaussian filter full-width half-maximum (FWHM) value 8 mm, permutation testP< 0.05, false discovery rate (FDR)qvalue 0.05,pixel cluster threshold value > 250 mm3.Final results were displayed with MRICRON(http://www.cabiatl.com/mricro/mricron),three-dimensional coordinates were mapped to the ICBM(International Consortium for Brain Mapping) anatomical brain template.

RESULTS

Retrieval results

A total of 137 English and 12 Chinese studies were retrieved in the first screening.According to the inclusion and exclusion criteria, 68 of these papers were deemed irrelevant to emotional stimulation on brain function, 40 papers involved bipolar depression patients as subjects,14 studies repeated the same data as other papers, and 10 papers did not present data as MNI or Talairach coordinates.These papers were all excluded from the analysis.As a result, 17 papers were included in the final analysis[9-25].A flow diagram of the retrieval process is shown in Figure 1.There were 261 patients with depression and 273 healthy controls, totally 201 coordinates.

Figure 1 Flowchart of literature retrieval and screening.

Baseline data analysis and quality evaluation

Among the 17 included papers, the emotional stimuli were mainly facial expressions, autobiographical memories and emotional words (Table 1).

Table 1 General information about the included studies

Meta-analysis resuts

Increased activation intensity in brain regions in response to negative emotional stimuli in depressive patients

A total of 17 trials and 52 coordinates were extracted from papers reporting that stimuli associated with sadness increased activation intensity in bilateral amygdala and right inferior frontal gyrus (BA45) in depressive patients compared with controls (Table 2, Figure 2A).

Decreased activation intensity in several brain regions in response to negative emotional stimuli in depressive patients

A total of 13 trials and 50 coordinates were extracted from papers reporting that negative emotional stimuli decreased activation intensity in the following brain regions in depressive patients: bilateral superior frontal gyrus (BA8), left middle frontal gyrus (BA8), paracentral lobule (BA5), right amygdala, right anterior cingulate gyrus (BA24) and left caudate, compared with controls(Table 3, Figure 2B).

Table 2 Brain regions exhibiting increased activation intensity in response to negative emotional stimuli in depressive patients compared with controls

Figure 2 Brain region activation changes in depressive patients in response to different types of stimuli.

Table 3 Brain regions exhibiting decreased activation intensity in response to negative emotional stimuli in depressive patients compared with controls

Decreased activation intensity in brain regions in response to positive emotional stimuli in depressive patients

A total of 10 trials and 85 coordinates were extracted from papers reporting that positive emotional stimuli decreased activation intensity in the following brain regions of depressive patients: the hypothalamus- dorsomedial nucleus, insula (BA13), left parahippocampal gyrus(BA30), right cerebellar anterior lobe, left precentral gyrus (BA6), right middle frontal gyrus (BA10) and left superior frontal gyrus (BA10) compared with control subjects (Table 4, Figure 2C).

Table 4 Brain regions exhibiting decreased activation intensity in response to positive emotional stimuli in depressive patients compared with controls

Increased activation intensity in brain regions in response to positive emotional stimuli in depressive patients

A total of three trials and 14 coordinates were extracted in this category.ALE analysis was not conducted because of the lack of available data.

Overlapping regions of activation in response to different types of emotional stimuli in depressive patients and controls

None of the included studies reported overlapping activated regions in response to different types of emotional stimuli between depressive patients and controls, indicating the independence of different emotional processing pathways (Figure 2D).

DISCUSSION

The current results revealed abnormal activation in patients with depression in the prefrontal cortex, basal ganglia and limbic lobe.This pattern is consistent with Shelineet al[26]proposed that the neuroanatomical pathology of the depression involves a neural network comprising the limbic lobe-cortex-striatum-globus pallidus-thalamus.Negative emotional stimuli were found to decrease activation intensity in the prefrontal cortex and basal ganglia, while increasing activation intensity in the limbic lobe.The present results indicated that different emotional processing pathways are relatively independent.Compared with the meta-analysis reported by Fitzgeraldet al[2], the current analysis has the following novel characteristics: (1) The inclusion of 12 papers published after 2006.(2) A random effects model was utilized for statistical analysis rather than the fixed effects model.The random effects model accounts for variation between studies, and includes experimental sample size in the analysis as a weight modulating factor.Thus, the analysis produces more objective and accurate results.(3) Improved statistical analysis methods require a larger sample size of coordinates, so the data set of findings of increased brain activation intensity in depressive patients in response to positive emotional stimuli could not be submitted to meta-analysis.(4) Increased activation in bilateral amygdala, because the majority of relevant papers published after 2006 have focused on the amygdala[17-18, 22-23].

In animal experiments, Gothardet al[27]reported that positive, negative and neutral stimuli all caused activation in amygdala neurons, and negative stimuli mobilized more cells.The current meta-analysis revealed the amygdala stimulation intensity was maximal in response to facial expressions of fear and disgust.The left amygdala is more strongly associated with language stimulation, while the right is more strongly associated with facial expression[28].Our finding showed negative emotional stimuli to elicit an abnormally high level of activation in the amygdala bilaterally in depressive patients, with a greater activation increase on the right.In contrast,positive stimulation induced no abnormal reaction in the amygdala in depressive patients, in accord with previous reports[26].The right amygdala was activated by facial expressions, while the left amygdala was stimu-lated by both language and sad facial expressions.This result is in accordance with the notion that the left amygdala is involved in language processing, while the right is involved in processing facial expression.Leeet al[19]reported decreased activation intensity in the bilateral amygdala in response to negative stimuli.However, this is an isolated finding, and such low probability events were excluded by ALE in the current study.

Further studies are required to conclusively determine the correlation between different stimuli, different emotional valence and amygdala responses.

The cingulate gyrus is an important region of emotional integration in the core area of the emotional recognition loop and the ventral emotional processing system (comprising the ventral anterior cingulate, amygdala and hypothalamus) and plays an important role in emotional processing[26].The left cingulate gyrus has been reported to exhibit particularly strong activation in response to negative stimuli, and as emotional processing dorsal system, the dorsal anterior cingulate cortex activation decreased in response to negative stimuli[17].The present study revealed that the anterior cingulate (BA24)was activated in a decreasing fashion by negative stimuli.This area contains the dorsal anterior cingulate according to the Talairach and Tournoux neuroanatomical stereotactic atlas.However, further studies are needed to verify the positive findings regarding the abnormal activation in the ventral anterior cingulate in depressive patients.

The prefrontal cortex, a major integration center in the brain, modulates the functions of the limbic system and subcortical tissues.This area is divided into ventral and dorsal regions, functioning as components of the ventral and dorsal emotional processing systems, respectively[26].The present findings revealed that negative stimuli decreased activation intensity in the left prefrontal lobe including left superior frontal gyrus (BA8), left middle frontal gyrus (BA8), and paracentral lobule (BA5).In addition, positive stimuli decreased the activation intensity in the left prefrontal lobe including the left precentral gyrus (BA6), left middle frontal gyrus (BA10) and left superior frontal gyrus (BA10).These results indicate that depressive patients exhibit lateralized abnormal activity in the prefrontal cortex, particularly a decrease in left prefrontal activation.Our findings are consistent with those of Frodlet al[28].Another clinical study reported that infarction in the left prefrontal cortex can induce depression[2].

The present findings also revealed that positive emotional stimulation elicited decreased activation intensity in the hypothalamus and lenticular nucleus, which function as a neural reward system in conjunction with the midbrain tegmentum, nucleus accumbens, orbitofrontal cortex and ventral medial prefrontal cortex.A number of studies have reported that this system is involved in positive emotional responses[1-2,26].The decreased activation of the hypothalamus in response to positive stimuli may be related to abnormal cognition and anhedonia in patients with depression.

The current study involved the following limitations.(1)The number of included papers was relatively small and the variation in the stimulations used in these studies was large.In addition, there was a relatively large amount of heterogeneity in trial results, where facial expressions were the most common type of stimulus used.(2) The majority of papers published after 2006 focused on the amygdala, so the final analysis results may be affected by statistical selection bias.

In summary, this study elucidated abnormal patterns of brain activity in depressive patients during emotional perception processing using a quantitative ALE method and meta-analysis.The results revealed that negative emotional stimuli elicit decreased activation intensity in the prefrontal cortex and basal ganglia in depressive patients, while increasing activation intensity in the limbic lobe (particularly the amygdala) compared to control subjects.The current findings suggest that these brain regions constitute an important neurofunctional loop for regulating brain function, and that abnormal connections among the prefrontal cortex, basal ganglia and limbic lobe may be involved in the neuropathological basis of depression.Increased understanding of the functional characteristics and anatomical connections between brain regions can help to further clarify the mechanisms of neurological pathogenesis underlying depressive mood disorders.

Authors contributions:Chenwang Jin was responsible for trial design and data analysis, and wrote the manuscript.Ming Zhang advised and checked the manuscript, and was in charge of funding.Dan Li and Shaohui Ma were responsible for data analysis and producing figures.Yuan Wang and Min Li collected and screened data.Lihua Liu was responsible for statistical analysis.

Conflict of interest:None declared.

Funding:This study was financially sponsored by the National Natural Science Foundation of China, No.30870686, Science and Technology Research Development Program of Shaanxi Province, No.2009K1601, and Innovation Funds of the First Affiliated Hospital of School of Medicine, Xi’an Jiaotong University, No.2009YK7.

[1]Drevets WC, Price JL, Furey ML.Brain structural and functional abnormalities in mood disorders: implications for neurocircuitry models of depression.Brain Struct Funct.2008;213(1-2):93-118.

[2]Fitzgerald PB, Laird AR, Maller J, et al.A meta-analytic study of changes in brain activation in depression.Hum Brain Mapp.2008;29(6):683-695.

[3]Costafreda SG.Pooling FMRI data: meta-analysis, mega-analysis and multi-center studies.Front Neuroinform.2009;3:33.

[4]Eickhoff SB, Laird AR, Grefkes C, et al.Coordinate-based activation likelihood estimation meta-analysis of neuroimaging data: a random-effects approach based on empirical estimates of spatial uncertainty.Hum Brain Mapp.2009;30(9):2907-2926.

[5]Li H, Chan RC, McAlonan GM, et al.Facial emotion processing in schizophrenia: a meta-analysis of functional neuroimaging data.Schizophr Bull.2010;36(5):1029-1039.

[6]Talairach J, Thournoux P.Co-planar Stereotaxic Atlas of the Human Brain.Stuttgart: Thieme Medical Publishers.1988.

[7]Turkeltaub PE, Eden GF, Jones KM, et al.Meta-analysis of the functional neuroanatomy of single-word reading: method and validation.Neuroimage.2002;16(3 Pt 1):765-780.

[8]Laird AR, Eickhoff SB, Kurth F, et al.ALE Meta-analysis workflows via the brainmap database: progress towards a probabilistic functional brain atlas.Front Neuroinform.2009;3:23.

[9]Kumari V, Mitterschiffthaler MT, Teasdale JD, et al.Neural abnormalities during cognitive generation of affect in treatment-resistant depression.Biol Psychiatry.2003;54(8):777-791.

[10]Canli T, Sivers H, Thomason ME.Brain activation to emotional words in depressed vs healthy subjects.Neuroreport.2004;15(17):2585-2588.

[11]Lawrence NS, Williams AM, Surguladze S, et al.Subcortical and ventral prefrontal cortical neural responses to facial expressions distinguish patients with bipolar disorder and major depression.Biol Psychiatry.2004;55(6):578-587.

[12]Gotlib IH, Sivers H, Gabrieli JD, et al.Subgenual anterior cingulate activation to valenced emotional stimuli in major depression.Neuroreport.2005;16(16):1731-1734.

[13]Keedwell PA, Andrew C, Williams SC, et al.The neural correlates of anhedonia in major depressive disorder.Biol Psychiatry.2005;58(11):843-853.

[14]Surguladze S, Brammer MJ, Keedwell P, et al.A differential pattern of neural response toward sad versus happy facial expressions in major depressive disorder.Biol Psychiatry.2005;57(3):201-209.

[15]Schaefer HS, Putnam KM, Benca RM, et al.Event-related functional magnetic resonance imaging measures of neural activity to positive social stimuli in pre- and post-treatment depression.Biol Psychiatry.2006;60(9):974-986.

[16]Fu CH, Williams SC, Brammer MJ, et al.Neural responses to happy facial expressions in major depression following antidepressant treatment.Am J Psychiatry.2007;164(4):599-607.

[17]Siegle GJ, Thompson W, Carter CS, et al.Increased amygdala and decreased dorsolateral prefrontal BOLD responses in unipolar depression: related and independent features.Biol Psychiatry.2007;61(2):198-209.

[18]Abler B, Erk S, Herwig U, et al.Anticipation of aversive stimuli activates extended amygdala in unipolar depression.J Psychiatr Res.2007;41(6):511-522.

[19]Lee BT, Seok JH, Lee BC, et al.Neural correlates of affective processing in response to sad and angry facial stimuli in patients with major depressive disorder.Prog Neuropsychopharmacol Biol Psychiatry.2008;32(3):778-785.

[20]Fu CH, Mourao-Miranda J, Costafreda SG, et al.Pattern classification of sad facial processing: toward the development of neurobiological markers in depression.Biol Psychiatry.2008;63(7):656-662.

[21]Dichter GS, Felder JN, Smoski MJ.Affective context interferes with cognitive control in unipolar depression: an fMRI investigation.J Affect Disord.2009;114(1-3):131-142.

[22]Walter M, Henning A, Grimm S, et al.The relationship between aberrant neuronal activation in the pregenual anterior cingulate,altered glutamatergic metabolism, and anhedonia in major depression.Arch Gen Psychiatry.2009;66(5):478-486.

[23]Norbury R, Selvaraj S, Taylor MJ, et al.Increased neural response to fear in patients recovered from depression: a 3T functional magnetic resonance imaging study.Psychol Med.2010;40(3):425-432.

[24]Almeida JR, Versace A, Hassel S, et al.Elevated amygdala activity to sad facial expressions: a state marker of bipolar but not unipolar depression.Biol Psychiatry.2010;67(5):414-421.

[25]Peng DH, Jiang KD, Xu YF, et al.The neural substrate of bias mood-processing in patients with first-episode depression: a functional magnetic resonance imaging study with emotional picture stimuli.Zhonghua Jingshen Ke Zazhi.2008;41(2):68-72.

[26]Sheline YI, Barch DM, Donnelly JM, et al.Increased amygdala response to masked emotional faces in depressed subjects resolves with antidepressant treatment: an fMRI study.Biol Psychiatry.2001;50(9):651-658.

[27]Gothard KM, Battaglia FP, Erickson CA, et al.Neural responses to facial expression and face identity in the monkey amygdala.J Neurophysiol.2007;97(2):1671-1683.

[28]Frodl T, Bokde AL, Scheuerecker J, et al.Functional connectivity bias of the orbitofrontal cortex in drug-free patients with major depression.Biol Psychiatry.2010;67(2):161-167.