Compassion Fatigue: What Karuna Knew About Burning Out
Caring may not be what wears carers down. Brain studies and a fifth-century meditation manual point at the same thing: pain taken in with nowhere to go.
Compassion fatigue looks more like empathic distress. Trained compassion recruits a different brain network and seems to protect against it.
- In a 2014 study, empathy training raised negative feeling and pain-network activity. Compassion training then reversed it in the same people.
- The fifth-century Visuddhimagga says compassion fails when it produces sorrow, and names grief as its near enemy.
- A 2026 meta-analysis of four trials and 296 nurses linked mindfulness programmes to lower compassion fatigue, though the trials disagreed widely.
Compassion fatigue is the exhaustion that builds in people who spend their days close to other people's pain: nurses, carers, therapists, parents of sick children. The name blames caring. Brain research from Tania Singer's lab points somewhere else. The drain seems to come from feeling someone's distress as if it were your own, and compassion, trained on purpose, uses a different circuit that appears to protect against it.
A nurse finishes a twelve-hour shift and sits in the car park for a while before starting the engine. She isn't crying, and she isn't thinking about anything in particular. She has nothing left for the people waiting at home. The part that stings is the conclusion she draws from it: maybe she cares too much, and caring is the thing she now has to ration.
Most of what she'll read online agrees with her. The research mostly doesn't.
What is compassion fatigue?
The term reached print in 1992, when Carla Joinson used it in the journal Nursing, writing about nurses worn down by the work. The trauma psychologist Charles Figley took it up a few years later and gave it a clinical shape. The signs people describe have stayed consistent since: dread before a shift, numbness during it, a short temper at home, sleep that doesn't restore, and a slow flattening of the feeling that brought someone into the work in the first place.
None of that is in dispute. The word worth arguing with is "compassion".
Name it, then turn
- Sit where you are, in the car or on a bench, before the next part of the day starts. Leave the engine off.
- Bring one person from today to mind. Notice where their distress landed in you: throat, chest, stomach.
- Say, under your breath: this is their pain, and I felt it with them.
- Instead of feeling it again, send a plain wish toward them. May this ease for you.
- Give yourself the same wish, then start the car.
Compassion fatigue vs empathic distress
In 2012 Olga Klimecki and Tania Singer published a book chapter with a pointed title: "Empathic distress fatigue rather than compassion fatigue?" Their argument rests on a distinction that sounds academic until you have felt both sides of it.
Empathy, in their usage, is resonance. You see someone hurting and a version of the hurt happens in you. Compassion is a response to that hurt: warmth toward the person and a wish for their suffering to ease. Empathy puts you inside the other person's pain. Compassion keeps you beside them, facing the same way.
On paper the two sound close. In a scanner they look different. Watching others suffer activates the anterior insula and the anterior midcingulate cortex, a network that also responds to your own pain, and it comes with a drop in mood. When Klimecki's group trained people in compassion, a separate network came online, including the medial orbitofrontal cortex and parts of the brain's reward system, regions tied to affiliation and positive feeling. Participants reported warmer feelings while watching the same distressing videos.
So the emptiness in the car park may be resonance with no exit. Pain taken in, shift after shift, with nothing to turn it into.
Nurses across the four trials pooled in a 2026 meta-analysis of mindfulness and compassion fatigue. The effect was large and the trials disagreed a lot, so read it as a direction.
What compassion training changes in the brain
The cleanest test came in a 2014 study from the Max Planck Institute in Leipzig. One group was first trained in empathic resonance, which meant practising feeling with people in distress. It did what it promised. Negative feeling went up, and so did activity in the insula and midcingulate cortex. A control group trained in memory skills showed no such change.
Then the same people were trained in compassion. The rise in negative feeling reversed. Positive feeling increased, and the brain recruited a reward and affiliation network that did not overlap with the empathy-for-pain one. The authors suggested compassion might work as a coping strategy for empathic distress.
A study led by Helen Weng in Richard Davidson's group at Wisconsin looked at behaviour instead of mood. After a short compassion training, participants gave more of their own money to a stranger they had watched being treated unfairly, in a game that had nothing to do with the training. The people who gave most showed changes in regions linked to emotion regulation and to understanding other people. The training made them more willing to act, which is the opposite of what the fatigue story predicts.
It succeeds when it makes cruelty subside and it fails when it produces sorrow.Buddhaghosa · Visuddhimagga IX.94 · trans. Ñāṇamoli
What the Visuddhimagga says about how compassion fails
Around fifteen hundred years before the scanners, the scholar Buddhaghosa compiled the Visuddhimagga, a meditation manual written in Sri Lanka. Its ninth chapter treats compassion, karuṇā, as a skill with a specification sheet. It lists what compassion is, what it does, and what gives rise to it. Then it adds something few modern definitions bother with: how compassion succeeds and how it breaks.
It succeeds, the manual says, when it makes cruelty subside. It fails when it produces sorrow.
The same chapter names compassion's near enemy, the state that looks like compassion from outside and replaces it without anyone noticing. That enemy is domanassa, grief of the ordinary household kind, the sorrow of seeing loss and taking it on. The manual says the two are easily confused because both see failure. The far enemy, cruelty, is obvious. The near enemy passes for virtue.
Read as a practice record from people who spent their lives watching their own minds, this is a working description of empathic distress. Sorrow that feels like caring is flagged as the failure. Warmth that wants suffering to ease is the thing to train. Klimecki and Singer reached the same line with fMRI.
It helps to be exact about what the manual left out. It never tells anyone to care less. The chapter is about extending care further, eventually to everyone. Its instruction is about the shape of the caring, so that the person doing it can stay in the room.
Compassion fatigue among nurses: a 2026 meta-analysis
The newest evidence came out this month. A meta-analysis in International Nursing Review pooled randomised trials of mindfulness programmes for nurses. Across four trials with 296 nurses, the programmes were linked to lower compassion fatigue scores. Across six trials with 518 nurses, burnout scores fell too. On paper the effect on compassion fatigue was large.
The trials also disagreed with each other a great deal. The authors say the evidence is too thin to treat mindfulness as a stand-alone fix for the workforce, and that it belongs alongside staffing, workload and access to psychological support. That caution deserves repeating. A breathing practice will not repair a ward that is two nurses short, and nobody should be told to meditate their way through an unsafe rota.
What the trials suggest is smaller and still worth having: a trained way of meeting pain can make the same amount of it cost less.
Compassion fatigue vs burnout
The two overlap, and the difference matters because the remedies differ. Burnout is usually a response to the job itself: too much work, too little control, effort that goes unrewarded. You can burn out in a role that involves no suffering at all. Compassion fatigue comes from repeated exposure to other people's distress. Secondary traumatic stress sits closer still, with intrusive images and a jumpy nervous system picked up from other people's trauma.
A carer often has some of all three. One way to sort them is to ask what drains you most. If it is the rota and the paperwork, the fix is structural. If it is the faces, the empathy side needs attention, and that is where compassion training has evidence behind it.
How to recover from compassion fatigue, one shift at a time
The practice below borrows from the compassion methods used in the training studies, cut down to fit into the few minutes before you drive home. It is not therapy. If you have slept badly for weeks, dread going in, or carry images you can't put down, talk to someone trained to help.
The change it asks for is small. When someone else's pain lands in you, notice that it has landed. Name it as theirs. Then turn toward them with a wish, instead of a second copy of their feeling.
Some people cry the first few times. That doesn't mean the practice has gone wrong, and tears in meditation often come when feelings held all day finally get some room. The old image of the second arrow fits here too. The pain you witnessed is the first arrow. The voice saying you should have done more is often the second.
Two misreadings are worth heading off. If compassion practice becomes a way of keeping patients at arm's length, it has turned into spiritual bypassing, calm used to avoid contact. And it isn't a call to endure anything at all. The classical idea of titiksha is about bearing what has to be borne, and a rota that breaks people doesn't qualify.
If your hardest cases are at home, the same method works with family, and we wrote about that in loving-kindness for difficult people.
The nurse in the car park is right that something needs rationing. What needs rationing is the absorbing. The caring can stay, and it holds up better when it is practised every day, because the shifts keep coming.
“Be patient toward all that is unsolved in your heart and to try to love the questions themselves.”Rainer Maria Rilke · Letters to a Young Poet
Questions
What is the difference between compassion fatigue and burnout?
Burnout is usually a response to the job itself: workload, lack of control, effort that goes unrewarded. It can happen in work that involves no suffering at all. Compassion fatigue comes from repeated exposure to other people's distress. Many carers have both, and asking whether the rota or the faces drain you more is a useful way to tell them apart.
Is compassion fatigue real?
The exhaustion is real and well documented, especially in nursing, since Carla Joinson named it in 1992. What researchers question is the name. Klimecki and Singer argue that the drain comes from empathic distress, feeling another person's pain as your own, and that compassion itself engages a different, more sustaining brain network.
What are the signs of compassion fatigue?
Common signs include dread before work, numbness while you are there, irritability at home, sleep that doesn't refresh you, and a fading of the feeling that drew you to the work. Intrusive images and a constant sense of alarm point more toward secondary traumatic stress, which is worth raising with a professional.
How to recover from compassion fatigue?
Start with the structural things you can change: rest, workload, supervision and support. Alongside them, compassion practice has lab evidence behind it. The short version is to notice when someone's pain has landed in you, name it as theirs, and turn toward them with a wish for their relief instead of taking the feeling on again.
Is compassion fatigue the same as secondary traumatic stress?
They overlap without being the same. Secondary traumatic stress looks like trauma symptoms picked up from someone else's trauma: intrusive images, avoidance, a jumpy nervous system. Compassion fatigue is the broader wearing-down that comes from sustained contact with suffering. Figley's framework treats secondary traumatic stress as one part of it.
Compassion has grief based on the home life as its near enemy, since both share in seeing failure. It succeeds when it makes cruelty subside and it fails when it produces sorrow.
Empathy training increased negative affect and brain activations in anterior insula and anterior midcingulate cortex. Subsequent compassion training could reverse the increase in negative affect and augment self-reports of positive affect.
Compassion training increased positive affective experiences, even in response to witnessing others in distress, and elicited activity in a network including the medial orbitofrontal cortex, putamen, pallidum and ventral tegmental area.
Compassion training increased altruistic redistribution of funds to a victim encountered outside of the training context.
For compassion fatigue, four studies involving 296 nurses contributed to the pooled analysis (SMD = -1.11, 95% CI -1.94 to -0.28; I2 = 91%). Current evidence remains insufficient to support routine implementation as an independent workforce intervention.
One of these in your inbox, each morning
A single contemplative reading at first light. No threads, no streaks, no notifications you'll resent. Unsubscribe in one tap, any morning.
One contemplative reading at first light. No threads, no noise, unsubscribe in a breath. By subscribing you accept our Privacy Policy.