Background The adverse life-long consequences of being looked after as a child are well recognised. However, systematic evidence on outcomes for looked-after children beyond the early adult years is currently very limited.

Methods Data were used from »1 20 000 dependent children (aged <18 and never-married) in the ONS Longitudinal Study (LS) in each of the census years 1971, 1981, 1991 and 2001 (total n=495,165). Separately for each census, logistic regression was used to compare odds of long-term limiting illness and self-rated health 10, 20 and 30 years later for LS members with different care experiences in childhood (parental household, relative household, non-relative household and residential care).

Results In this nationally-representative sample of children in England and Wales, approximately 1.4% of dependent children were looked after in non-parental households, 3664 children in relative households, 2351 in unrelated households and 1007 in residential care. From 1971 to 2001, there was a general trend of an increasing percentage of dependent children residing in relative households (0.4% to 1.2%) and decreasing percentage in residential care (0.3% to 0.1%)(p-values<0.001). At 10, 20 and 30-year follow-up, LS members who had resided outside the parental home in childhood, had higher odds of a long-term limiting illness and rating their health as ‘not good’ vs ‘good’; with the highest odds for residential care. For example, for LS members who were dependent children in 1991, odds of a long-term limiting illness 10 years later were 7.8 (95% confidence interval: 5.3, 11.4) higher for those who had resided in residential care, 1.9 (1.4, 2.7) higher in relative households and 2.4 (1.9, 3.1) higher in non-relative households, compared to residence in a parental household. For non-relative and residential care, odds ratios increased with each successive census year. The longer the follow-up period the weaker the strength of association, but with all associations remaining significant. Weaker associations were seen for self-rated health compared to long-term limiting illness.

Additional analysis to be presented will make step-wise adjustments for childhood demographics and social circumstances.

Conclusion Decades after children and youths are placed in out-of-home care; they are still likely to report worse health than children who grew up in a parental household. Future work from this project will investigate how much of these associations are due to socio-demographic selection into different care experiences, compared to effects of being in care.