4. Conclusions

We have found, in a community survey of infants who were full term births, with birthweights in the normal range, that about 3.5% have growth faltering in the first postnatal year (Skuse, Wolke & Reilly, 1992). To reach our case criteria the degree of growth failure had to be such as to warrant clinical concern, and is equivalent to that seen in previous investigations of hospitalised infants with the condition (see Skuse, 1993c). Yet at the time of our assessment less than 20% of the cases we identified had been referred to hospital for investigation (Skuse, Wolke & Reilly, 1992). Accordingly, the annual incidence figure we report of 3.5% of full term births is considerably higher than might be expected.

At approximately 15 months of age all case children were found to be underweight for their length to some degree. A variety of anthropometric variables which have been shown, in the developing world, to be weakly associated with poverty and malnutrition (e.g. Martorell, Mendoza & Castillo, 1988) were also obtained. They did not show any significant correlation with measures of psychosocial adversity in the families studied in this survey, but this is not surprising as the sample was relatively small. On a variety of criteria the population studied was socioeconomically disadvantaged and other indices of psychosocial adversity were positive indicating families were at relatively high risk of experiencing parenting difficulties. For example, a high proportion of mothers were depressed or suffering from other minor psychiatric disturbance, were living in unsatisfactory accommodation, had relatively low intellectual abilities and a lack of education.

By means of a simple arithmetical procedure it was possible to partition this sample of growth faltering infants into two subgroups: those in whom the onset of the condition was immediately after birth and those for whom it began three to six months later. Remarkably, the outcome in terms of cognitive and psychomotor development, but not in terms of anthropometric criteria, was much worse for those whose growth failure was early. In fact for those in the later group this outcome was not different from a normal comparison group drawn from the same population and closely matched on a wide range of criteria (see Skuse, Wolke & Reilly, 1992). Yet these children who faltered early came from relatively advantaged homes in which the burden of psychosocial adversity was lower than for the later growth faltering subjects.

In conclusion, serious growth faltering during the first postnatal year is more common among full term, otherwise healthy, infants living in a socioeconomically disadvantaged inner city environment than might be expected from the perspective of hospital practice. Their outcome, in terms of cognitive and psychomotor development in the second year of life, seems to be determined primarily be the timing, duration and onset of that growth failure (Skuse et al., 1993). Sustained failure to thrive through the whole period of early childhood, which leads eventually to stunting, is associated with an exceptionally poor outcome for mental abilities (Dowdney et al., 1987). The aetiology of such growth faltering is not simply a matter of nutrition, nor of poor parenting practices. An adequately broad perspective on the subject must also take into account the interaction between behavioural risk factors, a perspective that points the way towards the imperatives of any programme of preventive intervention.

Acknowledgements - This research was supported by grants from the Wellcome Trust, the Child Growth Foundation and the Newcome Educational Foundation, and the work was carried out at the Behavioural Sciences Unit, Institute of Child Health, London.

We thank the staff of Guy's Hospital Community Paediatric service, the Lewisham and North Southwark Health Authority Priority Care Unit and the local general practitioners who actively cooperated with our investigations. Further invaluable assistance with tracing subjects and records was provided by Mrs. Dorothy Gill and Mrs. Jennifer Smith, administrative assistant, who coordinated data collection and prepared the manuscript.