I am currently teaching a course on Human Brain Connectivity and I have asked students to ask me questions to go deeper into what we discussed during the contact sessions. Here one of the questions that came out.
Can you really get resting state in the brain, isn’t everyone thinking about something and that way doing something with their brain all the time?
As you feel it yourself when you lay down and stop moving, your mind is still wandering and doing all sorts of mental tasks, so what is it that resting state really measures? There is indeed a whole field of neuroscience that explores this along with mind-wandering, rumination, and what it means to measure the brain at rest.
In the jargon of the field, people talk about intrinsic activity (that is not driven by a stimulus) versus extrinsic (driven and/or time-locked with external stimuli). There is one network called the default mode network (DMN, more on this in contact session #4) that is found to be active during the lack of task. For this reason it has been called “task negative” in the early days of resting state research, but today we know that it is actually involved in many tasks.
So what is the brain doing during rest or lack of stimuli? Well clearly mental tasks with no external stimuli such as 1) remembering the events of your day 2) subtracting numbers, or 3) (silently) singing lyrics are producing different connectivity patterns that can be decoded with 84% accuracy (https://www.ncbi.nlm.nih.gov/pubmed/21616982).
However when asking the subjects to rest in the scanner they are usually not given specific instructions like “subtract numbers”. Some people like Jonathan Smalwood took things a bit further and started exploring the possibility that the activity at rest is involved with active rumination and mind-wadering (here a recent review https://www.ncbi.nlm.nih.gov/pubmed/25293689). Things however are not that simple, another recent meta-analysis has found that it’s not only in the DMN but other networks seems to be involved during mind wandering, also networks involved in active task taking (https://www.sciencedirect.com/science/article/pii/S1053811915001408).
A recent paper in PNAS http://www.pnas.org/content/113/48/13899 seems to actually propose that the active mind wandering is not DMN specific, leaving DMN to be more related to stable factors (put it simply, DMN seems to be more related to who you are rather than what you are specifically thinking about). [Check also the reply to that PNAS paper trying to reconcile past and current literature, http://www.pnas.org/content/early/2017/07/12/1705108114.short]
When it comes to clinical conditions, pain rumination at rest has been associated with stronger connectivity within DMN regions http://www.jneurosci.org/content/34/11/3969.short Here another paper comparing active rumination and “free” rest finds differences between healthy controls and major depressive disorder patients, with a note regarding the mood of the subjects having a stronger effect than the rumination itself. (https://www.sciencedirect.com/science/article/pii/S1053811914007691)
Finally, one might argue that the true resting brain is the sleeping brain. Tagliazucchi and others first learned the EEG pattern that sets in when people drift into sleep and then collected simultaneous fMRI + EEG to find when people were falling asleep during fMRI resting state paradigm. When applying this classifier to open datasets they also found that 30% of public resting state data is about people falling asleep in the scanner! https://www.sciencedirect.com/science/article/pii/S0896627314002505
So… is resting state the way to go? I think the answer with everything is always yes and no. Resting state is good, if combined with other basic tasks, as well as mood sampling of the participant and active control of the participants’ mental state (e.g. by monitoring eye movements during rest to see if person falls asleep and by asking the subject about her mental state after scanning). On the other hand, it is also very helpful that even somebody who cannot perform any task (a sedated patient or somebody in a coma [although it is more intriguing – and somewhat shocking – to put people in a coma in a scanner and show them a movie http://www.pnas.org/content/111/39/14277]) can also undergo resting state fMRI: the resulting network is anyway mostly related to the person itself rather than fluctuations so it can still work out as a “fingerprint” (https://www.nature.com/articles/nn.4135) for future diagnostic purposes.
Please contribute to the answer with comments below.