The importance of depression in children and adolescents, and the potential negative long-term sequelae are increasingly recognised. The article by Paterson et al1 in this issue of the Journalprovides some insights into depression in children aged 9 years in the ongoing Pacific Island Families Study.This study is the first large prospective longitudinal study of Pacific Island families born in Auckland, New Zealand and has as its aim to inform health intervention strategies for the Pacific Island population. These 1398 children and 1376 mothers have been followed up at 6 weeks (baseline), 1, 2, 4, 6 and 9 years.At this 9-year wave they report on a wide range of child and parent outcomes including maternal reports on their child's behaviour, maternal mental health including depression, partner violence, recent stressful life events, and parental alcohol and tobacco use.The children (858 children; 61% of the sample) were given a developmentally appropriate self-report screening questionnaire on depression. Information on bullying (both as perpetrator and victim) behaviour, involvement with gangs (e.g. hanging out, wearing gang colours, representing a gang in fights), the children's perceptions of their physical abilities, relationships with peers and family, general self-perception and school performance and teacher ratings of school performance and a standard global assessment of cognitive development were also collected. The associations of these factors with childhood depression were then investigated.Their main finding was that children involved in bullying, either as a victim or perpetrator, reported significantly higher levels of depressive symptoms than those not involved in bullying. The relationship between bullying and depression is complex but there is evidence from another internationally important and ongoing longitudinal study, the Christchurch Health and Development Study (CHDS), about long-term effects of bullying.The CHDS showed that if a parent or teacher reported that a child was a bully in middle childhood (ages 7 to 12) then as an adolescent and adult, they were at some risk (adjusted mean OR 1.3) of a range of mental health problems including depression.2 Moreover, parental reports of their child being a victim of bullying in early teen years in this study were associated with a range of mental health problems including depression and suicidality over adolescence and adulthood.The adverse effects of both bullying perpetration and victimisation have attracted increased public attention and concern over recent years. The recent youth 2012 survey gives some insight into current bullying in New Zealand schools.3 This nationally representative study on high school students in 2001, 2007 and now 2012 reported that there was little change in the proportion of students being bullied at school with around 7% of boys and 5% of girls reporting that they were bullied weekly or more often and 9% of students said they had been afraid that someone at school would hurt or bother them in the past year. Both of these figures were higher among younger students and bullying became less of a problem as students got older.School-based interventions tackling bullying in schools have been shown to improve not only the emotional, physical and social health of victims4 but also to have economic advantages with increased school attendance and attainment leading to better long term employment and earnings.Economic models show that each dollar invested in school-based interventions to reduce bullying result in $14 of net savings.5 Finland has implemented a nationwide programme in schools (the KiVa programme) with early results suggesting that if generalised to the Finnish population of 500,000 students there would be a reduction of 7500 bullied and 12,500 victims.6Depression in the pre-pubertal period is uncommon (prevalence rates of 1–3%)7 and affects boys and girls equally so that the prevalence of depression reported in Pacific children at 7% is relatively high. The prevalence of depression rises significantly through adolescence to that of adult rates with a strong female preponderance (2:1) emerging after puberty.Although depression runs in families, with offspring of parents with depression having 3–4 times the rates of depression as non-depressed parents, inherited factors only partially account for this. Indeed twin studies have shown that child onset depression has very low rates of heritability compared with the modest heritability (30–50%) of depression that has a late adolescent onset.8It is thought that depression in children often reflects stress within the family. Inherited factors appear to increase the risk indirectly through gene-environment interplay by a combination of increasing exposure to risky environments and by increasing the sensitivity of the brain to that risk through brain and neuroendocrine mechanisms. In this way, the risk of maternal (and probably paternal) depression are mediated through exposure to the environment that the depressed adult caregivers create around them.This environment may contain not only the exposure to the symptoms of depression in the adult, but also an increase in adverse chronic stressors, particularly those that affect relationships in the family and with peers such as bullying.Thus child onset depression is more strongly associated with family adversity, parental neglect and peer relationship problems. It is surprising then that in this study there was no association between child-reported depressive symptoms and maternal depression, stressful life events and family instability.There was also no association with socioeconomic factors with the exception of low maternal education, a well-known risk factor associated with a number of adverse child physical and mental health outcomes. Perhaps the low prevalence of parent self-reported depression may partially explain this.Children noted by their parents to have a range of internalising symptoms when they were aged 6 were also more likely to have depressive symptoms at aged 9, confirming the stability of these disorders in childhood.The Dunedin longitudinal study also looked at depression in their cohort of around 1000 (mainly Pakeha/New Zealand European) children at age 9 years born 40 years ago. They found that although parent reports overestimated depressive symptoms, parents were better than teachers in picking up on the irritable, non-compliant behaviour and somatic symptoms of their depressed children who had been identified by interview by a psychiatrist.They also found that depressed children had low self-perception but no evidence of cognitive impairment.9 Both the Dunedin study and CHDS have clearly shown that early onset emotional and behavioural problems persist and are related to a wide range of later adverse outcomes.10 These problems simply do not go away and point to a need for active intervention with this group of children.Pacific children at the other end of the spectrum with low levels of depressive symptoms had a range of protective factors including positive perception about themselves and their important relationships, and better scholastic performance. The importance of positive relationships in families and at school is once again underlined, in line with findings from other studies such as the series of Youth 2000 surveys.Longitudinal studies provide unique and important information on the trajectory of important determinants of health and wellbeing. This study, focusing on particular issues for Pacific children and their families, is a welcome addition to the suite of excellent longitudinal studies we have in New Zealand.