In this article
- A supporter's exhaustion is neither weakness nor a rarity
- What drains you faster than the illness does
- Three roles you cannot hold at once
- A boundary is not a wall
- How to say a boundary
- The guilt that arrives right behind the boundary
- When it stops being your job
- What actually helps the person you are supporting
- Why exhaustion comes out sounding like criticism
- Where this advice fails
- What to remember
“I don’t want to be a bad sister, but when I see her name on my phone, my heart sinks.” She said it quietly, as though confessing to something shameful. She had been supporting her sister for fourteen months. Driving her to the doctor, texting her every evening, cancelling her holiday twice. And now she had reached the point where a call from her own person felt like a weight.
I see as many people in that position in my room as I see people who are ill. They arrive with a question about what to say, what to do, how to be better at it. They almost never arrive with a question about the limit of what they can carry. That question feels improper to them — as though they were complaining about their ankle in the middle of someone else’s illness.
This piece is for you, the one doing the supporting. Not about which sentences to use — I have written about that elsewhere. About how much this may cost you, and what to do when it costs too much.
If you are thinking about taking your own life: in Slovenia, the crisis line 01 520 99 00 (19.00–7.00), the Samaritan confidential line 116 123 (24 hours), and in immediate danger 112.
A supporter’s exhaustion is neither weakness nor a rarity
The first thing I tell people in this position is a piece of data. Van Wijngaarden and colleagues measured the consequences of caring for someone with depression and found the burden on relatives comparable to that of caring for someone with schizophrenia — which surprises most readers. Depression is the “less dramatic” illness, so people expect caring for it to be easier. It is not.
As long ago as 1980, working with relatives of older patients, Zarit described what burden actually is: not the amount of work done, but the ratio between what a person carries and what they get back. That is why supporting someone through depression is so demanding. The work is invisible, there is no end date, and there is almost no feedback — because of the illness itself, the person cannot tell you it helped.
What drains you faster than the illness does
When I ask supporters what grinds them down most, they almost never list what you would expect. Not the driving, not the cooking, not the broken nights.
The first is unpredictability. Depression does not proceed in a straight line upwards. Three good days and then one like before. If you hung your hopes on the good days, a bad day does not return you to the start — it returns you below the start.
The second is the missing milestone. A broken leg has six weeks. Depression has no number to lean on. The nervous system tolerates a surprising amount if it knows for how long, and burns through its reserves faster without that.
The third and quietest is the disappearance of mutuality. Relationships replenish us because they run both ways. When a relationship becomes one-way for several months, the part that used to cover the costs is gone. You are not tired because you are giving. You are tired because nothing has come back for a long time.
Three roles you cannot hold at once
Most exhausted supporters have slid into roles nobody assigned. It helps to name them.
Therapist. You read articles, ask about childhood, try to understand the mechanism. The problem is not the knowledge but that a therapeutic relationship works because it is bounded: an hour, a fee, an end. At home there is no such frame, so the stance swallows the whole relationship.
Rescuer. You take responsibility for the outcome. If today is a bad day, that is your failure. This stance is the most exhausting and, at the same time, the least true, because you cannot determine the outcome — you can influence the environment, not the course of the illness.
The close person. The only role nobody else can take. A doctor can be found, a therapist can be found, but a sister or a friend or a father cannot. When I ask a supporter which role to keep, the answer is always this one — and it is always the one they dropped first.
A boundary is not a wall
In these conversations the word boundary sounds like disconnection, as though it meant “stop calling me”. In fact a boundary is a precise description of what you can do, and that precision is what makes support sustainable.

Unclear support is endless support. If you never say when you are available, you are available always — and since nobody can manage that, at some point you will drop out without warning. Then the person you support experiences exactly what they feared: that people leave. A boundary stated in advance guards against a sudden withdrawal. I have written about how one is built in the piece on setting boundaries.
How to say a boundary
Concretely, because abstractions do not help here. Three parts: what I can do, what I cannot, and what instead.
“After nine in the evening I will not be on the phone. If it gets bad then, call the Samaritan line on 116 123, and I will ring you first thing in the morning.”
“On Saturday I am with you. On Sunday I go swimming, because I need that in order to manage the following week.”
“I am not going to argue about whether you should go to therapy any more. I will say it once a month, and otherwise I love you whether you go or not.”
Every one of those sentences is painful for the person supporting — it sounds selfish. It is not. What is selfish is promising constant availability that will not exist in three months, and then vanishing.
The guilt that arrives right behind the boundary
Everyone who sets a boundary for the first time gets a wave of guilt in the hours after. This is predictable and not a sign that you did wrong. Guilt here is not a signal about the act but a habit of the relationship: you used to be always available, now you are not, and the system flags it.
It is worth distinguishing guilt that tells you something from guilt that merely hurts. The first appears when you have hurt someone and can be put right with an apology. The second appears when you have looked after yourself and has nothing to put right. I have written about that difference in the piece on healthy and toxic guilt, because for supporters it is the most useful tool I know.
When it stops being your job
This is the most practical part and I say it to every supporter. There are three levels, and each has a different addressee.
At the first level there are bad days, withdrawal, listlessness, tears. Here you are the right person: presence, concrete help, no advice.
At the second level the state drags on for weeks, the person is not working, not eating, not sleeping, or has started drinking. Here you are no longer the one who can help — here the addressee is a doctor. Your task changes: you are not carrying the illness, you are helping someone to the door. Which door is the right one and when the psychiatrist is the first step, I have set out in the piece on when to see a psychiatrist first.
At the third level there is immediate danger: they talk about not wanting to live, they say goodbye, they give things away. Here you do not weigh things up and you do not promise silence. You call for help. If someone asks you not to tell anyone, that is a promise you must not give.
What actually helps the person you are supporting
Briefly, because it is not the centre of this piece. Three shifts I see most often in my room.
Instead of “tell me if you need anything”, offer something concrete: “I’ll come Wednesday for the bins and stay half an hour.” Depression narrows the capacity to decide, so an open offer is often useless.
Instead of encouragement, say what you have noticed: “I noticed you have got up in the morning three days running. That is not nothing.” Encouragement pushes; noticing confirms.
And: do not correct the interpretation. When someone says they are a burden to everyone, the reply “no you’re not” has almost no effect. What works better is “I know that is how it looks to you right now, but it is not how it is for me”. Why certain well-meant sentences achieve the opposite of what is intended, I have explained in the piece on the symptom as a message.
Why exhaustion comes out sounding like criticism
There is a line of research that is uncomfortable for supporters, but better known than not. Hooley spent decades studying what happens in families around relapses of mental illness and found that a reasonably good predictor of relapse is so-called expressed emotion in relatives: frequent criticism, hostility and over-involvement. This is not an accusation aimed at families. It is a description of what an exhausted person looks like after caring too long without relief.
I read that finding the other way round. Criticism in these families is not a sign that relatives love too little, but that they have given for too long. Tiredness comes out in tone rather than in words — in the sigh before the question, in the fast answer, in the way the door closes. A person in depression hears that tone far more accurately than the content.
What matters here: the solution is not to pretend better. Suppressed irritation sounds the same as expressed irritation, only without the possibility of a conversation. What works better is to say where you are — “I am snappy today and it is not about you, I am overtired” — and to organise the rest that actually lowers the irritation. Over-involvement decreases when you get your own life back, not when you decide to be more patient.
Where this advice fails
Now, honestly, about the limits of what I have written.
Not every situation can be bounded. If it is your partner who is ill and you have two small children, the advice “keep two regular things of your own each week” is not useful but insulting. In cases like that a boundary is not an inner decision but a question of resources: who can come, who can pay, who can take the children. If those resources are absent, the honest therapeutic move is not a conversation about self-care but help in finding concrete relief.
Self-compassion is not the answer to every exhaustion. Neff’s line of research has shown that self-compassion is associated with less burnout in carers, and that holds. What does not hold is the idea that an exercise could replace rest. If you have not slept a full night in six months, your problem is not your relationship with yourself.
And the most uncomfortable: sometimes exhaustion is no longer exhaustion. Supporters often slide into depression themselves without noticing, because they have an external explanation for their state. If it lasts for weeks, if insomnia comes with it, and if nothing gives you pleasure even when you are on your own, that is not tiredness from caring. Then you are the one who needs treatment — I have set out the signs of burnout and of crossing that line separately.
What to remember
Supporting someone through depression is serious work with a measurable cost, not a favour to perform without mentioning it. If it drains you, that is not a sign that you are not good enough to them.
A boundary said in advance and precisely is safer than an unlimited promise that breaks in six months. Three parts: what I can do, what I cannot, what instead.
Drop the therapist and rescuer roles. The close person is the only one nobody else can take, and the only cost worth paying.
Know the three levels. Bad days are yours. Weeks with no movement are the doctor’s. Talk of not wanting to live is a phone call — at once, without weighing it up and without promising silence.
Sources
- van Wijngaarden, B., Schene, A. H., & Koeter, M. W. J. (2004). Family caregiving in depression: impact on caregivers’ daily life, distress, and help seeking. Journal of Affective Disorders. PubMed
- Zarit, S. H., Reever, K. E., & Bach-Peterson, J. (1980). Relatives of the impaired elderly: correlates of feelings of burden. The Gerontologist. PubMed
- Coyne, J. C., Kessler, R. C., Tal, M., Turnbull, J., Wortman, C. B., & Greden, J. F. (1987). Living with a depressed person. Journal of Consulting and Clinical Psychology. PubMed
- Hooley, J. M. (2007). Expressed emotion and relapse of psychopathology. Annual Review of Clinical Psychology. PubMed
- Neff, K. D., & Germer, C. K. (2013). A pilot study and randomized controlled trial of the Mindful Self-Compassion program. Journal of Clinical Psychology. PubMed
- NIJZ. Mental health — programmes and helplines. NIJZ
- Samaritan confidential line, 116 123. telefon-samarijan.si
Who is supporting you
If you have been caring for someone for months and are slowly running out, write to me. The person supporting needs a space of their own, not one taken from the patient.
Write to me





