1. Neuroscience

Anterior insular cortex plays a critical role in interoceptive attention

  1. Xingchao Wang
  2. Qiong Wu
  3. Laura Egan
  4. Xiaosi Gu
  5. Pinan Liu
  6. Hong Gu
  7. Yihong Yang
  8. Jing Luo
  9. Yanhong Wu  Is a corresponding author
  10. Zhixian Gao  Is a corresponding author
  11. Jin Fan  Is a corresponding author
  1. Beijing Tiantan Hospital, Capital Medical University, China
  2. China National Clinical Research Center for Neurological Diseases, China
  3. Capital Normal University, China
  4. Peking University, China
  5. The City University of New York, United States
  6. Icahn School of Medicine at Mount Sinai, United States
  7. The James J. Peter Veterans Affairs Medical Center, United States
  8. National Institute on Drug Abuse, United States
https://doi.org/10.7554/eLife.42265
Share this article
Cite this article
  1. Xingchao Wang
  2. Qiong Wu
  3. Laura Egan
  4. Xiaosi Gu
  5. Pinan Liu
  6. Hong Gu
  7. Yihong Yang
  8. Jing Luo
  9. Yanhong Wu
  10. Zhixian Gao
  11. Jin Fan
(2019)
Anterior insular cortex plays a critical role in interoceptive attention
eLife 8:e42265.
https://doi.org/10.7554/eLife.42265
  2. Download BibTeX
  3. Download .RIS

Abstract

Accumulating evidence indicates that the anterior insular cortex (AIC) mediates interoceptive attention which refers to attention towards physiological signals arising from the body. However, the necessity of the AIC in this process has not been demonstrated. Using a novel task that directs attention toward breathing rhythm, we assessed the involvement of the AIC in interoceptive attention in healthy participants using functional magnetic resonance imaging and examined the necessity of the AIC in interoceptive attention in patients with AIC lesions. Results showed that interoceptive attention was associated with increased AIC activation, as well as enhanced coupling between the AIC and somatosensory areas along with reduced coupling between the AIC and visual sensory areas. In addition, AIC activation was predictive of individual differences in interoceptive accuracy. Importantly, AIC lesion patients showed disrupted interoceptive discrimination accuracy and sensitivity. These results provide compelling evidence that the AIC plays a critical role in interoceptive attention.

https://doi.org/10.7554/eLife.42265.001

Introduction

Our brain consistently receives physiological signals arising from sensory inputs and our body. Although attention toward inputs from the external environment (i.e., exteroceptive attention) has been extensively investigated, the attentional mechanism of the awareness and conscious focus on bodily somatic and visceral signals or responses (i.e., interoceptive attention) has been less studied because of difficulties in its measurement (Brener and Ring, 2016; Ring et al., 2015; Craig, 2002; Craig, 2003; Craig, 2010; Critchley, 2005; Critchley, 2004; Farb et al., 2013a). Previous theories argue that subjective emotions arise from these bodily reactions and visceral experiences (Cannon, 1987; Critchley and Harrison, 2013; Damasio, 1996; Dolan, 2002; Tranel and Damasio, 1991) in which interoceptive awareness plays a critical role. Appropriate attention to bodily states and accurate perception of interoceptive information are essential in emotional awareness and in the maintenance of normal physiological conditions (Craig, 2002; Craig, 2003; Craig, 2010; Critchley, 2005; Wiens, 2005). The link between deficits in interoceptive attention and psychiatric symptoms may also be explained by the James–Lange theory of emotion (Cannon, 1987), the somatic marker hypothesis (Damasio, 1996; Damasio et al., 1991) for the embodied mind mediated by interoception (Garfinkel and Critchley, 2013), and the embodied predictive processing model (Allen et al., 2016; Allen and Friston, 2018; Barrett and Simmons, 2015; Seth, 2013; Seth and Critchley, 2013; Seth et al., 2011).

Recent human studies have emphasized the role of the insula in interoceptive representations (Daubenmier et al., 2013; Farb et al., 2013b; Ronchi et al., 2015). Neuroanatomical evidence, consistent with neuroimaging findings, suggests that the anterior insular cortex (AIC) is an important structure for encoding and representing interoceptive information (Craig, 2002; Craig, 2003; Craig, 2009; Critchley et al., 2004; Stephani et al., 2011). Although the AIC has been recognized as an interoceptive cortex (Craig, 2003; Critchley et al., 2004; Ernst et al., 2014; Singer et al., 2009; Terasawa et al., 2013), these findings remain equivocal because AIC activation seems ubiquitous across a wide range of tasks involving cognition, emotion, and other cognitive processes in addition to interoceptive attention (Allen et al., 2013; Seeley et al., 2007; Uddin et al., 2014; Yarkoni et al., 2011). Therefore, a task that selectively and reliably engages interoceptive attention needs to be employed. In addition, the correlational AIC activation found in functional neuroimaging studies alone does not provide causal evidence for its role in interoceptive attention, leaving the question of whether the AIC is critical in interoceptive attention unanswered. Studying patients with focal lesions in the AIC (Gu et al., 2012; Gu et al., 2015; Ronchi et al., 2015; Starr et al., 2009; Wang et al., 2014; Wu et al., 2019) would thus provide a unique opportunity to examine the necessity of the AIC in this fundamental process.

One challenge to the study of interoceptive attention is the vague nature of interoceptive awareness. According to the classic definition of attention by James (1890), only the contents that are clearly perceived and represented by the mind can be the target of attention. However, most existing tasks measuring interoceptive attention fail to meet this criterion (Ring et al., 2015). In contrast to exteroceptive attention toward external sensory inputs, precise measurements of interoceptive attention are difficult to obtain experimentally because of the imprecise perception of visceral changes, such as heart rate (Brener and Ring, 2016; Paulus and Stein, 2010; Ring et al., 2015; Windmann et al., 1999). Multiple sources of physical information contribute to bodily signals, and most of these sources of somatic feedback cannot be described accurately by mindful introspection in normal physiological states (Ring et al., 2015). This limitation impedes accurate measurement of interoceptive attention and examination of the neural mechanisms underlying this process. To overcome this barrier, a perceivable visceral channel needs to be used.

Breathing is an essential activity for maintaining human life and, more importantly, is an easily perceivable bodily signal. As an autonomous vital movement, breathing can be measured and actively controlled in humans (Daubenmier et al., 2013; Davenport et al., 2007). The unique physiological characteristics of respiration render breath detection an ideal method for measuring interoceptive accuracy and sensitivity (Garfinkel et al., 2015) and for exploring the neural activity underlying interoceptive attention. Thus, we designed a breath detection task to engage interoceptive attention (attention to bodily signals), in which participants were required to indicate whether a presented breathing curve is delayed or not relative to their own breathing rhythm (breath detection task, BDT), in contrast to engaging exteroceptive attention (attention to visual signals), in which participants were required to indicate whether a visual dot stimulus is flashed on the breathing curve (dot flash detection task, DDT). This design enabled us to examine the involvement of the AIC in interoceptive processing in healthy participants and the necessity of the AIC in this processing in patients with AIC lesions.

Basing from previous evidence (e.g., Critchley, 2004), we hypothesized that the AIC is critical for interoceptive attention to reach subjective awareness by integrating information from an individual’s homeostatic state and the external environment. We first conducted functional magnetic resonance imaging (fMRI) studies with two samples to map the neural substrates underlying interoceptive attention to internal bodily signals in contrast to exteroceptive attention to external visual signals in healthy participants while they performed the tasks. We then investigated the necessity of the AIC in interoceptive attention by assessing interoceptive attention in patients with focal AIC lesions (AIC group) in comparison to brain-damaged controls (BDC group, patients with lesions in areas other than insular- or somatosensory-related cortices) and matched neurologically intact normal controls (NC group). We predicted that the AIC is involved in interoceptive attention and that patients with AIC lesions would show deficits in performance on the interoceptive but not exteroceptive attention task.

Results

Behavioral results of the fMRI studies

Performance accuracy (%) and discrimination sensitivity (dʹ) in the BDT were 82.1 ± 14.7% and 2.2 ± 1.1 (mean ± SD) for the first sample, and 74.9 ± 9.6% and 1.6 ± 0.6 (mean ± SD) for the second sample, respectively, which were significantly above the chance levels (50% and 0 for accuracy and dʹ, respectively; For the first sample: t(43) = 14.51, p<0.001, Cohen’s d = 2.18 for accuracy and t(43) = 13.09, p<0.001, Cohen’s d = 2.0 for dʹ, respectively; For the second sample: t(27) = 13.77, p<0.001, Cohen’s d = 2.59 for accuracy and t(27) = 12.89, p<0.001, Cohen’s d = 2.67 for d', respectively), but lower than the DDT accuracy of 87.3 ± 9.8% and dʹ of 2.6 ± 0.8 for the first sample (t(43) = −2.36, p=0.02, Cohen’s d = 0.35 and t(43) = −2.31, p=0.03, Cohen’s d = 0.35, respectively) and accuracy of 80.9 ± 14.7% and dʹ of 2.2 ± 1.1 for the second sample (t(27) = −1.83, p=0.08, Cohen’s d = 0.35 and t(27) = −2.83, p=0.009, Cohen’s d = 0.50, respectively). Participants were slower in terms of reaction time (RT) (only for the first sample) and less biased in BDT than in the DDT (RT: t(43) = 2.89, p=0.006, Cohen’s d = 0.44 for the first sample, and t(27) = 0.6, p=0.55, Cohen’s d = 0.12 for the second sample; β: t(43) = −2.62, p=0.01, Cohen’s d = 0.39 for the first sample, and t(27) = −4.32, p<0.001, Cohen’s d = 0.80 for the second sample) (see Figure 1—figure supplements 1 and 2 for details of the behavior results for the first and second samples, respectively, in accuracy, RT, dʹ, and β. Data were plotted in R using ‘raincloud’ script (Allen et al., 2018a; Allen et al., 2018b); See Table 1 for the statistics of behavioral results for the first and second samples). The split-half reliability of the BDT and DDT were 0.86 and 0.85 for the first sample, and 0.68 and 0.89 for the second sample, respectively.

Figure 1 with 2 supplements see all
Download asset Open asset
Experimental setup, trial structure of the tasks, and stimulus conditions.

(a) The respiratory effort is converted into electronic signal changes using a respiratory transducer, amplified by BIOPAC, digitized using an A/D converter, and sent to the test computer for the final visual display as a dynamic breath curve, with or without a 400 ms delay. (b) This panel shows two trials for the breath detection task (BDT) and flash dot detection task (DDT) runs, respectively. Each trial begins with a 3 s blank display, followed by a 12 s display of respiratory curve presented with or without a 400 ms delay and with or without a 30 ms red dot flashed at a random position on the curve, and ends with a 3 s response window during which participants make a forced-choice button-press response to two alternative choices depending on the block type (BDT or DDT) to indicate whether the feedback curve is synchronous or delayed (for the BDT run) or whether a dot has appeared (for the DDT run). (c) The task represents a 2 × 2 × 2 factorial design with the factors of attention to breath or dot (block design), presence or absence of breath curve delay, and presence or absence of a dot flashed. The dashed line represents the actual breath curve, while the solid line represents the feedback breath curve displayed on the screen.

https://doi.org/10.7554/eLife.42265.002
Table 1
Statistics of behavioral results of the fMRI studies.
https://doi.org/10.7554/eLife.42265.008
First sampleSecond sample
DfTCohen’s dDfTCohen’s d
accuracyintero vs. 0.54314.51***2.182713.77***2.59
intero vs. extero43−2.36*0.3527−1.830.35
dʹintero vs. 04313.09***2.02712.89***2.67
intero vs. extero43-2.31*0.3527-2.83**0.50
βintero vs. extero43−2.31*0.3527−2.83**0.50
RTintero vs. extero432.89**0.44270.60.12

  1. * p<0.05; **p<0.01; ***p<0.001.

For the first sample, the relative interoceptive accuracy was negatively correlated with the subjectively scored difficulty of the BDT relative to the DDT (Pearson r = −0.43, corrected p=0.02, Bayes Factor (BF) = 10.38), but not significantly correlated with the ‘awareness of bodily processes’ subtest of the body perception questionnaire (BPQ) after correction for multiple comparisons (Pearson r = 0.27, corrected p=0.38, BF = 0.86). No significant correlations were observed between relative interoceptive accuracy and subjective emotion experiences, including trait positive affective experience (measured by Positive and Negative Affect Schedule, PANAS) (Watson, 1988) (Pearson r = 0.31, corrected p=0.20, BF = 1.38), anxiety (Pearson r = −0.006, p>0.9, BF = 0.19) or depression score (Pearson r = −0.002, p>0.9, BF = 0.19). For the second sample, however, we did not find significant correlations between relative interoceptive accuracy and scores of questionnaires (awareness of bodily processes: Pearson r = −0.17, corrected p>0.9, BF = 0.33; trait positive affective experience: Pearson r = 0.12, corrected p>0.9, BF = 0.27; anxiety: Pearson r = 0.29, corrected p=0.56, BF = 0.69; depression: Pearson r = 0.03, corrected p>0.9; note that we did not collect subjective rating of task difficulty in the second sample). In addition, we also calculated correlation coefficients between task performance and questionnaires by pooling the two samples (See Table 2 for Pearson correlation strength and Bayesian tests between all behavioral measures in the first sample, the second sample, and across the two samples).

Table 2
Pearson correlation coefficients (and Bayes Factors) between the behavioral measurements for the first, the second, and across the two samples.
https://doi.org/10.7554/eLife.42265.009
Relative accuracySubjective difficultyBPQPositive PANASHAMABDI
Relative accuracy-
1st sampleSubjective difficulty−0.43**
(10.38)		-
BPQ0.27
(0.17)		−0.15
(0.29)		-
Positive PANAS0.31
(1.38)		−0.04
(0.19)		−0.006
(0.19)		-
HAMA−0.006
(0.19)		−0.14
(0.28)		0.25
(0.69)		−0.12
(0.25)		-
BDI−0.002
(0.19)		−0.004
(0.19)		0.16
(0.32)		−0.06
(0.20)		0.70***
(>100)		-
Relative accuracy-
2nd sampleSubjective difficulty--
BPQ−0.17
(0.33)		--
Positive PANAS0.12
(0.27)		-0.07
(0.25)		-
HAMA0.29
(0.69)		-0.40
(1.90)		−0.034
(0.24)		-
BDI0.034
(0.24)		-0.075
(0.25)		−0.43
(2.84)		0.47*
(4.96)		-
Relative accuracy-
1st + 2nd samplesSubjective difficulty--
BPQ0.06
(0.17)		--
Positive PANAS0.25
(1.16)		-0.03
(0.15)		-
HAMA0.12
(0.25)		-0.31*
(4.91)		−0.09
(0.20)		-
BDI0.008
(0.15)		-0.14
(0.28)		−0.20
(0.56)		0.60***
(>100)		-

  1. * corrected p<0.05; ** corrected p<0.01; *** corrected p<0.001; value in brackets represents Bayes factor. BPQ, body perception questionnaire; PANAS, positive and negative affective schedule; HAMA, Hamilton anxiety scale; BDI, Beck depression inventory.

Imaging results of the whole brain analysis of the first fMRI study

Main effects of interoceptive attention and feedback delay, and their interaction

The main effect of interoceptive attention, compared to exteroceptive attention (BDT vs. DDT), was associated with enhanced activity in the cognitive control network (Fan, 2014; Wu et al., 2015; Xuan et al., 2016), including the AIC, the dorsal anterior cingulate cortex (ACC) and the supplementary motor area (SMA), and the superior frontal and the parietal cortices (the frontal eye field, FEF; and the areas near/along the intraparietal sulcus; Figure 2a, Table 3). In addition, this contrast revealed significantly less activation, or deactivation, in the core regions of the default mode network (Raichle et al., 2001), including the the ventral medial prefrontal cortex, the middle temporal gyrus (MTG), and the posterior cingulate cortex.

Figure 2
Download asset Open asset
Main effects and the interaction effect of the whole brain analysis of the first sample.

(a) Main effect of interoceptive vs. exteroceptive attention contrast (BDT vs. DDT). (b) Main effect of breath curve feedback condition (delayed curve vs. non-delayed curve). (c) Interaction between attention type and breath-curve feedback condition ([delayed – non-delayed] BDT – [delayed – non-delayed] DDT). Here we showed the left AIC for the visualization of the seed for the ROI analysis in the second fMRI sample, although the cluster with 210 voxels did not survive GRF correction. Red color represents an increased activation; Blue color represents a decreased activation. (d) Activation of the left and the right AIC under the four task conditions, and the pattern of the interaction.

https://doi.org/10.7554/eLife.42265.010
Figure 2—source data 1

CSV file containing data for Figure 2d.

https://doi.org/10.7554/eLife.42265.011
Table 3
Activation and deactivation of the brain regions involved in interoceptive attention (interoception – exteroception).
https://doi.org/10.7554/eLife.42265.012
MNI
RegionL/RBAXYZTZK
Positive
Cerebelum crus IL−30−70−2413.02Inf.73834
Middle occipital gyrusR1932−682211.99Inf.
Cerebelum crus IIL−20−78−4811.727.80
Inferior frontal gyrusR4452142411.247.63
Inferior parietal lobuleR4036−484411.197.62
Inferior parietal lobuleL40−38−464210.417.32
Postcentral gyrusR246−405410.297.27
Supramarginal gyrusR4048−344210.007.15
Superior occipital gyrusR722−72469.997.15
Cerebelum VIIBL−32−70−529.787.06
Superior parietal lobule (Intraparietal sulcus)R716−78529.697.02
Cerebelum VIIIR22−74−509.616.99
Middle frontal gyrusL46−4450129.206.80
Middle frontal gyrusR464242249.166.78
Supplementary motor areaR684768.926.68
Inferior occipital gyrusR3752−66−128.686.56
Cerebelum crus IIR2−76−368.666.56
Middle occipital gyrus (Intraparietal sulcus)R1932−76348.586.52
ThalamusR18−20208.556.50
Inferior temporal gyrusR2056−38−208.416.43
Inferior frontal gyrusR454438128.316.38
Superior parietal lobule (Intraparietal sulcus)L7−20−72468.216.33
Supplementary motor areaL6-2-4748.086.27
Inferior frontal gyrusL44−5412268.076.26
CaudateR16-8247.896.17
Anterior cingulate cortexR32218447.786.12
Vermis-2−74−127.766.10
Middle frontal gyrusR465014407.756.10
Middle frontal gyrusL46−4034347.726.08
Supramarginal gyrusL40−60−36287.475.95
Middle frontal gyrusR6282487.015.69
Anterior insular cortexR342046.985.68
Postcentral gyrusL2−62−26366.875.62
Inferior frontal gyrusL6−528126.845.59
Superior frontal gyrusL6−264666.735.53
Middle occipital gyrus (Intraparietal sulcus)L7−24−66366.665.49
Lingual gyrusL18−18−90−186.615.46
Superior parietal lobuleL1−24−44726.555.42
CaudateL-82246.455.37
Precentral gyrusL6−402566.235.23
Superior occipital gyrusL18−22−92286.205.21
Middle occipital gyrusL18−24−94166.095.14
Middle occipital gyrusR1830−86166.095.14
Fusiform gyrusL37−46−46−225.824.97
Anterior insular cortexL−302085.504.76
CuneusL190−88345.224.57
Superior parietal lobuleL5−18−60665.184.54
Fusiform gyrusR3744−32−204.964.39
Negative
Anterior cingulate cortexR32438-47.475.953232
Anterior cingulate cortexL32-638-47.105.94
Superior frontal gyrusL9−1638545.975.07
Medial superior frontal gyrusR321052205.334.65
Medial superior frontal gyrusL32-850265.324.63
Middle frontal gyrusL8−2430565.124.50
Superior frontal gyrusL9−2032484.544.08
PrecuneusL23−10−44406.455.37819
PrecuneusR236−60244.243.85
Middle temporal gyrusL21−60−10−145.895.02787

Activation in the AIC, the middle frontal gyrus, the SMA, and the temporal parietal junction was associated with the main effect of feedback delay (Figure 2b, Table 4). The regions showing the main effect of feedback delay also showed the interaction effect between attentional focus (interoceptively in the BDT and exteroceptively in the DDT) and feedback (with and without delay) (Figure 2c, Table 5). The task-induced responses extracted from the bilateral AIC, defined by the attention by feedback interaction map, shows the activation pattern under different task conditions (Figure 2d). In specific, (1) the bilateral AIC demonstrated greater activation during interoceptive processing (BDT) than during exteroceptive processing (DDT), irrespective of the feedback type; (2) the delayed trials induced greater activation in the bilateral AIC in comparison to the non-delayed trials only during interoceptive processing (BDT). The evidence of this interaction effect in the AIC suggests that the AIC was actively engaged in interoceptive processing.

Table 4
Activation and deactivation of the brain regions involved in feedback delay (delay – non-delay).
https://doi.org/10.7554/eLife.42265.013
MNI
RegionL/RBAXYZTZK
Positive
Anterior insular cortexR3026-45.264.60618
Inferior frontal gyrusR45422284.403.98
CaudateR82444.293.90
Inferior parietal lobuleL40−38−54425.234.58598
Angular gyrusR3944−44304.994.411317
Inferior parietal lobuleR4056−54444.173.80
Middle frontal gyrusR6348464.784.26780
Middle frontal gyrusR93418344.744.23
Middle frontal gyrusR463428324.323.92
Negative
Lingual gyrusL17−10−78-46.215.22443
Table 5
Activation of brain regions related to the interaction between interoceptive attention and feedback delay ([delayed – non-delayed] interoception – [delayed – non-delayed] exteroception).
https://doi.org/10.7554/eLife.42265.014
MNI
RegionL/RBAXYZTZK
Positive
Anterior insular cortexR282805.524.77516
Inferior frontal gyrusR474026−104.664.17
Middle frontal gyrusR94014405.364.672330
Supplementary motor areaR8422545.194.55
Anterior cingulate cortexR32636385.124.5
Superior frontal gyrusR8630444.714.21
Inferior frontal gyrusR454622164.504.05
Middle frontal gyrusR6344524.273.88
Supplementary motor areaL6−128523.643.38
Anterior cingulate cortexR321030283.493.25
Supramarginal gyrusR4054−46264.914.351748
Middle temporal gyrusR2166−32−104.704.20
Inferior parietal lobuleR1960−48424.564.10
Superior temporal gyrusR4258−40164.494.04

Correlation between interoceptive accuracy and AIC activation

Voxel-wise regression analysis revealed the relationship between the interoceptive task-induced activation strength (map of the interaction contrast) and participants’ interoceptive accuracy (performance accuracy in the BDT), with exteroceptive accuracy (performance accuracy in the DDT) controlled as a covariate. Higher interoceptive accuracy was associated with greater interaction effect of the bilateral AIC (and MTG) across participants (Figure 3a, Table 6). The AIC activation during the interoceptive processing involved attending to physiological signals and matching bodily signals to external visual input, which predicted individual differences in interoceptive attention (see Figure 3b for the illustration).

Figure 3
Download asset Open asset
Relationship between brain activation and behavioral performance across participants.

(a) This was revealed in a regression analysis of contrast images for the interaction between interoceptive attention deployment (BDT vs. DDT) and breath curve feedback condition (delayed vs. no-delayed), with performance accuracy on interoceptive and exteroceptive tasks as regressor-of-interest and covariate, respectively. AIC, anterior insular cortex; MTG, middle temporal gyrus. (b) Correlational patterns between the interaction effect of bilateral AIC activation and relative interoceptive accuracy. Data were normalized as z-scores.

https://doi.org/10.7554/eLife.42265.015
Figure 3—source data 1

CSV file containing data for Figure 3b.

https://doi.org/10.7554/eLife.42265.016
Table 6
Relationship between the interaction effect ([delayed – non-delayed] interoception – [delayed – non-delayed] exteroception) of the brain and behavioral performance (interoceptive accuracy) across participants.
https://doi.org/10.7554/eLife.42265.017
MNI
RegionL/RBAXYZTZK
Positive
Middle temporal gyrusR2054−20−103.853.53232
Middle temporal gyrusL22−48−24-23.693.41170
Anterior insular cortexL−4212-63.643.37168
Anterior insular cortexR4216-63.413.18119
Angular gyrusR2258−50263.102.92128

Functional and effective connectivity of the AIC

Psychophysiological interaction (PPI) analysis showed augmented connectivity between the right AIC (as the seed) and the SMA/ACC, the FEF, the inferior frontal gyrus (IFG), and the postcentral gyrus (PoCG) during interoceptive (versus exteroceptive) attention (BDT vs. DDT) in contrast to the reduced connectivity between the right AIC and visual cortices (VCs) modulated by interoceptive attention (Figure 4a, Table 7). This result indicates that an increase in activation in the right AIC was associated with a greater increase in activation in the FEF, the IFG, and the PoCG and a greater decrease in activation in the VCs under interoceptive attention compared with exteroceptive attention (Figure 4b). Similar PPI results were obtained when the left AIC was used as the seed (Figure 4—figure supplement 1).

Figure 4 with 2 supplements see all
Download asset Open asset
PPI and DCM results of the first fMRI sample.

(a) Regions showing positive (red) and negative (blue) associations with AIC activation modulated by interoceptive attention relative to exteroceptive attention (BDT vs. DDT). (b) An increase in activation in the right AIC was associated with an increase in activation in the postcentral gyrus (PoCG) and a decrease in activation in the visual cortex (VC, V2/3) under the condition of interoceptive attention compared with exteroceptive attention. (c) Five base models generated by specifying possible modulations of interoceptive and exteroceptive attention (BDT and DDT) on the four endogenous connections between ROIs. The model surrounded by a rectangle in dashed-line indicates the winning model out of 52 variant models revealed by random-effects Bayesian model selection (BMS). (d) Intrinsic efferent connection from the AIC to the PoCG was significant. The modulatory effect of interoceptive attention (BDT) on the connection from the AIC to the PoCG was significant. The modulatory effect of exteroceptive attention (DDT) on the connection from AIC to V2/3 was significant (uncorrected).

https://doi.org/10.7554/eLife.42265.018
Figure 4—source data 1

CSV file containing data for Figure 4b.

https://doi.org/10.7554/eLife.42265.021
Table 7
Positive and negative psychophysiological interaction effects with the right AIC as the seed.
https://doi.org/10.7554/eLife.42265.022
MNI
RegionL/RBAXYZTZK
Positive
Inferior frontal operculumR44528267.495.965895
Precentral gyrusR65810366.715.52
Insula cortexR380146.355.30
PutamenR208106.335.29
Rolandic operculumR48484106.015.09
CaudateR81045.865.00
Inferior frontal gyrusR454236104.353.94
Postcentral gyrusR4358−16326.956.552078
Supramarginal gyrusR266−22346.045.11
Superior temporal gyrusR4262−32205.284.61
Precentral gyrusL6−5810306.895.6311155
PutamenL−2010126.045.11
Supplementary motor areaL6-8-4645.905.02
CaudateL-81625.414.70
Triangle Inferior fronal gyrusL48−3832245.214.56
Superior temporal gyrusL44−48−42245.194.55
Insula cortexL−36-285.194.55
Supplementary motor areaR644645.194.55
Supramarginal gyrusL2−56−28405.134.50
Superior frontal gyrusL6−24-2584.734.22
Postcentral gyrusL3−56−20344.534.07
Middle frontal gyrusL6−28-8524.484.04
Middle temporal gyrusR3748−6085.444.72569
Cerebelum VIIbL−16−74−484.954.38427
Cerebelum VIIIL−24−66−524.754.24
Negative
CuneusL17−10−96167.305.855904
CuneusR1814−90286.805.40
Lingual gyrusR1814−62-26.055.11
Lingual gyrusL18−18−74-85.264.60
CalcarineL180−76185.114.49
Fusiform gyrusL18−24−80−164.954.38
CalcarineR1720−5464.724.22
Cerebelum Crus IL−38−78−184.373.95
Middle occipital gyrusL18−16−86-44.223.84

On the basis of the PPI results, VCs of the right V2/3 (x = 14, y = −90, z = 28 as indicated by negative PPI) and the right PoCG (x = 58, y = −16, z = 32 as indicated by positive PPI) were included in the dynamic causal modeling (DCM) model. Data from one participant were excluded because significant activation in the V2/3 region of interest could not be identified. For model comparison, random-effects (RFX) Bayesian model selection (BMS) indicated that the winning model (with an exceedance probability of 29.84%) was the one with the modulatory effects of interoceptive and exteroceptive attention (BDT and DDT) exerting on the connection from the AIC to the PoCG and from the AIC to V2/3 (Figure 4c and Figure 4—figure supplement 2). The BMS indicated that interoceptive and exteroceptive attention were achieved through modulating the top-down connectivity from the AIC to these two sensory cortices.

We performed parameter inference by using Bayesian model averaging (BMA), which considers uncertainty by pooling information across all models in a weighted fashion (Stephan et al., 2010). For BMA (Figure 4d), the modulatory effect of interoceptive attention (BDT) was significant on the connection from the AIC to the PoCG (t(42) = 4.85, Bonferroni corrected p<0.001). The modulatory effect of exteroceptive attention (DDT) on the connection from the AIC to the V2/3 was significant without correction (t(42) = 2.25, uncorrected p=0.03). The BMA results were consistent with the winning model selected by model comparison and the PPI results: the modulatory effect from the AIC to the PoCG was driven by interoceptive attention (BDT), whereas the modulatory effect from the AIC to the V2/3 was driven by exteroceptive attention (DDT). In addition, the BMA results highlighted the importance of the intrinsic efferent connection from the AIC to the PoCG in the network (t(42) = 3.61, Bonferroni corrected p=0.01).

Region-of-interest (ROI) analysis results of the second fMRI study

The interaction between attentional focus (interoceptively in BDT and exteroceptively in DDT) and feedback (with and without delay) was significant in both left and right AICs (left: F(1, 27)=6.12, p=0.020; right: F(1,27) = 5.88, p=0.022; Figure 5a), which confirmed the interaction effect in the bilateral AIC revealed by whole brain analyses of the first sample. The main effect of attentional focus (BDT vs. DDT) was significant in the right AIC with greater activation during the BDT than during the DDT (F(1, 27)=4.20, p=0.05) but not significant in the left AIC (F(1, 27)<1, p=0.51). The main effect of the feedback was not significant in either left or right AIC (left: F<1; right: F<1). In addition, similar to the results of the first sample, we found a significant correlation between the interaction effect of both left and right AICs and relative interoceptive accuracy (left: Pearson r = 0.32, p=0.050, one-tailed; right: Pearson r = 0.42, p=0.014, one-tailed; Figure 5b). In addition, we examined the pattern of the respiratory volume under BDT and DDT (see Figure 5—figure supplement 1). Despite the difference in the respiratory volume between interoceptive and exteroceptive conditions (BDT and DDT was significant, F(1,27) = 15.88, p<0.001), this difference was canceled out for the interaction effect (F < 1). These results further illustrated that the interaction effect in the AIC is not subject to the confounding of breathing effort difference between the two tasks.

Figure 5 with 3 supplements see all
Download asset Open asset
ROI results of the second fMRI sample.

(a) ROI analysis of the parameter estimates of the left and the right AIC under the four experimental conditions. Raincloud plots were used for visualization. (b) Correlation between the interaction effect of bilateral AIC and relative interoceptive accuracy. The values of the variable in b were normalized as z-scores.

https://doi.org/10.7554/eLife.42265.023
Figure 5—source data 1

CSV file containing data for Figure 5b.

https://doi.org/10.7554/eLife.42265.027

The whole brain analysis of the second fMRI sample showed significant overlap between the activations without and with physiological correction for the main and the interaction effects (see Figure 5—figure supplement 2). We further checked how much physiological noise impacted AIC activation by comparing the contrast maps without and with physiological correction at an extremely permissive threshold (p<0.05 uncorrected). The difference in signals of the AIC between the analyses with and without physiological corrections was only evident for the main effect of interoceptive vs. exteroceptive attention (BDT vs. DDT) but not for the interaction contrast, confirming that the interaction effect of the AIC was not significantly impacted by the physiological noises (see Figure 5—figure supplement 3). Altogether, these ROI results from the second sample confirmed that the AIC was actively engaged in interoceptive processing.

Lesion study results: the necessity of the AIC in interoceptive attention

Figure 6 shows the insular lesion overlap for the AIC patient group. The area with the most overlap was identified as the AIC according to the literature (Kurth et al., 2010; Naidich et al., 2004). We found a significant interaction effect between group (AIC, BDC, and NC) and task (BDT and DDT) in performance accuracy (F(2,21) = 5.19, p=0.015) and discrimination sensitivity (dʹ) (F(2,21) = 4.77, p=0.023). Planned simple comparisons were conducted between groups for each task. For the BDT, patients with AIC lesions had significantly lower performance accuracy (58%, t(13) = −3.47, p<0.001, BF = 14.71 compared with NC; t(8) = −2.35, p=0.009, BF = 3.95 compared with BDC) (Figure 7a) and discrimination sensitivity (dʹ) compared with the NCs and BDCs groups (t(13) = −3.62, p<0.001, BF = 13.78 compared with NC; t(8) = −2.22, p=0.013, BF = 3.40 compared with BDC) (Figure 7b), indicating diminished interoceptive attention. However, we did not find significant difference in accuracy between the NC and BDC groups (t(8) = 0, p=0.3; dʹ: t(8) = 0.112, p=0.23). For the DDT, the patients with AIC lesions did not show significant abnormalities in performance accuracy (AIC vs. NC: t(9) = 0.18, p=0.22, BF = 0.38; AIC vs. BDC: t(7) = −0.99, p=0.10, BF = 0.98), and in dʹ (AIC vs. NC: t(9) = 0.18, p=0.22, BF = 0.38; AIC vs. BDC: t(7) = −0.83, p=0.12, BF = 0.85) compared with the NC and BDC groups. We did not find significant interaction effect on β (F <1, p=0.65) (Figure 7d–f). A summary of the statistical results of the lesion study is provided in Table 8. Our results demonstrated significant impairment in discrimination ability when attending to bodily signals, but not to external visual input, in patients with AIC lesions.

Figure 6
Download asset Open asset
Reconstruction of anterior insular cortex lesions of six patients.

Red color indicates 100% overlap. Left lesions were flipped to the right side to map the lesion overlap.

https://doi.org/10.7554/eLife.42265.028
Figure 7
Download asset Open asset
Behavioral results of the lesion study.

(a, b, c) the interoceptive performance on the BDT, and (d, e, f) the exteroceptive performance on the DDT. On the BDT, patients with AIC lesions had significantly lower performance in accuracy and dʹ compared with the NC and BDC groups but did not show significant alteration in β during the BDT. On the DDT, patients with AIC lesions did not show significant abnormality in performance in accuracy, dʹ, and β compared with either the NC or BDC groups. NC, normal control; BDC, brain damage control. Dashed line: chance level. * p < 0.05; ** p < 0.01; *** p < 0.001.

https://doi.org/10.7554/eLife.42265.030
Figure 7—source data 1

CSV file containing behavioral data for lesion study.

https://doi.org/10.7554/eLife.42265.031
Table 8
Statistics of the results of the lesion study.
https://doi.org/10.7554/eLife.42265.032
Accuracydʹ
TBFTBF
BDTAIC vs. NC−3.47***14.71−3.62***13.78
AIC vs. BDC−2.35**3.95−2.22*3.40
BDC vs. NC00.420.110.43
DDTAIC vs. NC0.180.380.180.38
AIC vs. BDC−0.990.98−0.830.85
BDC vs. NC1.74*0.821.460.69

  1. * p<0.05; **p<0.01; ***p<0.001; one- tailed; BF, Bayes factor.

Discussion

Using fMRI, we showed that the AIC is involved in interoceptive attention towards respiration, with the underlying connectivity between the AIC and the somatosensory cortex and visual areas modulated by interoceptive and exteroceptive attention, respectively. Notably, we confirmed the necessity of the AIC in supporting interoceptive attention by showing reduced behavioral performance on the interoceptive task in patients with focal AIC lesions. Thus, this study demonstrates that the AIC plays a critical role in interoceptive attention.

The necessity of the AIC in interoceptive attention

Previous functional neuroimaging studies have shown that the insula is activated by autonomic arousal and emotional reactions (Craig, 2002; Craig, 2003; Critchley et al., 2004) and emphasized the central role of the insula in interoceptive awareness. The achievement of interoceptive awareness depends on the integration of afferent bodily signals with higher-order contextual information attributable to the AIC (Craig, 2002; Craig, 2009; Critchley, 2005; Damasio et al., 2000; Mutschler et al., 2009). In this study, the increase in neural activation in the AIC and other related brain structures when focusing on breath rhythm indicates that the AIC supports attention toward bodily signals. Most importantly, participants’ performance accuracy on the interoceptive task was significantly correlated with the activation of the AIC, further demonstrating the involvement of the AIC in interoceptive attention.

Anatomically, the insula receives thalamo-insular projections of the interoceptive pathways (Craig, 2002). The AIC encodes subjective feelings (Craig, 2003; Craig, 2009; Flynn, 1999) and is critical for instantaneous representation of the state of the body (Allen et al., 2016; Allen and Friston, 2018; Cao et al., 2014; Gu et al., 2015). During the BDT, this is achieved by attending to bodily signals (i.e., breath rhythm) and matching them to external visual feedback (i.e., the breath curve). The present results provide additional support to previous finding that the activation of the right AIC is related to accuracy in sensing the timing of one’s bodily signals, for example, heartbeat (Critchley et al., 2004). Consistent with the notion that the AIC contributes to accurate perception of bodily states (Bechara and Naqvi, 2004), the insula works as a hub to convey bodily information into internal feelings for maintaining homeostasis and to mediate the representations of visceral states that link to the representations of the external world (Farb et al., 2013a).

Our finding of a critical role of the AIC in interoceptive attention fits with a recent predictive coding account of the brain (Bastos et al., 2012; Friston and Kiebel, 2009; Rao and Ballard, 1999), which suggests that the brain actively tries to predict possible future states and to minimize the difference between actual and predicted states. In the context of interoceptive and embodied predictive coding (Allen and Friston, 2018; Gu et al., 2013), previous studies hypothesized that interoceptive predictions are computed within a network of brain regions with the AIC as the key node (Allen et al., 2016; Barrett and Simmons, 2015; Seth, 2013). Empirical evidence that directly supports this computational role of the insula is still rare. One such study using a tactile oddball paradigm and DCM of fMRI time series demonstrated that the AIC is the only region, among a network of body-related brain regions, that shows a reciprocal increase in connectivity with the somatosensory cortex (Allen et al., 2016). Our finding is consistent with this previous study and extends the role of the AIC in predictive coding to breathing-related interoception by using both fMRI and lesion approaches.

The role of the AIC in interoceptive attention identified by the fMRI studies was augmented by the data from patients with focal damage to the insula. Relative to non-insular lesion patients and healthy controls, AIC lesions led to a deficit in accuracy and sensitivity of interoceptive attention. These findings provide causal evidence demonstrating the critical role of the AIC in interoceptive attention. Traditionally, the insular cortex is considered as a limbic sensory region that participates in the intuitive processing of complex situations (Augustine, 1996; Butti and Hof, 2010; Menon and Uddin, 2010) by integrating the ascending visceromotor and somatosensory inputs with attention systems via intrinsic connectivity to identify and respond to salient stimuli (Menon and Uddin, 2010; Uddin, 2015). The AIC, in particular, is a node that mediates cognitive processes including bottom-up control of attention (Corbetta et al., 2002; Corbetta et al., 2008; Wu et al., 2015) and conscious detection of signals arising from the autonomic nervous system (Craig, 2002; Critchley, 2004). Therefore, the behavioral deficit of interoceptive attention in patients with AIC lesions is due to the disruption in the integration of the somatic and visceral inputs with the abstract representation of the present internal state (i.e., the saliency of a certain type of signals). Consequently, it leads to failure in discriminating whether the displayed respiratory curve is different from internal states.

Most previous lesion studies indicated interoceptive deficits with AIC lesions (Critchley and Garfinkel, 2017; García-Cordero et al., 2016; Ibañez et al., 2010; Ronchi et al., 2015; Starr et al., 2009; Terasawa et al., 2015; Wang et al., 2014), supporting the conclusion that interoceptive accuracy relies on a widely distributed network with the insular cortex as a key node (Craig, 2002; Critchley and Harrison, 2013). However, the preservations of interoceptive processing (Khalsa et al., 2009) and self-awareness across a large battery of tests (Philippi et al., 2012) were documented in one patient with bilateral insular damages. These studies are mostly based on subjective report focusing on ‘feeling/awareness’ (Khalsa et al., 2009) that might be compensated by other brain structures such as the brainstem and subcortical structures, for example, nucleus tractus solitaries, the parabrachial nucleus, area postrema and hypothalamus (Damasio et al., 2013), frontal and temporal regions, for example, amygdala, superior temporal gyrus, and temporal pole (García-Cordero et al., 2016; Shany-Ur et al., 2014). In the current study, the BDT challenged interoceptive attention that requires the integration of interoceptive awareness and accuracy. Our examination of interoceptive attention in patients with focal AIC lesions showed that lesions of the AIC were associated with a deficit in performance, indicating that the AIC is critical in supporting the precision of interoceptive processing.

Mechanisms of the AIC in relation to interoceptive attention

Interoceptive attention is the mechanism that coordinates the processing of bodily signals and higher-level representation of that information. The AIC reportedly encodes and represents bodily information (e.g., visceral states) and transmits this information to other neural systems for advanced computations in conscious perception and decision-making (Bechara and Naqvi, 2004; Flynn, 1999; Gu and FitzGerald, 2014). The AIC is a key node of the large-scale network that detects information from multiple sources including objective visceral signals, generates subjective awareness (Craig, 2009; Gu et al., 2012; Gu et al., 2015; Kleckner et al., 2017; Seeley et al., 2007), and responds to the switch between networks that supports internal oriented processing and cognitive control (Menon, 2011; Menon and Uddin, 2010; Sridharan et al., 2008). Supporting this argument, we showed that the AIC is intrinsically connected to the somatosensory area of the PoCG and that this connection is positively modulated by interoceptive attention (relative to exteroceptive attention).

Other higher-level areas, for example, the ACC/SMA, FEF, and IFG of the so-called cognitive control network (CCN) (Fan, 2014; Wu et al., 2018), are also involved in the interoceptive process. This is supported by the results of the enhanced functional connectivity between the AIC and these regions. Both somatosensory afferents and a network that includes the AIC and the ACC are possible pathways of interoceptive attention (Khalsa et al., 2009). The AIC may play a central role in integrating sensory signals from the PoCG and visual cortex and sends top-down signals that guide sensation and perception through a dynamic interaction with sensory or bottom-up information. Somatosensory information concerning the internal state of the body is conveyed through the PoCG, as well as the visual signals in V2/3 containing the majority of external information. The top-down modulation of the AIC in interoceptive attention is accomplished by augmenting the efferent signals to the somatosensory cortices. This result is consistent with the argument that a first-order mapping of internal feeling is supported by insular and somatosensory cortices (Damasio, 2003) and that somatosensory information critically contributes to interoceptive attention (Khalsa et al., 2009).

In the BDT, interoceptive attention reflects a combination of the attention to the internal bodily signal (i.e., the breath) and the external visual stimulus (i.e., the curve). To coordinate perceptual processing, the AIC may distribute and balance the processes of external and internal information. The winning model and parameter inference from DCM provide evidence that interoceptive attention is achieved mainly by modulating the connectivity between the AIC and the somatosensory areas (PoCG), while exteroceptive attention is primarily modulated via the connectivity between the AIC and V2/3. We propose that the dynamic adjustment of the connectivity of the AIC to sensory cortices is the foundation of interoceptive attention for bodily signals, which is critical for homeostatic regulation, and of exteroceptive attention for external objects or inputs.

Interoceptive task in the respiratory domain

Although the neural correlates of interoceptive awareness have been studied by other tasks, such as the heartbeat detection task (Bechara and Naqvi, 2004; Critchley et al., 2004; Khalsa et al., 2009; Ring et al., 2015), the error rate arising from the difficulty in heartbeat counting or non-sensory process confounds are inherent to these cardioception designs (Kleckner et al., 2015; Ring et al., 2015). In contrast to cardioception, breath can be clearly perceived and autonomously controlled. This feature enabled us to design a task measuring interoceptive attention, which requires that the target of interoception be clearly and vividly perceivable by our consciousness. The positive correlations between objective interoceptive accuracy during the BDT and subjectively scored difficulty of interoceptive task relative to exteroceptive task (i.e., interoceptive awareness) further demonstrates that the BDT is valid in assessing interoceptive attention. We developed the respiratory interoception task as a non-intrusive measurement with low cognitive load that is more practical for patients with focal brain damage than the demanding cardiac interoception tasks.

The BDT may not represent a pure probe of interoception because respiratory processes can also be tracked using exteroceptive and proprioceptive information. Thus, the participants possibly relied on a mix of interoceptive, exteroceptive, and proprioceptive information to perform the task. In our design, we included the DDT for a measure of exteroception so that the cognitive subtraction of DDT from BDT leaves the interoceptive and proprioceptive processing components of interoception (Gu et al., 2013). In the BDT, the delayed manipulation in our study was fixed to 400 ms, approximately 1/10 of an average cycle of normal healthy people (i.e., 3–4 s/cycle). This delayed duration can be manipulated according to each individual’s respiratory cycle in an effort to control subjective task difficulty across participants.

Interoceptive attention

Depending on the source of information, attention can be categorized into (1) interoceptive attention, which is directed toward bodily signals such as somatic and visceral signals (e.g., in a heartbeat detection or counting task); (2) exteroceptive attention, which is directed toward primary sensory inputs from outside (e.g., visual and auditory stimuli); and (3) executive control of attention, which coordinates thoughts and actions (e.g., in Color Stroop, flanker, and working memory tasks; see review Fan, 2014). Although extensive studies have focused on the attentional modulation of sensory and perceptual inputs and on the executive control of attention, interoceptive attention is difficult to study because the vast majority of intrinsic visceral activity, except breath effort, cannot be clearly perceived under normal conditions. Using the BDT to examine attentional deployment toward breath effort enabled us to reveal the neural mechanism of interoceptive attention. In general, the perceptible, controllable, measurable, and autonomous features of breathing guaranteed more accurate and reliable measurement of individual differences in interoceptive ability.

As a type of interoceptive attention that could be clearly perceived and autonomously controlled, breath plays a potentially important role in generating and regulating emotion. For example, mindfulness meditation, which is now well known for its role in emotion regulation and mental health (Khoury et al., 2015), can be viewed as a practice involving interoceptive attention. One of its primary methods is to bring one’s attention (the processing) and then awareness (the outcome) to the current experiences of the movement of the abdomen when breathing in and out or the breath as it enters and exits the nostrils. Considering the revealed neural mechanisms of interoceptive attention in this study, we predict that the AIC plays an important role in meditation. Findings that meditation experience is associated with increased gray matter thickness in the AIC (Lazar et al., 2005) and increased gyrification (increase in folding) of the AIC (Luders et al., 2012) support this prediction. Meditation training may enhance interoceptive attention to focus on bodily signals so that accurate feelings can be generated based on the bodily responses, meanwhile the mind can be released from an intensive involvement of exteroceptive and executive control of attention (the internal attention for the coordination of thought processing) that consumes the majority of mental resources.

Conclusion

This study provided important evidence of the involvement of the AIC in interoceptive attention by the fMRI studies and further demonstrated that the AIC is critical for the process by the lesion study. The converging evidence also suggests that interoceptive attention is achieved through top-down modulation from the AIC to the somatosensory and sensory cortices. In addition, the implementation of the interoceptive task extends the research on interoceptive processing into the respiratory domain with the validity and reliability demonstrated. It may have significant applications in studying issues related to interoceptive attention in patients with neuropsychiatric disorders, such as anxiety (Avery et al., 2014) and autism (Barrett and Simmons, 2015; Quattrocki and Friston, 2014), and in patients with substance use disorders (Sönmez et al., 2017).

Materials and methods

Task design

Task implementations

View detailed protocol

A respiratory transducer (TSD201, MRI compatible, BIOPAC Systems Inc), which was fastened around the participants’ upper chest, was utilized to record breathing effort by measuring thoracic changes in circumference during respiration. The signal for the change in circumference was sampled at 1000 Hz using the BIOPAC MP150/RSP100C system, passing through a DC amplifier with low-pass filtering at 1 Hz and high-pass filtering at 0.05 Hz, and gain set to 10 V. Analog signal was then digitized by an A/D converter (USB-1208HS-4AO, Measurement Computing, Inc) and sent to a USB port of the test computer (Figure 1a). The task program in E-Prime (Psychology Software Tools, Pittsburgh, PA, USA) served as an interface through which the digitized signal from the USB port was received and presented to the participants on the computer screen as a continuous blue breath curve extending from left to right as time elapsed (Figure 1b), which was representative of their breathing effort. The breath curve was presented either with or without a delay (Figure 1c).

For the engagement of interoceptive attention during BDT, the participants were required to judge whether the presented breath curve was delayed relative to the breath rhythm they perceived from their body. In half of the trials, the displayed breath curve was synchronized with the participant’s own respiration. In the other half, the displayed breath curve was delivered after a 400 ms delay period relative to the participant’s own respiration (i.e., the plotting of the point on the extending curve was actually the point saved 400 ms before the current time point). Note that the parameter of 400 ms delay was determined based on a proportion (~1/10) of an average respiratory cycle of normal healthy people (i.e., 3–4 s/cycle). For the engagement of exteroceptive attention, the DDT was performed. The participants were instructed to detect whether a red dot flashed on the respiratory curve at any time when the breath curve was displayed. In half of the trials, a red dot flashed (30 ms for the fMRI experiment, and 50 ms for the lesion study) at a randomized time point on the breath curve. Figure 1b illustrates the two tasks. The stimuli for these two tasks were the same, consisting of four trial types reflecting the combination of presence or absence of a delay and presence or absence of a dot (Figure 1c). The factor of attentional deployment involved directing attention interoceptively to ‘respiration’ or exteroceptively to ‘dot’ in the BDT and DDT, respectively. During the two tasks, the participants were instructed to breathe as usual without holding or forcing their breath.

The participants were asked to perform each task in a blocked fashion in the interoceptive and exteroceptive runs. The fMRI experiment consisted of two runs, with one run for the BDT and the other run for the DDT. Each run, which included 60 trials, began and ended with a 30 s blank display, and each trial lasted 18 s, with an average inter-trial interval of 2 s, for a total of 21 min per run. Each trial began with a 3 s "Relax" display, followed by a 12 s respiratory curve presented with or without a 400 ms delay, and ended with a 3 s response window during which the participants made a forced-choice button-press response, prompted by the presentation of two alternative choices to indicate their response (Figure 1b). After the fMRI scan, the participants were asked to indicate the subjective difficulty they felt for each task on a 1–10 scale, with higher value indicating higher difficulty. For the lesion study, the same tasks were employed, one run for each task, with 40 trials in each run.

Behavioral data analysis

Request a detailed protocol

Interoceptive attention is associated with the objective accuracy in detecting bodily signals, the subjective belief in one’s ability to detect bodily signals in general (i.e., sensibility), and the correspondence between objective accuracy and subject report (i.e., metacognitive awareness about one’s performance when detecting bodily signals) (Garfinkel and Critchley, 2013; Garfinkel et al., 2015). Objective interoceptive/exteroceptive accuracy was calculated as the overall correct response rate during the BDT/DDT. In addition, we used signal detection theory to index detection sensitivity and response bias. Signal detection theory characterizes how perceivers separate signal from noise, assuming that the perceiver has a distribution of internal responses for both signal and noise (Snodgrass and Corwin, 1988; Stanislaw and Todorov, 1999). A fundamental advantage of signal detection theory is the distinction between sensitivity (ability to discriminate alternatives) and bias (propensity to categorize ‘signal’ or ‘noise’). For the BDT, the sensitivity index (dʹ) was calculated as dʹ=Zhit rate – Zfalse alarm rate, where the hit rate is the proportion of trials with delayed breath curve and responded as ‘yes’, and the false alarm rate is the proportion of trials with non-delayed breath curve and responded as ‘yes’. A higher value of dʹ indicates a better interoceptive accuracy, whereas a value of 0 represents that the performance is at the chance level. The response bias index β, which represents the position of the subjective decision criterion, was defined as β = exp (dʹ×C), where C = -(Zhit rate +Zfalse alarm rate)/2. Index β corresponds to the distance of participants’ estimated criterion to ideal observer criterion, and a value of 1 indicates no bias. For the DDT, indices of dʹ and β were calculated using the same formula, with the dot present as ‘signal’ and dot absent as ‘noise’. Relative interoceptive accuracy was defined as the difference in performance accuracy between the BDT and DDT to control for non-specific performance effects (Critchley et al., 2004).

An individual’s subjective account of how they experience internal sensation and perception represents an alternative aspect of interoceptive processing, namely sensibility (Garfinkel et al., 2015). In the fMRI experiment, the subjective sensibility of interoceptive processing was measured using the self-report questionnaire of Body Perception Questionnaire (BPQ) (Porges, 1993). The subjective perception of one’s performance during the BDT represents the awareness aspect of interoceptive attention (Garfinkel et al., 2015), which was measured via the subjectively scored difficulty of the BDT relative to the DDT. Note that the confidence ratings of the performance would be more directly related to awareness of interoception, but the measures of subjective difficulty and confidence ratings should be closely related. The correlation between the relative interoceptive accuracy and these indices of subjective sensibility and awareness was calculated to examine the relationship between the perceived (subjective) and actual measured (objective) performance of interoceptive attention. In addition, we conducted BFs of these correlation coefficients using JASP (The JASP Team, 2018). A BF larger than three suggests a significant correlation, whereas a BF smaller than 1/3 indicates a null correlation.

fMRI experiments

Participants

The fMRI experiments included two samples of participants: the first sample included 44 adults (23 females and 21 males, mean age ± standard deviation: 21.43 ± 2.51 years, age range: 19–29 years), and the second sample included additional 28 adults (14 females and 14 males, mean age ± standard deviation: 21.93 ± 2.11 years, age range: 18–26 years). All participants underwent the same experimental procedures, except that pulse and respiratory signals were recorded for the second sample using the pulse sensor (Siemens Peripheral Pulse Unit, PPU_098) of the scanner and BIOPAC, respectively. All participants were right-handed (except for one participant), reported normal or corrected-to-normal vision, and had no known neurological or visual disorders. All participants completed questionnaires indexing subjective interoceptive sensibility (BPQ), symptoms of anxiety (Hamilton anxiety scale, HAMA; Hamilton et al., 1976), depression (Beck Depression Inventory, BDI; Knight, 1984), and positive and negative affective experience (PANAS) (Watson, 1988). They provided written informed consent in accordance with the procedures and protocols approved by The Human Subjects Review Committee of Peking University.

fMRI data acquisition and preprocessing

Request a detailed protocol

During functional scanning, the participants performed the BDT and DDT in separate runs that required them to attend to either their respiration or a visual flash dot, respectively. All neuroimaging data were acquired on a MAGNETOM Prisma 3T MR scanner (Siemens, Erlangen, Germany) with a 64-channel phase-array head-neck coil. During the tasks, blood oxygen level-dependent (BOLD) signals were acquired with a prototype simultaneous multi-slices echo-planar imaging (EPI) sequence (echo time, 30 ms; repetition time, 2000 ms; field of view, 224 mm ×224 mm; matrix, 112 × 112; in-plane resolution, 2 mm ×2 mm; flip angle, 90 degree; slice thickness, 2.1 mm; gap, 10%; number of slices, 64; slice orientation, transversal; bandwidth, 2126 Hz/Pixel; slice acceleration factor, 2). For the second cohort, the thickness was changed to 2 mm with a gap of 15%, and the number of slices was changed to 62. Field map images were acquired using a vendor-provided Siemens gradient echo sequence (gre field mapping: echo time 1, 4.92 ms; echo time 2, 7.38 ms; repetition time, 635 ms; flip angle, 60 degree; bandwidth, 565 Hz/Pixel) with the same geometry and orientation as the EPI image. A high-resolution 3D T1 structural image (3D magnetization-prepared rapid acquisition gradient echo; 0.5 mm ×0.5 mm × 1 mm resolution) was also acquired. Image preprocessing was performed using Statistical Parametric Mapping package (SPM12, RRID: SCR_007037; Welcome Department of Imaging Neuroscience, London, United Kingdom). EPI volumes were realigned to the first volume, corrected for geometric distortions using the field map, coregistered to the T1 image, normalized to a standard template (Montreal Neurological Institute, MNI), resampled to 2 × 2 × 2 mm3 voxel size, and spatially smoothed with an isotropic 8 mm full-width at half-maximum Gaussian kernel.

fMRI: analysis of the first sample

Image statistical parametric mapping

Request a detailed protocol

Imaging data from the two samples were analyzed separately and independently, with the exploratory whole brain analysis conducted with the first sample and the confirmatory ROI analysis conducted with the second sample. For the whole brain analysis of the first sample, statistical inference was based on a random-effects approach (Penny and Holmes, 2007), which comprised two steps: first-level analyses estimating contrasts of interest for each subject followed by second-level analyses for statistical inference at the group level. For each participant, first-level statistical parametric maps of BOLD signals were modeled using general linear modeling (GLM) with regressors defined for each run with the four trial types: 2 breath curve delay (non-delayed, delayed)×2 dot present (no dot, dot). Each trial was modeled as an epoch-related function by specifying an onset time and a duration of 12 s. The corresponding four regressors were generated by convolving the onset of each trial with the standard canonical hemodynamic response functions (HRF) with a duration of 12 s, that is, by convolving the trial block with HRF, equivalent to a box-car function. Six parameters generated during motion correction were entered as covariates of no interest. The time series for each voxel were high-pass filtered (1/128 Hz cutoff) to remove low-frequency noise and signal drift.

Contrast maps for interoceptive vs. exteroceptive attention (BDT – DDT), the presence of breath curve delay (delayed – non-delayed), and the interaction between them ([delayed – non-delayed] BDT – [delayed – non-delayed] DDT) for each participant were entered into a second-level group analysis conducted with a random-effects model that accounts for inter-subject variability and permits population-based inferences. The statistical maps were corrected for multiple comparisons using Gaussian random field (GRF) theory (T > 3.29, cluster-wise p<0.05, GRF corrected) with a minimum cluster size of 420 resampled voxels. Note that changes in neural activity revealed by the main effect of interoceptive vs. exteroceptive attention (the contrast of BDT vs. DDT) could also reflect task-specific effects, such as differences in task difficulty or respiratory characteristics (i.e., amplitude and frequency) between the two tasks, in addition to effects of change in attentional focus. Although the main effect of interoceptive vs. exteroceptive attention (the contrast of BDT vs. DDT) is subject to confounding by the task-specific effects, the interaction effect can disentangle those effects (i.e., cancel out the breathing effort difference between the two tasks). This interaction reflected the brain response when directing attention to the feedback mismatch during interoceptive processing while controlling for the non-specific effect (i.e., the physical difference in feedback stimulus between delayed and non-delayed curves during exteroceptive processing). Therefore, a positive interaction effect represents brain response to the interoceptive processing above and beyond the physical feedback difference.

Correlation between interoceptive accuracy and the interaction effect of the AIC

Request a detailed protocol

To test for a linear correlation between AIC activation and behavioral performance on the BDT, we entered each participant’s interaction contrast maps into the second-level random-effects group regression analysis, together with their individual accuracy in the BDT as the variable of interest and accuracy in the DDT as the covariate. Threshold of significance was GRF-corrected at p<0.05 (T > 2.42) with a cluster extent of 106 contiguous voxels (resampled), corrected using small-volume ROI correction. The mask image was generated from an anatomical template of the bilateral insular cortex based on the Automated Anatomical Labeling template (Tzourio-Mazoyer et al., 2002).

PPI analysis

Request a detailed protocol

PPI analysis provides a measure of change in functional connectivity between different brain regions under a specific psychological context (Friston et al., 1997). We conducted PPI analyses using a moderator derived from the product of the activity of a seed region (i.e., the AIC) and the psychological context (i.e., interoceptive in contrast to exteroceptive attention, BDT vs. DDT). The ROI selection was independent of the interoceptive attention process that was used as the psychological context: The left and right AICs were first identified from the main effect of the breath curve delay (the contrast of delayed versus non-delayed) in the GLM. We then conducted two whole-brain PPI tests for the right and left AIC, reflecting changes in functional connectivity between the seed region time series (physiological regressor) and other brain regions as a function of interoceptive relative to exteroceptive attention (BDT vs. DDT, psychological regressor). The AIC time series of each participant were extracted from a 6 mm-radius sphere centered at the peak of the AIC (right AIC: x = 30, y = 26, z = −4; left AIC: x = −30, y = 24, z = −4). The PPI term was calculated as an element-by-element product of the deconvolved physiological regressor and psychological regressor, which was then reconvolved with the canonical HRF. The generated PPI model included the PPI term, the physiological regressor, the psychological regressor, and nuisance regressors of six motion parameters. The threshold of significance for the second-level group data analysis of the images from the PPI regressor was determined the same as in the GLM. Regions identified as significant clusters have two possible interpretations: (1) the connectivity between the AIC and these regions was altered by the psychological context, or (2) the response of these regions to the psychological context was modulated by AIC activity. To simplify the explanation, we used the first interpretation throughout this article.

DCM analysis

Request a detailed protocol

DCM (Friston et al., 2003) is used to disambiguate different potential network structures by inferring hidden neuronal states from measurements of brain activity. DCM distinguishes between endogenous coupling and context-specific coupling, which could account for the effects of experimentally controlled network perturbations. Considering the inherent limited causal interpretability of the PPI analysis for the direction of interaction, we only conducted DCM to explain the potential mechanisms of interplay between AIC and other brain areas involved in interoceptive attention. The ROI of the right AIC in the DCM was the same as that in the PPI analysis. The other regions included in the DCM were selected based on significant positive and negative PPI results and with the coordinates of the ROIs identified by the group level T-contrast of all conditions versus baseline. Data from one participants were excluded from the DCM analysis because activity in one of the ROIs could not be identified.

A three-area DCM was specified for all participants with bidirectional endogenous connection between the right AIC and the other two ROIs, and with the main effect of ‘all stimuli’ as the driving input entering the other two ROIs. Five base models were generated by specifying possible modulations of interoceptive and exteroceptive attention (BDT and DDT, respectively) on the four endogenous connections between ROIs. These base models were then systematically elaborated to produce 52 variant models, which included all possible combinations of the modulation of interoceptive and exteroceptive attention (BDT and DDT, respectively) on endogenous connections between the right AIC and the two other ROIs.

Model comparison was implemented using random-effects BMS in DCM12 to determine the most likely model of the 52 models given the observed data from all participants (Stephan et al., 2009). The RFX analysis computes exceedance and posterior probabilities at the group level, and the exceedance probability of a given model denotes the probability that this model is more likely than all other models considered (Stephan et al., 2009). To summarize the strength of effective connectivity and its modulation quantitatively, we used random-effects BMA to obtain average connectivity estimates (weighted by their posterior model) across all models and all participants (Penny et al., 2010). We conducted one-sample t tests on the subject-specific BMA parameter estimates to assess their consistency across subjects with Bonferroni correction for multiple comparisons.

fMRI: ROI analyses of the second sample

Request a detailed protocol

Whereas whole brain analyses of the first sample aimed at identifying brain areas involved in interoceptive processing, ROI analyses of the second sample aimed to confirm that the effects found from the first sample were not confounded with the effects induced by other physiological signals. Change in BOLD signals can be due to direct neural activity (induced by experimental manipulation) or an indirect effect (such as vascular response, which would be considered as a confounding effect). For example, the cerebral vascular response is sensitive to the circulation of carbon dioxide (CO2) and oxygen (O2), and causes a change in global cerebral blood flow (CBF) and global BOLD signal. It is evident both in human and animals that the global CBF and global BOLD responses influence local stimulus-induced hemodynamic response to neural activation (Cohen et al., 2002; Friston et al., 1990; Ramsay et al., 1993). In general, a larger local stimulus-induced BOLD response occurs when global BOLD is lowered, whereas a smaller local stimulus-induced BOLD response occurs when global BOLD is elevated. In our study, the experimental manipulation of interoception was likely to cause a change in respiratory characteristics (i.e., circulation of CO2 and O2). The difference in physiologic states between the BDT and the DDT might cause a change in global BOLD signals. Thus, the effect resulting from local interoception-related BOLD responses would be confounded by the global hemodynamic influence.

To partial out the potential confounding, we processed physiological data, including cardiac pulsation and respiratory volume collected in the second sample, by using the PhLEM toolbox (http://sites.google.com/site/phlemtoolbox/). Physiological noise correction consisted of (1) regressing out time-locked cardiac and respiratory effects, and their interaction effect using a modification of the conventional RETROICOR approach (Brooks et al., 2008; Glover et al., 2000), and (2) regressing out low-frequency respiratory and heart rate effects using the RVHRCOR approach (Verstynen and Deshpande, 2011). In RETROICOR, a cardiac phase calculated from a pulse oximeter was assigned to each acquired image in a time series (Hu et al., 1995), and a respiratory phase was assigned to a corresponding image using the histogram equalized transfer function that considers both the respiratory timing and depth of breathing (Glover et al., 2000). The conventional RETROICOR approach (Glover et al., 2000) defines low-order Fourier terms (i.e., sine and cosine values of the principal frequency and the 2nd harmonic) to model the independent effects of the cardiac and respiratory fluctuation, which is considered insufficient to remove variations caused by physiological artifacts (Harley and Bielajew, 1992; Tijssen et al., 2014). Therefore, we used additional terms of higher-order Fourier expansions (i.e., to the 5th harmonics) in RETROICOR, and formed multiplicative sine/cosine terms that consider the interaction between cardiac and respiratory effects. In specific, the interaction terms were calculated by Sin(φc±φr) and Cos(φc±φr), where φc,r is the cardiac or respiratory phase, consisting of a mixture of third-order cardiac and second-order respiratory harmonics. In RVHRCOR, two nuisance regressors were generated by convolving respiratory variations (RVs) and heart rate (HR) with ‘respiration response function (RRF)’ and the ‘cardiac response function (CRF)’ respectively. In specific, RV was computed as the root-mean-square amplitude of the respiration waveform across a 6 s sliding window, and HR was computed as the inverse of the average beat-to-beat interval in a 6 s sliding window (Chang et al., 2009). Therefore, the physiological correction contained a total of 46 regressors, of which 20 were from independent time-locked cardiac and respiratory effects, 24 were from interaction terms, and two were from low-frequency RV and HR effects. Statistical parametric maps were generated using the same GLM as in the whole brain analyses, with motion parameters and these physiological regressors entered as covariates of no interest.

To avoid double dipping, we defined the ROIs based on the first sample. In specific, the ROIs (i.e., left and right AICs) were the clusters of the second-level group analysis results of the interaction effect ([delayed – non-delayed] BDT – [delayed – non-delayed] DDT). Parameter estimates were extracted from each ROI in the second sample under the four experimental conditions of each participant, and then entered into a two-way repeated-measures ANOVA. We also examined correlations between the interaction effect of each ROI and behavior measures (i.e., relative interoceptive accuracy) across participants.

To further examine the degree to which physiological correction impacted the whole brain activation, we conducted whole brain paired t-tests by comparing the contrast maps without and with physiological correction at an extremely permissive threshold (voxelwise p<0.05 uncorrected).

Lesion study

Brain lesion patient and control groups

Request a detailed protocol

Six male patients (33–53 years old, mean 42.17 ± SD 7.31 years) with focal unilateral insular cortex lesions participated in the lesion study (see Table 9 for patient characteristics). Two patients had a right-side lesion, and four patients had a left-side lesion. In addition, six patients with focal lesions in regions other than the insular cortex (i.e., temporal pole, n = 3, lateral frontal cortex, n = 2, and superior temporal gyrus, n = 1) were recruited as BDCs, and 12 neurologically intact participants were recruited as NCs. All lesions were resulted from the surgical removal of low-grade gliomas. All patients were recruited from the Patient’s Registry of Tiantan Hospital, Beijing, China. NC participants were recruited in the local community. All NC participants were right-handed, had normal color vision, and reported no previous or current neurological or psychiatric disorders. BDC patients matched with patients with insular cortex lesions in chronicity (t(10) = −0.36, p=0.38, BF = 0.48), and neither group significantly differed from the NC group in age (AIC vs. NC: t(9) = −1.01, p=0.18, BF = 0.61; BDC vs. NC: t(10) = −1.80, p=0.06, BF = 1.25) nor education (AIC vs. NC: t(10) = 0.77, p=0.25, BF = 0.52; BDC vs. NC: t(6) = −1.04, p=0.18, BF = 0.72). All six insular lesion patients were considered cognitively intact, as determined by Mini-Mental State Examination (MMSE), a measurement of cognitive impairment (Folstein et al., 1975), and the raw scores of MMSE were not significantly different from either BDCs (t(10) = −1.30, p=0.13, BF = 0.78) or NCs (t(10) = −1.76, p=0.06, BF = 1.17). Compared with the NCs, the patients with insular lesions did not show significant alteration in baseline mood indexed by the BDI score, compared to NCs (t(6) = 1.70, p=0.06, BF = 1.84) or BDCs (t(9) = 0.65, p=0.28, BF = 0.53). Demographic information of the groups can be found in Table 9. By chance, all the patients with AIC lesions were male. All participants were informed of the study requirements and provided written consent prior to participation. The patient study was approved by the Institutional Review Board of the Beijing Tiantan Hospital, Capital Medical University.

Table 9
Demographic characteristics of the participants in lesion experiment.
https://doi.org/10.7554/eLife.42265.007
Lesion lateralityLesion size (ml)Chronicity (months)Age (years)GenderEducation (years)MMSEBDI
IC1Right3.73839M15284
IC2Right5.5633M16281
IC3Left11.2938M12264
IC4Left9.01253M12268
IC5Left16.0651M16291
IC6Left9.23740M16260
BDC3 Left/3 Right18 ± 1421 ± 1639 ± 73F/3M12 ± 328 ± 12 ± 2
NCN/AN/AN/A46 ± 78F/4M14 ± 228 ± 11 ± 1

  1. IC, insular cortex; BDC, brain damage control; NC, normal control; MMSE, mini-mental state examination; BDI, Beck depression inventory.

Lesion reconstruction

Request a detailed protocol

Two neurosurgeons, blinded to the experimental design and behavioral results, identified and mapped the lesions of each patient onto a template derived from a digital MRI volume of a normal control (ch2bet.nii) embedded in the MRIcro program (RRID: SCR_008264; http://www.cabiatl.com/mricro/mricro/index.html). In each case, lesions evident on MRI were transcribed onto corresponding sections of the template to create a volume of interest image. This volume of interest image was then used to measure the location (in MNI coordinates) and volume (in mL) of individual lesions and to create within-group overlaps of lesions using the MRIcro program.

Behavioral data analysis of the lesion study

Request a detailed protocol

We used non-parametric analysis (Feys, 2016) to test the two-way interaction between group (AIC, BDC, and NC) and task (BDT and DDT) using R (R Development Core Team, 2013) because the small sample data sets did not meet the assumption of parametric tests. In specific, we used the npIntFactRep function (from the npIntFactRep package) that yielded an aligned rank test for interaction in the two-way mixed design with the group (AIC, BDC, and NC) as the between-subject factor and with the task (BDT and DDT) as the within-subject factor. If the interaction was significant, we used the non-parametric bootstrapping method to test the simple between group effects for each task separately. The bootstrapping procedure was conducted with 10,000 iterations (Hasson et al., 2003; Mooney and Duval, 1993). If the probability of obtaining the observed t-value was less than 5% (one-tailed), we considered the difference between the two groups to be significant. We used one-tailed tests because we hypothesized that lesions of a specific brain region (i.e., the AIC) would induce deficits in behavioral response. In addition, we calculated BFs with Cauchy prior to determine the relative strength of evidence for the null and alternative hypotheses (Dienes, 2014; Dienes and Mclatchie, 2018). The value of BF means that the data are BF times more likely under the alternative than under the null hypothesis. The standard value for assessing substantial evidence for the null is BF <1/3 and for the theory against null is BF >3, whereas values between 1/3 and 3 are counted as data insensitivity. The BFs were calculated using JASP (The JASP Team, 2018).

Data availability

Source data have been deposited in Dyrad (doi:10.5061/dryad.5sj852c), including behavioral data, fMRI data, and lesion patient data.

The following data sets were generated
    1. Wang X
    2. Wu Q
    3. Egan L
    4. Gu X
    5. Liu P
    6. Gu H
    7. Yang Y
    8. Luo J
    9. Wu Y
    10. Gao z
    11. Fan J
    (2018) Dryad Digital Repository
    Data from: Anterior insular cortex plays a critical role in interoceptive attention.
    https://doi.org/10.5061/dryad.5sj852c

References

    1. Damasio AR
    2. Tranel D
    3. Damasio HC
    (1991)
    Behavior: theory and preliminary testing
    Frontal Lobe Function and Dysfunction 217:.
    1. Damasio AR
    (1996) The somatic marker hypothesis and the possible functions of the prefrontal cortex
    Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences 351:1413–1420.
    https://doi.org/10.1098/rstb.1996.0125
  32. Book
    1. Damasio AR
    (2003)
    Looking for Spinoza: Joy, Sorrow and the Feeling Brain
    Houghton Mifflin Harcourt.
    1. Folstein MF
    2. Folstein SE
    3. McHugh PR
    (1975)
    "Mini-mental state". A practical method for grading the cognitive state of patients for the clinician
    Journal of Psychiatric Research 12:189–198.
    1. Friston K
    2. Kiebel S
    (2009) Predictive coding under the free-energy principle
    Philosophical Transactions of the Royal Society B: Biological Sciences 364:1211–1221.
    https://doi.org/10.1098/rstb.2008.0300
    1. Hamilton MC
    2. Schutte NS
    3. Malouff JM
    (1976)
    Sourcebook of Adult Assessment: Applied Clinical Psychology
    154–157, Hamilton anxiety scale (HAMA), Sourcebook of Adult Assessment: Applied Clinical Psychology, Springer.
  63. Book
    1. James W
    (1890)
    The Principles of Psychology
    New York: Holt and company.
  74. Book
    1. Mooney CZ
    2. Duval RD
    (1993)
    Bootstrapping: A Nonparametric Approach to Statistical Inference
    Newbury Park: Sage Publications.
    1. Naidich TP
    2. Kang E
    3. Fatterpekar GM
    4. Delman BN
    5. Gultekin SH
    6. Wolfe D
    7. Ortiz O
    8. Yousry I
    9. Weismann M
    10. Yousry TA
    (2004)
    The insula: anatomic study and MR imaging display at 1.5 T
    AJNR. American Journal of Neuroradiology 25:222–232.
  78. Book
    1. Penny W
    2. Holmes A
    (2007) Random effects analysis
    In: Friston K, Ashburner J, Kiebel S, Nichols T, Penny W, editors. Statistical Parametric Mapping: The Analysis of Functional Brain Images. Academic Press. pp. 156–165.
    https://doi.org/10.1016/b978-012372560-8/50012-7
    1. Porges S
    (1993)
    Body perception questionnaire
    Laboratory of Developmental Assessment, University of Maryland.
  83. Software
    1. R Development Core Team
    (2013) R: A Language and Environment for Statistical Computing
    R Foundation for Statistical Computing, Vienna, Austria.
    https://www.r-project.org/
    1. The JASP Team
    (2018)
    JASP
    JASP, 0.9.
  107. Book
    1. Tranel D
    2. Damasio H
    (1991)
    Somatic Markers and the Guidance of Behaviour: Theory and Preliminary Testing
    New York: Oxford University Press.

Decision letter

  1. Klaas Enno Stephan
    Reviewing Editor; University of Zurich and ETH Zurich, Switzerland
  2. Michael J Frank
    Senior Editor; Brown University, United States
  3. Klaas Enno Stephan
    Reviewer; University of Zurich and ETH Zurich, Switzerland
  4. Olivia K Faull
    Reviewer; University of Oxford, United Kingdom
  5. Micah Allen
    Reviewer; UCL, United Kingdom

In the interests of transparency, eLife includes the editorial decision letter and accompanying author responses. A lightly edited version of the letter sent to the authors after peer review is shown, indicating the most substantive concerns; minor comments are not usually included.

[Editors’ note: a previous version of this study was rejected after peer review, but the authors submitted for reconsideration. The first decision letter after peer review is shown below.]

Thank you for submitting your work entitled "Anterior insular cortex plays a critical role in interoceptive attention" for consideration by eLife. Your article has been reviewed by three peer reviewers, including Klaas Enno Stephan as the Reviewing Editor and Reviewer #1, and the evaluation has been overseen by a Senior Editor. The following individuals involved in review of your submission have agreed to reveal their identity: Olivia K Faull (Reviewer #2); Micah Allen (Reviewer #3).

Our decision has been reached after consultation between the reviewers. Based on these discussions and the individual reviews below, we regret to inform you that your work will not be considered further for publication in eLife.

All three reviewers found the paper of interest and regarded the novel paradigm as a promising basis for developing experimental probes of respiratory interoception. However, all reviewers also expressed significant methodological concerns, particularly with regard to the absence of physiological noise correction, that prohibit an interpretation of the current findings. Since a re-analysis of the data from scratch was recommended unanimously, the decision was to reject the paper at this stage; however, we would encourage you to submit the paper again de novo, provided that the revised analyses produce results that can be considered a major advance for the field. Below, a synthesis of the three reviews is provided that is meant to help you address the methodological issues and redo the analyses. This is deliberately kept brief, with most minor comments by the reviewers excluded, to help focus on the most essential revisions.

Critical issue:

1) The fMRI data analysis did not include any correction for physiological noise in the fMRI data. Due to the breathing-related task design, it is imperative that such a correction is performed; in its absence, the current results cannot be interpreted. This is because when directing conscious attention towards breathing, it is very likely that breathing patterns will be adjusted (even when not intended). Therefore, without correction, there is no way to disentangle the effect of a change in breathing from a change in interoceptive processing of respiration-related signals. Furthermore, changes in breathing can confound fMRI measurements, both due to direct effects on BOLD signals and due to indirect effects (e.g., B0 fluctuations). For these reasons, a RETROICOR (or similar) correction would be essential for the first-level analyses. This RETROICOR correction should not only include respiratory, but also cardiac signals, given the modulation of heartrate by breathing, the representation of cardiac activity in the insula, and the presence of large blood vessels (e.g., middle cerebral artery) near the insula that can produce vascular artefacts. Additionally, we would suggest calculation and inclusion of key respiratory parameters (e.g., rate and depth during both task conditions, measures of end-tidal CO2 if available) as regressors of no interest in the second-level analysis, to separate physiological and interoceptive aspects of changes in breathing.

Major issues:

2) Although not quite as critical as the lack of physiological noise correction, all reviewers also noted issues with regard PPI analysis. Generally, it would be helpful to clarify what type of PPI you are using. For example, the classical PPI that tries to capture interaction effects (Friston et al., 1997); or a PPI term within an extended statistical model that tests for context-dependent coupling over and beyond any other experimental effects. If you are going for the former, the seed region would be identified by one of the main effects (please note that your F-contrast is not, as stated in the paper, statistically independent from the t-contrast of the main effect of interoceptive vs. exteroceptive attention), and the PPI term would correspond to the interaction between the other main effect and the timeseries (see Friston et al., 1997 for details). If you have in mind the latter, it would be good to ensure that the PPI model contains all experimental effects (e.g., see https://fsl.fmrib.ox.ac.uk/fsl/fslwiki/PPIFAQ).

3) The Introduction and Discussion do not represent the literature on interoception well and should be revised carefully. This includes incorrect statements (such as the assertion that breathing "is the sole 'perceptible' internal bodily signal"), incorrect citation of literature (e.g., papers cited by Craig, Critchley, etc. are not about interoceptive "attention" but about conscious awareness of/sensitivity to interoceptive signals), and lack of references to key components of the literature (e.g., theoretical papers on different components of interoception and experimental papers on insula lesions).

4) The conceptual interpretation of the experimental paradigm as an "interoceptive attention" task should be revisited. While the task relies on shifting attention (between intero-and exteroceptive domains), it primarily serves to provide a measure of intero- and exteroceptive accuracy or sensitivity (as also reflected by your analysis in terms of d'), and appears to be more adequately described in these terms.

5) The proposed experimental paradigm is innovative and has much potential for future studies of respiratory interoception. However, there are some potential problems that may need consideration. First, the control condition appears to require a different cognitive process than the condition of interest. The latter requires a temporally extended matching process; the former requires a detection process that terminates once a dot has appeared. Second, the delay was a set interval of 400 ms, rather than a proportion of the individual's respiratory cycle. This may partially determine task difficulty and performance across individuals. Third, given that the task is novel, it would be important to see more details of task performance, e.g. plots of individual accuracy rates, analysis of reaction times and signal-detection theoretic considerations (i.e., where they more biased for either interoceptive or exteroceptive conditions?). The task seems very easy compared to standard heartbeat detection tasks: were there ceiling effects (i.e., did any participants have 100% accuracy)? Finally, the task does not represent a pure probe of interoception as respiratory processes can also be tracked using exteroceptive and proprioceptive information. It thus seems likely that participants relied on a mix of interoceptive, exteroceptive, and proprioceptive information for performing the task. These issues do not invalidate the task, but they deserve a critical and frank discussion so that the reader is aware of the limitations of the paradigm.

6) There are some issues with the statistical analysis and reporting. Exact p-values, test statistics, and standardized effect sizes should be reported for all analyses. Numerous tests are reported as one-sided; this needs to be justified or replaced by two-sided tests. Non-significant results should not be presented as evidence for the absence of a difference (e.g., in the lesion analysis); this corresponds to accepting the null hypothesis and should be replaced by a corresponding Bayesian test.

[Editors’ note: what now follows is the decision letter after the authors submitted for further consideration.]

Many thanks for submitting a revised version of your manuscript to eLife. It has now been seen by two of the three previous reviewers. Please excuse that this consultation process took longer than usual.

We were impressed by the effort you invested in acquiring an additional dataset with concomitant measures of cardiac and respiratory activity. However, we continue to think that the statistical analysis needs to account for task-induced variations in breathing which can profoundly impact on BOLD measurements. We did read the paper (Miller and Chapman) that you attached for justification of omitting respiratory measures from the statistical model but must confess that we did not find it very insightful in relation to the current problem; in particular, equating the current issue with "Lord's paradox" (which is a rather specific case) seems misleading.

The problem in your analysis is a very generic one: including or excluding a confound regressor that is correlated to a regressor of interest in a GLM amounts to an active decision how shared variance is interpreted – or, put differently, whether one wishes to maximise sensitivity or specificity of the analysis. We think that for a study that reports the effect of a cognitive intervention for the first time, specificity is more important: the reader would like to be assured that activations attributed to the cognitive intervention are not merely driven by physiological effects. We agree that the interaction effect should be protected against task-induced breathing changes. The main effect of task, however, is not; and it is arguably of greater importance for the message of the paper.

For these reasons, we are not convinced it is a good idea to pool the two groups and report analyses without including regressors that represent physiological (respiratory) noise. We also thought that the RETROICOR analysis presented in the response letter (2nd order respiratory regressors only and no cardiac-respiratory interactions) is unusually lenient.

In our view, these problems are too substantial to proceed with in-depth peer-review. If you would like eLife to continue considering the paper, we would recommend that the paper (i) reports analyses from both samples separately, (ii) discusses the potential problems of interpretation in the first sample, and (iii) includes a rigorous RETROICOR correction of breathing effects for the second sample. You could boost the statistical sensitivity of the second analysis by using the FWE-corrected activations from the first analysis in order to specify a mask for reducing the search volume for FWE correction in the second analysis. In this way, you would use the higher statistical sensitivity of the first analysis in order to identify regions where the cognitive process of interest may take place and then test in the second sample, with due consideration of potentially confounding effects, whether this can be corroborated.

We are very sorry that we cannot be more positive at this stage and understand that this must be disappointing for you, given the substantial effort you have invested in the revision of this paper. We do hope, however, that the recommendation above is helpful.

[Editors’ note: what now follows is the decision letter after the authors submitted for further consideration.]

Thank you for sending your article entitled "Anterior insular cortex plays a critical role in interoceptive attention" for peer review at eLife. Your article has been evaluated by three peer reviewers, one of whom is a member of our Board of Reviewing Editors, and the evaluation has been overseen by Michael Frank as the Senior Editor.

Major points:

1) The wording "… checked that the AIC ROI results were not dependent on the (independent) ROI selection.…" is confusing. Presumably you wanted to say something like "… checked how much the AIC ROI results were affected by physiological noise correction.…"? In direct relation to this point, it is rather surprising to see such little effects of physiological noise correction on insula activity. Typically, physiological noise regressors (RETROICOR) do explain a substantial amount of BOLD signal in the insula. The particular statistical test you used asks whether specific contrasts are altered by the inclusion vs. exclusion of physiological noise regressors (which is fine) but is not sensitive to the question whether insular activity is affected by physiological noise at all (as implied by your wording in the subsection “ROI analysis results of the fMRI study of the second sample”). As a sanity check, it would be worth performing an additional F-test spanning all RETROICOR regressors. If this test does not show significant insula activation, it would seem wise to double-check the RETROICOR analysis, in order to make sure there are no errors.

2) Materials and methods: "The corresponding four regressors were generated by convolving the onset vectors of each trial type with a standard canonical hemodynamic response function (HRF)". Was each trial modelled as an event or a block? The methods describe each stimulus period lasting 12 seconds, which would appear more akin to a block design for the GLM?

3) The value of the mention of CO2 and O2 in this manuscript is questionable – these effects would need to be accounted for by either measuring them and regressing them out, or using an approximation such as RVT (respiratory volume per unit of time) regressors, which do not appear to be used here. Standard cardiac and respiratory waveforms and harmonics do not account for these effects. This paradigm would likely induce very slight hyperventilation when attention is directed towards monitoring breathing curves, which would result in a decrease in expired CO2 over the 12 second stimulus period (and the resulting washout period), which would induce a global over-estimation of the BOLD activity related to the task. RVT regressors could be included in the RETROICOR to account for this. They actually mention that there is a difference in respiratory volume between tasks in the subsection “ROI analysis results of the fMRI study of the second sample”.

4) The main effect of the task (interoceptive attention vs exteroceptive attention) is very large (Figure 2). However, it should be noted that the participants found the interoceptive task more difficult than the exteroceptive task, and thus these differences in brain activity are very likely associated with task difficulty as well as the direction of attention. This is probably worth mentioning somewhere in the Discussion?

5) Please accept our apologies – we should have noted this earlier – but the analyses presented in the lesion study suffer from a major problem. The authors are conducting non-parametric tests between participants in the interoceptive and exteroceptive condition separately, and then interpreting the presence vs lack of significance as evidence for a specific effect of AIC lesions on interoceptive attention. This, however, is an erroneous conclusion (see also https://www.nature.com/articles/nn.2886), and instead it would be necessary to demonstrate a significant group (AIC vs control) by condition (intero vs extero) interaction. It is important to perform the correct test, especially because these are the most controversial findings of the paper. It is debatable whether previous lesion research has provided any valid evidence for AIC lesions on interoceptive sensitivity and emotional awareness (in a previous review, the authors were asked to consider this work, but this appears to have been ignored).

6) There are also some concerns – and, again our apologies for not having addressed this earlier – regarding the analyses of the correlation between task measures and questionnaires (subsection “Behavioral results of the fMRI studies”). Specifically, the use of multiple one-sided tests seems questionable and would need a strong motivation, and there is a lack of multiple comparison correction. There were also questions about power: if you pool the two samples you may have reasonable power (i.e., for an effect size of|r|=0.3, N=72 would give you 75% power of detecting a significant effect at α=0.05, two-sided) for a single test, but this would diminish when taking into account multiple comparisons (with lower α as a result). It seems fair to ask that the analysis is fully transparent with regard to power, uses two-sided tests and multiple comparison correction, and tones down the interpretation of the results considerably.

7) Following directly from the previous point, one of the reviewers made the following suggestion which we would suggest you consider: "My advice is the following: take all of the correlation analyses and throw them into a giant Bayesian correlation table with appropriate priors on correlation strength, and make these supplementary analyses flagged clearly as exploratory in nature. Flag up the ones that show Bayes factors greater than 3 under a two-sided test, and in particular discuss any who are implicated in both study 1 and study 2. Refocus the paper to emphasize the novelty and importance of understanding respiratory awareness and make it clear that the link to emotion is more speculative – a fascinating area that future large-scale studies can target by using the task presented here."

https://doi.org/10.7554/eLife.42265.037

Author response

[Editors’ note: the author responses to the first round of peer review follow.]

Critical issue:

1) The fMRI data analysis did not include any correction for physiological noise in the fMRI data. Due to the breathing-related task design, it is imperative that such a correction is performed; in its absence, the current results cannot be interpreted. This is because when directing conscious attention towards breathing, it is very likely that breathing patterns will be adjusted (even when not intended). Therefore, without correction, there is no way to disentangle the effect of a change in breathing from a change in interoceptive processing of respiration-related signals. Furthermore, changes in breathing can confound fMRI measurements, both due to direct effects on BOLD signals and due to indirect effects (e.g., B0 fluctuations). For these reasons, a RETROICOR (or similar) correction would be essential for the first-level analyses. This RETROICOR correction should not only include respiratory, but also cardiac signals, given the modulation of heartrate by breathing, the representation of cardiac activity in the insula, and the presence of large blood vessels (e.g., middle cerebral artery) near the insula that can produce vascular artefacts. Additionally, we would suggest calculation and inclusion of key respiratory parameters (e.g., rate and depth during both task conditions, measures of end-tidal CO2 if available) as regressors of no interest in the second-level analysis, to separate physiological and interoceptive aspects of changes in breathing.

We fully agree with the reviewers that the effect of a change in breathing patterns would not be disentangled from the effect of interoceptive processing we are interested in, which might be a confound. In the following, we respond to two aspects of this comment:

First, the effect of interoceptive processing can be isolated by using the interaction effect without correcting for physiological parameters.

We agree that the possible change in breathing patterns (i.e. the respiratory effort difference) might contribute to the difference in terms of activation in the key region of interest, i.e., the AIC. Therefore, analysis to partial out the impact of the effort difference using the physiological signals might be the solution. However, based on Miller, G. A., and Chapman, J. P. (2001). Misunderstanding analysis of covariance. Journal of abnormal psychology, 110(1), 40-48., this ANCOVA method is not an appropriate solution for this scenario. That is, when the covariate and the experimental treatment are related (share variance), the regression adjustment may remove part of the treatment effect or produce a spurious treatment effect. This would be a misuse of ANCOVA, as in the so-called “Lord’s Paradox”. In our studies, the manipulation of interoception inherently involved a change in breathing effort (as evident in the difference of respiratory volume between interoceptive and exteroceptive condition, t(27) = 3.90, p = 0.001, see Author response image 1). Therefore, removing variance associated with breathing effort in BOLD signals would not only remove considerable variance in BOLD signals associated with interoceptive attention, but may also result a spurious treatment effect. A change in breathing pattern should not be viewed as a covariate but rather as a feature inherent in the task, and regressing out the physiological signals would be an inappropriate use of analysis of covariance. Therefore, although we have the physiological signals collected for the new sample, we decided to use an alternative strategy (see below) rather than follow the suggested ANCOVA method for the final report. In the response letter, we report the results after RETROICOR correction as a comparison.

Although the main effect of interoceptive attention (the contrast of interoceptive versus exteroceptive condition) is subject to this confounding and cannot be solved using ANCOVA, the interaction effect ([delay – non-delayed]interoception vs. [delay – non-delayed]exteroception) should not be confounded by the change in breathing. This interaction reflects the brain response to the mismatch (delay versus non-delayed) under the interoceptive condition controlling for the non-specific effect (i.e., the difference in feedback stimulus under the exteroception task condition). Therefore, a positive interaction effect represents brain response to the interoceptive processing above and beyond the physical feedback difference. To further illustrate that the interaction effect of brain response is not subject to breathing effort difference, we showed the pattern of the respiratory volume under different experimental conditions (see Author response image 2). Although there was a change in the respiratory volume between interoceptive and exteroceptive task conditions (F(1,27) = 15.88, p < 0.001), this difference is canceled out by using the interaction effect (F < 1, BF = 0.025). In the original version of this manuscript, we did not explain this logic clearly or emphasize the interaction effect. In this revised manuscript, we have added a detailed description of this logic and used the interaction effect as the index of brain responses (i.e., brain activity, connectivity, and individual difference in terms of the relationship between interaction effect and interoceptive accuracy) to interoceptive processing (subsection “Image preprocessing and statistical parametric mapping”).

Author response image 1
Download asset Open asset
Activation maps with and without including individual heart rate and respiratory volume as covariates in the 2nd GLM.

(a) Main effect of interoceptive attention (interoceptive task vs. exteroceptive task). (b) Interaction between attention type and breath-curve feedback condition ([delayed – non-delayed]interoceptive task – [delayed – non-delayed]exteroceptive task).

Author response image 2
Download asset Open asset
The pattern of respiratory parameters under different task conditions.

Second, the brain activation patterns are almost the same with and without correcting physiological parameters, demonstrated in a new sample of N = 28.

We are thankful to the reviewers for pointing out that change in BOLD signals can both due to direct neural activity (induced by experimental manipulation), and due to indirect effect, such as vascular response (considered to be a confounding effect). Specifically, cerebral vascular response is sensitive to circulation of CO2 and O2, and causes a change in global cerebral blood flow (CBF) and global BOLD signal. It is evident both in human and animals that the global CBF and global BOLD response influence local stimulus-induced hemodynamic response to neural activation (Cohen et al., 2002; Friston et al., 1990; Ramsay et al., 1993; Sicard et al., 2005). Typically, a larger local stimulus-induced BOLD response occurs when global BOLD was lowered, while a smaller local stimulus-induced BOLD response occurs when global BOLD was elevated. In our study, the difference on physiologic states between interoceptive and exteroceptive task conditions might cause a change in global BOLD signals, and thus confounded the effect resulting from altered neural activity (truly experimental effect) with the effect resulting merely from global hemodynamic influence. For example, the increased respiratory depth under interoceptive task condition (even when not intended) might increase cerebral O2 level and lowered CO2 level, which leads to a reduction in global CBF and global BOLD. Therefore, the AIC activation identified from the main effect of interoceptive processing (interoception task vs. exteroception task) might just represent a change in global BOLD response (i.e. lower global BOLD leads to larger local BOLD response), rather than neural activation underlying interoceptive processing. To explore the potential effect resulted from respiratory and cardiac difference, it is worth comparing activation patterns with and without correcting physiological parameters.

The reviewers offered a method to resolve this issue, which is to regress out the physiological noises by including respiratory and cardiac signals. In the original experiment, we did not record those signals. To examine the impact of those potential confounding variables, we have collected a new sample of 28 participants using the same task with both respiratory and cardiac signals recorded. According to the reviewers’ suggestions, we applied the RETROICOR correction to the phasic aspect of cardiac (8 regressors) and respiratory (2 regressors) signals to regress out potential physiological influence, generated nuisance regressors for variations in breathing rate/volume (RV) (1 regressor) and heart rate (HR) (2 regressors) which were convolved with the “respiration response function” (RV+RRF correction) and “cardiac response function” (HR+CRF correction), respectively, in addition to the motion correction (6 regressors), for the first-level GLM. We also followed reviewers’ suggestions to include the parameters of individual heart rate and breathing volume in the second-level analysis. We also conducted the data analysis without RETROICOR correction as a comparison. Our results revealed almost the same activation patterns with and without the correction of physiological signals (see Author response images 1 and 3). Specifically, the anterior insular cortex was involved in the interoceptive processing in terms of the main effect of interoception and the interaction effect.

Author response image 3
Download asset Open asset
Activation maps with and without physiological correction for the 1st level GLM.

(a) Main effect of interoceptive attention (interoceptive task vs. exteroceptive task). (b) Interaction between attention type and breath-curve feedback condition ([delayed – non-delayed]interoceptive task – [delayed – non-delayed]exteroceptive task). (c) The interaction patterns of the left and right anterior insular cortex (AIC) activity without physiological correction. (d) The interaction patterns of the left and right AIC activity with physiological correction.

In addition, we conducted paired t-test on the interoception vs. exteroception and the interaction beta maps obtained without and with physiological correction (see Author response image 4). Indeed, the physiological parameters induced a global increase in BOLD signal, which was evident in a whole brain activation comparing “without physiological correction” with “with physiological correction” beta maps (Author response image 4A). However, the interoceptive processing induced neural activation still hold after physiological correction (see Author response image 1A and Author response image 2A). In contrast, the interaction effect is not subject to the physiological parameters induced global activation, as revealed by a “blank brain” using paired t-test (Author response image 4B).

Author response image 4
Download asset Open asset
Paired t-test on beta maps obtained without and with physiological correction.

(a) using interoceptive vs. exteroceptive contrast maps. (b) using interaction effect beta maps.

The main result of the new sample (n = 28) replicated the result of our original sample, which is that the AIC is involved in the interoceptive processing defined by the interaction effect. We have decided to combine the data from the original and the new samples (total n = 72) to enhance the statistical power. We have conducted the corresponding analyses and updated the Materials and methods and the Results. Most of the key findings from our original experiment still hold.

Major issues:

2) Although not quite as critical as the lack of physiological noise correction, all reviewers also noted issues with regard PPI analysis. Generally, it would be helpful to clarify what type of PPI you are using. For example, the classical PPI that tries to capture interaction effects (Friston et al., 1997); or a PPI term within an extended statistical model that tests for context-dependent coupling over and beyond any other experimental effects. If you are going for the former, the seed region would be identified by one of the main effects (please note that your F-contrast is not, as stated in the paper, statistically independent from the t-contrast of the main effect of interoceptive vs. exteroceptive attention), and the PPI term would correspond to the interaction between the other main effect and the timeseries (see Friston et al., 1997 for details). If you have in mind the latter, it would be good to ensure that the PPI model contains all experimental effects (e.g., see https://fsl.fmrib.ox.ac.uk/fsl/fslwiki/PPIFAQ).

We thank the reviewers for making the suggestion of the correct way to conduct the PPI analysis. Actually, our PPI analyses were intended to follow the classical PPI to capture the interaction effect in terms of a psychological process and brain activity. In this sense of a 2 × 2 factorial design, the physiological signals should be extracted from one main effect while the psychological variable should be the other. However, our original selection of seed region, which was the F-contrast of all condition, was not independent of the psychological context variable, which was the interoceptive versus exteroceptive attention. Therefore, it was incorrect. Following the correction of the reviewers, we have selected the seed region of AIC based on the main effect of feedback delay (the contrast of delayed versus non-delayed condition), and used the other main effect of interoception (the contrast of interoceptive versus exteroceptive attention) as the psychological variable (subsection “Image preprocessing and statistical parametric mapping”). One way to interpret the PPI result is that the connectivity between the AIC and other brain areas is modulated by attention to interoception in contrast to exteroception. We have reconducted PPI analyses with this correct method and updated the Results. The new results showed that there was enhanced connectivity between AIC and posterior central gyrus (as well as some regions of the cognitive control network such as frontal eye field and anterior cingulate cortex) modulated by interoceptive attention, while there was increased connectivity between AIC and visual cortex (i.e., V2/3) modulated by exteroceptive attention. Additionally, we also found that the PPI between AIC and visual cortex was predictive of individual differences in interoceptive accuracy.

3) The Introduction and Discussion do not represent the literature on interoception well and should be revised carefully. This includes incorrect statements (such as the assertion that breathing "is the sole 'perceptible' internal bodily signal"), incorrect citation of literature (e.g., papers cited by Craig, Critchley, etc. are not about interoceptive "attention" but about conscious awareness of/sensitivity to interoceptive signals), and lack of references to key components of the literature (e.g., theoretical papers on different components of interoception and experimental papers on insula lesions).

We have now revised the manuscript to make the terminology accurate. We agree that the “solo perceptible” is over strong, and we revised the statement about the breath as “the most perceptible internal bodily signal” in comparison to the heart beat which is used in almost all of the tasks examining interoception (Introduction, third paragraph). Regarding the terms of conscious awareness of/sensitivity to interoceptive signals and interoceptive attention, we define interoceptive attention as the “process” and conscious awareness as the final “outcome” of the attentional processing. We have made this point more clear by defining interoceptive attention as the attentional mechanisms in interoceptive awareness (Introduction, first paragraph). The citations we have are all for interoceptive awareness. We have also made this logic more clear in the Discussion by changing the word “interoceptive attention” to “interoceptive awareness” when referring to the outcome of the attentional process for a more appropriate use.

Regarding the literature review of insula, we have added some missed references (e.g., He et al., 2009, for the empirical study for lesion of insular cortex and Garfinkel et al., 2015, for the component of interoception (Introduction).

4) The conceptual interpretation of the experimental paradigm as an "interoceptive attention" task should be revisited. While the task relies on shifting attention (between intero-and exteroceptive domains), it primarily serves to provide a measure of intero- and exteroceptive accuracy or sensitivity (as also reflected by your analysis in terms of d'), and appears to be more adequately described in these terms.

As we explained regarding the issue of the definition of interoceptive attention, attention is the process and the awareness is the outcome. In most psychological experiments, outcome is use to index the underlying processing. Here, we followed the same logic by associating interoceptive attention with interoceptive accuracy/sensitivity (which are the outcomes). We have two task conditions in the Breath Detection Task (BDT) and Dot Detection Task (DDT). Our experimental design was block design with two tasks administered in separate runs/blocks. Interoceptive attention (and corresponding performance) is the focus, as reflected by the title of this manuscript, with exteroceptive attention as the contrast baseline. In the revised manuscript, we have made the description more clearly. For example, when referring to the concept, we used “interoceptive attention”; when referring to the task, we used “BDT” and “DDT”; when referring to performance, we used “interoceptive accuracy/sensitivity”.

5) The proposed experimental paradigm is innovative and has much potential for future studies of respiratory interoception. However, there are some potential problems that may need consideration. First, the control condition appears to require a different cognitive process than the condition of interest. The latter requires a temporally extended matching process; the former requires a detection process that terminates once a dot has appeared. Second, the delay was a set interval of 400 ms, rather than a proportion of the individual's respiratory cycle. This may partially determine task difficulty and performance across individuals. Third, given that the task is novel, it would be important to see more details of task performance, e.g. plots of individual accuracy rates, analysis of reaction times and signal-detection theoretic considerations (i.e., where they more biased for either interoceptive or exteroceptive conditions?). The task seems very easy compared to standard heartbeat detection tasks: were there ceiling effects (i.e., did any participants have 100% accuracy)? Finally, the task does not represent a pure probe of interoception as respiratory processes can also be tracked using exteroceptive and proprioceptive information. It thus seems likely that participants relied on a mix of interoceptive, exteroceptive, and proprioceptive information for performing the task. These issues do not invalidate the task, but they deserve a critical and frank discussion so that the reader is aware of the limitations of the paradigm.

We would like to thank the reviewers for the comment on the innovation of our task. Here we respond to the questions and the suggestion point by point.

1) We agree with the reviewers that the exteroceptive process would stop as soon as participants detected the dot. The same strategy would also apply to the interoceptive task: the interoception would terminate if the participant detected the non-matched/matched curve to their breath. The reviewers’ concern does exist if these two processes do not terminate at the same time point. However, we randomized the time point of the appearance of the dot trial by trial, although it is not guaranteed that the average termination time points of the two tasks are the same.

2) We agree with the reviewers that manipulating the feedback delay according to each individual’s respiratory cycle is a better way in terms of controlling for task difficulty across subjects. However, we did not do so by calculating immediate respiratory cycle online due to concern about precision issues arising from the variations of the breath curves and an appropriate implication of the algorithms. Note that the 400 ms delay was determined based on a proportion (~1/10) of an average cycle of normal healthy people which is 3~4 s/cycle. However, we agree that this fixed delay could not ensure equal subjective difficulties across participants. To acknowledge this issue and for future improvement, we have added a sentence in Materials and methods (subsection “Task implementations”, second paragraph) and in Discussion (Subsection “The interoceptive task in the respiratory domain”, last paragraph).

3) We have added the box plots for individual accuracy rate, reaction time, d-prime, and beta in the Supplementary Figure 1. Participants were less accurate and slower in the BDT than in the DDT, while were more biased in the DDT than in the BDT (see subsection “Behavioral results of the fMRI study”). There was only one participant who reached 100% accuracy in the BDT and one participant reached 100% accuracy in the DDT out of 72 participants. Here we provided the plots to show the distributions of the performance, which showed no evidence of a ceiling effect.

Author response image 5
Download asset Open asset

4) We thank the reviewers for making the insightful suggestion for a critical and frank discussion about the mixed nature of interoceptive signals with exteroceptive and proprioceptive information as in our measure using the breath detection task (BDT). The interoceptive attention is toward both interoceptive and proprioceptive information, but not exteroceptive information (Gu, Hof, Friston, and Fan, 2013). In our design, we have the contrast task of dot detection task (DDT) for a measure of exteroception, so that the subtraction of BDT and DDT leaves the components of interoceptive and proprioceptive processing which is the classical definition of interoception (of bodily somatic and visceral signals or responses). We have added a sentence in Discussion to acknowledge that: “The BDT does not represent a pure probe of interoception as respiratory processes can also be tracked using exteroceptive and proprioceptive information. It thus seems likely that participants relied on a mix of interoceptive, exteroceptive, and proprioceptive information for performing the task. In our design, we have the contrast task of dot detection task (DDT) for a measure of exteroception, so that the subtraction of BDT and DDT leaves the components of interoceptive and proprioceptive processing of interoception.”

6) There are some issues with the statistical analysis and reporting. Exact p-values, test statistics, and standardized effect sizes should be reported for all analyses. Numerous tests are reported as one-sided; this needs to be justified or replaced by two-sided tests. Non-significant results should not be presented as evidence for the absence of a difference (e.g., in the lesion analysis); this corresponds to accepting the null hypothesis and should be replaced by a corresponding Bayesian test.

We thank the reviewers for pointing out those inaccuracies in analyzing and reporting of the statistics. We have now reported exact p values, test statistics, and effect sizes for all analyses. In the revised manuscript, we have all statistical tests as two-sided except for lesion study. Although the significant differences between AIC lesion patients and normal controls survived two-tailed tests, we prefer to use one-tailed tests because we have the hypothesis that lesions of a specific brain region (e.g., AIC) would induce deficits in behavioral response. We have added a sentence in Materials and methods to justify the reason using one-tailed test in the lesion study (subsection “Behavioral data analysis of the lesion study”). Regarding to the non-significant results (e.g., in behavioral analyses of the fMRI and the lesion studies), we have followed reviewers’ suggestions to conduct Bayesian tests. We have updated the corresponding Materials and methods and Results in the revised manuscript.

[Editors' note: the author responses to the re-review follow.]

We were impressed by the effort you invested in acquiring an additional dataset with concomitant measures of cardiac and respiratory activity. However, we continue to think that the statistical analysis needs to account for task-induced variations in breathing which can profoundly impact on BOLD measurements. We did read the paper (Miller and Chapman) that you attached for justification of omitting respiratory measures from the statistical model but must confess that we did not find it very insightful in relation to the current problem; in particular, equating the current issue with "Lord's paradox" (which is a rather specific case) seems misleading.

The problem in your analysis is a very generic one: including or excluding a confound regressor that is correlated to a regressor of interest in a GLM amounts to an active decision how shared variance is interpreted – or, put differently, whether one wishes to maximise sensitivity or specificity of the analysis. We think that for a study that reports the effect of a cognitive intervention for the first time, specificity is more important: the reader would like to be assured that activations attributed to the cognitive intervention are not merely driven by physiological effects. We agree that the interaction effect should be protected against task-induced breathing changes. The main effect of task, however, is not; and it is arguably of greater importance for the message of the paper.

For these reasons, we are not convinced it is a good idea to pool the two groups and report analyses without including regressors that represent physiological (respiratory) noise. We also thought that the RETROICOR analysis presented in the response letter (2nd order respiratory regressors only and no cardiac-respiratory interactions) is unusually lenient.

In our view, these problems are too substantial to proceed with in-depth peer-review. If you would like eLife to continue considering the paper, we would recommend that the paper (i) reports analyses from both samples separately, (ii) discusses the potential problems of interpretation in the first sample, and (iii) includes a rigorous RETROICOR correction of breathing effects for the second sample. You could boost the statistical sensitivity of the second analysis by using the FWE-corrected activations from the first analysis in order to specify a mask for reducing the search volume for FWE correction in the second analysis. In this way, you would use the higher statistical sensitivity of the first analysis in order to identify regions where the cognitive process of interest may take place and then test in the second sample, with due consideration of potentially confounding effects, whether this can be corroborated.

We are very sorry that we cannot be more positive at this stage and understand that this must be disappointing for you, given the substantial effort you have invested in the revision of this paper. We do hope, however, that the recommendation above is helpful.

Thank you for making constructive suggestions. We have followed your recommendations to address the issues in the data analysis and reporting. We agree that compared to sensitivity, specificity should be prioritized when exploring the effect of cognitive processing for the first time. Therefore, it is necessary to ensure that the AIC activity reported in our study is attributed to interoceptive processing, rather than driven by physiological effects. Thus, we analyzed fMRI data of the two samples separately in the following three steps:

First, we re-conducted connectivity analyses (i.e., PPI and DCM) of the first sample. You kindly pointed out that the selection of the seed region in PPI analysis should be independent of the psychological variable. In the context of a 2 × 2 factorial design, the physiological signals should be extracted from one main effect while the psychological variable should be the other. Therefore, in the new analysis we selected the seed region of AIC based on the main effect of feedback delay (the contrast of delayed versus non-delayed condition), and used the other main effect of interoception (i.e. the contrast of interoceptive versus exteroceptive attention, which was orthogonal to the main effect of feedback delay) as the psychological variable. The new results from PPI analysis and DCM showed that there was enhanced connectivity between AIC and posterior central gyrus modulated by interoceptive attention, while there was decreased connectivity between AIC and visual cortex (i.e., V2/3) modulated by exteroceptive attention. These results were consistent with what we reported in the previous versions. The sections of connectivity analyses were updated accordingly in the manuscript.

Second, we discussed the potential problems of the interpretation in the first sample due to the confounding effect of physiological activity. Specifically, the experimental manipulation in our study would inherently cause a change in respiratory patterns (i.e. amplitude and frequency) between interoceptive and exteroceptive tasks. The difference on physiological states might cause a change in global BOLD signals, which would confound with the effect of interoceptive processing we are interested in. We have added a detailed discussion of this problem in Materials and methods, and described the solution to correct for this confounding effect in this section. Following your suggestions, we regressed out physiological artifacts in the second sample by applying a rigorous RETROICOR correction of the brain activation associated with physiological activity, which included 5th order respiratory and cardiac regressors, and cardiac-respiratory interactions (see details in Materials and methods).

Third, to maximize the specificity of brain response while also boosting the sensitivity, we conducted an ROI analysis in the second sample to confirm that the involvement of AIC in interoceptive processing was not subject to physiological artifacts. Specifically, we identified the ROI of the AIC from the interaction contrast in the first sample, and then extracted the signals of the ROI in the second sample after a rigorous RETROICOR correction was applied to regress out respiratory and cardiac noises. For the ROI analysis of the second sample, there was a significant interaction between attentional focus (interoceptive and exteroceptive) and feedback (with and without delay) in both left and right AIC (left: F(1,27) = 5.77, p = 0.024; right: F(1,27) = 5.73, p = 0.024; Figure 5A), which is a confirmation of what we found from the first sample. The correlation between the interaction effect of the right AIC activity and relative interoceptive accuracy was significant (Pearson r = 0.36, p = 0.03, one-tailed). In addition, the whole brain analysis of the second sample showed that there was the significant overlap in the main and the interaction effects with and without physiological noise correction (see Figure 5—figure supplement 2). The difference of the signals of the AIC between the analyses with and without physiological corrections was not significant, suggesting that the effect of the AIC was not significantly influenced by the physiological noises (see Figure 5—figure supplement 3). Altogether, these results confirm that the AIC was actively engaged in interoceptive processing. We have updated the corresponding sections in Materials and methods and Results.

[Editors' note: the author responses to the re-review follow.]

Major points:

1) The wording "… checked that the AIC ROI results were not dependent on the (independent) ROI selection.…" is confusing. Presumably you wanted to say something like "… checked how much the AIC ROI results were affected by physiological noise correction.…"? In direct relation to this point, it is rather surprising to see such little effects of physiological noise correction on insula activity. Typically, physiological noise regressors (RETROICOR) do explain a substantial amount of BOLD signal in the insula. The particular statistical test you used asks whether specific contrasts are altered by the inclusion vs. exclusion of physiological noise regressors (which is fine) but is not sensitive to the question whether insular activity is affected by physiological noise at all (as implied by your wording in the subsection “ROI analysis results of the fMRI study of the second sample”). As a sanity check, it would be worth performing an additional F-test spanning all RETROICOR regressors. If this test does not show significant insula activation, it would seem wise to double-check the RETROICOR analysis, in order to make sure there are no errors.

We thank you for your conscientious suggestions regarding RETROICOR correction. Following your suggestion, we have conducted an F-test across all RETROICOR regressors. The results revealed a significant physiological impact on the insula as well as the whole brain activation (please see Author response image 6), suggesting that the RETROICOR indeed explained a substantial amount of BOLD signal under the interoceptive condition (BDT) and exteroceptive condition (DDT). However, in our last analysis, contrasting BDT versus DDT did not reveal a significant impact of physiological noise correction on insula activation. We speculate that the reason might be that standard Fourier harmonics of RETROICOR correction do not account for all effects of physiological noises, i.e., CO2 and O2 effect caused by respiratory per unit time (RVT) as you suggested in the third comment below. Therefore, we have conducted a new physiological correction by including another two nuisance regressors of respiratory volume (RV) and heart rate (HR) that convolved with the “respiration response function (RRF)” and the “cardiac response function (CRF)” respectively (Verstynen and Deshpande, 2011). This new correction revealed a reduced activation in the insula as well as the whole brain under the main effect of interoceptive attention (the contrast of BDT vs. DDT) compared to no correction (see Figure 5—figure supplement 3A), suggesting that the RV/HR regressors indeed accounted for additional physiological noise induced by the RV and HR differences between the two tasks. However, the interaction contrast ([delayed – non-delayed] BDT – [delayed – non-delayed] DDT) was not much affected by the physiological correction, as revealed by an almost blank brain when comparing the contrast maps without and with physiological corrections (see Figure 5—figure supplement 3C), which further confirmed that the interaction contrast was not subject to the confounding effects caused by physiological changes between tasks. With this new physiological correction, we have re-conducted the ROI analyses of the second sample. The results of the ROI analysis were consistent with what we reported in the previous version of the manuscript. We have revised that misleading sentence and updated the Materials and methods and Results sections accordingly.

Author response image 6
Download asset Open asset
The F-test across all RETROICOR regressors.

Voxelwise p < 0.001.

2) Materials and methods: "The corresponding four regressors were generated by convolving the onset vectors of each trial type with a standard canonical hemodynamic response function (HRF)". Was each trial modelled as an event or a block? The methods describe each stimulus period lasting 12 seconds, which would appear more akin to a block design for the GLM?

We apologize that the sentence describing the generation of task-related regressors was not clearly written. We modelled each trial as a mini-block with a duration of 12 seconds. We have revised this sentence to “Each trial was modelled as an epoch-related function by specifying an onset time and a duration of 12 s. The corresponding four regressors were generated by convolving the onset of each trial with the standard canonical hemodynamic response functions (HRF) with a duration of 12 s, i.e., by convolving each trial block with HRF, equivalent to a box-car function.”

3) The value of the mention of CO2 and O2 in this manuscript is questionable – these effects would need to be accounted for by either measuring them and regressing them out, or using an approximation such as RVT (respiratory volume per unit of time) regressors, which do not appear to be used here. Standard cardiac and respiratory waveforms and harmonics do not account for these effects. This paradigm would likely induce very slight hyperventilation when attention is directed towards monitoring breathing curves, which would result in a decrease in expired CO2 over the 12 second stimulus period (and the resulting washout period), which would induce a global over-estimation of the BOLD activity related to the task. RVT regressors could be included in the RETROICOR to account for this. They actually mention that there is a difference in respiratory volume between tasks in the subsection “ROI analysis results of the fMRI study of the second sample”.

We are very thankful for your helpful suggestions on controlling for RVT. Following your suggestion, we have re-analyzed the data of the second fMRI study by adding another two nuisance regressors of RVe and HR according to Verstynen and Deshpande, 2011. The inclusion of those additional two regressors indeed accounted for the RV and HR difference between tasks, and we found that there was an impact of physiology correction on insula activation under the main effect of interoceptive attention (the contrast of BDT vs. DDT) (please refer to our response to the first comment above for details).

4) The main effect of the task (interoceptive attention vs exteroceptive attention) is very large (Figure 2). However, it should be noted that the participants found the interoceptive task more difficult than the exteroceptive task, and thus these differences in brain activity are very likely associated with task difficulty as well as the direction of attention. This is probably worth mentioning somewhere in the Discussion?

We agree with you that the main effect of the task (interoceptive attention vs exteroceptive attention) reflected both task-specific effect (i.e. task difficulty and respiratory characteristics difference between tasks) and attention deployment. We have discussed this point in the Materials and methods section after we described the specification of the contrasts and then clarified that the interaction contrast can disentangle the effects to some extent.

5) Please accept our apologies – we should have noted this earlier – but the analyses presented in the lesion study suffer from a major problem. The authors are conducting non-parametric tests between participants in the interoceptive and exteroceptive condition separately, and then interpreting the presence vs lack of significance as evidence for a specific effect of AIC lesions on interoceptive attention. This, however, is an erroneous conclusion (see also https://www.nature.com/articles/nn.2886), and instead it would be necessary to demonstrate a significant group (AIC vs control) by condition (intero vs extero) interaction. It is important to perform the correct test, especially because these are the most controversial findings of the paper. It is debatable whether previous lesion research has provided any valid evidence for AIC lesions on interoceptive sensitivity and emotional awareness (in a previous review, the authors were asked to consider this work, but this appears to have been ignored).

We thank you for pointing out the correct way to conduct the statistical analyses of our lesion study. We have followed your suggestion and conducted a non-parametric test for the interaction effect using R (please refer to “Nonparametric Tests for the Interaction in Two-way Factorial Designs Using R”, by Jos Feys). In specific, we used the npIntFactRep function (from the npIntFactRep package, see https://cran.r-project.org/web/packages/npIntFactRep/npIntFactRep.pdf) that yields an aligned rank test for an interaction in the two-way mixed design with the group (normal controls, AIC lesions, and brain damage controls) as the between-subject factor and with the task (BDT and DDT) as the within-subject factor. The results showed a significant interaction between group and task on performance accuracy (F(2,21) = 5.19, p = 0.015) and discrimination sensitivity (dʹ) (F(2,21) = 4.77, p = 0.023). By contrast, we did not find significant interaction effect on β (F(2,21) <1, p = 0.65). We then reported simple comparison effects with our original bootstrapping tests between groups under interoceptive and exteroceptive condition separately. We have updated the Materials and methods and Results sections accordingly.

We apologize that we failed to discuss the AIC lesion patient study you mentioned. We have examined the literature investigating the impact of AIC lesions on interoceptive processes and added our speculations regarding on the inconsistency to the Discussion section. Most previous lesion studies indicated the interoceptive deficits with AIC lesions (Critchley and Garfinkel, 2017; García-Cordero et al., 2016; Ibanez, Gleichgerrcht, and Manes, 2010; Ronchi et al., 2015; Starr et al., 2009; Terasawa, Kurosaki, Ibata, Moriguchi, and Umeda, 2015; Wang et al., 2014), supporting the notion that interoceptive accuracy relies on a widely distributed network with the insular cortex as a key node (Craig, 2002; Critchley and Harrison, 2013). However, the preservations of interoceptive (Khalsa, Rudrauf, Feinstein, and Tranel, 2009) and self-awareness across a large battery of tests (Philippi et al., 2012) were documented in one patient with bilateral insula damages. These studies are mostly based on subjective report focusing on the “feeling/awareness” (Khalsa et al., 2009) that might be compensated by other brain structures such as brainstem and subcortical structures, e.g., nucleus tractus solitaries, the parabrachial nucleus, area postrema and hypothalamus (A. Damasio, Damasio, and Tranel, 2012), frontal (IFG) and temporal regions, e.g., amygdala, superior temporal gyrus, and temporal pole (García-Cordero et al., 2016; Shany-Ur et al., 2014). In the current study, the BDT challenged interoceptive attention that requires the integration of interoceptive awareness and accuracy. Our examination of interoceptive attention in patients with focal AIC lesion showed that lesions of the AIC were associated with the deficit in the performance, indicating that the AIC is critical in supporting the precision of interoceptive processing.

6) There are also some concerns – and, again our apologies for not having addressed this earlier – regarding the analyses of the correlation between task measures and questionnaires (subsection “Behavioral results of the fMRI studies”). Specifically, the use of multiple one-sided tests seems questionable and would need a strong motivation, and there is a lack of multiple comparison correction. There were also questions about power: if you pool the two samples you may have reasonable power (i.e., for an effect size of|r|=0.3, N=72 would give you 75% power of detecting a significant effect at alpha=0.05, two-sided) for a single test, but this would diminish when taking into account multiple comparisons (with lower alpha as a result). It seems fair to ask that the analysis is fully transparent with regard to power, uses two-sided tests and multiple comparison correction, and tones down the interpretation of the results considerably.

We agree with you that one-sided tests without multiple comparisons correction are questionable when we do not have a strong hypothesis. In the revised version, we have added a correlational analysis by pooling the two samples (N = 72), while also reporting the results separately for each sample to keep consistent with the imaging results. For all the tests, we have used two-sided tests with multiple comparison corrections. Indeed, the correlations between interoceptive performance and questionnaire scores did not survive after the correction. Therefore, we have toned down the interpretations of interoceptive attention in relation to emotional awareness.

7) Following directly from the previous point, one of the reviewers made the following suggestion which we would suggest you consider: "My advice is the following: take all of the correlation analyses and throw them into a giant Bayesian correlation table with appropriate priors on correlation strength, and make these supplementary analyses flagged clearly as exploratory in nature. Flag up the ones that show Bayes factors greater than 3 under a two-sided test, and in particular discuss any who are implicated in both study 1 and study 2. Refocus the paper to emphasize the novelty and importance of understanding respiratory awareness and make it clear that the link to emotion is more speculative – a fascinating area that future large-scale studies can target by using the task presented here."

We appreciate your helpful suggestions. We have conducted Bayesian correlation analyses for sample 1 and 2, and also the pooled sample by specifying a beta prior. Then, we flagged up the ones showing Bayes factor greater than 3 under a two-sided test (please see Supplementary file 3). Results showed that the interoceptive performance was only robustly correlated with subjective difficulty but not with other questionnaires. Following your suggestions, we have toned down the interpretations in Discussion.

https://doi.org/10.7554/eLife.42265.038

Article and author information

Author details

  1. Xingchao Wang

    1. Department of Neurosurgery, Beijing Tiantan Hospital, Capital Medical University, Beijing, China
    2. China National Clinical Research Center for Neurological Diseases, Beijing, China
    Contribution
    Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Writing—original draft, Writing—review and editing
    Contributed equally with
    Qiong Wu
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-1281-1973

  2. Qiong Wu

    1. Beijing Key Lab of Learning and Cognition, School of Psychology, Capital Normal University, Beijing, China
    2. School of Psychological and Cognitive Sciences, Peking University, Beijing, China
    Contribution
    Data curation, Formal analysis, Investigation, Methodology, Writing—original draft, Writing—review and editing
    Contributed equally with
    Xingchao Wang
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0003-2597-4322

  3. Laura Egan

    Department of Psychology, Queens College, The City University of New York, New York, United States
    Contribution
    Methodology, Writing—original draft，Writing—review and editing
    Competing interests
    No competing interests declared

  4. Xiaosi Gu

    1. Department of Psychiatry, Icahn School of Medicine at Mount Sinai, New York, United States
    2. Nash Family Department of Neuroscience, Icahn School of Medicine at Mount Sinai, New York, United States
    3. The Mental Illness Research, Education, and Clinical Center, The James J. Peter Veterans Affairs Medical Center, New York, United States
    Contribution
    Methodology, Writing—original draft, Writing—review and editing
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-9373-987X

  5. Pinan Liu

    1. Department of Neurosurgery, Beijing Tiantan Hospital, Capital Medical University, Beijing, China
    2. China National Clinical Research Center for Neurological Diseases, Beijing, China
    Contribution
    Funding acquisition, Investigation, Writing—review and editing
    Competing interests
    No competing interests declared

  6. Hong Gu

    Neuroimaging Research Branch, Intramural Research Program, National Institute on Drug Abuse, Baltimore, United States
    Contribution
    Methodology, Writing—original draft，Writing—review and editing
    Competing interests
    No competing interests declared

  7. Yihong Yang

    Neuroimaging Research Branch, Intramural Research Program, National Institute on Drug Abuse, Baltimore, United States
    Contribution
    Methodology, Writing—original draft, Writing—review and editing
    Competing interests
    No competing interests declared

  8. Jing Luo

    Beijing Key Lab of Learning and Cognition, School of Psychology, Capital Normal University, Beijing, China
    Contribution
    Resources, Writing—original draft,Writing—review and editing
    Competing interests
    No competing interests declared

  9. Yanhong Wu

    1. School of Psychological and Cognitive Sciences, Peking University, Beijing, China
    2. Beijing Key Laboratory of Behavior and Mental Health, Peking University, Beijing, China
    Contribution
    Funding acquisition, Supervision, Investigation, Methodology, Writing—original draft, Writing—review and editing
    For correspondence
    wuyh@pku.edu.cn
    Competing interests
    No competing interests declared

  10. Zhixian Gao

    1. Department of Neurosurgery, Beijing Tiantan Hospital, Capital Medical University, Beijing, China
    2. China National Clinical Research Center for Neurological Diseases, Beijing, China
    Contribution
    Data curation, Funding acquisition, Supervision, Investigation, Writing—original draft, Writing—review and editing
    For correspondence
    gaozx@ccmu.edu.cn
    Competing interests
    No competing interests declared

  11. Jin Fan

    1. Department of Psychology, Queens College, The City University of New York, New York, United States
    2. Department of Psychiatry, Icahn School of Medicine at Mount Sinai, New York, United States
    3. Nash Family Department of Neuroscience, Icahn School of Medicine at Mount Sinai, New York, United States
    4. Friedman Brain Institute, Icahn School of Medicine at Mount Sinai, New York, United States
    Contribution
    Conceptualization, Supervision, Funding acquisition, Investigation, Methodology, Writing—original draft, Writing—review and editing
    For correspondence
    jin.fan@qc.cuny.edu
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0001-9630-8330

Funding

Beijing Municipal Administration of Hospitals (Youth Program QML20170503)

  • Xingchao Wang

National Natural Science Foundation of China (81600931)

  • Xingchao Wang

Capital Health Development Research Project of Beijing (2016-4-1074)

  • Xingchao Wang

Brain Research Project of Beijing (Z16110002616014)

  • Xingchao Wang
  • Pinan Liu

China Postdoctoral Science Foundation (2016M600835)

  • Qiong Wu

National Institute on Drug Abuse (1R01DA043695)

  • Xiaosi Gu

The Mental Illness Research, Education, and Clinical Center, James J. Peter Veterans Affairs Medical Center (MIRECC VISN 2)

  • Xiaosi Gu

National Institute on Drug Abuse (Intramul Research Program)

  • Hong Gu
  • Yihong Yang

National Basic Research Program of China (973 Program) (973-2015CB351800)

  • Yanhong Wu

National Natural Science Foundation of China (31771205)

  • Yanhong Wu

National Natural Science Foundation of China (61690205)

  • Yanhong Wu

National Natural Science Foundation of China (81328008)

  • Zhixian Gao
  • Jin Fan

National Natural Science Foundation of China (81729001)

  • Zhixian Gao
  • Jin Fan

National Institute of Mental Health (R01MH094305)

  • Jin Fan

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

Acknowledgements

We thank Dr. Thomas Beck and Dr. Tian-Yi Qian from Siemens Healthcare for providing the simultaneous multi-slice EPI sequence for fMRI data acquisition. This work was supported by the National Natural Science Foundation of China (grant number: 81729001, 81328008) to JF and ZG. JF was also supported by the National Institute of Mental Health of the National Institutes of Health (NIH) under Award Number R01 MH094305. YW was supported by the research grant of 973 (grant number: 973-2015CB351800) and National Natural Science Foundation of China (grant number: 31771205, 61690205). YY and HG were supported by the Intramural Research Program, National Institute on Drug Abuse, NIH. XW and PL were supported by Beijing Municipal Science and Technology Commission (grant number: Z161100002616014). XW was also supported by the National Natural Science Foundation of China (grant number: 81600931), Beijing Municipal Administration of Hospital’ Youth Programs (code: QML20170503) and Capital Health Development Research Project of Beijing, China (grant number: 2016-4-1074). QW was supported by China Postdoctoral Science Foundation (grant number: 2016M600835). XG was supported by the National Institute on Drug Abuse (grant number 1R01DA043695) and the Mental Illness Research, Education, and Clinical Center (MIRECC; VISN3), James J Peters VA Medical Center, Bronx, NY. Dr. Nicholas Van Dam and Evelyn Ramirez were involved at the early stage of the study on interoception. We also thank Shira Russell-Giller and Liat Kofler for their help on proof reading. The authors also thank the National Center for Protein Sciences at Peking University in Beijing, China, for assistance with the MRI data acquisition.

Ethics

Human subjects: All participants in fMRI study and in lesion study were gave written informed consent in accordance with the procedures and protocols approved by The Human Subjects Review Committee of Peking University and by The Institutional Review Board of the Beijing Tiantan Hospital, Capital Medical University, respectively.

Senior Editor

  1. Michael J Frank, Brown University, United States

Reviewing Editor

  1. Klaas Enno Stephan, University of Zurich and ETH Zurich, Switzerland

Reviewers

  1. Klaas Enno Stephan, University of Zurich and ETH Zurich, Switzerland
  2. Olivia K Faull, University of Oxford, United Kingdom
  3. Micah Allen, UCL, United Kingdom

Publication history

  1. Received: September 24, 2018
  2. Accepted: April 13, 2019
  3. Accepted Manuscript published: April 15, 2019 (version 1)
  4. Version of Record published: April 29, 2019 (version 2)

Copyright

This is an open-access article, free of all copyright, and may be freely reproduced, distributed, transmitted, modified, built upon, or otherwise used by anyone for any lawful purpose. The work is made available under the Creative Commons CC0 public domain dedication.

Metrics

  • 11,922
    Page views
  • 922
    Downloads
  • 59
    Citations

Article citation count generated by polling the highest count across the following sources: Scopus, Crossref, PubMed Central.

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Open citations (links to open the citations from this article in various online reference manager services)

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)

  1. Xingchao Wang
  2. Qiong Wu
  3. Laura Egan
  4. Xiaosi Gu
  5. Pinan Liu
  6. Hong Gu
  7. Yihong Yang
  8. Jing Luo
  9. Yanhong Wu
  10. Zhixian Gao
  11. Jin Fan
(2019)
Anterior insular cortex plays a critical role in interoceptive attention
eLife 8:e42265.
https://doi.org/10.7554/eLife.42265

Categories and tags

Research organism