Helping depressed adolescents acquire mindfulness via personalized real-time neurofeedback

Novel circuit-based intervention offers new opportunity for self-powered improvement
Helping depressed adolescents acquire mindfulness via personalized real-time neurofeedback

There is no debate that teenagers today face a mental health epidemic. Among serious mental illnesses, major depressive disorder (MDD) in particular affects 17% of teenagers and accounts for over half the mental health-related pediatric hospitalizations in the United States. Unfortunately, this already dire situation was further exacerbated recently by the COVID pandemic. Given that gold-standard depression treatments only work for a dismal 50% of adolescents, our team went on a scientific journey to search for novel treatments that target core mechanisms fundamental to MDD.

And we went to the brain. 

Depression, at the neural level, is consistently characterized by hyperactivation and hyperconnectivity among a set of regions belonging to the 'default mode network' (DMN), including key nodes in the medial prefrontal cortex (MPFC) and subgenual anterior cingulate cortex (sgACC). The DMN is thought to facilitate patterns of depressogenic, self-referential processing and a heightened focus on distressing emotional states. The sgACC is specifically considered a 'hub' in depression, proving as an effective direct or associated target in stimulation-based treatment options such as deep brain stimulation and transcranial magnetic stimulation

Instead of relying on externally delivered neurostimulation that may require invasive surgery and can only be performed at a clinic, we wondered, could we train individuals to volitionally modulate their own DMN? We drew inspiration from a well-established meditation strategy called 'mindfulness', which helps focus attention to the present and quiets the DMN. One reason that mindfulness may be hard to acquire is that people cannot continuously assess if they're practicing mindfulness effectively. Consider the types of physiological information we can gather to assess the status of our body when we exercise: heart rate, Galvanic skin response, respiration, and VO2max. These numbers are helpful because they allow us to understand the effect of physical exercise on our body and therefore facilitate more effective subsequent exercises. Wouldn't it be great if we had similar feedback that gives us updates on the status of our brain when we practice mindfulness? 

To tackle this exact challenge, our team developed a unique mindfulness-based, real-time neurofeedback (mbNF) paradigm (Figure 1) whereby people observe a visual display of their brain activity and practice mindfulness to volitionally reduce DMN activation relative to central executive network (CEN; important for attentional control) activation for 15 minutes.

Figure 1. During mbNF, participants were instructed to practice mindfulness to move the white dot on the screen up into the red circle. The movement of the white dot was dependent on a real-time analysis of the fMRI data that computed the difference in personalized DMN and CEN activations. When DMN activation is lower than CEN activation, the white dot moves up; when DMN activation is higher than CEN activation, the white dot moves down.

When a participant is successfully down-regulating their DMN and up-regulating their CEN, a white ball will move up, and vice versa (see video illustration below). Prior to this study, we have successfully applied this intervention in patients with schizophrenia which helped reduce their DMN connectivity as well as improve hallucination symptoms. 

The mindfulness-based neurofeedback (mbNF) intervention allows a participant to monitor and modulate their brain activity in real-time.

As a proof-of-concept, we tested this paradigm in nine adolescents previously diagnosed with depression and/or anxiety (at the time of the experiment, two had comorbid depression and anxiety, and one had anxiety) and found consistent and promising preliminary results (Figure 2). Reduced DMN connectivity (between the sgACC seed and the MPFC) was observed individually in all nine participants. 

Figure 2. (A) We used a 8 mm spherical seed in the sgACC. (B) There was reduced connectivity between sgACC seed and midline DMN regions after mbNF. (C) Reduced sgACC-MPFC connectivity was found in all participants. Each bar represents the change in functional connectivity strength in a participant. 

We also found that participants with greater reduction in DMN connectivity showed more increase in state mindfulness (e.g., 'I noticed emotions come and go', 'I noticed physical sensations come and go') after the experiment.

The numbers were encouraging and perhaps more encouraging were the words. Participants in this study and other mbNF studies we are currently running have frequently commented to us that they felt empowered by the feeling of control and felt confident they could translate the mindfulness exercise to their daily life.

Our team has recently been funded with a NIMH R61/R33 grant to compare mbNF head-to-head with mindfulness. We will also examine long-term benefits of our mbNF intervention, as other neurofeedback studies showed sustained or even improved symptoms post-neurofeedback. 

Given that DMN hyperactivation/hyperconnectivity has also been observed in youth at-risk for depression and schizophrenia, a major next step is to test the benefit of mbNF in preventing disease onset.

Across these studies, we aim to further examine baseline characteristics that will tell us who will benefit more from mbNF and when is the optimal time for mbNF intervention. Crucially, we are also adapting this approach using more affordable and wearable devices such as electroencephalogram (EEG) and functional near-infrared spectroscopy (fNIRS).

We ardently hope the future is not too far away when we can catch children before they fall and empower them with mind-calming skills that they can use for life.

If you're interested in participating in our ongoing studies, please contact us at

Please sign in or register for FREE

If you are a registered user on Research Communities by Springer Nature, please sign in