For children in social housing, to order BIRB-796 determine LBH589 biological activity whether there was a unique and 
significant effect of place (i.e., income quintile area) where the social housing unit was located. This analysis controlled for family-level income by using income assistance status of the family unit of each child as a surrogate for poverty of the individual. For complete immunizations at 2 years and school readiness outcomes, there was no statistically significant independent effect of income quintile area. However, for grade-9 completion, grade-12 high-school completion, and adolescent pregnancy, social housing residency in Q1 (and Q2 for the first 2) was associated with statistically significantly poorer outcomes compared with social housing residency in Q3, Q4, and Q5. For example, the odds ratio for grade-12 completion was only 0.34 (95 confidence interval = 0.15, 0.81) in Q1 and Q2 compared with Q5.Additional factors associated with poorer outcomes in at least 3 of the models in Table 3 include being on income assistance (i.e., living in poverty), having a mother who was younger at the 
birth of her first child, and being male for the education outcomes. Referring back to Table 2, a comparison of adolescents living in social housing by income quintile–lowest (Q1) to highest (Q5)–shows grade-9 credits at 36.9 versus 66.3 (i.e., 1.8 times more likely to complete grade 9 in Q5; P = .003), grade-12 completion at 26.5 versus 63.0 (2.4 times more likely to graduate from high school in Q5; P < .001), and adolescent pregnancy at 278.8 per 1000 versus 166.7 per 1000 (1.7 times more likely to have an adolescent pregnancy in Q1; P = .04).DISCUSSIONOverall, children living in social housing had poorer health and education outcomes than all other children, and in most cases when compared with those living in the same socioeconomic areas of Winnipeg. The implication is clear for government and community programs--health and education officials know exactly which housing is social housing, so targeted preventive health or educational programs could be provided to residents. When we compared children in social housing within different SES areas, outcomes were distinctively different for early childhood versus school age. For the 2-year-old complete immunizations and school readiness indicators,differences were similar by socioeconomic area of residence. In contrast, for school-aged indicators (i.e., grade 9, grade 12, adolescent pregnancy), the wealthier the area in which social housing was located, the better the outcomes for social housing residents. This was especially evident when we compared placement of social housing in the 2 poorest quintile areas (Q1, Q2) with placement of social housing in the middle and wealthiest income quintile areas (Q3---Q5), where children in the PubMed ID:http://www.ncbi.nlm.nih.gov/pubmed/19896111 latter showed improvements in positive outcomes. One could hypothesize that once children experience peer influence in school, their behavior may be affected by peers. So even though the outcomes of children in social housing were worse than their corresponding “all others” in wealthier neighborhoods (not surprising, when one considers the challenges of poverty), they did better compared with their social housing counterparts in poorer areas. Thus, “place of residence” of social housing may be important, with better outcomes for school-aged social housing residents for higher-SES areas. This would mirror the majority of studies earlier cited regarding the importance of neighborhood effects.For children in social housing, to determine whether there was a unique and significant effect of place (i.e., income quintile area) where the social housing unit was located. This analysis controlled for family-level income by using income assistance status of the family unit of each child as a surrogate for poverty of the individual. For complete immunizations at 2 years and school readiness outcomes, there was no statistically significant independent effect of income quintile area. However, for grade-9 completion, grade-12 high-school completion, and adolescent pregnancy, social housing residency in Q1 (and Q2 for the first 2) was associated with statistically significantly poorer outcomes compared with social housing residency in Q3, Q4, and Q5. For example, the odds ratio for grade-12 completion was only 0.34 (95 confidence interval = 0.15, 0.81) in Q1 and Q2 compared with Q5.Additional factors associated with poorer outcomes in at least 3 of the models in Table 3 include being on income assistance (i.e., living in poverty), having a mother who was younger at the birth of her first child, and being male for the education outcomes. Referring back to Table 2, a comparison of adolescents living in social housing by income quintile–lowest (Q1) to highest (Q5)–shows grade-9 credits at 36.9 versus 66.3 (i.e., 1.8 times more likely to complete grade 9 in Q5; P = .003), grade-12 completion at 26.5 versus 63.0 (2.4 times more likely to graduate from high school in Q5; P < .001), and adolescent pregnancy at 278.8 per 1000 versus 166.7 per 1000 (1.7 times more likely to have an adolescent pregnancy in Q1; P = .04).DISCUSSIONOverall, children living in social housing had poorer health and education outcomes than all other children, and in most cases when compared with those living in the same socioeconomic areas of Winnipeg. The implication is clear for government and community programs--health and education officials know exactly which housing is social housing, so targeted preventive health or educational programs could be provided to residents. When we compared children in social housing within different SES areas, outcomes were distinctively different for early childhood versus school age. For the 2-year-old complete immunizations and school readiness indicators,differences were similar by socioeconomic area of residence. In contrast, for school-aged indicators (i.e., grade 9, grade 12, adolescent pregnancy), the wealthier the area in which social housing was located, the better the outcomes for social housing residents. This was especially evident when we compared placement of social housing in the 2 poorest quintile areas (Q1, Q2) with placement of social housing in the middle and wealthiest income quintile areas (Q3---Q5), where children in the PubMed ID:http://www.ncbi.nlm.nih.gov/pubmed/19896111 latter showed improvements in positive outcomes. One could hypothesize that once children experience peer influence in school, their behavior may be affected by peers. So even though the outcomes of children in social housing were worse than their corresponding “all others” in wealthier neighborhoods (not surprising, when one considers the challenges of poverty), they did better compared with their social housing counterparts in poorer areas. Thus, “place of residence” of social housing may be important, with better outcomes for school-aged social housing residents for higher-SES areas. This would mirror the majority of studies earlier cited regarding the importance of neighborhood effects.