Does Divorce Damage Children's Mental Health?
It is one of the most common questions we hear in couples and family work. A parent will say some version of it near the end of a session: if we separate, what does that do to my children?
Sometimes the question arrives years later. A parent whose marriage ended a decade ago watches their teenager struggle and quietly wonders whether they are looking at the consequences of a decision they made long ago.
Research published late last month in Psychiatry Research offers something more useful than reassurance. It suggests that the question most worth asking is NOT whether a marriage ended, but what else the child was living alongside at the time — and whether any of it was ever addressed.
The study
Mary Kate Schilke, Philip Baiden and Esme Fuller-Thomson analysed responses from 128,264 adults who took part in a large American health survey between 2021 and 2024. Participants reported on their childhood family structure and on whether they had ever been diagnosed with depression.
The first finding looks like the one most people expect. Among adults whose parents divorced, 19.2% reported depression at some point in their lives. Among those whose parents remained continuously married, the figure was 14.1%. Adults whose parents never married sat at 18.7%.
A five-point gap is not nothing. Had the study stopped there, it would have confirmed the familiar story.
Looking at the findings in detail
They then asked what else had been present in those households.
When they accounted for whether a parent had a mental illness, the association between divorce and adult depression shrank by 68%. When they accounted for parental substance use, it shrank by 49%. And in the full analysis — adjusting for the whole range of difficult childhood experiences, along with social and economic circumstances and adult health behaviours — the link between childhood family structure and adult depression was no longer statistically significant.
In plain terms: most of what first looked like the footprint of divorce turned out to be better explained by other things happening in the same home.
It does not mean separation is painless, and it does not mean any individual child is unaffected. Plenty of children find a separation genuinely hard, and that difficulty is real whether or not it shows up in a survey two decades later.
What the study does suggest is that the ending of a marriage, on its own, may carry less of the long-term weight than we have tended to assume.
Why this is more hopeful than it sounds
Here is the part we find most useful clinically.
A marriage that has ended cannot be “unended” (if that was a word). If divorce itself were the mechanism, there would be nothing to do but brace.
But the factors that carried most of the association in this study are not like that. A parent's depression can be treated. A parent struggling with alcohol can get support. A household where distress goes unnamed can start naming it. These are circumstances that respond to attention.
That reframes the question a separating parent is actually facing. Not will this damage my child — but what is my child living alongside right now, and is any of it being looked after? That is a question with something to do at the end of it.
It also quietly makes the case for something parents often resist. A parent who gets their own support is not taking time away from their children. In a study where parental mental health accounted for most of the association, a parent looking after their own wellbeing starts to look less like self-indulgence and more like an intervention on the child's behalf.
What this study cannot tell us
We think it is worth being straightforward about the limits.
The data was cross-sectional — a single snapshot — so the researchers could not establish what came before what. Participants were adults recalling their own childhoods, and memory is imperfect, particularly for a difficult period. The category of "unmarried parents" grouped together households that may have had very little in common. And the study measured participants' adult circumstances, not the family's financial situation at the time of the separation, which may matter a great deal.
The sample was also American. Household structures, extended family involvement, and the practicalities of separation all look different here, and we would not assume the figures transfer to Asian or Middle Eastern contexts unchanged.
Relevance for families here
Some of what we see in practice sits alongside these findings.
Children tend to cope better when the conflict stays between the adults. Being asked to carry messages, take sides, or manage a parent's feelings about the other parent is its own kind of strain, and it can continue long after the legal process ends.
Children also tend to do better with straight answers at a level they can hold. Being told very little often leaves a child filling the gap with something worse, and frequently with a version in which they are somehow responsible.
For families in Singapore's international community there is an additional layer. A separation can mean one parent relocating, a visa or school place becoming uncertain, or a child losing a familiar country along with a familiar household. Those are real losses, and they are worth naming rather than folding quietly into the general upheaval.
Don’t wait for a crisis
You do not need a crisis to have a conversation.
Some parents come to us while a separation is still a question, to think it through with someone outside the family. Some come during it, wanting help with the practicalities of two households. Some come afterwards, because a child has become withdrawn or angry and they are unsure how much is ordinary.
We cannot tell you what the right decision is for your family, and we would be wary of anyone who claimed otherwise. What we can offer is a considered place to think, and a sense of what tends to help children through a transition like this one.
If that would be useful, get in touch and we can talk about where to start (click here). Contact us here.
This article discusses published research for general information. It is not a substitute for individual clinical advice, and no assessment or course of therapy can guarantee a particular outcome. The study described is reported here.
