Our concern is the transfer of mental illness from generation to generation and our interest is to learn how to contribute to breaking the cycle in everyday clinical practice in the health care system. We studied the effectiveness of two interventions, the more extensive Family Talk Intervention (FTI) and a short child-focused Let’s Talk about Children discussion (LT), when the interventions were carried out in psychiatric health services for adults. Both interventions were coupled with a guide book for parents.
As hypothesized, both interventions were associated with children’s mental health and the effects could be still documented at 1.5-year follow-up. The main aim of the interventions is to prevent the rise of children’s psychosocial symptoms, and thereby eventually prevent outbreak of depressive and anxiety disorders in the offspring. The interventions did not only prevent the rise in children’s symptoms, but even led to a decrease in these symptoms in both groups. As we were able to control for the parents’ depression, these positive changes were not due to alleviation of parental depression, which is known to be associated with symptom reduction in children [20, 21, 38].
Children’s prosocial behaviour improved and emotional symptoms, anxiety, and marginally, hyperactivity decreased both in the FTI and the LT, while the FTI was more effective in reducing emotional symptoms relative to the LT. The higher effectiveness of FTI was time dependent.
It is noteworthy that the positive changes in the FTI tend to happen during the first 4 months, while the change in the LT takes more time. The FTI involves the whole family and also directly children and the intended family process is initiated during the intervention itself. In contrast, the impact of the LT on children is mediated totally through parents. Parents are given information, which they most likely process within themselves before they involve children, and the family process itself is likely to be slower.
Our study is the first one to document the effectiveness of the FTI on children’s symptoms over and above the control intervention. The US and Finnish studies have methodological differences which may explain this. Our study was a questionnaire study, while the Beardslee et al. study was interview-based. These discussions with the families may have increased the effectiveness of the lecture group. They might also predispose to report bias, i.e. parents may less openly express negative views on the interventions, when they are interviewed in person. The second difference concerns the study settings. As our study was carried out in real-world conditions, the interventions were imbedded in the patient’s normal treatment process. Therefore, the interveners may have been able to link the psychoeducational material and intervention process more closely to the patient’s and the family’s personal experiences and life situation. This is, indeed, the aim in the FTI and might contribute to its effectiveness in our trial. The effects of the FTI might be even stronger in real-world than in highly controlled efficacy trials.
Both interventions were related to a significant decrease in children’s anxiety and marginal decrease in hyperactivity. Anxiety is linked with experiences of threat and insecurity about the present and the future, which often characterize families with parental mental illness [13, 14]. Hyperactivity in turn can be understood as a behavioural response to the fear, worry and relationship problems documented in families with parental depression [1]. Hyperactivity can evoke punitive parenting practices thus accelerating family problems.
The main focus of our preventive interventions is to help family members to come together to master the family situation and to find strategies to deal with problems in a constructive way. This might build up a sense of security for all family members. We have, indeed, previously reported that a majority of parents in both interventions experienced increased confidence in children’s and family future and a decrease in their worries [36]. The present findings suggest that our interventions are successful in actually alleviating children’s problems and enhancing promotive behaviour.
Promotion of protective factors is an important goal in preventive and promotion efforts. Prosocial behaviour provides means for children to solve interpersonal problems constructively and to strengthen their relationships. Connectedness to peers and family is essential for healthy development [25] and one of the key protective factors for children in families with parental mental health problems [3]. Both interventions promoted in the long run children’s prosocial behaviour and thereby our finding suggests that they have not only preventive but also promotion capacity.
The relative similarity in the effectiveness of the intensive, family-based FTI and the brief parent-based LT is relevant for preventive policies. The costs of the FTI exceed those of the LT, but the FTI brings more immediate improvements to the children. Families with depressed parents are not a homogeneous group, but differ in their resources and vulnerabilities. Ideally, depressed patients with children should be offered preventive alternatives tailored to their needs. Further research should shed light on the mechanisms of positive change and identify families who benefit from one intervention rather than from the other. These questions are important for clinical practice as well as for further intervention development.
It has been argued that intervention development should include its applicability in other than the original culture [27]. The FTI has been studied in middle class, predominantly white population [2, 7] as well as African-American and Latino populations in the USA with favourable results [29]. Our positive findings in the Finnish health care provide strong support for the applicability of the FTI in new cultural settings, albeit within the Western cultural sphere.
Apart from the actual effectiveness findings, the study provides evidence that attending to the preventive child mental health needs is possible and effective in services for adults. The preventive interventions were implemented in the finnish national health service and carried out as part of the treatment protocol for adult patients with mood disorder. The Effective Child and Family Programme [35] thus presents a new approach, even a paradigm change for psychiatric services by including promotion of parenting and prevention of child mental health problems in the treatment for the adult psychiatric patient.
Strengths and limitations of the study
The strength of our study is in its nature as a true effectiveness study carried out by ‘lay’ practitioners in the national health services. As such it forms the next step [23] from the highly controlled efficacy studies in the process of intervention development and implementation.
There are also limitations, the main concern being the initial refusal rate (about 55–60%) and drop-out rate during the study (about 25%). It is common that prevention trials report high refusal and attrition rates [11, 27]. Large fraction of those who have initially consented, never participate in the intervention or withdraw during the study [25]. It seems more difficult to promote subject involvement and adherence in preventive than in treatment trials. This is understandable as preventive interventions deal with predicted but not present problems. Furthermore, many subjects in risk groups never develop problems and their refusal to participate in preventive interventions can be considered rational. In addition, our study involved whole families. In family approaches, all family members have to consent to the study, which lowers the consent rate compared to interventions for individuals. This was also seen in our study.
On the other hand, in some families, the depression itself might contribute to lack of energy and motivation to participate. This was the case also in our study, since the initial level of depression and anxiety was higher in those who withdrew during the study than in those who persisted. It may interfere in the generalizability of our findings for parents with severe symptoms.
Furthermore, the effectiveness design in our study might have also contributed to the refusal and attrition. Our interventions were imbedded in the patients’ normal clinical visits and carried out by the health care staff, while in efficacy studies, e.g. the Beardslee et al. study [2, 7] the families receive visits from outside experts coming from a prestigious research centre. This might raise the families’ motivation and it presents one further difference between a real-world effectiveness and a controlled efficacy study. Finally, we offered no pay for participation, which is the standard in Finland.
A further limitation is the lack of a control, no-interference group. The LT was a very brief intervention, but it did depart from earlier practice in that children were systematically discussed and parents received the self-help booklet [35]. A no-interference control group would have been needed to document a true prevention effect, i.e. a possible rise of symptoms in the control group with no change in the study group. The decrease of symptoms presented here is a treatment rather than prevention effect, although it might prevent the development of full-blown disorder in a longer follow-up.
Finally, child reports should also be studied. It is important to learn about children’s own experiences of their well-being and family. Multiple informants also contribute to a more comprehensive understanding of family dynamics.