Should you stay with someone who has depression?

The association between depression and relationship distress is one of the better established findings in this area, and the important detail is that it is bidirectional rather than one-way. Depression makes relationships harder through withdrawal, irritability, loss of interest in shared activity and reduced sexual desire, all of which are symptoms rather than choices. But relationship distress also predicts the onset of depression. A community study following adults over twelve months found that those who were dissatisfied in their marriages were substantially more likely to experience a major depressive episode during that period than those who were not[1]. Reviews of the wider literature describe the same two-way pattern across a range of psychological conditions[2]. The practical implication is that asking which came first is often unanswerable and usually unhelpful. A couple can be in a loop where each is worsening the other, and that loop is a more accurate description of most situations than a story in which one person is ill and the other is coping.

Research on the partners and families of people with depression finds substantial burden, including distress, restriction of the caregiver's own activities, and effects on their own mental health[3]. This is worth stating plainly because the person in this position is frequently told, directly or by implication, that their difficulty is not the relevant one. It is relevant, both because they are a person and because a supporter who is depleted is less able to support anyone. Studies in this area also find that partners often receive very little information or support from services, and that their contribution is largely invisible in the treatment of the illness. None of this establishes that anyone should leave. It establishes that the exhaustion is a documented phenomenon with a literature behind it rather than a personal failing, which is usually not how it feels from inside.

The most useful fact here is that depression is a treatable condition with an expected course, not a fixed characteristic of a person. Most episodes remit, and treatment substantially improves the odds and shortens the time. More specifically relevant to this question, a meta-analysis of couple therapy as a treatment for depression found it comparable to individual therapy in reducing depressive symptoms, while also improving relationship satisfaction, which individual therapy does not reliably do[4]. That is a meaningful finding for someone in this situation. Where depression and relationship distress are entangled, treating them jointly is an evidence-based option rather than a compromise, and it addresses the loop described above rather than one side of it. This also reframes what you are actually deciding. The choice is often not between staying with an ill person and leaving, but between a situation where treatment is happening and one where it is not.

Two situations get discussed as though they were one. In the first, someone is unwell, knows it, is engaged with treatment however imperfectly, and the relationship is under strain during an episode with a plausible end. In the second, the illness has been present for years, treatment is refused or repeatedly abandoned, and one person has become the entire support structure, managing appointments, income, household and mood, while being told that raising any of it is unkind. The evidence about remission, about the effectiveness of treatment and about the value of joint work applies to the first with reasonable force. It applies to the second much more weakly, because it assumes a treatment process that is not occurring. Distinguishing which one you are in is more informative than assessing how severe the depression is, and it is a question about behaviour and direction rather than about diagnosis, which means you do not need clinical training to answer it.

Two failure modes sit either side of this question. One is the framing in which leaving is simply abandonment, which is used to make the question unaskable and leaves people carrying situations indefinitely without ever examining them. The other is the framing in which anything difficult is described as toxic and leaving is the healthy response, which treats an illness as a character flaw and asks nothing of the person doing the leaving. Neither is supported by anything. It is also worth being careful about the direction of blame in the bidirectional finding: saying that relationship distress predicts depression is not a way of telling someone their partner made them ill, nor a way of telling the depressed partner that they are the cause of the relationship's problems. It describes a loop that neither person chose. And the practical picture varies enormously with what else is true, since the same illness is a different proposition where there are children, where one partner is financially dependent, or where access to treatment is limited or expensive.