Familial aggregation of childhood health and the socioeconomic gradient of disease: a longitudinal population-based sibling analysis
This study explores the relationships that emerge between socioeconomic status (SES) and the prevalence of several health outcomes in children of different ages utilizing administrative data housed at The Manitoba Centre for Health Policy (MCHP). This research also determines the effect that family has on a child developing (or not developing) a specific health outcome. Finally, the relationship between prevalence and familial aggregation are examined. The Johns Hopkins ACG(r) Case-Mix System grouped various physician and hospital diagnosis codes into 32 Aggregated Diagnostic Groups (ADGs). Eight of these ADGs were assessed at four age groups (0-3, 4-8, 9-13 & 14-18) for each member of the final study population. Each member was assigned to one of six SES groups, five income quintile groups and one social assistance group. Familial aggregation was determined for eight selected ADGs using an intraclass correlation coefficient (ICC). Statistical contrasts were made for SA vs. Q1-Q5 and an overall linear trend (SA – lowest; Q5 – highest) to establish the SES differences for the prevalence and familial aggregation of a particular condition. Many of the conditions across SES had statistically significant (p<0.05) linear and SA vs. Q1-Q5 contrasts for 3 both ICCs and prevalence at all age groups. Of the eight ADGs that familial aggregation was calculated, chronic conditions related to the eye had the highest ICCs at all age groups. Injury ADGs had consistently lower ICCs for all age groups. Factors that affected the results of ICC estimation for binary outcomes include the number of bootstrap selections, the width of the age group and the event rate for the outcome of interest. Suggested future research includes a validity review of ICC estimates for binary outcomes, exploring the variables that may reduce or eliminate the SES gradient for ICCs and exploring the aggregation for different study samples within Manitoba.
ICC, Siblings, Family, SES, Correlation, Aggregation