Cullen et al ^{Reference Cullen, Fisher, Roberts, Pariante and Laurens1} describe childhood antecedents of schizophrenia: such prospective studies are rare. Retrospective research suggests that as the number of adverse childhood experiences increases, so does the risk for health problems, including alcohol misuse, ischaemic heart disease, suicide attempts and externalising behaviours.^{Reference Felitti, Anda, Nordenberg, Williamson, Spitz and Edwards2,Reference Hillis, Anda, Felitti and Marchbanks3} However, retrospective studies are prone to the biases associated with recalling early childhood. The best way to fully understand the mechanisms underpinning the relationship between adverse childhood experiences and later development is to follow children prospectively from early childhood.

We had a unique opportunity to achieve this in Glasgow because of the existence of the Women’s Reproductive Health Service (WRHS), which provides antenatal care for some of the most vulnerable women in Glasgow: those affected by problem drug or alcohol use or significant mental health or personality problems. This cohort is well characterised in terms of family adversity.

We conducted a feasibility study to see whether it was possible to assess the mental health of the children of very vulnerable mothers. We selected a random sample of ten women who had received antenatal care from the WRHS 7 years earlier. Of the ten children targeted, one was deceased, two had been adopted and one was uncontactable because the mother was in a woman’s refuge in a secret location. Of the remaining six, three opted out, one was uncontactable despite repeat attempts, and of the two whose mother provided consent, one then became uncontactable and the last opted out. Each woman received a minimum of ten phone calls and five attempted visits with a letter left each time (unless they had opted out in writing or by phone). Despite two members of staff working full time for 8 weeks, it was not possible to conduct any mental health assessments on these children of very vulnerable mothers. Our research team were able to meet with only two out of our target sample of ten women and did not succeed in assessing any of the children. In other words, despite persistent phone calls and home visits, eight of these vulnerable women and all of their children remain invisible.

The considerable resources available to our research team - including the potential to make multiple phone calls and visits - are not usually open to healthcare or social-care professionals. The question we then have to ask is, how do we reach these most vulnerable of families and safeguard the health of their children?