Probing consciousness in heart-brain interactions in severely brain-injured patients

Aims
It is suggested that the brain monitors and regulates visceral organs, such as the heart (1). However the relationship between this regulation and consciousness levels has not been fully explored. We tested whether electrocardiography (ECG) can be used as an index to characterize cognitive function in (un)consciousness, as measured by EEG during a validated paradigm that tests two levels (local/short or global/long) of auditory regularities (2, 3).
Methods
The local-global paradigm (2) was applied on 91 conscious subjects (MCS patients and controls) and 70 patients in vegetative state/unresponsive wakefulness syndrome (VS/UWS). Independent component analysis extracted the embedded ECG signal from the recorded EEG. Across all trials, the RR intervals, i.e. the intervals between the peaks of the QRS ECG complex, were correlated with the Coma Recovery Scale-Revised scores (CRS-R, 4). Between-group differences in the RR intervals which were locked to the last sound (either for the local or global breaking of auditory regularities) of the oddball paradigm were checked with Wilcoxon rank sum tests.
Results
A positive correlation between heart rate and CRS-R scores was found only in the VS/UWS group (r=0.31, p=0.008). Shortened RR intervals were identified only in conscious subjects (z=-2.94, p=0.003) when the trial corresponded to a global deviant. This effect was absent when the auditory stimulus corresponded to a local deviant.
Conclusions
Correlation between CRS-R and HR suggests that the diversity of behaviors characterizing conscious states (associated with cortical processing) does not necessarily translate into strong correlations with vegetative markers, such as heart rate. This effect might be associated with an overall deterioration of clinical condition. Violations of global regularities, which cause a long-lasting brainscale P3 response in conscious individuals, induces a top-down effect on this vegetative marker. Violations of local regularities, which cause a transient and local response confined to auditory cortical areas, did not produce this effect as predicted by our model (5). Our results suggest an interaction between consciousness level and the neural information related to cardiac monitoring, consistent with previous theoretical frameworks (1).
Ref
1. Park HD & Tallon-Baudry, C Philos Trans R Soc Lond B Biol Sci 369, 2014
2. Bekinschtein TA et al. Proc Natl Acad Sci USA 106, 1672–7, 2009
3. Sitt JD et al. Brain 137, 2258–70, 2014
4. Giacino JT et al. Arch Phys Med Rehab 85, 2020–9, 2004
5. Dehaene S & Naccache L, Cogn 79, 1–37 2001