1. The programme at a glance
Indication. A child or adolescent aged 4 to 17 whose parent, or the adult who is bringing them up, has a mental illness: depression, an anxiety disorder, bipolar disorder, psychosis, borderline personality disorder, an alcohol or drug use disorder, or a disorder that began around the time of a birth. The family comes of its own accord or is referred by the team treating the parent. The programme is first and foremost preventive: the child does not need to be doing badly to benefit from it, and if they have a disorder themselves, that disorder is treated with its own treatment (section 6).
Reference model. The evaluated preventive family interventions for these families. Family Talk, by William Beardslee and his colleagues, in which the clinician meets the parents and then each child, prepares with the parents what they will say, and leads a family meeting in which the parents talk about the illness (Beardslee et al., 2003; Solantaus et al., 2010). "Let's Talk about Children", by Tytti Solantaus, a brief, structured discussion with the parent about the child's everyday life and about what, around the child, can support them (Allchin and Solantaus, 2022). And the family cognitive behavioural intervention of Bruce Compas and his colleagues, a group programme for families in which a parent is depressed (Compas et al., 2009). This manual adapts their published principles to work with an individual family, without reproducing their materials.
Format. Three assessment sessions (Év1 to Év3): the ill parent, the other parent or the adult who brings up the child with them, then each child. Three sessions with the parents (P1 to P3), two with the child (E1 and E2), the family meeting (F1), the plan for difficult periods (F2), the school and the network (R), and the review (B). That makes twelve sessions for an only child, over three to four months, plus one more Év3 session for each additional child, as well as separate E1 and E2 sessions when the children's ages are far apart (section 31). A three-session module (A1 to A3) is added for the adolescent who needs it. Sessions last 60 minutes with the parents, except the first (Év1), which lasts 75 minutes; 45 to 50 minutes with a child or adolescent, and about 30 minutes with a child aged 4 to 6, whose parent may stay in the room; and 75 to 90 minutes for the family meeting, reduced to about 30 minutes when the only child is aged 4 to 6. Three booster sessions follow, at three, six and twelve months. A brief format of three meetings with the parent is suitable when the child is doing well and the parent is receiving treatment (section 28).
Mechanism targeted. The parent's illness cannot be made to disappear, but what the child experiences of it can be changed: that they understand what is happening to their parent and know that it is not their fault; that they keep their friends, their activities and adults to talk to; that everyday life holds together during difficult periods without their having to carry adult responsibilities; and that their parent receives treatment.
| Session |
Focus |
What should come out of it |
| Év1 |
The ill parent |
History gathered, safety assessed, written agreements |
| Év2 |
The other parent, or the adult who brings up the child |
Safety assessed, agreement to see the child |
| Év3 |
The child |
What they know, fear and carry, in their own words |
| P1 |
What protects a child |
The circle drawn, the protective factors counted |
| P2 |
Everyday life during the illness |
The minimum plan written, one task taken off the child, the words for E1 approved |
| E1 |
Understanding what is happening to my parent |
The child's questions written down |
| E2 |
What helps me |
Their sources of support named, an adult to talk to |
| P3 |
Preparing the family meeting |
What the parents will say, written for each child |
| F1 |
The family meeting |
The illness named by the parents, the questions asked |
| F2 |
The plan for difficult periods |
A written family plan, shared |
| R |
The school and the network |
One trusted adult per child |
| B |
Review and next steps |
Measures repeated, plan for what comes next |
| A1 to A3 |
The adolescent module |
A notebook kept by the adolescent |
What distinguishes this programme from the other manuals on this site. The child is not the identified patient: they come because their parent is ill, and the target is what is happening around them. The parents remain the ones who talk to the child: you help them do so without speaking in their place. And the parent's treatment is part of prevention: you work with the team treating them, with their agreement. The family takes away nine worksheets (section 35), and the guide When a parent has a mental illness: supporting the child and the teenager sets out the same ideas for them.
2. Before you begin
The five preliminary decisions
Is the child safe, and is the parent? A mental illness is not in itself a danger to the child; a severe deterioration, suicidal ideas in the parent or the child, delusional ideas concerning the child, domestic violence, substance use or serious neglect are. This is section 7, the only one that can suspend the programme.
Does the ill parent agree? Without their agreement, you do not talk to the child about their illness and you do not contact their team, except in case of danger (section 11).
Who can authorise the work with the child? The rules vary from country to country; as a precaution, the work with the child takes place with the agreement of the parents who hold parental authority, and the child's agreement is always sought.
Does the child need this programme, or treatment? An established disorder is treated with its own treatment, and the family work is carried out alongside it.
Is the parent stable enough? During a manic episode, acute psychosis or a suicidal crisis, the first concern is safety and the organisation of the child's life. Current depression is not a reason to wait if the parent can take part: in the Compas programme, the effect hardly depended on whether a parent had a depressive episode at the start (Compas et al., 2015). By contrast, in Garber's group programme for adolescents, the effect disappeared when the parent was depressed at the start (Garber et al., 2009). The authors therefore suggest treating the parent's depression in parallel (Brent et al., 2015), and this programme asks for it too.
What this programme does not treat
The parent's disorder is treated by their team: you do not become their therapist (section 11). An established disorder in the child falls within Treating depression in children and adolescents: a therapist's manual, Treating anxiety in children: a therapist's manual, Post-traumatic stress disorder in children and adolescents: a therapist's manual or Suicidal behaviour and self-harm in adolescents: a therapist's manual. A situation of danger falls first within child protection (section 7). A separation conflict in the foreground falls within Parental separation and divorce: supporting the child and the parents — a therapist's manual. Bereavement after a parent's death falls within Treating grief in children and adolescents: a therapist's manual.
How to use it
Read the whole manual before the first session, in particular sections 5, 7 and 11, and reread session F1 before session P3. The sessions follow a common framework: the aim, the steps with an approximate duration, what you say and, when a sentence is to be avoided, what you do not say, the task between sessions, common errors and the criterion for moving on, these last two headings being combined in the adolescent module.
Three warnings. Illness is not danger, and silence is not protection: seeing a danger in every ill parent drives families away, failing to see a real danger exposes the child, and silence leaves the child alone with hypotheses that are often worse than reality. Parents arrive afraid of being judged: SCIE, the British reference body for social work, observes that parents hesitate to ask for help for fear of losing custody of their children, and children hesitate to talk for fear of being separated from their family (SCIE, 2011). The measured effects are modest: the programmes reduce risk, but their effects on symptoms are small and often fade (section 9). Tell families so: the programme adds protective factors, it guarantees nothing.
3. Routes into the programme
1. The team treating the parent refers the family. Several texts ask clinicians who treat adults to take an interest in their patients' children. NICE, the British institute that publishes care guidelines, asks that the initial assessment of an adult with psychosis cover their responsibilities towards children (NICE, 2014a). In Norway, since 2010, the law has required every professional treating a patient with a mental illness or an addiction to find out whether they have minor children and to talk with them about the children's needs (Helsepersonelloven, § 10 a); in Finland, the law requires the support needs of these children to be assessed (Terveydenhuoltolaki 1326/2010, section 70). The parent is often relieved that someone is thinking about their children, and worried about what will be thought of them: the first meeting takes place with them alone, and the brief format is often enough.
2. The child is doing badly. They are referred for their own difficulties, and the parent's illness emerges during the assessment, often never named at home. The child's disorder is treated with its own treatment, family work is offered in parallel, and you check that the parent is receiving treatment, because the recovery of a depressed mother is accompanied, on average, by an improvement in her children (section 5). When it is the school that is worried, faced with a child who is exhausted, absent, or who says they look after their parent, you also go through the parents, and you alert the child protection services only if there is a danger.
3. A crisis or a hospital admission. SCIE describes a parent's hospital admission as a major crisis for the family: routines disrupted, adults overwhelmed, parents and children worried and powerless (SCIE, 2011). The urgency concerns the immediate situation: who is looking after the children, what they are told today, how they keep in touch with their parent, and whether any of them has seen something frightening (sections 7 and 30). The programme begins once the parent has stabilised.
4. The parent asks for help. A parent in treatment asks how to talk to their children about their illness: this is the most favourable situation, because their request concerns exactly what the programme works on.
A vignette. Inès, 9, has a stomach ache every morning. Her mother has been treated for depression for a year; nobody has told Inès, who thinks her mother is "cross with her" and checks every morning, before leaving, that she has got up.
The three triage questions. The first, "Is anyone in the family in danger at the moment?", is asked of each adult alone, then of the child alone. The second, "What do your children know about your illness?", is often answered with "nothing", which is almost always inaccurate. The third, "How is each of your children doing, at home, at school and with their friends?", separates the child who is doing well from the one whose difficulties call for an assessment of their own.
4. What the child goes through, by age
What children themselves say
A meta-synthesis of 22 qualitative studies describes children who are responsible and worried, who feel ashamed when their parent behaves differently, hide their family life for fear of being stigmatised or bullied, and put their own needs after the parent's (Dam and Hall, 2016). Canadian children aged 6 to 16 understood their parent's illness poorly and had received few factual explanations, which caused them unnecessary suffering; the youngest feared above all their parent's death, the oldest also feared becoming ill in turn (Mordoch, 2010). These children adapt by keeping watch over the parent and adjusting to their behaviour to preserve the bond, or by keeping a protective distance (Mordoch and Hall, 2008). The age-related pointers that follow come from practice: no evaluated programme offers a validated breakdown by age.
From 4 to 6. In practice, the child sees the changes without being able to explain them: a parent who stays in bed, who cries, who flares up, who talks strangely, or who disappears into hospital. They may believe that they made their parent ill, or that they will cure them by being good, and they react through their behaviour more than through words. They need short, repeated explanations, routines, to know who is looking after them, and to go on playing and seeing their friends: the youngest need to remain children (Dam and Hall, 2016).
From 7 to 11. In practice, the child compares their family with their classmates' families and asks their questions, often of someone other than their parents: "Is Daddy going to die?" They begin to hide the situation, to stop inviting friends round, to watch for their parent's mood, and they take on tasks: making their own after-school snack, looking after a younger sibling, reminding their parent about a medicine, comforting.
From 12 to 17. In practice, a question appears: "Am I going to become like him?" (Mordoch, 2010). Some adolescents take on an adult's role or become the parent's confidant; others take a distance that parents read as indifference, when it is often a way of protecting themselves. It is also the age at which the young person's own risk becomes visible: among the children of depressed parents followed for twenty years, the incidence of depression peaked between 15 and 20 (Weissman et al., 2006b).
At all ages, three themes recur. The child feels guilty, and Beardslee's programmes explicitly seek to reduce this guilt (Beardslee et al., 2003). They keep quiet out of shame: a review of 32 studies describes the stigma "by association" that these children experience, the stigma they are subjected to, the stigma they anticipate, the stigma they turn against themselves, and the discrimination linked to institutions (Dobener et al., 2022). And they want to count: children often want to be recognised as important to their parent's well-being, whereas adults seek to protect them from excessive responsibility (Gladstone et al., 2011). The programme recognises what the child does, and takes away what is not theirs to carry. At any age, the following should raise concern: difficulties that persist or spread, absences from school to stay with the parent, withdrawal from friends, talk of death, self-harm, substance use, and a child exhausted by what they carry.
5. What harms, what protects: the programme's model
The risk is real, and it is not a fate
It is higher than in other children. In a meta-analysis of 33 family high-risk studies, the children of parents with schizophrenia, bipolar disorder or major depression had a 32% probability of developing a severe mental illness in adulthood, a relative risk of 2.52 compared with controls (Rasic et al., 2014). The largest meta-analysis, of 211 studies, gives for the lifetime risk of having at least one mental disorder a relative risk of 3.0 and an absolute risk of 55% when the parent has an anxiety disorder, 2.1 and 55% when they have bipolar disorder, 1.9 and 51% when they have depression, 1.5 and 38% when they have a substance use disorder, and 2.6 and 17% when they have psychosis (Uher et al., 2023). These absolute risks cannot be compared from one disorder to another, because they come from different studies whose control groups do not themselves have the same rates: the low figure for psychosis does not mean that these children are the least exposed, and the relative risk better expresses the gap with other children.
It does not concern only the parent's disorder. The risk of having the same disorder as the parent is the most increased, multiplied by 3.59, but the risk of having another severe disorder is increased too, multiplied by 1.92 (Rasic et al., 2014). In a Danish cohort of 865,078 children, a severe disorder in a parent was associated with almost all mental disorders in the child, including those with no apparent link to the parent's (Dean et al., 2010), and the children of anxious parents have more anxiety and depressive disorders, without preferential transmission of their parent's specific anxiety disorder (Lawrence et al., 2019). The child does not inherit an illness: their vulnerability is partly specific to their parent's disorder, partly more general.
Many children do not develop their parent's disorder. In Rasic's meta-analysis, about two thirds of the children of parents with schizophrenia, bipolar disorder or major depression had not developed a severe disorder in adulthood (Rasic et al., 2014). The lifetime risk of having the same disorder as the parent was 8% for psychosis, 5% for bipolar disorder, 14% for depression, 31% for anxiety and 9% for substance use disorders (Uher et al., 2023). In the Danish registers, 7% of the children one of whose parents had been admitted to hospital for schizophrenia had received this diagnosis by the age of 52, compared with 0.86% of those with no affected parent: 93% had therefore not received it (Gottesman et al., 2010). For psychosis, one can thus say that the great majority of children do not develop their parent's illness, and for bipolar disorder that most do not, even with the least favourable figure, 36% at 28 (section 9). For depression, the message is more fragile: in the same meta-analysis by Uher, the cumulative curves reach 24% at 18 and 46% at 28 among the children of depressed parents, and in the Welsh cohort described below, only 20% of the adolescents of parents with recurrent depression remained in good mental health across three assessments over four years (Collishaw et al., 2016). With the family of a depressed parent, say that many children do not develop their parent's illness and that nothing is written in advance, rather than "most".
These figures are fragile. The associations between maternal depression and the child's difficulties are small in magnitude and vary according to many factors (Goodman et al., 2011), and 96% of the studies in Uher's meta-analysis had a high risk of bias. What you tell a family about them comes down to four sentences: the risk is higher; it does not concern only the parent's illness; many children do not develop their parent's illness, and nothing is written in advance; and what happens in the family matters.
How the risk is passed on
The Goodman and Gotlib model. In the children of depressed mothers, it describes four transmission pathways (heritability, an early neurobiological vulnerability, exposure to the parent's negative thoughts, behaviours and emotions, and a stressful life context), which are moderated by the father's health and involvement, the course and timing of the mother's depression, and the child's characteristics (Goodman and Gotlib, 1999).
Genes, in part. Heritability is estimated at 37% for major depression and 81% for schizophrenia (Sullivan et al., 2000, 2003). It describes a share of the variation within a population, not the probability that a given child will be ill, and these studies describe disorders that result from both genetic and environmental factors. An adoption study illustrates this: the depression of an adoptive mother, who shares no genes with the child, was associated with depression and behavioural disorders in the adolescent (Tully et al., 2008).
Parenting during the illness. Maternal depression is associated above all with more negative behaviour towards the child, somewhat less with disengagement, and weakly with less positive behaviour; these effects are strongest during an episode (Lovejoy et al., 2000). In mothers with borderline personality disorder, studies more often describe interactions that are insensitive, overprotective or hostile (Eyden et al., 2016).
The accumulation of difficulties. An isolated risk factor does not necessarily have a serious effect; risks add up, and an impact becomes much more likely when three or more factors combine, such as an addiction, domestic violence and a mental illness without support. That is why the diagnosis is not enough: the real impact of the symptoms on the family must be assessed (SCIE, 2011). Poverty is part of this: in the United Kingdom, exposure to maternal mental illness reaches 28.3% of children in the most deprived areas (Abel et al., 2019).
Adult responsibilities. SCIE uses the term young carer for a minor who takes on tasks and a responsibility normally assumed by an adult: paperwork, post, contacts with professionals, medicines, emotional support, housework (SCIE, 2011). "Parentification" reported in childhood has a small but reliable association with mental disorders in adulthood (Hooper et al., 2011), and young carers have on average poorer health than their peers, with weak evidence (Lacey et al., 2022). A British study points out that this risk is in no way inevitable, and also describes positive outcomes, such as a good-quality parent-child relationship (Aldridge, 2006).
What protects
Relationships, and thinking for oneself. Resilient adolescents of parents with a mood disorder were characterised by a deep understanding of themselves, a strong investment in their relationships, and the ability to think and act independently of their parent, even though many of them were looking after that parent (Beardslee and Podorefsky, 1988). In adolescents, social connectedness was more clearly associated with good adjustment than knowledge about the illness (Fraser and Pakenham, 2009).
Five protective factors that add up. In a Welsh cohort of 262 adolescents of parents with recurrent depression, five factors were associated with maintaining good mental health over four years: the expression of positive emotions by the depressed parent, support from the other parent, good-quality social relationships, a sense of self-efficacy and frequent physical activity. They added up: the proportion of adolescents who remained in good mental health rose from 4% with zero or one factor to 48% with all five, and the authors conclude that difficulties are "common but not inevitable" (Collishaw et al., 2016). Emotional support from the father, fostered by a good couple relationship, was associated with fewer depressive symptoms in these adolescents (Mahedy et al., 2018).
Understanding, and being able to talk about it. According to SCIE, children ask for information suited to their age, someone to talk to and the chance to see their friends, and helping each family member to understand the illness better helps everyone to cope (SCIE, 2011). Children fare better thanks to information, openness and trusting relationships (Dam and Hall, 2016).
Treating the parent also helps the child
This is the finding that most changes practice, and it concerns maternal depression: for other disorders, the link is plausible but has not been studied in the same way. In an American study of 151 mother-child pairs, remission of the mother's depression after three months of medication was accompanied by an 11% fall in diagnoses among children aged 7 to 17, compared with a rise of about 8% when the mother did not recover; no child who had no diagnosis at the start and whose mother had recovered developed a disorder within three months, compared with 17% of the children whose mother remained depressed (Weissman et al., 2006a). It is an observed association, not an effect established by randomisation, but the trials point in the same direction: psychotherapy for depressed mothers slightly improves their children's mental health (Cuijpers et al., 2026). And the link runs both ways: the course of the parent's symptoms predicted the child's, and vice versa, partly through the acceptance the parent showed the child (Garber et al., 2011).
The model that guides this programme
The risk is not a fate, and much of what weighs on the child can be changed: the duration of the illness, through the parent's treatment; everyday life during the bad periods; the child's sources of support; and what they understand of what is happening.
Protective factors add up, at least the five factors of the Welsh cohort. The programme seeks to strengthen those that can be strengthened, support from the other parent, relationships with peers, an activity, the parent's warm moments, and it adds its own targets, which have not been evaluated as protective factors: an accurate explanation, an adult to talk to, routines that hold, a plan for crises, a parent who is receiving treatment.
It is the parents who talk. This is a choice made by this programme, taken from the structure of Family Talk, more than a demonstrated finding. In the first Family Talk trial, the clinician-led intervention improved the children's understanding more than lectures intended for the parents, which argues for linking information to the family's experience and involving the children directly (Beardslee et al., 1997). And the more the parents changed the way they talked and behaved with the child, the more the child's understanding of the illness progressed (Beardslee et al., 2003), an association that does not say what causes what. On the child's symptoms, however, Family Talk did no better than the lectures (Beardslee et al., 2003), nor than "Let's Talk about Children", which does not bring the family together, in the Swedish study (Wirehag Nordh et al., 2023).
The circle. The illness changes the parent, through fatigue, irritability, withdrawal or unusual behaviour, and it changes everyday life. Without an explanation, the child fills the gap with their own hypotheses: they believe they are responsible, they are afraid that their parent will die or that they will become ill, they keep watch, they take on adult tasks, and they hide the situation out of shame. They then drift away from their friends and activities, that is, from what protects them most, and their own difficulties appear, which in turn weigh on a parent who is already fragile. You draw this circle with the parents in session P1, and each arrow gets its own session: the explanation (E1, P3, F1), support (E2, R), everyday life and tasks (P2, F2), and the parent's care (Év1, link with their team).