Causal mechanisms of risk transmission

The repeated finding of an association between parental depression and adverse child outcomes over the past 40 years has led many to the belief that the association must be causal. However, there are a number of problems with this assumption. First, we are at risk of forgetting that depression is associated with many other adversities in life, and so may be acting as a risk marker, rather than a causal risk factor. ^{Reference Thapar and Rutter4} The peril of this is that it can lead to simplistic, and usually only partially effective, solutions, as treating the parental depression alone may miss other key difficulties that need addressing. Indeed, the recent finding, published in this issue (Sellers et al), that even in a group of mothers with recurrent depression, additional psychopathology, such as alcohol use and antisocial behaviour, adds significantly to the risk for adolescent depression, serves as a timely reminder of the complexity of factors that interact to determine vulnerability in offspring. Second, much of the evidence linking parental depression and adverse child outcomes is correlational. Given the challenges of doing experimental work in this area, it is difficult to unpick the genetic and environmental contributions to such associations.

Intervention trials are clearly the best way to establish clear causal influences. However, these are challenging to undertake and the findings to date are somewhat mixed. To complement this, there have been some very elegant studies that have added strength to the notion that parental depression can directly have an impact on child outcomes. Novel designs such as those using in-vitro fertilisation (IVF) ^{Reference Rice, Harold, Boivin, Hay, van den Bree and Thapar5} have helped to begin the long task of establishing what is, and what is not, inherited through direct genetic transmission. Animal studies using cross-fostering designs have added to this, highlighting the importance of early maternal care (which is often disrupted in postnatal depression). ^{Reference Weaver, Cervoni, Champagne, D'Alession, Sharma and Seckl6} Unravelling the key mechanisms by which parental depression may influence child development is a final way of substantiating the causal argument. Here again, animal studies, including the work of scientists such as Meaney and colleagues, have meant that the science of epigenetics, which has excited many fields of medicine, also has increased relevance for those of us working in psychiatric research.

Clinical interventions

One key clinical benefit of an increased awareness of the impact of parental depression on child development is the potential to identify a group of children who are at an increased risk of adverse outcomes. In turn, this offers the opportunity to intervene and so potentially break the pattern of transfer of risk of depression from one generation to the next. The widespread screening of women for depression in the postnatal period, and increasingly during pregnancy in some countries, ⁷ has been a significant development. Clearly, identifying and treating parents with depression is of considerable worth, but the potential additional benefit of also improving outcomes for children is attractive. The challenge then is to intervene in an effective and timely manner to reduce both parental depression and also the associated risk in offspring. Evidence from some studies of older children suggests that the treatment of parental depression can lead to improvements in children's psychological functioning, ^{Reference Weissman, Pilowsky, Wickramaratne, Talati, Wisniewski and Fava8} and programmes focused on both parents and children in families where a parent has depression have also shown promising outcomes. ^{Reference Beardslee, Gladstone, Wright and Cooper9,Reference Solantus, Paavonen, Toikka and Punamaki10} Examples of the development of services for parents with mental illness, such as the NSPCC Parkside Parental Mental Health Service in London, UK, or the Children of Parents with a Mental Illness (COPMI) initiative in Australia (www.copmi.net.au), are still too few in number, but offer hopeful signs of more joined-up patterns of mental healthcare that may benefit children whose parents have more severe mental illness. However, it is worth noting that evidence supporting the impact of interventions for improving outcomes in younger children is less convincing. ^{Reference Forman, O'Hara, Stuart, Gorman, Larsen and Coy11}

Although there have been these positive clinical and research advances made in our understanding of the impact of parental depression on children and adolescents, there are some potential pitfalls associated with an increased focus on the wider-reaching impact of parental mental health. For example, the public health success of raising the profile of postnatal depression, so that it is now screened for, and frequently spoken about, has had undoubted benefits for the many women now receiving treatment. However, the impact in terms of understanding that the children of women with postnatal depression are at increased risk of cognitive, emotional and behavioural problems has had a perhaps more mixed outcome to date. The repetition of the maxim that children of women with postnatal depression are at increased risk for a range of problems, from lower IQ to antisocial behaviour, has in some cases had a stigmatising effect, leading to increased worry in women already experiencing anxiety and depression. The loose use of explanations of risk has led to some blindness to the fact, that although there may be an increased risk for such children, the vast majority of children of mothers with depression (probably over 80% for most outcomes) are relatively or completely unaffected.

The right response

So what are we to make of parental depression? Clearly there is some way to go in our understanding of the processes of risk transmission, and exciting scientific avenues are waiting to be explored. Further delineation of whether and when parental depression has a causal impact on child outcomes, and what the key mechanisms are, will lead to a better focus for our preventive efforts. But what should clinical services be doing in the here and now? Despite the challenges that we highlight, the realisation that one of our adult patients has children, should lead to consideration of the effect of that illness on the children at home and a more family-focused approach to treatment. Supportive interventions exist for children who are affected, ^{Reference Beardslee, Gladstone, Wright and Cooper9} and these can run alongside existing effective treatments for the parent. We can and should be implementing these today.