Title: Racial and Ethnic Disparities in Adolescent Health
1Racial and Ethnic Disparities in Adolescent Health
- Robert Wm. Blum, M.D., M.P.H., Ph.D.
- Professor and Director
- Division of General Pediatrics Adolescent
Health - University of Minnesota
- Prepared for
- 16th Annual UIC MCH Leadership Conference
- Translating Research into MCH Public Health
Practice - April 28, 2003
2There is extensive data comparing adolescent risk
behaviors among various ethnic groups
3 Poverty Rates by Ethnicity
- In 1992, 21.9 of all Youth Lived in Poverty
Percentage
Single Parent
4Homicide Rates
5Sexual Debut Among 15-19 Year Old Females
Percentage
Hispanic
African American
White
6Cigarette Smoking by Race
Percentage
7Summary
- Prevalence differences in adolescent behaviors
among races has lead to the conclusion that race
(or its associated cultures) are important
explanations of those differences.
8Add-Health StudyCurrent Analyses
- To what extent does race/ethnicity explain
differences in adolescent risk behaviors? - To what extent is race confounded by income
and/or family structure? - How much can adolescent risk behaviors be
explained by ethnicity, income and family
structure together?
9Add-Health Sample Design
129 Schools (79) 90,118 (75.6)
In-School Sample (1994)
Saturation High income African American
Twins Disability
In-Home Samples (1995)
N15,243
Core In-school survey school rosters
Special
Teen N12,105 (79.5)
Parent N10,471 (86.5)
Wave 1 (1994-5)
Wave 2 (1996)
Teen
10Sample Description
11Family Variables
- Income (Parent Report)
- lt 20,000
- 21,000 - 40,000
- gt 40,000
12Ethnicity/Race(Adolescent Report)
- White (Non-Hispanic)
- Hispanic (all races)
- African American/Black (Non-Hispanic)
- Multiple Races primary ethnic identity
13Family Structure(Adolescent Report)
- Two resident parent (biologic,
recombined, adoptive) - Single resident parent (biologic, adoptive)
14Adolescent Risk Behaviors
- Cigarette Smoking A 7-category composite
variable from never smoked to smoked gt1 pack/day
in the past 30 days. - Alcohol Use An 8-category variable from
never/almost never used to daily/almost daily use
alcohol in the past year. - Suicidality A 5-category scale from seriously
thought about committing suicide to attempted 1,
2, or more times in the past year.
15Adolescent Risk Behaviors
- Violence An 8-item scale from a physical
fight, a group fight, injured someone, threatened
someone with a weapon in a fight, to shot or
stabbed someone in the past year. - Sexual Debut Ever had intercourse.
16Findings Intercorrelation of Family Structure
and Income
Income Over 40,000 Among Single Parent Families
Percent
17Findings Intercorrelation of Family Structure
and Income
Income Under 20,000 Among Single Parent Families
Percent
18High Risk Behavior Income Grade
a plt.05 b plt.01 c plt.001 X2 (2df) Test for
difference among three groups on dichotomized
risk behavior variables. Note Calculated on
subsample of youth in three ethnic groups who
have resident parents and information on
household income (n8,486).
19High Risk Behavior Ethnicity Grade
aplt.05 bplt.01 cplt.001 X2 (2df) Test for
difference among three groups on dichotomized
risk behavior variables Note Calculated on
subsample of youth in three ethnic groups who
have resident parents and information on
household income (n8,486).
20High Risk Behavior Ethnicity Income
aplt.05 bplt.01 cplt.001 X2 (2df) Test for
difference among three groups on dichotomized
risk behavior variables. Low 20,000
household income Moderate 21,000-40,000
household income High 40,000 household
income.
21High Risk Behavior Ethnicity Income
aplt.05 bplt.01 cplt.001 X2 (2df) Test for
difference among three groups on dichotomized
risk behavior variables. Low 20,000
household income Moderate 21,000-40,000
household income High 40,000 household
income.
22High Risk Behavior Ethnicity Income
aplt.05 bplt.01 cplt.001 X2 (2df) Test for
difference among three groups on dichotomized
risk behavior variables. Low 20,000
household income Moderate 21,000-40,000
household income High 40,000 household
income.
23Cigarette Smoking
24Cigarette Smoking
- When taken together race, incomeand family
structure explain 2 of variance (R2.020) for
younger teens and 6.1 for high school students.
Gender does not contribute explanatory power.
25Alcohol Use
26Alcohol Use
- Race, income and family structure
- explain 1 of variance for younger adolescent
drinking behavior and 2.5 for older teens.
Gender is non-contributory for older adolescents.
27Violence Perpetration
28Violence Perpetration
- Most of the variance (4.6 for younger teens
and 5.0 for older) is explained by gender. Race,
income and family structure together account for
an additional 3.5 and 3.9 of variance in
violence perpetration, respectively, for younger
and older teens.
29Suicide Attempt
30Suicide Attempt
- Gender explains less than 1 of variance for
younger teens and only 1.3 for older
adolescents. Race, income and family structure
when taken together add no explanatory power to
understanding suicide attempts.
31Sexual Debut
32Sexual Debut
- While an R2 statistic cannot be calculated
using logistic regression, using an improvement
in X2 fit, it is estimated that race, income and
family structure explain an estimated 8.5 of
variance among younger teens and 3.5 for older
adolescents.
33Summary
- When taken together, race, income and family
structure explain between 0.7 and 8.5 of the
variance in the adolescent risk behaviors.
34Which are the factors that matter most across
ethnic groups?
35Cigarette Smoking and Alcohol Use
36Cigarette Smoking and Alcohol Use
37Violence, Suicide Risk, Sexual Debut
38Violence, Suicide Risk, Sexual Debut
39Violence, Suicide Risk, Sexual Debut