Untangling cross-frequency coupling in neuroscience

https://doi.org/10.1016/j.conb.2014.08.002Get rights and content

Highlights

  • Fundamental caveats and confounds in the methodology of assessing CFC are discussed.

  • Significant CFC can be observed without any underlying physiological coupling.

  • Non-stationarity of a time series leads to spectral correlations interpreted as CFC.

  • We offer practical recommendations, which can relieve some of the current confounds.

  • Further theoretical and experimental work is needed to ground the CFC analysis.

Cross-frequency coupling (CFC) has been proposed to coordinate neural dynamics across spatial and temporal scales. Despite its potential relevance for understanding healthy and pathological brain function, the standard CFC analysis and physiological interpretation come with fundamental problems. For example, apparent CFC can appear because of spectral correlations due to common non-stationarities that may arise in the total absence of interactions between neural frequency components. To provide a road map towards an improved mechanistic understanding of CFC, we organize the available and potential novel statistical/modeling approaches according to their biophysical interpretability. While we do not provide solutions for all the problems described, we provide a list of practical recommendations to avoid common errors and to enhance the interpretability of CFC analysis.

Section snippets

Cross-frequency coupling (CFC): how much is that in real money?

One of the central questions in neuroscience is how neural activity is coordinated across different spatial and temporal scales. An elegant solution to this problem could be that the activity of local neural populations is modulated according to the global neuronal dynamics. As larger populations oscillate and synchronize at lower frequencies and smaller ensembles are active at higher frequencies [1], CFC would facilitate flexible coordination of neural activity simultaneously in time and

Caveats and confounds of the CFC analysis

In this section we concentrate on what we call the classical CFC analysis — it is illustrated and explained in Figure 1. Any result of this analysis can be used to classify different conditions but only as a marker that is devoid of concrete and clear physiological interpretation. To give a physiological interpretation to CFC, one needs to know the set of potential mechanisms responsible for neural coupling. This set of mechanisms is only beginning to emerge (discussed below). We now discuss

Organization of modeling/statistical approaches to CFC

Until now we have focused on what we call the classical approach (see Figure 1) to assess phase–amplitude CFC effects consisting of: first isolating frequency components, second assessing their dependencies, and third computing P-values based on surrogates. We have described how difficult it is at this stage to draw any conclusions about the biophysical mechanisms underlying these measures. However, different frameworks exist to assess relationships among rhythmic processes from experimental

Practical recommendations

As previously discussed we will need progress in several directions to establish phase–amplitude or other types of CFC as a fundamental mechanism in coordinating neuronal activity. Together with experimental and modeling advances, stricter standards in the use of CFC metrics are also necessary. Below we list practical recommendations to avoid some of the mentioned confounds and increase the specificity of the most popular phase–amplitude CFC metric (see Figure 1). Rather than a comprehensive

Conclusions

CFC might be a key mechanism for the coordination of neural dynamics. Several independent research groups have observed CFC and related it to information processing, most notably to learning and memory [4, 5, 6, 7, 8••]. Recently, CFC has also been used to investigate neurological and psychiatric disorders [9, 10, 11•, 12, 13, 14, 15]. Thus, CFC analysis is potentially a promising approach to unravel brain function and some of their pathologies.

In the present manuscript we have reviewed some

Conflict of interest statement

Nothing declared.

References and recommended reading

Papers of particular interest, published within the period of review, have been highlighted as:

  • • of special interest

  • •• of outstanding interest

Acknowledgements

This work was supported by the Max Planck Society. RV, VP and MW were supported by LOEWE grant ‘Neuronale Koordination Forschunsgschwerpunkt Frankfurt (NeFF)’. RV also acknowledges the support from the Hertie Foundation and the projects PUT438 financed by the Estonian Research Council and SF0180008s12 of the Estonian Ministry of Science and Education. VP received financial support from the German Ministry for Education and Research (BMBF) via the Bernstein Center for Computational Neuroscience

References (38)

  • K.J. Miller et al.

    Power-law scaling in the brain surface electric potential

    PLoS Comput Biol

    (2009)
  • G. Buzsaki

    Rhythms of the brain

    (2006)
  • A. Bragin et al.

    Gamma (40–100 Hz) oscillation in the hippocampus of the behaving rat

    J Neurosci

    (1995)
  • R.T. Canolty et al.

    High gamma power is phase-locked to theta oscillations in human neocortex

    Science

    (2006)
  • N. Axmacher et al.

    Cross-frequency coupling supports multi-item working memory in the human hippocampus

    Proc Natl Acad Sci U S A

    (2010)
  • J.E. Lisman et al.

    Storage of 7 ± 2 short-term memories in oscillatory subcycles

    Science

    (1995)
  • A.B. Tort et al.

    Theta–gamma coupling increases during the learning of item-context associations

    Proc Natl Acad Sci U S A

    (2009)
  • J. Lopez-Azcarate et al.

    Coupling between beta and high-frequency activity in the human subthalamic nucleus may be a pathophysiological mechanism in Parkinson's disease

    J Neurosci

    (2010)
  • S.A. Shimamoto et al.

    Subthalamic nucleus neurons are synchronized to primary motor cortex local field potentials in Parkinson's disease

    J Neurosci

    (2013)
  • Cited by (402)

    View all citing articles on Scopus
    View full text