Article Text

Download PDFPDF

Physical illness in looked-after children: a cross-sectional study
  1. Alice Martin1,
  2. Tamsin Ford2,
  3. Robert Goodman3,
  4. Howard Meltzer4,
  5. Stuart Logan5
  1. 1Department of Paediatrics, Royal Devon and Exeter Hospital, Exeter, Devon, UK
  2. 2Child Health Group, University of Exeter Medical School, Exeter, Devon, UK
  3. 3King's College London, Institute of Psychiatry, London, UK
  4. 4Department of Health Sciences, University of Leicester, Leicester, UK
  5. 5Peninsula CLAHRC, University of Exeter Medical School, Exeter, UK
  1. Correspondence to
    Dr A Martin, Department of Paediatrics, Royal Devon and Exeter Hospital, NHS Foundation Trust, Barrack Road, Exeter, Devon EX2 5DW, UK; alice.martin{at}


Objective To compare the reported point prevalence of chronic physical illness among children looked after by local authorities with those living in their own homes.

Design Cross-sectional study, using questionnaire data from a national survey.

Setting The UK.

Participants Random samples of children aged 5–15 years. Children looked after were selected from Department of Health databases, stratified according to placement type. The child benefit register was the sampling frame for children in their own homes, weighted to match the child population demographic and compensate for response variability.

Main outcome Carer-reported prevalence of 10 physical illnesses.

Results Data were collected on 1253 looked-after children and 10 438 children in their own homes. There were lower rates of asthma, eczema and hay fever reported among looked-after children compared with children at home (ORs, adjusted for age, gender and ethnicity, were 0.63, 0.61 and 0.36, respectively). Epilepsy, cystic fibrosis and cerebral palsy were more commonly reported in looked-after children (adjusted ORs 4.13, 4.2 and 7.26, respectively). There was no difference in the proportions of children in the two groups reporting glue ear, diabetes mellitus, spina bifida or cancer.

Conclusions Looked-after children have an increased prevalence of some physical illnesses. The results also suggest that there may be significant unmet need, with health professionals and carers failing to identify other illnesses. The lower reported prevalence of atopic conditions may reflect a truly lower occurrence of such diseases in looked-after children; this requires further work to explore.

  • Comm Child Health
  • Epidemiology
  • General Paediatrics

Statistics from

What is already known on this topic

  • Children looked after by local authorities have a higher prevalence of physical illnesses than those in their own homes.

  • Looked-after children have frequent changes in caregivers and healthcare providers, making continuity of healthcare difficult.

  • Looked-after status is associated with low socioeconomic class which contributes to physical ill health.

What this study adds

  • Cerebral palsy, epilepsy and cystic fibrosis have a higher reported prevalence in looked-after children. By contrast, atopic illnesses have a lower reported prevalence.

  • There may be significant unmet need if health professionals and carers fail to identify chronic conditions. This emphasises the importance of transferring accurate information.

  • The lower atopic disease in looked-after children is intriguing but requires independent confirmation.


Looked-after children are some of the most vulnerable people in our society. Their adverse social circumstances and age mean they are dependent on others, frequently professionals, for their care. Despite this there is relatively little evidence regarding their health. In the UK looked-after children are all those in the care of the local authority, either voluntarily or following a court order, although the parental responsibility may still be shared with the parents.1 There are approximately 67 050 looked-after children (57 per 10 000 children) in England.2 There has been long-standing concern that looked-after children have worse physical health than the general paediatric population, which seems likely considering their adverse personal and social circumstances. Observational studies, initially from the USA, have reported a high prevalence of physical health problems in children in care.3–5 These studies suggest that 80–90% of looked-after children have at least one physical health problem and a majority (50–60%) has multiple health needs. Viner and Taylor have reported that these children have worse physical health into adulthood.6 Similar results were reported by Nathanson and Tzioumi among a population of looked-after children in Australia. They found that 31% of 122 looked-after children seen in an assessment clinic required referral for specialist medical attention. Unfortunately their conclusion that this is higher than the general paediatric population is not supported by data. The only comparative data given are for hearing, visual and speech problems.7 Williams et al8 in 2001 reported that looked-after children in Wales have lower immunisation rates, inadequate dental care, higher illegal drugs use and smoking and received less health education compared with matched controls.

A disproportionate number of looked-after children come into the care system from less affluent families; their initial sociodemographic situation itself predicts a higher prevalence of chronic physical ill health.9 ,10 There are several other possible reasons why looked-after children may have worse physical health than their peers living in private households; failure of advocacy with rapid changes of primary caregivers and poor communication between carers,1 more frequent changes in location of community services such as general practitioner8 and non-adherence in children who may have emotional and behavioural problems. It is important to recognise that some children with chronic health problems can be more difficult to care for, which in families who are struggling to cope with other multiple stressors, may itself precipitate a need for local authority care.11

In this study we compared the prevalence of carer-reported (and thus identified) chronic physical illness among looked-after children compared with the general population using data from national surveys conducted by the Office of National Statistics. Comparative data on the association of psychiatric disorders with personal and social factors from the same study have been published.12

The estimate of prevalence will depend on the ‘true’ prevalence and the likelihood of identification and recording of the diagnosis. We recognise that conditions will be differentially identified between the two populations due to variations in their severity. Therefore, we have examined the reported prevalence of a range of conditions of different severity. We hypothesised that the reported point prevalence of the 10 physical illnesses will be higher in the looked-after population of children.


Study design

This cross-sectional study used data from two surveys conducted by the Office for National Statistics on behalf of the British Department of Health.13–16 The surveys were carried out between 2002 and 2004. Each survey used the same study team and asked parents or carers to report the presence of the same physical illnesses, listed in tables 2 and 3. Therefore the outcome was the reported point prevalence of illness, according to parents and carers.


Children looked after by local authorities

It is a statutory requirement that all local authorities annually inform the Department of Health about children under their care. Random samples of children (aged 5–17 years) looked after by local authorities were selected from the relevant Department of Health databases of England, Wales and Scotland. Figure 1 (reproduced with permission from the British Journal of Psychiatry) illustrates the complex survey design and response rates. The sampling was stratified according to placement type (ie, foster care or residential home, etc) and therefore we have data on children from a range of care settings. Following random selection, a contact person in each administrative area was sent child summary forms for each child selected. After obtaining whatever consent the local authority deemed necessary, the contact person was responsible for ensuring that the child's social worker completed the information on the child summary forms and for returning the completed forms to the Office for National Statistics. Similar proportions of children from each country were deemed ineligible, and response rates among eligible children were very high. Data were collected on 1543 looked-after children, 90% of those eligible (n=1706), see figure 1. In order to match the age range of the private household survey (5–15 years), the current analyses excluded looked-after children aged 16–17 years (n=290); in total, data on 1253 looked-after children were analysed.

Figure 1

Diagram illustrating the sample selection and response rate from the surveys of children looked after by local authorities in England, Wales and Scotland, listing the mail reasons for ineligibility in each country. Reproduced with permission from the British Journal of Psychiatry.

Private household sample

The child benefit register was used as the sampling frame for England, Wales and Scotland, and 14 250 children were sampled by the postal sector. To improve the representativeness of the data, samples were weighted with differential sampling between postal sectors to enable the sample to match the age and sex distribution of 5–15-year-olds in the population at the time of the survey. Further weights compensated for response variability, for example, poorer response in inner cities and among less affluent families.13 Information was collected on 10 438 (83%) of the 12 529 eligible children. Figure 2 shows the reasons for ineligibility, including opting out by responding to a free phone number on the initial letter sent to 14 250 houses.13

Figure 2

Diagram illustrating the sample selection and response rate from the surveys of children in private households, listing the main reasons for ineligibility.


All the surveys used the same questions to establish if the parents or carers recognised a range of physical illnesses in their children, which are available as appendices to the original survey reports.13–16 The 10 illnesses were; asthma, eczema, hay fever, glue ear, epilepsy, diabetes mellitus, cystic fibrosis, spina bifida, cancer and cerebral palsy. The parents and carers’ response options were ‘yes’ or ‘no’. The outcome was detected by parent or carer reporting, there were no medical consultations to validate this.

Parents and carers also reported demographic details such as age (in years), gender and ethnicity (15 categories were used). The socioeconomic status of families in the general population survey was classified by National Statistics Socio-economic Classification, based on the occupation of the head of the household. For looked-after children however no classification of socioeconomic status was available. Due to the small number of children from some ethnic groups, we redefined as white and non-white for the analysis.


All analyses were conducted in STATA V.11.1.17 We compared the odds of reporting each illness between looked-after children and those in the general population and calculated 95% CIs for each OR. We used a multiple logistic regression analysis to adjust for age, sex and ethnicity.

Recognising the likelihood that the majority of looked-after children come from deprived backgrounds, we conducted a secondary analysis which compared the ORs for each condition among looked-after children and children from the most deprived group on the general population. This group was defined as households where the head of the household had either never worked or worked in an unskilled occupation. Where there were less than 10 children in any cell in these secondary analyses the data are not reported.

Approval for the original surveys was granted by London MultiCentre Research Ethics Committee and the Local Research Ethics Committees for each area in whichthey were undertaken, while approval for secondary analysis of these data was granted by the Peninsula Medical School Research Ethics Committee in 2007.


The surveys included 1253 looked-after children and 10 438 children living in their own homes. The age range was 5–15 years and the mean ages were similar in the two groups. There were a higher proportion of men in the group of looked-after children, which would be expected from registers of children who are looked-after children.2 There were only 122 missing entries for the questions on physical illnesses (1% of total), no missing entries for age or gender and only seven for ethnicity (table 1).

Table 1

Baseline demographic data

Reported prevalence of physical conditions

Statistically significantly lower rates of asthma, eczema and hay fever were reported among children in the care of the local authority compared with children living in their own homes while epilepsy, cystic fibrosis and cerebral palsy were statistically significantly more commonly reported among in looked-after children (table 2). We found no difference in the proportions of children in the two groups reported as having glue ear, diabetes mellitus, spina bifida or cancer. These differences persisted after adjustment for age, gender and ethnicity.

Table 2

ORs for physical illness in the looked-after children and those in private households

In the secondary analysis comparing looked-after children with those from the least advantaged socioeconomic group of children in their own homes, we found a broadly similar pattern although the differences between groups were no longer statistically significant for eczema or epilepsy and the numbers of children with cystic fibrosis and cancer were too small to compare (table 3). Again these results remained similar after adjustment for age, gender and ethnicity.

Table 3

Subgroup analysis


The conditions we looked at fall into three groups: those more commonly reported in looked-after children, those less commonly reported and a group where we found no statistically significant difference. Cerebral palsy, epilepsy and cystic fibrosis were all more frequently reported in looked-after children. This is consistent with other studies that suggest that serious, chronic illnesses, particularly those causing disability, contribute to the range of factors that may lead a child to requiring local authority care.11 This may in part reflect the extra stresses that caring for these children can place on their families, increasing the likelihood of needing support from social services. It has also been reported that children from poorer families are at higher risk of conditions such as cerebral palsy,9 although not cystic fibrosis, and social disadvantage is in itself a risk factor for children requiring local authority care. Confounding by socioeconomic status may therefore contribute to this finding. When we compared looked-after children with those whose parents had either never worked or worked in unskilled occupations, we still found increased rates of reporting of cerebral palsy, while the OR for epilepsy was attenuated and no longer statistically significant. There were too few children with cystic fibrosis for this secondary analysis.

Asthma, eczema and hay fever were all substantially less commonly reported among looked-after children, results which are highly statistically significant. It is unclear whether these findings represent a true difference in prevalence or are explained by bias in diagnosis or failure to communicate health information as the child moves between care settings. Differences in the prevalence of atopic disease have been related to social factors in some epidemiological studies including reports of an inverse relationship between family size and the prevalence of hay fever, eczema and allergen specific skin positivity.18–20 One proposed mechanism is the ‘hygiene hypothesis’, which, although controversial, is based on strong epidemiological data and a plausible biological mechanism.21 This suggests that more frequent exposure to infections in young children related to overcrowding reduces the likelihood of the development of atopic disease. It is tempting to postulate that looked-after children may have been at higher risk of infection due to adverse social conditions. However, the secondary analysis showed similarly lower rates of reported atopic disease in looked-after children compared with children from the least advantaged groups. There are a number of possible explanations for this. First, social disadvantage may not be the key factor. Second, the hygiene hypothesis may not apply to ‘ordinary’ poor households while still applying to the very chaotic households where many looked-after children spend their early years. Third, the low rates of atopic disorder primarily reflect under-recognition by carers of looked-after children, either because no diagnosis has been made, possibly as a consequence of a failure to consult regarding symptoms, or because carers are not adequately aware of the child's medical history. It has previously been reported in the UK that health professionals carrying out statutory health assessments frequently do not have complete past health data available 1 ,22 which suggests that the latter is a plausible explanation.

The small numbers of children in the sample with cancer, spina bifida and diabetes mellitus provide little power to explore differences in reported rates. We found no difference in reported rates of glue ear between groups and because it is a relatively common condition the CIs round the estimates are narrow. The known associations between social disadvantage9 and glue ear meant that we would have anticipated that at least in the primary analyses we would have seen higher rates among looked-after children. This may add credence to the hypothesis that there is a risk of lack of recognition of less serious health problems among looked-after children.

These data support previous reports that looked-after children are at increased risk of a number of chronic health problems, highlighting the importance of careful health assessments, accurate clinical documentation and effective transfer of information to other professionals and to the children's carers.

Strengths and limitations

This is the largest UK epidemiological study of looked-after children. The findings are strengthened by the use of random sampling and inclusion of young people in a range of placements. The study had high response rates, 83% of eligible children in the private households survey and 90% in the looked-after children survey. These features minimise selection bias and increase the likelihood that conclusions can be extrapolated to the whole population of looked-after children. Unlike most previous studies of looked-after children the design allows the comparison of looked-after children with children of a similar age who live in private households.

Another strength of this study is the completeness of data and therefore no measures were needed to address the effect of missing data.

Health data are derived from questionnaires completed by parents or carers and hence the degree to which findings reflect true estimates of prevalence depends on the accuracy of their reporting and, for looked-after children, the effectiveness of communication of health data to those with parental responsibility.


These results reinforce previous work suggesting an increased prevalence of some significant chronic conditions among looked-after children. They also suggest that there may be significant unmet need with health professionals and carers failing to identify other chronic health conditions such as asthma.

Looked-after children frequently move between healthcare providers and care settings. Our findings highlight the importance of accurate clinician documentation and transfer of information between clinicians and to non-healthcare professionals and families. They also emphasise the need for those providing care to consider the possibility of undiagnosed or unrecorded health problems.



  • Contributors HM designed the study and led a group at the Office of National Statistics in carrying out the surveys for data collection. AM performed the data analysis. AM and SL drafted this paper. All authors contributed to reviewing and critically developing the paper. HM was the guarantor.

  • Funding We acknowledge funding support for Stuart Logan by the National Institute for Health Research (NIHR) Collaboration for Leadership in Applied Health Research and Care (CLAHRC) for the South West Peninsula. AM completed this work while supported by a NIHR Academic Clinical Fellowship in Paediatrics. The views expressed in this publication are those of the authors and not necessarily those of the NHS, the NIHR or the Department of Health in England.

  • Competing interests None.

  • Ethics approval London MultiCentre Research Ethics Committee, Local Research Ethics Committees, Peninsula Medical School Research Ethics Committee.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles