Skip to main content

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

  • Article
  • Published:

The neuronal basis of fear generalization in humans

Nature Neuroscience volume 18, pages 1811–1818 (2015)Cite this article

Subjects

Abstract

Organisms tend to respond similarly to stimuli that are perceptually close to an event that predicts adversity, a phenomenon known as fear generalization. Greater dissimilarity yields weaker behavioral responses, forming a fear-tuning profile. The perceptual model of fear generalization assumes that behavioral fear tuning results from perceptual similarities, suggesting that brain responses should also exhibit the same fear-tuning profile. Using fMRI and a circular fear-generalization procedure, we tested this prediction. In contrast with the perceptual model, insula responses showed less generalization than behavioral responses and encoded the aversive quality of the conditioned stimulus, as shown by high pattern similarity between the conditioned stimulus and the shock. Also inconsistent with the perceptual model, object-sensitive visual areas responded to ambiguity-related outcome uncertainty. Together these results indicate that fear generalization is not passively driven by perception, but is an active process integrating threat identification and ambiguity-based uncertainty to orchestrate a flexible, adaptive fear response.

This is a preview of subscription content, access via your institution

Access options

Buy this article

  • Purchase on Springer Link
  • Instant access to full article PDF
Buy now

Prices may be subject to local taxes which are calculated during checkout

Figure 1: Experimental procedure and autonomic and behavioral generalization profiles.
Figure 2: Brain areas exhibiting fear-tuned responses.
Figure 3: Correlation between SI and neuronal and perceptual tuning parameters.
Figure 4: Effect of conditioning on insular multivariate response patterns.

Similar content being viewed by others

References

  1. Pavlov, I. Conditioned Reflexes: an Investigation of the Physiological Activity of the Cerebral Cortex (Oxford University Press Humphrey Milford, 1927).

  2. Bass, M.J. & Hull, C.L. The irradiation of a tactile conditioned reflex in man. J. Comp. Psychol. 17, 47–65 (1934).

    Article  Google Scholar 

  3. Guttman, N. & Kalish, H.I. Discriminability and stimulus generalization. J. Exp. Psychol. 51, 79–88 (1956).

    Article  CAS  PubMed  Google Scholar 

  4. Ghirlanda, S. & Enquist, M. A century of generalization. Anim. Behav. 66, 15–36 (2003).

    Article  Google Scholar 

  5. Shepard, R.N. Toward a universal law of generalization for psychological science. Science 237, 1317–1323 (1987).

    Article  CAS  PubMed  Google Scholar 

  6. Tenenbaum, J.B. & Griffiths, T.L. Generalization, similarity, and Bayesian inference. Behav. Brain Sci. 24, 629–640 (2001).

    Article  CAS  PubMed  Google Scholar 

  7. Lissek, S. et al. Generalization of conditioned fear-potentiated startle in humans: experimental validation and clinical relevance. Behav. Res. Ther. 46, 678–687 (2008).

    Article  PubMed  PubMed Central  Google Scholar 

  8. Resnik, J., Sobel, N. & Paz, R. Auditory aversive learning increases discrimination thresholds. Nat. Neurosci. 14, 791–796 (2011).

    Article  CAS  PubMed  Google Scholar 

  9. Schechtman, E., Laufer, O. & Paz, R. Negative valence widens generalization of learning. J. Neurosci. 30, 10460–10464 (2010).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  10. Holt, D.J. et al. A parametric study of fear generalization to faces and non-face objects: relationship to discrimination thresholds. Front. Hum. Neurosci. 8, 624 (2014).

    Article  PubMed  PubMed Central  Google Scholar 

  11. Dunsmoor, J.E., Prince, S.E., Murty, V.P., Kragel, P.A. & LaBar, K.S. Neurobehavioral mechanisms of human fear generalization. Neuroimage 55, 1878–1888 (2011).

    Article  PubMed  Google Scholar 

  12. Dunsmoor, J.E. & Murphy, G.L. Categories, concepts, and conditioning: how humans generalize fear. Trends Cogn. Sci. 19, 73–77 (2015).

    Article  PubMed  PubMed Central  Google Scholar 

  13. Dymond, S., Dunsmoor, J.E., Vervliet, B., Roche, B. & Hermans, D. Fear generalization in humans: systematic review and implications for anxiety disorder research. Behav. Ther. 26, 561–582 (2015).

    Article  Google Scholar 

  14. Lissek, S. et al. Overgeneralization of conditioned fear as a pathogenic marker of panic disorder. Am. J. Psychiatry 167, 47–55 (2010).

    Article  PubMed  Google Scholar 

  15. Cha, J. et al. Circuit-wide structural and functional measures predict ventromedial prefrontal cortex fear generalization: implications for generalized anxiety disorder. J. Neurosci. 34, 4043–4053 (2014).

    Article  CAS  PubMed  Google Scholar 

  16. Lissek, S. et al. Generalized anxiety disorder is associated with overgeneralization of classically conditioned fear. Biol. Psychiatry 75, 909–915 (2014).

    Article  PubMed  Google Scholar 

  17. Greenberg, T., Carlson, J.M., Cha, J., Hajcak, G. & Mujica-Parodi, L.R. Ventromedial prefrontal cortex reactivity is altered in generalized anxiety disorder during fear generalization. Depress. Anxiety 30, 242–250 (2013).

    Article  PubMed  Google Scholar 

  18. Charney, D.S., Deutch, A.Y., Krystal, J.H., Southwick, S.M. & Davis, M. Psychobiologic mechanisms of posttraumatic stress disorder. Arch. Gen. Psychiatry 50, 295–305 (1993).

    CAS  PubMed  Google Scholar 

  19. Lissek, S. Toward an account of clinical anxiety predicated on basic, neurally mapped mechanisms of Pavlovian fear-learning: the case for conditioned overgeneralization. Depress. Anxiety 29, 257–263 (2012).

    Article  PubMed  PubMed Central  Google Scholar 

  20. Lissek, S. et al. Neural substrates of classically conditioned fear-generalization in humans: a parametric fMRI study. Soc. Cogn. Affect. Neurosci. 9, 1134–1142 (2014).

    Article  PubMed  Google Scholar 

  21. Grupe, D.W. & Nitschke, J.B. Uncertainty is associated with biased expectancies and heightened responses to aversion. Emotion 11, 413–424 (2011).

    Article  PubMed  PubMed Central  Google Scholar 

  22. Lake, J.I. & LaBar, K.S. Unpredictability and uncertainty in anxiety: a new direction for emotional timing research. Front. Integr. Neurosci. 5, 55 (2011).

    Article  PubMed  PubMed Central  Google Scholar 

  23. Grupe, D.W. & Nitschke, J.B. Uncertainty and anticipation in anxiety: an integrated neurobiological and psychological perspective. Nat. Rev. Neurosci. 14, 488–501 (2013).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  24. Grillon, C., Baas, J.P., Lissek, S., Smith, K. & Milstein, J. Anxious responses to predictable and unpredictable aversive events. Behav. Neurosci. 118, 916–924 (2004).

    Article  PubMed  Google Scholar 

  25. Hsu, M., Bhatt, M., Adolphs, R., Tranel, D. & Camerer, C.F. Neural systems responding to degrees of uncertainty in human decision-making. Science 310, 1680–1683 (2005).

    Article  CAS  PubMed  Google Scholar 

  26. Nader, K. & Balleine, B. Ambiguity and anxiety: when a glass half full is empty. Nat. Neurosci. 10, 807–808 (2007).

    Article  CAS  PubMed  Google Scholar 

  27. Kanwisher, N., McDermott, J. & Chun, M.M. The fusiform face area: a module in human extrastriate cortex specialized for face perception. J. Neurosci. 17, 4302–4311 (1997).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  28. Haxby, J.V. et al. Distributed and overlapping representations of faces and objects in ventral temporal cortex. Science 293, 2425–2430 (2001).

    CAS  PubMed  Google Scholar 

  29. Greenberg, T., Carlson, J.M., Cha, J., Hajcak, G. & Mujica-Parodi, L.R. Neural reactivity tracks fear generalization gradients. Biol. Psychol. 92, 2–8 (2013).

    Article  PubMed  Google Scholar 

  30. Roy, M., Shohamy, D. & Wager, T.D. Ventromedial prefrontal-subcortical systems and the generation of affective meaning. Trends Cogn. Sci. 16, 147–156 (2012).

    Article  PubMed  PubMed Central  Google Scholar 

  31. Dunsmoor, J.E., Kragel, P.A., Martin, A. & LaBar, K.S. Aversive learning modulates cortical representations of object categories. Cereb. Cortex 24, 2859–2872 (2014).

    Article  PubMed  Google Scholar 

  32. Kriegeskorte, N. Representational similarity analysis – connecting the branches of systems neuroscience. Front. Syst. Neurosci. 2 4 10.3389/neuro.06.004.2008 (2008).

  33. Visser, R.M., Scholte, H.S. & Kindt, M. Associative learning increases trial-by-trial similarity of BOLD-MRI patterns. J. Neurosci. 31, 12021–12028 (2011).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  34. Edeline, J.M. & Weinberger, N.M. Receptive field plasticity in the auditory cortex during frequency discrimination training: selective retuning independent of task difficulty. Behav. Neurosci. 107, 82–103 (1993).

    Article  CAS  PubMed  Google Scholar 

  35. Whalen, P.J. et al. Masked presentations of emotional facial expressions modulate amygdala activity without explicit knowledge. J. Neurosci. 18, 411–418 (1998).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  36. Craig, A.D. How do you feel? Interoception: the sense of the physiological condition of the body. Nat. Rev. Neurosci. 3, 655–666 (2002).

    Article  CAS  PubMed  Google Scholar 

  37. Paulus, M.P. & Stein, M.B. An insular view of anxiety. Biol. Psychiatry 60, 383–387 (2006).

    Article  PubMed  Google Scholar 

  38. Etkin, A. & Wager, T.D. Functional neuroimaging of anxiety: a meta-analysis of emotional processing in ptsd, social anxiety disorder, and specific phobia. Am. J. Psychiatry 164, 1476–1488 (2007).

    Article  PubMed  PubMed Central  Google Scholar 

  39. Kriegeskorte, N., Formisano, E., Sorger, B. & Goebel, R. Individual faces elicit distinct response patterns in human anterior temporal cortex. Proc. Natl. Acad. Sci. USA 104, 20600–20605 (2007).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  40. Herry, C. et al. Processing of temporal unpredictability in human and animal amygdala. J. Neurosci. 27, 5958–5966 (2007).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  41. Badia, P., Harsh, J. & Abbott, B. Choosing between predictable and unpredictable shock conditions: Data and theory. Psychol. Bull. 86, 1107–1131 (1979).

    Article  Google Scholar 

  42. Grillon, C. et al. Increased anxiety during anticipation of unpredictable aversive stimuli in posttraumatic stress disorder but not in generalized anxiety disorder. Biol. Psychiatry 66, 47–53 (2009).

    Article  PubMed  PubMed Central  Google Scholar 

  43. Büchel, C. & Dolan, R.J. Classical fear conditioning in functional neuroimaging. Curr. Opin. Neurobiol. 10, 219–223 (2000).

    Article  PubMed  Google Scholar 

  44. LaBar, K.S., Gatenby, J.C., Gore, J.C., LeDoux, J.E. & Phelps, E.A. Human amygdala activation during conditioned fear acquisition and extinction: a mixed-trial fMRI study. Neuron 20, 937–945 (1998).

    Article  CAS  PubMed  Google Scholar 

  45. Büchel, C., Morris, J., Dolan, R.J. & Friston, K.J. Brain systems mediating aversive conditioning: an event-related fMRI study. Neuron 20, 947–957 (1998).

    Article  PubMed  Google Scholar 

  46. Maloney, L.T. & Yang, J.N. Maximum likelihood difference scaling. J. Vis. 3, 573–585 (2003).

    Article  PubMed  Google Scholar 

  47. Watson, A.B. & Pelli, D.G. QUEST: a Bayesian adaptive psychometric method. Percept. Psychophys. 33, 113–120 (1983).

    Article  CAS  PubMed  Google Scholar 

  48. Aguirre, G.K., Mattar, M.G. & Magis-Weinberg, L. de Bruijn cycles for neural decoding. Neuroimage 56, 1293–1300 (2011).

    Article  PubMed  Google Scholar 

  49. Pelli, D.G. The VideoToolbox software for visual psychophysics: transforming numbers into movies. Spat. Vis. 10, 437–442 (1997).

    Article  CAS  PubMed  Google Scholar 

  50. Benedek, M. & Kaernbach, C. Decomposition of skin conductance data by means of nonnegative deconvolution. Psychophysiology 47, 647–658 (2010).

    PubMed  PubMed Central  Google Scholar 

  51. Mumford, J.A., Poline, J.-B. & Poldrack, R.A. Orthogonalization of regressors in fMRI models. PLoS ONE 10, e0126255 (2015).

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  52. Kruschke, J.K. Doing Bayesian Data Analysis: a Tutorial with R and BUGS (Academic Press, 2011).

  53. Desmond, J.E. & Glover, G.H. Estimating sample size in functional MRI (fMRI) neuroimaging studies: statistical power analyses. J. Neurosci. Methods 118, 115–128 (2002).

    Article  PubMed  Google Scholar 

  54. Efron, B. & Tibshirani, R. An Introduction to the Bootstrap (Chapman & Hall, 1994).

Download references

Acknowledgements

We thank S. Schall, J. Caplan, S. Geuter, T. Kietzmann, A. Etkin, B. Knutson and K. Friston for invaluable comments, and L. Kampermann for her assistance during the data acquisition. S.O. and C.B. are supported by the DFG SFB TRR 58.

Author information

Authors and Affiliations

  1. Department of Systems Neuroscience, University Medical Center Hamburg-Eppendorf, Hamburg, Germany

    Selim Onat & Christian Büchel

Authors
  1. Selim Onat
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Christian Büchel
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

S.O. and C.B. conceived and designed the study. S.O. acquired the data. S.O. and C.B. analyzed and interpreted the data. S.O. drafted the manuscript. C.B. and S.O. revised the manuscript.

Corresponding author

Correspondence to Selim Onat.

Ethics declarations

Competing interests

The authors declare no competing financial interests.

Integrated supplementary information

Supplementary Figure 1 Perceptual organization of face stimuli.

(a) Eight faces forming a circular similarity continuum along gender (x-axis) and identity dimensions (y-axis) (Right: most male face; Left: most female; Top/Bottom: two different facial identities). (b) Recovered perceptual organization of the stimulus set in a two-dimensional perceptual space. Distances between nodes (black dots) represent perceptual similarity between faces that are most likely to have caused the observed binary responses during the two-alternative forced choice task. Ellipses represent 95% chi-square confidence intervals (p =.05), computed following an affine inter-subject alignment. (c) The same perceptual analysis conducted after aligning faces with respect to the CS+ face (red: CS+; cyan: CS–). Circular color code represents the angular distance from the CS+ face. Distances (black arrows) between the CS+ (red node) and all other faces were used to estimate the perceptual tuning at the single-subject level shown in (e). (d) The negative log-likelihood of one-, two- or three-dimensional perceptual models averaged across subjects (mean ± SEM). Decreasing values indicate higher probability for the model. A one-dimensional perceptual model performed worse than the two- and three-dimensional models, whereas the increase from two- to three-dimensions was negligible. (e) Colored bars represent the estimated perceptual distances (shown with black lines in (c)) between the CS+ face and all other faces for two volunteers.

Supplementary Figure 2 Organization of single trials.

(a) The temporal structure of a single trial ending with an UCS delivery. (b) Oddball stimuli consisted of a randomly selected (each time anew) face with artificially added freckles (TR: repetition time; UCS: Unconditional stimulus).

Supplementary Figure 3 Computation of single-subject autonomic fear generalization profiles using skin conductance responses.

(a) From raw electrodermal recordings (not shown), we extracted the phasic skin conductance responses (SCR) using a deconvolution technique50 on a single-subject basis. Time-course of evoked skin conductance responses, averaged across trials during the test phase is depicted for a representative subject. Note that stimulus presentation lasted 1.5 seconds. We computed generalization profiles by averaging the signal within the temporal interval of interest (vertical gray dashed lines) separately for each of the eight individual faces. (b) Fear-tuning from two volunteers. Left panel: An example of wide autonomic fear-tuning (σ =.93 radians) derived from the data shown in (a). Right panel: Fear-tuning from another volunteer exhibiting sharper fear-tuning (σ =.57 radians). Error bars: SEM across trials. Black curves: Best fitting Gaussian model (p <.05, likelihood-ratio test).

Supplementary Figure 4 Brain areas with Gaussian fear tuning.

(a) Horizontal sections showing statistically significant clusters that exhibited fear tuning consistent with a Gaussian model (P <.05, corrected). (b) Fear tuning profiles of peak-voxels within each cluster (depicted in a) are shown for baseline (gray bars) and test phases (red bars, Mean ± SEM). Red curves (Gaussian model) and black horizontal lines (null model) represent the best-fitting model for each phase. In all cases the Gaussian model performed better than the null model only during the test phase but not before. These profiles are sorted from left to right with decreasing selectivity, that is with increasing tuning-width parameter. We refer to the two loci located in the insular region as anterior insula (aIC, shown in the left panel in a) and frontal operculum (shown in the second panel from left in a). These were the only brain areas with a positive amplitude parameter (first and fifth panels in b). (L: Left; R: Right; A: Anterior; P: Posterior).

Supplementary Figure 5 Anterior insula (aIC) is more selective than behavioral and autonomic response profiles.

(a) Posterior distribution of the tuning-width hyper-parameter (σaIC) fitted to generalization profiles recorded at aIC during the test phase. Posterior distributions were obtained using a hierarchical Bayesian model-fitting scheme. Red dotted lines indicate the average value of posterior distributions for σSCR and σRating parameters representing the tuning-width of SCR and shock expectancy ratings. Values correspond to 94th and 98.2th percentiles, respectively. These distributions are shown in (b) in cyan and brown colors, respectively. The average of these three posterior distributions was 0.46 for aIC, 0.60 for SCR and 0.67 for expectancy ratings. (c) Pair-wise differences of single-subject tuning-width parameter between aIC and SCR (left bar) and Ratings (right bar, Mean ± SEM). Positive values indicate smaller tuning-width parameters for aIC. Note that even though single subject parameters are assumed to originate from the same hyper-parameter distribution (shown in (a) and (b)), the pair-wise differences between different sources can be assumed to be independent (t-test (28) = 8.02 and 7.58; p <.001; SCR: skin conductance responses; aIC: anterior insular cortex).

Supplementary Figure 6 Correlation between sharpening index (SI) and neuronal/autonomic tuning parameters.

(a-c) Relationship between SI and the tuning strength (α) of anterior insular cortex (aIC, a), infero-temporal cortex (ITC, b) and ventromedial prefrontal cortex (vmPFC, c). Each point represents a single subject (n = 29). For vmPFC the relationship between SI and tuning-specificity is also shown (d). In the case of ITC, α parameter represents the tuning strength of the cosine model, whereas for all other areas it is the amplitude of the Gaussian model. (e,f) Horizontal lines inside vertical bars depict the correlation between SI and different tuning parameters (e: tuning-strength; f: tuning-specificity) for all areas that exhibited significant fear tuning (listed in Supplementary Table 1). The vertical extent of the bars represents bootstrap confidence intervals at alpha = 0.01 (significant correlations are shown in red). Correlation between fear-tuning parameters of SCR, as well as perceptual tuning (Percept. B, Percept. T: Perceptual tuning before baseline and following test phase, respectively) is also shown. Tuning-specificity arising from the integration of fear responses in aIC and ITC was significantly correlated with SI (last column in f).

Supplementary Figure 7 Evoked multivariate activity patterns within insular region of interest (ROI).

(a) Top panel: Insular ROI for the representational similarity analysis. Bottom panel: Same ROI as a binary glass brain. (b) Average-intensity projection of evoked activity-levels within the insular ROI in response to 8 different faces (columns) before (top three rows) and after conditioning (bottom three rows). 4th and 8th columns depict responses to CS+ and CS–, respectively. Note that the shape of the ROI matches the ones shown within the glass brain in (a). Color values represent beta-weights averaged across all subjects. Note the similarity of CS+ pattern with UCS pattern shown in Figure 4b (D: dorsal, L: left, R: right, A: anterior, P: posterior).

Supplementary Figure 8 Multivariate pattern similarity between activity evoked by different faces and UCS.

First panel: Same data as in Fig. 4d shown for comparison. Second, third and fourth panels depict similarity between activity patterns evoked by different faces and UCS for hippocampus, vmPFC and ITC, respectively. Error bars represent 95% bootstrap confidence intervals. Only in the aIC, a Gaussian model (black curve) performed significantly better than the null model (horizontal line, P < 0.005, likelihood-ratio test; vmPFC: ventromedial prefrontal cortex; ITC: inferotemporal cortex; aIC: anterior insula).

Supplementary Figure 9 Fear tuning in amygdala.

Temporal evolution of fear tuning in right amygdala during the test phase in [16.5 -7.5 -16.5] (xyz in mm). Fear tuning is shown for early (first panel), middle (second panel) and late temporal (third panel) windows. The x-axis represents responses to different faces. Responses to CS+ and CS– are shown at the 4th and 8th columns, respectively. Black curves and lines indicate best fitting models (Gaussian or null models; P <.01, likelihood-ratio test). The Gaussian model explained fear-tuning significantly better (P = 0.025, likelihood-ratio test) than the null model only in the early temporal window.

Supplementary Figure 10 Field of view during fMRI acquisition.

The field of view included the complete ventral part of the brain and was automatically co-registered across participants before the start of the scanning sessions to minimize loss of voxels due to lack of overlap.

Supplementary Figure 11 Hierarchical model used for parameter estimation.

Hierarchical dependencies between different parameters of the Bayesian model used for parameterization of fear generalization profiles. yi represents the ith observation recorded in response to ith face, which is xi radians distant from the CS+ face. The generalization profile of each single participant, p, within a given source, s (e.g. insula, or skin conductance responses), is modeled with αp,s and σp,s parameters of a Gaussian function, G. The error is assumed to be normally distributed (the “~” sign indicates stochastic relationships) with parameters μi and τs. τs is inverse of noise variance and specific for a given source, s. τs is half-Cauchy distributed (Gelman A., Bayesian Analysis 1, 515-534, 2006) with scale value of 5. μi is deterministically (the “=” sign indicates equalities) related to αp,s and σp,s parameters. In a given source s, αp is assumed to be distributed normally with parameters and , whereas the tuning-width parameter, σp, is assumed to originate from a log-normal distribution with parameters and .

Supplementary information

Supplementary Text and Figures

Supplementary Figures 1–11 and Supplementary Tables 1 and 2 (PDF 1675 kb)

Supplementary Methods Checklist (PDF 391 kb)

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Onat, S., Büchel, C. The neuronal basis of fear generalization in humans. Nat Neurosci 18, 1811–1818 (2015). https://doi.org/10.1038/nn.4166

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1038/nn.4166

This article is cited by

Search

Advanced search

Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing